Cognitive Linguistics
Cognitive Linguistics
Cognitive Linguistics
Tony Veale
484
Tony Veale
Editors
Gitte Kristiansen
Michel Achard
Rene Dirven
Francisco J. Ruiz de Mendoza Ibanez
Mouton de Gruyter
Berlin New York
Cognitive Linguistics:
Current Applications and
Future Perspectives
edited by
Gitte Kristiansen
Michel Achard
Rene Dirven
Francisco J. Ruiz de Mendoza Ibanez
Mouton de Gruyter
Berlin New York
484
Tony Veale
Mouton de Gruyter (formerly Mouton, The Hague)
is a Division of Walter de Gruyter GmbH & Co. KG, Berlin
ISBN-13: 978-3-11-018950-6
ISBN-10: 3-11-018950-X
ISSN 1861-4078
Copyright 2006 by Walter de Gruyter GmbH & Co. KG, D-10785 Berlin
All rights reserved, including those of translation into foreign languages. No part of this book
may be reproduced or transmitted in any form or by any means, electronic or mechanical,
including photocopy, recording, or any information storage and retrieval system, without
permission in writing from the publisher.
Typesetting: Drlemann Satz GmbH & Co. KG, Lemfrde
Printed in Germany
Acknowledgements
The editors of this volume would like to thank those who made this publication possible.
First and foremost we are deeply grateful to Anke Beck for her interest
in this project.
We also wish to express our gratitude to all the contributors for their insights and professionalism and for complying with strict deadlines in
spite of their many other commitments.
Finally, a sincere thanks goes to Birgit Sievert, Wolfgang Konwitschny,
Monika Wendland, Julia Ulrich, Ursula Kleinhenz, Jennifer Mand and
Kirstin Brgen at Mouton de Gruyter for their efficient help and fruitful
cooperation.
vi
vii
Table of contents
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . .
List of contributors . . . . . . . . . . . . . . . . . . . . . . . . . .
Introduction: Cognitive Linguistics: Current applications and
future perspectives . . . . . . . . . . . . . . . . . . . . . . . . . .
Gitte Kristiansen, Michel Achard, Ren Dirven and
Francisco J. Ruiz de Mendoza Ibez
v
ix
21
51
81
. . . . . . . . . . . . . . . . . . . . . 147
viii
Table of contents
ix
List of contributors
Michel Achard
Department of Linguistics
Rice University
Houston, TX
Seana Coulson
Department of Cognitive Science
University of California
San Diego, CA
USA
USA
e-mail: [email protected]
e-mail: [email protected]
John A. Barnden
School of Computer Science
University of Birmingham
U. K.
e-mail: J.A.Barnden@
cs.bham.ac.uk
Ren Dirven
Mechelen
Belgium
e-mail: [email protected]
Frank Boers
Department of Applied
Linguistics
Erasmushogeschool
Brussels
Belgium
e-mail: frank.boers@docent.
ehb.be
Institute of Education
University of Antwerp
e-mail: [email protected]
Cristiano Broccias
Department of Linguistic
and Cultural Communication
Universit di Genova
Italy
e-mail: Cristiano.Broccias@
unige.it
Charles Forceville
Department of Media Studies
Universiteit van Amsterdam
The Netherlands
e-mail: [email protected]
Margaret H. Freeman
Myrifield Institute for Cognition
and the Arts
Heath, MA
USA
e-mail: [email protected]
Dirk Geeraerts
Department of Linguistics
University of Leuven
Belgium
e-mail: dirk.geeraerts@arts.
kuleuven.be
List of contributors
USA
Gary B. Palmer
Department of Anthropology
and Ethnic Studies
University of Nevada
Las Vegas, NV
e-mail: [email protected]
USA
e-mail: [email protected]
Susan Goldin-Meadow
Department of Psychology
University of Chicago, IL
USA
e-mail: [email protected]
Terry Janzen
Department of Linguistics
University of Manitoba
Canada
e-mail: [email protected]
Gitte Kristiansen
Department of English Language
and Linguistics
Universidad Complutense
de Madrid
Madrid
Spain
e-mail: [email protected]
Klaus-Uwe Panther
Department of English
and American Studies
Universitt Hamburg
Germany
e-mail: [email protected]
Marcus Perlman
Department of Psychology
University of California
Santa Cruz, CA
USA
e-mail: [email protected]
Tim Rohrer
Colorado Advanced Research
Institute
Boulder, CO
USA
e-mail: [email protected]
Seth Lindstromberg
Hilderstone College
Kent
U.K.
e-mail: [email protected]
Seyda
zalskan
Department of Psychology
University of Chicago, IL
USA
e-mail: [email protected]
xi
List of contributors
John R. Taylor
Department of English
University of Otago
New Zealand
e-mail: john.taylor@stonebow.
otago.ac.nz
Tony Veale
School of Computer Science
and Informatics
University College Dublin
Ireland
e-mail: [email protected]
xii
List of contributors
Introduction
Introduction
Cognitive Linguistics: Current applications and
future perspectives
Gitte Kristiansen, Michel Achard, Ren Dirven and
Francisco J. Ruiz de Mendoza Ibez
1.
The collective volume Cognitive Linguistics: Current Applications and Future Perspectives brings together specific case studies and critical overviews
of work in a variety of CL strands. Written by prominent researchers, the
chapters of the volume thus provide the scientific community with an updated survey of recent research in Cognitive Linguistics. Most authors
furthermore go beyond the more immediate scope of describing or
exemplifying state-of-the-art research (e.g. by providing the reader with a
generally accessible synthesis or a specialized case study) and explicitly
address a number of pressing questions pertaining to future perspectives
and future research agendas.
Together with its companion volume Cognitive Linguistics: Basic Readings (Cognitive Linguistics Research 34, edited by Dirk Geeraerts), it constitutes a highly informative resource for linguists and scholars in neighbouring disciplines, and in general for any scholar wishing to become
familiar with what Cognitive Linguistics is all about. Whereas Cognitive
Linguistics: Basic Readings offers an introductory survey of the foundational concepts of Cognitive Linguistics, the present volume focuses on
more recent theoretical developments, illustrates the many fields of application that CL already covers (both within linguistics and in an interdisciplinary environment), and identifies the future research trends that CL
is now heading for.
At the same time, the present volume is the very first issue in the new
book series Applications of Cognitive Linguistics (ACL). In collaboration
with its sister series Cognitive Linguistics Research, ACL offers a platform
for high quality work which applies the rich framework developed in Cognitive Linguistics to a wide range of different fields of application. These
fields include descriptive linguistics, cultural linguistics, language acquisition, discourse studies, sociolinguistics, visual communication, stylistics,
poetics, pedagogical linguistics, computational linguistics, signed language
and still many other fields, often within an interdisciplinary framework.
The goals of ACL will be summarised in section 3 of this introduction.
First and foremost, however, as the subtitle suggests, the volume overviews and explores the major avenues of the cognitive linguistic enterprise
at present and towards the future. Over the last two, perhaps even three,
decades, Cognitive Linguistics has gradually but firmly established itself
as a complete and innovating discipline, but certainly not one which for
these reasons has ceased to evolve, nor to expand. The contributions in
this volume testify to the existence of a number of different strands, the
most important of which may be summarised as follows. First, a number
of basic concepts (cf. the twelve cornerstones of Cognitive Linguistics described and exemplified in Cognitive Linguistics: Basic Readings) are currently being critically refined or even challenged. Second, the fact that
Cognitive Linguistics is fully committed to the analysis of meaning in all
its various facets naturally lends itself to multiple applications in all those
areas of human communication where meaningfulness is relevant. Hence,
the range of applications is expanding, not only across disciplines, but
also within linguistics itself. This volume is a representative, but necessarily non-exhaustive illustration of this trend: while it covers a broader
perspective than most other readers or introductions to Cognitive
Linguistics, it obviously cannot reflect all of the many burgeoning applications. And last, but not least, there is a call for increasing methodological depth, to ensure a firm position for CL within Cognitive Science at
large.
The volume is structured in six main sections. While the first three sections (chapters 18) focus on fundamental theoretical and methodological
issues, the last sections (chapters 915) either discuss or exemplify the extent to which the cognitive linguistic framework can be applied in a variety
of neighbouring disciplines (i.e. interdisciplinary or transdisciplinary applications) or to a wide range of practical language-related phenomena (applied linguistics). The distinction between these two main areas (theory and
method vs. interdisciplinary applications and applied linguistics) is not,
however, meant as a strictly separated one. On the contrary, application
perspectives are legion in the first part, and basic theoretical issues naturally pervade the more application-oriented sections of the second part.
The first section, The cognitive base, addresses issues which in various ways relate to the basic tenet that Cognitive Linguistics is fundament-
Introduction
ally usage-based. The section opens with a chapter on linguistic methodology and emphasizes the necessity of an empirical revolution. In the rest
of section 1, the mental lexicon is seen as the highly flexible result of numerous usage events, and the various grammar models such as cognitive
grammar, construction grammar, and radical construction grammar are
revisited in terms of notational variants. The second section, The conceptual leap, looks into the embodied basis and nature of conceptual
metaphor, the semiotic and indexical basis of conceptual metonymy, and
the rhetorical potential of conceptual blending. The third section, The
psychological basis, exhibits a possible implementation of the empirical
revolution advocated in chapter 1 to the area of metaphor research.
Sections 4 to 6 exemplify a representative selection of fields of application. The first comes to the fore in section 4, Go tell it on the mountain, which focuses on language, culture and thought. It opens with a dialogue between cognitive linguistics and anthropological linguistics, thus
emphasizing the cultural core of language and the linguistic core of culture. Another long-established field is second and foreign language education (or pedagogical linguistics), which can boast of a long tradition of
empirical research, but which is still in the process of discovering the
motivational potential of a cognition- and usage-based approach to
foreign language teaching and learning. Section 5, The verbal and
beyond, extends the research on verbal modes of expression to other
modes of conceptual self-expression. The cognitive-linguistic approach is
applied to the visual modes of expression as exploited in signed languages
and in the manipulation of visual metaphors in the media-dominated
communication landscape. Similarly, the imaginative realities of poetic
text worlds are explored in a cognitive analysis of Robert Frosts poem
Mending Wall, a choice that serves as a symbol and pun for the wall
between linguistics and literary studies. In section 6, Virtual reality,
computer linguistics is looking at cognitive linguistics and lays down the
conditions that have to be fulfilled in order to make it testable for computational experimentation.
It must be clear that these fields of application do not exhaust the possibilities for interdisciplinary cognitively inspired explorations. To name
just one area, cognitive sociolinguistics is only indirectly represented in
the context of cultural linguistics. This issue will be taken up in the last
part of this introduction. Let us begin, however, by introducing the fifteen
chapters in slightly further detail.
2.
Introduction
Introduction
zalskan and
Susan Goldin-Meadow first discuss simple analogical mappings in terms
of one of the earliest metalinguistic abilities which children master and
which thus constitute the stepping-stones for the development of metaphorical thinking. The authors then present the results of research which
investigated the emergence and later development of similarity mapping
abilities (X is like Y) in spontaneous speech and gesture, using longitudi-
Introduction
10
Introduction
11
ate sensations, feelings, and images in language that enable the mind to
encounter them as phenomenally real. The choice of Mending Wall as a
case study undoubtedly contributes to rendering the fall of the wall an
even more symbolic and entertaining process.
2.6. Virtual reality as a new experience
The study of the computational applications to CL is a comparatively new
area of enquiry which poses a challenge both for CL, which needs to be
tested in terms of its adaptability to computational requirements, and for
computer science, which needs to assess its own ability to cope with the
descriptive and explanatory richness of an approach to language that explicitly ties together body, mind, thinking and language.
In this context of mutual challenges, John Barndens chapter, entitled
Artificial intelligence, figurative language and cognitive linguistics, discusses the relationship between CL and Artificial Intelligence. The
chapter is focused on Barndens well-known ATT-Meta project, which is
both an explanatory approach and a computer program capable of performing reasoning tasks involved in metaphorical utterances. The overall
project is related to but at the same time challenges some of the central assumptions in Lakoff s (1993) conceptual metaphor theory. However, as
Barnden himself remarks, ATT-Meta is in some respects fairly close to
Gradys theory of primary metaphor (Grady 2005). Thus, rather than a
mapping from machines to minds, ATT-Meta prefers to postulate
transfer rules from physical operation to mental operation. Also significantly, the ATT-Meta approach questions the usefulness of the notion
of domain in computational implementations because of the high degree
of arbitrariness involved in determining domain boundaries. This leads
Barnden together with other AI theorists to disregard domain mappings and see metaphor as a matter of transferring from a pretence to
reality. This move has implications for the traditional distinction between
metaphor and metonymy in terms of domain-external versus domain-internal mappings. For Barnden, metaphor and metonymy are merely vague
notions playing a heuristic role, while the real phenomena lie at a deeper
level involving conceptual distance, similarity, contiguity, and connections between source and target, an approach which thus ties in with
Panthers criticism of the unreliability of the notion of domain.
Unlike Barndens contribution, the chapter by Tony Veale, entitled
Computability as a test on linguistics theories, focuses more on computational assessment than on mutual feedback between CL and AI. The
12
3.
Introduction
13
The available CL theoretical know-how, constructs and general characterisations of language are well-suited for explorative applications in the
research areas traditionally known as descriptive linguistics, text linguistics, psycholinguistics, applied linguistics, sociolinguistics, anthropological linguistics, etc. It is indeed the case that such applications are currently in fast expansion in many fields, as the chapters just surveyed in this
introduction amply show. The CL usage-based approach furthermore
naturally calls for a non-restrictive interpretation of the term application: one of the main assumptions of ACL is that linguistic research
must be firmly grounded in usage-based, descriptive data, which may,
more often than not, imply an interdisciplinary approach to the object
under scrutiny. In language pedagogy, for instance, linguistics meets with
psycholinguistics and educational psychology to study the combined processes of natural and conducted acquisition of second and/or foreign languages; hence the label pedagogical linguistics is out of place. In cultural
linguistics, the deeper relation between language and culture is explored,
which covers and overlaps with various scientific disciplines or fields such
as (cognitive) anthropology, cross-cultural semantics, intercultural pragmatics, cross-cultural communication, social psychology and still other
disciplines. In computational linguistics, amongst many other possibilities, theoretical models such as cognitive grammar, conceptual metaphor
theory, blending theory, etc. are caught in self-monitoring artificial intelligence programs to test the viability of the constructs used.
As to the scope of ACL the following may serve as an attempt to spell
out some examples of ACLs orientations, at best signposting some major
areas on the map, but certainly not laying out a complete map.
One major application of CL is the area of descriptive linguistics covering either the description of parts or the whole of the grammar of given
languages, or else the description of language as used in given domains of
human experience. The grammatical descriptions can focus on any level
of linguistic organization, including syntax (e.g. Cornillie fc.), lexis, morphology and phonology, and thereby concentrate on one language in particular or be conceived cross-linguistically as contrastive studies. Descriptive applications can also focus on the exploration of theoretical models
in actual language use: for example, on views of mental motion and summary scanning, on prototype theory, on construction grammar, on conceptual metaphor or metonymy (as already practiced in the many discourse studies of political (e.g. Musolff 2004), economic (e.g. Koller 2004)
and even architectural communication (e.g. Caballero 2006) or on conceptual blending (as illustrated by Coulsons chapter in this volume and in
14
various other studies). Further burgeoning fields in descriptive applications of CL are cognitive discourse studies, either in the tradition of
blending theory or in the search for coherence relations (e.g. Sanders and
Gernsbacher, eds. 2004). As shown in Freemans chapter in this volume,
highly interesting developments are also taking place at the interface of
literary studies and cognitive linguistics, known as cognitive poetics or
stylistics, thereby offering many bridges to both disciplines.
Another major field is constituted by the pedagogical application of CL,
which is now becoming a novel area of research in the cognitive-linguistic
world (e.g. Achard and Niemeier, eds. 2004), and one which naturally
finds an interdisciplinary basis in psycholinguistics, educational psychology, theoretical and descriptive linguistics, also pointed out by Boers and
Lindstromberg in this volume. It is probably because of their inherently
interdisciplinary orientation that pedagogical applications of linguistics,
or the somewhat all-encompassing and hence misleading term applied
linguistics, has seldom (except by Ptz et al., eds. 2001) been brought together with descriptive linguistics in one organizational framework, be it
journals, series of volumes, or conferences. Still, since Cognitive Linguistics
offers such a strong conceptual unity in its approach to language and linguistics, it is entitled to accept the challenge of rallying such seemingly diverse applications of Cognitive Linguistics into one conceptual organic
whole. The interdisciplinary character of pedagogical linguistics, or language pedagogy to use a more generally accepted and less misleading
label may necessitate a multi-layered foundation, but in a sense this
equally applies to several other applications, such as cognitive sociolinguistics and cultural linguistics, two other fields we still want to go into.
In cognitive sociolinguistics (e.g. Kristiansen and Dirven fc.) several
branches of traditional sociolinguistics are redefined from a cognitive
perspective. Cognitive sociolinguistics combines the usage-based CL
tenet with fine-grained research on language variation and examines the
correlations with cognitive models, now widened as cultural models (e.g.
Dirven et al., eds.). Sociolinguistic research which at one end tends to
overlap with cultural linguistics further includes work on language policy,
global language variation and ideology in socio-politico-economic areas
such as political, educational, and corporate-culture frameworks.
Cultural linguistics is, as just stated, a close neighbour of cognitive sociolinguistics. Cultural linguistics makes use of the analytical tools developed within the framework of CL in order to explain many aspects of
human languages in their cultural contexts, thereby exhibiting clear ties
with work in cognitive anthropology, where much emphasis is being put
Introduction
15
on the cognitive grounding of social structure and on how cultural schemas relate to action. In spatial and social cognition, on the other hand,
cross-cultural studies with an emphasis on language acquisition have provided us with evidence for the social grounding of cognitive structure.
Studies on cultural beliefs, values and norms (e.g. Holland and Quinn
1987; Goddard, ed. 2006) likewise constitute an interesting area of research in constant evolution. It may also be interesting to note that proponents of cultural linguistics have recently begun to explore the way their
theoretical persuasion applies to language learning, language teaching
and cross-cultural communication. Relevant issues in this applied perspective are the study of meaning negotiation as adjustment of cultural
schemas, the analysis of the impact of conceptual metaphor and metonymy in misinterpretation in cross-cultural encounters, the study of cultural models, cultural keywords and culture-specific collocations in L2
varieties of English.
In conclusion, it should be emphasised that, in spite of the great diversity of research fields, one central cohesive factor remains and binds all
strands and models within CL together: the ubiquity of the link between
language in use and other cognitive faculties. Descriptive applications use
the research tools of CL and explore their validity in the description of
single languages or in cross-linguistic explorations. Language pedagogy
may be as diversified as the functions of language in usage events, but language learning is probably the most spectacular place to look for the integration of all cognitive faculties, a new nascent set of schemas derived
from language usage events and social interaction. Cultural-specific categories and models not only necessitate the adoption of psychologically
real constructs, but also and far beyond that the adoption of models
capable of accounting for culturally distributed knowledge, that is, common cultural knowledge that is not necessarily possessed by all single
members of a community, but only possessed by the collectivity of the cultural community.
Although this aspect has not been restated in this final section on the
new series ACL, it must be clear that in line with the whole orientation
of this volume only empirical research, possibly including computational verification of any linguistic or cultural model, is the pre-condition
and the ultimate test guaranteeing that the constructs work and can survive in the long-term future both of cognitive linguistics and its diverse
fields of application.
16
References
Achard, Michel, and Susanne Niemeier (eds.)
2004
Cognitive Linguistics, Second Language Acquisition, and Foreign Language Teaching. Studies on Language Acquisition 18. Berlin/New
York: Mouton de Gruyter.
Attardo, Salvatore, Christian F. Hempelmann, and Sara Di Maio
2002
Script oppositions and logical mechanisms: Modeling incongruities
and their resolutions. Humor: International Journal of Humor Research
15 (1): 346.
Caballero, Rosario
2006
Re-Viewing Space: Figurative Language in Architects Assessment of
Built Space. Applications of Cognitive Linguistics 2. Berlin/New York:
Mouton de Gruyter.
Cornillie, Bert
ForthEvidentiality and Epistemic Modality in Spanish (Semi-) Auxiliaries: A
coming Cognitive-Functional Approach. Berlin/New York: Mouton de Gruyter.
Croft, William
2001
Radical Construction Grammar: Syntactic Theory from a Typological
Perspective. Oxford: Oxford University Press.
Dirven, Ren, Roslyn Frank, and Martin Ptz (eds.)
2003
Cognitive Models in Language and Thought: Ideology, Metaphors and
Meanings. Cognitive Linguistics Research 24. Berlin/New York: Mouton de Gruyter.
Dodge, Ellen, and George Lakoff
2005
Image schemas: From linguistic analysis to neural grounding. In From
Perception to Meaning: Image Schemas in Cognitive Linguistics,
Hampe, Beate (ed.), 5791. Cognitive Linguistics Research 29. Berlin/
New York: Mouton de Gruyter.
Fauconnier, Gilles, and Mark Turner
2002
The Way We Think: Conceptual Blending and the Minds Hidden Complexities. New York: Basic Books.
Geeraerts, Dirk (ed.)
2006
Cognitive Linguistics: Basic Readings. Cognitive Linguistics Research 34. Berlin/New York: Mouton de Gruyter.
Goddard, Cliff (ed.)
2006
Ethnopragmatics: Understanding Discourse in Cultural Context. Applications of Cognitive Linguistics 3. Berlin/New York: Mouton de
Gruyter.
Goldberg, Adele E.
1995
Constructions: A Construction Grammar Approach to Argument Structure. Chicago: University of Chicago Press.
Grady, Joseph
2005
Primary metaphors as inputs to conceptual integration. Journal of
Pragmatics 37 (10): 15951614.
Introduction
17
18
19
20
21
Abstract
The future of linguistics is likely to be determined by methodological issues:
whether linguistics will be able to turn its current theoretical chaos (however creative it may be) into a situation of cumulative development will crucially depend
on the possibility of finding a common ground that allows competing accounts to
be compared in an objective way. This chapter explores the position of Cognitive
Linguistics vis vis the methodological evolution of linguistics: what is the current state of affairs in Cognitive Linguistics, and in what direction could the cognitive approach contribute to the methodological progress of linguistics? The first
section introduces the concept of empirical research, using a non-linguistic
example. The second section sketches a scenario for the evolution of linguistics at
large, highlighting the specific role of method as a safeguard of theoretical comparability. The third section describes the particularly advantageous position of
Cognitive Linguistics, i.e. the fact that a number of theoretical features of Cognitive Linguistics seem to be specifically favorable to the methodological development that is meant here. The fourth section takes into account the other side of the
coin: if we consider the actual state of affairs in Cognitive Linguistics, the situation is definitely a mixed one, with a growing concern for empirical method on
the one hand and waving of hands at the other. If this heterogeneous situation is
not accidental, we should be able to identify theoretical features that have exactly
the opposite effect of the ones discussed in the third section features, that is, that
go against the tendency towards well-grounded empirical research. The overall
future of Cognitive Linguistics, needless to say, will then be determined by the
extent to which it can resolve the methodological tensions that seem to derive
from its theoretical assumptions.
Keywords: methodology; empirical research; experimentation; corpus linguistics;
introspection; future developments; cognitive science; psychology; statistics
1.
22
Dirk Geeraerts
tice that the students first names come in clusters. In one year, your attention is caught by a frequent group of Heleens or Inekes, in another, by
a conspicuous cluster of Rons or Lens or Georges. As a professor of linguistics, youre an inquisitive type of person, so you wonder whether this
is a coincidence. What could be the case? One possibility would be that the
clustering of proper names in your class is a faithful reflection of what is
happening in the population at large: there are fashions in giving first
names that change over the years, and the sample of the population that
attends your class reproduces the overall tendencies. In that case, there is
something specific about the age of your students: they were born in a
period in which a particular name was trendy. But there is at least one
other possibility to consider: it could be that the frequency of specific
names in your class differs from what you could expect given the distribution of names in the population as a whole. In that case, there would be
something specific about your class: for one reason or another, it attracts
people with a certain name.
How could you decide between the two options? You would want to
make a number of comparisons: on the one hand, you would compare the
distribution of names in your class with the distribution of names in the
entire population of the same age group, and on the other, you would
compare the distribution of names in one year or one period with the distribution of names in another year or period. In both comparisons, you
would not just want to know whether the distributions differ, but specifically also whether the differences you observe are a coincidence or not.
That is, of course, where statistics will help you: a statistical test will allow
you to see whether a difference you detect is due to chance or not.
If it is due to chance, you will leave the matter for what it is, but if it is
not, your curiosity will get the better of you, and you will start wondering
about an explanation. For instance, if there are fashions that change
through time (as there would be), what is it that determines them? What is
it that makes Heleen popular in a particular period? A hypothesis that you
are likely to come up with is that naming fashions of this kind are determined by the degree of exposure certain names get in the media and in
popular culture: popular sports people or popular characters from television fiction people whose name gets around may well provide a
model for naming new-borns. Again, you would want to put the hypothesis to the test: you may have to put in a considerable effort to find the
right kind of data, but you would want to find out whether one or another
Heleen was popular some twenty years before you noticed all those Heleens in your class. Or rather, you would not want to do this for just one of
23
the names, but you would want to reach a general conclusion: if a certain
name is popular in the media in a certain period, does that lead to an increase in the number of babies receiving that name in the following years?
And again, you would need some statistics to ascertain whether the influence you think you see might not just be a coincidence.
We could go on like this, surely, investigating explanatory questions
and research techniques for our hypothetical Christian Name Conundrum, but that is not the point of the exercise. The point is this: the procedure illustrated in the previous pages is typical of what we call empirical
research, and it involves a way of thinking that we all understand and
that we all recognize as being superior to an attitude in which you would
simply say: I notice these clusters of names in my classes. I conclude that
they are due to fashion. Before you can really conclude that, you have to
take a number of steps, steps of the type that we explored above.
This chapter, then, is all about the position of empirical research in
Cognitive Linguistics. It will argue, first, that linguistics in general would
benefit from more research of this type; second, that Cognitive Linguistics has an excellent starting-point to contribute to such an increase in
the use of empirical studies; and third, that in spite of that advantageous
basis, Cognitive Linguistics has not yet fully realized the breakthrough towards the massive use of empirical methods. What the chapter will not do,
by the way, is discuss the practicalities of doing empirical research in linguistics. Although, in terms of boosting empirical studies, it would be extremely useful to do so, it is simply beyond the scope of the present text to
offer a practically oriented primer in experimental or corpus-based
studies in linguistics. It is hoped that the references to empirical studies
scattered throughout the text will help the reader to get started. Further,
see Gonzalez, Mittelberg, Coulson, and Spivey (forthcoming) for a booklength introduction to different empirical methods for Cognitive Linguistics.
As a crucial backdrop for the following discussion, let us now summarize the main features of empirical research that we can derive from our
simple example.
1. Empirical research is data-driven. You cannot easily draw conclusions from single cases and isolated observations, and the more data
you can collect to study a particular phenomenon, the better your conclusions will get. The observations could come from many sources, by the
way: you could collect them as they exist (as in our Christian Name Conundrum), but you could also elicit them by doing experimental research,
or by doing survey research we will come back to that.
24
Dirk Geeraerts
25
anecdotal observation, tried to interpret it, turned the possible interpretations into alternative hypotheses, tested the hypotheses, and so arrived
at a new, more secure and better established observation. But this observation was only the starting-point for a new attempt at interpretation,
which led to a new hypothesis about media influence, which led to new research efforts. Just like it is misguided to think that empirical, data-driven
research automatically gives you all the answers, it is misguided to think
that it immediately gives you the final answer.
2.
Let us now turn to linguistics: why could we say that more empirical research of the type illustrated and demarcated above would be beneficial for
the development of linguistics? Theoretical linguistics is characterized by
the existence of a multitude of theories: Universal Grammar in the generativist sense, a family of models claiming the label of Functional Grammar, smaller trends like Role and Reference Grammar or HPSG and of
course Cognitive Linguistics, in a number of versions that may each
branch off into subversions, and then into alternative realizations of
subversions. There is Cognitive Linguistics Langacker-style (Cognitive
Grammar in the narrow sense), and there is Cognitive Linguistics Talmystyle to some extent using the same vocabulary, involving notions like the
figure/ground opposition, but still sufficiently different to be easily recognizable by anyone familiar with the relevant literature. And then there is
Construction Grammar, which may or may not include Cognitive Grammar (Langacker-style), but which certainly encompasses different approaches like Fillmore and Goldberg-style Construction Grammar, Radical Construction Grammar, and even Embodied Construction Grammar.
Superficially, it might seem that the existence of different theories is
not really typical of linguistics, as the foregoing passage may seem to suggest. In psychology (to take a discipline that is not too far removed from
linguistics), one may come across different theories about how categories
are stored in the mind: you get exemplar models that assume that the bulk
of the work is done via the knowledge we have of typical members of a
category, and you get concept-based models that rather start from the assumption that there is an abstract, schematic definition of a concept, and
then you get mixed models that explore different combinations of exemplar- and concept-based models. Is there no difference with linguistics,
then?
26
Dirk Geeraerts
27
28
3.
Dirk Geeraerts
29
30
Dirk Geeraerts
embodied nature of the mind), it also involves society and culture. In this
sense, an interest in language variation forms an integral part of Cognitive
Linguistics. Now, language variation has been studied in Cognitive Linguistics and adjacent approaches primarily from three points of view:
from a diachronic perspective, including grammaticalization research (see
e.g. Geeraerts 1997; Bybee 2001; Hopper and Traugott 1993; Heine 1991),
from a comparative and anthropological point of view (see e.g. Pederson
1998; Kvecses 2000; Levinson 2003), and from a developmental point of
view (see e.g. Tomasello 2003). Language-internal variation and sociolinguistic diversity has been much less studied, but still, we may note a
number of developments within Cognitive Linguistics that are likely to
contribute to an increased interest in sociolinguistic research.
First, there is the interest in cultural models and the way in which they
may compete within a community: see e.g. many of the papers collected in
Dirven, Frank, and Ptz (2003). Specifically, it has been pointed out
(Geeraerts 2003) that such models may also characterize the beliefs that
language users entertain regarding language and language varieties. In
this way, Cognitive Linguistics may link up with existing sociolinguistic
research about language attitudes.
Second, there is a growing tendency in the theoretical conception of
language entertained by Cognitive Linguistics to stress the social nature
of language. A number of researchers have started to investigate social
variation from the perspective of Cognitive Linguistics: see e.g. the work
by Kristiansen on phonetic variation (2003). In the realm of lexicological
investigations, the research tradition started in Geeraerts, Grondelaers,
and Bakema (1994) and continued in works such as Geeraerts, Grondelaers, and Speelman (1999), or Speelman, Grondelaers, and Geeraerts
(2003), integrates theoretical ideas from Cognitive Linguistics with a variational perspective and a corpus-based methodology.
Third, the usage data that Cognitive Linguistics, as a self-styled usagebased approach, is confronted with, automatically include sociolinguistic
variation. Assuming that the primary empirical source for a usage-based
approach to linguistics consists of corpus materials, then more often than
not, the corpus will not be internally homogeneous: because the texts collected for the corpus come from various sources, it will not be known in
advance whether the variation that may be observed in the corpus is due
to lectal factors or not. (The term lectal is used here to refer to all types of
language varieties or lects: dialects, regiolects, national varieties, registers,
styles, idiolects, whatever). As such, filtering out the effects of such factors will be necessary for any cognitive linguistic attempt to analyse the
31
usage data even if the analysis is not a priori interested in lectal variation.
The importance of socially and culturally determined variation for
Cognitive Linguistics has an obvious methodological correlate: these socially oriented forms of Cognitive Linguistics would naturally support the
adoption of the quantitative, empirical methodology that is dominant in
sociolinguistic research at large.
We may conclude, in short, that using the main types of empirical
methodology that were mentioned in the first section is motivated by the
theoretical stance of Cognitive Linguistics: its very cognitive nature, its
usage-based perspective, and its contextualizing approach may lead in
various ways to the use of experimental data, corpus material, survey
techniques, and advanced quantitative analysis.
4.
32
Dirk Geeraerts
33
search: the general terms data and empirical on the one hand, and the methodologically more specific terms corpus and experiment(al) on the other.
The Cognitive Linguistics Bibliography is a magnificent research instrument compiled predominantly by Ren Dirven; it is included in the subscription to the journal Cognitive Linguistics. The version of the bibliography used for Table 1 is the version of September 2005, which includes
a total of 7000 entries.
Table 1. Presence of terms related to empirical methods in the Cognitive Linguistics Bibliography, divided over five-year periods
19851989
n=766
corpus
experiment(al)
empirical
data
total
4
15
24
21
64
0.5 %
2.0 %
3.1 %
2.7 %
8.3 %
19901994
n=1140
18
46
59
69
151
1.6 %
4.0 %
5.2 %
6.0 %
13.2 %
19951999
n=1881
68
119
116
151
357
3.6 %
6.3 %
6.2 %
8.0 %
19.0 %
20002004
n=2314
215
214
213
249
648
9.3 %
9.2 %
9.2 %
10.8 %
28.0 %
Coarse though the calculation may be, the table does show that there is a
continuous increase in the number of books and articles that refer to empirical approaches: the number of bibliographical entries that cites either
of the four terms (the row labeled total) rises steadily to 28 % in the final
period, and the growth occurs with each of the four terms separately. The
separate growth curves for the individual terms are fairly even, except in
the case of corpus, where the line shows a marked increase from the penultimate to the last period. This is not altogether surprising: the massive
availability of corpus materials is a relatively recent phenomenon, specifically if we take into account the use of the internet as a corpus.
The impression we get from Table 1 is strengthened by further indications of an intensification of the appeal of empirical methods within
Cognitive Linguistics. It suffices to have a look at some of the thematic
symposia organized over the last few years: the Empirical Methods in
Cognitive Linguistics workshop held at Cornell University in May 2003,
the theme session on the use of corpora in Cognitive Linguistics organized
during the ICLC8 in Logrono, July 2003, the workshop on the necessity of
experimental methods in semantics organized during the 20th Scandinavian Conference in Linguistics at the University of Helsinki in January
2004, the choice of experimental and empirical methods as the focus
theme of the 7th Conference on Conceptual Structure, Discourse, and
Language at the University of Alberta in October 2004, and the corpus-
34
Dirk Geeraerts
based theme session Language System and Language Use at the ICLC9 in
Seoul, July 2005 all of these testify to the growing importance that the
Cognitive Linguistics community attaches to an empirical methodology.
4.3. Room for expansion in the use of an empirical methodology
Table 1 also suggests that the empirical approach, although on the rise,
does not yet dominate the methodological landscape in Cognitive Linguistics. This impression may be further supported if we have a look at the
program of the ICLC9 conference in Seoul, July 2005. Table 2 gives the
distribution of the talks over three methodological classes: experimental
research (which in this case is intended to include the use of all elicitation
techniques, neurological evidence, and survey methods), corpus-based research, and traditional research. In terms of data gathering, the latter approach basically uses examples drawn from the linguistic knowledge of
the researcher. The figures involve the plenary talks at the conference, and
the talks from the general session, i.e. posters and theme session papers
have not been included. Since the data are predominantly based on the abstracts of the talks (attending all of them would have been impossible), a
certain degree of caution is necessary: it need not be the case that the actual method of research is well described in the abstract (and actual talks
may differ somewhat from what is announced in the abstracts). Given this
proviso, a number of interesting conclusions present themselves.
Table 2. Methodological breakdown of plenary and general session talks at the
ICLC9, Seoul, July 2005
experimental
General Session
Plenary Session
31
2
17.6 %
20.0 %
corpus-based
50
1
28.4 %
10.0 %
traditional
95
7
53.9 %
70.0 %
First, the figures confirm the conclusion drawn from Table 1 with regard
to the dominance of the traditional method: there is an outspoken interest
in empirical methods, but they do not yet represent the majority of cases.
Admittedly, it would be wrong to expect a 100 % presence of empirical research: theory formation does require pure conceptual analysis, internal
and external comparison of results, meta-reflection on methods and terminology, or synthetic overviews that do not immediately involve new research and there will always be talks of that kind at conferences. There
should be no empirical imperialism, then: one of the great things about
Cognitive Linguistics is the room it has always left for creativity and the
35
bold development of novel ideas, and it should remain that way if the approach is to keep its dynamism. However, at least half of the talks in the
category labeled traditional are not of that kind. They are descriptive
talks that lend themselves to the use of an empirical methodology. In this
sense, a further growth of the empirical trend in Cognitive Linguistics is
definitely possible.
Second, the picture emanating from the plenaries is more conservative
than that from the general session papers. The majority of the plenary
talks, i.e. those of Eve Sweetser, Gilles Fauconnier, Ron Langacker, Len
Talmy, Keedong Lee and Gnther Radden were based on a traditional
methodology of conceptual analysis, without much explicit attention for
empirical data of the type discussed here. The plenary talks by Melissa
Bowerman and Seana Coulson, by contrast, were firmly placed within the
experimental tradition, while Suzanne Kemmers talk used state of the art
corpus methods. Despite its promising title, John Taylors paper on The
mental corpus basically used the corpus idea as a metaphor for the organization of knowledge in the mind but did not provide systematic indications how this metaphor could be linked to actual corpus research. As
such, it is here counted with the traditional category.
Third, if we look more closely at the disciplines and domains of research covered by the different methodologies, we may note that experimental methods and corpus research are more firmly entrenched in some
fields than in others. In foreign language learning, language acquisition
research, and sign language and gesture research, for instance, experimental methods are well entrenched. This is a reflection of the importance of
such methods in the fields in question at large: if experimental testing of
the efficiency of a given teaching method is a standard practice in second
language learning research generally, it is no surprise that cognitive linguists working in that domain will adopt an experimental methodology.
Similarly, using corpora is normal in historical linguistics, language acquisition research, and text linguistics and discourse analysis. If we take
this skewed distribution over disciplines into account, we may again conclude that there is room for a further expansion of an empirical methodology in what is usually thought of as the core area of linguistics, i.e. synchronic grammar.
Fourth, an analysis of the two empirical classes in Table 2 in terms
of substance rather than frequency further supports the idea that an expansion of the methodological repertoire of Cognitive Linguistics is possible. To begin with, we may note that some methodological formats are
relatively underrepresented. The experimental techniques that are being
36
Dirk Geeraerts
37
derived ones. Although corpus-illustrated work constitutes a big step forward in comparison with introspective work, there are two points with regard to which a more advanced way of dealing with corpus materials is
possible, and indeed necessary. Both points, needless to say, link up with
the basic characteristics of an empirical approach as identified in section 1.
A first point concerns the systematicity with which the data are gathered and/or analyzed. Typically, the argumentation in corpus-illustrated research takes one of two forms. When a theory predicts a phenomenon, the corpus presence of the phenomenon is taken as a confirmation
of the theory. When a theory excludes a phenomenon, the corpus presence
of the phenomenon is taken as a falsification of the theory. In both cases,
however, a problem arises when the frequency of the selected corpus data
is low: what if the confirming or falsifying data are marginal with regard
to the alternatives? Because corpus evidence is seldom a question of black
or white (of one hundred percent versus zero percent of occurrences), a
proper analysis of corpus materials requires that alternative patterns are
systematically explored and that their frequencies are statistically evaluated. Grammatical models like Cognitive Linguistics and Construction
Grammar in particular, which accept the existence of networks in which
elements or levels may have different degrees of salience, should be particularly sensitive to differences in the frequency of alternative expressions. That is to say, a usage-based linguistics needs quantification and
statistical analysis.
A second point with regard to which a more advanced form of corpus
analysis may be envisaged is the interpretation of the corpus data. In a corpus-illustrated approach, the corpus data are used to restrict a process of
intuitive interpretation: examples of a linguistic phenomenon (say, a construction) are retrieved from the corpus, and the set of examples is then
analyzed largely intuitively. The availability of large corpora may allow for
a more advanced approach. A given interpretation may be turned into a
set of expectations with regard to the behavior of the entity under investigation, and these expectations may then be tested against the corpus. That
is to say, advanced corpus research needs to pay specific attention to the operationalization of hypotheses. If we interpretatively assume that a linguistic entity has this meaning or that function, what can we expect about
its observable behavior in actual language use? The further steps for a corpus approach follow logically: how can we identify that behavior in the
corpus data, and when can we say that the initial hypothesis is correct?
Within the broad domain of Cognitive Linguistics, a corpus-based
methodology as meant here may be found in various fields (the references
38
Dirk Geeraerts
are exemplary, not exhaustive): in language acquisition research (Tomasello 2003); in language change research (Bybee 2001; Krug 2000); in language variation research (Grondelaers, Speelman and Geeraerts 2002;
Speelman, Grondelaers and Geeraerts 2003); in functional discourse
analysis (Sanders and Gernsbacher 2004), and in quantitative construction grammar (Gries 2003; Stefanowitsch 2003; Stefanowitsch and Gries
2003; Tummers, Speelman and Geeraerts 2005).
But if we take the Seoul conference as representative for Cognitive Linguistics as a whole, this type of advanced corpus research represents only
a small minority of the corpus-based category in Table 2: not more than
10 or 15 % of the corpus-based papers in the General Session would seem
to move beyond the corpus-illustrated level as defined above. It is here in
particular that the raw figures in Table 2 appear to be somewhat deceptive: yes, corpus materials are being widely used, but from the point of
view of an empirical revolution in linguistics and the characteristics of
genuinely empirical research as identified in section 1, the sophistication
with which they are being analyzed could still be improved.
In addition, note that the dichotomy between corpus-illustrated and
more advanced forms of corpus-based linguistics obviously only defines
the endpoints of a continuum. A lot of case studies occupy an intermediary position between both extremes. In particular, studies that do not yet
use quantitative forms of analysis may display a lot of variation with respect to the systematicity with which the empirical data are collected and
scrutinized, ranging from an anecdotal selection of examples from a corpus to a systematic scanning of a corpus to uncover the patterns and uses
of a linguistic phenomenon. Good examples of what may be achieved
with a thorough but not specifically quantified examination of corpus
materials include Deignan and Potter (2004), Heyvaert (2003), Maldonado (2002) and of course, there are traditional branches of linguistics,
like historical lexicography, that testify to the value of in-depth qualitative
analyses of large-scale corpus data (cp. Geeraerts 1997). But just like lexicography has, in recent decades, fruitfully enriched its traditional corpusbased approach with quantitative methods of information retrieval, it is
to be expected that theoretical linguistics too might profit from further
sophistication in the analysis of corpus data. The existence of a continuum, then, should not surprise us. If the spread of empirical methods
in Cognitive Linguistics is a gradual process, it is normal that there will be
researchers, and areas of research, that have progressed further in the empirical cycle than others.
5.
39
Empirical methods, even though they are receiving an unquestionably rising amount of attention, do not yet constitute the dominant tendency in
Cognitive Linguistics, and considerable progress can still be made in the
direction of a full-fledged implementation of the empirical research program. What could be the reasons for this situation a situation that
contrasts somewhat with the expectations that follow from the theoretical
self-conception of Cognitive Linguistics? Two fundamentally different
groups of reasons have to be brought into the picture. One has to do with
rather more mundane, practical reasons, with the way, that is, in which
Cognitive Linguistics is actually being practised and the way it presents
itself to young researchers. The other group concerns reasons of principle:
could there be fundamental features of Cognitive Linguistics that slow
down the introduction of empirical methods? In particular, couldnt we
say that the emphasis on meaning in Cognitive Linguistics would naturally favor the use of introspection, as the only reliable method of direct
access to semantic phenomena? If this is correct, there would be a methodological tension at the heart of Cognitive Linguistics, because we established in section 3 that a number of other characteristics of Cognitive Linguistics are particularly favorable towards the adoption of empirical
methods. We will argue, however, that the theoretical reasons in favor of
an introspective methodology are not compelling.
5.1. Reasons of practice
Two practical factors may be invoked to explain why Cognitive Linguistics has not yet progressed in the direction of empirical research to the
full extent that might perhaps be expected on the basis of its theoretical
position: the fact that training in experimental techniques or corpus research is not a standard part of curricula in linguistics, and the recognition that the public image that Cognitive Linguistics presents to young
researchers and newcomers does not contain a commitment to empirical
research as a conspicuous feature.
The first factor is undoubtedly the major one: coming from traditional
curricula in language and linguistics, many cognitive linguists are simply
at a loss on how to get started on empirical research of an experimental or
a corpus-based kind. (And, to repeat, the present paper will not immediately help them, because it was conceived as a metatheoretical overview
and not as a practical guide.) The practical steps to be taken are indeed nu-
40
Dirk Geeraerts
merous and often cumbersome. In the case of corpus research, you need to
collect appropriate corpus materials, retrieve the relevant observations
from the corpus, code the observations for all the features that you think
might be relevant in the analysis, perform the right quantitative analysis
and then start your second round in the empirical cycle with an interpretative effort. In the case of experimental research, the practical difficulties
are similar: you need to set up an experimental design with the right stimuli (something which you may achieve only after an initial pilot study),
find an appropriate set of experimental subjects, carry out the actual experiments, analyze them with the relevant techniques and then start your
second round in the empirical cycle with an interpretative effort.
It is no surprise then, that linguists trained in a more purely analytic
paradigm need time and help to get acquainted with this form of doing research. At the same time, there are clear indications that many within the
Cognitive Linguistics community are willing and waiting to receive such
an initiation. A good example is the appeal of the collostructional approach introduced by Stefanowitsch and Gries (2003). It applies existing
methods of statistical collocation analysis (i.e. methods that measure the
strength of association that exists between two elements) to the relationship between the lexical items that occupy the slots in a construction and
the construction as such. Cast in a readily applicable format, the method
has soon attracted a number of scholars working within Construction
Grammar: see Gries and Stefanowitsch (2006) for a sample. From the perspective of the present chapter, the appeal of the collostructional method
shows that there is indeed a demand for the adoption of empirical techniques thus strengthening the assumption that a lack of familiarity is
one of the main explanations for the situation sketched in the previous
section.
At the same time, the collostructional example reveals a potential
danger: not much would be achieved if cognitive linguists, out of ignorance of alternatives, would simply equate empirical research with readily
available specific techniques like the collostructional analysis. What we
should achieve, in the long run, is for cognitive linguists to be aware of the
full range of empirical techniques, experimental as well as corpus-based.
But that will require a reorganization of the average curriculum and a
long-term investment in empirical training.
A second practical factor that should be mentioned involves the
example provided by the leading authorities in Cognitive Linguistics:
what is the image that newcomers and young researchers get when they
are confronted with the work of the most highly valued researchers in the
41
42
Dirk Geeraerts
require a thorough investigation of actual usage, and in any case a methodology that goes beyond just enumerating examples. Next, concluding
from the mere presence of a form in the language to a certain way of
thinking requires at least that the productivity of the image inherent in the
message is established, i.e. that it is made clear that the metaphor is not a
dead one but actually influences or reflects, to a greater or a lesser extent,
thinking. To be explicit, the point made here is not that attempts to establish an effect of language on thinking by means of adequate techniques do
not exist, but rather that such an approach is far from dominant in research in the realm of Conceptual Metaphor Theory.
In other words, what is probably the most broadly attractive and most
widely known form of research in Cognitive Linguistics is at the same
time one of the most facile ones in methodological terms, in spite of the
long-standing efforts of people like Ray Gibbs (and more recently, researchers like Lera Boroditsky, 2001) to provide empirical underpinnings
for Conceptual Metaphor Theory. To the extent that this contributes to a
certain image of what it involves to do interesting work in Cognitive Linguistics, this is not advantageous for the expansion of empirical methods
in Cognitive Linguistics.
5.2. Reasons of principle
In addition to this absence of empirical tendencies in the dominant public
image of Cognitive Linguistics both in terms of the attitude of its leading figures and in terms of its best known fields of application there
might be a more principled reluctance with regard to cutting edge empirical work. Cognitive Linguistics considers itself to be a non-objectivist theory of language, whereas the use of corpus materials involves an attempt
to maximalize the objective basis of linguistic descriptions. Is an objectivist methodology compatible with a non-objectivist theory? Isnt any attempt to reduce the role of introspection and intuition in linguistic research
contrary to the spirit of Cognitive Linguistics, which stresses the semantic
aspects of the language and the meaning of linguistic expressions is the
least tangible of linguistic phenomena. Because meanings do not present
themselves directly in the corpus data, will introspection not always be
used in any cognitive analysis of language? (For an explicit defense of
such a position, albeit in terms of intuition rather than introspection,
see Itkonen 2003.)
There seems to exist a tension, in other words, between a broad methodological tendency in Cognitive Linguistics that considers introspection
43
the most or perhaps the only appropriate method for studying meaning,
and a marginal but increasing tendency to apply empirical methods that
are customary in the other cognitive sciences.
To summarize: the dissemination of empirical methods in Cognitive
Linguistics is far from complete, because some of the leading figures adhere (if not in principle, at least in practice) to a more traditional linguistic
methodology; because the most broadly appealing branches of Cognitive
Linguistics tend to be the methodologically least constrained ones; because the use of empirical methods is insufficiently incorporated in Cognitive Linguistics programs; and maybe also because there is a certain degree of theoretical skepticism with regard to empirical methods.
Because the latter argument is clearly the most fundamental one (it is
the only one that could not be countered on a very practical level), let us
try to assess its weight. Just how essential is a principled skepticism with
regard to empirical methods? The argument may in fact be construed in
two ways, a minimalist and a maximalist one. The minimalist position entails that empirical methods may never completely replace a more interpretative approach: even though it is granted that empirical methods yield
highly valuable circumstantial evidence with regard to the semantics of
natural language, there is an irreducible core to meaning that cannot be
accessed from the outside, with the indirect evidence provided by objective
methods, but that necessarily has to be experienced subjectively. The
maximalist position involves the view that a hermeneutic, interpretative
methodology is not only inevitable but also vastly superior when it comes
to the study of meaning, and that the addition of empirical evidence is
basically superfluous. The minimalist reading of the skeptical argument,
in short, implies that empirical evidence is not sufficient for a proper study
of meaning, while the maximalist reading implies that it is not necessary.
Both arguments can be easily countered. The minimalist reading of the
skeptical argument will not stand in the way of an attempt to maximize
the benefit that one can get from empirical evidence. Even if you accept
that subjectivity cannot be reduced for a 100 %, the practical imperative
for research will be to push the employment of methods aspiring to objectivity as far as it can gets that is to say, the practical consequences of the
minimalist position will be virtually non-existent: you would still try to
pursue the empirical line of research as far as you can take it.
The maximalist reading of the skeptical argument, on the other hand,
will have difficulty to stand up against the line of reasoning developed in
the second section of this paper. If it is indeed important for the further
progress of linguistics to achieve a higher degree of intertheoretical com-
44
Dirk Geeraerts
6.
45
tion is fruitful in the long run, a number of recommendations for the future development of Cognitive Linguistics may be spelled out.
First, if it is correct that the spread of empirical methods is hampered
by insufficient knowledge of approaches and techniques, didactic efforts
are necessary to make cognitive linguists acquainted with corpus analysis,
experimental design, survey techniques, and statistical processing of data.
A program in Cognitive Linguistics worthy of that name should resolutely
opt for advanced training in empirical linguistics. In the process, it will
have to be made clear what the essence of empirical research really is, so
that the misconceptions that now sometimes mar the debate may be
avoided. No, empirical research does not involve abandoning theory
formation in favor of purely descriptive research: rather, it involves trying
to provide proof for theories, and from there, refining the theories. No,
empirical research does not imply that intuition and interpretation have
no role to play in linguistic research: rather, it implies that interpretation
is but one step in the empirical cycle of successful research. And no, empirical research is not about restricting the investigation to one kind of
method or technique: rather, it is about using experimental, and corpusbased, and other empirical approaches in combination to achieve maximally reliable results.
Second, the standards of scholarship in Cognitive Linguistics will gradually have to be raised to such a level that unsubstantiated claims will become unfashionable. This is not to say that the bold and inspired analytical and theoretical work that sparked off Cognitive Linguistics to begin
with, should be ruled out: that would be a counterproductive check on the
creativity of the community of cognitive linguists (and if there is one thing
you cannot deny them, its creativity). Nor is this to say that the desired
level should be achieved in one go and simultaneously by all: what we are
talking about is a medium-term development that leaves room for different paces of development, just like it is in theory development. But even if
the speed with which Cognitive Linguistics will get there may be moderate
and variable according to the circumstances, the direction of the development should be clear, and all those who have responsible positions in Cognitive Linguistics leading authors, editors of journals and books, referees, conference organizers should consciously and consistently pay more
attention to empirical grounding, to keep that development going.
And third, Cognitive Linguistics (and linguistics at large, for that
matter) will have to learn how to live with a slower pace of development.
Empirical thoroughness and reproducibility come at a price: that of painstaking and time-consuming effort. While the late 20th century in linguis-
46
Dirk Geeraerts
tics may have had a tendency to prefer conceptual audacity and the construction of grand theories, these are values that we associate primarily
with 19th century philological accuracy and meticulousness. Perhaps,
then, the early 21st century may profit from a return of some the 19th century spirit
References
Barlow, Michael, and Suzanne Kemmer (eds.)
2000
Usage-based Models of Language. Stanford: CSLI.
Berman, Ruth A., and Dan I. Slobin
1994
Relating Events in Narrative: A Crosslinguistic Developmental Study.
Hillsdale: Lawrence Erlbaum Associates.
Boroditsky, Lera
2001
Does language shape thought? Mandarin and English speakers conceptions of time. Cognitive Psychology 43: 122.
Bybee, Joan
2001
Phonology and Language Use. Cambridge: Cambridge University
Press.
Croft, William, and Esther J. Wood
2000
Construal operations in linguistics and artificial intelligence. In Meaning and Cognition, Liliana Albertazzi (ed.), 5178. Amsterdam/Philadelphia: Benjamins.
Deane, Paul D.
1992
Grammar in Mind and Brain: Explorations in Cognitive Syntax. Berlin/
New York: Mouton de Gruyter.
Deignan, Alice, and Liz Potter
2004
A corpus study of metaphors and metonyms in English and Italian.
Journal of Pragmatics 36: 12311252.
Dirven, Ren, Louis Goossens, Yvan Putseys, and Emma Vorlat
1982
The Scene of Linguistic Action and its Perspectivization by Speak, Talk,
Say, and Tell. Amsterdam/Philadelphia: Benjamins.
Dirven, Ren, Roslyn Frank, and Martin Ptz (eds.)
2003
Cognitive Models in Language and Thought. Ideology, Metaphors and
Meanings. Berlin/New York: Mouton de Gruyter.
Dirven, Ren, and John Taylor
1988
The conceptualisation of vertical space in English: the case of tall. In
Topics in Cognitive Linguistics, Brygida Rudzka-Ostyn (ed.), 379402.
Amsterdam/Philadelphia: Benjamins.
Geeraerts, Dirk
1997
Diachronic Prototype Semantics. Oxford: The Clarendon Press.
2003
Cultural models of linguistic standardization. In Cognitive Models in
Language and Thought. Ideology, Metaphors and Meanings, Ren
47
48
Dirk Geeraerts
Kristiansen, Gitte
2003
How to do things with allophones: linguistic stereotypes as cognitive
reference points in social cognition. In Cognitive Models in Language
and Thought: Ideologies, Metaphors, and Meanings, Ren Dirven, Roslyn Frank, and Martin Ptz (eds.), 69120. Berlin/New York: Mouton
de Gruyter.
Krug, Manfred
2000
Emerging English Modals: A Corpus-Based Study of Grammaticalization. Berlin/New York: Mouton de Gruyter.
Kuhn, Thomas S.
1962
The Structure of Scientific Revolutions. Chicago: The University of Chicago Press.
Labov, William
1972
Some principles of linguistic methodology. Language in Society 1:
97120.
Lakoff, George
1990
The Invariance Hypothesis. Cognitive Linguistics 1: 3974.
Langacker, Ronald W.
1988
A usage-based model. In Topics in Cognitive Linguistics, Brygida
Rudzka-Ostyn (ed.), 127161. Amsterdam/Philadelphia: Benjamins.
1999 Reference point constructions. In Grammar and Conceptualization,
Ronald W. Langacker, 171202. Berlin/New York: Mouton de
Gruyter.
2005
Construction Grammars: cognitive, radical, and less so. In Cognitive
Linguistics. Internal Dynamics and Interdisciplinary Interaction, Francisco J. Ruiz de Mendoza Ibez, and M. Sandra Pea Cervel (eds.),
101159. Berlin/New York: Mouton de Gruyter.
Levinson, Stephen
2003
Space in Language and Cognition. Cambridge: Cambridge University
Press.
Maldonado, Ricardo
2002
Objective and subjective datives. Cognitive Linguistics 13: 165.
Nagel, Ernst
1961
The Structure of Science. London: Routledge and Kegan Paul.
Nuyts, Jan
2005
Brothers in arms? On the relations between Cognitive and Functional
Linguistics. In Cognitive Linguistics. Internal Dynamics and Interdisciplinary Interaction, Francisco J. Ruiz de Mendoza Ibez, and M. Sandra Pea Cervel (eds.), 69100. Berlin: Mouton de Gruyter.
Pederson, Eric
1995
Language as context, language as means: Spatial cognition and habitual language use. Cognitive Linguistics 6: 3362.
1998
Spatial language, reasoning, and variation across Tamil communities.
In Language and Location in Space and Time, Petr Zima, and Vladimr
Tax (eds.), 111119. Mnchen: Lincom Europa.
49
Rudzka-Ostyn, Brygida
1988
Semantic extensions into the domain of verbal communication. In
Topics in Cognitive Linguistics, Brygida Rudzka-Ostyn (ed.), 507554.
Amsterdam/Philadelphia: Benjamins.
Sanders, Ted J. M., and Morton Ann Gernsbacher
2004
Accessibility in text and discourse processing. Discourse Processes 37:
7989
Sandra, Dominiek, and Sally Rice
1995
Network analyses of prepositional meaning: Mirroring whose mind
the linguists or the language users? Cognitive Linguistics 6: 89130.
Schulze, Rainer
1988
A short story of down. In Understanding the Lexicon: Meaning, Sense
and World Knowledge in Lexical Semantics, Werner Hlllen, and Rainer
Schulze (eds.), 394410. Tbingen: Max Niemeyer Verlag.
Speelman, Dirk, Stefan Grondelaers, and Dirk Geeraerts
2003
Profile-based linguistic uniformity as a generic method for comparing
language varieties. Computers and the Humanities 37: 317337.
Stefanowitsch, Anatol
2003
Constructional semantics as a limit to grammatical alternation: The
two genitives of English. In Determinants of Grammatical Variation,
Gnter Rohdenburg, and Britta Mondorf (eds.), 413444. Berlin/New
York: Mouton de Gruyter.
Stefanowitsch, Anatol, and Stefan Th. Gries
2003
Collostructions: Investigating the interaction between words and constructions. International Journal of Corpus Linguistics 8: 209243.
Tomasello, Michael
2003
Constructing a Language. A Usage-based Theory of Language Acquisition. Harvard: Harvard University Press.
Tummers, Jos, Kris Heylen, and Dirk Geeraerts
2005
Usage-based approaches in Cognitive Linguistics: A technical state of
the art. Corpus Linguistics and Linguistic Theory 1: 225261.
Tummers, Jos, Dirk Speelman, and Dirk Geeraerts
2005
Inflectional variation in Belgian and Netherlandic Dutch: A usagebased account of the adjectival inflection. In Perspectives on Variation.
Sociolinguistic, Historical, Comparative, Nicole Delbecque, Johan van
der Auwera, and Dirk Geeraerts (eds.), 93110. Berlin/New York:
Mouton de Gruyter.
50
Dirk Geeraerts
51
Abstract
In this paper I take issue with the standard view of polysemy as the association of
a single word-form with a multitude of distinct meanings. First, I review different
approaches to polysemy, concluding that it may not be legitimate to attempt to determine just how many meanings a word has, and what these meanings might be. I
suggest that such a programme may be the outgrowth of a particular conception
of language, which I characterize as the dictionary+grammar book model. This
model is undermined by the pervasiveness of the idiomatic in language. The
matter is illustrated by a corpus-based study of (some uses of) the expression all
over. The expression has a range of uses which cannot be derived on the basis of
independently identifiable meanings of its parts. All over, in turn, participates in
larger expressions whose semantic values likewise cannot be compositionally derived. The creative extension of these idiomatic uses is also documented. These
findings lend support to the view that knowing a word involves knowing the usage
range of the word, rather than through the association of the word with a fixed
number of determinate meanings.
Keywords: polysemy; lexicon; polyseme; metonymy; idiom; construction; creativity; compositionality; network
1.
Some years ago I wrote a paper entitled How many meanings does a
word have? (Taylor 1992). I began by juxtaposing some remarks by John
Searle on the verb to open and observations by George Lakoff on the word
window.
x
Lakoff (1987: 416) pointed out that window can be used to refer to different categories of things. If someone throws a rock and breaks a
window, what is broken is (most probably) a glass panel, whereas if
someone paints a window, what they paint is (most probably) the
wooden structure encasing the glass panels. Deliver the window, brick up
the window, sit in the window again refer to different kinds of entity. For
Lakoff, this was evidence that the word was polysemous, that is, it had
distinct, though closely related meanings.
52
x
John R. Taylor
Searle (1983: 1458), discussing the verb open, noted that a person who
opens a door and a surgeon who opens a wound do very different things
to the entities in question. You would not, for example, open a door by
making incisions in it with a surgical scalpel. The truth conditions for
open a door and open a wound the kinds of things that a person
would have to do in order for the expressions to properly apply to their
actions are therefore different in the two cases. Searle, however, does
not draw the conclusion that open is polysemous. Rather, he claims that
the verb has a fixed and constant value. Truth conditions emerge as a
consequence of the way in which we interpret the expressions, relative
to what he calls the Background, that is, our knowledge of cultural
practices, how we interact with the world, and how the world is constituted. Searle illustrates the role of interpretation relative to the Background on the example Sam opened the Sun. We can, Searle claims,
readily understand this expression on the basis of the semantic content
of the component words, but we cannot generate a set of truth conditions for it. Opening the Sun is not an established practice, and we have
no idea what a person who opened the Sun is supposed to have done.
53
open the window would have n1 n2 possible readings, and could only be
understood as the result of a complex process of disambiguation. This
conclusion is patently counterintuitive. If anything, the more context
that is added to a word, the less likely it is to be perceived as ambiguous.
The number of distinct polysemes of a word is not only likely to be very
large, it may well be indefinitely large. For example, the number of different things that can be opened, and the different sets of truth conditions associated with these various opening activities, is in principle
open-ended. Tolerating large-scale polysemy is one thing, admitting indefinitely large polysemy would border on the absurd (Searle 1983: 146).
Searles position, which assumes a single semantic value for a word, is also
not without its problems.
x
54
John R. Taylor
The above remarks suggest that tolerance for extreme polysemy on the
one hand, and attempts to eliminate polysemy on the other, are both seriously flawed. Both, it will be noted, raise questions of psychological
plausibility. Can we reasonably suppose that speakers of a language have
stored, and access, and make choices between large numbers of polysemes, and that the understanding of even very banal expressions requires
a massive amount of disambiguation? Can we, on the other hand, suppose
that speakers have stored schematic meanings for each lexical item (meanings which may actually be quite difficult to articulate) and that they
really do apply to these all manner of elaboration strategies? Observe that
the question of cognitive economy cuts both ways. Unitary meanings
economize on storage but necessitate the computation of contextual interpretations. Polysemy economizes on the computation of contextual
variants but increases storage demands; it also requires disambiguation
procedures. We should also bear in mind that while a streamlined lexicon
might appeal to some theorists on account of its elegance, or parsimony, there are no principled grounds for setting an upper limit on the
amount of polysemy that the lexicon can tolerate, or for deciding, a priori,
what an optimal amount of polysemy might be.
Given the problems associated with rampant polysemy on the one hand
and unitary abstract senses on the other, a reasonable approach would be
to seek a compromise. We can imagine two kinds of compromise:
x
55
senses are not feasible. In other words, we strive for as little polysemy as
possible, but permit as much as necessary, arriving at an account which
eschews both rampant polysemy and strict monosemy.
Having it both ways. We allow both fine-grained semantic distinctions
and more general, schematic meanings which capture the commonality
amongst groups of specific meanings. The meanings, at various levels of
schematicity, would be typically displayed in a network format, with
more specific meanings standing as elaborations of more schematic
meanings.
I consider these two approaches below, and some of the issues that they
raise.
2.
56
John R. Taylor
kind of trajector, its manner of motion, and the kind of landmark entity,
in question.
Take, as another example, the variable reference of the word window,
cited above. This might be handled in terms of the active zone phenomenon (Langacker 1990). Whenever an entity is involved in a predication,
typically only some facet, or facets, of the entity will participate in the
predication, whereby the relevant active zone can be inferred on the basis
of general conceptual knowledge of the entity and of the kind of predication that it participates in. Given what we know about windows, and
about what it means to paint something, we can reasonably infer that
paint a window implicates only the paintable facets of the window,
namely, its wooden frame. Likewise, when we brick up a window, we
can infer that the reference of window has got to be a vacant gap in a wall,
not a wooden frame, or a pane of glass.
Appeal to the active zone phenomenon begs the question of what the
basic, or real meaning of a word is. What kind of thing does window
before the active zone phenomenon kicks in refer to? It cannot be the
sum of its active zones wood, glass, and a hole in a wall do not add up to
a coherent category. Rather, we want to say that the word window provides
access to a body of conceptual knowledge, concerning the typical location, the typical function, and the typical manner of construction of the
things in question. It would follow that window, as a decontextualized
word, does not actually refer to a thing at all, but rather stands at the hub
of a conceptual network; for contextualized uses, the relevant concept
would be conjured up in accordance with the specifications of a particular
scene (cf. Barsalou 1987). Our impression that the word does have a determinate reference would be due to the fact that certain contextualized
uses such as open the window frequently re-occur, thereby entrenching
the relevant conceptualization.
As Langacker noted, the active zone phenomenon gradually shades
into the more general process of metonymy (see also Taylor 2003b: 127).
The named entity functions as a reference point for accessing the intended
referent. The intended referent may be a part of the named entity, or it
may be something that is closely associated with it within a given knowledge configuration. Interestingly, processes of metonymic extension typically apply across the board to words of a certain semantic category. A
well-known example concerns the names of institutions. University can be
used to refer to the institution, to the building which houses the institution, and even to a person who acts on behalf of the institution. Compare:
The University was founded in the 19th century, I parked near the University,
57
58
John R. Taylor
path). But while you can walk towards the sea or drive towards the
mountains, but it would be odd to say of someone that they live towards the sea, or live towards the mountains.3
These latter cases would be particularly problematic for attempts to
systemically regulate the incidence of polysemy. They concern, not so
much the availability of different senses on the basis of general polysemization processes, but the non-occurrence of otherwise expected senses.
To the extent that the different senses of museum, etc., are predictable
from general processes at work in English, we could reduce the amount
of polysemy in the lexicon by not separately listing the different senses.
But in the case of government and parliament, we would need to flag these
words as not undergoing the polysemization process. Paradoxically, the
words with fewer different uses would need a more complex representation than the ones which undergo regular polysemization. But even
flagging a word as an exception might not be fully adequate in the case
of towards. It is not that towards does not allow the place as end-point
reading; what are odd, are some specific examples which invoke this
reading, such as its occurrence with live in connection with a geophysical
landmark.
3.
59
I have characterized and presented arguments for the network approach to polysemy elsewhere (Taylor 1995). The network model cannot,
however, be regarded as the last word on polysemy. In recent years, the
model has in fact come in for some substantial criticism. These are both
methodological and conceptual in nature:
(i) Methodological problems. Suppose two linguists are working independently on the same word and each makes a proposal for a polysemy network. Suppose, furthermore, that the proposed networks differ in important respects. One linguist identifies a large number of highly specific senses
which the other ignores, or there is disagreement about how particularized
senses are to be allocated to more schematic senses and how these schematic senses are to be characterized. What criteria do we apply for evaluating the two accounts? The example is by no means hypothetical. Consider, for example, the still ongoing story of over.4 What has driven the
discussion of this word, for more than two decades since Brugmans (1981)
pioneering dissertation, is disagreement over these methodological questions. How many different senses does the word really have and how are we
to decide the matter? What is the content of these different senses? Should
we accord representational status to highly particularized senses? To what
extent is it feasible to collapse different uses under more schematic representations and what is the content of the proposed schemas? Further questions have been, how these different sense clusters relate to each other, and
which, if any, of the proposed senses can be regarded as the most central.
(ii) The second issue concerns the ontological status of networks. Is a network supposed to have psychological reality? Are networks meant as hypotheses about speakers knowledge of words? If so, how can the psychological reality of a proposed network be tested? If not, what is a network
supposed to be representing? What is the relevance of the networks in language acquisition? At what point in the network is a word accessed in production and reception? These are still very much open questions (Sandra
& Rice 1995; Hallan 2001; Rice 2003; Langacker 2006).
4.
In view of the kinds of problems that arise when one attempts to give a detailed semantic description of a word, it comes as no surprise that different
scholars (and lexicographers) often fail to converge on a common answer
60
John R. Taylor
to the question of just how many different meanings a word has and what
these meanings are supposed to be (Fillmore & Atkins 2000). Given this
state of affairs, there are, broadly speaking, two approaches that an analyst
might take. One would be to continue refining the criteria by which two
uses are to be regarded as elaborations of the same meaning or representative of distinct polysemes. But given the less than unanimous outcomes of
research to date, it is not surprising that a number of scholars have recently
come to question the traditional notion of polysemy (namely, as the association of a single word-form with more than one distinct meaning), and the
reification of meaning which the notion entails (Allwood 2003; Geeraerts
1993; Zlatev 2003; Langacker 2006). It may be well to stand back a little
from the discussion and consider the theoretical issues that drive it.
Elsewhere (Taylor 2003a, c) I have suggested that a desire to associate a
word with a fixed number of distinct meanings derives from a specific
conceptualization of the object of investigation. I have referred to this
(Taylor, in press, a) as the generative model of language, or, more colloquially, as the dictionary+grammar book model. According to this
model, knowledge of a language resides in two components. One component is the dictionary, or lexicon, which lists the basic meaning-bearing
units of the language (prototypically, the words). The other component is
the syntax, which lists the rules whereby smaller elements can be combined. The rationale for this view of linguistic knowledge is explained in
the following remarks by Jackendoff:
Since the number of possible utterances in a human language is unlimited, language users cannot store them all in their heads. Rather, knowledge of language requires two components. One is a finite list of structural elements that
are available to be combined. This list is traditionally called the lexicon, and
its elements are called lexical items. The other component is a finite set of
combinatorial principles, or a grammar. To the extent that speakers of a language (or a dialect) are consistent with one another we can speak of the
grammar of a language as a useful approximation to what all its speakers
have in their heads. (Jackendoff 2002: 39)
For a first approximation, the lexicon is the store of words in long-term memory from which the grammar constructs phrases and sentences. (Jackendoff
2002: 130)
61
according to the following rules of grammar (you hand her an equally massive
English grammar). [B]etween them they would tell the reader how to make
any conceivable English sentence. (Baker 2001: 534)
62
John R. Taylor
63
In the remaining part of this chapter I wish to demonstrate that searching for the meaning of a word may be a futile exercise, once we take account of the contexts in which the word is used. This is because contextualized uses may take on semantic values that cannot be compositionally
derived. The process, moreover, can be recursive, in that a contextualized
use may itself have contextual variants. As we consider the properties of
ever larger configurations, the semantic contribution of component words
will become less and less of a concern.
5.
If it is true that language use is riddled with idioms, of various kinds, the
question of how many different meanings a word has may not be a legitimate thing to ask. What we should be concerned with is not the meanings
of a word as such, but the uses of a word, in various kinds of linguistic
context. For an illustration of this, we need go no further than the example
of over.
There is general agreement that, of the various senses of over, one has to
do with a trajector covering, or occluding, a landmark. Lakoff cites the
following examples:
(1) a. She spread the tablecloth over the table. (Lakoff 1987: 419)
b. The board is over the hole. (Lakoff 1987: 427)
Here, the trajector is at least two-dimensional and extends across the
boundaries of the [landmark] (p. 426), a view endorsed by Tyler and
Evans (2001), who note that the trajector is larger, or perceived to be
larger than the [landmark], such that the trajector physically intervene[s] between the viewer and the [landmark], thereby obscuring the
landmark from view (p. 752).
Here are some examples of the covering sense of over, taken from the
British National Corpus (BNC):5
(2) a. I held my hand over my face for a few moments before speaking.
b. His black trousers were held together by safety-pins, paperclips
and needles. Over this he had a black mackintosh; there were five
belts strapped around his waist and a sort of grey linen nappy attached to the back of his trousers.
c. Tugging her jeans over her thighs, she said:
d. She folded her scarf over her blonde hair.
64
John R. Taylor
It will be noted, to begin with, that none of these examples exhibits the
covering sense with the pristine clarity of Lakoff s (invented) examples
(nor, it might be added, of the invented examples which other scholars,
such as Tyler and Evans 2001, and Deane 2005, have cited in this connection). There are, moreover, some subtle differences between the uses in (2).
In (a) the idea seems to be that the trajector (a hand) partially obscures
the landmark (a face) partially, because a hand is smaller than a face and
cannot fully obscure the face from view. In (b), the mackintosh is a large,
outer garment, which wraps around other clothing, such as the trousers
and the belts, and hides them from view. In (c), the trajector is moved into
a position which covers the landmark, though not with the idea of obscuring it, but rather of enclosing it, and adapting to its shape. In (d), the relevant point seems to be the positioning of the scarf, rather than the obscuring of the hair.
Rather than dwelling on the covering sense of over which, to be sure,
would merit a much closer examination than what it has been accorded in
the already voluminous literature on the word I want to move on to the
use of over in a larger expression. Some 4.2 % of all uses of over in the
BNC occur in the phrase all over. Here are some examples:
(3) a. More often because of his impatience, it blew back in his face,
dusting him all over.
b. Lamarr jumped back dropping the broken pieces and with blood
all over his hand and face.
c. Shes got the dress on and shes taken her hat off so her hairs all
over her shoulders.
d. I still weigh 66 kilos and still have a galaxy of spots around my
nose and all over my chin.
e. Oh, God, I thought, she can see the lipstick all over me.
f. She was trembling all over, holding on to the car door for support,
despising herself for her own weakness.
All over is commonly analyzed as a variant of covering over. Here, I
want to raise the question of the contribution of all. An obvious answer,
and one consistent with the compositionality principle, would be that all
has a quantificational sense, in that the coverage, or occlusion, is total.
The meaning of all over, therefore, would be a function of the covering
meaning of over in combination with the quantifying meaning of all (Lakoff 1987: 428). There are several problems with this account.
65
(i) All is not readily omitted in some of the examples with all over, suggesting that all over functions as a unit:
(4) a. *It blew back in his face, dusting him over.
b. *Oh, God, I thought, she can see the lipstick over me.
(ii) Equally, over, in (2), can often not be quantified with all:
(5) a. * I held my hand all over my face for a few moments before speaking.
b. *Tugging her jeans all over her thighs, she said:
c. *She folded her scarf all over her blonde hair.
Note, in connection with (5a), that the strangeness of the sentence is not
due to the factual impossibility of a hand covering all of a persons face. I
held both my hands all over my face (thereby totally obscuring my face) is
equally bizarre.
(iii) All does not readily contrast with other quantifying expressions,
such as half, partly, somewhat, mostly, etc.
(6) *Oh, God, I thought, she can see the lipstick {half, partly, somewhat,
mostly} over me.
The oddness of half over, in this example, in not due to the impossibility of
the expression per se. There are 32 occurrences of half over in the BNC (as
against 5479 occurrences of all over). Here are some examples:
(7) a. She pulled the sheet half over her head, to be ready to jump out,
and she waited.
b. Maggie looked up, peering at him, the towel like a white tent over
her head, her long hair half over his face.
c. Meredith, stranded inelegantly with a foot on the wooden step
and the other leg half over the topmost bar, said crossly, I cant
see it.
d. Allain moved until his body was half over her.
Here, some notion of quantification does seem to be involved. Half can be
replaced by partly, it can be omitted without too much of a strain on acceptability, and even, in the case of (7b), it might be replaced by all. But if
66
John R. Taylor
67
(v) The major objection to the compositional account of all over, however, is that covering, whether quantified or not, is hardly the appropriate way to characterize the examples in (3). In fact, the examples in (3) do
not exhibit a covering relation at all, in the strict sense that the trajector
occludes the landmark. Rather, as suggested above, what unifies the
examples is some notion of the distribution, or dispersal of the trajector
entity within the boundary of the landmark, or, in the case of (3f), with
the manifestation of a process at different places on the landmark. The
point has been developed by Queller (2001: 58), who drew attention to the
following contrast:
(10) a. This tablecloth has got bloodstains all over it.
b. ?? This tablecloth has got red squares all over it.
According to Queller, example (10b) is unacceptable because the red
squares are a design feature of the tablecloth. Sentence (10a), in contrast,
is fine, and has to do with the random dispersal of the stains on the surface of the cloth, whereby, as noted above, the stains, as a mass, do not
strictly cover the landmark a point also acknowledged by Lakoff
(1987: 428). Moreover, the stains are perceived as being an unwanted imperfection on the cloth, or, at least, a temporary blemish. If we now go
back to examine the examples in (3), we note that the idea of random distribution, in association with temporary blemish or dishevelment, applies
in all cases.
What this means, I think, is that the semantic value of all over cannot be
computed compositionally, from one of the meanings (the covering meaning) of over, in association with all. All over would have the status of an idiomatic expression. As such, it has its own specific conditions of use. In this
connection, it is interesting to take note of a property which is shared by all
of the BNC examples with all over cited above and which distinguishes
these examples from the invented sentences in (1). This is, that the attested
examples all make reference to the human body, or parts of it, either as trajector or as landmark, or as both, or, in the case of (2b), to an article of
clothing located on a human body. As far as I am aware, this distinctive feature of the dispersal sense of all over has not previously been commented
upon, and further corpus research would be needed in order to substantiate
the association. If substantiated, this property would contribute to what
Stubbs (1995) has referred to as the semantic profile of the expression.
The expression, to be sure, has other semantic profiles. Particularly prominent are uses of all over which have a geopolitical entity as the landmark.
68
John R. Taylor
(11) a. Why are managers from all over Britain and sometimes even from
abroad, even from Japan, attracted to start industries on these
trading estates?
b. People must be hunting all over the country for him.
c. Police all over the country are known to want the DNA register.
d. Pietro Metastasio (whose texts were set to music all over Europe by
the best composers of the time)
e. The worship of, of professed belief in, a god of some kind is to be
found in varying degrees of sincerity all over the world.
The discovery that all over functions as a unit, with its own distinctive
semantic value(s), may not, in itself, be all that remarkable. The matter becomes more interesting, however, when we realize that this process of
idiomaticization is recursive. Just as over combines with all to give an
expression with its own distinctive and idiomatic value(s), so all over can
combine with other items to produce expressions which also have their
distinctive values.
As many as 8.6 % of all occurrences of all over in the BNC are in the
context all over the place. The place in question is not specified precisely.
(12) a. Why somebodys left an old choc-ice on the floor, and its run all
over the place.
b. Although Shaker has become fashionable and Shaker style furniture is appearing all over the place, very little of it bears much relation to the real thing.
c. We started using it and interestingly enough that typeface is really
popular now, you see it in the United States everywhere, its all
over the place.
d. There is no way that a farrier can shoe a horse that is so angry or
frightened that it is rearing and leaping all over the place.
In (12a-c), all over the place conveys that the designated process is manifested at numerous places (within a contextually defined region); everywhere (= at all places within the defined region) offers itself as a suitable
(though perhaps hyperbolic) paraphrase. Substitution with everywhere
would not be possible in (12d), however. Here, the idea is that the activity
(rearing and leaping) is taking place in a random and unpredictable
manner. This notion is exploited in the following:
69
(13) Yesterdays polls looked much of a muchness they were not all over
the place but concealed politically crucial variations.
If the results of an opinion poll or any other set of data, for that matter
are all over the place, the idea is that there is no obvious trend; that data
points appear to be random and show no clear pattern. A specialization
of this idea is exemplified by the following (from Queller 2001: 62):
(14) a. This paper is all over the place.
b. You are all over the place.
In saying (14a) to a student, the instructor conveys that the essay lacks
structure; it jumps, seemingly at random, from one point to another, without an overall plan. If the student is said to be all over the place (14b),
the idea, again, is that the students work, or thought processes, lack direction and coherence.
At this point, it may be well to recall the earlier citation from Jackendoff. Jackendoff presented the standard argument for the dictionary+grammar book model of language. This has to do with the fact of linguistic creativity. The fact that speakers are constantly creating novel
sentences means that speakers to do not learn sentences, they learn the
principles by which sentences can be generated. It follows so the argument goes that speakers need to learn a (finite) set of meaning-bearing
units (the words) and a finite set of rules for combining the words. Idioms,
being non-compositional, would constitute a non-essential supplement to
the lexicon and the syntax. This view, however, ignores the fact that idioms
can also be used creatively. The usage range of an idiom can be extended
to new contexts (as when all over the place is extended from its more basic
meaning everywhere to conveying random distribution or directionless
behaviour). Alternatively, the slots in an idiom can be filled by different
lexical items. And, indeed, there exist a number of variants of all over the
place; these include all over the show, all over the shop, and all over the map.
All over the show occurs three times in the BNC; in none of these is a
show (in the basic, literal sense of the word) implicated:
(15) a. Squealing and squawking and leaving its dirty nappies all over the
show.
b. Speaking of Deano everyone I talked felt that as we had paid
2.7 million for a centre forward wouldnt it be nice if hung around
near the goal or in the box for that matter instead of running all
70
John R. Taylor
over the show and when we finally get the ball in the box theres nobody there.
c. My legs were all over the show and Ive got to stand tall tomorrow.
The first example can be paraphrased by everywhere, while the second
conveys running in random directions, without purpose. The third elaborates on a specific case of random behaviour, namely, that the person did
not have control over the movements of his/her legs.
All over the shop is attested 9 times. The first example below aligns itself
with those uses of all over which take a geopolitical entity as its landmark,
while the second is a variation on the theme of random, directionless activity. Neither of the examples involves reference to a shop.
(16) a. Wisely cautious at first ; awed, probably, by sheer numbers (because they were coming in, now, from all over the shop, from Canada, from Palestine) ; German society duly broadened itself to
let the newcomers in
b. The movies dilemma is that its luxurious pace ambles all over the
shop, never really getting anywhere or bringing any of the characters to a satisfying conclusion
All over the map is attested only twice in the BNC. The first example involves, literally, places on a map of a terrain. The second, however,
refers again to the random distribution of data.
(17) a. In the east and centre of the county we find a close network of narrow, winding lanes, wandering from hamlet to hamlet and farm to
farm, churches standing alone, isolated houses dotted all over the
map, many of them called Hall or Old Hall ; significant names
b. Er what I have chosen to do is to do what we were doing in the
nineteen eighties, which is to take the simplest measure of the
maximum flow and the detrusor pressure at that volume and when
we do that, you can see that these patients, instead of being a
single group of patients with a single kind of bladder pressure and
flow, these patients are all over the map.
Google returned over 1.3 million hits for all over the map the figure no
doubt inflated by the existence of an atlas of that name. The following,
however, clearly elaborate on the idea of random distribution, or directionless behaviour.
71
(18) a. When it comes to interests, I am all over the map. I read and
enjoy all genres of novels, but most of my reading is nonfiction.
b. I am all over the map these days. I cant seem to settle into one project.
c. Bond Yields in the Euro Zone Are All Over the Map.
d. The next largest group comes from what he called New Age
cults, whose belief systems are all over the map.
A particularly interesting variant is all over the paddock (not attested in
the BNC). This appears to be a peculiarly Australasian expression, paddock, in Australian and New Zealand English, corresponding to standard
English field, meadow. Google searching suggests that the expression may
be (or may have been) specifically associated with the behaviour of players
on a rugby field:
(19) a. This set the tone for the rest of the game with tries being ran in
from all over the paddock against a side which battled bravely.
b. It was the sort of thing the big men hate to see and no one was
bigger on the night than Springbok lock Botha who was all over
the paddock.
c. The determined and dogged Irish werent giving up and had the
Aussies running ragged all over the paddock
d. We had them all over the paddock, particularly our ruckman LS
Craig We had winners like LSPT Robert Tarjani all over the
paddock.
e. The Welsh scrum didnt push the Wallabies all over the paddock because they were a set of England-style behemoths, it was because
they were better coached.
Not surprisingly, all over the paddock has uses which have nothing to do
with a paddock, or rugby field. In fact, I first noticed this expression in
a commentary on the 2005 general election in New Zealand, where it was
stated that the initial election results were all over the paddock. This
usage is clearly a variation on the one cited in (13). In contrast, the following, from a New Zealand politicians weblog, retains the image of the paddock (or rugby field), fusing it with the notion of the Prime Minister randomly changing her account in response to the news channels persistent
questioning.
72
John R. Taylor
(20) I remember when TV One chased Prime Minister Jenny Shipley all
over the paddock over whether she had discussed politics when having
dinner with Saatchi chief Kevin Roberts.
All over enters into another set of idiomatic expressions. About 1 % of all
uses of all over in the BNC occur in the context written all over. Consider,
first, the following example.
(21) My partner for the event was the marvellous American player Muffin
Spencer-Devlin, who turned up with a bag which had Save the
Whales written all over it.
There is no suggestion, here, that all of the bags (outer) surface was
covered by the inscription (this would be a nonsense, anyway), nor that
the lettering was randomly displayed on the bags surface (though it could
be the case that Save the Whales was inscribed more than once on the
bag). The idea, instead, is that the wording Save the Whales was prominently displayed on the bag, visible to any observer.
Of special interest are cases in which there is no literal writing on a surface; rather, what seems to be involved are certain features of an entity
which are salient to any observer. The following (BNC) examples elaborate
on the idea that the prominent feature is (literally) inscribed on the entity.
(22) a. Glazed, red-rimmed, they stare off into some distant past that has
mid-1980s written all over it.
b. And its got faulty written all over the box, you know.
In the following, the entity bears the name of its author (or instigator):
(23) a. Daves rhythm work has got his name written all over it.
b. This has Paul written all over it.
Zooming in still further on the idiomatic, we find expressions in which the
writing is located on a persons face, this making the persons thoughts
or emotional state apparent to an observer. The inscription may even take
the form of a quoted expression (24b):
(24) a. He shook his head, doubt written all over his face
b. Mr Kopek, I didnt know you had it in you, was the phrase I
could see written all over her expression.
73
Queller proposes that written all over (ones face) is associated with the
unintentional expression of emotion, pointing out that the phrase is more
likely to be used of guilt and disappointment, which a person might prefer
to keep hidden, than of rage or pleasure, which a person freely expresses.
He (2002: 57) glosses the phrase as someones attempt to present a front
of composure or nonchalance is being messed up by an unconscious
or reflexive dispersal of emotion over his/her face. While some BNC
examples, such as (25a), might conform with this account, others, e.g.
(25b), dont.
(25) a. Though Jackson was the first to congratulate Olympic champion
McKoy, the disappointment was written all over his face.
b. As William and Harry boarded the 10.30am BA flight, excitement
was written all over their faces.
There is no reason to suppose, in (25b), that William and Harry would
want to hide their excitement behind an expression of composure or nonchalance. These examples, rather, all elaborate on the notion of the face
as interface. A persons face is where emotions and intentions are made
manifest, and where they can be read by others. This conceptualization
of the face is prominent in the following (Googled) examples:
(26) a. He approaches slowly, anxiety visible all over his face.
b. The Conductor was sitting with satisfaction clearly visible all over
his face
c. Connor peers at him, disgust legible all over his face.
d. the agony was prominent all over his face as was the disappointment.
I mentioned above the possibility of the idioms being creatively extended.
Google searching discovered numerous variations on the expression
written all over (ones) face. Here is a small sample:
(27) a. Frustration was scrawled all over his face.
b. This is a comic, then, who walks on stage with please like me
scrawled all over his face.
c. Mr Eun glanced at the half-empty classroom, and then back at
me, the usual kids these-days look scribbled all over his face.
d. He leant over with compassion printed all over his face in italic
type.
74
John R. Taylor
e.
f.
g.
h.
i.
The man has rising star printed all over his face.
arrogance inscribed all over his face
I cant believe it, he moaned, betrayal stamped all over his face.
His love of life was stamped all over his face.
Selariu dragged himself up to complete his routine but his disappointment was etched all over his face.
j. The delight is visibly etched all over his face.
k. Gourry looks down at me, confusion etched all over his face.
l. Simon looked up from his computer screen, irritation engraved all
over his face.
m.The Source? Simple. Its Dan Rathers daughter. It was typed all
over his face when he defended the story Friday.
n. Zainab looked up at her mother, helplessness portrayed all over her
face.
o. with a terrifying earnestness pictured all over my face
p. Contempt was depicted all over his face.
It would, I think, be futile to approach these examples enquiring which
sense of over (or, indeed, which sense of all over) is being exemplified. The
examples play on the face as interface metaphor, in association with
the entrenched expression (have something) written all over ones face.6
6.
Concluding remarks
75
open a book, speakers are readily able to describe the action, they can perform it, and mime it. But to state the meaning of open, abstracting away
from the action performed on the book, is not only a difficult task, it is
also an artificial one. Speakers may, no doubt, perceive some commonality between the different kinds of opening events. However, this perception of commonality is by no means necessary to guarantee mastery of the
word speakers merely need to have learned the relevant collocations.
Appeal to established usage patterns and their associated practices readily
resolves the disambiguation problem (or, rather, non-problem) that was
raised in opening sections of the present chapter. Open the window is not
perceived as being ambiguous, and no choices amongst the various senses
of open and window are involved. This by no means entails that open and
window have fixed invariant senses, merely that open the window has a
learned sense, it is associated with an embodied practice, and, as Searle
pointed out, it comes with easily statable truth-conditions. New uses,
when encountered, will be interpreted on the basis of their similarity with
already familiar uses. In the case of Searles example open the Sun, speakers would need to determine which of the established uses of open is to be
taken as a model for interpreting the expression. They might construct
some fantastic science-fiction scenario, in which opening the Sun would
be analogous to, say, opening a wound. Or they might well conclude that
the expression is simply nonsense.
In this paper I have focussed on corpus evidence for the reality of usage
patterns, showing that at least some uses of over resist a compositional
analysis. This casts doubt on the idea that the word might have a finite
number of discrete meanings. What was true of over, is no less true of all
over. This expression also has a number of particularized uses, with their
own distinctive semantic values, which, once again, resist a strict compositional analysis.
Wittgensteins Blue Book opens with the question: What is the meaning of a word? (Wittgenstein 1958: 1). Wittgenstein proposed to bring
this rather abstract question a question which, as he put it, is liable to induce in us a kind of mental cramp down to earth by operationalizing it, that is, by replacing it with a more practical question, concerning
how one might set about determining the meaning of a word. His answer,
as is well-known, was that meaning was to be studied by examining how
language is used: to say that a word has a meaning is to say that it has a
use in our language (p. 69). The aphorism, to be sure, begs the question
of what use in [a] language entails, and how it is to be characterized.
This chapter can be read as an attempt to come to grips with this issue,
76
John R. Taylor
with respect to a (very) limited range of uses of over, and of all over. The
picture that emerges is that a speaker of a language does not learn a list of
words (with their meanings) and a list of rules for combining words, as
proposed in the earlier citations from Jackendoff and Baker. A speaker
has learned a vast number of word uses, some more frequent than others,
to be sure, and some more idiomatic than others. This approach need
not be in conflict with the argument from linguistic creativity, which
is standardly put forward in support of the dictionary+grammar book
model of language. On the contrary. Idiomatic expressions can themselves
be creatively extended to new usage situations, and are subject to creative
modification. Creativity need not be the bogey of a non-generative, usagebased linguistics.
Notes
1. The two-level approach is of interest, in that it attempted to address some
of the problems raised above in connection with Searles proposal for unitary
word meanings. The two-level theorists did make proposals for unitary semantic representations for a selected number of words, above all, the prepositions, and for a few other examples, such as institution-type words. In addition, mechanisms were proposed for explicitly deriving conceptual (that is,
contextualized) interpretations from these semantic representations. I have
critically reviewed the two-level approach elsewhere (Taylor 1994, 1995,
2003b). One of my criticisms was that the mechanisms for conceptual elaboration often failed to predict the sometimes quite specific properties of the
words in context. These specific properties could only be derived by appeal to
ad hoc principles of conceptual interpretation. Appeal to such ad hoc principles, however, would be tantamount to recognizing the distinctiveness and
independence of the different readings of the words in question, that is, that
the words were indeed polysemous.
2. Ren Dirven (personal communication) has suggested that government and
parliament behave differently from museum, university, etc., because they
focus on the body politic rather than on the institutions per se.
3. A Google search (April 14, 2006) conformed these intuitions. The phrase
stand towards the back scored 213 hits (alongside 410 hits for stand toward the
back), as against no hits for live toward(s) the sea/mountains.
4. The literature on over is vast. For some recent contributions, see Tyler and
Evans (2001), Queller (2001), and Deane (2005), and the references cited
therein.
5. The examples were generated using the Simple Search facility of the British
National Corpus at http://sara.natcorp.ox.ac.uk/. This facility returns a maximum of 50 randomly selected instances of the search item.
77
6. The examples in (27), as well as earlier examples with written all over, would
profit from an analysis in terms of conceptual blending (Fauconnier & Turner
2003). This is not explored here, for lack of space.
References
Allwood, Jens
2003
Meaning potentials and context: Some consequences for the analysis of
variation in meaning. In Cognitive Approaches to Lexical Semantics,
Hubert Cuyckens, Ren Dirven, and John R Taylor (eds.), 2965. Berlin/New York: Mouton de Gruyter.
Baker, Mark
2001
The Atoms of Language: The Minds Hidden Rules of Grammar. Oxford:
Oxford University Press.
Barsalou, Lawrence
1987
The instability of graded structure: Implications for the nature of concepts. In Concepts and Conceptual Development: Ecological and Intellectual Factors in Categorization, Ulrich Neisser (ed.), 101140. Cambridge: Cambridge University Press.
Bierwisch, Manfred
1981
Semantische und konzeptuelle Reprsentation lexikalischer Einheiten.
In Untersuchungen zur Semantik (= Studia Grammatica XXII), R. Ru
z icka, and Wolfgang Motsch (eds.), 6199. Berlin: Akademiie-Verklag.
Brugman, Claudia
1981
Story of OVER. MA thesis, University of California, Berkeley.
Chomsky, Noam
1980
Rules and Representations. New York: Columbia University Press.
Croft, William
1998
Linguistic evidence and mental representations. Cognitive Linguistics
9: 151173.
Cuyckens, Hubert, and Britta Zawada
2001
Introduction. In Polysemy in Cognitive Linguistics, Hubert Cuyckens,
and Britta Zawada (eds.), ix-xxvii. Amsterdam/Philadelphia: Benjamins.
Deane, Paul
2005
Multimodal spatial representations: On the semantic unity of over. In
From Perception or Meaning: Image Schemas in Cognitive Linguistics,
Beate Hampe (ed.), 23582. Berlin: Mouton de Gruyter.
Fauconnier, Gilles, and Mark Turner
2003
The Way We Think. New York: Basic Books.
Fillmore, Charles, and Beryl Atkins
2000
Describing polysemy: The case of crawl. In Polysemy: Theoretical and
Computational Approaches, Yael Ravin, and Claudia Leacock (eds.),
9110. Oxford: Oxford University Press.
78
John R. Taylor
Geeraerts, Dirk
1993
Vaguenesss puzzles, polysemys vagaries. Cognitive Linguistics 4:
223272.
Goldberg, Adele
2006
Constructions at Work: The Nature of Generalization in Grammar. Oxford: Oxford University Press.
Hallan, Naomi
2001
Paths to prepositions? A corpus-based study of the acquisition of a
lexico-grammatical category. In Frequency and the Emergence of Linguistic Structure, Joan Bybee, and Paul Hopper (eds.), 91120. Amsterdam/Philadelphia: Benjamins.
Jackendoff, Ray
2002
Foundations of Language. Oxford: Oxford University Press.
Lakoff, George
1987
Women, Fire, and Dangerous Things: What Categories Reveal about the
Mind. Chicago: University of Chicago Press.
Langacker, Ronald W.
1987
Foundations of Cognitive Grammar. Vol. I: Theoretical Prerequisites.
Stanford: Stanford University Press.
1990
Active zones. In Concept, Image, and Symbol: The Cognitive Basis of
Grammar, Ronald W. Langacker, 17788. Berlin: Mouton de Gruyter.
2006
On the continuous debate about discreteness. Cognitive Linguistics 17:
107151.
Moon, Rosamund
1998
Fixed Expressions and Idioms in English: A Corpus-based Approach.
Oxford: Oxford University Press.
Nunberg, Geoffrey, Ivan Sag, and Thomas Wasow
1994
Idioms. Language 70: 491538.
Pustejovsky, James
1991
The generative lexicon. Computational Linguistics 17: 409441.
Queller, Kurt
2001
A usage-based approach to modeling and teaching the phrasal lexicon.
In Applied Cognitive Linguistics II: Language Pedagogy, Martin Ptz,
Susanne Niemeyer, and Ren Dirven (eds.), 5583. Berlin/New York:
Mouton de Gruyter.
Rice, Sally
2003
Growth of a lexical network: Nine English prepositions in acquisition.
In Cognitive Approaches to Lexical Semantics, Hubert Cuyckens, Ren
Dirven, and John R Taylor (eds.), 243280. Berlin/New York: Mouton
de Gruyter.
Sandra, Dominiek, and Sally Rice
1995
Network analyses of prepositional meaning Mirroring whose mind
the linguists or the language users? Cognitive Linguistics 6: 89130.
Searle, John
1983
Intentionality: An Essay in the Philosophy of Mind. Cambridge: Cambridge University Press.
79
80
John R. Taylor
Wray, Alison
2002
Formulaic Language and the Lexicon. Cambridge: Cambridge University Press.
Wunderlich, Dietrich
1993
On German um: Semantic and conceptual aspects. Linguistics 31:
111133.
Zlatev, Jordan
2003
Polysemy or generality? Mu. In Cognitive Approaches to Lexical Semantics, Hubert Cuyckens, Ren Dirven, and John R Taylor (eds.),
447494. Berlin/New York: Mouton de Gruyter.
81
Abstract
This paper offers a short description and critical appraisal of four cognitive approaches to grammar, Langackers Cognitive Grammar, Goldbergs Construction Grammar, Crofts Radical Construction Grammar and Fauconnier and
Turners Blending Theory. It first points out that the term grammar is polysemous, having both a narrow/traditional/descriptive sense (grammar as syntax
plus morphology) and a broad/generative/cognitive sense (grammar as a theory
of language). Both interpretations are taken into account. In a narrow sense, the
present paper tries to evaluate how the four models define word classes and syntactic functions and how they handle specific constructions such as change constructions and noun phrases. In a broad sense, this contribution argues that Cognitive Grammar remains the most innovative and comprehensive cognitive
theory of grammar and that the other models can, to some extent, be regarded as
notational variants. They highlight different (though related) facets of the shared
conceptualization of language as a taxonomic and diffuse network, i.e. of language as a structured inventory of conventional linguistic units where much is a
matter of degree.
Keywords: grammar; construction; Cognitive Grammar; Construction Grammar;
Radical Construction Grammar; Blending Theory; network; syntax-lexicon continuum; grammatical function; word class; polysemy; usage-based model; change
construction; noun phrase; resultative construction; caused motion construction
1.
Introduction
82
Cristiano Broccias
does not subscribe to the existence of generative rules either. Rather, it contends that syntax and the lexicon form a continuum of constructions ranging from very specific elements (e.g. cat, kick the bucket) to increasingly
more general patterns (e.g. noun, transitive construction). General patterns
(or schemas) capture the commonality of various specific instantiations
(e.g. kick the bucket and stroke the cat are both instantiations of the transitive construction) and thus serve the same purpose as rules in generative
grammars, i.e. they account for our ability to combine linguistic elements
in a regular way. Despite the agreement on the non-derivational (or monostratal) nature of grammar since rules are not posited, constructions are
not taken to be derived by manipulating underlying forms cognitive linguists do not always view grammar identically, as will be shown below.
The present overview begins by presenting some of the core notions of
Cognitive Grammar (see Langacker 1987, 1990, 1991, 1999), which offers
the clearest definition of what is meant by grammar within a cognitive theory of language (see section 2). It should become apparent from the discussion that grammar can (and will) be used in two main senses: a narrow
sense, corresponding to the descriptive definition (i.e. syntax plus morphology), and a broad sense, as in generative and cognitive approaches. In
the latter sense, grammar refers generally to how language works and is
not restricted to syntax and morphology (so one can speak of a Cognitive
Grammar approach to English phonology, for example). The reader
should also be aware that a similar distinction applies to the notion of
construction, which will figure prominently in this paper. In a narrow (or
traditional) sense, a construction refers to the combination of syntactic
phrases but, in a broad (or cognitive) sense, construction is a synonym of
(any kind of) linguistic unit, as will be explained in section 2.
The title (and hence the topic) of the present paper refers mainly to
grammar in the narrow sense. That is, my primary concern will be problems which are part and parcel of traditional grammar, e.g. the definition
of word classes and grammatical functions, noun phrases and change
constructions (i.e. caused motion constructions and resultative constructions). Nevertheless, in order to deal with such problems in a principled
way, I will of course have to discuss cognitive approaches to grammar in
the broad sense, i.e. grammar as a theory of language. But I will do so only
insofar as they are relevant to the problems discussed here.
After having examined Cognitive Grammar, I will illustrate and discuss
in section 3 the basic tenets of Goldbergs Construction Grammar (see
Goldberg 1995). Section 4 will summarize the fundamental assumptions
behind Crofts Radical Construction Grammar (see Croft 2001). Section 5
83
will briefly review how grammar (in the traditional sense) is handled in
Fauconnier and Turners Blending Theory (see Fauconnier and Turner
2002). Section 6 will draw the main conclusions. In particular, I will point
out that, despite some (important) differences, a coherent picture emerges
from the various cognitive approaches examined. The importance of the
various theories lies in the fact that they tend to highlight different (although related) facets of the shared conceptualization of language as a
taxonomic and diffuse network.
2.
Cognitive Grammar
84
Cristiano Broccias
The Cognitive Grammar definition of grammar has important consequences. First, linguistic units can be of any length. Grammar includes
symbolic assemblies of any internal complexity rather than just atomic
units like cat. Complex constructions like the ditransitive construction
(e.g. He gave her a present) are part and parcel of grammar and are not
taken to be projected from a verbs argument structure (as in generative
grammar). Second, linguistic units can be of any degree of specificity.
Grammar includes both very specific patterns (e.g. I love you, which can
be considered an entrenched unit in the English language although its
meaning is transparent, unlike idioms such as kick the bucket, give up the
ghost, etc.) and general schemas that may subsume them (e.g. grammar
includes the transitive construction, of which I love you is an elaboration
or instantiation). From these two observations regarding complexity and
specificity, it follows that there are no clear boundaries between what are
traditionally referred to as the lexicon, morphology and syntax (this is
known as the syntax-lexicon continuum hypothesis). These labels refer to
symbolic assemblies differing from one another only in terms of structural complexity and schematicity (or specificity), as is illustrated in
Table 1 below (adapted from Croft 2001: 17). Indeed, any entrenched
symbolic assembly is regarded as a construction in Cognitive Grammar
(as well as in Radical Construction Grammar, see below) irrespective of
either its internal complexity or its specificity. More generally, Cognitive
Grammar constantly underlines that much in language is a matter of degree. Since grammar is an integral part of human cognition and since
human cognition rests, among other abilities, on categorization (which
operates by way of instantiation and extension relations rather than feature-listing), sharp boundaries should not be expected in grammar.
Table 1. The syntax-lexicon continuum.
Construction type
syntax
idiom
I love you,
black cat
morphology
noun-s
word class
word/lexicon
85
The Cognitive Grammar definition of grammar also implies that linguistic analyses should always be checked against two omnipresent fallacies in linguistic argumentation, namely the exclusionary fallacy and the
rule/list fallacy. The gist of the former is that one analysis [] for a linguistic phenomenon necessarily precludes another (Langacker 1987:
28). For example, if we take the noun computer and say that this word is
derived by rule (i.e. the suffix -er is added to the verb base compute), we
cannot account for the fact that computer means more than something
that computes. By contrast, if we regard it as an item listed in the lexicon
(on the assumption that the lexicon is an independent level of the linguistic system corresponding to an unstructured inventory of linguistic
forms), we fail to capture the productivity of the verb + -er derivational
pattern.
The rule/list fallacy is the assumption that particular statements (i.e.
lists) must be excised from the grammar of a language if general statements (i.e. rules) that subsume them can be established (Langacker 1987:
29). For example, the fact that most plural nouns can be derived from the
corresponding singular forms by adding -s does not necessarily imply that
specific plural forms like cats, dogs, etc. are not listed in the grammar. In
other words, the best grammar is not the shortest grammar according to
Cognitive Grammar since redundancies are an integral feature of the
Cognitive Grammar model. Both computer and cats are analysed as linguistic units in Cognitive Grammar since they are probably accessed
automatically or, at least, without much constructive effort. Further, they
are regarded as instantiations of (i.e. are categorized by) two other linguistics units, namely the verb + -er schema and the noun-s schema, respectively. This also means, incidentally, that what are traditionally called
morphemes (see also Table 1) are in fact schemas which involve at least
one more (phonologically unspecified) element, e.g. a verb in the case of
-er and a noun in the case of -s.
Although the symbolic links between the semantic pole and the phonological pole of a construction may often be regarded as arbitrary (e.g. it is
arbitrary that a doer is often signalled morphologically by way of the -er
morpheme), Cognitive Grammar does not neglect the issue of motivation
behind symbolic links. For example, the form of the prepositional dative
construction (e.g. He sent a letter to Susan) is not arbitrary but emphasizes the path traversed by the letter with Susan as a goal (Langacker
1987: 39), as is signalled by the motion preposition to. Conversely, the
double object construction (e.g. He sent Susan a letter) is said to emphasize the possessive relation between Susan and the letter by way of their
86
Cristiano Broccias
juxtaposition and linear order (Langacker 1987: 39). Cognitive Grammar claims that the meaning of these two constructions is not necessarily
identical because, among other things, they impose different construals
(i.e. views) on a common conceptual content. The prepositional dative
construction construes transfer of possession in motion terms while the
double object construction construes transfer of possession in terms of its
outcome, i.e. the establishment of a relation of possession (or, more generally, control) between the indirect object and the direct object.
2.1. Word classes and grammatical functions in Cognitive Grammar
The view of language as an integral part of human cognition rather than
as an independent module has also important repercussions for the analysis of traditional notions such as grammatical (or syntactic) functions
(also known as grammatical/syntactic relations or roles) and word classes
(also known as parts-of-speech or lexical categories). In particular, Cognitive Grammar, adopts (at least, in some cases) a universalist stance towards both grammatical functions and word classes (but not in the same
sense as in Radical Construction Grammar, as will be pointed out below).
In Cognitive Grammar, subject for example is regarded as a universal
notion ultimately based on the (perceptual) distinction between figure
and ground, viz. trajector and landmark, configurations in a scene. As is
well-known from Gestalt psychology (see Ungerer and Schmid 1996: Ch. 4
for an overview), certain objects function as primary foci of attention (e.g.
smaller and/or moving objects as opposed to larger, static objects). A sentence like The table is under the book (employed to convey the scene depicted by The book is on the table) is felt to be deviant because it reflects a
non-canonical perception of the intended spatial configuration. A smaller
object like a book is usually chosen as the primary focus of attention (or
trajector, in Cognitive Grammar terminology) whereas a larger object like
a table functions as a secondary focus of attention (or landmark, in Cognitive Grammar terminology).
Cognitive Grammar proposes that subject is a universal category in
that it can be defined schematically by characterizing it as a clause-level
trajector. Similarly, object can be defined schematically as a clause-level
secondary figure or landmark.
In addition to its schematic conception of grammatical relations, Cognitive Grammar also provides a schematic characterization of word
classes. Cognitive Grammar contends that parts-of-speech can not only
be defined prototypically (e.g. a noun prototypically refers to a physical
87
object) but can also be described in a manner which is valid for all its
other instantiations (i.e. a noun describes a thing, see below). In other
words, an invariant can be found for all major lexical categories. The
Cognitive Grammar analysis of (the semantic pole of) word classes is
summarized diagrammatically in Figure 1 (shaded boxes indicate concepts that are possible semantic poles for words; for example, the higher
category of entities, unlike that of things, does not correspond to any particular word class). We start with the very general category of entities,
which can be divided into things and relations. A thing, a technical term
in Cognitive Grammar, is a set of interconnected entities and is the semantic pole of the noun class. For example, the noun team (see Langacker
1987: 197) profiles a set of entities (represented in Figure 2a as the dashed
boxes connected to each other) rather than singling out any constitutive
member. The bold circle visually represents such a set in Figure 2a. Relations, by contrast, profile connections between entities (as well as the
entities themselves), as is shown for two arbitrary entities e1 and e2 in Figure 2b. Relations can be either processes (or temporal relations) or atemporal relations. In order to distinguish between the two, Cognitive Grammar contends that we can process complex scenes in two different ways.
The different facets of a scene (e.g. the constitutive parts of a kissing
event) are conceptualized either successively (as in a motion picture) or as
a single Gestalt (as in a photo with multiple exposures, where the various
parts are all made available simultaneously). The former mode of cognitive processing is called sequential scanning while the latter is referred to as
summary scanning. Processes are said to involve sequential scanning: the
different facets of a process are accessed sequentially, one after the other.
Processes, depicted schematically in Figure 2c, constitute the semantic
pole of verbs (e.g. kiss). The bold (temporal) line in Figure 2c represents
the fact that the relation in question has a positive temporal profile, i.e. is
scanned sequentially. By contrast, atemporal relations have a null temporal profile. This means that they are scanned summarily. Atemporal
relations come in two types: stative relations and complex atemporal
relations. Stative relations involve a single, stable configuration through
time and correspond to the semantic pole of adjectives (i.e. stative relations whose trajector is a noun and whose landmark is a region on a
scale), adverbs (i.e. stative relations whose trajector is a process and whose
landmark is a region along a scale), and static prepositions (such as in, as
opposed to into). Complex atemporal relations (e.g. into) are made up of
more than one configuration over time and such configurations or facets
are scanned in summary fashion.
88
Cristiano Broccias
Figure 2. Schematic representations of things (a), relations (b) and processes (c)
2.2. Discussion
One of the features setting Cognitive Grammar apart from other cognitive models is the contention that notions such as subject and object
can be defined schematically, i.e. independently of specific forms (or realizations) in a way that encompasses all their uses across all languages. Indeed, such a position may appear controversial (especially in the light of
89
typologically-oriented research, see the discussion of Radical Construction Grammar below). Cognitive Grammar claims for example that there
should be analyzed as a subject in a sentence like There are a lot of people
in the lift (i.e. a presentational there-construction). There, in the case at
hand, is categorized as a subject because it refers to an abstract setting
which is used as a first trajector, i.e. an initial reference point for accessing
the described situation, or a spotlight of prominence (see also Langackers
2005b analysis of it-subjects). Observe that, under this analysis, the trajector (there) refers to an entity larger than that evoked by the landmark
(a lot of people). That is, the schematic definition of subject does not
necessarily pick out smaller entities as trajectors (cf. also the setting-subject construction The garden is swarming with bees, where the trajector the
garden refers to a location and the smaller and moving entities, the bees,
are the clause-level landmark). Related to this is the question of whether
the analysis of presentational there constructions poses any problems for
a sentence such as Down the hill came a strange procession (an inversion
construction, see Chen 2003). Is down the hill also to be categorized as a
subject? Down the hill, like there, refers to a setting, although a concrete
one, and, like there, can be taken to be the starting point for accessing the
configuration depicted by the clause (see Langacker 2005b). Still, down
the hill is categorized as a landmark in Cognitive Grammar.
To be sure, the Cognitive Grammar analysis of the notion of subject
defies traditional analyses. Traditionally, subject identification is based
on morphological and/or syntactic criteria (e.g. verb-agreement and
word-order, respectively, in English). Consequently, distributional facts
point to both there and the postcopular noun phrase (e.g. a lot of people)
as having subject status. The morphological criterion (at least in Standard
English, cf. colloquial Theres a lot of people in the lift where a lot of people
does not agree with the copula) identifies a lot of people as a subject. The
syntactic criterion (cf. Are there a lot of people in the lift?) forces us to conclude that there is also a subject. In the case of the inversion construction,
by contrast, distributional facts would lead us to identify a strange procession as the subject.
At the very least, one can conclude that the Cognitive Grammar analysis is conceptual, not grammatical (in the traditional sense). That is, the
Cognitive Grammar analysis is not necessarily meant to pick out those
subjects identified by formal criteria. After all, distributional evidence
simply points to the fact that some elements have a special status compared to others. The deeper and more interesting question is why this is so.
Cognitive Grammar contends that such special (formal) behaviour is
90
Cristiano Broccias
3.
Construction Grammar
91
92
Cristiano Broccias
Figure 3a depicts the schema for the caused motion construction. Construction Grammar distinguishes between the semantics (Sem) and the
syntax (Syn) of a given construction. The exact details of the representational conventions used in Construction Grammar are not our concern
here. It will suffice to point out that the semantics of the caused motion
construction indicates that an argument (cause) causes another entity
(theme) to move towards a certain position (i.e. goal). The causal relation
and the ensuing movement, i.e. the overall caused motion predicate, are
abbreviated as cause-move. Cause is realized syntactically as a subject
(SUBJ), goal as an oblique (OBL, that is, an adjective phrase or a prepositional phrase), and theme as a direct object (OBJ). The construction
also specifies (by way of the middle tier) that specific instantiations of
the construction are obtained by merging the schematic caused motion
predicate (cause-move) with a specific predicate (pred), which corresponds syntactically to a verb (V), e.g. slice in our example. This is
shown in Figure 3b. Slice (as well as making reference to a sharp instrument as part of its semantic base) takes two arguments, a slicer argu-
93
ment (i.e. the cause argument), Pat, and a sliced argument (the theme
argument), the carrots. The merger between the caused motion construction and slice results in the specific construction Pat sliced the carrots into
the salad.
3.2. Networks
As in Cognitive Grammar, grammatical knowledge is represented in networks. Instead of categorizing relations, Construction Grammar makes
use of inheritance links (as in computer science) to relate constructions to
one another. Four major types of inheritance links are recognized: polysemy links, metaphorical extension links, subparts links and instance links
(see Goldberg 1995: 7481). Before briefly illustrating them, it should be
observed that inheritance links are treated as objects in Construction
Grammar (see Goldberg 1995: 7475). This differs from Cognitive Grammar (and Radical Construction Grammar), where the only elements
postulated in grammar are the semantic and phonological poles of a construction plus their symbolic links. Like other cognitive models, however,
Construction Grammar allows for multiple inheritance, i.e. a construction
can be categorized by two (or more) independently established constructions. For example, the intransitive motion construction (e.g. The car
screeched around the corner) inherits from both the intransitive construction (since the intransitive motion construction is quite trivially intransitive) and the motion construction (since motion is predicated of one of the
constructional elements, i.e. the subject, via an oblique phrase).
Polysemy links and metaphorical links roughly correspond to extension
relations in Cognitive Grammar. Construction Grammar claims for
example that the double object construction has a variety of senses which
can be viewed as extensions (via polysemy links) from the central sense
[a]gent successfully causes recipient to receive patient (see Goldberg
1995: Fig. 2.2). A metaphorical link connects the caused motion construction and the resultative construction (see (1d) and (1e) above) in that
the latter is said to have originated from the former via the metaphorical
construal of states (e.g. crazy in Chris drove Pat crazy) as locations. A subpart link is posited when a construction is a proper subpart of another
construction (Goldberg 1995: 78). The intransitive motion construction
is linked to the caused-motion construction via such a link since the
former is a proper subpart of the latter (the only missing element in the intransitive motion construction is the cause argument; for a different
view see Broccias 2003: Ch.5). Finally, instance links correspond to elab-
94
Cristiano Broccias
orative categorization in Cognitive Grammar. This obtains when a construction specifies another construction in more detail (e.g. Chris drove
Pat crazy instantiates the schematic resultative construction).
It should be pointed out that Construction Grammar, at least in its
1995 version, also makes use of constraints to motivate the (alleged) impossibility of some patterns. Among them is the constraint (see Broccias
2003: 213218 for a review of Goldbergs 1995 constraints) that states that
no intermediary time intervals are possible between the causing event and
the resultant event in a resultative construction (as in Chris shot Pat dead,
where death is said to necessarily occur immediately after the shooting
event). Even if we accepted the descriptive correctness of this and other
proposed constraints (but see Broccias 2003: Ch. 6), it remains to be established how they are represented in the grammar. One might claim that,
in a truly cognitive model of grammar, constraints cannot be independent
objects in the grammatical system. Constraints should simply amount to
paraphrases of schemas which subsume a large amount of specific instances (see e.g. Taylors 2002: Ch. 5 discussion of phonological constraints as phonological schemas).1 Broccias (2003) argues, for example,
that Goldbergs (1995) constraints for the caused motion construction
and the (transitive) resultative construction are just paraphrases of the
conceptual import of the schema to which both construction types can be
related, i.e. the billiard-ball schema.2
3.3. Discussion
There are various differences between Construction Grammar, on the one
hand, and other cognitive models like Cognitive Grammar and Radical
Construction Grammar, on the other. The previous subsection has already
pointed out the possibly problematic postulation of constraints or generalizations, which cannot be assumed as independent objects in Cognitive
Grammar, for example. Similarly, the fact that constructions are related to
one another does not rest in Cognitive Grammar on the postulation of
links as objects as is the case in Construction Grammar. Relatedness in
Cognitive Grammar is a necessary corollary of the language-independent
cognitive ability of categorization, which structures any type of information in terms of a standard (of comparison) and a target (of conceptualization), i.e. all information is stored in networks by default. There is
no need to postulate links as independent objects in Cognitive Grammar.
Construction Grammar also makes use of abstract predicates (e.g.
cause, move, see Figure 3 above) and roles (both semantic, e.g. theme, and
95
syntactic, e.g. subj) but their nature is not detailed any further. That is,
Construction Grammar, unlike Cognitive Grammar and Radical Construction Grammar, does not take a stance on their universality (or lack
thereof).
Even more importantly, although constructions (like linguistic symbols
in Cognitive Grammar) are defined as pairs of form and meaning, form is
equated with syntax in Construction Grammar. That is, syntax is taken to
symbolize (i.e. express) the semantic level, see Figure 3 above. Langacker
(2005a) objects to this analysis because it implies that syntax is treated as
a level independent of semantics (as in modular approaches). In Cognitive
Grammar what is (traditionally) called syntax simply amounts to a collection of (sufficiently) abstract schemas, i.e. assemblies of semantic and
phonological structure. A pattern like subj verb obj obl (instantiated for
example by Pat sliced the carrots into the salad) cannot be taken as pure
form in Cognitive Grammar. Subject and object, for example, refer
to, respectively, the trajector and landmark of the relation designated by
the verb and hence they profile concepts (although their phonological
pole is maximally non-specific). Similarly, verb stands for a pairing of
meaning and form. Its semantic pole designates a sequentially scanned relation and its phonological pole is non-specific (although we know,
among other things, that most verbs usually take a final alveolar plosive
segment when used to refer to the past, etc.). In sum, in Cognitive Grammar what symbolizes the semantic pole is the phonological pole, not syntax, because syntax in Cognitive Grammar is not pure form.
We can now turn to the issue of lexical (vs. constructional) polysemy,
which also sets Construction Grammar apart from other cognitive models. As was pointed out in section 3.1, Construction Grammar aims to
minimize lexical polysemy. Slice in (1) is analysed as having only one
meaning, roughly cut with a sharp instrument. The fact that slice can be
used with a caused motion meaning, see (1b), is attributed to the interaction between the verb and the caused motion construction. Construction
Grammar does not assume that this type of interaction results in the
verbs acquiring a new meaning (i.e. in the verb being polysemous). More
generally, such an analysis implies that Construction Grammar subscribes to a parsimonious view of grammar (i.e. the best grammar is the
shortest grammar). A new (e.g. caused motion meaning) for slice is not
postulated because this meaning can be retrieved from our knowledge
that slice can be used in a caused motion construction. Even if we accepted this position as correct (i.e. cognitively plausible), at closer inspection
there seems to be no clear theoretical advantage (in terms of parsimony)
96
Cristiano Broccias
97
4.
98
Cristiano Broccias
objects as direct objects. The problem here is that the choice of passivization (over, for example, obligatory presence vs. absence of a preposition
after the verb) as the defining criterion for direct object status is not theoretically motivated. Further, there is no a priori reason why we should
group different cases into the same category. In fact, a Radical Construction Grammar analysis would claim that the three sentences at hand contain three different types of objects whose properties overlap only partially: John immediately follows the verb and can become the subject of a
passive sentence; 80 kg cannot become the subject of a passive sentence
but, like John, occurs immediately after the verb; this house, like John, can
become the subject of a passive sentence but, unlike John and 80 kg, is separated from the verb by a preposition.
(2) a.
a.
b.
b.
c.
c.
The second reason why Radical Construction Grammar is radical is because constructions are taken to be the basic units of syntactic representation. This assumption is also shared by Cognitive Grammar and Construction Grammar (at least in its latest implementations) and was
illustrated in Table 1 above.
Thirdly, Radical Construction Grammar is radical because syntactic
relations are claimed not to exist. Croft argues that the term syntactic
role should be used instead. He points out that the term subject, for
example, can be used in two different ways (see e.g. Croft 2001: 24 and
Croft and Cruse 2004: 285286). Subject can be used to refer to the syntactic relation between an element of a construction and another element
(the verb), e.g. we say that Heather is the subject of sings in Heather sings.
Subject can also be used to refer to the role that an element has in a construction, i.e. the part-whole (or meronomic) relation of an element with
the whole construction. For example, we say that Heather has the role
subject in the intransitive construction Heather sings. The crucial point
here is that Croft uses the term syntactic relation in a traditional (formal)
sense, whereby a syntactic relation like subject is taken to be constructionindependent by definition cf. the Government and Binding Theory configurational definition of subject as the Specifier of an Inflectional Phrase
99
or [Spec, IP] and does not necessarily symbolize any semantic content.
Since the only entities which should be posited in grammar according to
Croft are constructions (as well as relations among them), it follows that
syntactic relations (as construction-independent objects) must be dispensed with. Accordingly, Radical Construction Grammar only uses the
term role, which refers to the relation between an element and the whole
construction rather than the relation between an element and another
element within the same construction.
Finally, constructions are said to be language-specific: no two constructions can be assumed to be identical across languages. This assumption is
also shared by Construction Grammar and Cognitive Grammar.
In sum, Radical Construction Grammar can be described as a usagebased model: grammar is maximalist and bottom-up, rather than minimalist and top-down (like formal approaches). It consists of taxonomic
hierarchies of constructions which include very specific constructions
(e.g. kick the bucket, kick the habit) as well as more general schemas abstracting away their commonalities (e.g. a transitive schema for kick, a
general transitive schema, etc.). A self-explanatory example of the kind of
taxonomy assumed in Radical Construction Grammar is offered in Figure 4 below (adapted from Croft 2001: Figure 1.15, p.56).
100
Cristiano Broccias
Although Radical Construction Grammar claims that grammatical categories are construction and language-specific, Croft (2001) shows that
syntactic roles can be compared across languages once semantic notions
(e.g. actor event, undergoer event, etc.) are appealed to. In a similar vein,
he argues that a universal characterization of parts of speech is possible.
(For reasons of space, I will only discuss parts-of-speech here.) Crofts
analysis is based on the interaction between discourse functions and semantic classes. Croft recognizes three discourse (or pragmatic/communicative) functions (see Croft 2001: 66), namely the propositional acts of
predication (e.g. This intelligent detective is young, where young is predicated of this intelligent detective), reference (e.g. this intelligent detective,
which picks out a certain referent) and modification (e.g. intelligent detective, where intelligent modifies our cognitive file for the detective in
question, i.e. it enriches the detectives identity by adding a property to the
features already stored in our mind concerning the detective). As these
examples illustrate, we tend to use certain constructions to code such discourse functions (the constructions in question are called propositional
act constructions). Croft hypothesizes that the typological prototypes of
the referring, attributive, and predicating constructions are the semantic
classes of objects, properties and actions, respectively, i.e. we prototypically refer to objects, modify cognitive files by attributing properties to
entities and predicate actions of entities. It should be observed that these
semantic classes are only a small subset of the semantic classes of words
found in human languages (and are defined in terms of four semantic
properties, namely relationality, stativity, transitoriness, gradability, see
Croft 2001: 87). Parts-of-speech in a given language can therefore be represented onto a bi-dimensional space (a semantic map) defined by the
dimensions discourse functions and semantic classes. Nouns, adjectives and verbs are viewed as the (prototypical) pairings of reference/object, modification/property and predication/action, respectively. This is
shown in Table 2 (adapted from Croft 2001: Table 2.3, p.88. Unmarked
means that no derivational morphemes are employed, e.g. cat is an unmarked noun while happiness is a marked, deadjectival noun).4
101
semantic class
discourse function
reference
modification
predication
objects
unmarked nouns
genitive,
adjectivalizations,
PPs on nouns
predicate nominals,
copulas
properties
deadjectival nouns
unmarked
adjectives
predicate adjectives,
copulas
actions
action nominals,
complements,
infinitives, gerunds
participles,
relative clauses
unmarked verbs
4.2. Discussion
Radical Construction Grammars assumptions are very similar to those
made in Cognitive Grammar. Indeed, they are radical especially if compared to Goldbergs Construction Grammar, rather than Cognitive
Grammar. Construction Grammar, unlike Cognitive Grammar, does not
seem to take a stand on the first three reasons mentioned in the previous
subsection. This observation may lead us to view Radical Construction
Grammar as a notational variant of Cognitive Grammar applied to typological research (since the main focus of Croft 2001 is typology). Nevertheless, Langacker (2005a) argues that there are some substantive differences between the two models.
The main difference between the two grammars (see also Broccias
2003) resides in the fact that grammatical categories are given a schematic
characterization in Cognitive Grammar whereas they are claimed to be
construction and language-specific in Radical Construction Grammar.
Radical Construction Grammar, at most, offers a prototype description
of parts of speech. By contrast, Cognitive Grammar claims that a schematic definition (hence, a definition valid for all cases) can be offered for
both parts of speech (see Figure 1) and (at least some) grammatical relations (i.e. in terms of trajector/landmark alignment, see section 2.2). As
was pointed out in section 2.3, the schematic characterization provided in
Cognitive Grammar is not usage-driven but is advanced as a matter of
conceptual necessity, i.e. it stems from our cognitive abilities. This
amounts to saying that Cognitive Grammar, unlike Radical Construction
Grammar, is, at the same time, a usage-based or bottom-up model and a
universalist or top-down model. Like Radical Construction Grammar,
102
Cristiano Broccias
103
for example an action verb like sleep (always) represented in our mind as a
relation, akin, one may speculate, to to be asleep?).
Alongside grammatical categories, a further difference between Radical
Construction Grammar and Cognitive Grammar involves the use of the
term syntactic representation, which seems to be employed in Crofts
theory quite similarly to Construction Grammar. Croft subscribes to the
Cognitive Grammar and Construction Grammar view of constructions as
(partially arbitrary) symbolic units (or pairings of form and meaning).
Rather controversially, however (see also Broccias 2003), he views form as
including syntactic, morphological and phonological properties. This, as
was pointed out in section 3.3 for Construction Grammar, is potentially
problematic because syntax in a cognitively oriented theory of language
does not exist as an independent grammatical level. (Remember that
Croft himself denies the existence of syntactic relations because they are
taken to be independent of meaning.) Nevertheless, since the formal pole
of constructions is said to include syntactic, morphological and phonological properties (and not only syntactic properties as in Construction
Grammar), one may speculate that Crofts use of the term syntax may
just be a convenient label to refer to word order patterns in a construction.
All in all, a fundamental contribution of Radical Construction Grammar has not only been to have shifted the focus of scholarly research from
lexemes to constructions (in the traditional sense) but, most of all, to have
underlined the need for a truly usage-based or bottom-up model in language description.
5.
Blending Theory
104
Cristiano Broccias
(semantically) simple cases like red ball and more exotic cases like land
yacht (used to refer to a large, luxurious car) we resort to identical complex mental operations.
Consider the noun-noun combination land yacht. The conceptual pole
of this noun phrase results from the blending of two input spaces, as is
shown in Figure 5 (adapted from Fauconnier and Turner 2002: Fig.17.1,
p. 357). One input involves water, a skipper, a course, a yacht and a tycoon; the other space includes land, a driver, a road, a car and an owner.
These elements from the two inputs are put into correspondence with one
another (so that water corresponds to land, skipper to driver, etc.) and
some of them (e.g. tycoon/owner and car/yacht) are projected into the
blend so as to create emergent structure: a land yacht names a new object,
which is neither a generic car nor a yacht. (Cases like computer mentioned
in section 2.1 above would be handled similarly. Computer involves the
blending of an input space referring to the activity of computation, symbolized by the verb compute, and an input space referring abstractly to an
entity performing an action, symbolized by -er. Crucially, the blended
noun, computer, has features which are not inherited by either input because a computer is more than something that computes. In other words,
the blend contains emergent structure.)
105
Let us now analyse the adjective-noun combination red ball. This noun
phrase is deceptively simple. It could mean a pen whose case is red, a pen
whose ink is red, a pen whose cap is red, a pen used to record the activities of a team dressed in red (see Fauconnier and Turner 2002: 355) and
so on. In other words, we integrate (as was the case with land yacht) two
input spaces, one containing the colour red and the other containing the
pen plus, if necessary, relevant frames (i.e. scenarios) in which the pen
plays some role (e.g. a frame describing the use of the pen to record the
activities of a team). The operation of conceptual integration takes place
by virtue of the correspondences established between the two inputs (e.g.
between the colour of the pen cap and the colour red) and by projecting
the pen and the colour red into the blend. The fact that certain interpretations seem to be simpler and more accessible than others is simply due
to the existence of strong defaults: [] simplicity of form, conventionality of vocabulary, and frequency of use [do not] indicate simplicity of
underlying conceptual organization (Fauconnier and Turner 2002:
365).
More generally, Fauconnier and Turner demonstrate that the elements
within any construction should be taken as prompts for the activation of
complex blending operations. One more example will suffice. Consider
copular constructions like The child is safe, The beach is safe, and The
shovel is safe. Blending Theory contends that it is not strictly speaking
correct to say that safe attributes a property to the subject referent (cf. the
Radical Construction Grammar characterization of predication in the
previous section). After all, the first example means that the child is not
likely to be harmed (i.e. the child is a potential victim), while the second
and third examples mean that the beach and the shovel are not likely to
cause harm (but observe that beach refers to a location while shovel refers
to an instrument). In Fauconnier and Turners analysis (see Fauconnier
and Turner 2002: 2526 and 354) safe is a prompt for evoking an abstract
frame of danger (i.e. an input space containing, among other things, a victim, a location and an instrument) which is blended with the current situation of the child playing on the beach (i.e. another input space) so as to
obtain a counterfactual blend (where the child is harmed while playing on
the beach). To put it differently, the counterfactual blend is disanalogous
to the current situation since the child is not harmed in the real world.
The counterfactual blend is blended, in its turn, with the current situation
into a new blend where safe signals the compression of the relation of disanalogy between the counterfactual blend and the current situation: the
child is not harmed while playing on the beach.
106
Cristiano Broccias
107
He sneezed the napkin off the table. The former sentence does not mean literally that I performed any action on the pan (I could have forgotten
about the pan and let the water in it evaporate). However, in the blend we
108
Cristiano Broccias
construe the real world event sequence in terms of the subjects acting on
the pan itself. It is not simply a matter of the constructions imposing a
causal interpretation on the specific verb being used (as Goldberg claims)
or of categorizing the event referred to by the verb as a force-dynamic
event (as Langacker claims, see also Broccias 2003). We also construe the
whole unintegrated event sequence as a very compact event, where results
are construed as obtaining immediately because some force-dynamic action is being carried out on a manipulable entity.
There is, however, a potential problem with Fauconnier and Turners
approach (see Broccias 2003: Ch.5 and Broccias 2004). It is not clear why
the parts in the event sequence represented as input 2 in Figure 6 are not
treated individually as input spaces which are blended into the force-dynamic scenario represented as an input in Figure 6. In fact, this alternative analysis would be consonant with the Blending Theory analysis of,
for example, noun phrases such as red ball, where red and ball are viewed
as belonging to two different input spaces. By the same token, the verbal
event and the change event (i.e. the change of state or place of the affected
entity) could be treated as two different inputs. The relation of causation
existing between the two is compressed into a force-dynamic interaction
in the blend (the reader is referred to Broccias 2003 for an analysis along
these lines).
Be that as it may, Blending Theory is a powerful tool for the detailed description of the conceptual operations underlying the construction and
interpretations of both simple and complex expressions. Nevertheless, a
more detailed discussion of the limits and implementation of Blending
Theory should be the focus of future research (see e.g. Ruiz de Mendoza
Ibezs critique of the idea of emergent structure: Ruiz de Mendoza
Ibez 1998, 2002, Ruiz de Mendoza Ibez and Pea Cervel 2005).
6.
Conclusion
Although the present overview could not obviously do justice to its complex articulation, Langackers theory is by far the most comprehensive
theory of grammar available in the cognitive linguistic camp. To some extent, other cognitive models such as Goldbergs Construction Grammar,
Radical Construction Grammar and Blending Theory (as applied to the
investigation of linguistic structures) can be seen as notational variants
of Cognitive Grammar. The similarities are numerous. They all share,
among other things, the assumption that language must not be studied in-
109
dependently of general cognitive abilities. In fact, general cognitive abilities are taken to shape language. They also agree on the fact that language does not merely consist of a syntax plus a lexicon. Rather, language
is viewed as a taxonomic hierarchy, where sharp distinctions should not
be assumed; language is a diffuse network since much in language is a
matter of degree. This means that linguistic units (or constructions) can
be of any length and of any degree of complexity. Further, cognitive theories subscribe to the view that constructions are often ready-made pairings of form and meaning. We do not need to generate all their various
parts every time we want to use them but they may be available to us automatically, as gestalts (i.e. we are often unaware of the contribution of their
constitutive parts).
Still, important differences exist among the various theories. I have
tried to point out that the nature of word classes and grammatical relations is still a matter of controversy. Similarly, the relation between constructional meaning and lexical meaning is still the topic of heated debate.
Nevertheless, the theories briefly illustrated in this paper alongside Cognitive Grammar serve a very important purpose if they are all given due
consideration (especially) when specific phenomena are studied. They
highlight certain facets of grammar (in the Langackerian sense) that may
not have been worked out fully in the Cognitive Grammar model. This is
most evident in the case of change constructions.
Blending Theory reminds us that when we examine complex expressions we should distinguish between two conceptual levels, the unblended (or unintegrated) level and the blended (or integrated) level. The
latter can convey new meaning (with respect to the input spaces, cf. Langackers scaffolding metaphor mentioned in note 7), as is indeed the case
with change constructions. An unintegrated, complex sequence of events
is conceptualized as (or categorized by) a force-dynamic event. Thus, we
construe the unintegrated causal sequence as implying instantaneous causation (cf. the discussion of I boiled the pan dry above).8
Goldbergs analysis brings two important points to the fore. We should
investigate the relation between the meaning of a construction and the
meanings of its parts, e.g. the verbs meaning in a change construction. In
this respect, Langacker urges us to be cautious. The overlap between constructional and verbal meaning may be a matter of degree and, ultimately,
we may add, the issue should be the object of psycholinguistic investigation. To be sure, in some cases, even if a verb is entrenched in a specific
construction, this does not (at least intuitively) result in the verbs acquiring a new sense.9 For example, if we consider the two constructions I
110
Cristiano Broccias
laughed him out of patience and I laughed him into patience, we may find
it intuitively odd to claim that laugh should be categorized as a verb of
removal in the former case (since the object referent, metaphorically
speaking, moved out of patience) and as a verb of creation in the latter
case (since the object referent, metaphorically speaking, moved into a region of patience). Rather, we categorize laugh as a force-dynamic process
by virtue of our knowledge that laughing can affect people surrounding
us.
Goldbergs other important point is that constructions can be polysemous. This is indeed the case with change constructions. Consider the
examples in (3).
(3) a.
b.
c.
d.
In (3a), kiss receives a force-construal. It is categorized as a force (because we know that kissing engenders some emotional response) causing
anxiety to be removed from someone. Hence, (3a) is a prototypical
example of a change construction.10 By contrast, (3b) shares its form with
prototypical change constructions (cf. He was booed off the podium,
which may have a causal interpretation) but only has a temporal interpretation, as the prepositional phrase at the end makes clear (see Broccias
2005b). The team would have abandoned the pitch even if the fans had
not booed at them because it was half-time. The fans booing simply unfolded as the team left the pitch. Hence, (3b) shares with prototypical
change constructions only unity of time, not (immediate) causation. (Remember that, at the blended level, change constructions construe the
causing event and the caused event as obtaining one immediately after
the other even if this is not the case in the real world.) (3c) can be analysed as a change construction (i.e. the river became solid because it
froze) but also shows that certain change phrases (e.g. solid) may be better
analysed as intensifiers. After all, if speakers understand freeze as to become solid, to say that the river became solid because it froze (i.e. became solid) is a tautology. Hence, solid may be interpreted as meaning
something akin to completely, i.e. it intensifies the event denoted by
the verb by specifying that the freezing affected all parts of the river. (3d)
also illustrates redundancy of the change phrase (i.e. up) with respect to
the verbal meaning. After all, lift already means to move up. But up
111
Notes
1. Taylor (2002) stresses that phonology is autonomous, i.e. it does not necessarily symbolize the semantic pole and it obeys constraints of its own. Despite
this, phonological constraints are analysed as phonological schemas. Their
nature, therefore, is in no way different from that of constraints pertaining to
symbolic units: both are schemas abstracted away from specific instances.
2. The billiard-ball model captures our conceptualization of events in terms of
112
3.
4.
5.
6.
7.
8.
9.
10.
Cristiano Broccias
energetic interactions between two entities resulting in some change (of either
state or location) of the affected entity, see Langacker (1991: 13).
The idea that the degree of overlap between verbal meaning and constructional meaning is a matter of degree is expanded on in Langacker (2005b). In
this paper, he claims that an element which regularly fills the verb position in
a construction tends to develop a meaning congruent with that position.
Table 2 is not the complete map of English part of speech constructions (see
Croft 2001: Fig.2.3, p.99). But it will suffice for our present purposes.
Such a view derives from the Aristotelian analysis of copular sentences like
Socrates is mortal as involving the ascription of a quality (e.g. that of being
mortal) to the subject referent. This analysis is then extended to non-copular
sentences (e.g. The children went to school). But observe (see Table 2) that,
paradoxically (with respect to the history of the notion of predication), noncopular sentences are not treated as the prototypical constructions coding the
propositional act of predication in Crofts theory.
My understanding of Crofts analysis is that a clear-cut distinction between
the three types of discourse functions and the three types of semantic classes is
possible. Indeed, the distinction between predication and modification may
seem intuitively straightforward. Consider, however, a sentence like John lives
poorly: what is the function of poorly in this sentence? Is it predicating or modifying?
The view that the overall meaning of a construction is obtained by summing
arithmetically the meanings of its constitutive parts is known as compositionality and is expressed by the building-block metaphor in Langackers
analysis. He dubs the view that constitutive parts are just prompts the scaffolding metaphor.
The reader should observe that the present formulation is meant to be noncommittal about the issue of whether the integrated sequence should be
viewed as input 1 or just as a blend.
I am not suggesting that this is Langackers position (cf. especially Langacker
2005b). Rather, I am making a more general point concerning the usefulness
of Goldbergs view that a construction can impose a meaning (or construal)
on a verb used in it (without the verbs acquiring this meaning).
Incidentally, (3a) also shows that the subcategorized object of the verb (i.e.
him, cf. Ske kissed him) is not always inherited as an object in the construction
(although it shows up in the construction as the object of the complex preposition out of). The constructional object anxiety is not a possible object for
the verb in isolation (cf. *She kissed the anxiety). This type of change construction is called an asymmetric construction by Broccias (2003).
113
References
Broccias, Cristiano
2003
The English Change Network. Forcing Changes into Schemas. Berlin/
New York: Mouton de Gruyter.
2004
Review of Gilles Fauconnier and Mark Turner. 2002. Cognitive Linguistics 15(4): 575588.
2005a Review of Gnter Radden and Klaus-Uwe Panther (eds.) 2004. Studies
in Linguistic Motivation. Berlin/New York: Mouton de Gruyter. In:
LINGUIST List: Vol-16872. Tue Mar 22 2005.
2005b Non-causal change constructions. In Modelling Thought and Constructing Meaning: Cognitive Models in Interaction, Annalisa Baicchi, Cristiano Broccias, Andrea Sans (eds.), 7688. Milano: FrancoAngeli.
Broccias, Cristiano, and Willem Hollmann
2005
Do we need summary and sequential scanning in (Cognitive) grammar? Ms., Universit di Genova/Lancaster University.
Chen, Rong
2003
Inversion. A Ground-before-Figure Construction. Berlin/New York:
Mouton de Gruyter.
Croft, William
2001
Radical Construction Grammar. Syntactic Theory in Typological Perspective. Oxford: Oxford University Press.
Croft, William, and D. Alan Cruse
2004
Cognitive Linguistics. Cambridge: Cambridge University Press.
Duffley, Patrick J.
2005
The English gerund-participle in Cognitive Grammar. Paper presented
at the 32rd LACUS Forum, Dartmouth College, Hanover.
Evans, Vyvyan, and Andrea Tyler
2004
Spatial experience, lexical structure and motivation: the case of in. In
Studies in Linguistic Motivation, Gnter Radden, and Klaus-Uwe
Panther (eds.), 157192. Berlin/New York: Mouton de Gruyter
Fauconnier, Gilles, and Mark Turner
1996
Blending as a central process of grammar. In Conceptual Structure, Discourse and Language, Adele Goldberg (ed.), 113130. Stanford: Center
for the Study of Language and Information.
2002
The Way We Think. Conceptual Blending and the Minds Hidden Complexities. New York: Basic Books.
Fillmore, Charles
1988
The mechanisms of Construction Grammar. Berkeley Linguistics Society 14: 3555.
Fillmore, Charles, and Paul Kay
1996
Construction Grammar. Ms., University of California at Berkeley, Department of Linguistics.
Goldberg, Adele
1995
Constructions. A Construction Grammar Approach to Argument Structure. Chicago: University of Chicago Press.
114
Cristiano Broccias
115
116
Cristiano Broccias
117
118
119
Abstract
Cognitive Linguistics presently suffers from three dogmas. First, we all too often
present embodiment as if it were an eliminativist project, requiring that linguists
translate their findings into the language of neurons and brain structures and
thereby diminishing the role of socio-cultural embodiment. Second, we consider
embodiment as if it were something static or fixed, when in fact it is temporally dynamic in ontogenetic, historical and phylogenetic terms. Third, we often confuse
embodiment with debates over consciousness. While it is perhaps true that many
of Cognitive Linguistics early findings were the result of phenomenological conscious introspection, the advent of explicitly cross-disciplinary empirical and experimental methods has begun to reshape the field. After tracing the history of the
embodiment hypothesis from its origins in conceptual metaphors and related evidence, I propose a broad-based theoretic framework for integrating cross-disciplinary research. The challenges posed by disparate methodologies prompt a reconsideration of how the evidence for embodiment is converging. I conclude
with reflections on what Cognitive Linguistics must do to re-envision itself as one
of the cognitive sciences, including identifying some key areas in which we do not
yet have enough explicitly cross-methodological projects.
Keywords: embodiment; cognitive science; image schemas; eliminativism; development; consciousness
1.
Introduction
No one would dispute that as academics we run the risk of becoming lost
in our thoughts. One of the merits of a proper philosophical education is
that we acquire a keen sense of when theorists become so engrossed with
their own theories that they lose touch with the chaos of the real world. Or
perhaps it is more accurate to note that such theorists do not lose touch
with the real world in its entirety, but spend their days exploring the nooks
and crannies of a metaphoric monastery built upon the craggy point of
their original insight into the real world. Thus Spinoza the lens-grinder
constructed a rationalist philosophy replete with optical metaphors for
120
Tim Rohrer
121
primitives onto language. The obvious implication that many more socioculturally-minded cognitive linguists draw is that theories of language will
be eliminated in favor of theories of how brain structure determines language. Just as the symbolic logicians sought to replace Shakespeares love
sonnets with logical formulae, so do some cognitive linguists run the risk
of advocating translating polysemy, syntax and grammatical construals
into talk of neural models, neural circuits, axonal firings and parietal-hippocampal networks.
Cognitive linguists are right to question this dogma on both philosophical and substantive grounds. However, oftentimes the reaction to
this dogma is a counterproductive iconclasm, in which no shrift is given to
evidence of how the brain shapes language. It is equally unpardonable to
commit the opposite sin and eliminate our neurophysiological embodiment in favor of our sociocultural embodiment, though I know of no one
proclaiming that cognitive linguistics isnt cognitive linguistics if it ignores relevant facts about the sociocultural context. The right antidote
to such extreme formulations is to realize that even brain structures are
not isolated, static entities, but are embedded within a nervous system and
body that interacts with both physical and sociocultural space. As the
American pragmatist philosopher Dewey famously asserted, to see the
organism in nature, the nervous system in the organism, the brain in the
nervous system, the cortex in the brain is the answer to the problems
which haunt philosophy (Dewey 1925: 198); or, and more appropriately
for the present conversation, haunt cognitive linguistics. In studying
language and conceptual structure with the new tools of cognitive neuroscience, we do not necessarily have to engage in an eliminativist reductionism of language to the biophysical. From the perspective of American
pragmatism, both the brain and the sociocultural situation have their
roles in the interactions which shape language.
Embodiment as temporally static. The second false dogma dovetails neatly
with the first. For example, if we aim to offer some insights to linguistics
from the study of brain structures, it would help if we understood how
brain structures develop and change across time. A brain is fundamentally an organic entity; it grows, matures, changes, suffers injury and recovers, adapts in response to a changing environment; it evolves over generations. It is altogether too facile to take findings about brain structure in
right-hand dominant adults and generalize them inappropriately. If we
fail to attend to the limitations of the methods and the scope of current
brain research, we risk reifying brain structure into a static causal force
122
Tim Rohrer
that determines language, rather shaping and motivating it. Brain structure is dynamic; and while it is certainly not infinitely malleable there are
important differences between the adult brain and the developing brain,
between the adult brain and the aging brain, between the injured brain
and the uninjured brain, between brains of left- and right-hand dominant
individuals, and so on and so forth. This organic dynamicity has real impact on cognition and language.
When we see the brain and organism in context, we realize that embodiment is not static. We are fundamentally live creatures who are embodiedin-time; unlike Athena, we do not spring full-grown from the head of
Zeus. As newborns we know no language and have only the barest rudiments of what will eventually become the preconceptual basis for adult
language. If in Cognitive Linguistics we hold that semantic structure is
largely shaped by preconceptual structures such as image schemas, then
we must inquire as to how these image schemas are formed on a developmental timeline and whether they might differ cross-linguistically, however slight a difference that might be. For example, Bowerman and Choi
(2003) have shown that while nine-month old Korean- and English-speaking infants can make the same spatial discriminations, at eighteen months
their acquisition of language has solidified their spatial categories enough
so that they are no longer able to make the discriminations which their
language does not. To me this suggests that the developmental course of
image schemas is such that their neural instantiations in the brain
hardens both with age and exposure to the dominant language of the sociocultural environment. This hypothesis about the neural structure
underlying image schemas is further warranted given the evident other
differences between early and late language learning experiences in humans, and given what we know from animal research into how different
spatial schemas cause different patterns of growth in juvenile versus adult
brains.1 Similarly, languages change across time; every individual speaker
is situated within a specific sociocultural and historical context that influences how their metaphors are shaped (Geeraerts and Grondelaers
1995). Thus, there are multiple temporal forces at work, from phylogenetic change to historical language change to learning and other ontogenetic changes.
Embodiment as consciousness (or as unconscious). Having skirted the
shoals of an eliminativist Scylla, we must also avoid being engulfed by the
Charybdis of consciousness. A standing temptation of academics is to assume that because we can describe and study some mental process, we are
123
(or can be) ordinarily conscious of it. This is in fact not the case. For
example, consider how we process sound waves into phonemes. We do not
consciously attend to the peaks and valleys of the sound waves at different
frequencies, but the fact that we do parse sound waves into phonemes
implies that our brains do something very much like that in the auditory
neural maps for pitch. However, conscious introspection is almost entirely
useless when it comes to determining the sound shape of a phoneme at
different frequencies; instead, scientists are forced to rely on sophisticated
electronic measuring equipment. Such processes are certainly beneath the
level of ordinary conscious awareness, and inaccessible to consciousness
without the tools of science.
Nonetheless, Zlatev has recently argued that Lakoff and Johnsons notion of an embodied Cognitive Unconscious is seriously mistaken and
conceptually incoherent (Zlatev in press). He rejects the existence of
unconscious mental processes entirely, arguing that every mental process
must at least be in principle accessible to conscious introspection. Following Searle (1992), he asserts that if we cannot be aware of a mental process
then it is neurophysiological and not mental. However, note that this argument is merely definitional it simply equates introspective consciousness with all mental processes, a philosophical move which is both trivial
and unproductive.
There is more merit to other parts of Zlatevs argument that mental
processes are in principle accessible to consciousness, such as those concerning learning and development. To return to the previous example, we
are not ordinarily aware of the sound shape that distinguishes one phoneme from another. However when as adults we grapple with hearing a
new phoneme while learning a new language, we do become much more
conscious of subtle differences in the shape of its sound. We might also become aware that it is easier to learn it if we also try to speak it and associate its sound with the motor movements which produce it. Eventually we
will automate this mental process well enough that we can listen and converse in the new language without having to attend to it. Moreover, at
some point, as infants we likely had to consciously attend to learning
each and every phoneme we now have automated though the infants
consciousness indubitably has a very different quality to it than how we
would typically construe as consciousness in the case of the adult cognition; i.e., as a discursive reflective introspection.2
While the philosophical arguments concerning consciousness are notoriously interminable, Zlatev is right that embodiment theory must explicate the relationships between at least some of its major theoretical con-
124
Tim Rohrer
2.
125
Lakoff and Johnson (1980: 117119) identified three sources for these
more concrete concepts. They argued these more concrete concepts constitute the natural kinds of experience and are comprised of experiential
gestalts more basic than other concepts because they are the natural
products of our bodies, our interactions with the physical environment,
126
Tim Rohrer
and our interactions with other people in our culture. Reserving judgment
for future research, they also indicated that while some of these natural
kinds of experience might be universal, others might very well vary from
culture to culture. They explicitly pointed out that they were using the
terms nature and natural in a sense which at least encompassed the
possibility of cultural variation, and not in the sense of a rigid nature
culture distinction. Lakoff and Johnson concluded this section by arguing that these more concrete concepts can be used in the metaphorical
definition of more complex concepts. In short, they argued that these
three natural kinds of experience experience of the body, of the physical
environment, and of the culture are what constitute the basic source domains upon which metaphors draw.
2.2 Elaborations and extensions of the embodiment hypothesis
Over the ensuing twenty years, the notions of experientialism, embodiment, and a directionality to conceptual metaphor received much scrutiny, generated much controversy, and consequently received much elaboration. More systematic surveys undertaken during the mid-1980s at
Berkeley and elsewhere showed that bodily source domains were prevalent not only for the semantics of English, but also for languages as distant from it as Japanese and Mixtec. However, it is equally important to
note that the languages did vary cross-culturally as to which particular
bodily source domains were used to understand a given target domain,
and with respect to how these patterns were represented linguistically.
With respect to historical semantic change, Sweetser (1990) has argued
that the direction of such change is motivated by the embodiment hypothesis. For instance, she documented a directionality within Indo-European
languages for metaphors such as knowing is seeing, arguing that the
terms which came to be the ordinary ones for abstractions such as knowing were at an earlier time restricted to embodied perceptual capabilities,
such as seeing, grasping, hearing, smelling, tasting, and feeling. In a now
standard example, she traces the transition of the Indo-European root
*weid see through the Greek eidon to see and, in its perfective form oida
sight, know, to the English terms idea, wit, and witness, which retain
none of their visual sensibility to most native English speakers (Sweetser
1990: 2348). By observing how a wide range of embodied perceptual
terms systematically lose their perceptual connotations as they acquire
their intellectual meaning, she proposed that there exists a large-scale
temporal constraint on the directionality of semantic change.
127
In the preface to The Body in the Mind, Johnson (1987: xiixiii) presented six converging bodies of evidence for the embodiment hypothesis
understood as a directional constraint on meaning. This list included not
only cross-cultural research on metaphor and historical semantic change
but also work on prototypes in categorization, the framing of concepts,
polysemy, and inferential patterns in metaphor. Near the same time, other
research in Cognitive Linguistics (such as Langackers (1987) Cognitive
Grammar) contributed to an increasing focus on the role of the body in
shaping linguistic and conceptual structure generally, and not just within
a thread of semantic theory. In work that also appeared that same year,
Lakoff (1987) characterized the experientialism (or experiential realism)
at the core of the embodiment hypothesis as including
everything that goes to make up the actual or potential experiences of either individual organisms or communities of organisms not merely perception,
motor movement, etc., but especially the internal genetically acquired makeup
of the organism and the nature of its interactions in both its physical and social
environments. (Lakoff 1987: xv)
128
Tim Rohrer
129
130
Tim Rohrer
131
132
Tim Rohrer
3.
133
134
Tim Rohrer
gation and define the theoretical constructs that they develop, and these
are a product of the physical scale at which the measurement is taking
place. The final two columns acknowledge this by specifying some of the
relevant theoretic constructs and the various methodologies operative at
each level of investigation.
This framework can be used to situate the wide methodological array of
studies on various topics of interest to cognitive linguists, such as metaphor, mental imagery, categorization, frames of reference, emotions, and
so on. This type of theoretical framework is now fairly common within
much of cognitive science, but Cognitive Linguistics has been slow to give
explicit attention to the problem of how we are to theoretically situate and
reconcile these different levels of investigation (see Table 1 below).
I have included just a single level of cultural and communicative analysis, but by no means should this be taken as indicative of its importance
relative to the other levels. Of course, one could argue for a multiplicity of
levels embedded within this one, though they might not be clearly differentiated from one another in terms of physical scale. In choosing to include a general level situated at a meter and up on the physical size axis I
mean to emphasize only that human beings should be considered not
simply in terms of physiological size, but also in terms of the standard
scale of their interactional distance in speaking and interacting with one
another. At this level of the framework the physiological structures column reads Multiple Central Nervous Systems, but that awkward term
is intentionally inadequate so as to emphasize that the physiology is less
relevant here what primarily matters on this level are the social and cultural interactions between human beings. Investigations at the cultural
level are occasionally given short shrift by some versions of embodied
cognitive science, but generally this has been and should remain a strong
thrust of future research in Cognitive Linguistics.
Note also that difficult phenomena such as cultural and linguistic
norms, or individual consciousness and awareness, are situated at the
physical scale at which they are measured and observed, rather than attempting to place them on (or reduce them to) a lower level of investigation. Nonetheless, it is certainly possible and sometimes useful to ask,
for example, how long the neural auditory processing of a phonemic slip
takes before impacting conscious speech, or how the linguistic norm of
forming the English past tense might be performed in a neurocomputational model. However, research in embodied cognitive science should not
seek to reduce such phenomena to another level but should instead bridge
across these levels in important ways for example, the linguistic corpora
135
Verbal report,
observational
neurology, and
psychiatry, cognitive and developmental studies
examining reaction time (RT)
Conceptual
metaphor mappings, primary
metaphor, conceptual blends,
disanalogy,
image schemas,
topological
maps
Lesion analysis,
neurological dissociations, neuroimaging with
fMRI and PET,
ERP methods,
neurocomputational simulations
Image schemas,
primary metaphor, topographic maps,
convergence
zones
Orientation-tuning cells; ocular
dominance columns
Electrocellular
recording, anatomical dyes,
neurocomputational simulations
Electrocellular
recording, anatomical dyes,
neurocomputational simulations
NeurotransNeuro-pharmamitter, synapse, cology, neurion channels
ochemistry, neurophysics
136
Tim Rohrer
used to train the neurocomputational model should be based on naturalistic recordings of an actual childs utterances rather than text harvested
from internet newsgroups, and so on.
While the table depicting the theoretic framework is designed to give an
overview of the relationship between body, brain and culture, this representation is not as illustrative for issues pertaining to evolutionary, developmental and historical time scales, which may be considered at any of
these levels. However, this failing is more a limitation of the imagery of a
two-dimensional table than of the theoretic framework itself. If we were to
add another axis for time perpendicular to the surface plane of the chart,
we could imagine this framework as a rectangular solid (Figure 1). Embodiment is not temporally static, and when we strip away the veneer a
finely detailed veneer, to be sure of synchronic studies, we see that those
results are the result of processes of learning and development, and of history and evolution.
4.
137
There can be no doubt that most of the ordinary research tools of Cognitive Linguistics are at the communicative and cultural level of analysis, the
level whose undeniable phenomenological components perhaps most
lends itself to two standard critiques. First, there is a methodological
question about how one is to go about doing the analysis, for example, as
to how one knows when a conceptual metaphor is correctly mapped; second, there are legitimate questions as to how empirical such methods are.
While the objection that such analyses are not empirical at all can be at
least partially rebutted by the argument that the sentences or phrases were
empirically gathered from naturalistic contexts, even beginning to answer
the first objection is not so easy. The problem is that an appeal as to what
counts as a grammatically well-formed sentence, a plausible construal in
cognitive grammar, or a correct conceptual metaphor mapping seems to
be little more than an introspective gut-check by the practitioners themselves. Of course, one might respond that there is a body of previous work
to which a new analysis must either conform, make fine small distinctions
from, or challenge and then reorganize satisfactorily. For instance, conceptual metaphors, whether gathered ad hoc (as in Lakoff and Johnson
1980) or systematically gathered from a defined corpus (as in Rohrer
1995), are supposed to yield relatively stable conceptual mappings. However, both in my experience teaching conceptual metaphor analysis in
seminars to both advanced students and scholars from other disciplines,
and in a careful reading of the conceptual metaphor literature, there appears a wide variety as to how the source and target domains of the same
of a conceptual metaphor are named, or as to the exact counterparts or
order to the conceptual mappings of the metaphor. In its strongest form,
this objection is that such methods are not just non-algebraic, but almost
entirely ad-hoc.6
Within Cognitive Linguistics I see my colleagues make three types of responses to these metatheoretical critiques. The first is to ignore such criticisms and continue to do the work, using the methodologies as they are,
perhaps continually refining them according to the community aesthetic7
of what counts as a methodological innovation, or perhaps improving
them incrementally importing methods from the related social sciences
such as inter-coder reliability ratings for corpus analysis, and so forth. A
second approach is to confront these questions using a philosophical defense of the importance of sociocultural factors or the irreducibility of
consciousness to the methods of Cognitive Linguistics. For example, Zla-
138
Tim Rohrer
tevs arguments (in press) have the following philosophical shape: first,
language is intrinsically a social behavior for communicating; second, language is intentional, in the sense that every word is a communicative act
consciously intended by the speaker to share or convey some experience
with (or to) the hearer; third, language must be a triadic communicative
system, which is to say that the meanings of its signs are consciously conventional across speakers in a population of n >= 3 speakers; ergo, all
judgments (by either laypeople or linguists) about whether a sentence is
grammatical, or whether the grammar cognitively fits the situation, or
whether the a term has been used in an appropriate semantic context are
intrinsically a conscious act. Other theorists have staked out similar
philosophical positions, arguing that it is primarily sociocultural factors
of language acquisition that shape spatial relation schemas (for example,
Bowerman and Choi 2003; Sinha and Jensen de Lopez 2000), or arguing
that the historical context determines metaphor mappings (Geerarts and
Grondelaers 1995).
The third response, which I would take to be the standard response of
Cognitive Linguistics, is to cite and discuss converging bodies of evidence for embodiment from the other cognitive sciences. It is easy to see
how such evidence could be converging under an eliminativist rubric,
because Cognitive Linguistics and the other cognitive sciences would then
converge on an underlying neurophysiological reality. But methodological pluralism implies that these methodologies are independent and
that there is no guarantee that their results can be integrated. If we are not
converging on the same ontologically real material, as eliminative materialism would have it, then what are we converging upon? The answer,
I think, is a guiding idea or vision: embodiment. Our convergence is more
on a set of theoretical assumptions and methodological biases than any
findings.
Recent years have witnessed numerous attempts (Hutchins 1995; Lakoff and Johnson 1999; Gibbs 1994; Clark 1998; Fauconnier and Turner
2002) to retell the history of cognitive science so that that an embodied vision of cognitive science can both incorporate past labors and findings
and shape new research questions that might bridge the past with the future. While attempts at such grand unifications are laudable, the persuasiveness of the case for convergence likely depends on whether and how
accurately the authors have described the evidence gathered outside their
own traditions, and whether the conceptual leaps across these traditions
are kept suitably small. If not, we are still faced with the possibility that
the convergence of the evidence will be construed as a myth and the evi-
139
dence will turn out to be very heterogeneous (Zlatev in press: 19), creating unbridgeable conceptual chasms between rival theoretic monasteries.
Yet there is still more that cognitive linguists can do. It is now, more
than ever, particularly important for cognitive linguists to reach out to the
other cognitive sciences. If we set out to design collaborative research projects that are deliberately cross-methodological, we will no longer be
forced to rely on evidence gathered from disparate traditions. We have already begun to see movements in this direction. For example, within psychology, Matlock, Ramscar and Boroditsky have investigated whether
spatial metaphors for time alter our inferences about time (2005). Matlock and Richardson (2004) have investigated whether fictive-motion language induces eye-movements. Amorim, Isableu and Jarraya (in press)
have shown that adding a cylindrical head to the standard ShepardMetzler cubes used in mental rotation/mental imagery experiments produces both visual and motoric embodiment effects. Within cognitive
neuroscience, Coulson and van Petten (2002) have investigated the neural
processing of conceptual blends using event-related potentials (ERPs),
while Rohrer (2001b) has investigated whether the sensorimotor cortex responds to both literal and metaphoric body-part sentences using both
ERPs and functional magnetic resonance imaging (fMRI). But more explicitly cross-methodological efforts are needed, particularly for testing
within experimental developmental psychology (see for ideas Mandler
2005) and in primatological studies of sign language use (see for ideas Tomasello, Savage-Rumbaugh and Kruger 1993 and Fouts, Jensvold and
Fouts 2002). Only when Cognitive Linguistics takes its hypotheses to the
other cognitive sciences and develops collaborative, cross-methodological
studies will cognitive linguistics become not just a listening but a speaking
member of the family of cognitive sciences.
Notes
1. I have in mind several series of studies, but in particular the Knudsen and colleagues studies of juvenile and adult barn owls, some of whom who wore prismatic glasses that distort the visual field in the xy-plane (e.g. 23 degrees to the
right) results in an bifurcated neural growth pattern in the cross-modal map of
the ICX in their brains. (Knudsen 2002, 1998). To make the connection between this and language more explicit: Just as the juvenile owls who wore
glasses in their youth become bi-visual, so do the bilingual speakers who are
140
2.
3.
4.
5.
6.
Tim Rohrer
exposed to more than one language from birth are more likely to become
more adroit speakers of those languages than those who start off monolingual
and acquire a second or third language later in life. It seems likely, then that
the many different neural structures underlying second language acquisition
also grow differently, particularly when we are talking about distinguishing
the subtle sorts of spatial relations differences that underlie using the spatial
relations terms perfectly in a second language. For further discussion, see
Johnson and Rohrer (in press) and Rohrer (2005).
Infants have what appears to be a near-innate ability to distinguish between
the full range of human phonemes on a perceptual basis, but between months
1012 lose that raw perceptual ability (Bohn 2000: 610). Instead, they automate and zoom in on those contrasts which are functionally relevant (i.e.
phonemically distinctive) in their own peculiar linguistic environment (Kristiansen 2003: 85). Such infant evidence is not only an example of how the categorization of phonemes in the infant mind emerges from the perceptual categorization processes available in the infant brain, but also demonstrates that
the infants mental categorization of phonemes is neither a result of nor is accessible to the processes of conscious introspection that Zlatev takes as essential
for language. (I wish to thank Gitte Kristiansen and Ren Dirven for their
help in elucidating this point.)
Another source of inspiration that Johnson acknowledges for the idea of
image schemas came from the German school of gestalt perception. Many of
the notions of pattern completion in visual imagery, of the felt sense of aesthetic organization, as well as pattern completions in auditory imagery and
music that contributed to their cross-modal character, were inspired by observations expressed in (for example) Wertheimer (1923).
For example Johnson and Rohrer (in press) discuss the social and cultural dimensions of embodiment in detail, arguing that they are continuous with the
physical and neural substrates. Virtually all such attempts to place embodied
cognition within an evolutionary context begin with the physical and neural
and trace their way up to the social and cultural. This habitual mode of presentation is perhaps a continuing source of the mistaken dogma that embodiment leads to eliminativism.
The conflation hypothesis was arrived at by Christopher Johnson analyzing
language acquisition in a well-known corpus of a single child, Shem (Johnson
1997, 1999). Single subject data, while empirical, is of questionable value
standing alone. The conflation hypothesis is also at odds with many known
facts of the initial development of neural structures; however, it is possible
that at a later stage of cognitive development children might find it difficult to
differentiate image schematic features which are well-integrated in the underlying cross-modal maps. If found, such evidence would certainly lend credence to Zlatevs (in press, 2005) hypothesis that language socialization is
required for image-schematic thought.
In this sense, research in computational modeling such as that of Veale (this
vol.) can serve as a useful methodological precision check.
141
References
Amorim, Michel-Ange, Brice Isableu, and Mohammed Jarraya
in press Embodied spatial transformations: Body analogy for the mental rotation. Journal of Experimental Psychology: General.
Bailey, David
1997
When push comes to shove: A computational model of the role of
motor control in the acquisition of action verbs. Ph.D. diss., University
of California at Berkeley.
Bohn, Ocke-Schwen
2000
Linguistic relativity in speech perception. An overview of the influence
of language experience on the perception of speech sounds from infancy to adulthood. In Evidence for Linguistic Relativity, Susanne Niemeier, and Ren Dirven (eds.), 112. Amsterdam: Benjamins.
Bowerman, Melissa, and Soonja Choi
2003
Space under construction: Language-specific spatial categorization in
first language acquisition. In Language and Mind, Dedre Gentner, and
Susan Goldin-Meadow (eds.), 387427. Cambridge, Mass.: The MIT
Press.
Churchland, Paul
1981
Eliminative materialism and the propositional attitudes. Journal of
Philosophy 78: 6790.
1989
A Neurocomputational Perspective: The Nature of Mind and the Structure of Science. Cambridge, Mass.: The MIT Press.
Clark, Andrew
1998
Being There: Putting Brain, Body, and World Together Again. Cambridge, Mass.: The MIT Press.
Coulson, Seana, and Cynthia Van Petten
2002
Conceptual integration and metaphor: An event-related potential
study. Memory & Cognition 30 (6): 958968.
Dewey, John
1925
Experience and Nature. In John Dewey, The Later Works, 19251953,
vol. 1, Jo Ann Boydston (ed.). Carbondale: Southern Illinois University Press, 1981.
Dodge, Ellen, and George Lakoff
2005
Image schemas: From linguistic analysis to neural grounding. In From
Perception to Meaning: Image Schemas in Cognitive Linguistics, Beate
Hampe (ed.), 5791. (Cognitive Linguistics Research 29.) Berlin/New
York: Mouton de Gruyter.
142
Tim Rohrer
Edelman, Gerald M.
1992
Bright Air, Brilliant Fire: On the Matter of Mind. New York: Basic
Books.
Fauconnier, Gilles, and Mark Turner
2002
The Way We Think: Conceptual Blending and the Minds Hidden Complexities. New York: Basic Books.
Feldman, Jerome, and Srini Narayanan
2004
Embodied meaning in a neural theory of language. Brain and Language
89 (2): 385392.
Fouts, Roger S., Mary Lee A. Jensvold, and Deborah H. Fouts
2002
Chimpanzee signing: Darwinian realities and Cartesian delusions. In
The Cognitive Animal: Empirical and Theoretical Perspectives in Animal
Cognition, Bekoff, Marc, Colin Allen, and Gordon M. Burghardt
(eds.), 285291. Cambridge, Mass.: The MIT Press.
Gallese, Vittorio, and George Lakoff
2005
The brains concepts: the role of the sensory-motor system in conceptual knowledge. Cognitive Neuropsychology 22: 455479.
Geeraerts, Dirk, and Stefan Grondelaers
1995
Looking back at anger: Cultural traditions and looking back at anger:
Cultural traditions and metaphorical patterns. In Language and the
Cognitive Construal of the World, John R.Taylor, and Robert E. MacLaury (eds.), 153179. (Trends in Linguistics Studies and Monographs 82.) Berlin/New York: Mouton de Gruyter.
Gibbs, Raymond W.
1994
ThePoetics of Mind: Figurative Thought, Language, and Understanding.
Cambridge: Cambridge University Press.
Grady, Joseph E.
1997
THEORIES ARE BUILDINGS revisited. Cognitive Linguistics 8:
267290.
2005a Primary metaphors as inputs to conceptual integration. Journal of
Pragmatics 37 (10): 15951614.
2005b Image schemas and perception: Refining a definition. In From Perception to Meaning: Image Schemas in Cognitive Linguistics, Beate Hampe
(ed.), 3555. (Cognitive Linguistics Research 29.) Berlin/New York:
Mouton de Gruyter.
Grady, Joseph E., and Christopher Johnson
2002
Converging evidence for the notions of subscene and primary scene.
In Metaphor and Metonymy in Comparison and Contrast, Ren Dirven,
and Ralf Prings (eds.), 533554. Berlin/New York: Mouton de
Gruyter.
Hauk, Olaf, Ingrid Johnsrude, and Friedemann Pulvermller
2004
Somatotopic representation of action words in human motor and premotor cortex. Neuron 41: 301730.
Hutchins, Edwin
1995
Cognition in the Wild. Cambridge, Mass.: The MIT Press.
143
James, William
1900
Psychology (American Science Series, Briefer Course) New York:
Henry Holt and Company.
Johnson, Christopher R.
1997
The acquisition of the Whats X doing Y construction. In Proceedings of the 21st Boston University Conference on Language Development 2, Somerville, Mass.: Cascadia Press.
1999
Metaphor vs. conflation in the acquisition of polysemy: The case of
see. In Cultural, Psychological and Typological Issues in Cognitive Linguistics. Selected Papers of the Bi-annual ICLA Meeting in Albuquerque,
July 1995, Masako K. Hiraga, Chris Sinha, and Sherman Wilcox (eds.),
155169. (Current Issues in Linguistic Theory 152.) Amsterdam/Philadelphia: Benjamins.
Johnson, Mark
1987
The Body in the Mind: The Bodily Basis of Meaning, Imagination, and
Reason. Chicago: University of Chicago Press.
2005
The philosophical significance of image schemas. In From Perception to
Meaning: Image Schemas in Cognitive Linguistics, Beate Hampe (ed.),
1534. (Cognitive Linguistics Research 29.) Berlin/New York: Mouton
de Gruyter.
Johnson, Mark, and Tim Rohrer
In press. We are live creatures: Embodiment, American pragmatism, and the
cognitive organism. In Body, Language, and Mind. Vol. 1: Embodiment,
Tom Ziemke, Roslyn Frank, and Ren Dirven (eds.). Berlin/New York:
Mouton de Gruyter.
Knudsen, Eric I.
2002
Instructed learning in the auditory localization pathway of the barn
owl. Nature 417 (6886): 322328.
1998
Capacity for plasticity in the adult owl auditory system expanded by juvenile experience. Science 279 (5356): 15311533.
Korzybdski, Alfred
1948
On structure. In Selections from Science and Sanity (1972 ed 7th edition). Lakeville, Illinois: International Non-Aristotlean Publishing.
Kosslyn, Steven
1980
Image and Mind. Cambridge, Mass.: Harvard University Press.
1994
Image and Brain: The Resolution of the Imagery Debate. Cambridge,
Mass.: The MIT Press.
Kosslyn, Stephen, William L. Thompson, Irene J. Kim, and Nathaniel M. Alpert
1995
Topographic representations of mental images in primary visual cortex. Nature 378: 496498.
Kristiansen, Gitte
2003
How to do things with allophones. Linguistic stereotypes as cognitive
reference points in social cognition. In Cognitive Models in Language
and Thought. Ideology, Metaphors and Meaning, Ren Dirven, Roslyn
Frank, and Martin Ptz (eds.), 69120. (Cognitive Linguistics Research 24.) Berlin/New York: Mouton de Gruyter.
144
Tim Rohrer
Kuhn, Thomas
1970
The Structure of Scientific Revolutions, 2nd edition, enlarged. Chicago:
University of Chicago Press.
Lakoff, George
1987
Women, Fire and Dangerous Things: What Categories Reveal about the
Mind. Chicago: University of Chicago Press.
Lakoff, George, and Mark Johnson
1980
Metaphors We Live by. Chicago: University of Chicago Press.
1999
Philosophy in the Flesh: The Embodied Mind and its Challenges to Western Thought. Chicago: University of Chicago Press.
Langacker, Ronald W.
1987
Foundations of Cognitive Grammar. Vol. 1.: Theoretical Prerequisites.
Stanford: Stanford University Press.
1991
Foundations of Cognitive Grammar. Vol. 2: Descriptive Application.
Stanford: Stanford University Press.
Mandler, Jean
2005
How to build a baby: III. Image schemas and the transition to verbal
thought. In From Perception to Meaning: Image Schemas in Cognitive
Linguistics, Beate Hampe (ed.), 137164. (Cognitive Linguistics Research 29.) Berlin/New York: Mouton de Gruyter.
Matlock, Teenie, Michael Ramscar, and Lera Boroditsky
2005
The experiential link between spatial and temporal language. Cognitive
Science 29: 655664.
Matlock, Teenie, and Daniel Richardson
2004
Do eye movements go with fictive motion? In Proceedings of the 26th Annual Conference of the Cognitive Science Society, Kenneth Forbus, Dedre
Gentner, and Terry Regier (eds.), Mawhaw, N.J.: Lawrence Erlbaum
Meltzoff, Andrew
1993
Molyneuxs babies: Cross-modal perception, imitation and the mind of
the preverbal infant. In Spatial Representation: Problems in Philosophy
and Psychology, Naomi Eilan, Rosaleen McCarthy, and Bil Brewer
(eds.), 219235. Oxford: Blackwell.
Narayanan, Srini
KARMA: Knowledge-based active representations for Metaphor and
1997
aspect. MS thesis, University of California at Berkeley.
Posner, Michael I., and Marcus E. Raichle
1997
Image of Mind. New York: W. H. Freeman; Scientific American
Reddy, Michael J.
1993
The conduit metaphor: A case of frame conflict in our language about
language. In Metaphor and Thought, Andrew Ortony (ed.). Cambridge:
Cambridge University Press.
Regier, Terry
1992
The acquisition of lexical semantics for spatial terms: A connectionist
model of perceptual categorization. Ph.D. dissertation. University of
California at Berkeley.
145
The Human Semantic Potential: Spatial Language and Constrained Connectionism. Cambridge, Mass.: The MIT Press.
Rohrer, Tim
1995
The metaphorical logic of (political) rape: The new wor(l)d order.
Metaphor and Symbolic Activity 10 (2): 115137.
2001a Pragmatism, ideology and embodiment: William James and the philosophical foundations of cognitive linguistics. In Language and Ideology.
Volume 1: Theoretical Cognitive Approaches, Ren Dirven, Bruce Hawkins, and Esra Sandikcioglu (eds.), 4981. (Current Issues in Linguistic
Theory 204.) Amsterdam/Philadelphia: Benjamins.
2001b Understanding through the body: fMRI and of ERP studies of metaphoric and literal language. Ms., Paper presented at the 7th International Cognitive Linguistics Association Conference, July 2001.
2005
Image schemata in the brain. In From Perception to Meaning: Image
Schemas in Cognitive Linguistics, Beate Hampe (ed.), 165196. (Cognitive Linguistics Research 29.) Berlin/New York: Mouton de Gruyter.
in press. The body in space: Embodiment, experientialism and linguistic conceptualization. In Body, Language and Mind, vol. II, Roslyn Frank,
Enrique Bernrdez, Ren Dirven and Tom Ziemke (eds.). Cognitive
Linguistics Research. Berlin/New York: Mouton de Gruyter.
Searle, John
1992
The Rediscovery of the Mind. Cambridge, Mass.: The MIT Press.
Shepard, Roger N., and Jacqueline Metzler
1971
Mental rotation of three-dimensional objects. Science 171: 701703.
Sinha, Chris, and Kristine Jensen de Lpez
2000
Language, culture, and the embodiment of spatial cognition. Cognitive
Linguistics 11: 1741.
Stern, Daniel
1985
The Interpersonal World of the Infant. New York: Basic Books.
Sweetser, Eve E.
1990
From Etymology to Pragmatics: Metaphorical and Cultural Aspects of
Semantic Structure. Cambridge: Cambridge University Press.
Talmy, Leonard
1985
Force dynamics in language and thought. In Papers from the Twentyfirst Regional Meeting of the Chicago Linguistic Society (2): Possession
in Causatives and Agentivity, William H. Eikfont, Paul D. Kroeber, and
Karen L. Peterson (eds.). Chicago: Chicago Linguistic Society.
2000a Toward a Cognitive Semantics, Vol. I: Concept Structuring Systems.
Cambridge, Mass.: The MIT Press.
2000b Toward a Cognitive Semantics, Vol. II: Typology and Process in Concept
Structuring. Cambridge, Mass.: The MIT Press.
Tomasello, Michael, Sue Savage-Rumbaugh, and Ann C. Kruger
1993
Imitative learning of actions on objects by children, chimpanzees, and
enculturated chimpanzees. Child Development 64 (6): 16881705.
146
Tim Rohrer
Wertheimer, Max
1923
Untersuchungen zur Lehre von der Gestalt II. Psycologische Forschung
4: 301350. (Translation published in A Source Book of Gestalt Psychology, Ellis, W. (ed.) (1938), 7188). London: Routledge & Kegan
Paul).
Zlatev, Jordan
1997
Situated Embodiment: Studies in the Emergence of Spatial Meaning.
Stockholm: Gotab Press.
2003
Polysemy or Generality? Mu. In Cognitive Approaches to Lexical Semantics, Hubert Cuyckens, Ren Dirven, and John Taylor (eds.),
447494. Berlin/New York: Mouton de Gruyter.
2005
Whats a schema? Bodily mimesis and the grounding of language. In
From Perception to Meaning: Image Schemas in Cognitive Linguistics,
Beate Hampe (ed.), 313342. (Cognitive Linguistics Research 29.) Berlin/New York: Mouton de Gruyter.
in press. Embodiment, language, mimesis In Body, Language and Mind, vol. 2.
Tom Ziemke, Roslyn Frank, Ren Dirven (eds.). Berlin/New York:
Mouton de Gruyter.
147
Abstract
In this chapter a view of metonymy as an indexical relation between source and
target meaning is presented and contrasted with metaphor, which is regarded as a
specific kind of iconic relation. It is shown that notions such as domain, subdomain, single domain, and separate domains, which have often been used as
definitional criteria for distinguishing metonymy from metaphor, are unreliable
because they are cover terms for heterogeneous concepts and conceptual relations.
Furthermore, it is argued that metonymy is a kind of meaning elaboration whose
result is a conceptually prominent target meaning, an integrated whole that contains the backgrounded source meaning and novel meaning components resulting
from the process of elaboration. Finally, the role of context in the interpretation of
metonymy in usage events is explored in two case studies. A detailed analysis of
two contextualized metonymically used expressions, suggests that one important
function of metonymy is to provide generic prompts. These serve as inputs for additional pragmatic inferences that flesh out the specifics of the intended utterance
meaning.
Keywords: conceptual prominence, domain, generic prompt, iconic relation, indexical relation, meaning elaboration, pragmatic inference
1.
Introduction
The last two decades have seen a remarkable renewed interest in figures of
speech, in particular, metaphor and metonymy. The interpretation of
metaphor and metonymy has been a major concern in psycholinguistics,
pragmatics, and, of course, cognitive linguistics (see Gibbs 1994). In cognitive linguistics the focus of analysis has shifted from the stylistic and
embellishing function of metaphor and metonymy in discourse to their
role in human thinking and reasoning, and their impact on language
structure and language use. The study of language use is of common interest to both pragmatics and cognitive linguistics. In the latter, the analysis of usage events, i.e. actual utterances in their full phonetic detail and
contextual understanding (Langacker 2000: 2), has, in recent years,
moved from a mere declaration of intention to a number of full-blown
148
Klaus-Uwe Panther
studies based on large corpora (for an overview see Geeraerts 2005). This
chapter does not present a corpus-based quantitative analysis of the function of metonymy in usage events, but carries out an in-depth conceptual
and pragmatic analysis of a few illustrative examples that involve metonymic reasoning. My aim is to show that metonymic meanings provide
generic prompts that are fleshed out on the basis of background knowledge (world knowledge), the situation of the utterance and the linguistic
context (co-text) in which the metonymic expression occurs. My analysis
proceeds from the local, i.e. intra-clausal context, in which the metonymic
expression occurs, to the larger discourse and the extralinguistic context.
The method of analysis pursued in this chapter is thus complementary
to that of large-scale corpus studies. There are some obvious drawbacks
to my methodology one being that any kind of generalization is highly
risky. However, large-scale corpus-based studies also have their disadvantages: the higher the quantity of data to be covered by the analyst, the
greater the danger that their conceptual and pragmatic analysis will be
somewhat shallow and that important features of the data will be overlooked.
More specifically, the chapter is organized as follows. In section 2, I
present my own view of metonymy as a figure of thought and language,
and identify some typical properties of metonymy. I share a substantial
set of assumptions with other researchers, but I also see some problems
with the current use of key notions such as domain, subdomain, single
domain, and separate domains, which figure prominently as criteria for
distinguishing metaphor from metonymy. Rather than defining metaphor
and metonymy in terms of the notions separate domains and single domain, respectively which need further clarification I propose that the
difference between metaphor and metonymy resides in the type of semiotic relation between their respective source and target. I contend that
metaphor is an iconic relation and metonymy an indexical relation. This
semiotic approach, in the spirit of C. S. Peirce, and more recently Masako
Hiraga (1998, 2005) has the advantage that some properties of metonymy
follow quite naturally from the premise that it is an indexical relation. I
also argue that metonymy is a type of meaning elaboration, a term that
comes close to, but is not identical with, what Francisco Ruiz de Mendoza
calls domain expansion (see e.g. Ruiz de Mendoza Ibez and Otal
Campo 2002, Ruiz de Mendoza Ibez and Pea 2005). Section 2 also addresses some criticism that has been aimed against the approach to metonymy advocated by the present author and Linda L. Thornburg (Haser
2005).
149
2.
150
Klaus-Uwe Panther
what is known as the paradigmatic axis of language, whereas words occurring contiguously in a syntactic arrangement exemplify the syntagmatic
axis. As to similarity and contiguity of meaning, Jakobson reports a psychological test in which children were asked to provide spontaneous verbal responses to the stimulus word hut. Some subjects showed a predilection for words that were either similar or opposite in meaning such as
cabin, hovel (synonyms), den, burrow (called metaphors by Jakobson), palace (antonym). Other subjects preferred semantically contiguous concepts, as exemplified in words like thatch, litter, and poverty. According to
Jakobson, the former subjects, whose responses were based on similarity
(and contrast) manifested a metaphoric mode of thinking whereas the
latter, whose thinking relied on semantic contiguity, a metonymic mode of
reasoning.
At first sight, Jakobsons view seems to be at odds with both the rhetorical tradition and cognitive linguistic theories of metonymy. Jakobson
relates metonymy to the syntagmatic axis of language, but in cognitive accounts of metonymy it is quite common to regard metonymy as an operation of substitution, i.e. a paradigmatic relation, a view that is reflected
in the notation x for y, which is commonly used for notating metonymies.
However, on closer inspection, Jakobsons conception of metonymy is
compatible with the substitution view. The crucial distinction involved
here is that between positional similarity/contiguity and semantic similarity/contiguity. In Jakobsonian terms, one could say that metonymy involves semantic contiguity, which manifests itself as positional similarity.
A simple referential metonymy such as Mary is not in the phone book substitutes Marys name (the metonymic target) for the metonymic vehicle
Mary. The metonymic operation occurs in a specific syntactic position in
the sentence and is therefore paradigmatic, but the relation between the
metonymic source and its target is one of semantic contiguity.
In principle, Jakobsons conception of metonymy is not restricted to
referential metonymy and it is thus quite compatible with cognitive linguistic views of metonymy. However, at least as far as his examples of semantic contiguity are concerned, Jakobsons approach to metonymy is far
too unconstrained. Consider e.g. the words that were associatively elicited
in the experiment mentioned above from the stimulus word hut: thatch,
litter, and poverty. None of these words seems to be able to be used metonymically for hut:
(2) *The old man lived in a little thatch/litter/poverty. (for hut)
151
As pointed out above, the present author views metonymy not only as a
referential shift phenomenon, but assumes that metonymic operations are
active also on the levels of predication and illocution (see e.g. Panther
2005). In line with other cognitive approaches, especially Radden and Kvecses (2006), I regard metonymy as a cognitive operation through which
one concept (the source) provides access to another concept (the target).
Furthermore, I assume that the target meaning resulting from a metonymic shift is an elaboration of the source meaning. The metonymic shift
can be regarded as a substitution operation, but one in which the source
meaning does not vanish but remains part of the conceptual structure of
the target meaning.2 What I regard to be the typical metonymic relation is
diagrammed in Figure 1, which is explained in more detail below.
152
Klaus-Uwe Panther
153
nymy, involves an indexical relationship. In a linguistic metonymy, a linguistic form, which I call linguistic vehicle, denotes a source meaning. The
relation between linguistic vehicle and source meaning is in general symbolic (indicated by a thick line in Figure 1), i.e. there is no natural connection between the form (linguistic vehicle) and its (literal) content. The
source meaning, in turn, functions as a conceptual vehicle (alternatively, a
thought vehicle) that provides cognitive access to the target meaning.4
The conceptual vehicle is thus a signifier that indexically evokes its signified, i.e. target meaning.5 The link between source meaning and target
meaning can thus be thought of as a semiotic relation, viz. an indexical relation. In the framework adopted here, indexicality is a pointing-to relation between one concept and another concept.6 In other words, different from the usually arbitrary (symbolic) relation between linguistic
vehicle and source meaning, the relation between source meaning and target meaning is motivated.7 In Figure 1 the indexical relation between the
conceptual vehicle and the target meaning is indicated by an open-headed
arrow. As already pointed out above, Figure 1 also encapsulates the idea
that the identification of the metonymic target meaning may be dependent
on other conceptual components in the conceptual domain. Their potential contribution to the identification of the target meaning is symbolized
by dotted-lined arrows. It should be borne in mind that the target conceptually integrates the source meaning, allocating it a backgrounded status.
Another important property of metonymy graphically represented by
means of bold type in Figure 1 is the conceptual prominence of the target.
In other theories, e.g. Croft (2002), this corresponds roughly to what he
calls domain highlighting, i.e., an originally backgrounded subdomain
within a domain becomes foregrounded through the metonymic operation.8 The property of conceptual prominence holds for prototypical
cases of metonymy. Some cases that, in the literature on metonymy, are
cited as typical examples of metonymy are consequently classified as
more peripheral in my approach:
(4) Bush invaded Iraq.
Sentence (4) is usually regarded as a typical example of referential metonymy of the type controller for controlled (cf. Lakoff and Johnson
1980: 38). However, it is clearly a source-prominent metonymy: the target,
i.e. troops, air force, etc. that were actively involved in the invasion of Iraq
remain indeterminate and backgrounded (see Panther 2005: 371274 for
detailed discussion).
154
Klaus-Uwe Panther
The conception of metonymy presented here is, in some respects, different from the theories of metonymy developed by researchers such as
Antonio Barcelona, William Croft, Francisco Ruiz de Mendoza, Gnter
Radden and Zoltn Kvecses. It is also, at least partially, at odds with Lakoff and Johnsons (1999) approach to metaphor and metonymy (see especially section 2.4 below). To list two differences very briefly:
i. In other theories, metonymy is regarded as linking one subdomain to
another subdomain within the same domain (e.g. Barcelona 2002;
Croft 2002); as a mapping from a subdomain to its containing matrix
domain, or vice versa (e.g. Ruiz de Mendoza Ibez and Otal Campo
2002); or as a mapping from part to whole, whole to part, or part to
part (Radden and Kvecses 2006). In the present approach, metonymy
is regarded as a means of semantic enrichment or elaboration. This
means that through a metonymic operation a source concept is expanded so that, as a result, the target concept is construed as conceptually more complex. At the same time, the source concept is an integrative part of the target concept.
ii. The approach advocated here assumes that the metonymic relation is
contingent, i.e. not any meaning that is mentally accessible from a certain conceptual reference point gives rise to a metonymic relation.
More precisely, I do not regard relations that are conceptually necessary as possible metonymies. In other words, mental access of concepts
based on semantic entailment and semantic presupposition is not considered to be a mode of metonymic reasoning here.
Contingence as a fundamental property of metonymy has been emphasized in publications such as Panther and Thornburg (2003) and Panther
(2005). The notion of contingence, as used in the above-cited publications,
has sometimes been misunderstood. Contingence means simply that the
metonymic relation per se is defeasible or cancelable (which distinguishes
it from entailment). Contextual factors may however override defeasibility of the metonymic relation and more or less enforce (coerce) a metonymic interpretation. For example, in an utterance such as The bathtub is
overflowing, the noun phrase the bathtub is metonymically elaborated into
something like the water in the bathtub. The above utterance thus metonymically conveys The water in the bathtub is overflowing. This interpretation is virtually enforced or coerced by the context (here the predicate expression is overflowing), but this fact does not invalidate the point
that the metonymic relation between the source meaning bathtub and
155
the target water (in the bathtub) is incidental, i.e. logically not necessary.
No doubt, there is a tight (factual) relation between bathtubs and water in
that the primary function of bathtubs is for people to take baths in.
Nevertheless, bathtubs can be used for many other purposes. In contrast,
the relation between e.g. broccoli and vegetable is hyponymic and therefore in the taxonomy of vegetables necessary by definition. The sentence They had broccoli with their meal entails They had a vegetable with
their meal, but broccoli is not a metonym for vegetable (see Bierwiaczonek 2005 for an analogous argument). In this connection, note that for
Charles S. Peirce an indexical relation is related to its object not by
rational necessitation but by blind fact (see fn. 3). If it is assumed that
metonymy is a case of indexicality, the contingence of the metonymic relation follows automatically.
A case that, at first sight, seems to be based on a hyponymic relation between source and target is given by Radden and Kvecses (2006: 11).9
These authors posit a metonymy member of a category for the category that they claim accounts for uses of aspirin for pain-relieving tablet. Obviously, under one construal, the meaning of aspirin entails painrelieving tablet.10 I agree with Radden and Kvecses that there is a real
metonymy here, but it seems that it is not captured by the metonymic principle member of a category for the category.11 Notice first that the
relation between the concepts aspirin and pain-reliever is not one of
meaning elaboration, which I claim is a typical characteristic of metonymy, because the feature pain-reliever is already a part of the definition of (one sense of) aspirin. The metonymic relation is between aspirin
and its co-hyponyms, i.e. other pain-relievers (including aspirin itself). The
paraphrase given by Radden and Kvecses, viz. any [italics mine,
K.-U.P.] pain-relieving tablet, is quite revealing in this respect. It
amounts to saying aspirin or any medication that has the same effect as
aspirin. In other words, what is evoked metonymically is not the superordinate category/class pain-reliever, but, via co-hyponymic extension,
any member of this category/class (see Figure 2 below).
That metonymy is a matter of contingence rather than conceptual
necessity is also advocated by Carita Paradis. She contrasts sentences
such as (5) with sentences of type (6) (Paradis 2004: 246):
(5) The red shirts won the match.
(6) The court had to assume that the statement of claim was true.
156
Klaus-Uwe Panther
The noun phrase the red shirts in (5) metonymically indexes the team
wearing the red shirts, but this target meaning is not entailed by the
meaning of the linguistic vehicle. In (6), as Paradis points out, the intended target meaning is something like staff (i.e. the personnel working
in the court), but this meaning is a facet of the overall meaning of court,
the other three facets being administrative unit, building, and interior outfit (251). According to Paradis, individual facets are entailed
by the overall meaning of a lexical item. This might be too strong a statement, but it is certainly a default assumption (generalized conversational
implicature in terms of Levinson 2000) that a court is staffed (with judges
and other personnel). However, the same kind of connection does not obtain between the concepts red shirts and team wearing red shirts. The
difference between metonymy and facets, as has been pointed out by a
number of linguists, is also reflected in grammatical structure. In the case
of metonymy, it is usually the target meaning that, because of its conceptual prominence (highlighting), determines grammatical and/or distributional properties of a sentence, whereas a switch from one lexical facet to
another within a sentence usually goes unnoticed and is not considered to
be anomalous.
(7) I listened to some/*a Mozart last Sunday.
(8) The red shirts, who/*which delivered a great game last Saturday, have
qualified for the Champions League.
(9) Put this heavy book back on the shelf; it is rather unreadable. (Cruse
2000: 114)
(10) The court, which is located on Russell Square, has ruled in Marys
favor.
157
In (7), Mozart is metonymically interpreted as music composed by Mozart and it is interpreted as an abstract substance that determines the
choice of some over the determiner a, which requires a noun with the feature countable. Thus the metonymic target determines the grammatical
structure of the noun phrase some Mozart. Similarly, in (8) it is the conceptual nature of the target team wearing red shirts that determines the
selection of the relative pronoun who, excluding the choice of which. Sentences (9) and (10) are examples of facet shifting. Sentence (9) starts out
with the reading of book as physical object and, in the second clause,
switches to the facet content. In (10), the overall meaning of court is first
narrowed down to the facet building motivating the use of the relative
pronoun which and then interpreted as staff (e.g. judge) under the influence of the verb form has ruled.
To summarize the main points, I have based my characterization of metonymy on the semiotic notion of indexicality. This is not necessarily incompatible with a definition of metonymy in terms of single domainhood,
but it seems extremely difficult to establish criteria that would allow one
to distinguish reliably between separate conceptual domains and single
conceptual domains. In the case of (8) above, one can view red shirts as belonging to the domain of clothing, which is relatable to, but also separate
from, the domain of humans (or the narrower domain of athletes). It is
difficult to pin down the notions single domain and separate domains
in a way that does not seem to be more or less ad hoc. Nevertheless these
notions are considered by many scholars to be essential in telling metaphor apart from metonymy. I turn to this problem in more detail in the
following section.
2.2. Domains and subdomains
As mentioned in section 2.1, standard definitions of metonymy often rely
on the notion of single domain or domain matrix within which the metonymic mapping is supposed to operate (see e.g. Radden and Kvecses
2006).12 More specifically, Francisco Ruiz de Mendoza and his coauthors
have forcefully argued in a number of publications (see e.g. Ruiz de Mendoza Ibez and Dez Velasco 2002; Ruiz de Mendoza Ibaez and Otal
Campo 2002) that metonymy amounts to two kinds of operation: domain
expansion (source-in-target metonymy) or domain reduction (targetin-source metonymy). Ruiz de Mendoza and his collaborators reject partto-part metonymies, claiming that they can be reduced to either domain
expansion or domain reduction.
158
Klaus-Uwe Panther
159
rooms a subset of the set of buildings. The relation between noses and
faces and rooms and houses, respectively, is not one of class inclusion.
Again, it seems to be a matter of perspective that determines what is regarded as the superordinate domain and what as the subdomain. One
could argue that a proper understanding of house includes some understanding of the concept room, kitchen, attic, basement, etc.; thus
house would constitute a superordinate domain vis--vis room. Conversely, one could also argue that room requires reference to concepts like
house, buidling, cottage, palace and the like, and postulate that
room is a superordinate domain from that perspective. The person-attribute relation in (iv) seems to be a clear case, at first sight. Intuitively,
one tends to think that an attribute is subordinate to the person that has
it. Even if this perspective is the normal one, it does not exclude another
perspective where e.g. potbelly is a superordinate domain comprising
the class of people who display the attribute of having a potbelly. Finally,
consider relations such as those exemplified by (v), which also have been
claimed to exhibit a domainsubdomain structure. What is the criterion
for deciding that location is the superordinate domain of institution,
which itself is supposed to be the superordinate domain of people (working at the institution)? True, 10 Downing Street, which refers to a location
in London and a house at this location, can metonymically evoke the institution and ultimately the human beings working in the institution, but
the same arguments as above again apply: One can see an institution as
having as one of its features a location; there is no inherent necessity to
view institution as being a subdomain of location, although it can certainly be construed that way.13
Croft (2002: 166) seems to offer a viable solution to the problem of distinguishing domains from subdomains. He proposes the following definition for domain: We can [] define a domain as a semantic structure that
functions as the base for at least one concept profile (typically, many profiles). The two notions of profile and base (adopted from Langacker
1987) are especially relevant here: According to Croft, a base is the domain immediately presupposed by the profiled concept (167). Croft calls
the base of a profiled concept its base domain. This definition seems, at
least in part, to coincide with Ruiz de Mendozas use of the term matrix
domain. Notice that Croft uses the term presupposition to characterize
the relation between profile and base. It does not become clear, however,
whether he has the notion of semantic presupposition in mind; if so, it
would render statements about domain structure testable by means of the
well-known (internal) negation test for presupposition.14 But the appli-
160
Klaus-Uwe Panther
cation of this test leads to unsatisfactory results: For example, the notion
arc is, in Crofts sense, profiled against, or presupposes, the base circle.
However, neither from the proposition This is an arc nor from its negation
This is not an arc follows the proposition This is a circle. As another
example consider the relations between nose, face, and head. Does nose
presuppose face, which, in turn, would presuppose head? The (frame-internal) negation test fails here: Neither from x is a nose nor from x is not a
nose does it follow that x is a face; analogously, for face and head. The
presupposition test works somewhat better with taxonomies on condition
that negation is restricted to frame- or domain-internal negation: From
This is a rose the proposition This is a flower follows; frame-internal negation (This is not a rose) would also imply This is a flower. In conclusion,
even Crofts attempt at providing a more constrained characterization of
the relation between domains and subdomains is inconclusive. Presupposition, as defined in standard semantics, does not provide a sufficient
criterion for distinguishing superordinate domains from subordinate domains.
An additional complication results from the fact that Croft different
from e.g. Ruiz de Mendoza does not propose a characterization of metonymy in terms of domain expansion and domain reduction but in terms of
highlighting an originally secondary subdomain: [] metonymy makes
primary a domain that is secondary in the literal meaning (Croft 2002:
179). This conception of metonymy is not compatible with the idea that
metonymy is a domain-subdomain or subdomain-domain relationship.
What counts as a domain and a subdomain might be a matter of perspective or construal on the part of the conceptualizer. Take the example
of french fries, which is supposedly a subdomain of the domain of customers (see Ruiz de Mendoza Ibez and Otal Campo 2002: 132 for an analogous example). I think it is just a matter of perspective to reverse the subdomaindomain relationship in this case. French fries are a specific kind
of cooked food (potato sticks fried in oil), often (but not necessarily)
served in fast food restaurants to customers who are in a hurry and do not
want to spend much time and money on a meal. There is thus a whole
knowledge structure concerning french fries and, one could argue that
part of this knowledge structure is the concept of customer. From this
perspective, french fries constitutes the superordinate domain of restaurant customer. There is thus again no clear objective sense that
would allow one to distinguish domains from their subdomains.
The preceding discussion should not be interpreted as implying that I
argue against the existence of frames, domains, etc., as such. On the
161
162
Klaus-Uwe Panther
The conceptual frames for tiger (target) and lion (source) overlap considerably. Still, it is possible to construe a metaphoric sense for (11) where
properties of the concept lion are mapped onto tigers. It thus does not
seem to be a necessary condition for metaphors that the source domain
and the target domain be disjunct. As for metonymy, I suggest a semiotically based conception of metaphor, which is represented in Figure 3. Following Hiraga (1998), I propose that metaphor is based on an iconic relation between source meaning and target meaning. The source meaning
has some frame or domain structure that is iconically replicated in the target domain. This approach is, to a certain extent, in line with traditional
cognitive metaphor theory as advocated by Lakoff and Johnson (1980).
The present approach differs from Lakoff and Johnson in that it treats the
source meaning of a metaphor, in analogy to metonymy, as a thought vehicle (signifier) for accessing a target meaning (signified). The conception
of metaphor as an iconic relationship accords with traditional accounts of
metaphor in that it relies on some notion of similarity between source and
target. Hiraga (1998) suggests that in metaphor the resemblance between
source and target is created rather than given, i.e. even a priori dissimilar
domains can be connected through metaphoric mappings. The resemblance in this case is structural rather than a direct imitation of (perceived) reality in the signifier of the linguistic sign, as in onomatopoeia
or other kinds of sound symbolism. The structural resemblance is one between the two conceptual frames, the source and the target, as has been
demonstrated by Lakoff and Johnson with illuminating examples. The
view of metaphor as a special case of iconicity (isomorphism) avoids the
pitfalls of having to come up with some criterion for distinguishing single
domains from separate domains. Figure 3 diagrams the basic metaphoric
relation. Note that the target meaning is printed in bold as in the case of
metonymy, indicating that it is conceptually prominent. For convenience
of graphic representation, source and target domain are represented as
distinct in Figure 3 (see below).
2.4. Primary metaphors
Lakoff and Johnson (1999: 4573), who rely on e.g. Grady (1998), assume
the existence of a set of metaphorical primitives that they call primary
metaphors. Primary metaphors are supposedly based on experiential
correlations, which are originally indistinguishable for the child and become only experientially and conceptually dissociated at a more advanced
age. This dissociation is a prerequisite for their development into primary
163
Examples given by Lakoff and Johnson for such primary metaphors are,
among others, affection is warmth, closeness is intimacy, happy is
up, states are locations, purposes are destinations, and knowing
is seeing.16 In this section, I want to defend the idea that these correlations are first and foremost the experiential basis for conceptual metonymy, which, in turn, may become the basis for the construction of metaphors (see also Radden 2002, who develops a similar argument).
Let us consider the experiential correlation between closeness and
intimacy in more detail. Given this correlation, it is not surprising, that
people spontaneously infer that two people who are located close to each
164
Klaus-Uwe Panther
165
166
Klaus-Uwe Panther
ing between meaning something different from what is said and meaning
something in addition to what is said in their approach to metonymy.
Haser points out that, in a case of genuine metonymy, such as the German
expression ein Glas trinken drink a glass, the container (the glass) stands
for its liquid contents (water, milk, beer, wine, etc.). Haser claims that
Panther and Thornburg classify as metonymies phenomena that do not
involve substitution of meaning but rather addition of meaning.
The pragmatic types of metonymy distinguished by Panther and
Thornburg are illustrated by the following sentences (see e.g. Panther and
Thornburg 2003; Panther 2005; Panther and Thornburg In Press):
(17) A lot of Ph.D.s have been written about this topic. (referential: doctoral degree for doctoral thesis)
(18) The President was clear on the matter. (predicational: manner of
speaking for speaking)17
(19) I would like you to close this window. (illocutionary: wish for request)
Consider now (20), which Panther and Thornburg regard as inducing a
predicational metonymy resulting in the interpretation (21):
(20) He was able to finish his dissertation.
(21) He finished his dissertation.18
Haser (2005: 18) claims that the relation between (20) and (21) is one of
addition rather than of substitution so that the almost automatic inference (21) from (20) would not be analyzable as a metonymic relation
contrary to Panther and Thornburgs analysis. Furthermore, Haser
(rightly) points out that (21) entails (20). The entailment relation between
target and source does, however, according to her, not hold in classical
cases of metonymy as in (22), which metonymically evokes (23):
(22) She is listening to Schubert all day.
(23) She is listening to music composed by Schubert all day.
Obviously, (23) does not entail (22). From this fact Haser concludes that
the relation between (22) and (23) is not of the same kind as that between
(20) and (21). Haser draws the conclusion that since the relation between
(22) and (23) is metonymic, the relation between (20) and (21) cannot be
metonymic.
167
Hasers argumentation is however not stringent. If one wants to establish target-source entailment as a criterion for non-metonymicity, what
should be compared is the relation between the denotata of the nominal expressions Schubert and of its metonymic target music composed by Schubert not the relation between the two propositions expressed by sentences (22) and (23), respectively. The metonymic target music composed
by Schubert expresses the underlying proposition Schubert composed
music. From this latter proposition the proposition follows that Schubert
existed.19 Thus, the relationship between was able to finish his dissertation
and finished his dissertation in (20) and (21) is analogous to that between
the existential presupposition Schubert existed (following from Schubert)
and the proposition Schubert composed music (following from music
composed by Schubert), in (22) and (23), respectively. Just as ability is a prerequisite for action, the existence of a person is a prerequisite for that person to be a composer. Table 1 summarizes these analogies.
Table 1. Analogies between referential and predicational metonymy
REFERENTIAL METONYMY PREDICATIONAL METONYMY
Source
Schubert
Target
More generally, nobody has ever suggested, to my knowledge, that a metonymic target must not entail its source. Why non-entailment from target
to source should be a condition for metonymicity remains unclear. I do
not see any reason for introducing such a condition as a defining criterion
of metonymy.
Another problem with Hasers approach is her claim that the target of
(20), i.e. (21), adds meaning to what is said, whereas in her view a genuine
metonymy does not add meaning but substitutes one meaning for another. Consider the expression drink a glass again, which everybody regards as a metonymical expression. One condition for employing this metonymy crucially involves the use of a glass; drinking directly from the tap
or from ones cupped palms does not count as drinking a glass. One
might object that drink a bottle (of wine) does not necessarily, and even
168
Klaus-Uwe Panther
not normally, mean that the drinker drinks from a bottle. However, the
bottle as a container is still crucially involved, because if somebody drinks
directly from the wine barrel, this activity will not count as drink a bottle
even if the drinker consumes the quantity contained in a bottle. To come
back to drink a glass, the metonymic meaning can be paraphrased as
drink (approximately) the quantity of liquid contained in a glass and,
most probably, from a glass. The point is that there is not just substitution of one meaning for another but meaning elaboration of the source
concept glass (as a container).20 The source meaning is a conceptual
component of the target meaning. This is also what Panther and Thornburg assume for cases of predicational metonymy such as (20) and (21):
The source meaning expressed by (20) becomes an integrative part of the
target meaning expressed by (21). The metonymic reading induced by (20)
can be paraphrased as was able to finish his dissertation and actually
managed to finish his dissertation. The meaning component ability is
combined with the meaning component sucessful action in the metonymic target, but ability becomes a backgrounded meaning component
whereas the sucessful action sense is conceptually prominent.
In conclusion, the problems that Haser has with Panther and Thornburgs metonymy theory are based on a false presumption, viz. that these
authors regard added meanings as metonymies. Target meanings are
not additions, but elaborations of the source meanings, which remain a
conceptual part of the target meaning.
Another related point of Hasers critique concerns the relation between
metonymy and implicature. Implicature, she seems to assume, is added
conceptual material, on top of what is said by the speaker, whereas metonymy is presumably a substitution relation. I believe this to be a fictitious
opposition that collapses on closer inspection. Take a standard implicature based on the Gricean maxim of quantity such as some +> not all
(as is common, the symbol +> is used here to represent the implicature
relation). This looks like a case of addition, but the overall (including the
implicated) meaning of some could alternatively be formulated as meaning elaboration: some +> some & not all.21 Furthermore, the not all
component seems to be a conceptually prominent meaning in most cases
whereas the meaning component some is more backgrounded (but still
activated). Consider the following utterance:
(24) At midnight, some of the guests left the party.
169
In many contexts, in utterances like (24) the meaning that not all guests
left the party (and probably most did not leave it) is conceptually prominent. The conceptual link between some and not all is by many speakers
felt to be so tight that the two meanings are almost indistinguishable.
Conceptual tightness of source and target senses is also a property of metonymy.
There are thus some interesting analogies between metonymy and implicature, which have led some scholars to assume that metonymic principles actually underlie many implicatures (see e.g. Barcelona 2003:
passim; Ruiz de Mendoza Ibaez and Pea 2005: 274).
3.
In this section, I relate the theoretical discussion of metonymy in the previous sections to the more down-to-earth question of how metonymy
functions in actual discourse. The section is devoted to the interpretation
of two contextualized examples of metonymic usage in written discourse.22 In the glossary compiled by Larry Gorbet for Langackers first
volume of his pioneering Foundations of Cognitive Grammar (1987: 494) a
usage event is defined as
a symbolic expression assembled by a speaker in a particular circumstance for a
particular purpose; the pairing of a detailed, context-dependent conceptualization and (in the case of speech) an actual vocalization.
The analysis that ensues considers the particular circumstance of the utterances, their particular purpose and the discourse context in which they
are embedded.
3.1. What happened at Pearl Harbor?
The first metonymy-in-use example involves an Internet text about a commemoration ceremony (in 1999) of the Japanese attack on Pearl Harbor
in 1941, which started World War II in the Pacific. For methodological
reasons, I isolate the metonymic sentence to be analyzed looking at it first
from a null-context perspective, before re-embedding it in a larger discourse context:23
(25) A Pearl Harbor will never happen again.
170
Klaus-Uwe Panther
The leading question is: What does a metonymic inference per se convey?
On its own, the noun phrase Pearl Harbor in (25) is a straightforward
example of the metonymy place for event at place. In terms of generative grammar (25) involves a violation of a selection restriction because
the verb phrase will never happen again normally selects a subject that denotes an event. Minimally, readers/hearers with no historical knowledge
will identify a metonymic interpretation on the basis of their knowledge
that Pearl Harbor is a place; the verb phrase, which contains the verb
happen, will coerce the reading that something happened at that place.
The metonymic interpretation Something happened at Pearl Harbor is
not more than a prompt for further interpretive work. Furthermore, there
is also a metonymic inference triggered by the indefinite article a: Pearl
Harbor is conceptualized as a representative event for a whole class of
events with similar properties. Lakoff (1987: 8788) calls such metonymically based conceptualizations paragons (for an in-depth analysis of paragons, see Barcelona 2004). To summarize, (25) metonymically evokes the
following propositions:
(26) a. Some event happened at Pearl Harbor.
b. The event at Pearl Harbor is representative of other events of the
same kind.24
The full details of the Japanese attack on Pearl Harbor can only be identified on the basis of encyclopedic knowledge, which the historically informed reader might already have at her disposal (additional conceptual
components in Figure 1 above) or which is provided by the surrounding
discourse context. In (27) the discourse context, in which (25) is embedded, is already significantly larger:
(27) I return each year for this ceremony aboard Taney, said principal
speaker VADM Douglas J. Katz, USN(Ret.). We are here on the eve
of the millennium as the greatest, most powerful nation on earth.
From World War II until today, we still have the greatest military in
the world, but we need to be more alert than ever as we move into the
21st century so that a Pearl Harbor will never happen again [italics
mine] we must never, ever forget.
From (27), at least the following additional inferences about the event that
occurred at Pearl Harbor can be drawn:
171
172
Klaus-Uwe Panther
bridging inferences make it clear that the country attacked was the U.S.).
The inference in (30b), viz. that 101 American warships were destroyed in
the attack, follows from the information that Taney [] is the last floating survivor of the 101 warships [].
The preceding analysis supports my thesis that metonymies are
prompts that induce fairly general inferences, which are in need of further
inferential elaboration unless the reader has at her disposal a rich knowledge data-base that enables her to fill in the details on the basis of the metonymy alone. The question remains as to the status of the inferences
listed in (28ad) and (30ab). Are they metonymic inferences or inferences of a different kind? Some of them are definitely not metonymic as
e.g. the presuppositional inference included in (30a); others are very specific and context-driven. Some researchers have treated low-level inferences as metonymic (see Barcelona 2003), but the question as to their cognitive status has to be left open here.
3.2. Red hot doorbuster
As a second case study, I discuss in some detail the expression red hot
doorbusters found in the advertising section of a local American newspaper (see Photo 1 below). This expression exhibits a much more complex
conceptual structure than the expression Pearl Harbor. There is not only
metonymy involved in this case but metaphor as well. The importance of
context in the interpretation of a fully specified meaning and the role of
metonymy and metaphor as guiding conceptual principles towards a specific interpretation is demonstrated.
A photograph of the advertisement is found below.25 For heuristic reasons I assume a language user who is not familiar with the word doorbuster, i.e. encounters the term for the first time, but recognizes the word
as containing the free morphemes door (noun), bust (verb), and the derivational morpheme -er, which has been shown to exhibit a high degree of
polysemy (see Panther and Thornburg 2001, 2002) (see Photo 1 below).
What kind of information would help a reader who is not familiar with
the expression doorbuster to work out the intended meaning extremely
low prices for desirable products sold for a limited time or desirable
products at extremely low prices for a limited sales period? The interpretative work of the reader is first of all facilitated by the linguistic context,
in particular what follows the expression red hot doorbusters. The noun
phrase these specials has as its coreferential antecedent red hot doorbusters. The term special is well entrenched in English and is found in
173
ordinary dictionaries with the meaning a lower price than usual for a particular product for a short period of time (Longman Dictionary of Contemporary English, 2003) (LDCE), s.v. special). It is actually, for all practical purposes, synonymous with the expression red hot doorbuster.
Moreover, the restricted time period of the sale is redundantly conveyed
by the verb phrase wont last long.
In what follows, I analyze the meanings of red hot and doorbusters. Section 3.2.1 contains some reflections on the conceptual structure of red hot;
section 3.2.2 analyzes this meaning in conjunction with the metonymic
and metaphoric structure of doorbusters.
3.2.1. Red hot
The adjectival expression red hot is a well-entrenched lexical item. It has
at its syntactic head hot, which is modified by red. The compound expression is literally applied to objects or substances such as poker, metal,
and rock. The attribute red denotes the effect of extreme heat; a red hot
object or substance exhibits a glowing or shining red color. However, the
174
Klaus-Uwe Panther
meaning component shiny red often gives way to the meaning very hot,
i.e. extremely high on the heat scale in utterances like Be careful with
these plates theyre red hot (LDCE). The target meaning of red hot as a
modifier of doorbusters is to put things simply something like very desirable. I contend that this meaning comes about through metonymic
elaboration of the source meaning and metaphoric mapping. First, the
metonymic elaboration in the source domain is an inference from the
modifier red to an intensifying meaning very, highly, extremely, i.e. an
endpoint on a scale of intensity. If something is red hot (visual stimulus),
then it is easily inferred to be very hot. Second, red hot also suggests, e.g.
in the context of iron poker, a high degree of salience in terms of visibility.
These two metonymically derived meaning components are then metaphorically mapped onto the conceptual entities very desirable and
higly salient on price scale, a meaning component that is reinforced
by doorbuster. The latter actually narrows down highly salient on the
price scale to the low end of the price scale, i.e. inexpensive and
affordable.
The conceptual structure of red hot (doorbusters) is diagrammed in Figure 4. At the most general level, there is a metaphoric mapping from the
domain of sense perception to the sale domain. Both the source and the
target domain are inferentially elaborated; it is worth noting that the target domain contains a metonymic inference from cause to effect: the de-
175
sirability of the sale object is caused by the highly salient feature low
price.
The effect for cause metonymy is necessary in order to reconcile the
(target) meanings of red hot very exciting and doorbusters cheap prices
for products temporarily on sale. Cheap prices are not intrinsically exciting; they are meant to cause excitement in the customer. Thus doorbuster has an impact on the conceptual structure of the target domain of
its modifier red hot.
From a different perspective, which integrates the notion of scale (left
implicit in Figure 4), the figurative meaning of red hot as a modifier of
doorbuster can be represented as relating three scales: a heat scale (in the
source domain), a desirability scale and a price scale, the latter two inversely and causally related to each other in the target domain.26 These
scales focus on endpoints. In the source domain, the modifier red metonymically intensifies the degree of heat encoded in hot. The heat scale is
metaphorically mapped onto a desirability scale in the target domain.
High heat corresponds to high desirability. The base desire in desirability
is here understood as denoting not only a propositional attitude, but also,
in the context of frantic consumerism, a strong emotion. Temperature
words are quite systematically used in English and other languages to describe emotional states; there is thus a generic metaphor (which itself
might be metonymically motivated) for the use of red hot. In the target domain, as already noted, there is a remarkable inverse relationship between
the desirability scale and the price scale. A high value on the desirability
scale corresponds to a low value on the price scale, and vice versa. Moreover, as previously mentioned, the two scales are causally related: desirability is (at least partially) caused by the reduced cost of the sales item
and thus a strong incentive for customers to buy a product. This inverse
(causal) relationship is indicated by a dotted arrow in Figure 5 (see below).
3.2.2. (Red hot) doorbuster
In this section the leading question is: How can one make sense of doorbuster as a conceptually motivated morphological structure? Let us look
at the components (constituent morphemes) of this word and the meanings that are usually assigned to them:
176
Klaus-Uwe Panther
177
178
Klaus-Uwe Panther
hot pizza is not as literal as red hot poker because red hot pizza lacks a
fiery glow even though it may have just come out of the oven. Furthermore, the modifier red hot, combined with a noun denoting food,
strongly suggests that the food item has been freshly prepared and should
be consumed within a short time period. These attributes are paralleled in
the meaning of red hot doorbuster: freshly (recently) prepared food corresponds to items with new low prices; for immediate consumption corresponds to for immediate purchase; and in both domains red hot is associated with the scalar value highly desirable.
4.
179
been used by cognitive linguists over and over again to make important
conceptual distinctions, in particular of course, as operational criteria to
delineate metaphor from metonymy. This is not to challenge the notions
of frame or domain as such, but to urge cognitive semanticists to develop
more precise criteria for these key concepts.
Another concern of this chapter has been to demonstrate that metonymy accounts for what, to my mind, has been erroneously termed primary metaphor. Given a conception of metaphor as an iconic mapping,
i.e. a case of structural resemblance between a source domain and a target
domain, it does not make much sense to call (the verbalization of) basic
experiential correlations metaphoric. Primary metaphors look exactly
like the indexical relations that are exploited by metonymy.
A third intention of this chapter has been to defend the view of metonymy advocated here against criticism to the effect that metonymy is a
wastebasket category lumping together heterogeneous conceptual phenomena. This view rests on problematic premises of what constitutes
added meanings in contrast to substituted meanings. It also erroneously assumes that metonymic targets should not entail their sources.
Finally, I have tried to show, by means of two detailed case studies, how
metonymic (and metaphoric) meanings are worked out in context, i.e.
how metonymy functions in usage events. The aim has been to demonstrate that discourse context, as discussed in connection with Figure 1, is
absolutely crucial in the interpretation of newly encountered expressions
like Pearl Harbor (as a historic event) and red hot doorbusters (as special
sale items/prices). Preexisting metonymic principles will guide the hearer
or reader in their interpretive efforts, but only the context will allow the
hearer to draw the right inferences leading to a sufficiently constrained
reading. Knowledge about the semantics of related constructions provides additional input for a plausible interpretation.
Notes
1.
180
Klaus-Uwe Panther
181
13. See e.g. the analysis proposed for Wall Street by Ruiz de Mendoza Ibez and
Dez Velasco (2002: 513) and Ruiz de Mendoza Ibez and Otal Campo
(2002: 65), where it is assumed that the location Wall Street is the matrix domain for an institution, which itself constitutes the matrix domain for people
(working in the institution).
14. The kind of negation that is relevant for semantic presupposition is frame-internal negation (see Fillmore 1982).
15. This conclusion supersedes the definition of metonymy provided most recently in Panther (2005) and Panther and Thornburg (2003).
16. As an aside, it should be noted that for some of these examples it is hard to see
how they involve a correlation between a sensorimotor operation and subjective experience or judgment. For example, closeness, as a spatial relation, does
not seem to be describable as a sensorimotor operation.
17. This example I have taken from Brdar and Brdar-Szab (2003: 241). The metonymy where the manner of action stands for the action itself has been explored in some detail from a cross-linguistic perspective by the same authors
(Brdar-Szab and Brdar 2002).
18. I now think that the metonymic relation is better described as obtaining between be able to (source meaning) and manage to (target meaning), but this
change is irrelevant to the argument made by Haser.
19. Whether there is a separate relation of semantic presupposition that is distinct
from entailment is controversially discussed in the literature. For overviews,
see Kempson (1975) and Wilson (1975).
20. Note that glass also involves a metonymically exploited relation between substance and container made of substance.
21. The idea that metonymy involves integration of the source meaning into the
target meaning has also been put forward by Radden and Kvecses (2006).
22. A usage event comes very close to what in anthropological linguistics is referred to as a speech event and in pragmatics as a speech act.
23. The text from which examples (25), (27) and (29) are taken is available at:
http://www.uscg.mil/reserve/magazine/mag1999/dec1999/Taney.htm. Retrieved May 21, 2006.
24. Note that I analyze a Pearl Harbor in way analogous to the case of aspirin
(Figure 2). The focus is not on the class of events as such but on the any individual event that is like Pearl Harbor in significant respects.
25. By necessity Photo 1 is in black and white. In the original ad the words red
hot doorbusters are in orange against a red field; the words 4 hours only!
are against a blue field.
26. This alternative analysis as well as Figures 5 and 6 were developed with Linda
Thornburg. Much of section 3.2 would not have been possible without her
competence as a native speaker of English and her expertise as a linguist. I am
grateful to her for inspiring ideas during the writing of this article.
27. This definition is proposed as the basic sense of -er nominals in Panther and
Thornburg (2001, 2002).
28. Alternatively one might also assume a hyperbolic interpretation, based on a
182
Klaus-Uwe Panther
metonymic shift from a lower value on a behavioral scale, i.e. a relatively civilized way of rushing into a store, to a higher scalar value, i.e. stampede-like behavior that does not shy away from rudeness, aggression and physical force.
References
Barcelona, Antonio (ed.)
2000
Metaphor and Metonymy at the Crossroads. (Topics in English Linguistics 30.) Berlin/New York: Mouton de Gruyter.
Barcelona, Antonio
2002
Clarifying and applying the notions of metaphor and metonymy within
cognitive linguistics: an update. In Metaphor and Metonymy in Comparison and Contrast, Ren Dirven, and Ralf Phrings (eds.), 207277.
(Cognitive Linguistics Research 20.) Berlin/New York: Mouton de
Gruyter.
2003
The case for a metonymic basis for pragmatic inferencing: Evidence from
jokes and funny anecdotes. In Metonymy and Pragmatic Inferencing,
Klaus-Uwe Panther, and Linda L. Thornburg (eds.), 81102. (Pragmatics & Beyond New Series 113.) Amsterdam/Philadelphia: Benjamins.
2004
Metonymy behind grammar: The motivation of the seemingly irregular grammatical behavior of English paragon names. In Studies in Linguistic Motivation, Gnter Radden, and Klaus-Uwe Panther (eds.),
357374. (Cognitive Linguistics Research 28.) Berlin/New York: Mouton de Gruyter.
Bergman, Mats, and Sami Paavola (eds.)
2003 ff. The Commens Dictionary of Peirces Terms: Peirces Terminology in His
Own Words. Helsinki: University of Helsinki. [Available at: http://www.
helsinki.fi/science/commens/terms/index2.html. Retrieved April 23,
2006.]
Bierwiaczonek, Bogusaw
2005
On the neural and conceptual basis of semantic relations. In MetonymyMetaphor Collage, Elzbieta Grska, and Gnter Radden (eds.),
1136. Warsaw: Warsaw University Press.
Brdar, Mario, and Rita Brdar-Szab
2003
Metonymic coding of linguistic action in English, Croatian and Hungarian. In Metonymy and Pragmatic Inferencing, Klaus-Uwe Panther,
and Linda L. Thornburg (eds.), 241266. (Pragmatics & Beyond New
Series 113.) Amsterdam/Philadelphia: Benjamins.
Brdar-Szab, Rita and Mario Brdar
2002
Manner-for-activity metonymy in cross-linguistic perspective. In Cognitive Linguistics Today, Barbara Lewandowska-Tomaszczyk, and K.
Turewicz (eds.), 235257. Frankfort am Main: Peter Lang.
Buchler, Justus (ed.)
1955
Philosophical Writings of Peirce. New York: Dover.
183
Croft, William
2002
The role of domains in the interpretation of metaphors and metonymies. In Metaphor and Metonymy in Comparison and Contrast, Ren
Dirven, and Ralf Phrings (eds.), 161205. (Cognitive Linguistics Research 20.) Berlin/New York: Mouton de Gruyter.
Cruse, D. Alan
2000
Meaning in Language: An Introduction to Semantics and Pragmatics.
Oxford: Oxford University Press.
Dirven, Ren, and Ralf Prings (eds.)
2002
Metaphor and Metonymy in Comparison and Contrast. (Cognitive Linguistics Research 20). Berlin/New York: Mouton de Gruyter.
Fauconnier, Gilles, and Mark Turner
2002
The Way We Think: Conceptual Blending and the Minds Hidden Complexities. New York: Basic Books.
Fillmore, Charles
1982
Frame semantics. In Linguistics in the Morning Calm, The Linguistic
Society of Korea (ed.), 111137. Seoul: Hanshin.
Geeraerts, Dirk
2005
Lectal variation and empirical data in Cognitive Linguistics. In Cognitive Linguistics: Internal Dynamics and Interdisciplinary Interaction,
Francisco J. Ruiz de Mendoza Ibez, and M. Sandra Pea Cervel
(eds.), 163189. (Cognitive Linguistics Research 32.) Berlin/New York:
Mouton de Gruyter.
Gibbs, Raymond W., Jr.
1994
The Poetics of Mind: Figurative Thought, Language, and Understanding.
Cambridge: Cambridge University Press.
Grady, Joseph E.
1998
The conduit metaphor revisited: A reassessment of metaphors for communication. In Conceptual Structure, Discourse and Language 2, JeanPierre Koenig (ed.), 205218. Stanford: CSLI.
Haser, Verena
2005
Metaphor, Metonymy, and Experientialist Philosophy: Challenging
Cognitive Semantics. (Topics in English Linguistics 49.) Berlin/New
York: Mouton de Gruyter.
Hiraga, Masako K.
1998
Metaphor-icon link in poetic texts: A cognitive approach to iconicity.
The University of the Open Air 16: 9523. [Available at: http://www.conknet.com/~mmagnus/SSArticles/hiraga/hiraga.html. Retrieved May 26,
2006]
2005
Metaphor and Iconicity: A Cognitive Approach to Analyzing Texts. New
York: Palgrave Macmillan.
Jakobson, Roman
2002
The metaphoric and metonymic poles. In Metaphor and Metonymy in
Comparison and Contrast, Dirven, Ren, and Ralf Prings (eds.),
4147. (Cognitive Linguistics Research 20.) Berlin/New York: Mouton
de Gruyter.
184
Klaus-Uwe Panther
Kempson, Ruth M.
1975
Presupposition and the Delimitation of Semantics. Cambridge: Cambridge University Press.
Lakoff, George
1987
Women, Fire and Dangerous Things: What Categories Reveal About the
Mind. Chicago: The University of Chicago Press.
Lakoff, George, and Mark Johnson
1980
Metaphors We Live By. Chicago: The University of Chicago Press.
1999
Philosophy in the Flesh: The Embodied Mind and Its Challenge to Western Thought. New York: Basic Books.
Langacker, Ronald W.
1987
Foundations of Cognitive Grammar. Volume 1: Theoretical Prerequisites.
Stanford, CA: Stanford University Press.
2000
Grammar and Conceptualization. (Cognitive Linguistics Research 14.)
Berlin/New York: Mouton de Gruyter.
Leezenberg, Michiel
2001
Context of Metaphor. Amsterdam: Elsevier.
Levinson, Stephen C.
2000
Presumptive Meanings: The Theory of Generalized Conversational Implicature. Cambridge, MA: MIT Press
Panther, Klaus-Uwe
2005
The role of conceptual metonymy in meaning construction. In Cognitive Linguistics: Internal Dynamics and Interdisciplinary Interaction,
Francisco J. Ruiz de Mendoza Ibez, and M. Sandra Pea Cervel
(eds.), 353386. (Cognitive Linguistics Research 32.) Berlin/New York:
Mouton de Gruyter.
Panther, Klaus-Uwe, and Linda L. Thornburg
1998
A cognitive approach to inferencing in conversation. Journal of Pragmatics 30: 755769.
1999
The potentiality for actuality metonymy in English and Hungarian. In Metonymy in Language and Thought, Klaus-Uwe Panther, and
Gnter Radden (eds.), 333357. (Human Cognitive Processing 4.) Amsterdam/Philadelphia: Benjamins.
2001
A conceptual analysis of English -er nominals. In Applied Cognitive
Linguistics II: Language Pedagogy, Martin Ptz, Susanne Niemeier,
and Ren Dirven (eds.), 149200. (Cognitive Linguistic Research 19.2.)
Berlin/New York: Mouton de Gruyter.
2002
The roles of metaphor and metonymy in English -er nominals. In Metaphor and Metonymy in Comparison and Contrast, Ren Dirven, and
Ralf Prings (eds.), 279319. (Cognitive Linguistics Research 20.) Berlin/New York: Mouton de Gruyter.
2003
Introduction: On the nature of conceptual metonymy. In Metonymy
and Pragmatic Inferencing, Klaus-Uwe Panther, and Linda L. Thornburg (eds.), 120. (Pragmatics & Beyond New Series 113.) Amsterdam/
Philadelphia: Benjamins.
185
186
Klaus-Uwe Panther
187
Abstract
Cultural models are idealized cognitive models of sociocultural phenomena. In
this chapter, I suggest that the utility of cultural models lies in the ability of individual culture members to adapt overly simplistic models to our ever-changing
social world via conceptual blending, a set of cognitive operations for combining
concepts from different domains. A number of rhetorically motivated examples
are analyzed to show how speakers use conceptual blending to integrate concepts
with different affective valences, often so that the desired course of action is seen
as consistent with their audiences value system. Processes of conceptual blending are argued to mediate the exploitation of stable conventional mapping
schemes in order to adapt shared cultural models to the idiosyncratic needs of individuals.
Keywords: compression; conceptual blending theory; conceptual metaphor theory; cultural models; mental spaces; rhetoric
1.
Introduction
188
Seana Coulson
189
the idiosyncratic meanings speakers derive in situated instances of language use, between public, idealized meaning of cultural models and their
private meanings for individual culture members.
2.
Conceptual blending
Conceptual blending theory offers a general model of meaning construction in which a small set of partially compositional processes operate in
analogy, metaphor, counterfactuals, and many other semantic and pragmatic phenomena (Coulson and Oakley 2000, 2006; Fauconnier and
Turner 1998, 2002). In this theory, understanding meaning involves the
construction of blended cognitive models that include some structure
from multiple input models, as well as emergent structure that arises
through the processes of blending. Discussed at length in Fauconnier and
Turner (2002), blending theory describes a set of principles for combining
dynamic cognitive models in a network of mental spaces (Fauconnier
1994), or partitions of speakers referential representations.
Mental spaces contain partial representations of the entities and relationships in any given scenario as perceived, imagined, remembered, or
otherwise understood by a speaker. Mental spaces (Fauconnier 1994) can
be thought of as temporary containers for relevant information about a
particular domain. Spaces are structured by elements that represent each
of the discourse entities, and simple frames to represent the relationships
that exist between them. Frames are hierarchically structured attributevalue pairs that can either be integrated with perceptual information, or
used to activate generic knowledge about people and objects assumed by
default. Cultural models are a special kind of frame that deals with socially
relevant topics. Finally, mappings are abstract correspondences between elements and relations in different spaces. For example, in a tale about the
life of a butterfly, we might have one mental space to represent a time in the
past, and one for the present. We could then posit two elements, b in the
Past Space, and b in the Present Space. Elements b and b would be linked
via an identity mapping to represent the fact that they are the same individual. However, the properties of the Past b would be those of a caterpillar, while the properties of Present b would be those of a butterfly.
A development of mental space theory, conceptual blending is intended
to account for cases in which the content of two or more mental spaces is
combined to yield novel inferences. Central to conceptual blending theory
is the notion of the conceptual integration network, an array of mental
190
Seana Coulson
spaces in which the processes of conceptual blending unfold. These networks consist of two or more input spaces structured by information from
discrete cognitive domains, a generic space that contains structure common to the inputs, and a blended space that contains selected aspects of
structure from each input space along with any emergent structure that
arises out of the imaginative processes of blending. The first process is
called composition, and involves the juxtaposition of information from
different spaces; completion, as in pattern completion, occurs when part
of a cognitive model is activated and results in the activation of the rest of
the frame. Finally, elaboration is an extended version of completion that
results from mental simulation, or various sorts of physical and social interaction with the world as construed with blended concepts.
Blending involves the activation of cognitive models in integration networks, the establishment of partial mappings between models in the networks different spaces, and the projection of conceptual structure from
space to space. Take, for example, the blended concept of a snowflake
kid, as described in a BreakPoint Online article, published on a pro-life
Christian associations website (Morse 2005). Snowflake kids are children
conceived through in vitro fertilization and subsequently implanted in unrelated adoptive mothers with the help of organizations such as the
Snowflake Frozen Embryo Adoption agency. This agency allows infertile
couples the opportunity to be implanted with embryos leftover from fertility clinics, and serves to rebut the case of biologists who advocate the
use of such embryos for scientific research. Debate over the legitimacy of
research with human embryos is quite fierce in the United States due to a
lack of consensus over their moral status.
Table 1. Spaces in the Snowflake Kid blend
Snowflake
Space
Past
Space
Blended
Space
Present
Space
Snowflake(s)
Embryo(e)
SnowflakeKid(sk)
Kid(k)
Frozen(s)
Frozen(e)
Frozen(sk)
Unique(s)
Unique(e)
Unique(sk)
Unique(k)
Organism(e)
Person(sk)
Person(k)
As described by Coulson and Pascual (2006), the snowflake kid concept is a blend with three input spaces, the domain of snowflakes, the past,
and the present. Each column in the table below shows the information
191
192
Seana Coulson
3.
Conceptual blending theory assumes many of the same claims as conceptual metaphor theory, such as the idea that metaphor is a conceptual as
well as a linguistic phenomenon, and that it involves the systematic projection of language, imagery, and inferential structure between domains.
The two approaches are largely complementary, with similar assumptions
about the relationship between language and conceptualization, but with
different emphases and foci (see Grady, Oakley, and Coulson 1999). For
example, in contrast to the emphasis on conventional metaphors in conceptual metaphor theory, conceptual blending theory is intended to capture spontaneous, real-time processes that can yield short-lived and novel
conceptualizations. Further, blending theory reveals connections between
the cognitive underpinnings of metaphor and a variety of other linguistic
phenomena handled by mental space theory (conditionals, counterfactuals, metonymy, etc.), making it especially amenable to the characterization of rhetorical discourse.
Fauconnier and Turner (2002) propose that metaphoric utterances are
represented in conceptual integration networks in which the source and
target domain each structure one input space, the generic space represents
abstract commonalities in the inputs, and the blended space inherits
structure from its inputs as well as containing emergent structure of its
own. Rather than emphasizing the extent to which metaphorical utterances instantiate entrenched mappings between source and target domains, conceptual integration networks only represent those cognitive
models that are particularly relevant to the mapping supported by the utterances. While mappings in the integration network require knowledge
of conceptual metaphors such as those studied by Lakoff and colleagues
(e.g. Lakoff and Johnson 1980), blending theory is best suited for representing the joint influence of input domains and the origin of emergent inferences in particular metaphoric utterances.
One motivation for blending theory is the observation that metaphoric
expressions often have novel implications that cannot be traced back to
either the source or the target domain. For example, That surgeon is a
butcher, customarily connotes the surgeons incompetence, in spite of the
193
Surgeon Input
Space
Blended Space
Butcher Input
Space
Generic Space
Surgeon(s)
Surgeon(s/b)
Butcher(b)
Agent
Patient(p)
Patient(p/a)
Animal(a)
Patient
Scalpel(k)
Blade(k/c)
Cleaver(c)
Cutting-Instrument
Goal:
Heal Patient
Goal:
Heal Patient
Goal:
Kill Animal
Means:
Precise Cuts
Means:
Slashing Cuts
Means:
Slashing Cuts
194
Seana Coulson
butcher of Srebenica that recruit structure and imagery from the butchery domain, but do not connote incompetence. Differences in the implications of the butcher metaphor in the domains of medicine and the
military highlight the need for an account of their underlying conceptual
origin.
Consequently, blending theory can address the meaning construction
in metaphoric expressions that do not employ conventionalized mapping
schemes. For example, the italicized portion of this excerpt from an interview with philosopher Daniel Dennet involves a metaphorical blend,
Theres not a thing thats magical about a computer. One of the most brilliant things about a computer is that theres nothing up its sleeve, (Edge 94,
November 19, 2001). The input domains here are Computers and Magicians, and the blend involves a hybrid model in which the computer is a
magician. However, the connection between these two domains arises
purely from the co-text of this example, as there is no conventional COMPUTERS ARE MAGICIANS mapping in English.
Blending can also be used to explain how a number of different kinds of
mappings can be combined to explain the meaning of a particular
example such as the following (from Grady, Oakley, and Coulson 1999):
With Trent Lott as the Senate Majority Leader, and Gingrich at the helm in the
House, the list to the Right could destabilize the entire Ship of State.
195
overall political orientation of government. Further, the standard association between conservatism and the right as against liberalism and the left
is clearly not based on the ship model, as it is frequently encountered in
other contexts. However, because the scenario in the blend involves spatial motion, the literal notion of rightward movement is integrated with
the other structure in the blend to yield a cognitive model of a ship piloted
by New Gingrich that lists to the right. This example thus demonstrates
that while the linguistic system provides speakers with a set of coherent
mappings between the source and target domains, speakers need not
maintain this coherence in the models set up during the comprehension of
utterances that exploit these conventionalized mappings.
4.
Humorous blends
Blending is common in social and political humor and often involves the
projection of people into new contexts where the humorists point can be
clearly illustrated. Humorous examples highlight the human ability to derive meaningful information from partial, non-systematic correspondences in structure, and even to exploit accidental characteristics of input
frames. As Freud noted, joking provides a relatively safe arena for expressing aggressive, insulting, or otherwise socially unacceptable utterances. Blending and the cognitive abilities that support it are crucial in
this respect by enabling us to frame taboo topics in terms and domains
that are not taboo.
For example, a 1997 cartoon by Chip Bok points up a contrast in thenPresident Bill Clintons alleged penchant for dishonesty and that attributed to his 18th century political counterpart George Washington. On
the left-hand side of the cartoon, George Washington says, I cannot tell
a lie. Addressing Washington from the right-hand side of the cartoon,
Clinton says, If everyones on record denying it youve got no problem.
In some ways, Boks cartoon is similar to a blend discussed by Fauconnier
and Turner (1996) in which a modern-day philosopher engages in an imagined debate with Immanuel Kant. In the Kant blend, the professor projects Kant into the modern era so that he can get the esteemed Germans
reactions to ideas composed after his death.
In Boks cartoon, however, both Washingtons and Clintons lines were
allegedly uttered by the parties in question. Washingtons lines come
straight from the historical record, while Clintons were attributed to him
by his former mistress, Gennifer Flowers. As for Washington, legend has
196
Seana Coulson
it that when he was a boy, he chopped down a cherry tree on his fathers
farm. When Washingtons father discovered what had happened, he went,
furiously, to his family and demanded to know who had chopped down
the tree. Knowing that he would likely receive a spanking for his efforts,
Washington stood up and said, I cannot tell a lie. It was I who chopped
down the cherry tree.
On the other hand, Flowers attributed the remark in the right-hand
panel of the cartoon to Clinton in her declaration in the Paula Jones harassment suit against him. Worried about whether people would perceive
her relationship to then-Governor Clinton as contributing to her successful career in Arkansas state government, Flowers approached him to discuss whether she should admit to their adulterous affair. Flowers declaration suggests Clinton told her to deny it, and that he said, If everyones
on record as denying it, youve got no problem.
The composition of the two mens utterances in the blend results in the
activation of a conversation frame and provides a context in which
George Washington and Bill Clinton can interact. The mere juxtaposition of the two statements points up a contrast between the two presidents: Washington claimed to be incapable of dishonesty, while Bok portrays Clinton as someone all too willing to lie. Moreover, knowledge that
Clintons remark was originally addressed to Gennifer Flowers allows
the reader to set up a mapping between Washington in the blend, and
Flowers in the Clinton input. This makes Clinton look bad because it reinforces the idea that hes told people to lie in the past, and, further, it
suggests that he would attempt to corrupt someone as upstanding as
George Washington.
Moreover, it is only in the blend where Clinton addresses Washington
rather than Flowers that Clintons remark can evoke a misunderstanding of Washingtons cherry tree. While I cannot tell a lie, was presumably intended to mean that the guilt Washington would experience from
lying precluded him from doing so, Clintons remark suggests a different
interpretation. Rather than guilt over lying, Clinton believes Washington
fears reprisal for being caught. By offering Washington advice on how to
lie without getting caught, in the cartoon Clinton presents himself as
someone likely to behave (and likely to have behaved) in accordance with
his own advice. Consequently, it prompts retrospective projections to
Clintons input space, framing denials of the real Clintons misdeeds as
fabricated, and further framing him as untrustworthy. Though the cartoon is probably motivated by the disanalogy between George Washingtons reputation for honesty and that of Bill Clinton, the interactive frame
197
set up in the blend provides a context in which unique structure can arise
(see Coulson 2003).
Indeed, conceptual blending theory is particularly well-suited for the
analysis of persuasive efforts that involve the combination of a number of
different kinds of mappings. This is frequently the case in political cartoons which directly represent the contents of a blended space and invite
the viewer to unpack it into its inputs. Coulson (2005) describes a cartoon
by Blake Carlson that appeared in The Arizona Republic in November
2001 around the time of the American Thanksgiving holiday in which it
is customary to have a turkey dinner. In the cartoon, American president
George W. Bush stands at the head of a dinner table about to carve a turkey labeled Afghanistan. Seated at the table are a number of men in
turbans with name-cards that read the names of Afghani ethnic groups
(e.g. Uzbeks, Tajiks, Pashtuns, Hazara). Carving knife in hand, Bush says
Look, Im doin my best here, but there are only two of them. You cant
ALL have a drumstick!
The cartoon presents the viewer with the blend and invites her to unpack the input spaces in the network. The inputs in this case are Thanksgiving dinner and the Taliban overthrow in Afghanistan. Part of understanding this blend involves appreciating the mappings between Dad
carving the turkey on Thanksgiving, Bush carving the turkey in the cartoon, and Bush dividing the control of Afghanistan. Children compete
over turkey legs in the Thanksgiving input, the Afghanis compete over
turkey legs in the blended space, and (many more) Afghanis compete over
control of Afghanistan in the Taliban overthrow input. By having a single
Afghani metonymically represent each ethnic group (the Uzbeks, Tajiks,
etc.), the cartoonist here employs what Fauconnier and Turner (2002)
refer to as compression to human scale2. In this way complex sociopolitical struggles over land rights are understood (and potentially misunderstood) by analogy to everyday experience with dinner table disputes.
One interesting facet of blending in political cartoons is that some of
the actors in the blend play themselves. While some of the mappings between the blend and the Taliban overthrow space involve analogy mappings (between the turkey and Afghanistan), others involve identity mappings (Bush maps to himself). Presumably, by restricting the need to make
an analogical inference to one key concept (in this case, competition over
a limited resource) and utilizing identity mappings between the actors in
the cartoon and the actors in the political arena (i.e. having Bush carve the
turkey in the cartoon), the cartoonist facilitates appreciation of the cartoon. Political cartoons and rhetorically motivated discourse prompt us
198
Seana Coulson
5.
199
Persuasive absurdity
In fact, the conceptual integration of two scenarios into an absurd scenario in a blended space is a common argumentative tactic. Take the
example below, from an actual jury deliberation in a high-profile murder
case, videotaped and partly broadcast by an American television station:
Juror 7: Barbara Davis [victim] doesnt get to come when she turns 72 in that
casket she doesnt get to come out of that casket.
Juror 11: Yeah, but theres not a guarantee that technically hell get out either. []
Juror 13: If she has no chance of getting out of the casket why should he have a
chance of getting out of jail?
Juror 2: Exactly. (JurDel.B., p. 22)
Coulson and Pascual (2006) argue that, in this blend, the victims casket is
construed as a prison from which there is no possibility of release. The
disanalogy between the victims future and the Hypothetical Future of
her murderer at age 72 is used to argue for giving the murderer a life sentence. In this blend, the victim has the animate and intentional properties
she had in the Past Space, before the crime occurred. Further, in spite of
her death, the victim ages in an analogous fashion to the prisoner. However, unlike the prisoner, when the victim turns 72 she does not get to
come out of the casket. Thus the middle-aged murder victim is construed
in the blend as being 72 years old, and actively desiring to get out of her
casket. However, she is unable to do so because she is dead.
Unlike the examples in the previous section, the unreal image evoked by
this blend was not presented for embellishment or humorous purposes.
On the contrary, the aging corpse in Juror 7s blend was used to argue for
the appropriateness of a life sentence for the convicted murderer. In spite
of the absurdity of the image in the blend, or indeed perhaps because of it,
Juror 7s argument is compelling. Interestingly, the speakers goal in this
example was not to draw attention to the absurdities in the blended space,
but rather to the inferences and conceptualizations that emerge from it.
That is, the speaker is not attempting to argue for the absurdity of a dead
individual becoming older and trying to escape from her coffin on the day
her murderer is released from prison. Rather, the intended absurdity is the
idea of giving a more severe punishment to the victim than to the man
who killed her.
Examining other instances of blends produced by speakers under the
demands of real-time conversation suggests the comprehension and production of these blends is supported by extensive appeal to conventional
200
Seana Coulson
201
202
Seana Coulson
In fact, the full meaning of Clarkes remarks is not apparent until the
succeeding discourse where he points out that the CIA observed a meeting of high-level Al Quaeda operatives where the attendees were assumed
to be planning an attack, and that two of these people were known to have
entered the United States prior to September 11, 2001. His point is that
these facts alone constituted actionable intelligence and thus the analogy
to the game of connect the dots is not well supported. Because Clarke
objects to the very premise of the mappings in the connect-the-dots blend,
he explicitly disputes the applicability of the term connect, recruiting
instead the conventionalized fictive interaction blend. While fictive interaction is compatible with Clarkes construal of certain facts as being independently meaningful, it is less compatible with the mapping between
facts and dots. Moreover, the impact of additional information about the
target domain on the interpretability of the phrase screaming dots suggests that what is crucial for the argument is not the structure in the
blended space, but rather the mappings between blended structures and
concepts in the other spaces in the network.
Indeed, analysis of persuasive texts intended to actually change
peoples behavior ultimately depends on the audiences internalization of
the relevant cultural models. For example, Coulson and Oakley (2006)
discuss an example of a (real) direct mail advertisement soliciting contributions to a religious organization. The package includes a letter, a
prayer page to send with donations, a return envelope for the prayer page,
and a purple sealed envelope bearing a message from Jesus Christ. The
letter urges its recipient to perform a number of concrete actions in order
to show her faith, and be blessed by Jesus. In particular the reader is instructed to:
1. Place the purple sealed envelope under his or her pillow
2. Sleep on this purple point of contact just like the children of Israel did
when God instructed them to do so (Numbers 15:38,39)
3. Mail back the prayer page with a donation to the Ministry.
4. Open the purple sealed envelope to receive the purple point of contact
blessing.
As one aspect of its persuasive effort, the letter invokes a blend of conceptual structures from cognitive domains associated with organized religion. For example, the letter repeatedly appeals to a metaphoric construal of making a monetary donation as sowing a seed. For example,
towards the end of the letter proper, we read:
203
We believe you are going to sow a seed so God can bless you with a harvest. God
said, Give and it shall be given unto you Luke 6:38. We pray that you will
sow $5.00, $10.00, $20.00, or more. Let God lead you. Our prayer is that, by
faith, what you sow will start being returned to you before the seventh day of
next month, as God sees fit. He knows best how and when to let it begin. Let us
pray over this last page and purple sealed word. Let us bow our heads in prayer
shall we?
204
Seana Coulson
by faith. Because it is a seed of faith, the coming harvest depends on receiving a blessing from the Lord. Moreover, receiving the blessing depends on following a particular series of instructions outlined in the letter:
mailing in the donation, sleeping on the (enclosed) purple envelope, and
opening the purple envelope after sunset on the following day.
Coulson and Oakley (2006) argue that the desired rhetorical effect of
this letter depends on the existence of systematic correspondences between
the three input spaces displayed in the table. Besides conventional agricultural metaphors for investment (e.g. investments that grow), the letter
authors are exploiting conventional agricultural metaphors for spirituality
(e.g. spiritual growth). The former play into the readers greed, while the
latter are reminiscent of the Bible and bolster the legitimacy of the religious organization that mailed the letter.3 The integration of these three domains results in a scenario where the reader can satisfy her greed in a virtuous way. Thus the inputs in this blend are being exploited not only for
their inferential possibilities, but also for their sociocultural significance.
Table 3. Input spaces in faith sowing blend
Material Space
Spiritual Space
Agricultural Space
mail $5
make an offering
sow a seed
sleep on an envelope
cultivate a seed
receive money
receive blessing
reap harvest
While the letter clearly establishes the mappings between sending the
money, sowing a seed, and making an offering to God, establishing the
blend goes further. Without the blend, there is no way that anyone would
believe that sending off $5 could result in financial gain. Similarly, the
reader will not carry around the purple envelope or sleep on it unless she
believes the action will have the spiritual and/or the monetary results implied in the blend. So, to reiterate, anyone who performs the actions described in the letter will do so because they have adopted the blend where
mailing $5 is sowing a faith seed, sleeping on the envelope is an act of
faith, and that the ultimate result of these actions will be a monetary
blessing from God. Moreover, the difference between someone who does
and someone who does not carry out the instructions has little to do with
the mappings (presumably anyone can figure out what one is supposed to
do and why), and everything to do with integrating and elaborating the
structure in the blend until it becomes a motivating frame.
6.
205
Conclusion
The examples above show how speakers use conceptual blending to integrate concepts with different affective valences, often so that the desired
course of action is seen as consistent with their audiences value system.
We have also seen how compression is used to simplify complex causal relationships, both so they can be more readily understood, and so that they
can be construed with motivational human scale frames. Persuasive discourse in particular demonstrates the fact that our concepts have affective
and motivational properties as well as abstract, inferential ones. Moreover, neither is set in stone as speakers frequently employ conceptual
blending processes to reconstrue a particular action to alter its inferential,
affective, sociocultural, and even spiritual significance.
Be they serious or playful, communicative acts between humans inevitably involve a balancing act between the static, shared set of idealized
cultural models and the more dynamic needs of individuals faced with the
realities of a social world that is considerably more complex. Because our
construals of particular current events derive their social significance
from the cultural models they evoke, even humorous framings serve to reinforce the status of those models as interpretive resources, and create social pressure to conform at least somewhat, and in certain restricted circumstances, to the guidelines laid out in those models.
Political jokes and cartoons can thus be seen as part of the larger social
negotiation of the proper application of cultural models. By projecting
prominent personalities into new contexts, cartoons promote particular
construals of events and well-known people. They reinforce the availability of cultural models, and perhaps even police their use. Moreover, in exploiting the fortuitous structure that arises in blended spaces, humorous
examples allow us to test the flexibility of our conceptual system, to navigate the space of possible construals, and to explore the radically different
social and emotional consequences they can trigger.
However, for all their importance, the frames and cultural models
evoked in discourse do not have a deterministic impact on either our behavior or our construals. People construct blends that are likely to be persuasive, given certain assumptions about how frames and cultural models
affect behavior. However, the success of such efforts depends on the extent
to which the individual accepts the validity of the model both in general
and in its particular application. Moreover, people do not respond to each
others rhetorical efforts reflexively, but rather agentively with attempts to
reframe evoked models in ways more consistent with their own rhetorical
206
Seana Coulson
goals (see Coulson 2001 for examples). As Shore (1996: 372) writes, processes by which cultural models are brought to mind are activities of an
active, intentional, and opportunistic intelligence, not a passive recording
device.
The difference between blends we act on and those we do not depends
as much on the ontology supported by our cultural values and practices
as it does on the structural correspondences between the representations
in the different domains. For example, the success of rhetorical efforts to
reify a blend like sowing a faith seed will depend in a complex way on the
character of their appeal to social roles and previously established cultural practices as well as the extent to which the audience has internalized
cultural values. While conceptual integration accounts for some of the
mental operations necessary to incite action, the roots of action extend
beyond the individuals nervous system as conceptual blends are intimately intertwined with human doings. In this sense the reason of conceptual blending is indeed the slave to the passion of deeply felt motivational frames.
We have seen throughout how processes of conceptual blending mediate
the exploitation of stable conventional mapping schemes and adapt
shared cultural models to the idiosyncratic needs of individuals.
Notes
1. Note that there is no particular theoretical significance to the use of tables as
opposed to circles (as in Fauconnier and Turner 2002) to outline conceptual
integration networks. They are notational variants. Many people find tables
easier to produce and comprehend, however.
2. Conceptual metonymy is a common and theoretically important example of
the more general phenomenon of compression (see Fauconnier and Turner
2002). See Barcelona (2003) for a discussion of the role of conceptual metonymy in conceptual blending.
3. Indeed, the New Testament contains several different versions of the parable of
the Sower, about a sower who sows seed on a path, rocks, thorns, and good
dirt and only the latter grew into plants that yielded fruit. One of the few
parables in the Bible that is actually followed by exegetical notes from Jesus,
the parable is typically interpreted as being about Jesus (the Sower), the gospel,
or Gods message (the seeds), and the inability and/or unwillingness of many
people to understand Gods message. See Wierzbicka (2001: 257265) for insightful analysis of this parable. Despite the disanalogies between our letters
appropriation of this metaphor and the parable of the Sower (e.g. Jesus the
Sower of Gods message is presumably not a good mapping for the sinner this
207
References
Barcelona, Antonio
2003
The case for a metonymic basis of pragmatic inferencing. In Metonymy
and Pragmatic Inferencing, Klaus-Uwe Panther, and Linda L. Thornburg (eds.), 81104. Amsterdam: Benjamins.
Coulson, Seana
2001
Semantic Leaps: Frame-shifting and Conceptual Blending in Meaning
Construction. Cambridge: Cambridge University Press.
2003
Reasoning and rhetoric: Conceptual blending in political and religious
rhetoric. In Research and Scholarship in Integration Processes, Elzbieta
Oleksy, and Barbara Lewandowska-Tomaszczyk (eds.), 5988. Lodz,
Poland: Lodz University Press.
2005
Whats so funny? Cognitive semantics and jokes. Cognitive Psychopathology/Psicopatologia Cognitive 2(3): 6778.
Coulson, Seana, and Todd Oakley
2000
Blending basics. Cognitive Linguistics 11(3): 175196.
2006
Purple persuasion: Conceptual blending and deliberative rhetoric. In
Cognitive Linguistics: Investigations across Languages, Fields, and
Philosophical Boundaries, June Luchjenbroers (ed.), 4765. Amsterdam: Benjamins.
Coulson, Seana, and Esther Pascual
2006, For the sake of argument: Mourning the unborn and reviving the dead
in press through conceptual blending. Annual Review of Cognitive Linguistics 3.
Fauconnier, Gilles
1994
Mental Spaces: Aspects of Meaning Construction in Natural Language.
Cambridge/New York: Cambridge University Press.
Fauconnier, Gilles, and Mark Turner
1994
Conceptual projection and middle spaces. La Jolla, CA: Cognitive
science technical report 9401.
1998
Conceptual integration networks. Cognitive Science 2(1): 133187.
2000
Compression and global insight. In (Special issue: Conceptual Blending, Seana Coulson, and Ted Oakley (guest eds.),. Cognitive Linguistics
11(3): 283304.
2002
The Way We Think: Conceptual Blending and the Minds Hidden Complexities. New York: Basic Books.
Grady, Joseph, Todd Oakley, and Seana Coulson
1999
Conceptual blending and metaphor. In Metaphor in Cognitive Linguistics, Raymond W. Gibbs, and Gerard J. Steen (eds.), 101124. Amsterdam/Philadelphia: Benjamins.
208
Seana Coulson
Lakoff, George
1993
The contemporary theory of metaphor. In Metaphor and Thought, Ortony, Andrew (ed.), 202251. Cambridge: Cambridge University Press.
Lakoff, George, and Mark Johnson
1980
Metaphors We Live By. Chicago: University of Chicago Press.
Morse, Anna
2005
Snowflake kids: What should we do with frozen embryos? BreakPoint Online, June 22, 2005.
Oakley, Todd, and Seana Coulson
under Connecting the dots: Mental spaces and metaphoric language in disreview course. In Mental Spaces Approaches to Discourse and Interaction,
Todd Oakley, and Anders Hougaard (eds.). Amsterdam/Philadelphia:
Benjamins.
Pascual, Esther
2002
Imaginary Trialogues: Conceptual Blending and Fictive Interaction in
Criminal Courts. Utrecht: LOT Dissertation Series, n. 68.
Shore, Bradd
1996
Culture in Mind: Cognition, Culture, and the Problem of Meaning. Oxford: Oxford University Press.
Wierzbicka, Anna
2001
What Did Jesus Mean? Explaining the Sermon of the Mount and the
Parables in Simple and Universal Human Concepts. Oxford/New York:
Oxford University Press.
209
210
211
Abstract
The cognitive linguistic revolution has had a major, yet contested, impact on psycholinguistic studies of metaphor understanding. Despite the tremendous success,
and increasing popularity of cognitive linguistic work on metaphor, psychologists
have been particularly critical of many of the central claims about conceptual
metaphor, and its possible implications for theories about conceptual structure
and metaphor understanding. Outside of psychology, scholars voice concern
about the lack of explicit criteria for identifying conceptual metaphors from a systematic analysis of conventional expressions, how to distinguish one conceptual
metaphor from another, and the fact that much cognitive linguist research is not
based on the analyses of speech and writing in real cultural contexts. We briefly
outline several steps that cognitive linguists can adopt within the context of their
research to deal with some of these problems. We then go on to describe recent experimental work from psycholinguistics that partly stems from cognitive linguistic
ideas and research, and is, indeed, quite consistent with newer claims about metaphor made in cognitive linguistics. The conclusion provides our impressions of the
present state-of-affairs about the application of cognitive linguistics research to
psycholinguistic accounts of metaphor understanding.
Keywords: psycholinguistics; psychological reality; methodology; experimental
re-orientation of CL; conceptual metaphor theory; psycholinguistic critique of
introspection; setting up of hypothesis; testing of hypothesis
1.
Introduction
The cognitive linguistic revolution has had a major, yet contested, impact
on psycholinguistic studies of metaphor understanding. Cognitive linguistic studies have demonstrated that metaphor is not just a special linguistic device, perhaps reflecting one-shot mental mappings, but arises
from enduring patterns of metaphorical thought, or conceptual metaphor, many of which are grounded in embodied and neural activity. Evidence in favor of conceptual metaphor theory comes from a variety of
sources, including systematic patterns of linguistic expressions, novel ex-
212
2.
213
214
Grady 1997; Lakoff and Johnson 1999 for how these may be related to an
underlying primary metaphor ORGANIZATION IS PHYSICAL STRUCTURE).
In a similar vein, Haser (2005) notes that the level of generality for the
source domain of a conceptual metaphor seems, to her, to be specified in
an arbitrary manner. Thus, the ARGUMENT IS WAR metaphor might
better be expressed using a more general source domain such as FIGHT,
even though this level of abstraction makes it far less clear whether we are
still speaking about arguments metaphorically. Many people would feel
quite comfortable subcategorizing argument as literally a kind of fight.
Haser does not sufficiently acknowledge how the work on primary
metaphors handles some of these problems in specifying the exact level
for a source domain in conceptual metaphor theory. But her observations
about source domain indeterminacy raise a related issue in that cognitive
linguists do not typically explore alternative accounts for their data. For
example, Haser (2005) proposes, similar to Murphy (1996), that metaphorical expressions get extended through a chain of family resemblances.
For instance, an expression such as win an argument could be linked to
other expressions by the domain of GAMES, while it could also be linked
to expressions by the domain of WAR. This expression related to WAR
could then systematically relate to an extended use of fortify, which also
then links to BUILDINGS, and so on. In this way, Haser claims to account
for the wide variety of existing metaphorical expressions about ARGUMENTS, while eliminating the problematic issue of their indeterminate
source domains. Haser contends that her approach is more theoretically
parsimonious than the cognitive perspective, as well as free from theory-internal contradictions (Haser 2005: 1920).
At the very least, linguistic analyses alone are unable to determine what
putative conceptual metaphors underlie the use of particular linguistic expressions. As Murphy (1996) advises, metaphorical linguistic patterns
might indeed be related to underlying metaphorical conceptual structure;
however, there must be independent evidence for such structure.
Underlying these criticisms of cognitive linguistic research is the skepticism that many scholars have about drawing inferences about the nature of
thought from linguistic analyses. In many cases, cognitive linguists assume
a direct, or motivated, relationship between systematicity in language with
peoples underlying cognitive reality, including their unconscious mental
representations and brain structures (Dodge and Lakoff 2005). Concepts
are talked about in metaphorical language because people think about
these ideas metaphorically. At the same time, the complexity of metaphori-
215
216
claims simply lack empirical support and should not be treated as either
psychologically real (Glucksberg 2001; Murphy 1996) or philosophical
(Haser 2005) accounts of metaphoric language and thought.
2.3 What is metaphor understanding?
Finally, cognitive linguistic discussions of metaphor understanding typically do not distinguish between different aspects of linguistic experience.
The claim that conceptual metaphors are automatically accessed during language use does not imply that people always compute, or access en
bloc, conceptual metaphors when interpreting real speech, especially familiar or conventional language. Similarly, complex blending theory analyses of the meanings associated with different metaphorical constructions do not imply that (a) people necessarily infer all those meaning
ordinarily during language processing, or (b) engage in the exact set of
blending operations posited by the theory, especially, again, when the expressions being understood are conventional. In general, cognitive linguistics accounts of metaphor typically do not acknowledge the complex
ways that language understanding can be characterized, ranging from
very fast, unconscious comprehension processes to slower, more reflective, interpretation activities (Gibbs 1994).
2.4 Methodological suggestions
There are several ways that cognitive linguistic work on metaphor can be
presented in a way to make this research more amenable to empirical validation (adapted from Gibbs in press).
1. Dont rely solely on your own intuitions. Systematically ask others, including nonlinguists, for their judgments, and report the degree of
similarity of response.
2. Dont just use made-up examples Examine a large sample of linguistic
materials related to the topic or phenomenon of interest. Although individual case studies and the analysis of individual exceptions to
broader trends are always interesting and, indeed, often important, determining the extent to which some linguistic phenomenon exists, and
its frequency, is critical to establishing the legitimacy of ones intuitions
about some aspect of language.
3. Be as concrete as possible about the exact criteria used for classifying
some language item as an instance of some larger linguistic or conceptual category. For instance, describe exactly what criteria have been
217
218
3.
Our discussion of some criticisms of cognitive linguistic research on metaphor is partly offset by a wide variety of psycholinguistic evidence that
supports basic tenets of conceptual metaphor theory. Indeed, many psycholinguistic studies provide evidence suggesting that accessing conceptual metaphors is part of the interpretation and processing of many kinds
of metaphorical language (Boroditsky and Ramscar 2002; Gibbs 1994,
2006; Langston 2002), even if there remains much debate within psychology on the merits of this work (Glucksberg 2001; Murphy 1996).
For the most part, psycholinguistic research has followed cognitive linguistic ideas that characterize the activation of conceptual metaphor during metaphor understanding as a purely cognitive process. Thus, understanding the conventional phrase Our relationship has hit a dead-end
street is partly accomplished through the activation of the conceptual
metaphor LIFE IS A JOURNEY in long-term memory. This enduring
chunk of metaphorical knowledge has a source domain (e.g., JOURNEY)
that is grounded in the pervasive bodily experience, or image-schema, of
SOURCE-PATH-GOAL. But the entire process of accessing a specific conceptual metaphor during verbal metaphor understanding is mostly
viewed as activating abstract, schematic, disembodied knowledge that is
not tied to ongoing bodily action
But there is now an emerging body of empirical work suggesting that
understanding what any word or phrase means requires that listeners engage in an experiential/embodied simulation of the described situation
(Gallese 2005). Research from cognitive linguistics, known as simulation semantics, gives a primary role to embodied simulations in drawing appropriate inferences from various metaphorical and non-metaphorical language (Bergen 2005; Feldman and Narayanan 2004; Gallese
and Lakoff 2004). For example, concepts are physical objects that can be
touched, held on to, juggled, and dropped. When hearing grasp the
concept listeners engage in, or imagine engaging in, a relevant body action, such as grasping, that facilitates metaphorical construal of the abstract notion of concept as a physical entity, such that concepts can
be things that are grasped, held on to, dropped, misplaced, chewed on,
and so on. In this way, the recruitment of embodied metaphors in some
aspects of verbal metaphor understanding is done imaginatively as
people recreate what it must be like to engage in similar actions. The
remainder of this paper describes various new experimental research,
employing rather different methodologies, which provide additional be-
219
220
chew on the gum), where peoples images should focus on the procedural characteristics of the concrete actions (i.e., moving their mouths
as they chew the gum), peoples mental images for metaphorical phrases
should show an analogical understanding of how abstract domains, such
as ideas or concepts, can be actively structured in terms of embodied
source domains (i.e., chewing on something to get more out of it).
The results of a first study showed that when participants were presented specific phrases that were either metaphorical (e.g., grasp the concept) or nonmetaphorical (e.g., grasp the branch), given ten seconds to
form a mental image of that action, and asked What is particularly noticeable in your image? they gave a wide variety of responses to this question, and these could be roughly divided into two groups. The first set of
answers made some specific reference to the participants actually participating in the action. For example, when one participant was given the
metaphorical phrase chew on the idea, she noted, My jaw goes up and
down as I chew. People actually gave far more of these specific references
to participating in the action responses for the nonmetaphors (63 %) than
to the metaphors (29 %).
But for the metaphors, people gave significantly more conceptualized
description of the action responses (71 %) than they did for the nonmetaphors (37 %). For instance, when one person was presented with the metaphor stretch for understanding, he said that the most noticeable thing
in his image was there is much stretching going in both in terms of the
ideas being stretched out to see if they are true and me stretching to better
see of examine the idea. The participant essentially noted that IDEAS
ARE OBJECTS which can be physically inspected, by stretching them out
to more effectively examine them, and that UNDERSTANDING IS
GRASPING enables the person to extend his or her body to better control
the object, and thus better understand it. This response provides an excellent example of how embodied metaphors constrain the mental images
people construct when hearing metaphorical action statements.
When participants were next asked, Why is this concept (e.g., idea)
sometimes associated with this action (e.g., chewing)? they gave two
types of responses. The first set of responses focused on providing a concrete explanation of the relevant process or action. For instance, when one
participant heard the nonmetaphorical phrase chew on the gum. she responded with That is what you do with gum-chew on it. But for the
metaphors, people specifically provided an analogous, conceptual explanation as to why some concept was sometimes associated with some action or process. For example, for the metaphorical phrase chew on the
221
idea, one person said Chewing is related to a slow methodological activity and it could be related to turning something over in your mind to
better understand it. Overall, people gave analogous, conceptual explanations significantly more often to the metaphors (77 %) than to the nonmetaphors (36 %). This result is as expected, assuming that peoples mental images for metaphorical action phrases are constrained by their
embodied, metaphorical understandings of the target domains referred to
in these expressions (e.g., ideas, concepts, feelings).
A second study more directly explored the possibility that people imagine metaphorical actions by engaging in embodied simulations of the
actions referred to in the metaphorical statements. Participants once
again heard different metaphorical and nonmetaphorical expressions,
formed mental images for these phrases, and then answered a series of
questions about their images. In Experiment 2, however, people also participated in one of three enactment conditions in which they first did one
of three things: (a) watched the experimenter make a bodily action relevant to the main verb in each statement (e.g., making a stretching motion
before forming a mental image for the phrase stretch for understanding), (b) watched the experimenter make a relevant bodily action, which
they then imitated, before being given ten seconds to form their mental
image for a phrase, or (c) watched the experimenter make a relevant
bodily action, then imagined themselves doing the same action, before
forming a mental image for the phrase. These three experimental treatments were referred to as the watching, imitating, and imagining conditions, respectively.
Analysis of the responses to the conceptual explanations for the metaphors, collapsed across all three enactment conditions, showed that 78 %
of these referred to additional bodily actions and consequences of these
actions related to the main verb in each metaphorical phrase. This
proportion was higher than that obtained in Experiment 1 (48 %). For
example, when one participant was given the phrase put your finger on
the truth (in the imagine condition), she replied I guess being able to
touch the truth is an important thing, being able to relate to it, being able
to actually see that it is a physical thing and can be examined. This evidence reflects the product of the embodied simulation the participant
constructed that made this impossible action plausible and meaningful.
The extent to which people ordinarily engage in imagistic processes
during online metaphor comprehension is unclear. But the Gibbs et al.
studies show that asking people to form explicit mental images for metaphoric and nonmetaphorical language can reveal significant differences in
222
peoples intuitions about why these phrases have the specific meanings
they do, including the important role of embodied metaphorical thought
in constructing simulations of different metaphorical actions. Having
people engage in different kinds of bodily action does not interfere with
the creation of metaphorical images, but seems to highlight the imaginative reasons for why impossible metaphorical action phrases are meaningful and plausible.
3.3 Bodily movement and metaphor comprehension
In most theories of metaphor understanding, when people read grasp
the concept they must inhibit the physical meaning of grasp to properly infer its abstract, metaphoric meaning (see Glucksberg 2001). Under
this view, having people make a grasping motion before reading grasp
the concept should interfere with their immediate processing of the
phrase. Yet one recent set of studies demonstrated this was not true. Wilson and Gibbs (2006) hypothesized that if abstract concepts are indeed
understood as items that can be acted upon by the body, then performing
a related action should facilitate sensibility judgments for a figurative
phrase that mentions this action. For example, if participants first move
their arms and hands as if to grasp something, and then read grasp the
concept, they should verify that this phrase is meaningful faster than
when they first performed an unrelated body action. Engaging in body
movements associated with these phrases should enhance the online simulations that people create to form a metaphorical understanding of abstract notions, such as concept, even if concepts are not things that
people can physically grasp.
In fact, a computerized reading-time study showed that participants responded more quickly to the metaphorical phrases that matched the
preceding action (e.g., the motor action grasp was followed by grasp the
concept), than to the phrases that did not match the earlier movement
(e.g. the motor action kick was followed by grasp the concept). People
were also faster in responding to the metaphor phrases having performed
a relevant body movement than when they did not move at all. In short,
performing an action facilitates understanding of a metaphoric phrase
containing that action word. A second study showed that same pattern of
bodily priming effects when participants were asked to imagine performing the actions before they made their speeded responses to word strings.
This result reveals that real movement is not required to facilitate metaphor comprehension, only that people mentally simulate such action.
223
224
225
226
4.
Conclusion
Cognitive linguistic ideas on metaphor understanding have certainly generated a great deal of discussion and experimental research within psycholinguistics. There remain important criticisms of cognitive linguistic
methods, and distrust of its theoretical conclusions about metaphorical
thought and language use. But it is also clear that much research has been
reported which is deeply consistent with aspects of conceptual metaphor
theory, especially in regard to newer advances on metaphor understanding as an embodied simulation process. The experimental studies described in this chapter employ a variety of methods to explore various
parts of peoples immediate comprehension of and intuitions about different metaphoric language. Yet the findings all point to the intriguing idea
that interpreting some metaphorical language involves the construction
of embodied simulations related to the actions mentioned (e.g., moving
along in a good direction). This conclusion is consistent with an emerging literature in psycholinguistics on the importance of embodiment in
nonmetaphorical linguistic processing (Gibbs 2006a; Pecher and Zwaan
2005). In this way, the application of cognitive linguistic ideas on metaphor are supported by experimental research on language comprehension, more generally.
Despite its success, cognitive linguistic research will continue to be contested by various scholars in different academic disciplines, until it addresses some of its methodological weaknesses, or at least better articulates the methods implicit in much of this work. Psycholinguists should
aim to explain systematic metaphor patterns and novel metaphorical extensions, for example, as discovered by cognitive linguists. In this way,
psychologists and others must focus more of their attention to the empirical contributions of cognitive linguistic research. But cognitive linguists
should also acknowledge the limitations of their own research, as should
all scientists, and seek to ground their empirical findings on firmer
grounding, especially in the ways that linguistic evidence is gathered, analyzed, and explained in reliable, falsifiable ways.
227
References
Barcelona, Antonio
2002
Clarifying and applying the notions of metaphor and metonymy within
cognitive linguistics: An update. In Metaphor and Metonymy in Comparison and Contrast, Ren Dirven, and Ralf Prings (eds.), 207277.
Berlin/New York: Mouton de Gruyter.
Bergen, Benjamin
2005
Mental simulation in literal and figurative language understanding. In
The Literal and Nonliteral in Language and Thought, Seana Coulson, and
Barbara Lewandowska-Tomaszczyk (eds.), 255280. Frankfurt: Lang.
Boroditsky, Lera, and Michael Ramscar
2002
The roles of body and mind in abstract thought. Psychological Science
13: 185189.
Dirven, Ren
2002
Metonymy and metaphor: Different mental strategies of conceptualization. In Metaphor and Metonymy in Comparison and Contrast, Ren
Dirven, and Ralf Prings (eds.), 75112. Berlin/New York: Mouton de
Gruyter.
Dodge, Ellen, and George Lakoff
2005
Image schemas: From linguistic analysis to neural grounding. In From
Perception to Meaning: Image Schemas in Cognitive Linguistics, Beate
Hampe (ed.), 5792. Berlin/New York: Mouton de Gruyter.
Fauconnier, Gilles, and Mark Turner
2002
The Way We Think. Conceptual Blending and the Minds Hidden Complexities. New York: Basic Books.
Gallese, Vittorio
2005
Embodied simulation: From neurons to phenomenal experience. Phenomenology and the Cognitive Science 4: 2348.
Gallese, Vittorio, and George Lakoff
2005
The brains concepts: The role of the sensory-motor system in conceptual knowledge. Cognitive Neuropsychology 22: 455479.
Gibbs, Raymond W.
1994
The Poetics of Mind: Figurative Thought, Language, and Understanding.
New York: Cambridge University Press.
2006a Embodiment and Cognitive Science. New York: Cambridge University
Press.
2006b Walking the talk while thinking about the talk: Embodied interpretation of metaphorical narratives. Manuscript submitted for publication.
in press Introspection and cognitive linguistics: Should we trust our own intuitions? Annual Review of Cognitive Linguistics.
Gibbs, Raymond W., Jessica Gould, and Michael Andric
in press Imagining metaphorical actions: Embodied simulations make the impossible plausible. Imagination, Cognition, & Personality.
228
Glucksberg, Sam
2001
Understanding Figurative Language: From Metaphors to Idioms. New
York: Oxford University Press.
Haser, Verena
2005
Metaphor, Metonymy, and Experientialist Philosophy: Challenging
Cognitive Semantics. Berlin/New York: Mouton de Gruyter.
Keysar, Boaz, Yeshayahu Shen, Sam Glucksberg, and William S. Horton
2000
Conventional language: How metaphoric is it? Journal of Memory and
Language 43: 576593.
Lakoff, George, and Mark Johnson
1980
Metaphors We Live by. Chicago: University of Chicago Press.
1999
Philosophy in the Flesh: The Embodied Mind and its Challenge to Western Thought. New York: Basic Books.
Langacker, Ronald W.
1987
Foundations of Cognitive Grammar. Vol. 1. Theoretical perspectives.
Stanford: Stanford University Press.
Langston, William
2002
Violating orientational metaphors slows reading. Discourse Processes
34: 281310.
Murphy, Gregory
1996
On metaphoric representations. Cognitio 60: 173204.
Pecher, Diane, and Rolf A. Zwaan (eds.)
2005
Grounding Cognition: The Role of Perception and Action in Memory,
Language, and Thinking. New York: Cambridge University Press.
Ritchie, David
2003
ARGUMENT IS WAR Or is it a game of chess? Multiple meanings
in the analysis of implicit metaphors. Metaphor and Symbol 18:
125146.
Ruiz de Mendoza Ibez, Francisco J.
2000
The role of mappings and domains in understanding metonymy. In
Metaphor and Metonymy at the Crossroads: A Cognitive Perspective,
Antonio Barcelona (ed.), 109132. Berlin/New York: Mouton de
Gruyter.
Steen, Gerard J.
2002
Metaphor identification: A cognitive approach. Style 3: 286417.
Stefanowitsch, Anatol
2004
Happiness in English and German: A metaphorical-pattern analysis.
In Language, Culture, and Mind, Michel Achard, and Suzanne Kemmer
(eds.), 137150. Stanford: CSLI Publications.
Vervaeke, John, and John Kennedy
1996
Metaphor in language and thought: Falsification and multiple meanings. Metaphor and Symbolic Activity 11: 273284.
Wilson, Nicole, and Raymond W. Gibbs
2006
Real and imagined body movement primes metaphor comprehension.
Manuscript submitted for publication.
229
Abstract:
Similarity mapping (e.g., the butterfly is like a rainbow) is one of the earliest metalinguistic abilities children master and is likely to constitute a stepping-stone for
the development of analogical and metaphorical mapping abilities. We investigated the initial emergence and later development of this meta-linguistic ability in
childrens speech and gesture, using longitudinal observations of 40 Englishspeaking children from ages 1;2 to 2;10. We focused on the construction X IS
LIKE Y as the use of the word like is one of the earliest signs of similarity mapping ability in young children. Our results showed that children began to produce
similarity mappings routinely by age 2;2. They initially conveyed only the source
domain of the mapping in their speech and relied on nonverbal modalities (i.e.,
gesture and communicative context) to convey the target domain. However, with
increasing age, children showed a greater tendency to simultaneously convey both
the source and target domains of the similarity mapping in their speech. Moreover, the onset of X IS LIKE Y constructions in childrens speech was preceded
by the onset of gesture-speech combinations expressing similarity relations without the word like. Thus, gesture appears to be at the cutting edge of early language
development it both predates and serves as the supporting context for oncoming
changes in speech.
Keywords: similarity mapping; x is y like construction; acquisition of constructions; early gesture; development of similarity mapping ability; language development; gesture-speech combination; early metaphor
1.
230
Seyda
similarity mappings at younger ages. We explore this possibility by looking not only at childrens speech but also at their gestures.
Children begin to produce combinations in which gesture conveys different information from speech (eat + point to cookie) between ages
1;21;6, and gesture-speech combinations of this sort index oncoming
changes in speech: Children take the developmental step that allows them
to combine words with gestures to create a sentence-like meaning (e.g.,
ride + point to bike) several months before they take the step that enables them to combine words with other words to create a sentence (e.g.,
ride bike). The question we ask here is whether children use gesture in
combination with speech to convey similarity-based comparisons (like
ice-cream cone + point to mushroom; I go like that over my cup + rotate
finger as if stirring) before they convey these comparisons in speech
(ice-cream cone is like mushroom; I move my hand like I am stirring). To explore this question, we use longitudinal observations of 40 typically-developing children, all raised as monolingual English speakers. The
children were videotaped for 90 minutes in their homes every four months
while interacting with their primary caregivers.
To preview our results, we found that children are able to produce similarity mappings in their spontaneous speech two years earlier than they
produce them in experimental settings. Initially, children convey only the
source domain of the mapping in speech, typically relying on gesture to
convey the target domain. Thus, gesture grounds childrens early similarity mappings in the here-and-now. Even more interesting, the onset of
gesture-speech combinations expressing similarity relations without the
word like (e.g., doggie + point to cat) heralds the onset of X IS LIKE Y
constructions in the childrens speech. In other words, children use the
juxtaposition of gesture and speech to convey a similarity relation before
they use the word like to do so. Moreover, gesture-speech combinations
conveying a similarity relation without like decline in frequency just when
the X IS LIKE Y construction appears in the childrens speech.
Our results place gesture at the cutting edge of early language development gesture both presages oncoming changes in childrens speech and
serves as a forerunner of linguistic advances. At a point when children do
not have the words to express similarity mappings, gesture provides them
with a tool to express such mappings. And by doing so, it acts as a harbinger of change in the childs developing language system.
2.
231
Children spontaneously produce a variety of novel expressions that highlight similarities between objects during the preschool years (Billow 1981;
Chukovsky 1968; Carlson and Anisfeld 1969; Elbers 1988; Winner 1979).
They use speech to describe a bald man as having a barefoot head, a mint
candy as making a draft in the mouth (Chukovsky 1968), or a spinning top
as wobbling like a snake (Winner 1979). At these early ages, children also
use gesture and speech together to convey similarities between objects
(Billow 1981; Elbers 1988; Winner 1979). They point to an upward facing
pacifier and call it a candle, point to a car-shaped bread crumb and say car
that (Elbers 1988), hold up a rubber animal next to hair and say it is going
to eat some grass (Billow 1981), or hold up a horn upside down and turn it
in circles while uttering the word mixer (Winner 1979).
Experimental work on childrens understanding of similarity mappings
also points to early onset. Four- to five-year-old children can build simple
mappings between two objects based on feature-based similarity (Billow
1975; Epstein and Gamlin 1994; Gardner, Kircher, Winner and Perkins
1975; Mendelsohn, Robinson, Gardner and Winner 1984; Vosniadou and
Ortony 1983; Winner, McCarthy and Gardner 1980) and between two
events based on action-based similarity (Dent 1984, 1987). For example,
, children were
when asked to complete the statement, An eye is like a
more likely to choose a similarity-based match (e.g., a button) than an
anomalous match (e.g., a fork; Epstein and Gamlin 1994). Similarly, when
asked to pick two objects that go together, children were more likely to
group a cone-shaped block with a toy rocket-ship which was similar in
shape, rather than matching it with another block of a different shape
(Winner, McCarthy and Gardner 1980). Five-year-olds could also provide similarity-based interpretations when asked about expressions that
involved comparisons between objects such as her hair is spaghetti, a cloud
is like a sponge, or the butterfly is like a rainbow (Gardner, Kircher, Winner
and Perkins 1975; Gentner 1988; Billow 1975; Malgady 1977). For
example, they make sense of the statement a cloud is like a sponge by saying that both clouds and sponges are round and fluffy (Gentner 1988), or
they complete the statement he looks as gigantic as by selecting from
among multiple choice alternatives an ending that draws on a similaritybased comparison he looks as gigantic as a double-decker cone in a babys
hand (Gardner, Kircher, Winner and Perkins 1975).
Gardner and his colleagues (Gardner, Winner, Bechhofer and Wolf 1978)
suggest that the ability to make similarity-based comparisons emerges in-
232
Seyda
3.
233
ness crawls through the body) from among forced-choice alternatives, and
by age 5;0, they can provide explicit verbal explanations for these metaphorical expressions (e.g., time crawls by means it goes slowly, ideas slip
from the mind means you forget them; zalskan 2002, 2003, 2005, under
review). Thus mappings based on object similarity and simple analogical
relations, along with familiar metaphorical mappings, are within the communicative repertoire of the preschool child.
However, not surprisingly, the ability to understand more complex
analogical and metaphorical mappings involving less familiar domains
(e.g., mapping physical sensations onto psychological traits) or higher
order relations (i.e., mapping situations based on common higher order
relations, such as the similarity between an apple falling from a tree permitting a cow to reach it and a book falling from a shelf permitting a child
to reach it; Gentner and Rattermann 1990) increases with age and
achieves adult-like quality somewhere between ages 10;014;0 (Asch and
Nerlove 1960; Cicone, Gardner and Winner 1981; Gardner et al. 1975;
Schechter and Broughton 1991; Winner, Rosentiel and Gardner 1976).
But the ability to understand an analogy or metaphor is not determined
solely by a childs age; the nature of the conceptual domain also matters
(Gentner and Rattermann 1991; Keil 1986). For example, five-year-old
children can correctly map animate terms onto cars (e.g., the car is thirsty), but have difficulty understanding metaphors that involve mappings
between taste terms and people (e.g., she is a bitter person; Keil 1986).
From this perspective, the development of analogical or metaphorical
ability is a learning process that extends well into adulthood and shows
different developmental trajectories for different conceptual domains
based on ones knowledge of the domain.
4.
The research just reviewed clearly shows that children can form similarity
mappings by preschool age when tested in an experimental setting. However, we might find that children are able to convey similarity mappings
even earlier if we look at their spontaneous language, particularly when
that language is produced along with gesture, simply because a childs initial grasp of an idea is often evident first in gesture. For example, children
begin to produce combinations in which gesture conveys different information from speech (eat + point to cookie) somewhere between ages
234
Seyda
235
Table 1. The sample of children classified according to ethnicity and family income
Family income
Parents ethnicity
African Asian Caucasian Hispanic Mixed Total
American
$15,000-$34,999
$35,000-$49,999
$50,000-$74,999
$75,000-$99,999
$100,000 or more
Total
24
40
5.
A broad look at the childrens speech showed that they steadily increased
their speech production over time. As can be seen in Table 2, children produced more communicative acts containing speech (F(5,170)=84.09,
p<0.001), more different word types (F(5,170)=174.74, p<0.001), and
more words overall (i.e., tokens, F(5,170)=95.32, p<0.001) with increasing
age. There was a significant increase in the number of communicative acts
containing speech and word types between ages 1;6 and 1;10 (ps<0.001,
Schffe), and significant increases in all three measures from age 1;10 to
2;2 (ps<0.001, Schffe). In addition, children continued to increase their
word tokens and word types from age 2;2 to 2;6 (ps<0.01, Schffe).
Childrens verbal lexicons showed a steep increase from 11 word types and
44 word tokens at age 1;2 to 239 word types and 1741 word tokens at
age 2;10, and the majority of the children (37/40) were producing multiword combinations by age 1;10.
Children also increased their gesture production over time. They produced more communicative acts with gesture (F(5,170)=10.22, p<0.001),
more gesture tokens (F(5,170)=10.82, p<0.001), and more gesture-speech
combinations (F(5,170)=34.29, p<0.001) with increasing age. There were
significant increases in the mean number of communicative acts with ges-
236
Seyda
ture and in gesture tokens between ages 1;2 and 1;6 (ps<0.05, Schffe),
and a significant increase in the mean number of gesture-speech combinations between ages 1;6 and 1;10 (p<0.01, Schffe). By 1;2, only half of
the children (21/40) were producing gesture-speech combinations, but by
1;6, all but one child were combining gesture with speech. Thus, overall
children showed steady increases in their speech and gesture production
over time.
Table 2. Summary of childrens speech and gesture production a.
1;2
1;6
1;10
2;2
2;6
2;10
Speech b.
Mean number of communicative acts containing
speech (SD)
38 (44) 157 (125) 351 (249) 549 (257)
44 (53) 179 (142) 479 (401) 1054 (658) 1475 (795) 1741 (789)
11 (12)
35 (25)
91 (62)
162 (79)
213 (76)
239 (74)
Percentage of children
producing at least one
two-word combination c.
25 %
(10/40)
63 %
(25/40)
93 %
(37/40)
97 %
(37/38)
100 %
(37/37)
100 %
(38/38)
53 (36)
90 (63)
117 (73)
126 (88)
115 (62)
105 (59)
54(36)
91(64)
119(75)
131(90)
123(68)
112(65)
6(9)
30(32)
67(49)
95(66)
97(55)
88(51)
98 %
(39/40)
98 %
(39/40)
97 %
(37/38)
100 %
(37/37)
100 %
(38/38)
Gesture
Percentage of children
producing at least one
52 %
gesture-speech combination (21/40)
a.
b.
c.
SD = standard deviation
All speech utterances are included in the top part of this table, even those produced with
gesture.
As with most longitudinal designs, we missed a few sessions for some of the children,
which led to slight differences in sample size at later data points. Therefore, the total
number of children with respect to sample size at each data point is provided in parentheses.
6.
237
238
Seyda
239
(3) Like a mean monster + child has been talking about tornados
[2;10]
Tornado is like a mean monster
Thus, childrens initial uses of the construction showed an effect of domain type (F(2, 68)=16.17, p<0.0001). As Figure 2 illustrates, once
children began producing the X IS LIKE Y construction, they were significantly more likely to express the source domain on its own than either
the target domain on its own (p<0.0001, Schffe) or the source and target
domains together (p<0.03, Schffe).
There was also an effect of age in the childrens expression of source and
target domains (F(5, 170)=6.77, p<0.0001), which interacted with domain
type (F(10, 340)=6.88, p<0.0001). Children increased the number of X IS
LIKE Y constructions they produced containing the source domain on
its own over time (F(5, 170) = 8.25, p<0.0001) and the source and target
domains together (F(5, 170) = 4.66, p<0.001), with significant changes
from age 2;2 to age 2;10 for both (ps<0.05, Schffe). However, no such developmental trend was observed for the target domain on its own whenever the children expressed the target domain of a similarity mapping,
they also expressed the source domain (see examples 46).
(4) That one like this one + child is holding an ornament and looking at another ornament in her mothers hand [2;2]
Childs ornament is like mothers ornament
(5) Bunny looks like me + child is talking about a bunny character on tv [2;6]
Bunny looks like child
(6) The cow is like that cow there + child is talking about the two
cows in the two puzzles he is playing with [2;10]
The cow in one puzzle is like the cow in the other puzzle
Children also showed developmental changes in the linguistic means
they used to express the source and target domains of the X IS LIKE Y
construction. As can be seen in Table 3, they predominantly relied on demonstrative (this, that, these, those) and personal (it, he, she, me) pronouns to express the source domain and reliably increased their use of
pronominal references over time (F(5, 170)=7.95, p<0.0001, M=0.68 at
age 2;2 vs. M=4.71 at age 2;10). They used explicit nouns (e.g., baby,
spider) or verbs (e.g., climb, eat) to express the source domain far less
often. Nevertheless, their use of nominal devices also increased reliably
240
Seyda
Figure 2. Mean number of X IS LIKE Y constructions for which only the source
(hatched bars), only the target (gray bars), or both the source and the
target domains (black bars) were encoded in speech.
with age (F(5, 170)=4.81, p<0.001, M=0.16 at age 2;2 vs. M=2.42 at age
2;10) (see examples 79).
(7) Looks like it + child is comparing bingbong in book to squirrel
outside [2;2]
Bingbong looks like squirrel
(8) Just like Dori + child is comparing her voice to that of dori in
movie [2;6]
Childs voice like Doris voice
(9) Look like a whale + child is looking at a whale picture in
album [2;10]
Whale picture looks like a whale
The pattern was slightly different for the target domain. Children used
both linguistic devices (pronominal and nominal) at roughly equal rates
at each age. Their use of nouns and verbs to express the target domain in-
241
creased reliably over time (F(5, 170)=11.56, p<0.0001; M=0.08 at age 2;2
vs. M=1.42 at age 2;10), but their use of pronominal reference to express
the target domain did not (F(5, 170)=1.97, ns; M=0.13 at age 2;2 vs.
M=1.39 at age 2;10) (see examples 1012).
(10) Do you close it like this? + child closes door by pushing it with
foot [2;6]
Do you close door like you close it by pushing it with foot?
(11) Baby like the one at grandmas + child is playing with a baby
doll [2;2]
Baby doll is like the baby doll at grandmothers house
(12) Dad is like D and dog + child is learning how to spell letter d
[2;10]
D in dad is like D in dog
Table 3. Types of linguistic means children used to encode source and target domains of the X IS LIKE Y construction in their speech a.
1;2
1;6
1;10
2;2
2;6
2;10
0 (0)
0 (0)
0 (0)
0 (0)
0.03 (0.16)
0 (0)
0 (0)
0 (0)
0 (0)
0 (0)
0 (0)
0 (0)
Source domain
Target domain
a.
SD = standard deviation
242
Seyda
243
(17) We have to rock her like this + child rocks doll gently in the toy
car seat [2;10]
We have to rock the doll like rock her gently
In the third type of action-based comparison, the source domain was conveyed with a demonstrative pronoun (like this/that) and the target was
conveyed by a bleached verb (go, do) (see examples 1820).
(18) I go like this + child hops up and down like a frog [2;6]
I go like I hop like a frog
(19) You do it like this + child paints picture by stamping on it [2;6]
You do the picture like you paint by stamping on it
(20) No go like that + child rides bike by walking his feet on the
sides [2;10]
Go like you ride bike by walking your feet on the sides
Of the three types of action-based comparisons, the first type in which the
target had to be inferred from context was the most frequent and the first
to emerge. As Table 4 illustrates, children used the like+demonstrative
pronoun construction at age 2;2 (M=0.42) and increased their use of this
construction over time (M=2.21 at age 2;10). Like this/that constructions were followed by strong verb+like this type constructions. Only a
few children produced this construction at age 2;2 (M=0.05) but these
constructions became more frequent at age 2;6 (M=0.46) and by age 2;10,
half of the children were using the strong verb+like this construction
(M=1.08). The bleached verb + like this/that type constructions emerged
later. Children did not begin to use bleached verbs to express the target
domain until age 2;6 and then only 8 of the 40 children used this construction (M=0.97). Children remained relatively stable in their use of
bleached verbs to mark the target domain through age 2;10 (M=0.68).
Table 4. Mean number of different types of action-based comparisons in childrens
speech a.
1;2
1;6
1;10
2;2
2;6
2;10
0 (0)
0.03 (0.16)
0 (0)
0.42 (1.0)
0.76 (1.4)
2.21 (3.06)
0 (0)
0 (0)
0 (0)
0 (0)
0 (0)
0 (0)
a.
0 (0)
244
Seyda
The feature-based comparisons highlighted similarities in perceptual features (e.g., shape, size, color) between two entities. The most typical linguistic construction for feature-based comparisons consisted of a noun
encoding the source domain without any mention in speech of the target
domain; the verb look often accompanied these constructions (see
examples 13, repeated below).
(1) Look like a house + child is looking at a church painting [2;2]
Church painting looks like a house
(2) Just like potato-head + child is playing a computer game similar
to potato-head [2;6]
Computer game is like potato-head
(3) Like a mean monster + child has been talking about tornados
[2;10]
Tornado is like a mean monster
In the second type of feature-based comparison, both the source and target
domains were encoded in a noun or a pronoun, frequently accompanied by
the verb look. Interestingly, most of these feature-based comparisons highlighted the general appearance of the two objects, without specifying the
exact basis of the comparison, as can be seen in examples 21-to-23.
(21) That looks like Scooby + child is looking at a dog picture [2;2]
Dog picture looks like Scooby the dog
(22) It looks like a skirt + child is holding her underskirt [2;6]
Underskirt looks like a skirt
(23) I want to make a boat like that one + child has been building boats
with blocks [2;10]
I want to make a boat like my previous boat
There were, however, a few feature-based comparisons, which specified the
basis of comparison between the two objects. The two dimensions that were
explicitly mentioned were color (examples 2425) and size (example 26).
(24) It is like blue like this one + child points to blue scribble while
playing with a blue toy [2;6]
Blue toy is blue like the blue scribble
(25) It is brown like my hair + child is drawing with a brown crayon
[2;10]
Brown crayon is brown like childs brown hair
245
246
Seyda
247
1;6
1;10
2;2
2;6
2;10
Number of action-based
comparison that are accompanied by an action specifying the source domain
1 (100 %)
17 (94 %)
63 (78 %)
126 (83 %)
Number of action-based
comparison that are accompanied by an action specifying the target domain
2 (2 %)
9 (6 %)
18
81
151
Number of feature-based
comparison that are accompanied by a present object
specifying the source
domain
7 (50 %)
15 (38 %)
61(44 %)
Number of feature-based
comparison that are accompanied by a present object
specifying the target
domain
10 (70 %)
28 (72 %)
96 (69 %)
14
39
139
a.
Percentages of action-based or feature-based comparisons that are accompanied by an action or an object are provided in parenthesis. For action-based comparison percentages
were computed by dividing the total number of action-based comparisons that are accompanied by an action by the total number of action-based comparisons. For feature-based
comparisons, percentages were computed by dividing the total number of feature-based
comparisons that are accompanied by a present object by the total number of featurebased comparisons.
248
7.
Seyda
We begin by asking how often children used gesture (as opposed to other
non-verbal cues) to specify either the target or the source domain in an X
IS LIKE Y construction. Figure 4A shows childrens mean production of
gesture in X IS LIKE Y constructions and Figure 4B shows the number
of children who produced gesture in these constructions over the six time
periods. As the figures illustrate, gesture was used in X IS LIKE Y constructions as soon as children began producing these construction in
speech, that is, at age 2;2. The number of times the children used gesture in
the X IS LIKE Y construction increased steadily over time (F(5,
170)=4.61, p<0.001), as did the number of children who used gesture.
Four children used gesture in these constructions at age 2;2 (M=0.16);5 at
2;6, one third of the children (N=12/37) used gesture in at least one instance of the construction (M=0.62); and by age 2;10, almost half of the
children (N=17/38) used gesture in the construction (M=1.21). Thus, gesture was used more and more often in the X IS LIKE Y constructions
over time.
249
Figure 4. Childrens mean production of X IS LIKE Y construction in their gesture-speech combinations (Panel A) and the number of children who
produced at least one instance of X IS LIKE Y construction in a gesture-speech combination (Panel B).
250
Seyda
(34) Is that look like grandpa Johndeer? + child points at tractor picture [2;6]
Does the tractor picture look like grandpas Johndeer tractor?
(35) Shelley has sticky tape just like this one + child points at sticky
tape on desk [2;10]
Shelley has sticky tape just like the sticky tape on desk
(36) I can hug a baby like this one + child points at baby mouse in book
[2;6]
I can hug a baby like the baby mouse in book
(37) I eat like this + child performs the iconic gesture of eating
with fist by moving fist back and forth to mouth [2;2]
I eat like I eat with my fist
(38) Like this + child performs the iconic gesture of playing the
accordion by moving arms inward and outward [2;6]
I play the accordion like I play it in a particular style
(39) Give it to me like this + child performs the iconic gesture of
putting pretend necklace on neck by holding two pinched
fingers in air [2;6]
Give the necklace to me like you put it around my neck
(40) I go like that over my cup + child performs the iconic gesture of
dropping something into the cup by opening a closed fist in
air (iconic) [2;10]
I go like I drop something into my cup
(41) Knocked over like this + child performs the iconic gesture of
swaying by moving arm outward forcefully (iconic) [2;10]
I knocked it over like I knocked it over by swaying it
In terms of specifying a source or a target domain that was not encoded in
speech, pointing gestures were used to convey the entity in the target domain that was not mentioned in speech (examples 4245) and iconic gestures were used to convey the action in the source domain that was not
mentioned in speech (example 46).6
(42) Like ice-cream cone + child points to mushroom [2;2]
Mushroom is like ice-cream cone
(43) Like a sun + child points to circular object on tv [2;10]
Object on TV is like a sun
(44) Like a sheep + child points to sheep picture [2;10]
Sheep picture is like a sheep
251
8.
252
(47)
(48)
(49)
(50)
(51)
(52)
(53)
(54)
(55)
(56)
(57)
(58)
(59)
(60)
(61)
Seyda
As can be seen in Figure 5 (see below), the frequency of childrens gesturespeech combinations expressing similarity relations without the word like
changed significantly over time (F(5, 170)= 3.45, p<0.01), increasing from
age 1;2 (M=0.40) to age 1;6 (M=2.23). Childrens production of these
combinations remained relatively unchanged between ages 1;6 and 2;6,
but began to decline thereafter. Importantly, the decline at age 2;6 coincided with an increase in X IS LIKE Y constructions. Thus, the
children became less likely to produce gesture-speech combinations expressing similarity relations without the word like at just the point when
they began producing similarity-based mappings in appropriate syntactic
packaging (i.e., with the word like).
Consistent with the hypothesis that gesture is playing a bootstrapping
role in the emergence of X IS LIKE Y constructions, the onset of gesturespeech combinations expressing similarity relations without the word like
routinely preceded the onset of X IS LIKE Y constructions in speech. Of
the 40 children in our sample, 29 (73 %) produced a gesture-speech combination of this sort before producing an X IS LIKE Y construction in
speech, compared to only one child who displayed the opposite pattern
(29 vs. 1, X 2(1) = 38.88, p<0.001). Of the remaining 10 children, 5 produced gesture-speech combinations of this sort but had not yet produced
the X IS LIKE Y construction in speech; we expect these children to produce X IS LIKE Y in speech in subsequent sessions. The remaining 5
253
254
9.
Seyda
Conclusion
In the process of learning a language children have to learn not only to map
a word onto a particular referent or an event (e.g., calling a red block a red
block) but also to articulate correspondences between objects or events
based on commonalities in their features or relational structure (e.g., a red
block is red like a red crayon). In this chapter, we examined the beginnings
of this mapping ability in a sample of English-speaking children from
ages 1;2 to 2;10, as indexed by their use of the X IS LIKE Y construction.
Our analysis showed an early onset of the X IS LIKE Y construction.
Children began to produce the construction in their spontaneous speech
routinely by age 2;2 two years earlier than they have been found to produce such similarity mappings in experimental settings. However, these
early uses of the X IS LIKE Y construction were highly constrained. First,
the source and target domains of the mapping were typically bound by the
here-and now, involving objects that were present in the immediate environment or actions that were often being executed by the child. Second,
the linguistic forms that the children used to describe either domain were
context dependent. Children typically used pronominal references to encode the source and/or the target domain and relied on non-verbal cues
(i.e., ongoing action, gesture) to clarify these references. Thus, at the early
ages, speech provided a skeletal structure for the X IS LIKE Y construction and children relied on non-verbal means to flesh out this structure.
Children produced two types of X IS LIKE Y constructions, those involving comparison of an object to another object (i.e., feature-based
comparison) and those involving comparison of an action to an ongoing
action (i.e., action-based comparison). In line with earlier work (Gentner
and Rattermann 1991), childrens initial feature-based comparisons were
typically holistic and global, involving mappings based on overall appearance of objects (e.g., pizza picture looks like pizza) rather than particular
aspects or dimensions of the objects (e.g., blue scribble is blue like a blue
stripe). Similarity mappings in which the children made it clear which part
or dimension of the objects they were highlighting were infrequent in our
data. Indeed, in most instances, the dimension of similarity had to be inferred from non-verbal cues and other aspects of context. Moreover, all
instances of the X IS LIKE Y construction involved similarity mappings,
rather than more complex types of mappings (e.g., analogical, metaphorical), providing further support for the hypothesis that similarity mappings between objects act as precursors to more complex mapping types
(Gentner 1988; Vosniadou 1987).
255
However, unlike earlier work, a large portion of the X IS LIKE Y constructions the children in our study produced involved action-based comparisons, which functioned in slightly different ways from feature-based
comparisons. Action-based comparisons provided children with a linguistic frame that they could use to specify an action for which they
lacked a lexical item. The children framed the action-based comparison in
general terms, using phrases such as like this or go like this and then acting
out or gesturing the specific action they intended to convey (e.g., climb
up a ladder; move hand forward forcefully to indicate throwing). Interestingly, even in cases where the children produced a strong
verb to encode the action (e.g., I run/eat like this ), they often produced an action or a gesture to provide a more detailed rendition of the
action specified by the verb (e.g., make doll run rapidly with outstretched legs; perform the iconic gesture of moving fist back
and forth to mouth to convey eating with fist).
This type of framing is not unique to similarity mappings and is commonly observed in metaphorical types of mappings at later ages. As
shown in earlier work (zalskan 2003, under review), when 3- and
4-year-old children are asked to provide verbal explanations for different
types of metaphorical mappings (e.g., How does time fly? What does it
mean when ideas escape from your mind?), they typically produce a like
this/that construction in speech and act out whole body gestures to
further specify the source domain (Like this + child crawls on floor
to indicate hours crawling by; Like that + child moves arms up
and down to indicate ideas flying by). This type of response typically
disappears by age 5;0 when children begin to produce more elaborate verbal descriptions, along with semantically well-integrated gestures (e.g.,
time drips by means it goes really slowly like that + child moves finger
downward with small pauses like dripping water; zalskan
2002, under review). Thus, the particular way children framed actionbased comparisons in our study might be their initial step on the way to
verbally more elaborate X IS LIKE Y constructions. Moreover, gesture
and bodily action might be signaling the childs readiness to take the next
step towards more complete linguistic constructions.
Indeed, as shown in this paper, gesture played two important roles in
childrens production of X IS LIKE Y constructions. First, gesture
served as the supporting context for childrens early X IS LIKE Y constructions. Children initially expressed either the target or the source domains of the similarity mapping, and used gesture to convey the other domain. Even in cases where children expressed both of the domains in
256
Seyda
257
Acknowledgements
We thank Kristi Schoendube, Jason Voigt, Becky Free, and Laura
Schneidman for their administrative, technical and coding support, and
the project research assistants, Karyn Brasky, Elaine Croft, Kristin
Duboc, Jennifer Griffin, Carrie Meanwell, Erica Mellum, Molly Nikolas,
Jana Oberholtzer, Lillia Rissman, Becky Seibel, Kevin Uttich, and Julie
Wallman for their help in collecting and transcribing the data. The research presented in this paper was supported by grant #PO1 HD40605
to Goldin-Meadow.
Notes
1. Communicative acts are words or gestures, alone or in combination, which
were preceded and followed by a pause, a change in conversational turn, or a
change in intonational pattern.
2. The word like in childrens speech became polysemous at age 2;2, functioning
as a verb in some instances (e.g., I like ice-cream) and as a comparative marker
in others (e.g., ice-cream cone is like mushroom). Beginning at age 2;6, a few
children occasionally used like as a discourse marker. In this paper, we focus
exclusively on the uses of like as a comparative marker.
3. The only exception was one child who produced one instance of such a construction at age 1;6.
4. Speech is provided in italics and any nonverbal information (e.g., gesture,
communicative context) is given in small caps; the age of each child is indicated in brackets. Each example is followed by a gloss that approximates what
the child is intending to convey in his/her communication.
5. The only exception was one child who produced one instance of the construction in a gesture-speech combination at age 1;6; this was the same child who
also produced one instance of the construction in speech at age 1;6.
6. Children also produced several gesture-speech combinations that drew on a
similarity-based comparison by using linguistic devices other than the word
like. In these instances, gesture encoded the target domain of the mapping,
and speech conveyed information about the source domain (e.g., It is chocolatekind of chocolate + child holds up brown marker; Color is an apple
+ child points to a green balloon that looks like a green apple). We
first observed such instances at age 2;6, after the onset of the X IS LIKE Y
construction in childrens speech.
258
Seyda
References
Asch, Solomon E., and Harriet Nerlove
1960
The development of double function terms in children. In Perspectives
in Psychological Theory, Bernard Kaplan, and Seymour Wapner (eds.),
4760. New York: International Universities Press, Inc.
Billow, Richard M.
1975
A cognitive developmental study of metaphor comprehension. Developmental Psychology 11(4): 415423.
1981
Observing spontaneous metaphor in children. Journal of Experimental
Child Psychology 31: 430445.
Capirci, Olga, Jana M.Iverson, Elena Pizzuto, and Virginia Volterra
1996
Gestures and words during the transition to two-word speech. Journal
of Child Language 23: 645673.
Chukovsky, Kornei
1968
From Two to Five. Berkeley: University of California Press.
Carlson, Patricia, and Mosche Anisfeld
1969
Some observations on the linguistic competence of a two-year-old
child. Child Development 40: 565575.
Cicone, Michael, Howard Gardner, and Ellen Winner
1981
Understanding the psychology in psychological metaphors. Journal of
Child Language 8: 213216.
Clark, Eve V.
1973
What is in a word? On the childs acquisition of semantics in his first
language. In Cognitive Development and the Acquisition of Language,
Timothy E. Moore (ed.), 65110. New York: Academic Press.
Dent, Cathy H.
1984
The developmental importance of motion information in perceiving
and describing metaphoric similarity. Child Development 55:
16071613.
1987
Developmental studies of perception and metaphor: The twain shall
meet. Metaphor and Symbolic Activity 2 (1): 5371.
Elbers, Loekie
1988
New names from old words: Related aspects of childrens metaphors
and word compounds. Journal of Child Language 15: 591617.
Epstein, Ruth L., and Peter J. Gamlin
1994
Young childrens comprehension of simple and complex metaphors
presented in pictures and words. Metaphor and Symbolic Activity 9 (3):
179191.
Gardner, Howard
1974
Metaphor and modalities: How children project polar adjectives onto
diverse domains. Child Development 45: 8491.
Gardner, Howard, Mary Kircher, Ellen Winner, and David Perkins
1975
Childrens metaphoric productions and preferences. Journal of Child
Language 2: 125141.
259
2002
Metaphors we move by: A crosslinguistic-developmental analysis of
metaphorical motion events in English and Turkish. Ph. D. diss., Department of Psychology University of California, Berkeley, CA.
260
2003
Seyda
261
262
Gary B. Palmer
263
Abstract
In language and culture research, one often finds cognitive linguists and cognitive
anthropologists working on the same problems. This chapter reviews a few topics
to which both cognitive linguists and cognitive anthropologists have made significant research contributions. These topics are (1) Language as culture, (2) Disentangling psychological unity from cultural diversity: The case of experiential language, (3) Distributed knowledge versus consensus in language communities; (4)
Complex categories, (4.1) Reduplication as a complex category, (4.2) Noun classifiers as complex categories, (5) The new relativity in spatial orientation, and (6)
The origins of language.
Keywords: language and culture; cultural variation; cultural distribution; psychological unity; categorization; complex category; classifier system; reduplication;
origin of language
264
Gary B. Palmer
their findings shows that they have made significant contributions and
conducted fruitful dialogues, which have been supported and joined by
the many others who explore the nexus of language, and culture and cognition. This chapter reviews a few of the topics of mutual interest.
In section (1) Language as culture, it is argued that linguistic categories
are predominantly cultural even while they are constrained by biology
and universal physical experiences. Section (2) Disentangling psychological unity from cultural diversity examines this important cultural issue as it
applies to language, in part by comparing the Tagalog and English models
of cognitive process. Section (3) Distributed knowledge versus consensus in
language communities examines the linguistic side of another important
and related cultural issue. The two issues intersect because distributed
knowledge is the primary form of cultural diversity within language communities, however they may be understood and defined, and because arguments for consensus may be based on assumptions of psychological
unity. Section (4) Complex categories, shows that a cultural perspective on
language helps to explain two widespread and important grammatical
phenomena: reduplication and systems of noun classifiers. Subsection
(4.1) Reduplication as a complex category, describes Nahuatl reduplication as a complex category in both its phonology and its semantics. In
(4.2) Noun classifiers as complex categories, I argue from Shona data that
each classifier in a system of classifiers is a polycentric linguistic category
based on multiple interconnected cultural scenarios. Similar arguments
are applied to the language of spatial orientation, which is discussed in (5)
The new relativity in spatial orientation. This section revisits the question
of psychological unity and proposes a general framework of cognitive
maps useful for cross-linguistic, ergo cross-cultural, descriptions of the semantics of spatial orientation. In (6) The origins of language, cognitive linguistic and cultural theory are combined to argue that human language
originated in the ceremonials of Homo heidelbergensis, an immediate precursor to our own species. Section (7) presents conclusions.
1.
Language as culture
265
266
Gary B. Palmer
267
2.
Having made the case for a broad, inclusive field of cultural linguistics in
the previous section, I will admit to overstating the freedom of cultural
forms from biological and environmental constraints. Innate cognitive
abilities and universal experiences play important roles in supporting and
constraining the myriad categories of thought and language (Langacker
1994). Human cognition has evolved for language and cultural variation,
not independently nor in opposition to them.6 On the list of good candidates for innate image-schemas one finds horizontal and vertical orientations of viewed objects, color foci, small numbers, animacy, possession
of bodily parts, and a template for the human face, which enables us to
recognize thousands of faces, though each may be viewed only briefly
(Ehrsson 2004; Haxby et al. 2001; Gordon 2004). Johnson (2005) listed
several image-schemas, but he seems not to regard these simply as innate
structures. Instead, they are seen as properties of bodily experience in the
environment: image schemas are not to be understood either as merely
mental or merely bodily, but rather as contours of what Dewey (1925)
called the body-mind. Johnson mentioned the schemas container,
right and left, front and back, near and far, horizon, center-periphery, compulsion, attraction, blockage of movement, verticality, rectilinear motion, source-path-goal, and scalarity.
In spatial cognition, Langacker (2002: 139) has proposed several basic
notions expected to figure prominently in the locative systems of all
languages. These are separation, inclusion, proximity, line, surface,
source-path-goal. Other concepts proposed by Langacker for describing
constructions in general include profiling, trajector/landmark align-
268
Gary B. Palmer
269
and a significant share of uniformity. Since both biology and culture are
sources of both uniformity and diversity, there is no lack of interesting
work for analysts.
In their analyses of language data, cognitive linguists and anthropologists must often try to sort out the innate from the emergent and the
emergent from the culturally determined. Langacker has commented on
the complexity of the problem:
Any new experience then draws upon a combination of innate abilities and established mental structures from all previous strata, so that it becomes rather
pointless (if not impossible) to fully disentangle the innate from the learned, the
pre-cultural from the cultural, the linguistic from the non-linguistic. (Langacker 1994: 51)
Langacker is no doubt correct to doubt the possibility of full disentanglement, but the position holds a potential contradiction, since he himself
has done as much as anyone to call attention to the importance of innate
cognitive abilities in constraining the variety of grammatical constructions. To recognize such abilities and base a theory of language on them is
to intimate that some disentangling has been done. Furthermore, the
growing body of interesting new analyses based on the principles of cognitive linguistics suggest that the effort has achieved a measure of success.
In cognitive anthropology, the problem of identifying innate categories
has been defined most clearly in cross-linguistic studies of basic color
terms, the main point of reference being Berlin and Kay (1969), who proposed an evolutionary sequence in color naming and universal constraints on the naming of colors. A limitation on these findings is that
they apply only to basic color terms, which are usually only a small percentage of color terms found in any language. Conklin (1964) showed how
the full range of color terms incorporates other dimensions in addition to
the color spectrum, dimensions such as moisture and temperature, or as I
encountered in a Salish language, the texture of the underside of a leaf.
Using the notion of construal from cognitive linguistics, MacLaury
(1992) proposed that cross-linguistic variations in the extensions of color
terms reflect differences in the construal of color fields. He argued that
color categorization is analogous to spatial orientation, so that people are
able to determine their position in color space by adopting a particular
point of view or vantage point. Wierzbicka (1990) seeks the origin of
color terms in the projection of innate foci of color perception onto common experiences of sky, land, the sun, fire and blood, an approach which
resembles the broader theory of Alverson (1991: 112) who would derive
270
Gary B. Palmer
schemas from the primal scene of the dial, orbit, or trajectory of sun and
moon, whose light is both a point and a sweep or array; cloud cover;
altitude; courses of movement/travel through the trajectory; barriers to sensory or locomotor access; and behavior of entities occupying this scene. While one normally thinks of Wierzbickas essentialist approach as diametrically opposed to the Lakoff and Johnson (1980)
concept of image-schemas as emergent categories, if Wierzbickas primes
are actually derived from experience in the world, then the two approaches have something important in common. It appears from the diversity of conclusions in these studies that neither the causes of variations
nor of universals in color terms are well established.7
The language of emotion is a second domain in which much work has
been motivated by the issue of psychological unity versus cultural variation (Kvecses and Palmer 1999; Kvecses, Palmer, and Dirven 2002).
Kvecses (2000, 2005) has aimed at discovering conceptual universals
that motivate metaphors of emotions, especially the emotions of anger
and love. He has found that English, Hungarian, Chinese, and Japanese
all have the container metaphor as the ontological basis for their talk
about anger (2000: 146). In this model, human bodies are containers of
anger, which is conceptualized as a heated substance in a closed container.
The explanation offered by Kvecses for uniformity in these four languages and others, is that people everywhere experience the same physiological responses to the causes of anger (2000: 156). A major component
of this response is a rise in blood pressure, which motivates the image of a
pressurized container. Kvecses (2000: 159) generalized that it is conceptual metonymies (the physiological correlates of anger) that motivate conceptual metaphors of emotion. This explanation fits the emergent/embodied explanation of category formation and psychological unity.
Kvecses is not fixated on language universals. He asked How can we
construct a comprehensive theory that can account for both the universality
and the variation in our use of metaphor? (Kvecses 2005: 3, emphasis in
original) He explains variation as the expression of cultural models and in
fact he was one of the first to introduce cultural models into cognitive linguistics (Lakoff and Kvecses 1987). For example, he accepted the view of
Geeraerts and Grondelaers (1995) that the Euro-American concept of
anger comes from the classical-medieval notion of the four humors. He observed that the Japanese concept of anger involves keeping the real, true
self (honne) hidden in the body container (hara stomach/bowels) by exercising self control. The Chinese concept of anger involves the rise of qi,
energy that flows through the body, which upsets the desired balance of
271
yin and yang postulated by Chinese philosophy and medicine. Cross-linguistic variations in choices of specific metaphors for anger derive from
these differing cultural models (Kvecses 2005: 234235). Support for
cultural influences on these metaphors is provided by Yu (1998).
Closely related to the problem of emotion language is the problem of
language about thinking. I will refer to the language of emotion and
thought together as experiential language. The two domains of experience
are so closely entwined in English that one might conclude that the field is
thoroughly culturally structured (Damasio 1994). Many terms have both
emotional and nonemotional meanings. To see how this semantic conflation manifests linguistically, it may help to consider a few examples
such as the Tagalog definitions in (14) from Palmer (2003a). Emphasis
has been added to the translations:
(1) Ang akala ay isang pag-iisip na binabase sa pakiramdam
spc8 idea inv one-lg thinking lg based dr feeling
Akal is a way of thinking based on a feeling
(2) Ang batid ay isang pag-iisip na may kahalong pakiramdam.
spc know inv one-lg thinking lg ex combined-lg feeling
Batid is a thinking combined with feeling.
(3) Ang intensyn ay pag-iisip at pakiramdam
spc intention inv thinking and feeling
bago nagpaplano ng isang bagay.
before planning gn one-lg thing
An intention is thinking and feeling before planning of a thing.
(4) Ang nahihirapan ay maaaring mangyari sa damdamin, pag-iisip, o
pisikal
spc hardship inv possibly event dr emotion thoughts or physical
kapag nabibigatan ang isang tao sa kanyang ginagawa
if weighed.down spc one-lg person dr his do.(x)
o nararamdaman.
or feel.(x)
Nahihirapan is possibly an event in the emotions, thoughts, or physical
[being] if a person is weighed down by what he does or feels.
The examples show that concepts that one can translate as idea, know, intention, and hardship, are all defined in Tagalog to predicate a combination of thinking and feeling (pagisip and pakiramdam/damdamin).
Lee (2003) reported a similar mixing of thinking and feeling in the mental
models of her English test subjects:
272
Gary B. Palmer
Analytical thought is conceived as paradigmatically active, definable, and separable from other internal behavior. By contrast, feelings and other kinds of
thinking, especially contemplative activity, seem less easily definable and separable from each other A strong negative/positive polarity generated from
within the feelings/emotions range provides an alternative axis of organization
in some personal models, penetrating what has more traditionally been regarded as the domain of thought (2003: 247).
But, it would be no surprise to find that many, or even all, cultures recognize and lexicalize two regions on a gradient that extends from mental
events that are accompanied by intense physiological arousal (emotions)
to those with relatively little physiological arousal (thoughts). Damasio
(1994: xiv; cited in Lee 2003: 247) has argued on evolutionary grounds
that reason is more effective under the governance of emotion and feeling.
This would not necessarily imply that the distinction between thought
and emotion is sufficiently clear and discrete in any language to qualify
the two terms as lexical primitives, but it may be true that the two concepts, however delineated, are necessary to the explication of other concepts, as required by NSM theory (Goddard 2002: 6). Just as likely, this is
another instance of complex categories that needs further description and
analysis.
273
Putting aside the issue of whether the distinction between thinking and
feeling is universally discretely defined, there is the related question of
whether cultures and languages vary in their models of experiential processes. The following quote from Palmer (2003a: 271) illustrates one such
model:
Roy DAndrade (1987, 1995[a]) has discovered what he calls a major direction
of causation in English as follows: action > perception > thought (> feeling) >
wish > intention > action. Feelings may be bypassed, so that wishes result directly from thoughts (1995[a]: 162). Feelings may also stimulate expressive reflex acts such as sneezing, smiling, and crying. [Brackets added.]
DAndrade presented the model with a number of qualifications concerning the order of elements in the process, but let us assume that the model is
sufficiently well-defined to enable cross-linguistic comparison. Analysis
of a sample of Tagalog statements about causation yields the findings
summarized as follows:
Thoughts or thinking are most likely to arise from effort or from other
thoughts, but they may also result from states or feelings. The fact that thoughts
arise from effort follows from the common notion that the mind and thinking
are used. Thought may produce action or further thought, but seldom is it said
to produce feelings. Thought and feeling are not tightly linked in these expressions. To the extent that they are, the direction of causation often runs in
the opposite direction from DAndrades model for English. The definitions
contain no instance of perception motivating thought.
In these expressions, feelings result most commonly from actions, events, states
and changes of state, and from other feelings, not, as in English, from thoughts
and perceptions. Feelings typically produce bodily responses, actions, thoughts,
and other feelings. It is as though two stages in the English model were omitted
from Tagalog (Palmer 2003a: 274).
Where DAndrades model for English has six stages, the Tagalog model
seems to require only three, with one path running through thought, the
other through feeling. Action and perception lead to thought in English,
and then on to wishes, intentions, and further actions. In Tagalog, action
and several other kinds of mental experience lead to either thought or
feeling, and then on to more thoughts or more feelings, or actions and
bodily responses. Both models have a potential pathway with action at
both the initiation and the completion of the scenario. The Tagalog model
can be related to traditional philosophy of mind and body in which the
thoughts pertain to outward conscious being (malay), while feelings are
274
Gary B. Palmer
inner events (niloloob). There is little causal interaction between the two
domains. Both respond to events outside the person and lead to actions
directed outside the person.
The Tagalog model of experiential process was arrived at by applying
standard methods of ethnoscience and anthropology combined with concepts from cognitive linguistics: Native language question frames were
used to elicit definitions which were found to fit standard analytical
frameworks in this case, a taxonomy and a process model; and published ethnographic documentation was reviewed in search of cultural
schemas and scenarios that would help to explain the models (See Palmer
2003a for references).
To summarize, it may be the case that thought and feeling or emotion
are semantic universals, as claimed by Wierzbicka. Nevertheless, thought
and feeling must always be embedded in cultural scenarios that mix the
two domains (Lutz 1988). Lexical domains of experience profile elements
and construals of such mixed socio-cultural models. Case studies from a
wider sample of languages would be welcome, as would further comparative studies such as that of Fortesque (2001). The next section looks
at the issue of how much consensus or diversity may exist within a language community in the cultural schemas and scenarios that constitute
the semantic pole of grammar.
3.
275
276
Gary B. Palmer
277
4.
Complex categories
A good way to approach complexity in language is to begin with Langackers (1987: 58) notion of a symbolic unit of language as a phonological unit linked to a semantic unit. The phonological unit might be a
speakers concept, more or less concrete or abstract, of the sound of a
morpheme, a word, or a phrase structure. The semantic unit could be an
elementary concept or some construal of a complex situation. To say that
they are linked suggests that the sound may evoke the thought, or the
thought the sound. There may be variants of the semantic unit that are
linked to the same phonological unit. This is a precise way of describing
polysemy or roughly speaking, what Geeraerts and Grondelaers (2004)
have called semasiology, which in traditional grammar would be called
homonymy. It may also be the case that a single semantic unit is linked to
278
Gary B. Palmer
279
280
Gary B. Palmer
281
282
Gary B. Palmer
As in Tuggys study of Nahuatl reduplication, the symbolic sub-network (lexical domain) of a Shona classifier obeys no clear rules of taxonomy or deduction. There is no necessary set of features that a nominal
must possess to belong to the class. It must only have a direct or indirect
semantic connection to one of the central scenarios of the language and
culture. The salient connection of a nominal to a semantic domain of related scenarios qualifies it for the special grammatical treatment given to
its class. The concordial forms constitute the phonological pole of the
classifier as a complex category. The semantic pole is the class sub-network of central categories and their satellites.13 Whole classes may also
manifest oppositions vis a vis one another, such as earth/sky or male/female, giving moisture/accepting moisture. The oppositions are harder to
isolate from the noise of the semantic sub-networks, but they appear more
clearly when the central scenarios have been delineated.
How might the theory of classifiers as complex categories contribute to
our understanding of the consensus problem? To the extent that classifiers
cue evaluations, they would function similarly to evaluative reduplications, but that is only a part of their semantic content, which has more
to do with central cultural scenarios. Classifiers clearly function to evoke
large frames of reference, domains of experience, or what were formerly
referred to as cultural themes. The use of classifiers increases the accessibility of related knowledge (as culturally defined). They may, for example,
direct the recall of specific events that fit the frame of reference. Each interlocutor in a verbal exchange may choose to share her own partial
knowledge of the evoked domain or domain related events and thereby
contribute to knowledge consensus in the community or, without sharing
information, she may choose to act as intended by her interlocutor or to
act constructively in that domain. In the latter case, unshared knowledge
of the interlocutors is integrated for community purposes. Of course, she
may choose to neither share nor act, or to act disfunctionally, in which
cases there is no net gain to either consensus or the integration of diversity. Similar reasoning may apply to the figurative usages of spatial language which were brought to our attention so forcefully by Lakoff and
Johnson (1980).
5.
283
The language of spatial orientation has long fascinated cognitive anthropologists and linguists (Bhler, K. 1982 [1934]; Hallowell 1955; Friedrich
1971; Haugen 1969; Basso 1984; Hunn 1990; Haviland 1993, 1996; Senft
1997). This domain of language can reveal how people conceptualize their
surroundings, make their way from one place to another, and direct one
another to places. It can reveal how they think about the relative positions
of objects and the manipulation of objects. The concepts of cognitive linguistics, particularly figure/ground alignment (or Langackers trajector/
landmark alignment) and Langackers concept of the search domain have
found extensive application and they have been particularly helpful in enabling enlightening comparisons of spatial language (Talmy 1983, 2005;
Brugman 1983, 1988; Lakoff 1987; Casad 1988, 1993; Casad and Langacker 1985; Heine 1997; Levinson 1996a, b, 1997; Ogawa and Palmer
1999; Occhi, Palmer and Ogawa 1993).
In an important chapter summarizing research results from Max
Planck Institute for Psycholinguistics at Nijmegen, Levinson (1996b)
concluded that the language of spatial orientation is characterized by a
high degree of relativity. He found no universal coordinate systems nor
any universal grammatical patterns beyond the very abstract notion of
figure and ground orientation conventionalized in clause structure. Even
such seemingly basic and inescapable notions as left/right are not universally recognized and evidently not necessary to spatial orientation. Languages vary in the degree to which they profile and elaborate figures,
grounds, and the relations that connect them, and they vary in the sorts of
spatial information they include in each of these. For example, a positional adjective in Tzeltal Mayan includes information on the shape of the
figure whether it is a container with a mouth that is wider than any other
part, or that has an upright cylindrical body, or that has a narrow neck.
They vary in the extent to which they lexicalize orientations in intervening
space itself as opposed to contact with various parts of grounds. They
vary according to whether spatial language is ego-centric and deictic (relating to the bodily orientations of persons in the speaking situation),
which gives directions such as left/right, or whether it is based on absolute
orientations, which gives directions such as uphill/downhill or north/
south). The Guugu Yimithirr give directions only in absolute terms.
Guugu Yimithirr spatial language relates to a cultural model that constrains orientation even in the absence of speech. This has been demonstrated with clever experiments in which subjects are shown a set of ob-
284
Gary B. Palmer
jects in a certain relation, the objects are rearranged in their absence, and
then they are asked to restore the objects to the original positions. Dutch
subjects consistently restore them to their original positions relative to the
left or right of the subject. The Guugu Yimithirr restore them to their
original positions relative to north/south or east/west.
Contra Levinson, Talmy (2005) proposed a universal fundamental system of spatial schemas that constrain the semantics of closed-class forms
such as spatial prepositions, noun affixes, and verb satellites. In his framework, scenes segment into Figure, Ground, and Secondary Reference Object. Components of scenes may have zero, one, two, or four dimensions
(i.e. they may be a point, line, plane, or volume). Schemas have number:
one, two, between two, among, or amidst. Figures and Grounds may be
stationary or moving. Pathways may be bounded or unbounded. Motions
or locations may be rectilinear or radial. Regions of one, two, or three dimensional space may be compact or diffuse. Volumes may be vacant, solid,
liquid, or (possibly) fire, a category that occurs in Atsugewi, in which
caw specifies motion into a fire. Other schemas pertain to relations between components of scenes, force dynamics, and cognitive/affective states.
Relations among components include relative orientation (parallel, perpendicular, and oblique), degree of remove, and degree of dispersion
(sparse or dense). Degree of remove includes contact (coincidence and adjacency) and non-contact (proximal, medial, distal). Other categories in
his framework are not listed here because they seem more arguable. Talmy
does not argue that these schemas are innate only that they are universal
and basic. Perhaps a few are innate while others are emergent.
From the various findings of both universalists and relativists, one can
infer that, cross-linguistically, space may be thought of as a complex field
of stationary or moving figures and grounds which may be conceptualized
as having shapes and content, inherent spatial orientations (front, back,
etc.), as having parts (head, arm, buttocks, etc.), or as having positions
and postures (upright, bunched together) and mutual orientations (parallel, perpendicular). Entities in the spatial field may be conceptualized at
varying degrees of concreteness or schematicity. Figures and grounds relate to one another by being located at some orientation in space, or in
some degree of contiguity, which may involve localized or distributed contact with various types of adhesion (Ogawa and Palmer 1999). Conventional spatial constructions in any particular language may (or may not)
profile any of these in various combinations. Even languages from the
same language family, such as English and German, may use quite different principles in giving directions or explaining spatial events (Carroll
285
1997). Talmys contraints on the closed-class forms may apply, but systematic cross-linguistic comparisons will be needed to establish their
validity and universality.
Palmer (in press) presents a framework suitable for capturing the relativity of spatial orientation. The framework distinguishes two kinds of
cognitive maps involved in orientation: object maps and view maps. View
maps, the most interesting from a linguistic standpoint, are cognitive
models of the field of view of speakers or other participants. Their substantive conceptual content is primarily relational with inherent abstract
figures and grounds. Complex orientational predications, such as The
ball is to the right of the tree are based on focus chains embedded in view
maps (Langacker 2000). Focus chains are connected instances of trajector/landmark relations. In the example sentence, the owner of the view
map, whether speaker or other participant in the discourse, could only be
known from context.
Object maps are cognitive models of the spatial qualities of things, such
as the shape of a vase or the parts of an automobile. Utterances that imply
certain orientations of the bodies of speakers and participants or of the
speech situation itself, are termed deictic. Large scale orientations, such as
cardinal directions, uphill/downhill, upstream/downstream, and inland/
ocean are qualities of macro-maps, which differ from ordinary object
maps only in existing on a grander scale and consequently having greater
stability and availability for reference. It is only in this sense that they are
absolute. Given this framework, one would expect different languages
and cultures to assign orientational qualities to objects of sizes varying
from parts of the body, whether human or animal, at the small end of the
scale, to landforms, rivers, the prevailing weather patterns, and the cosmos including its supernatural terrain, at the large end of the scale. Some,
like the Guugu Yimithirr, may use only a single macro-map, but it is probably more typical to for spatial language to recruit multiple types of maps.
The Belhare of Eastern Nepal have four kinds of orientational maps,
which have been labeled ecomorphic, personmorphic, geomorphic, and physiomorphic (Bickel 1997). Bickel (2000: 178) has argued strongly for the
mutual influence of grammar and cultural practices:
Indeed, the distinction between various readings of up and down in language is recapitulated almost to perfection in cultural practices: the difference
between hill-based (geomorphic) or verticality-based uses of linguistic morphemes, for instance, exactly parallels the difference between the practices mentioned above and another set of practices which are performed along vertical,
286
Gary B. Palmer
sky-ward trajectories. Trajectories of the latter sort are found, for instance, in
architectural elevations of religious buildings or in upward gestures as symbols
of gratitude. They are culturally most dense in the notion of raising ones soul
(samet phokma) in shamanist curing rituals. Shock and certain illnesses can
cause ones soul to fall down, and without the necessary remedies, to get lost.
All curing of such maladies and conditions is centered around invoking an upward trajectory.
Bickels argument is explicitly based upon the Whorfian theory that grammar is related to habitual thought and behavior. It resembles Lakoff s
(1987) argument that Djirbal classifiers relate to domains of experience,
i.e. the myth of the sun and the moon and my own, discussed above, that
the central categories in Shona noun classifiers are cultural scenarios of
domestic life and ritual practices. In my approach, cultural scenarios are
defined as schematizations of social actions and events (Palmer 1996).
Bickel offered this general statement of his theory:
The relationship between language and culture thus turns from a relationship
between modes of expression into a relationship between the semantic constraints imposed by a specific linguistic category and the habitus of socio-cultural practice within which language use is embedded. Together, the boundaries
of linguistic signs that segment experience, and the habitus that schematizes
practice, channel and guide cognition into a heightened attention to specific experiences and actions. (Bickel 2000: 165)
Bickels observations suggest again that a large domain of grammar is recruited to cue and sustain central cultural/cognitive frames of reference.
The many observations by linguists and social anthropologists of the interpenetration of spatial, ritual, and social categories become explainable
in terms of this function. To the extent that the linguistic correlation of
domain structures is shared in the community of speakers, consensus is
fostered, but structural correlation also enables the harmonization of actions while leaving much knowledge in the community uncommunicated
and unshared. The problem for the linguist is to discover, through elicitation and ethnography, the relations between figurative usages of spatial
expressions, which evoke scenarios and evaluations, and their actual social contexts. Like reduplications and classifiers, closed class spatial terms
acquire affective or evaluative connotations (Lakoff and Johnson 1980;
Talmy 2005).
Given the extreme degree of interesting variation evident in the spatial
grammar of the few languages that have been studied and the thousands
of languages remaining, it is hoped that cognitive linguists and cognitive
287
6.
288
Gary B. Palmer
289
seasonal cycles of migration, childbirth, childcare, real and mock captures, and narrow escapes from predators. Salient images deriving from
these activities were presented to audiences in ceremonials consisting of
mimetic performances with dance, music, and vocalizations that created a
forcing environment for language genesis. Highly iconic vocalizations of
performers and audience responses were mapped to mimetic imagery by
something akin to fast mapping that we see in language development in
children (Nelson 1996). Ceremonial mimesis also motivated the narrative
combination of proto-symbols in constructions and their entrenchment
and standardization through repetition and mimicry.
Enhanced intersubjectivity involving joint attention frames, intention
reading, and conceptual role substitutions was a key to symbol genesis.
Intersubjectivity was also supported by ceremonial contexts. Performers
and audiences shared joint attention frames and understood performances to have communicative intentions. The first linguistic constructions
grammaticalized16 to produce proto-morphology, and consequently, protogrammar, at which point, selection for recombinance became possible
(Talmy 2004). Speech and song enhanced the effectiveness and status of
skillful and dominant performers, so that sexual selection favored the
genes of good speakers and singers. Symbols generated first in other activities, such as child care, grooming, tool making, and hunting, could be
brought to ceremonials, shared more widely, and standardized.
Thus, cognitive linguistics provides a new perspective on the emergence
of language. But applying the new perspective requires supporting information derived from archaeological research into proto-human cultures and comparative studies of the vocalizations, proto-cultures, and
cognition of primates.
7.
Conclusions
290
Gary B. Palmer
the biological, experiential, and cultural dimensions of language. A cognitive linguistic model of experiential process in Tagalog was compared
to DAndrades model for English, revealing fundamental similarities
and differences between the two languages and cultures. In both languages, thoughts are distinguished from emotions and feelings, but the
theories of cognitive process differ. It seems incontrovertible that language communities manifest a high degree of diversity, but the problem
of how diversity is balanced with consensus needs further clarification
and research. Clues to how grammars participate in the integration of diversity in community knowledge are found in studies of complex categories and the language of spatial orientation. Rapid advances in the
theory of complex categories are leading to useful descriptions of grammatical domains such as reduplication and classifier systems. Classifier
systems narrow the focus of discourse by cueing common frames of reference that may intentionally evoke recall of specific events. The amazing
relativity of spatial language can be described with a new framework consisting of cognitive maps, including object maps and viewer maps, which
contain focus chains. Conventional spatial expressions may, like classifiers, evoke larger sociocultural frames of reference. Finally, it is proposed that verbal symbols and proto-language originated with Homo heidelbergensis in the forcing environment of proto-ceremonials consisting
of mimetic presentations of salient cultural and natural scenarios. This
theory of language origins is supported by cognitive theory, archaeology,
and paleontology.
Cognitive linguistics is proving its value by producing insightful applications to the descriptions, analyses, and comparisons of both western
and non-western languages (e.g. Casad and Palmer 2003). Its value has
been seen by a number of linguistic and/or cognitive anthropologists.17 I
have found in my own research involving three non-Indo-European language families (Salishan, Western Austronesian, and Bantu) that a version of cognitive linguistics that emphasizes the cultural origins of linguistic categories enables cross-linguistic comparisons and insights into
both grammar and world view that were unavailable to other paradigms
in anthropology or linguistics. Conversely, it appears from the vantage
point of anthropology that cognitive linguists could do more to incorporate the frameworks of cognitive anthropology, such as those of ethnoscience, and they could make more use of ethnography to discover cultural schemas and scenarios, especially those that have non-linguistic as
well as linguistic expressions, for these will provide additional evidence
for proposed cognitive linguistic models.
291
Notes
1. Gary B. Palmer is Professor Emeritus at the University of Nevada, Las Vegas.
The author thanks Ed Shoben for his comments on the manuscript.
2. I am indebted to Gitte Kristiansen for pointing out the use of the culture concept in Lakoff and Kvecses (1987) and for reminding me of recent conference
sessions which all reflect a growing interest in culture or language-internal
variation: Language, Culture, and Cognition: An International Conference
on Cognitive Linguistics, Catholic University of Portugal-Braga, July 1618,
2003; Session on Lectal Variation and Conceptualisation, International Cognitive Linguistics Association Conference, Seoul, Korea, July 1722, 2005;
Language, Culture and Mind: Integrating Perspectives and Methodologies in
the Study of Language, University of Portsmouth, July1820, 2004, and Paris
July 1720, 2006.
3. At the occasion of the 100th anniversary of Humboldts death several symposia revived the popularity of linguistic relativity. An example is the 26th International LAUD Symposium, Humboldt and Whorf Revisited: Universal and
Culture-Specific Conceptualizations in Grammar and Lexis, Mlheim/Ruhr,
April 14, 1998. See Niemeier and Dirven (2000).
4. Keesing (1987: 369) listed as synonyms ethnoscience, ethnographic semantics,
and the new ethnography. One also sees the term ethnosemantics. I am unaware of
any clear boundaries between ethnosemantics and ethnoscience. Ethnoscience
has been harshly criticized (e.g. Keesing 1987), but its methods and assumptions
are not hard to find in recent work, such as that of Lpez et al. (1997).
5. For canonical examples of componential and taxonomic analysis, see Tyler
(1969) and Casson (1981). For a recent example, see Kronenfeld (1996). For
discussions of cultural logic in terminological systems, see Lakoff (1987: 24)
and Palmer (1996: 9899) on what Lakoff called generative categories. For an
excellent extended presentation of the theoretical framework of ethnoscience,
see DAndrade (1995a). See also, Colby (1996).
6. Human cognition is built for culture, and thus built for enculturated variation (Levinson 1996b: 177).
7. For further discussion of color terms, including both basic and non-basic
terms and refinements to the theory based on more recent research, see
Palmer (1996: 8088) and Foley (1997: 150165).
8. dr: directional; ex: existential; gn: genitive; inv: inverse; lg: ligatur; spc: specific.
9. I am indebted to Ren Dirven for calling this to my attention.
10. Any knowledge existing outside of human minds, say in books or other artefacts, can not function as knowledge without the existence of human minds
that comprehend the design of the artefactual knowledge.
11. Matrices and their eigenvalues fall under the topic of linear algebra. The formal mathematical derivation of the model appears in Batchelder and Romney
(1988), but this in turn is premised on derivations of supporting formulas
found in other sources. I am grateful to Roy Ogawa for leading me through the
linear algebra.
292
Gary B. Palmer
293
References
Alverson, Hoyt
1991
Metaphor and experience: looking over the notion of image-schema. In
Beyond Metaphor: The Theory of Tropes in Anthropology, James W.
Fernandez (ed.), 94117. Stanford: Stanford University Press.
Arbib, Michael
2005
From monkey-like action recognition to human language: An evolutionary framework for neurolinguistics. Behavioral and Brain Sciences
28 (2): 105124.
Armstrong, David F., William C. Stokoe, and Sherman E. Wilcox
1995
Gesture and the Nature of Language. Cambridge: Cambridge University Press.
Bachelder, William H., and A. Kimball Romney
1988
Test theory without an answer key. Psychometrika 53: 7192.
Basso, Keith H.
1984
Stalking with stories: names, places and moral narratives among the
Western Apache. In Text, Play and Story, Edward M. Bruner (ed.),
1955. Washington, D.C.: American Ethnological Society. Reprinted
in Keith H. Basso, 1990, 99137.
Berlin, Brent
1992
Ethnobiological Classification: Principles of Categorization of Plants and
Animals in Traditional Societies. Princeton: Princeton University Press.
Berlin, Brent, and Paul Kay
1969
Basic Color Terms: Their Universality and Evolution. Berkeley: University of California Press.
Bickel, Balthazar
1997 Spatial operations in deixis, cognition, and culture: where to orient
oneself in Belhare. In Language and Conceptualization, Jan Nuyts, and
Eric Pederson (eds.), 4683. Cambridge: Cambridge University Press.
2000
Grammar and social practice: On the role of culture in linguistic
relativity. In Evidence for Linguistic Relativity, Susanne Niemeier, and
Ren Dirven (eds.), 161191. Amsterdam/Philadelphia: Benjamins.
Bickerton, Derek
1998
Catastrophic evolution: The case for a single step from protolanguage
to full language. In Approaches to the Evolution of Language: Social and
Cognitive Bases, James R. Hurford, Michael Studdert-Kennedy, and
Chris Knight (eds.), 341358. Cambridge: Cambridge University Press.
Biesele, Megan
1998
Women Like Meat: The Folklore and Foraging Ideology of the Kalahari
Ju/Hoan. Johannesburg, South Africa: Witwatersrand University
Press / Bloomington: Indiana University Press.
Boas, Franz
1966
[1911] Introduction to the Handbook of American Indian Languages.
Lincoln: University of Nebraska Press.
294
Gary B. Palmer
Bohn, Ocke-Schwen
2000
Linguistic relativity in speech perception: An overview of the influence
of language experience on the perception of speech sounds from infancy
to adulthood. In Evidence for Linguistic Relativity, Susanne Niemeier,
and Ren Dirven (eds.), 128. Amsterdam/Philadelphia: Benjamins.
Borofsky, Robert
1994
On the knowledge and knowing of cultural activities. In Assessing Cultural Anthropology, Robert Borofsky (ed.), 331348. New York:
McGraw-Hill.
Brown, Cecil H.
1984
Language and Living Things: Uniformities in Folk Classification and
Naming. New Brunswick, N.J.: Rutgers University Press.
Brugman, Claudia
1983
The use of body-part terms as locatives in Chalcatongo Mixtec. In Survey of California and Other Indian Languages, Alice Schlichter, Wallace
Chafe, and Leanne Hinton (eds.), 235290. Studies in Mesoamerican
Linguistics. Report no. 4. Berkeley: Survey of California and Other Indian Languages.
1988
The Story of Over: Polysemy, Semantics, and Structure of the Lexicon.
New York: Garland Publishing.
Bhler, Karl
1982
[1934] The deictic field of language and deictic words. Reprinted in
Speech, Place and Action, Robert J. Jarvella, and Wolfgang Klein (eds.),
930. Chichester: John Wiley & Sons.
Carroll, Mary
1997
Changing places in English and German: language-specific preferences
in the conceptualization of spatial relations. In Language and Conceptualization, Jan Nuyts, and Eric Pederson (eds.), 137161. Cambridge:
Cambridge University Press.
Casad, Eugene
1988
Conventionalization of Cora locationals. In Topics in Cognitive Linguistics, Brygida Rudzka-Ostyn (ed.), 345378. Amsterdam/Philadelphia: Benjamins.
1993
Locations, paths, and the Cora verb. In Conceptualizations and
Mental Processing in Language, Richard A. Geiger, and Brygida
Rudzka-Ostyn (eds.), 593645. Berlin/New York: Mouton de Gruyter.
Casad, Eugene, and Ronald Langacker
1985
Inside and outside in Cora grammar. International Journal of
American Linguistics 51: 247281.
Casad, Eugene, and Gary B. Palmer (eds.)
2003
Cognitive Linguistics and Non-Indo-European Languages. Berlin/New
York: Mouton de Gruyter.
Casson, Ronald W. (ed.)
1981
Language, Culture and Cognition: Anthropological Perspectives. New
York: Macmillan.
295
Colby, Benjamin N.
1996
Cognitive Anthropology. In Encyclopedia of Cultural Anthropology,
Volume 1, David Levinson, and Melvin Ember (eds.), 209215. New
York: Henry Holt and Company.
Conklin, Harold
1964
Hanuno color categories. In Language in Culture and Society: A
Reader in Linguistics and Anthropology, Dell Hymes (ed.), 189192.
New York: Harper and Row.
Croft, William
2001
Radical Construction Grammar: Syntactic Theory in Typological Perspective. Oxford/New York: Oxford University Press.
Crystal, David
1980
A Dictionary of Linguistics and Phonetics. 3rd Ed. Oxford: Basil Blackwell.
Damasio, Antonio R.
1994
Descartes Error: Emotion, Reason and the Human Brain. London:
Macmillan.
DAndrade, Roy G.
1987
A folk model of the mind. In Cultural Models in Language and Thought,
Dorothy Holland, and Naomi Quinn (eds.), 112148. Cambridge:
Cambridge University Press.
1995a The Development of Cognitive Anthropology. Cambridge: Cambridge
University Press.
1995b Moral Models in Anthropology. Current Anthropology 36 (3): 399408.
Deacon, Terrence W.
1997
The Symbolic Species: The Co-Evolution of Language and the Brain.
New York/London: W. Norton & Company.
Dewey, John
1958
[1925] Reprint. Experience and Nature. New York: Dover Publications.
Original edition, Chicago/ London: Open Court, 1925. Cited in Johnson
(2006).
Dirven, Ren
2005
Major strands in cognitive linguistics. In Cognitive Linguistics. Internal
Dynamics and Interdisciplinary Interaction, Francisco J. Ruiz de Mendoza Ibaez, and M. Sandra Pena Cervel (eds.), 1768. CLR 32. Berlin/New York: Mouton de Gruyter.
Dirven, Ren and Ralf Prings (eds.)
2002
Metaphor and Metonymy in Comparison and Contrast,. Berlin/New
York: Mouton de Gruyter.
Dirven, Ren, and Marjolijn Verspoor (eds.)
2004
Cognitive Exploration of Language and Linguistics. Second Revised Edition.
(Cognitive Linguistics in Practice 1). Amsterdam/Philadelphia: Benjamins.
Donald, Merlin
1991
Origins of the Modern Mind: Three Stages in the Evolution of Culture
and Cognition. Cambridge, Mass.: Harvard University Press.
296
Gary B. Palmer
Dunbar, Robin
1998
Theory of mind and the evolution of language. In Approaches to the
Evolution of Language: Social and Cognitive Bases, James R. Hurford,
Michael Studdert-Kennedy, and Chris Knight (eds.), 92110. Cambridge: Cambridge University Press.
Enfield, Nick J. (ed.)
2002
Ethnosyntax: Explorations in Grammar and Culture. Oxford: Oxford
University Press.
Ehrsson, H. Henrik, Charles Spence, and Richard E. Passingham
2004
Thats my hand! Activity in premotor cortex reflects feeling of ownership of a limb. Science 305: 875877.
Falk, Dean
2004
Prelinguistic evolution in early hominins: Whence motherese? Behavioral and Brain Sciences 27: 491541.
Fauconnier, Gilles, and Mark Turner
2002
The Way We Think: Conceptual Blending and the Minds Hidden Complexities. New York: Basic Books.
Foley, William A.
1997
Anthropological Linguistics: An Introduction. Oxford: Blackwell.
Friedrich, Paul
1979
Language, Context, and Imagination. Stanford: Stanford University Press.
1971
The Tarascan Suffixes of Locative Space. Bloomington: Indiana University / The Hague: Mouton & Co.
Geeraerts, Dirk, and Stefan Grondelaers
1995
Looking back at anger: Cultural traditions and metaphorical patterns.
In Language and the Cognitive Construal of the World, John R. Taylor,
and Robert E. MacLaury (eds.). 153179. Berlin/New York: Mouton
de Gruyter.
2004
Whats in a word: Lexicology. In Cognitive Exploration of Language and
Linguistics. Second Revised Edition, Ren Dirven and Marjolijn Verspoor (eds.), (Cognitive Linguistics in Practice 1), 2550. Amsterdam/
Philadelphia: Benjamins.
Goddard, Cliff
2002
The search for the shared semantic core of all languages. In Meaning
and Universal Grammar Theory and Empirical Findings, Cliff Goddard,
and Anna Wierzbicka (eds.), 540. Volume I. Amsterdam/Philadelphia: Benjamins.
2003
Thinking across languages and cultures: Six dimensions of variation.
Cognitive Linguistics 14 (2): 109139.
Goodenough, Ward H.
1965
Yankee kinship terminology: A problem in componential analysis.
American Anthropologist 67/5, pt. 2, (Special Publication): 259287.
Reprinted in Cognitive Anthropology, Stephen A. Tyler (ed.), 255288.
New York: Holt, Rinehart & Winston, 1969.
297
Gordon, Peter
2004
Numerical cognition without words: Evidence from Amazonia.
Science 306: 496499.
Hallowell, A. Irving
1955
Ch. 9, Cultural factors in spatial orientation. Culture and Experience,
184202. New York: Schocken Books.
Haviland, John C.
1993
Anchoring, iconicity, and orientation in Guugu Yimithirr pointing gestures. Journal of Linguistic Anthropology 3(1): 345.
1996
Projections, transpositions, and relativity. In Rethinking Linguistic
Relativity, John J. Gumperz, and Stephen C. Levinson (eds.), 271323.
Cambridge: Cambridge University Press.
Haugen, Einer
1969
The semantics of Icelandic orientation. In Cognitive Anthropology,
Stephen A. Tyler (ed.), 330342. New York: Holt, Rinehart, and Winston.
Haxby, James V., M. Ida Gobbini, Maura L. Furey, Alumit Ishai, Jennifer L.
Schouten, and Pietro Pietrini
2001
Distributed and overlapping representations of faces and objects in
ventral temporal cortex. Science 293: 24252430.
Heine, Bernd
1997
Cognitive Foundations of Grammar. Oxford: Oxford University Press.
Hewes, Gordon
1973
Primate communication and the gestural origin of language. Current
Anthropology 14: 524.
1994
The gestural origin of language and new neurological data. Studies in
Language Origins 3: 293307.
Hockett, Charles F., and Robert Ascher
1964
The human revolution. Current Anthropology 5:13347.
Hopper, Paul, and Elizabeth Traugott
1993
Grammaticalization. Cambridge: Cambridge University Press.
Humboldt, Wilhelm von
1972
[1836] translated by George C. Buck, and Frithjof A. Raven. Linguistic
Variability and Intellectual Development. Philadelphia: University of
Pennsylvania Press (Orig. 1836. ber die Verschiedenheit des menschlichen Sprachbaus und ihren Einfluss auf die geistige Entwicklung des
Menschengeschlechts. Berlin: Royal Academy).
Hunn, Eugene
1990
Nchi-wna, the big river: Mid-Columbia Indians and their land / Eugene S. Hunn with James Selam and family. Seattle: University of Washington Press.
Hutchins, Edwin
1996
Cognition in the Wild. Cambridge, Mass./London: The MIT Press.
Jespersen, Otto
1922
Language: Its Nature, Development, and Origin. London: George Allen
& Unwin Ltd/New York: Henry Holt and Company.
298
Gary B. Palmer
Johansson, Sverker
2005
Origins of Language: Constraints on Hypotheses. Amsterdam/Philadelphia: Benjamins.
Johnson, Mark
2006
The philosophical significance of image-schemas. In From Perception
to Meaning: Image Schemas in Cognitive Linguistics, Hampe, Beate (in
cooperation with Joseph E. Grady) (ed.). (Cognitive Linguistics Research 29). Berlin/New York: Mouton de Gruyter.
Kay, Paul
1966
Comment on Ethnographic semantics: A preliminary survey. Current
Anthropology 7 (1): 2023. Reprinted in Cognitive Anthropology, Stephen
A. Tyler (ed.), 7892. New York: Holt, Rinehart & Winston, 1969.
Keesing, Roger
1987
Models folk and cultural: Paradigms regained. In Cultural Models
in Language and Thought, Dorothy Holland, and Naomi Quinn (eds.),
369393. Cambridge: Cambridge University Press.
1992
Anthropology and linguistics. In Thirty Years of Linguistic Evolution:
Studies in Honor of Ren Dirven on the Occasion of His Sixtieth Birthday, Martin Ptz (ed.), 593609. Philadelphia: Benjamins.
1997
Constructing space in Kwaio (Solomon Islands). In Referring to Space:
Studies in Austronesian and Papuan Languages, Gunter Senft (ed.),
127141. Oxford: Clarendon Press.
King, Barbara J.
1996
Syntax and language origins. Language and Communication 16
(2):193203.
Klein, Richard G.
1989
The Human Career: Human Biological and Cultural Origins, 2nd ed.
Chicago: University of Chicago Press.
Knight, Chris
1998
Ritual/speech coevolution: A solution to the problem of deception. In
Appraoches to the Evolution of Language, James R. Hurford, Michael
Studdert-Kennedy, and Chris Knight (eds.), 6891. Cambridge: Cambridge University Press.
Koerner, E. F. Konrad
2000
Towards a full pedigree of the Sapir-Whorf hypothesis: From Locke
to Lucy. In Explorations in Linguistic Relativity, Martin Ptz, and Marjolijn Verspoor (eds.), 124. Amsterdam/Philadelphia: Benjamins.
Kvecses, Zoltn
2000
Metaphor and Emotion: Language, Culture, and Body in Human Feeling. Cambridge: Cambridge University Press.
2005
Metaphor in Culture: Universality and Variation. Cambridge: Cambridge University Press.
Kvecses, Zoltn, and Gary B. Palmer
1999
Language and emotion concepts: What experientialists and social constructionists have in common. In Languages of Sentiment, Gary B.
299
300
Gary B. Palmer
Lee, Penny
1996
The Whorf Theory Complex: A Critical Reconstruction. Amsterdam/
New York: Benjamins.
2003
Feeling of the mind in talk about thinking in English. Cognitive Linguistics 14/2,3: 221249.
Levinson, Stephen C.
1996a Frames of reference and Molyneuxs question: Crosslinguistic evidence.
In Language and Space, Paul Bloom Mary A.Peterson, Lyn Nadel, and
Merril Garrett (eds.), 109170. Cambridge, MA: The MIT Press.
1996b Relativity in spatial conception and description. In Rethinking Linguistic Relativity, John J. Gumperz, and Stephen C. Levinson (eds.),
177202. Cambridge: Cambridge University Press.
1997
From outer to inner space: Linguistic categories and non-linguistic
thinking. In Language and Conceptualization, Jan Nuyts, and Eric Peterson (eds.), 1345. Cambridge: Cambridge University Press.
Li, Charles N.
2002a Some issues concerning the origin of language. In Complex Sentences in
Grammar and Discourse, Joan Bybee, and Michael Noonan (eds.),
203221. Amsterdam/Philadelphia: Benjamins.
2002b Missing links, issues and hypotheses in the evolutionary origin of language. In The Evolution of Language out of Pre-language, Talmy Givn,
and Bertram F. Malle (eds.), 83106. Amsterdam/Philadelphia: Benjamins.
Livingstone, Frank B.
1973
Did the Australopithecines sing? Current Anthropology 14 (12): 2529.
Lounsbury, Floyd
1964
A formal account of the Crow- and Omaha-type kinship terminologies.
In Explorations in Cultural Anthropology, W. H. Goodenough (ed.),
351394. New York: McGraw-Hill. Reprinted in Cognitive Anthropology, Stephen A. Tyler (ed), 212254. New York: Holt, Rinehart & Winston, 1969.
Lpez, Alejandro, Scott Atran, and Edward E. Smith
1997
The tree of life: Universal and cultural features of folkbiological taxonomies and inductions. Cognitive Psychology 32: 251295.
Lucy, John
1992a Language Diversity and Thought: A Reformulation of the Linguistic
Relativity Hypothesis. Cambridge: Cambridge University Press.
1992b Grammatical Categories and Cognition: A Case Study of the Linguistic
Relativity Hypothesis. Cambridge: Cambridge University Press.
Lutz, Catherine A.
1988
Unnatural Emotions: Everyday Sentiments on a Micronesian Atoll and
their Challenge to WesternTtheory. Chicago: University of Chicago Press.
MacLaury, Robert E.
1992
From brightness to hue: An explanatory model of color-category evolution. Current Anthropology 33: 137186.
301
Marwick, Ben
2003
Pleistocene exchange networks as evidence for the evolution of language. Cambridge Archaeological Journal 13 (1): 6781.
Mathews, Holly F.
1987
Intracultural variation in beliefs about gender in a Mexican community. American Behavioral Scientist 31: 219233.
Nelson, Katherine
1996
Language in Cognitive Development: The Emergence of the Mediated
Mind. Cambridge: Cambridge University Press.
Niemeier, Susanne, and Ren Dirven (eds.)
2000
Evidence for Linguistic Relativity. Amsterdam/Philadelphia: Benjamins.
Nishimura, Takeshi, Akichika Mikami, Juri Suzuki, and Tetsuro Matsuzawa
2003
Descent of the larynx in chimpanzee infants. PNAS 100 (12): 69306933.
Nuyts, Jan, and Eric Pederson (eds.)
1997
Language and Conceptualization. Cambridge: Cambridge University
Press.
Occhi, Debra J.
1999
Sounds of the heart and mind. In Languages of Sentiment: Cultural
Constructions Of Emotional Substrates, Gary B. Palmer, and Debra J.
Occhi (eds.), 151170. Amsterdam/Philadelphia: Benjamins.
Occhi, Debra J, Gary B. Palmer, and Roy H. Ogawa
1993
Like Hair, or Trees: Semantic Analysis of the Coeur dAlene Prefix ne
amidst. In Proceedings of the 1993 Annual Meeting of the Society for
the Study of the Indigenous Languages of the Americas, July 24, 1993,
and the Hokan-Sioux Conference, July 3, 1993, Columbus, Ohio. Report 8, Margaret Langdon (ed.), 4058. Survey of California and Other
Indian Languages. Berkeley, CA: Berkeley Linguistics Department.
Ogawa, Roy H., and Gary B. Palmer
1999
Langacker Semantics for Three Coeur dAlene Prefixes Glossed as
On. In Issues in Cognitive Linguistics, Leon de Stadler, and Christopher Eyrich (eds.), 165224. Berlin/NewYork: Mouton de Gruyter,.
Palmer, Gary B.
1996
Toward a Theory of Cultural Linguistics. Austin: University of Texas
Press.
2003a Talking about thinking in Tagalog. Cognitive Linguistics 14 (2): 251280.
2003b The Tagalog prefix category PAG-: Metonymy, polysemy, and voice. In
Cognitive Linguistics and Non-Indo-European Languages, Gene Casad,
and Gary B. Palmer (eds.), 193221. Berlin/New York: Mouton de
Gruyter.
In press Cognitive Linguistics and Anthropological Linguistics, Chapter 40, In
Handbook of Cognitive Linguistics, Dirk Geeraerts, and Hubert Cuyckens (eds.). Oxford: Oxford University Press.
Palmer, Gary B. and Claudia Woodman
1999
Ontological Classifiers as Polycentric Categories, as Seen in Shona
Class 3 Nouns. In Explorations in Linguistic Relativity, Martin Ptz,
302
Gary B. Palmer
303
can Anthropologist 67/5, pt. 2, (Special Publication): 288308] Reprinted in Cognitive Anthropology, Stephen A. Tyler (ed.), 288311.
New York: Holt, Rinehart & Winston, 1969.
Senft, Gunter, (ed.)
1997
Referring to Space: Studies in Austronesian and Papuan Languages. Oxford: Clarendon Press:.
Sharifian, Farzad
In press Distributed, emergent cultural cognition, conceptualisation, and language. In Body, Language, and Mind (Vol. 2): Sociocultural Situatedness, Roslyn M. Frank, Ren Dirven, Tom Ziemke, and Enrique Bernrdez (eds.). Berlin/New York: Mouton de Gruyter.
Sharifian, Farzad, and Ahmad R. Lofti
2003
Rices and Waters: The mass/count distinction in Modern Persian.
Anthropological Linguistics 45(2): 226244.
Shore, Bradd
1996
Culture in Mind: Cognition, Culture, and the Problem of Meaning. Oxford: Oxford University Press.
Sinha, Chris
2004
The evolution of language from signals to symbols to system. In Evolution of Communication Systems: A Comparative Approach, D. Kimbrough Oller, and Ulrike Griegbel (eds.), 217236. (Vienna Series in
Theoretical Biology). Cambridge, Mass. The MIT Press.
Sinha, Chris, and Kristine Jensen de Lpez
2000
Language, culture and the embodiment of spatial cognition. Cognitive
Linguistics 11(1): 1741.
Talmy, Leonard
1983
How language structures space. In Spatial Orientation: Theory, Research and Application, Herbert L. Pick, and Linda P. Acredolo (eds.),
225320. New York: Plenum Press.
2005
The fundamental system of spatial schemas in language. In From Perception to Meaning: Image Schemas in Cognitive Linguistics, Beate
Hampe (in cooperation with Joseph E. Grady) (ed.), 199234. Berlin/
New York: Mouton de Gruyter.
In press Recombinance in the evolution of language. In Proceedings of the 39th
Annual Meeting of the Chicago Linguistic Society: The Panels, Jonathon E. Cihlar, David Kaiser, Irene Kimbara, and Amy Franklin (eds.).
Chicago: Chicago Linguistic Society.
Tomasello, Michael
1999
The Cultural Origins of Human Cognition. Cambridge, MA: Harvard
University Press.
Trabant, Jrgen
2000
How relativistic are Humboldts Weltansichten? In Explorations in
Linguistic Relativity, Martin Ptz, and Marjolijn Verspoor (eds.),
2544. Amsterdam/Philadelphia: John Benjamins.
304
Gary B. Palmer
Tuggy, David
2003
Reduplication in Nahuatl: Iconicities and paradoxes. In Cognitive Linguistics and Non-Indo-European Languages, Eugene H. Casad, and
Gary B. Palmer (eds.), 91133. Berlin/New York: Mouton de Gruyter.
Tyler, Stephen A. (ed.)
1969
Cognitive Anthropology. New York: Holt, Rinehart and Winston.
Ungerer, Friedrich, and Hans-Jrg Schmid
1996
An Introduction to Cognitive Linguistics. London/New York: Longman.
Wallace, Anthony F. C
1961
Culture and Personality. New York: Random House.
Whorf, Benjamin
1956
Language, Thought and Reality: Selected Writings of Benjamin Lee
Whorf. John B. Carroll (ed.), Cambridge, MA: The M.I.T. Press.
Wierzbicka, Anna
1990
The meaning of color terms: Semantics, culture, and cognition. Cognitive Linguistics 1: 99150.
1996
Semantics: Primes and Universals. New York: Oxford University Press.
Wong, Kate
2004
Baby talk beginnings: Infant pacification may have led to the origin of
language. Scientific American 291 (1): 30.
Yao, Shunchiu
1989
Moulded gestures and guided syntax: Scenario of a linguistic breakthrough. In Studies in Language Origins, Volume 1, Jan Wind, Edward
G. Pulleyblank, Eric de Grolier, and Bernard H. Bichakjian (eds.),
3442. Amsterdam/Philadelphia: John Benjamins.
Yu, Ning
1998
The Contemporary Theory of Metaphor. A Perspective from Chinese.
Amsterdam/Philadelphia: Benjamins.
Zlatev, Jordan
2002
Mimesis: The missing link between signals and symbols in phylogeny
and ontology? In Mimesis, Sign and Language Evolution, Anneli Pajunen (ed.), Publications in General Linguistics 3, University of Turku,
Finland.
n.d
Embodiment, language, and mimesis. Unpublished ms. in possession of
the author.
305
Abstract
In this chapter we examine the contributions that Cognitive Linguistics (CL) can
make to second or foreign language teaching. A lot of CL premises are found to be
fully compatible with those of modern strands of education-oriented applied linguistics. However, the CL feature that seems to have the greatest potential as a
source for complementing current language pedagogy is its quest for linguistic
motivation, i.e. for explainable meaning-meaning connections (e.g. polysemy),
meaning-form connections (e.g. iconicity) and form-form connections (e.g. alliteration). Presenting segments of language as motivated is likely to be beneficial for
learners as it may enhance comprehension, retention, cultural awareness and positive affect. We survey the calls that have been made for the introduction of CL insights in language teaching and evaluate the empirical evidence of the pedagogical
effectiveness of CL approaches. Although most controlled experiments reported
so far have tended to be small-scale, taken collectively they are beginning to constitute a fairly robust body of evidence in favour of CL-inspired pedagogy. It needs
to be acknowledged, however, that the evidence collected so far is largely confined
to the pedagogical exploitation of figurative thought, and this mostly with regard
to polysemous words (especially prepositions and particles) and idiomatic expressions. We conclude by proposing an agenda for future research.
Keywords: applied linguistics; second/foreign language pedagogy; linguistic motivation; figurative thought; empirical evidence
1.
306
1. Language is an integral part of cognition as a whole, rather than a separate and unique faculty (Langacker 1987). From this it follows that
language learning is likely to benefit from the application of established, relatively general theories of cognitive processing. especially
those which concern memory. These include dual coding theory (Clark
and Paivio 1991; Paivio 1971, 1986), which holds that association
through figurative thought of verbal information with a mental image
facilitates recall, and trace theory, which holds that repetition of encounters with a given linguistic instantiation strongly tends to entrench
its traces in memory (Baddeley 1990; Cohen, Eysenck and LeVoi 1986;
Squire and Kandel 2000).
In applied linguistics (AL), explicit discussions of dual coding and
trace theory do not abound but they are widely assumed. Stevick (1986)
offers an early and still perhaps the most explicit account of applications of both theories in L2 pedagogy.
2. In a dynamic usage-based model of language (Langacker 1988, 2000;
Tomasello 2000), there is no innate language-acquisition faculty which
leads to essentially the same results in L1 acquisition (being most importantly the acquisition of syntax and phonology) regardless of
whether input is very rich or rather impoverished and imbalanced. Instead, acquisition is markedly influenced throughout early childhood
and beyond by quantity and quality of input. Frequency of encounter is
key if a language user encounters a linguistic unit often enough, it
eventually becomes a standard item in the learners linguistic inventory.
Concomitant with such learning is the formation in the mind of an
over-arching schema which can be said to sanction its instantiations.
At the level of groups of language users, frequency of occurrence of
understood linguistic forms (be they morphemes, words, phrases or
patterns) determines the likelihood of their becoming entrenched in a
language communitys linguistic repertoire.
A number of applied linguists have adopted a usage-based model of
language (e.g. Ellis 1998, 2005; Skehan 1998), along with the explicit
belief that the frequency of expressions and patterns in input has an effect on acquisition in L2, as evidenced by features of inter-language
(e.g. Ellis 2002).
3. The grammar-lexis dichotomy at the heart of both structural and generative linguistics is rejected. Instead, language is viewed as a non-dichotomous, structured inventory of conventional symbolic units of
varying complexity (Langacker 1987, 1991). This new conception of
307
language as a continuum of meaningful forms has led to a rapid increase in research into semi-fixed multiword expressions (e.g. idioms),
which had hitherto received the attention of relatively few linguists (e.g.
Makkai 1972) owing to the dubious legitimacy of these forms within
the then widely accepted lexis-grammar dichotomy.
No doubt in part thanks to fruitful computer-aided corpus studies
(e.g. Sinclair 1991), a growing number of applied linguists have become
active in the study of collocation in general and of multiword units in
particular (e.g. Howarth 1998; Nattinger and DeCarrico 1992; Pawley
and Syder 1983; Schmitt 2004; Wray 2002), with the contribution of
such units to oral fluency being a topic of particular interest (e.g. Boers,
Eyckmans, Kappel, Stengers and Demecheleer Forthcoming; Oppenheim 2000; Segalowitz 2000).
4. The distinction between linguistic, or dictionary, meaning (the theoretical ideal of which is word meanings that are equivalent to necessary
and sufficient conditions) and wider, encyclopaedic meaning is a fallacy (Haiman 1980; Langacker 1987). Rather, words and patterns
too are conceived of as nodes in elaborate semantic networks such
that activation of one node (by a usage event) may trigger the activation
of many other nodes in the network. Put simply, words and other constructions in the mental lexicon carry many more associations (ranging
from the culturally shared to the idiosyncratic) than earlier theory
allowed.
In AL, this view corresponds to language instruction that aims at:
a) depth (in addition to breadth) of linguistic knowledge (e.g. Read
2004; Vermeer 2001)
b) an enhanced general language awareness on the part of the language
learner (e.g. James and Garrett 1995)
c) pedagogical exploration of the inter-subjectivity of personal word
meanings (e.g. Morgan and Rinvolucri 2004: 118).
5. Figuration (especially metaphor and metonymy) plays an immense role
in the semantics of natural languages and in patterns of thought (Gibbs
1994; Johnson 1987; Lakoff 1987; Lakoff and Johnson 1980).
The view that metaphor is of major importance in everyday usage
also began to make headway in AL during the late 1980s. Low (1988)
seems to have been one of the first (at least in English) to offer a to-do
list. Additional AL work on metaphor, up to the end of the 1990s, is excellently and extensively surveyed in Cameron and Low (1999). Classroom or self-study materials combining the recent interests in multi-
308
309
Given just these points of contact (and we are likely to have overlooked
others) it should not be surprising that mutually enriching research has already been carried out in both CL and education-oriented AL. Still, as we
suggested at the very outset of this introduction, there is plenty of scope
for advancement.
1.2. The notion of motivation in cognitive linguistics
So far, we have focused on areas of common ground, but now it is time to
consider key elements of CL which are under-appreciated both in AL and
in fields which AL may serve (e.g. language pedagogy and materials design) namely the element of categorisation by prototype and the related
concept of (semantic) motivation. As to the latter, something in language
is motivated when it is neither arbitrary nor (fully) predictable either
(Lakoff 1987: 346). While the majority of applied linguists will claim, for
example, that the form-meaning connections in language are arbitrary
(e.g. Lewis 1997: 1719) with the exception, perhaps, of onomatopoeic
words cognitive linguists, building on their view that language is a reflection of general cognitive processes, will consider it their task to look
for (retrospective) explanations for form-meaning connections in language. Any finding that portions of natural language are motivated rather
than arbitrary should be taken as a cordial invitation for educational linguists to investigate the pedagogical potential of presenting linguistic phenomena as motivated in contexts of second or foreign language learning.
Radden and Panther (2004) propose a taxonomy of linguistic motivation which turns on whether a process of motivation involves (1) meaning-meaning connections, (2) form-form connections, or (3) form-meaning connections.
1.2.1. Meaning meaning connections
The majority of CL investigations of motivation pertain to meaningmeaning relations which is to say in terms from Langacker (1987) that
they focus on the semantic poles of symbolic units. (This is also the focus
of the great majority of applications of CL in foreign language pedagogy;
see sections 1.4 and 2 below). These investigations include studies of polysemy, where peripheral senses of a word are found to be extended from a
central or prototypical sense via image-schema transformations or via
metonymy and metaphor. Prepositions (including particles of phrasal
verbs), typically being highly polysemous, have often served as showcase
310
311
While a great deal of early cognitive semantic work focuses on polysemy displayed by prepositions and particles, lexical items from other
word classes have also been given cognitive semantic treatment in recent
years (e.g. da Silva 2003; Tuggy 2003). In addition, a number of linguists
have argued that linguistic phenomena on the (outdated) borderline between lexis and grammar, such as modal verbs and other grammaticalised words, can be motivated by tracing the grammaticalised form to its
prototypical lexical meaning (e.g. Fisher 1999; Lehti-Eklund 2003;
Sweetser 1990).
Still with respect to meaning-meaning connections, a second fashionable type of cognitive semantic case study focuses on figurative idioms.
Idioms, which structural and generative linguists have traditionally
treated as semantically arbitrary and non-decomposable, have been
shown by cognitive semanticists to be linguistic instantiations of overarching conceptual metaphors (CMs) or conceptual metonymies (Lakoff
and Johnson 1980), which are grounded in physical or social experience.
For example, idiomatic expressions such as He reached his boiling point
and She flipped her lid are instantiations of the CM anger is a hot fluid
in a container (Kvecses 1986). Apart from everyday idiomatic language, conceptual metaphor theory (CMT) has also been applied to the
analysis of conventional figurative expressions in a panoply of text genres
of various specialised target domains, such as economics (e.g. Boers 1999;
Jkel 2003, Chung 2003a, 2003b, Herrera 2000, 2003, Koller 2000, 2003,
2004, White 1999, 2003), architecture (Caballero 2003a, 2006), medicine
(Salager-Meyer 1990), politics (Charteris-Black 2004, Chilton 1987, 1993,
1995, 1996, 3002, 2004, 2005, Cienki 2004, Lakoff 2002, Morgan 1997,
2000, Musolff 1996, 2000, Rohrer 2000, Schffner 1995, 1996), and language teaching (see Low 2003, for a survey of the latter).
Although it is acknowledged that the correspondence of idioms to their
CMs is sometimes far from self-evident, quantitative studies of the instantiations of given CMs have pointed to cross-linguistic or cross-cultural variation in the popularity of congruent metaphors and metonymies
(Boers 2003; Charteris-Black 2003; Deignan 2003).
While the CM approach to idioms tends to look at idioms from a synchronic perspective, other cognitive semantic investigations take a more
diachronic approach. Boers and Stengers (forthcoming), for example,
motivate the semantics of idioms by tracing them back to their historicalcultural-etymological origins. This exercise allows for a matching of
idioms with quite specific source domains (e.g. matching Win hands down
with the source domain of horse racing) and this in turn allows for cross-
312
313
314
same source domain. One such expression, to remain with the case of athletics, could be quick off the mark. Relating new words to already acquired
knowledge is believed to foster learning (e.g. Skmen 1997).
Secondly, we may hypothesise that motivating L2 phenomena can be
beneficial to retention. Thinking about the motivation for a given meaning-meaning, form-form or form-meaning connection is an example of
what in AL literature is called elaboration (e.g. Barcroft 2002). Semantic
elaboration refers performance of a mental operation with regard to the
meaning of words or constructions; structural elaboration refers to performance of a mental operation with regard to the form of words or constructions. Elaboration is known to foster learning as it involves processing information at a relatively deep level, which according to
levels-of-processing theory (Cermak and Craik 1979) increases the likelihood of the information being retained in memory. One type of deeplevel processing is exemplified by dual coding (see above, section 1, point
1). Since many types of linguistic motivation rely heavily on figurative
thought (i.e. metaphor and metonymy), learners are likely to reap the
benefits of dual coding when that motivation is revealed to them. Dual
coding is likely to take place in, for example, the teaching of figurative
idioms such as jump the gun, if the learner is encouraged to associate the
idiom with the concrete scene of its original literal use (i.e. an image of the
scene in the form of schematic visual and motoric images). Dual coding is
also likely to occur when drawings or pictorials are used to represent the
meaning of words (e.g. drawings of the TR-LM configurations of a preposition) and constructions (e.g. a time-line to contrast the scope of different tenses). As is well known, the use of pictorial representations is also
common practice in CL.
Thirdly, we may hypothesise that showing segments of the L2 to be
motivated can contribute to a heightened pragmatic and cultural awareness on the part of learners. Recognition of CMs, for example, may help
learners comprehend texts in general and recognise persuasive rhetoric in
particular. The ability to identify the source domains behind sets of
idioms may help students recognise (historically) culturally salient fields
of experience through confrontation, for example, with (a) the large
number of sailing, card-playing, horse-racing and hunting idioms in English; (b) the large number of idioms derived from religion in Spanish; (c)
the large number of food-related idioms in French; and so on.
Finally, an awareness that a second or foreign language need not be
learned entirely via a long and daunting road of blind memorisation must
be an encouraging thought to learners who recognise that large segments
315
of the target language actually make sense. This encouragement can lead
to positive affect, which is obviously beneficial to longer-term learning
behaviour (Arnold 1999)
1.4. Perception of CL in periodicals for teachers
Despite its pedagogical potential, CL has left surprisingly few traces in
periodicals for teachers, so far. Looking back to the 1980s, Lindstromberg
(1991a) found very few examples of any applications of CL in TESOL.
The new survey we carried out for the purpose of this chapter1 suggests
that the intervening years have seen relatively little increase in the influence of CL on day-to-day L2 pedagogy, with the sole exception of a few
traces we found of CMT, a strand of CL dating from Lakoff and Johnson
(1980).
A particularly stark example of CLs unimpressive impact on TESOL is
provided by the popular-academic TESL-EJ, which seems not to have
had a single article on any aspect of CL, including CMT, since its founding in 1997 (and, of all places, it is hosted at Berkeley, the base camp of
CMT!). ELT Journal, a similarly popular-academic journal for teachers,
has featured only a few articles explicitly drawing on CMT (Boers and
Demecheleer 2001; Deignan, Gabrys and Solska 1997) or CMT plus versions of prototype theory (Boers and Demecheleer 1998; Lindstromberg
1996). The more popularising periodicals that we have looked at provide a
few additional examples, such as Hannan (1998), a two-page account of
the semantics of up and down in phrasal verbs, and Ponterotto (1994),
Baker (1998) and Rundell (2002), three very short articles recommending
Lakoff and Johnson (1980).
Of all the for-teachers periodicals we looked at, the eclectic online journal Humanising Language Teaching has profiled CMT most often: Boers
and Lindstromberg 2006; Holme 2001; Lindstromberg 2001b,c,d, 2002;
Lindstromberg and Boers 2005; Littlemore 2001b, 2004a, 2005; Stengers,
Eyckmans, Horemans and Boers 2004; Rundell 2001, 2005a.
Compared to the situation in general-English teaching, CMT is slightly
better represented in periodicals devoted to the teaching of English for
specific purposes (ESP) (e.g. Caballero 2003b; Lindstromberg 1991b), especially in the fields of English for business and economics, where the
figurative nature of a vast number of (more or less) technical terms can
be revealed by tracing the terms (e.g. cashflow, human capital, fledgling
companies, elastic demand) back to their literal origins and/or by grouping
them under common CMs (e.g. Boers 2000a; Charteris-Black 2000; Char-
316
teris-Black and Ennis 2001; Charteris-Black and Musolff 2003; Littlemore 2002; White 2003). For example, a chronic deficit, an arthritic market, the economy is recovering, we need to slim down our labour force, the
right economic remedy, well have to amputate loss-making departments,
etc. could be grouped under the CM economics is health care. Likewise, a flourishing firm, pruning expenses, a branch of our company, etc.
could be grouped under economics is gardening. Having students categorise figuratively used words or expressions under such metaphor
themes or source domains is believed to facilitate retention both via the
likelihood of dual coding occurring (i.e. the likelihood of an association
between the words and a mental image) and via the likelihood of deep
processing occurring (i.e. the likelihood of cognitive effort being invested
in the categorising task). Empirical support for such beliefs in the pedagogic effectiveness of CL-inspired treatments is surveyed in section 2.
Finally, a particularly important source of information for teachers and
learners is that of learners dictionaries. Lindstromberg (2001), taking on
as a case study on how prepositions were treated in UK published monolingual learners dictionaries, provides a stark example of the scant impact
of CL in pedagogy-oriented lexicography at his time of writing. None of
the dictionaries under examination was found to consistently present
senses in the order of likely relatedness or provided any hints at the motivated nature of the polysemy, despite the fact that by 1997 there was available a relatively jargon-free CL-influenced depiction of more or less the
entire contemporary system of English spatial prepositions (Lindstromberg 1997), which was addressed to ESOL teachers, translators and lexicographers. Fortunately, it seems the tide may now slowly be turning; as
evidenced by Rundell (2005b), a dictionary of phrasal verbs which has
plainly taken CMT (and prototype theory) into account in its semantic
maps of the meanings of individual prepositions.
2.
In this core section of the chapter, we shall survey in more detail some
of the proposals for second or foreign language instruction that have been
put forward by cognitive linguists in recent years. The rationale behind
CL-inspired language pedagogy was outlined by several scholars (focusing on various segments of language) in Ptz, Niemeier and Dirven
(2001a, 2001b) and in Achard and Niemeier (2004). Crucially, we shall
here survey the available evidence of the surplus effectiveness of the pro-
317
318
319
sises that beyond implies some distance between the TR and the LM, a
feature which makes beyond a likely candidate to instantiate the metaphor
abstract inaccessibility is distance. Moreover, although these students had been given only this core spatial sense of beyond, they significantly outperformed their peers who had been given access to complete
dictionary definitions including explanations of the figurative senses.
Kvecses and Szab (1996, revisited in Kvecses 2001) focused on English phrasal verbs with up and down. In the experimental condition,
15 Hungarian students were asked to study ten phrasal verbs accompanied by CL explanations that raised the students awareness of the conceptual metaphors (henceforward: CMs) underlying the given phrasal
verbs (more is up, happy is up, etc). Under the control condition, another 15 Hungarian students were asked to study the same phrasal verbs
accompanied by L1 translations. In an immediate post-test (a gap-filling
exercise) the experimental students outperformed the control students by
almost 9 %. The immediate post-test also targeted ten more up / down
phrasal verbs that had not been included in the instruction stage and, interestingly, the experimental students outperformed their control peers by
almost 25 % on these items, which suggests that the experimental students
managed to transfer the acquired knowledge of the chosen CMs to their
interpretation of instantiations they had not previously studied.
One of the experiments reported by Boers (2000b) focused on phrasal
verbs instantiating a wider range of CMs (including, visible is out, e.g.
find out, turn out; visible is up, e.g. look it up, show up, etc). In the experimental condition, 39 French-speaking students were asked to study a list
of 26 phrasal verbs that were grouped under various CMs and accompanied by a synonym. In the control group, another 35 French-speaking students were asked to study the same phrasal verbs listed alphabetically and
accompanied by more elaborate explanations (e.g. several synonyms) copied from a well-known grammar book. In an immediate post-test (a textbased gap-filling exercise) that targeted ten of the phrasal verbs studied,
the experimental students outperformed their control peers significantly.
The post-test also targeted ten more phrasal verbs that were not included
in the list the students had been asked to study. Unlike Kvecses and
Szabs (1996) set-up, these extra items did not instantiate the same CMs
as the phrasal verbs studied. Students post-test scores on this sub-set of
the gaps showed no difference between the experimental and the control
group. In other words, these results show no evidence of any transfer
strategy beyond the CMs that experimental students were made conscious
of.
320
The finding by both Kvecses and Szab (1996) and Boers (2000b) that
a presentation of phrasal verbs in terms of CMs is measurably beneficial
to at least short-term retention is certainly encouraging, although it
does not pinpoint whether the benefits are due to figurative thought (i.e.
mental imagery resulting in dual coding) or simply to the fact that organised vocabulary is easier to learn than random lists (e.g. Schmitt 1997;
Skmen 1997), and CMs can indeed help organise vocabulary. We acknowledge too, that the findings do not guarantee that the whole category
of prepositions and phrasal verbs lends itself well to the proposed strategy. Neither do the results imply that CL analyses of prepositional polysemy can be copied straight into course materials or lesson plans. CL-inspired pedagogues and materials writers would need to take at least the
following points into consideration:
1. There are several aspects of prepositional polysemy which cognitive
linguists tend to disagree about, e.g. what, in a given case, the prototype is; how many distinct senses/schemas there are; how these relate to
the prototype; and how they relate to each other. Rice (1996, 2003) and
Sandra and Rice (1995) report psycholinguistic evidence of structured
polysemy, but not at the level of the multiple fine-grained distinctions
among senses that some cognitive linguists have been wont to propose.
Rice and Sandra (op. cit.) therefore caution that purely introspective
linguistic analysis may result in portrayals of semantic networks that
may be lacking in inter-subjectivity. Even if semantic networks of polysemes are streamlined for teaching purposes (e.g. Evans and Tyler,
forthcoming), full agreement may not be realistically achievable, given
that an individuals language resources are usage-based. From a
learners perspective, however, the benefits of presentation of meaning
extensions as motivated may well depend on the perceived plausibility
of the motivations.
2. Not all explicated motivation is equally likely to facilitate learning.
Sometimes the motivation of meaning extensions is pretty straightforward, but at other times it may be quite abstract and may even come
across to learners as far-fetched. In a small-scale experiment, Condon
and Kelly (2002) used an early manuscript of a chapter to be included
in Rudzka-Ostyn (2003) to teach 19 university students (the experimental group) phrasal verbs with out (e.g. break out, pick out, rule out,
check out, set out, stand out). A control group (14 students) was taught
the same phrasal verbs along the lines of the treatment in Collins Cobuild Phrasal Verb Wordbook (Goodale 1998). In a post-test measuring
321
students recollection of the meaning of the phrasal verbs taught, experimental students did better than the control students, but only on
the phrasal verbs whose motivation had been straightforward and concrete enough to be illustrated with a drawing during the instruction
stage.
3. If they are to be embraced by the teaching community, applications of
CL must be learner-friendly and adapted to the target audience. For
one thing, CL-inspired materials writers may need to modify their jargon and refrain from using terms such as trajector. More importantly,
they may need to keep in mind that relatively few L2 learners are inclined to engage in the kind of prolonged and intensive semantic analyses that linguists find so fascinating. An informal study by Kurtyka
(2001) found that Rudzka-Ostyns (2003) materials for teaching and
learning phrasal verbs were evaluated quite positively by a group of
Polish EFL teachers, but they also reported that students at relatively
low levels of proficiency felt the materials were too demanding.
4. The perceived relevance of explicit instruction of the multiple senses of
prepositions may be undermined by the fact that most prepositions are
high-frequency items that can be acquired incidentally, in a rough and
ready fashion, given ample enough exposure to the target language.
Also, intermediate and advanced students are often already familiar
with the meanings of the prepositions taught, especially if cognates are
available in the students L1. Lowie and Verspoor (2004) report an experiment the results of which show significant effects of both frequency
of occurrence and cognates in students mastery of prepositions. Beginning learners will tend to rely on L1 cognates, but as learners progress and experience more exposure to L2, the frequency effect tends to
gain in power and overrides the cognate-effect. However, many languages have no straightforward counterparts of prepositions at all, let
alone cognates. And even across such moderately related languages as
French and English, misunderstandings about prepositions can persist
at extremely high levels of proficiency (Coppieters 1987).
2.1.2. Beyond prepositions and phrasal verbs
The presentation of polysemous words in terms of meaning extensions
from a central or prototypical sense has been applied to other classes of
expressions than prepositions, typically by motivating the metaphoric/
metonymic connections between the literal sense of a word and its figurative extensions (e.g. MacLennan 1994; Scott 1994). Segments of lan-
322
323
hind figuratively used polysemes facilitates vocabulary retention. However, when the same students were given a similar task one year later, the
experimental group did not do any better than the control group anymore
(Boers 2004).
In an experiment with the participation of 78 Dutch-speaking students
of English, Verspoor and Lowie (2003) first confronted experimental
group learners with the core senses of polysememes in order to test their
ability to interpret and remember more peripheral senses in the radial network. The researchers chose 18 polysemous words whose core sense had
given rise to a meaning extension that is still relatively close to the core
sense, but also had given rise to a meaning extension occupying a more
peripheral position in the network. Experimental students were asked to
guess the meaning of the core sense before attempting to figure out the
relatively closely connected meaning extension. Control students were
first asked to guess the meaning of the more peripheral sense of the word
before tackling the less peripheral one. Experimental students were found
to be significantly more likely to correctly interpret the relatively closely
connected meaning extension and also to remember its exact meaning in a
delayed post-test (after two weeks).
Csbi (2004) reports an experiment set up to measure the pedagogic effect of explicit CL explanations of the semantic networks of the English
verbs hold and keep. Participants were 52 Hungarian secondary school
pupils. The experimental group was told that the core meaning of hold
involves an agents hand (as in She held the purse in her right hand ). Meaning extensions were explained via conceptual metonymies and metaphors
such as the hand for control and possessing is holding (e.g. The terrorists held them hostage and He did not hold the right certificate). They
were told that the core meaning of keep implies a durative state of possession (as in You can keep the change), but without implying the use of
hands. The durative component was then used to explain other uses of
keep as in Earn enough to keep a family and Keep that dog out of my study.
In addition to verbal explanations, the experimental condition involved
some pictorial support, e.g. the teacher illustrating core senses by miming
or by drawing on the blackboard. The control group was presented with
the same examples of hold and keep, but these were explained by means of
translations into Hungarian. An immediate post-test and a delayed posttest (one day after instruction) in which students were asked to fill in gaps
with either hold or keep showed that the experimental students were significantly more likely to make the right choice of lemma. These results
suggest that even fairly abstract motivations of meaning extensions can be
324
beneficial to learners. The better performance by the experimental students may partly be due to the extra cognitive effort they invested in comprehending the examples on the basis of the motivations, which presumably involved deeper processing than going over the translations. It may
also partly be due to the pictorial support of the explanations, which was
absent in the control condition.
To evaluate the effects of visual, and possibly motoric, support, Lindstromberg and Boers (2005b) conducted an experiment focusing on 24
manner-of-movement verbs (e.g. hobble, stagger, teeter, veer, saunter,
slump, flit). In the experimental condition, students took turns miming
the verbs so that their peers could guess what verb was being acted out. In
the control condition, students explained the meaning of the verbs verbally so that their peers could guess what verb was being explained. An immediate post-test (a gap-filling exercise) targeting the verbs used in their
literal senses revealed that the experimental students were significantly
more likely to recollect them. In a delayed post-test, the students were
presented with the same verbs used in their figurative senses accompanied
by an L1 translation, which was deliberately kept fairly vague (e.g. it did
not hint at the level of hostility in He hurled insults at his girlfriend, or at
the incidental nature of He stumbled on a crucial piece of evidence). Students were asked to evaluate the proposed translations. The results show
that students who had watched each other mime the literal senses of the
verbs were significantly more likely to spot the lack of precision in the
translations of the metaphorical uses of the verbs. Students were especially likely to correctly question the translations of the verbs they had
acted out themselves. The significantly better scores under the miming
condition in the post-test that was administered to measure students recollection of literal uses of the verbs, suggests a strong mnemonic effect of
dual coding. The significantly better scores in the test that was administered to measure in-depth comprehension of the figurative uses of the
verbs, suggests that students are often capable of transferring their enhanced comprehension of core senses onto their appreciation of metaphorical extensions from those core senses, which confirms the results of
the Verspoor and Lowie (2003) study.
2.2. CL approaches to idioms
Idioms are traditionally described as dead metaphors, the idea being
that they have become so standardised that few people are normally aware
of their figurative nature. Cognitive linguists, however, have shown that
325
the imagery behind an idiom can easily be resuscitated either by enhancing peoples awareness of the underlying CMs or by tracing the idiom
back to its original, literal context (e.g., Gibbs 1994). Lazar (1996) is an
early call for promoting metaphor awareness in teaching and learning
figurative expressions (though not yet from a CL perspective at the time),
and Lazar (2003) is possibly the first EFL exercise book with a strong
focus on raising awareness of metaphor.
Boers and Lindstromberg (2006) propose classroom activities whereby
students (in accordance with research results, see below) first hypothesise
about the origin of an idiom and then try to work out its figurative meaning on the basis of that original use. This problem-solving task is meant
to engage students in deep processing; additionally, the aim is to foster
dual coding by asking volunteers to mime or draw the literal senses of the
idioms. Littlemore and Low (2006) describe a variety of more incidental
classroom interactions in which students who encounter an unfamiliar
figurative expression gradually figure out its meaning via the teachers
hints at relevant associations with the literal, sometimes culture-specific,
origin of the expression.
Empirical evidence of the usefulness of CM as a categorising principle
behind sets of figurative idioms comes from a number of different sources.
Boers (2000b) reports an experiment with 118 Dutch-speaking secondary school pupils who were presented with a list of 18 expressions (borrowed from Kvecses 1986) such as She blew up at me, Her comments
added fuel to the fire and Dont bite my head off. In the experimental condition, these expressions were grouped under headings referring to CMs,
i.e. anger is a hot fluid in a container, anger is fire and angry
people are dangerous animals. In the control condition, the same lexis
was organised under functional headings, referring to whether the expressions were used to describe sudden anger, a slow build-up of anger, or
angry personalities. This was done to ensure the same degree of organisation of the input under both conditions. In an immediate post-test (a
text-based gap-filling exercise targeting ten of the expressions), the experimental group outperformed the control group significantly. However, the
data also hint at the possibility that an enhanced awareness of congruent
CMs may lead to a greater inclination for L1-to-L2 transfer, and thus also
to a greater risk of erroneous L1 transfer at the level of linguistic form
(e.g. add *oil to the fire and dont bite my *nose off).
Learners are typically wary of using idioms in their target language
(Irujo 1993). In this connection, Deignan, Gabrys and Solska (1997) propose cross-linguistic comparisons of figurative idioms (English and
326
327
not yet know (e.g. pass the baton and a dummy run). The experimental students were given the supplementary task of writing down a hypothesis
about the origin of each idiom, while the control students were asked to
invent a communicative situation in which each idiom could be used. It
was assumed both tasks would involve a comparable investment of cognitive effort (i.e. comparable levels of deep processing), but it was hoped as
well that the hypothesising about the literal origin of an idiom would call
up a picture of a concrete scene in the learners mind, resulting in dual
coding (e.g. associating pass the baton with the scene of a relay race). In an
immediate post-test targeting the idioms in a gap-filling exercise, the experimental students significantly outperformed their control peers, which
may be indicative of the mnemonic power of dual coding. In a delayed
post-test (after a week) in which students were asked to explain the meaning of the ten idioms, the experimental group again did significantly
better.
Encouraged by the mnemonic effect observed in the Boers (2001) study,
Boers, Demecheleer and Eyckmans (2004a) and Boers, Eyckmans and
Stengers (Forthcoming) conducted a series of larger-scale experiments
with the aid of an on-line programme of exercises they developed to help
Dutch-speaking students comprehend and remember a total of 400 English idioms (see also Stengers, Eyckmans, Horemans and Boers 2004). In
this programme, each idiom is presented to the students in three types of
exercises: a multiple-choice exercise (the origin exercise) where the student is asked to tick the right source domain of the idiom (e.g. What domain of experience do you think the expression To be on the ropes comes
from? Sports, food or sailing?), a multiple-choice exercise (the meaning
exercise) where the student is asked to tick the right dictionary-like definition of the idiom, and finally a gap-filling exercise where the student is
invited to fill in (a keyword of) the idiom inserted in a suggestive context.
After each exercise the student is given feedback with the correct answer.
In the case of the origin exercise, the feedback gives a brief explanation
of the literal, historical-cultural-etymological origin of the idiom, which
is meant to help the student associate the expression with a concrete scene
(for example an association of To be on the ropes with the scene of boxing
where one fighter is in trouble). Boers, Demecheleer and Eyckmans
(2004a) found that students who had had access to the origin exercises
but not the meaning exercises were significantly more likely to be able to
reproduce the idioms in the gap-filling exercise than their colleagues who
had had access to the meaning exercises but not the origin exercises. In
a second experiment they found that apart from a couple of exceptions
328
(see below) the mnemonic effect of giving explanations about the literal
origins of idioms was just as strong for opaque idioms (defined here as
idioms whose source domain the student was unable to guess, e.g. the
source domain of card games behind To follow suit) as for transparent
idioms (e.g. To break ranks). These results too offer strong support for the
dual coding hypothesis. In a third experiment, Boers, Eyckmans and
Stengers (Forthcoming) found that this mnemonic effect was optimised
significantly by giving students the origin exercise before the meaning
exercise, so that they could use the newly acquired knowledge of the literal
sense of the idiom to figure out its figurative sense. We hypothesise that
this optimal sequence of events adds deeper processing to the learning
process (in the form of insightful problem-solving rather than blind guessing in the meaning exercise). The finding that students are significantly
more likely to identify the correct dictionary-like definition in the meaning exercise after they have been told about the literal origin of the idioms
shows that these idioms became motivated for L2 learners too. In addition, interviews with the students suggest that the high scores obtained
under this approach contribute to positive affect and thus to willingness
to continue learning.
Boers and Demecheleer (2001) report a small-scale experiment the results of which suggest that cross-cultural variation in the salience of (congruent) source domains for metaphor can impact learners ability to interpret idioms, and that therefore learners comprehension of idioms could
benefit additionally from cultural knowledge (in its widest sense). More
conclusive evidence of the effect of cultural variables on learners ability
to comprehend and also remember L2 idioms is offered by Boers, Demecheleer and Eyckmans (2004b). In a retrospective analysis of a large
databank of students scores on on-line English idiom exercises (see
above), it was found that idioms derived from rather culture-specific
source domains (such as cricket and horse-racing) were significantly
harder for Dutch-speaking students to interpret and remember. The lack
of familiarity with a given source domain may obviously reduce the likelihood of dual coding taking place in the learners mind (if a verbal explanation fails to call up a mental picture in the first place).
All the above-mentioned studies examine the pedagogic potential of
motivating the relations between literal and figurative senses through
imagery. This type of motivation may help learners comprehend and remember idioms. What it cannot do, however, is motivate the precise lexical make-up of idioms. Boers and Lindstromberg (2005b) and Boers and
Stengers (forthcoming) investigate the phonological motivation behind
329
the lexical selection in idioms. It appears that in the process of standardising word combinations, the phonological properties of one word may
determine the choice of another. Hand counts in idiom dictionaries reveal, for example, that the lexical selection in up to 20 % of English idioms
may be influenced by the appeal of alliteration and/or rhyme (e.g. from
pillar to post, through thick and thin, go with the flow). A retrospective
analysis of the performance of several generations of students in on-line
exercises targeting 400 English idioms (see above) shows that idioms with
alliterating word combinations are significantly more likely to be retained
than idioms without apparent phonological appeal. Moreover, this effect
is observed even though the instructional programme was set up to draw
the students attention to imagery rather than sounds, i.e. it was set up to
stimulate semantic elaboration (via the etymological feedback) in the
students minds rather than structural (or, more precisely, phonological)
elaboration. By occasionally raising learners awareness of catchy sound
patterns, the mnemonic effect of alliteration can be enhanced significantly
(Boers and Lindstromberg 2005b). With a view to optimising the memorability of catchy sound patterns, Lindstromberg and Boers (2005a) propose various classroom activities that go beyond mere noticing and which
can target not just idioms but compounds and collocations as well. It is
clear, though, that we are only just beginning to appreciate the potential
scope of phonological (and phonetic) motivation in natural language and
in language pedagogy.
2.3. CL approaches to constructions
Although within CL there are various opinions on the nature of constructions (see Goldberg 1995, Croft 2001, Langacker 2005), it is correct
to say that CL treats grammatical constructions on a par with lexis, in the
sense that, like words, constructions are symbolic units comprising a semantic and a phonological pole. The difference is that constructions are
more schematic than words and may more clearly function as overarching
templates sanctioning a greater variety of lexically elaborated instantiations. Dirven (1989), Hllen (1992), Langacker (2001), Meex and Mortelmans (2002), Taylor (1993), Turewicz (2000) are among those who have
conjectured that this treatment of grammatical constructions as meaningful makes CL an appealing source for the design of a pedagogical grammar.
A speakers choice of a given construction to present a state of affairs
reflects the speakers construal of that state of affairs. However, languages
330
obviously differ with respect to the kinds of constructions that are available for purposes of construal and with respect to the relative coverage
by congruent (or equivalent) constructions. For example, French does not
have the progressive tenses that English does. Conversely, English does
not have the subjunctive tenses that French does. Some applied cognitive
linguists (e.g. Meex and Mortelmans 2002; Niemeier 2004) have outlined
contrastive approaches to pedagogical grammar as a way to make
learners aware of the conceptualisations in L2 that differ from their L1.
An additional way of helping learners get a grip on L2 grammar is to
point out cases of the operation of the same cognitive principle in different segments of the language. For example, the bounded versus unbounded categorisation of things (resulting in the grammatical distinction between countable and uncountable nouns) also applies to actions,
resulting in the grammatical distinction between simple and progressive
tenses in English (Dirven 1989).
Much of the pedagogic appeal of cognitive grammar lies in its treatment of constructions as usage-based. Like words, constructions are subject to processes of elaboration and schematisation. Through usage
events, constructions may develop meanings and uses that do not correspond completely with their prototypical meaning or usage. However, if
the novel use occurs frequently enough to become entrenched, it will
eventually be fully sanctioned by an adapted, overarching schema (Langacker 1987). Several cognitive linguists (e.g. Hllen 1992; Langacker
2001) have called for grammar teaching whose first stage is attempting to
heighten awareness of the prototypical usage of each target construction
(e.g. a tense) and whose later stages involve showing how other usages derive from the prototype. This cognitive sequence from typical to special
cases is certainly not a new idea in grammar teaching (as a quick glance
through best-selling grammar books will reveal). The novelty lies in the
presentation of non-prototypical uses of a given construction not as exceptions to the rule, but rather as motivated extensions from the prototype. Thus, to a far greater extent than has been true of other approaches,
CL-inspired pedagogy has been conceived to help learners see that the L2
is informed by principles that make sense and are useful to know because
they facilitate learning. There is, additionally, an emphasis on (i) the language users needs to construe and produce L2 discourse, (ii) the manner
in which these needs are met by the available repertoire of L2 constructions, and (iii) the fact that under certain conditions that repertoire can be
stretched and adapted (Taylor 1993). For example, many so-called uncountable nouns in English can become countable if the language user de-
331
cides to reinterpret concepts that are conventionally construed as unbounded and construe them as bounded (and thus countable) instead (e.g.
behaviours).
The cognitive sequence from-prototype-to-extensions does not necessarily imply that cognitive grammarians endorse explicit, sequential teaching of discrete grammar points. According to Achard (2004), the conception of grammatical constructions as radial categories organised around
a deeply entrenched prototype (i.e. one that is frequent, physically
grounded, and/or early-learned) and slightly less entrenched (e.g. less frequent) peripheral uses is perfectly compatible with relatively un-programmed inductive grammar acquisition as proposed, for example, by
Natural Approach theorists (Krashen and Terrell 1983; see, however,
critique of Krashen in Taylor 1993) and by Prabhus (1987) early version
of Task-Based Learning. The idea here is that through exposure to large
amounts of input, learners tend most often and/or most perceptually saliently to encounter uses that correspond to the prototype, while they will
encounter the less typical instantiations less often (and/or less saliently)
and automatically consider these as more peripheral cases. At the same
time, the confrontation with atypical instances can be used to raise students appreciation of the way native speakers bend the grammar-book
rules in order to present the way they construe a given state of affairs. In a
similar vein, the use of authentic materials for grammar acquisition instead of discrete-point grammar teaching is favoured by Grundy (2004),
who uses the figure-ground notion from CL to demonstrate how important context (the ground) is in learners comprehension and intake (of segments) of the target language.
The case-studies used to back up the above-mentioned proposals for a
pedagogical cognitive grammar are a bit too complex to be summed up in
this survey chapter. For easy exemplification we therefore turn to Littlemore and Low (2006), who give straightforward examples of figurative
thinking behind grammatical choices. One of these examples is the use, in
academic writing, of tenses to make references which may reflect authors
evaluation of the work that they are referring to by creating either a sense
of proximity or a sense of distance. For example, the present tense in
Smith (1998) suggests that reflects closer mental proximity (and thus
more agreement) than the past tense in Smith (1998) suggested that
Another example is the choice of the demonstrative this (rather than that
or the indefinite article) in joke openings like There was this Englishman , which creates closer proximity between the listener and the protagonist in the created mental space.
332
The above survey shows that a considerable number of appealing proposals for CL-inspired pedagogical grammar have been put forward in recent years. It is unfortunate, however, that to our knowledge at least no
reports have yet been published of any empirical studies conducted to assess the pedagogical effectiveness of CL-inspired treatments in comparison with other, traditional instruction of L2 grammar.
2.4. CL approaches to reading comprehension
Reading research is a very vast research area in psycholinguistics and applied linguistics. The contribution that CL can offer to it will typically
investigate the potential role of figurative language present in texts as a
potential facilitator (or obstacle) in the comprehension process. Sometimes texts are structured around a recurring metaphor theme, i.e. an
overarching CM that helps lend rhetorical coherence and through its
linguistic instantiations lexical cohesion to a text. Allbritton, McKoon
and Gerrig (1995) conducted a series of experiments in which respondents
were asked to read texts, some of which were structured around an overarching metaphor and some of which were not. The respondents were
then presented with small segments from the texts and asked if they could
remember encountering these in the reading task. The segments taken
from the texts with a recurring metaphor theme turned out significantly
more likely to be recognised than those taken from other text fragments.
In other words, the presence of an overarching CM facilitates readers
(receptive) recollection of words or phrases they come across in a given
text. While this kind of recollection can only be an indirect measure of a
readers comprehension of a text, it can nevertheless be indicative of the
extent to which intake has taken place. If recurring metaphor themes facilitate this intake, then their presence could be exploited when reading
texts in a second or foreign language, too. The only prerequisite seems to
be that the language learners need to be able to recognise the CM, and to
do that they need to be able to recognise at least a couple of its linguistic
instantiations in the text. Again, some CM instruction may be useful, but
to our knowledge there have been no reports of controlled experiments set
up to measure the effect of CM instruction on learners intake of CMstructured texts.
An area in reading comprehension where metaphor awareness may be
especially useful to readers is when metaphors are used as a persuasive or
manipulative device, as in political rhetoric. Experiments have shown that
the CMs used by authors to frame problems and dilemmas can have a pro-
333
334
335
wards this objective include cross-cultural comparisons of CMs and metonymies, and cross-cultural comparisons of the prototypical members of
categories and the inclusion vs. exclusion of peripheral members to those
categories.
Goddard (2004) even proposes classroom use of semantic primes and
cultural scripts (e.g. Wierzbicka 1996, 1997) to raise students awareness
of subtle cross-cultural differences in the connotations of congruent concepts. For example, he claims that the concept freedom in British English
seems constrained by the idea that ones free actions should not be bad for
others. This cultural constraint could help motivate a whole panoply of
tentative phrases, such as Would you mind ?) which have so far been
relegated to folk wisdom about the oddly polite nature of English discourse.
Neither Niemeier (2004) nor Goddard (2004) offer any empirical evidence for the pedagogical effectiveness of their proposed classroom activities, but students who invest cognitive effort in them can certainly be
expected to engage in deep processing and associative thinking, which is
bound to be beneficial for learning.
3.
336
337
338
339
Notes
1. With respect to CL in the L2 classroom, our impressions stem from:
(a) our necessarily limited base of contacts with other teachers (e.g. colleagues, participants on training courses and workshops, and teachers we meet
at TESOL and AL conferences);
(b) our inevitably lacunaed acquaintance with published classroom and study
materials (mostly for English as a target language); and
(c) our screening of the following periodicals for teachers, which we class as:
(i) popular-academic: ELT Journal (UK) and TESL-EJ (US)
(ii) popularising: Forum (USA), English Teaching Professional (UK) and
Modern English Teacher (UK); and
(iii) mixed: Humanising Language Teaching (UK).
2. Throughout this chapter, the term significant is used to refer to statistical significance, i.e. with p-values smaller than .05. Actually, in most of the instances
where we have used the term in connection with the surveyed experiments, significance levels at p < .01 or at p < .001 were obtained.
References
Achard, Michel
2004
Grammatical instruction in the Natural Approach: a Cognitive Grammar view. In Cognitive Linguistics, Second Language Acquisition, and
Foreign Language Teaching, Michel Achard, and Susanne Niemeier
(eds.), 165194. (Studies on Language Acquisition 18.) Berlin/New
York: Mouton de Gruyter
Achard, Michel, and Susanne Niemeier (eds.)
2004
Cognitive Linguistics, Second Language Acquisition, and Foreign Language Teaching. (Studies on Language Acquisition 18.) Berlin/New
York: Mouton de Gruyter.
Allbritton, David W., Gail McKoon, and Richard Gerrig
1995
Metaphor-based schemas and text representations: making connections through conceptual metaphors. Journal of Experimental Psychology: Learning, Memory and Cognition 21: 612625.
340
Arnold, Jane
1999
Affect in Language Learning. Cambridge: Cambridge University Press.
Athanasiadou, Angeliki
2004
Teaching temporal connectors and their non-temporal extensions. In
Cognitive Linguistics, Second Language Acquisition, and Foreign Language Teaching, Michel Achard, and Susanne Niemeier (eds.), 195210.
(Studies on Language Acquisition 18.) Berlin/New York: Mouton de
Gruyter
Baddeley, Alan
1990
Human Memory: Theory and Practice. Needham Heights, MA: Allyn
and Bacon.
Baker, Judith
1998
Metaphor. English Teaching Professional 7: 1314.
Barcroft, Joe
2002
Semantic and structural elaboration in L2 lexical acquisition. Language Learning 52: 323363.
Boers, Frank
1996
Spatial Prepositions and Metaphor: A Cognitive Semantic Journey along
the up-down and the front-back Dimensions. Tbingen: Gunter Narr
Verlag.
1997
No pain, no gain in a free-market: A test for cognitive semantics?
Metaphor and Symbol 12: 231241.
1999
When a bodily source domain becomes prominent: the joy of counting
metaphors in the socio-economic domain. In Metaphor in Cognitive Linguistics, Raymond W. Gibbs, and Gerard J. Steen (eds.), 4756. (Current
Issues in Linguistic Theory 175.) Amsterdam/Philadelphia: Benjamins.
2000a Enhancing metaphoric awareness in specialised reading. English for
Specific Purposes 19: 137147.
2000b Metaphor awareness and vocabulary retention. Applied Linguistics 21:
553571.
2001
Remembering figurative idioms by hypothesising about their origin.
Prospect 16: 3543.
2003
Applied linguistics perspectives on cross-cultural variation in conceptual metaphor. Metaphor and Symbol 18: 231238.
2004
Expanding learners vocabulary through metaphor awareness: What
expansion, what learners, what vocabulary? In Cognitive Linguistics,
Second Language Acquisition, and Foreign Language Teaching, Michael
Achard, and Susanne Niemeier (eds.), 211234. (Studies on Language
Acquisition 18.) Berlin/New York: Mouton de Gruyter.
Boers, Frank, and Murielle Demecheleer
1998
A cognitive semantic approach to teaching prepositions. English Language Teaching Journal 53: 197204.
2001
Measuring the impact of cross-cultural differences on learners comprehension of imageable idioms. English Language Teaching Journal 55:
255262.
341
342
Brugman, Claudia
1981
Story of Over. M.A. thesis. University of California, Berkeley.
Byram, Michael
1997
Teaching and Assessing Intercultural Communicative Competence.
Clevedon: Multilingual Matters.
Caballero, Rosario
2003a Metaphor and genre: The presence and role of metaphor in the building review. Applied Linguistics 24: 145167.
2003b How to talk shop through metaphor: Bringing metaphor research into
the classroom. English for Specific Purposes 22: 177194.
2006
Re-Viewing Space: Figurative Language in Architects Assessment of
Built Space. (Applications of Cognitive Linguistics 2.) Berlin/New
York: Mouton de Gruyter.
Cacciari, Cristina
1993
The place of idioms in a literal and metaphorical world. In Idioms: Processing, Structure, and Interpretation, Cristina Cacciari, and Patrizia
Tabossi (eds.), 2756. Hillsdale, NJ: Erlbaum.
Cadierno, Teresa
2004
Expressing motion events in a second language: A cognitive typological approach. In Cognitive Linguistics, Second Language Acquisition,
and Foreign Language Teaching, Michael Achard, and Susanne Niemeier (eds.), 1349. (Studies on Language Acquisition 18.) Berlin/New
York: Mouton de Gruyter
Cadierno, Teresa, and Karen Lund
2004
Cognitive linguistics and second language acquisition: Motion events in
a typological framework. In Form-Meaning Connections in Second Language Acquisition, Bill VanPatten, Jessica Williams, Susanne Rott, and
Mark Overstreet (eds.), 139154. Hillsdale, N.J.: Lawrence Erlbaum.
Cameron, Lynne, and Graham Low
1999
Metaphor. State of the art survey. Language Teaching 32: 7796.
Cermak, Laird S., and Fergus I. M. Craik
1979
Levels of Processing in Human Memory. Hillsdale, N.J.: Lawrence Erlbaum.
Charteris-Black, Jonathan
2000
Metaphor and vocabulary teaching in ESP economics. English for Specific Purposes 19: 14965.
2004
Politicians and Rhetoric: Persuasive Power of Metaphor. Basingstoke,
U.K.: Palgrave MacMillan.
Charteris-Black, Jonathan, and Timothy Ennis
2001
A comparative study of metaphor in Spanish and English financial reporting. English for Specific Purposes 20: 24966.
Charteris-Black, Jonathan, and Andreas Musolff
2003
Battered hero or innocent victim? A comparative study of metaphors
for euro trading in British and German financial reporting. English for
Specific Purposes 22: 153176.
343
Chilton, Paul
1987
Metaphor, euphemism and the militarization of language. Current Research on Peace and Violence 10: 717.
1996
Security Metaphors. Cold War Discourse from Containment to Common
House. New York: Peter Lang.
2002
Do something! Conceptualising responses to the attacks of 11 September 2001. Journal of Language and Politics 1: 181195.
2004
Analysing Political Discourse: Theory and Practice. London/New York:
Routledge (Taylor and Francis).
2005
Manipulation, memes and metaphors: The case of Mein Kampf. In Perspectives on Manipulation and Ideologies: Theoretical Aspect, Louis de
Saussure and Peter Schulz (eds.), Amsterdam; Philadelphia: Benjamins.
Chilton, Paul, and Mikhail Ilyin
1993
Metaphor in political discourse: The case of the Common European
House. Discourse and Society 4: 731.
Chilton, Paul, and George Lakoff
1995
Foreign policy by metaphor. In Language and Peace, Christina
Schffner, and Anita. L. Wenden (eds.), 3759. Dartmouth, U.K.: Aldershot.
Chung, Siaw Fong, Kathleen Ahrens, and Chu-Ren Huang
2003
ECONOMY IS A TRANSPORTATION DEVICE: Contrastive representation of source domain knowledge in English and Chinese. Proceedings of the Special Session of UONLP, 2003 International Conference on Natural Language Processing and Knowledge Engineering
(NLP-KE), Beijing. Taiwan: Association for Computational Linguistics and Chinese Language Processing.
2003
ECONOMY IS A PERSON: A Chinese-English corpora- and ontological-based comparison using the conceptual mapping model. Proceedings of the 15th ROCLING Conference, National Tsing-Hwa University, Taiwan. Taiwan: Association for Computational Linguistics
and Chinese Language Processing.
Cienki, Alan J.
2004
Bushs and Gores language and gestures in the 2000 US presidential
debates: A test case for two models of metaphors. Journal of Language
and Politics 3: 409440.
Clark, James M., and Allan Paivio
1991
Dual coding theory and education. Educational Psychology Review 3:
233262.
Cohen, Gillian, Michael W. Eysenck, and Martin E. LeVoi
1986
Memory: A Cognitive Approach. Milton Keynes: Open University
Press.
Condon, Nora, and Peter Kelly
2002
Does cognitive linguistics have anything to offer English language
learners in their efforts to master phrasal verbs? ITL Review of Applied
Linguistics 137/138: 205231.
344
Coppieters, Ren
1987
Competence differences between native and near-native speakers. Language 63 (3): 54473.
Croft, William
2001
Radical Construction Grammar: Syntactic Theory from a Typological
Perspective. Oxford: Oxford University Press.
Csab, Szilvia
2004
A cognitive linguistic view of polysemy in English and its implications
for teaching. In Cognitive Linguistics, Second Language Acquisition,
and Foreign Language Teaching, Michel Achard, and Susanne Niemeier
(eds.), 233256. (Studies on Language Acquisition 18.) Berlin/New
York: Mouton de Gruyter.
Danesi, Marcel
1995
Learning and teaching languages: the role of conceptual fluency. International Journal of Applied Linguistics 5: 320.
Da Silva, Augusto Soares
2003
Image schemas and category coherence: The case of the Portuguese
verb deixar. In Cognitive Approaches to Lexical Semantics, Hubert
Cuyckens, Ren Dirven, and John R. Taylor (eds.), 281322. (Cognitive
Linguistics Research 23.) Berlin/New York: Mouton de Gruyter.
De Knop, Sabine, Antoon De Rijcker, Birgitta Meex, and Tanja Mortelmans (eds.)
in preparation Pedagogical Grammar. (Applications of Cognitive Linguistics.)
Berlin/New York: Mouton de Gruyter.
Deignan, Alice
2003
Metaphorical expressions and culture: An indirect link. Metaphor and
Symbol 18: 255271.
Deignan, Alice, Danuta Gabrys, and Agnieszka Solska
1997
Teaching English metaphors using cross-linguistic awareness-raising
activities. ELT Journal 51: 352360.
Dewell, Robert
1994
Over again: Image-schema transformations in semantic analysis. Cognitive Linguistics 5: 45180.
Dirven, Ren
1989
Cognitive linguistics and pedagogic grammar. In Linguistic Theorizing
and Grammar Writing, Gerhard Leitner, and Gottfried Graustein
(eds.), 5675. Tbingen: Max Niemeyer.
2001
English phrasal verbs: Theory and didactic application. In Applied
Cognitive Linguistics II: Language Pedagogy, Martin Ptz, Susanne
Niemeier, and Ren Dirven (eds.), 327. (Cognitive Linguistics Research 19b.) Berlin/NewYork: Mouton de Gruyter.
Dirven, Ren, and John R. Taylor
1994
English modality: A cognitive-didactic approach. In Perspectives on
English. Studies in Honour of Professor Emma Vorlat, Keith Carlon,
Kristin Davidse, and Brygida Rudzka-Ostyn (eds.). 542556. Leuven/
Paris: Peeters.
345
Ellis, Nick
1998
Emergentism, connectionism and language learning. Language Learning 48: 631664.
2002
Frequency effects in language processing. A review with implications
for theories of implicit and explicit language acquisition. Studies in
Second Language Acquisition 24: 143188.
Eubanks, Philip
2000
A War of Words in the Discourse of Trade: The Rhetorical Constitution
of Metaphor. Carbondale, Ill.: Southern Illinois University Press.
Evans, Vyvyan, and Andrea Tyler
2004
Spatial experience, lexical structure and motivation: the case of in. In
Studies in Linguistic Motivation, Gnter Radden, and Klaus-Uwe
Panther (eds.), 157192. (Cognitive Linguistics Research 28.) Berlin/
New York: Mouton de Gruyter:
forthcoming Applying cognitive linguistics to pedagogical grammar: The English prepositions of verticality. The Brazilian Journal of Applied Linguistics.
Fisher, Olga
1999
On the role played by iconicity in grammaticalisation processes. In Form
Miming Meaning: Iconicity in Language and Literature, Max Nnny, and
Olga Fisher (eds.), Amsterdam/Philadelphia: Benjamins.
Flower, John
2002
Phrasal Verb Organiser. Boston, MA: Thomson Heinle.
Gibbs, Raymond W.
1994
The Poetics of Mind: Figurative Thought, Language and Understanding.
Cambridge: Cambridge University Press
Goddard, Cliff
2000
Communicative style and cultural values: cultural scripts of Malay
(Bahasa Melayu). Anthropological Linguistics 42: 81106.
2004
Cultural scripts: A new medium for ethnographic instruction. In Cognitive Linguistics, Second Language Acquisition, and Foreign Language
Teaching, Michael Achard, and Susanne Niemeier (eds.), 143163.
(Studies on Language Acquisition 18.) Berlin/New York: Mouton de
Gruyter.
Goldberg, Adele E.
1995
Constructions: A Construction Grammar Approach to Argument Structure. Chicago: University of Chicago Press.
Goodale, Malcolm
1998
Collins Cobuild Phrasal Verb Workbook. London: Harper Collins Publishers.
Gries, Stefan Th., and Stefanie Wulff
2005
Do foreign language learners also have constructions?: Evidence from
priming, sorting, and corpora. In Annual Review of Cognitive Linguistics 3, Ruiz de Mendoza, Francisco. J. (ed.), 182200. Amsterdam/
Philadelphia: Benjamins.
346
Grundy, Peter
2004
The figure / ground Gestalt and language teaching methodology. In
Cognitive Linguistics, Second Language Acquisition, and Foreign Language Teaching, Michel Achard, and Susanne Niemeier (eds.), 119142.
(Studies on Language Acquisition 18.) Berlin/New York: Mouton de
Gruyter.
Haiman, John
1980
Dictionaries and encyclopaedias. Lingua 50: 32957.
Hannan, Peter
1998
Particles and gravity: Phrasal verbs with up and down. Modern English
Teacher 7: 2127.
Herrera, Honesto, and Michael White
2000
Cognitive linguistics and the language learning process: A case from
economics. Estudios Ingleses de la Universidad Complutense 8: 5578.
2002
Business is war or the language of takeovers. In Pragmatica y Analisis
del Discurso. Panorama Actual de la Linguistica Aplicada: Conocimiento, Procesamiento y Uso del Lenguaje, vol. 1, Mercedes Forns
Guardia., Juan Manuel Molina Valero, and Lorena Perez Hernandez
(eds.), 231239. Rioja, Spain: Universidad de la Rioja.
Holme, Randal
2001
Metaphor, language learning, and affect. Humanising Language Teaching 3. www.hltmag.co.uk. Pilgrims.
Howarth, Peter
1998
Phraseology and second language proficiency. Applied Linguistics 19:
2244.
Hllen, Werner
1992
Kognitive Linguistik Eine Mglichkeit fr den Grammatikunterricht. Ein neuer Vorschlag zur Lsung eines alten Problems. In Grammatica vivat. Konzepte, Beschreibungen und Analysen zum Thema
Fremdsprachengrammatik. In memoriam Hartmut Kleineidam, Alberto Barrera-Vidal, Manfred Raupach, and Ekkehard Zfgen (eds.),
1530. Tbingen: Gunter Narr.
Hunston, Susan
2002
Corpora in Applied Linguistics. Cambridge: Cambridge University
Press
Irujo, Suzanne
1993
Steering clear: Avoidance in the production of idioms. International
Review of Applied Linguistics 31: 205219.
Jkel, Olaf
2003
Motion metaphorized in the economic domain. In Motivation in Language (Studies in Honor of Gnter Radden), Hubert Cuyckens, Thomas
Berg, Ren Dirven, and Klaus-Uwe Panther (eds.), 297318. Amsterdam/Philadelphia: Benjamins.
James, Carl, and Peter Garrett (eds.)
1995
Language Awareness in the Classroom. London: Longman.
347
Johnson, Mark
1987
The Body in the Mind: the Bodily Basis of Meaning, Imagination, and
Reason. Chicago/London: University of Chicago Press.
Katz, Albert N.
1983
What does it mean to be a high imager? In Imagery, Memory and Cognition, John C. Yuille (ed.), 3963. Hillsdale, N.J.: Lawrence Erlbaum.
Kellerman, Eric
1978
Giving learners a break: Native language intuitions, a source of predictions about transferability. Working Papers on Bilingualism, 15: 5992.
Koller, Veronika
2002
A Shotgun Wedding: Co-occurrence of war and marriage metaphors
in mergers and acquisitions discourse. Metaphor and Symbol 17:
179203.
2003
Metaphor clusters, metaphor chains: Analyzing the multifunctionality
of metaphor in text. metaphorik (electronic journal) 5/2003.
2004
Businesswomen and war metaphors: Possessive, jealous and pugnacious? Journal of Sociolinguistics 8: 323.
Kvecses, Zoltn
1986
Metaphors of Anger, Pride and Love: A Lexical Approach to the Study of
Concepts. Amsterdam/Philadelphia: Benjamins.
2001
A cognitive linguistic view of learning idioms in an FLT context. In Applied Cognitive Linguistics II: Language Pedagogy, Martin Ptz, Susanne Niemeier, and Ren Dirven (eds.), 87115. (Cognitive Linguistics
Research 19b.) Berlin/NewYork: Mouton de Gruyter.
2005
Metaphor in Culture: Universality and Variation. Cambridge: Cambridge University Press.
Kvecses, Zoltn and Peter Szab
1996
Idioms: A view from Cognitive Semantics. Applied Linguistics 17:
326355.
Kvecses, Zoltn, Mariann Tth, and Bulcs Babarci
1996
A Picture Dictionary of English Idioms. Budapest: Etvs University Press
Kramsch, Claire
1993
Context and Culture in Language Teaching. Oxford: Oxford University
Press.
Krashen, Stephen D., and Tracy D. Terrell
1983
The Natural Approach: LanguageAcquisition in the Classroom. Hayward, CA: Alemany Press.
Kurtyka, Andrzej
2001
Teaching phrasal verbs: a cognitive approach. In Applied Cognitive Linguistics II: Language Pedagogy, Martin Ptz, Susanne Niemeier, and
Ren Dirven (eds.), 2954. (Cognitive Linguistics Research 19b.) Berlin/New York: Mouton de Gruyter.
Lakoff, George
1987
Women, Fire and Dangerous Things: What Categories Reveal about the
Mind. Chicago/London: University of Chicago Press.
348
2002
349
1999
350
351
Palmer, Gary
1996
Toward a Theory of Cultural Linguistics. Austin: Texas University Press.
Pawley, Andrew, and Frances H. Syder
1983
Two puzzles for linguistic theory: native-like selection and native-like
fluency. In Language and Communication, Jack C. Richards, and Richard W. Schmidt (eds.), 191225. London: Longman.
Ponterotto, Diane
1994
Metaphors we can learn by: how insights from cognitive linguistic research can improve the teaching/learning of figurative language. Forum
32: 27.
Prabhu, Nagore S.
1987
Second Language Pedagogy. Oxford: Oxford University Press.
Ptz, Martin, Susanne Niemeier, and Ren Dirven (eds.)
2001a Applied Cognitive Linguistics I: Theory and Language Acquisition. Berlin/New York: Mouton de Gruyter.
2001b Applied Cognitive Linguistics II: Language Pedagogy. (Cognitive Linguistics Research 19b.) Berlin/New York: Mouton de Gruyter.
Radden, Gnter, and Klaus-Uwe Panther
2004
Introduction: reflections on motivation. In Studies in Linguistic Motivation, Gnter Radden, and Klaus-Uwe Panther (eds.), 146. (Cognitive Linguistics Research 28.) Berlin/New York: Mouton de Gruyter:
Radden, Gnter, and Klaus-Uwe Panther (eds.)
2004
Studies in Linguistic Motivation. (Cognitive Linguistics Research 28.)
Berlin/New York: Mouton de Gruyter:
Read, John
2004
Plumbing the depths: how should the construct of vocabulary knowledge be defined? In Vocabulary in a Second Language: Selection, Acquisition and Testing, Paul Bogaards, and Batia Laufer (eds.), 209228.
Amsterdam/Philadelphia: Benjamins.
Rice, Sally
1996
Prepositional prototypes. In The Construal of Space in Language and
Thought, Martin Ptz, and Ren Dirven (eds.), 135165. (Cognitive
Linguistics Research 8.) Berlin/New York: Mouton de Gruyter.
2003
Growth of a lexical network: Nine English prepositions in acquisition.
In Cognitive Approaches to Lexical Semantics. Hubert Cuyckens, Ren
Dirven, and John R. Taylor (eds.), 243260. (Cognitive Linguistics Research 23.) Berlin/New York: Mouton de Gruyter
Riding, Richard, and Stephen Rayner
1998
Cognitive Styles and Learning Strategies: Understanding Style Differences in Learning and Behaviour. London: David Fulton Publishers Ltd.
Riggenbach, Heidi (ed.)
2000
Perspectives on Fluency. Ann Arbor: University of Michigan Press.
Robins, Shani, and Richard E. Mayer
2000
The metaphorical framing effect: metaphorical reasoning about textbased dilemmas. Discourse Processes 30(1): 5786.
352
Rohrer, Tim
1995
The metaphorical logic of (political) rape: The new wor(l)d order.
Metaphor and Symbolic Activity 10: 115137.
Rudzka-Ostyn, Brygida
1983
Cognitive Grammar and the structure of Dutch uit and Polish wy. Trier:
LAUT (Linguistics Agency of the University of Trier) Series A 109.
2003
Word Power: Phrasal Verbs and Compounds. A Cognitive Approach. Berlin; New York: Mouton de Gruyter.
Rudzka-Ostyn, Brygida, Paul Ostyn, Pierre Godin, and Francis Degreef
1991
Woordkunst. Une Synthse Cognitive et Communicative du Lexique de
Base du Nerlandais. Brussels: Plantyn.
Rundell, Michael
2001
Cats, conversations, and metaphor. Humanising Language Teaching, 3.
www.hltmag.co.uk. Pilgrims.
2002
Metaphorically speaking. English Teaching Professional 23: 21.
2005a Why are phrasal verbs so difficult? Humanising Language Teaching, 7.
www.hltmag.co.uk. Pilgrims.
Rundell, Michael (ed.)
2005b Macmillan Phrasal Verbs Plus. Hampshire: Macmillan.
Salager-Meyer, Francoise
1990
Metaphors in medical English prose: A comparative study with French
and Spanish. English for Specific Purposes 9: 14559.
Sandra, Dominiek, and Sally Rice
1995
Network analyses of prepositional meaning: Mirroring whose mind
the linguists or the language users? Cognitive Linguistics 6: 89130.
Schffner, Christina
1991
Zur Rolle von Metaphern fr die Interpretation der aussersprachlichen
Wirklichkeit. Folia Linguistica 25: 7590.
1995
Metapher als Bezeichnungsbertragung?. In Wort und Wortschatz:
Beitrge zur Lexikologie, Inge Pohl, and Horst Ehrhardt (eds.),
175184. Tubingen: Max Niemeyer.
1996
Building a European house? Or at two speeds into a dead end? Metaphors in the debate on the United Europe. In Conceiving of Europe
Unity in Diversity, Andreas Musolff, Christina Schffner, and Michael
Townson (eds.), 3159. Aldershot, U.K.: Dartmouth.
Schmitt, Norbert
1997
Vocabulary learning strategies. In Vocabulary: Description, Acquisition
and Pedagogy, Norbert Schmitt, and Michael McCarthy (eds.),
199236. Cambridge: Cambridge University Press.
Schmitt, Norbert (ed.)
2004
Formulaic Sequences: Acquisition, Processing and Use. Amsterdam/
Philadelphia: Benjamins.
Scott, Mike
1994
Metaphors and language awareness. In Reflections on Language Learning, Leila Barbara, and Mike Scott (eds.), 89104. Clevedon, U.K.:
Multilingual Matters.
353
Segalowitz, Norman
2000
Automaticity and attentional skill in fluent performance. In Perspectives on Fluency, Heidi Riggenbach (ed.), 200219. Ann Arbor: University of Michigan Press.
Sinclair, John
1991
Corpus, Concordance, Collocation. Oxford: Oxford University Press.
Skehan, Peter
1998
A Cognitive Approach to Language Learning. Oxford: Oxford University Press.
Skoufaki, Sophia
2005
Use of conceptual metaphors: A strategy for the guessing of an idioms
meaning? In Selected Papers on Theoretical and Applied Linguistics
from the 16th International Symposium, April 1113, 2003, Marina
Mattheoudakis, and Angeliki Psaltou-Joycey (eds.), 542556. Thessaloniki: Aristotle University of Thessaloniki.
Skmen, Anita J.
1997
Current trends in teaching second language vocabulary. In Vocabulary:
Description, Acquisition and Pedagogy, Norbert Schmitt, and M.
McCarthy (eds.), 237257. Cambridge: Cambridge University Press.
Speake, Jennifer (ed.)
1999
Oxford Dictionary of Idioms. Oxford: Oxford University Press.
Squire, Larry R., and Eric R. Kandel
2000
Memory: From Mind to Molecules. New York: Scientific American Library.
Stengers, Hlne, June Eyckmans, Arnout Horemans, and Frank Boers
2004
Idiom Teacher: Problems in designing a CALL package. Humanising
Language Teaching 6. www.hltmag.co.uk. Pilgrims.
Stengers, Hlne, Frank Boers, and June Eyckmans
2005
Putting memory theories to the test with an on-line instrument for
teaching idioms. Paper presented at The many faces of phraseology conference, Louvain-la-Neuve, Belgium: Universit Catholique de Louvain-la-Neuve.
Stevick, Earl
1986
Images and Options in the Language Classroom. Cambridge: Cambridge University Press.
Sweetser, Eve
1990
From Etymology to Pragmatics: Metaphorical and Cultural Aspects of
Semantic Structure. Cambridge: Cambridge University Press.
Taylor, John R.
1988
Contrasting prepositional categories: English and Italian. In Topics in
Cognitive Linguistics, Brygida Ruzdka-Ostyn (ed.), 299326. Amsterdam/Philadelphia: Benjamins.
1993
Some pedagogical implications of cognitive linguistics. In Conceptualizations and Mental Processing in Language. A Selection of Papers from
the First International Cognitive Linguistics Conference in Duisburg,
354
355
356
357
358
359
Abstract
Signed languages are articulated with gestures of the hands and face in an articulation space that is visible to the addressee. In many ways, signed languages and
spoken languages pattern similarly, although the phonemic material of each is
rather obviously radically different. Evidence has been mounting in cognitive linguistic research that while human, embodied conceptual structures and domains
are expressed in linguistic structure across both spoken and signed languages, the
articulation system of signed language enables a higher level of iconic relationship
between the conceptual world and linguistic expression. This plays out in numerous ways, including the appearance of iconic relationships at all levels of grammar
and discourse, in metonymic and metaphoric relationships, and in so-called simultaneous constructions that include body partitioning, in which various parts of
the signers body represent different entities simultaneously, and the expression of
shifting points of view in discourse. In signed languages, visible articulation is particularly suited to the mapping between conceptual structure and linguistic expression because in both action schemas and visible attributes, visual imagery is
highly salient, and the hands and bodies of signers are well-suited to reflecting this
imagery iconically.
Keywords: signed language; gesture; iconicity; metaphor; simultaneity; visible articulation
1.
Introduction1
360
Terry Janzen
2.
Visible articulation
In the late 1950s and early 1960s William Stokoe presented evidence that,
far from being holistic, non-linguistic gestures, the signs that Deaf people
articulated with their hands and used with each other patterned in ways
analogous to the words of spoken languages (Stokoe 1960; Stokoe, Casterline and Croneberg 1965). Of particular relevance was the analysis of
individual signs as resulting from the combination of definable elements
that belonged to a restricted number of categories, those of handshapes,
locations (i.e., points of contact or positions in space), and movement
types. These were thought to combine in arbitrary ways in much the same
way as do the arbitrary sounds that together make up spoken words. Over
the next number of years much effort was subsequently expended produc-
361
ing evidence that not only did individual signs conform to the organizing
principles of linguistic material, despite the obvious difference in means
of articulation, but so too did constructions larger than the word. Phrase
structure and sentence-level organization were thought to pattern no differently from that of spoken languages. At the level of articulation, however, differences between signed and spoken languages are profound.
A second crucial difference is that signed languages are seen and not
heard. The reception of linguistic structure visually has numerous consequences for those structures, some specifics of which are the subject of
this chapter, for example various iconic structures (section 4 below),
structures with multiple, simultaneously articulated elements (section 6),
and body partitioning and perspective coding constructions (section 6.1).
In the absence of the ability to perceive language as sound, Deaf communities worldwide have developed signed languages that are not substitutes
for spoken languages, that is, they are not visual versions of spoken languages but have evolved to take advantage of the complementary properties of both vision and the physiology of the human body (Armstrong,
Stokoe and Wilcox 1995). One advantage of vision over hearing for language reception is that many aspects of a visual stimulus such as a motor
action can be perceived simultaneously without compromising the perception of the signal as a whole. As a non-linguistic example, when we
watch someone jump to catch a ball, we can see (among other things) the
persons legs bend and then extend to propel him into the air, the arms
reaching toward the ball, and the ball arching toward the person all as
simultaneous actions, and we create an action schema for what the combined set of actions mean. In a visible linguistic sequence, we may see a
persons two hands, each having a distinct, albeit changing shape configuration, moving independently through space, perhaps colliding, along
with some particular facial gestures and intentionally directed eye gaze.
The ability to manipulate parts of the body simultaneously in this way
and to perceive and process visible components in such combinations has
a profound effect on how linguistic constructions can be composed. For
several decades researchers have maintained that the most significant differences between signed and spoken languages are found at the phonetic
level whereas at higher levels of organization the differences all but disappear (Brentari 1998). However, we are now seeing evidence that highly
complex morphological and syntactic constructions may also show some
unique patterning unlike what is attested for spoken languages (cf. Meier
2002). One major consequence of the features of signed language articulation combined with visual perception is that semantic categories can at
362
Terry Janzen
times be much more gradient than is the case for spoken language (Talmy
2003, 2006). Talmy suggests that this is what takes place within the classifier systems of signed languages, where the hands as objects move about
the signers articulation space as representations of objects and their relationships in some actual space.
Describing the linguistic system of a signed language has not been without its problems. Stokoe and his colleagues (1965) delineated categories of
phonemic units2 that were understood to be finite sets of meaningless
parts handshapes, locations, and movements that combined to create
meaningful words, as duality of patterning has always maintained is the
case in spoken language. But more recent work (e.g., Brentari 1998; Liddell 1984, 2003; Liddell and Johnson 1989) has suggested that these fairly
simple categories are not entirely comparable to the phonemes of spoken
languages. Rather, some elements such as handshape and location are
better understood as features of timing units and, in certain cases, movement features are often transitional and epiphenomenal. The distinction
between what is gestural and what is linguistic can be difficult to determine; locations that are not contact points on the body have been argued
to be gestural rather than purely linguistic (Liddell 2003). At higher levels
of grammar, descriptive problems also arise. For example, in complex
morphological forms each hand may contribute a meaningful unit distinct from the other, and further, meaning bearing facial gestures may cooccur. It is usually the case that such facial gestures constitute bound
morphology (Janzen 2004b; Wilcox 2004a), but how such morphology is
structurally bound to the root articulated by the hands is unclear. A
further difficulty arises when each of the two hands articulates a different
root, such as in classifier-based verb complexes. In many such cases, the
two hands refer to two distinct entities involved together in some event,
for example as agent and patient nominals, but whether they should be
considered as separate morphological units or as parts of a larger, single
complex is uncertain (Janzen, to appear).
Visible articulation means that the hands and body form signs that
symbolize concepts in the human mind. As we will see in the discussion
below, a visual orientation to ones world of experience, which includes
experiencing linguistic form, allows for the signer to map the conceptualization of events and space onto a visible articulation space. It should be
no surprise, then, that the potential for iconic relationships is high: what is
conceptualized in the mind and what is symbolized in articulation can be
closely aligned. Wilcox (2004a) notes that the hands as physical objects in
space can be manipulated by the signer to reflect features of objects and
363
3.
The relationship between gesture and signs can be approached from several angles. In evolutionary terms it has been argued that gesture has
played a crucial role in the development of all language (e.g., Armstrong,
Stokoe and Wilcox 1995; Corballis 2002; Wilcox 2002). Regarding the relationship between gesture and speech, there is evidence that the two are
an integrated system (Kendon 2004; McNeill 1992), which begs the question of whether a similar integration is possible for gesture and signed languages. Emmorey (1999) argues that it is clear that signers do gesture, but
because the medium is the same (hands, generally), gestures and linguistic
material do not occur simultaneously. Sandler (2006) shows that in Israeli
Sign Language a modality switch takes place: because the hands are occupied with linguistic expression, gestures are simultaneously found on
the mouth.
364
Terry Janzen
365
tween what is gesture and what is within a signed language system does
not lead us in the right direction. Continuous accounts (e.g., Armstrong,
Stokoe and Wilcox 1995; Janzen and Shaffer 2002; Wilcox 2002, 2004b)
suggest generally that what originates as gestural eventually regularizes
and conventionalizes as lexical or grammatical items (cf. Haiman 1994)
but that the transition from gesture to language is not abrupt. In grammaticization terms, overlap may occur, so that different senses and functions of an item co-occur in usage simultaneously (Hopper 1991). In the
case of the development of signs and grammatical morphemes in a signed
language, this overlap may mean that an item that has been regularized
into the linguistic system may still be present as a gesture. To cite one
example, Janzen (1998a, 1999) and Janzen and Shaffer (2002) propose
that both yes/no question and topic marking morphology in ASL developed out of a more generalized questioning gesture that includes a facial
expression with the eyebrows raised. While this facial feature is required
by the grammar in these constructions, it is the case that raising the eyebrows as a questioning gesture occurs frequently for both hearing and
Deaf people in North America and elsewhere.
4.
366
Terry Janzen
logical in tops. But even in lexical items in ASL we find handshapes that
are quite obviously iconic, and thus they bring a certain amount of meaning to the composite word. The verb MEET, for example, uses extended
upright index fingers, and the noun BOOK has two flat /B/ handshapes
that iconically represent the flatness of book pages. Janzen (1998b) suggests that in the verb UNDERSTAND, which involves a handshape change
from a closed fist (/S/) to an extended index finger (/1/), the initial closed
fist handshape metaphorically represents a closed state of not (yet)
understanding and the final extended index finger handshape represents
the realized state of subsequently having understood. Note that Wilcox
(2000) describes numerous instances in which an idea or thought is metaphorically represented by a /1/ handshape.
Iconicity abounds at higher levels as well. Morphological categories associated with verbs are often tightly bound to verb forms, such as aspect
and agreement morphemes that alter the movement path and the beginning and endpoint locations of the path respectively. At the sentence level,
ASL and many other signed languages use topic-comment constructions
that iconically represent known information first, new information second (Ingram 1978; Janzen 1998a). Shaffer (2004) demonstrates that
modal verbs are positioned pre-verbally when their scope is just the action
coded by the verb, and clause-final when the entire proposition is under
their scope. Even at the discourse level, Winston (1995) for example shows
that space is used iconically as a visual representation of items being compared even if the items are highly abstract concepts, and Sallandre (2006)
reports that in French Sign Language (LSF) narratives, the iconic use of
space serves an organizing function for the narrative events because threedimensional articulation space preserves a three-dimensional real-world
event. In fact, the iconic representation of spatial elements has been described in some detail, for example in Emmorey (2002) for ASL and in
Engberg-Pedersen (1993) for Danish Sign Language. Talmy (2003, 2006)
lists some thirty spatial categories that may be represented in the linguistic
structure of signed languages, many of which are combinable in, for
example, classifier constructions. Talmy concludes that the iconicity available in a signed language in terms of how space is represented in linguistic
form is much greater than for spoken language because there is virtually
no way to represent space structurally with the vocal apparatus.
Taub (2001) gives an inventory of iconic categories in ASL that includes, in part, form-meaning relationships in shape (i.e., a handshape or
the path that the hand takes represents the shape of a referent), movements, locations as representing locations in mental spaces, size, number,
367
and temporal ordering. Taub defines iconicity as not an objective relationship between image and referent; rather, it is a relationship between
our mental models of image and referent (Taub 2001: 19) and as the
existence of a structure-preserving mapping between mental models of
linguistic form and meaning (2001: 23). Such relationships are largely
motivated by our embodied human experiences. Taub claims that references (or sources) in iconic structures are much more abundant in the visual, as opposed to auditory, modality of signed language, leading to an
abundance of iconic structures in these languages as opposed to spoken
languages: Thus, users of a visual/gestural language will be able to draw
on a far wider range of sensory images than will users of aural/oral languages. A much greater percentage of the languages concepts have the potential for iconic representation (2001: 61). In comparing the prevalence
of iconic structures in signed versus spoken languages, Taub says that it is
a consequence of the fact that most phenomena do not have a characteristic noise to be used in motivating a linguistic form [in spoken languages] (2001: 3). I would suggest, however, that it is not so much a characteristic noise but that, in our observation of, or interaction with, an
object or event, it is the saliency of a sound associated with the item that
might spark a linguistic string naming the item to be iconic. As sound salience increases for spoken language users, imagic iconicity (Haiman
1980) emerges. Taub uses the example of crash, suggesting that the forceful impact of one object on another is one concept that has a characteristic sound, and thus we choose (in English) an iconic word that resembles
this characteristic sound. But here we might say that the sound of such
physical impact is highly salient (but not characteristic) in our experience of the event. Saliency in this respect is scalar: the more salient a
sound is to our experience, the more likely it is that we might choose an
iconic symbol for it.
Much about our embodied experience in the world is visual, thus we
might say that visual aspects of what we experience are often highly salient. So for a language with visible articulation, the visual imaging that
translates into iconic structure takes two general forms: action schemas
and visible attributes. But once again, this can extend to highly abstract
notions metaphorically (cf. Johnson 1987 on image schemas). Fictive
motion, for example, refers to expressions that represent a movement in
the conceptual domain where none actually exists in the real world (Wilcox 2003). Wilcox describes such expressions in signed language where a
movement is iconically associated with a source domain that does not
itself move. The ASL sign ALWAYS has a large circular movement path
368
Terry Janzen
that represents, according to Wilcox, un-ending or imperfective time. Wilcox also cites Shaffers (2002) analysis of CANT in ASL, which she describes as articulated action on one hand that encounters a stationary
barrier articulated by the other. Of interest is that these examples have an
iconic property that relates linguistic form to conceptualized form and
not real-world form because no real-world form exists.
4.1. Cognitive iconicity
Wilcox (2004a) proposes a novel view of iconicity, referred to as cognitive iconicity, which Wilcox defines not as a relation between the form
of a sign and what it refers to in the real world, but as a relation between
two conceptual spaces. Cognitive iconicity is a distance relation between
the phonological and semantic poles of symbolic structures (2004a:
122). In much of language, the phonological pole and semantic pole of
symbolic structure, both conceptual entities, reside in very different regions of conceptual space. When this is the case, the linguistic item has no
formal resemblance to its meaning. However, if the two poles are in close
proximity, it means that the conceptualizer construes the two as largely
similar in some iconic way. Wilcox uses the example of the ASL sign
SLOW, which can be tempered to mean VERY-SLOW. The phonological
pole for SLOW includes a slowly articulated path movement, but for
VERY-SLOW, the movement increases in speed and intensity so as to
quickly reach its final ending point. Items such as this have long been considered examples of a general principle of historical change from iconic to
arbitrary in the lexicon of signed language (Frishberg 1975) in which the
iconicity seen in the lexical word is submerged when it appears in a grammatical structure (Klima and Bellugi 1979), but Wilcox argues that while
certain iconic properties may submerge when individual items are inserted
into grammatical constructions, others emerge. For VERY-SLOW, the increased intensity in the movement iconically represents the conceptualized notion of intensity which may be applied to a host of adverbs and adjectives. The difference between SLOW and VERY-SLOW is a difference in
what is profiled, either slowness or the intensity of the attribute. The
super-imposition of pervasive iconic structures has also been noted in
French Sign Language and are central to Sallandre and Cuxacs (2002)
theory of the organization of signed language grammar.
5.
369
370
6.
Terry Janzen
371
hand (usually the non-dominant hand). All in all, Leeson and Saeed conclude that what these simultaneous construction strategies have in common is that the dominant hand marks foregrounded information while
the non-dominant hand marks backgrounded information.
6.1. Blended spaces, body partitioning and the expression of point
of view
In many types of simultaneously articulated constructions, different parts
of the signers body represent different entities which are described as participating together in a single event. This is the position of Dudis (2004a,
2004b), who terms what is taking place in the signers articulation as
body partitioning. Dudis shows that in addition to the two hands being
able to depict two distinct entities, the face may also participate. In one
example from Dudis (2004a: 235) in which a librarian glares at a library
patron, the signers body depicts the patron while the signers right hand
forms the verb STARE-AT, signifying the librarians gaze toward the patron. But in addition to this, the signers facial gestures further depict the
librarians expression, while his eye gaze depicts the patron looking back
at the librarian by being directed toward the verb articulated on the right
hand. The combined result constitutes a blended space conceptually for
the signer and addressee and demonstrates the complexity possible within
a simultaneously articulated construction. This further suggests that such
constructions may be constrained in part by the ability of humans to view
and assimilate multiple aspects of visual images cognitively, and when the
mode of expression allows it, to articulate this complexity in an iconically
simultaneous manner.
The expression of point of view in verb complexes in ASL is similarly
complex and exemplifies an iconic relationship between how the signer
construes a scene and the use of space. In such cases (Janzen 2004a, 2005,
to appear) the signers own vantage point coincides with some entitys perspective on a scene which may or may not be her own. In one type of construction, the signer mentally rotates her conceptualized space so that
others perspectives align with her own stance and view of the scene,
whereas what has typically been described in the literature is the case
where the signer shifts in her physical space to align her body with the spatially located position of some other referent. In a mentally rotated space,
every point of reference also rotates. In the narratives that are described
in Janzen (2004a, 2005) the signer enacts the utterances and actions of
two interacting referents positioned opposite one another and rotates the
372
Terry Janzen
entire conceptualized space, inviting the addressee to view the scene from
one referents perspective and then the other. The result is that the signer
maintains two spaces mentally: each is an arrangement of entities from
one of the referents vantage points. In order to keep track of the signers
spatial referencing, the addressee must understand which viewed space is
being referred to at each moment in the discourse.
7.
Conclusion
In this discussion I have outlined several aspects of signed language expression that exemplify the visual nature of its articulation and reception.
Early descriptions of signed languages concentrated on their organizational similarities with spoken languages, but clearly there are important
differences to examine. These differences neither detract from the reality
that signed languages are complex linguistic systems nor make them
anomalous. Rather, we can learn much about human conceptual organization through the study of signed language because the mapping between
conceptual space and articulation space is so often expressly iconic. A
cognitively oriented approach to research on signed languages allows us
to learn much about human conceptual structures.
What might be future directions for cognitive linguistic research on
signed language? Most certainly we are experiencing a resurgence of analyses of iconic relationships between conceptual and linguistic structure,
but with the shroud lifting, as Phyllis Wilcox so eloquently puts it, perhaps we have yet to see through the veil. Signed languages seem to be the
epitome of embodied language: physical and visual surely there is more
to discover here. As well, we have begun to uncover the gestural roots of
both lexicon and grammar in signed languages. The gesture-language interface that Wilcox (2002) speaks of is particularly interesting when the
medium for both gesture and language is the same. Cognitive linguistics
embraces rather than rejects the connection between language and gesture, and so we might expect further enlightening work in this regard. Finally, vision is taking an increasingly prominent role in the description of
language structure and language usage altogether. In light of this, language that is purely of a visual nature has much to offer our understanding of the embodied experience of interactions with each other, our environments, and our linguistic expression of that experience.
373
Notes
1. There are many colleagues with whom I have had helpful discussions on cognitive linguistics and signed language, but in particular, I am indebted to Sherman Wilcox and Barbara Shaffer for their insights, without which this chapter
could not have been written. Any errors, of course, remain my own.
2. Stokoe et al. (1965) refer to these categories as aspects and to cheremes as
opposed to phonemes. Phonologists since Stokoe, however, have tended to retain terminology more widely used in the discipline of phonology.
References
Armstrong, David F., William C. Stokoe, and Sherman E. Wilcox
1995
Gesture and the Nature of Language. Cambridge: Cambridge University Press.
Battison, Robbin
1978
Lexical Borrowing in American Sign Language. Silver Spring, MD: Linstok Press.
Brentari, Diane
1998
A Prosodic Model of Sign Language Phonology. Cambridge, MA: MIT
Press.
Browman, Catherine P., and Louis Goldstein
1989
Articulatory gestures as phonological units. Phonology 6: 201251.
Boyes Braem, Penny
1981
Features of the handshape in American Sign Language. Ph.D. diss.,
University of California, Berkeley.
Corballis, Michael C.
2002
From Hand to Mouth: The Origins of Language. Princeton/Oxford:
Princeton University Press.
Dudis, Paul G.
2004a Body partitioning and real-space blends. Cognitive Linguistics 15(2):
223238.
2004b Depiction of events in ASL: Conceptual integration of temporal components. Ph.D. diss., University of California, Berkeley.
Emmorey, Karen
1999
Do signers gesture? In Gesture, Speech, and Sign, Lynn Messing, and
Ruth Campbell (eds.), 133159. New York: Oxford University Press.
2002
Language, Cognition, and the Brain: Insights from Sign Language Research. Mahwah, NJ: Lawrence Erlbaum.
Engberg-Pedersen, Elisabeth
1993
Space in Danish Sign Language: The Semantics and Morphosyntax of
the Use of Space in a Visual Language. Hamburg: Signum-Verlag.
Frishberg, Nancy
1975
Arbitrariness and iconicity: Historical change in American Sign Language. Language 51: 676710.
374
Terry Janzen
Haiman, John
1980
The iconicity of grammar: Isomorphism and motivation. Language 56:
514540.
1994
Ritualization and the development of language. In Perspectives on
Grammaticalization, William Pagliuca (ed.), 328. Amsterdam/Philadelphia: Benjamins.
Hopper, Paul J.
1991
On some principles of grammaticization. In Approaches to Grammaticalization: Volume 1, Focus on Theoretical and Methodological Issues,
Elizabeth C. Traugott, and Bernd Heine (eds.), 1735. Amsterdam/
Philadelphia: Benjamins.
Hulst, Harry van der, and Anne Mills
1996
Introduction. Issues in sign linguistics: Phonetics, phonology and morpho-syntax. In Sign Linguistics: Phonetics, phonology and Morpho-syntax, Harry van der Hults, and Anne Mills (eds.), Lingua (special issue)
98 (13): 317.
Ingram, Robert M.
1978
Theme, rheme, topic, and comment in the syntax of American Sign
Language. Sign Language Studies 20: 193218.
Janney, Richard W.
1999
Words as gestures. Journal of Pragmatics 31: 953972.
Janzen, Terry
1998a Topicality in ASL: Information ordering, constituent structure, and
the function of topic marking. Ph.D. diss., University of New Mexico,
Albuquerque, NM.
1998b Multi-level iconic relationships in American Sign Language grammar.
In Proceedings of the first High Desert Linguistics Society conference,
April 34, 1998, Catie Berkenfield, Dawn Nordquist, and Angus
Grieve-Smith (eds.), Vol. 1, 159172. Albuquerque, NM: High Desert
Linguistics Society.
1999
The grammaticization of topics in American Sign Language. Studies in
Language 23 (2): 271306.
2004a Space rotation, perspective shift, and verb morphology in ASL. Cognitive Linguistics 15 (2): 149174.
2004b The expression of grammatical categories in signed languages. Rome,
Italy. Lingue verbali e lingue dei segni: Confronti di strutture, construtti e metodologie (Verbal and Signed Languages: Comparing
Structures, Constructs and Methodologies). International Colloquium, Universit Roma Tre, October 4 5, 2004.
2005
Perspective Shift Reflected in the Signers Use of Space. CDS/CLCS
Monograph Number 1. Dublin: Centre for Deaf Studies, University of
Dublin, Trinity College.
to appear Perspective shifts in ASL narratives: The problem of clause structure. In
Language in the Context of Use: Usage-Based Approaches to Language
and Language Learning. Selected Papers from Georgetown University
375
376
Terry Janzen
377
378
Terry Janzen
379
Abstract
Conceptual metaphor theory (CMT) has over the past 25 years amply sought to
underpin the claim that humans pervasive use of verbal metaphor reflects the fact
that they think largely metaphorically. If this tenet of CMT is correct, metaphor
should manifest itself not just in language but also via other modes of communication, such as pictures, music, sounds, and gestures. However, non-verbal and
multimodal metaphor have been far less extensively studied than their verbal
sisters. The present article provides a review of work done in this area, focusing on
a number of issues that require further research. These issues include the proposal
to distinguish between monomodal and multimodal metaphor; reflections on the
difference between structural and creative metaphor; the question of how verbalizations of non-verbal or conceptual metaphors may affect their possible interpretation; thoughts as to how similarity between target and source is created; and
suggestions about the importance of genre for the construal and interpretation of
metaphor.
Keywords: Monomodal and multimodal metaphor; pictorial metaphor; structural
versus creative metaphor; similarity in metaphor; genre.
1.
Introduction
Andrew Ortonys edited volume Metaphor and Thought (1979) and Lakoff
and Johnsons monograph Metaphors We Live By (1980) were milestone
publications in the sense that they marked the switch from research into
metaphor as a primarily verbal to a predominantly conceptual phenomenon. The conceptual metaphor theory (CMT), as the LakoffianJohnsonian model is habitually referred to, has been a very productive
one (e.g., Gibbs 1994; Johnson 1987, 1993; Kvecses 1986, 2000, 2002;
Lakoff 1987, 1993; Lakoff and Johnson 1999, 2003; Lakoff and Turner
1989; Sweetser 1990; Turner 1996). A key notion in this theory is that the
mind is inherently embodied, reason is shaped by the body (Lakoff and
Johnson 1999: 5; Chapter 3). Very briefly, what this means is the follow-
380
Charles Forceville
ing. Human beings find phenomena they can see, hear, feel, taste and/or
smell easier to understand and categorize than phenomena they cannot.
It is perceptibility that makes the former phenomena concrete, and the
lack of it that makes the latter abstract. In order to master abstract concepts, humans systematically comprehend them in terms of concrete concepts. Thus abstract concepts such as life, time, and emotions are systematically understood in terms of concrete phenomena. Life is
understood as a journey (Hes without direction in his life; Im at a crossroads in my life) but also, for instance, as a story (Tell me the story of
your life; Lifes a tale told by an idiot ). Time is comprehended in
terms of spatial motion (The time for action has arrived; Time is flying
by; He passed the time happily). Emotions are typically represented by
drawing on the domain of forces. (I was overwhelmed; I was swept off my
feet; examples from Kvecses 2000; Lakoff 1993; Lakoff and Johnson
1980). Conceptualisations of many phenomena, CMT proposes, have
deeply entrenched metaphorical forms, in which the metaphors target
(topic, tenor) is abstract and its source (vehicle, base) is concrete. A metaphors interpretation boils down to the mapping of pertinent features
from the source to the target; a mapping that in the case of entrenched
metaphors such as the above occurs automatically. Since concreteness
is apprehended perceptually, metaphorical source domains are strongly
rooted in the functioning of the human body. Metaphorical reasoning is
thus governed by the arch metaphor mind is body (Lakoff and Johnson
1999: 249). A more recent development rooted in CMT is blending theory (Fauconnier and Turner 2002). Rather than postulating a target and
a source domain, it presents two (or more) input spaces. The input
spaces have both shared and unique characteristics, and it is this combination that allows for the construal of a so-called blended space. Blending theory is a, mainly descriptive, model claiming to be superior to metaphor theory in being able to account for ad hoc forms of creativity,
metaphorical and otherwise. Hitherto it cannot quite convince (for a critical review, see Forceville 2004a), but new work, taking into account pragmatic rhetorical factors, is promising (see Coulson and Pascual, in press;
Terkourafi and Petrakis, under review).
CMT has inspired conferences (e.g., those organized by the International Cognitive Linguistics Association, and the Researching and Applying Metaphor [RaAM] association), journals (e.g., Metaphor and Symbol, Cognitive Linguistics), as well as empirical research (for references see
Gibbs 1994; Lakoff and Johnson 1999: 58788). Its importance is evident:
if CMT is basically correct, it provides crucial insights into what, thanks
381
382
Charles Forceville
2.
In order to distinguish multimodal metaphor from monomodal metaphor, it should first be further clarified what is meant by mode. This is
no easy task, because what is labelled a mode here is a complex of various
factors. As a first approximation, let us say that a mode is a sign system interpretable because of a specific perception process. Acceptance of this
approach would link modes one-on-one to the five senses, so that we
would arrive at the following list: (1) the pictorial or visual mode; (2) the
aural or sonic mode; (3) the olfactory mode; (4) the gustatory mode; and
(5) the tactile mode. However, this is too crude a categorization. For instance, the sonic mode under this description lumps together spoken language, music, and non-verbal sound. Similarly, both written language and
gestures would have to be part and parcel of the visual, since one cannot
hear, smell, taste, or touch either conventionally written language or gestures (although a blind person feels Braille language and, by touch, perceives certain gestures for instance those of a statue). If justice is to be
done to these distinctions (between images and gestures, between spoken
and written language, between spoken language, sounds, and music),
other factors need to be taken into account, such as the manner of production (e.g., printed versus Braille letters in relief on paper; signs made
with parts of the body versus signs whose use is governed by the grammar
and vocabulary rules of a natural language). There are other problematic
issues. For instance, what is music and what mere sound may differ
from one culture, or period, to another. Similarly, it is impossible to assess
objectively where music shades off into sound effect. And is typeface to
be considered an element of writing, of visuals, or of both? In short, it is at
this stage impossible to give either a satisfactory definition of mode, or
compile an exhaustive list of modes. However, this is no obstacle for
postulating that there are different modes and that these include, at least,
383
the following: (1) pictorial signs; (2) written signs; (3) spoken signs; (4)
gestures; (5) sounds; (6) music (7) smells; (8) tastes; (9) touch.
We can now provisionally define monomodal metaphors as metaphors
whose target and source are exclusively or predominantly rendered in one
mode. The prototypical monomodal metaphor is the verbal specimen that
until recently was identical with metaphor tout court, and which has
yielded thousands of studies (de Knop et al. 2005; Shibles 1971; Van
Noppen et al. 1985; Van Noppen and Hols 1990). A type of monomodal
metaphor that has more recently become the subject of sustained research
is pictorial or visual metaphor.1
An early discussion of metaphor in pictures is Kennedy (1982). This
perception psychologist takes metaphor in the all-encompassing sense
of what literary scholars call a trope or a figure of speech (and which
Tversky 2001 calls figures of depiction) and identifies some 25 types, including for instance metonymy, hendiadys, and litotes. Kennedys
attempts to describe an extensive catalogue of figures of depiction using
the tenor/vehicle distinction that Richards (1965) specifically coined for
metaphor are sometimes strained. There are other problems. It is a
matter for debate whether the names he selects for his examples are always
necessarily the best ones, and each trope is illustrated with one or two
examples only, making generalizations difficult (the same problem also
adheres to Durand 1987). This having been said, Kennedy makes a
number of points that are illuminating for a theory of pictorial/visualmetaphor-in-the-narrow-sense. In the first place he argues that for a phenomenon to be labelled a visual metaphor it should be understandable as
an intended violation of codes of representation, rather than as being due
to carelessness or error. Secondly, Kennedy emphasizes that target and
source are, in principle, irreversible, which ensures that what he labels
metaphor remains commensurate with a generally accepted criterion in
theories of verbal metaphor. Thirdly, Kennedy introduces the helpful notion of runes: the kind of non-iconic signs used profusely in comics and
cartoons to indicate speed, pain, surprise, happiness, anger and many
other phenomena by means of straight or squiggly lines, stars, bubbles
etc. surrounding characters or moving objects (see also Smith 1996). In
later work, Kennedy elaborates on his theoretical work in various experiments. Kennedy (1993) reports, among other things, how congenitally
blind children metaphorically draw a spinning wheel.
Whittock (1990) describes cinematographic metaphor. While his numerous examples are subsumed under ten subtypes, and thus are less
wide-ranging than those by Kennedy, they still go beyond metaphor-in-
384
Charles Forceville
385
3.
Lakoff and Turner (1989) have argued that not only metaphors occurring
in everyday verbal communication can be traced back to conceptual
metaphors, but also those in artistic texts, specifically poetry. Particularly
when poems thematize abstract concepts such as life and death, they cannot but draw on the same conceptual metaphors that permeate non-artistic language. Thus Lakoff and Turner cite, and richly illustrate, many
passages featuring life and time as metaphorical targets (life is a jour-
386
Charles Forceville
387
more, or less, instances of a given metaphor. Gevaert (2001), basing herself on corpus-based data, questions for instance the embodied, timeless status of anger is heat claimed for conceptualisations of anger in
Lakoff (1987). She demonstrates, among other things, that in the Old
English period swelling was a much more important source domain in
anger metaphors than heat, and speculates that the latters growing
popularity in more recent periods may well be due to the humoral theory
that dominated mediaeval times; that is, to cultural no less than embodied
knowledge (see also Gevaert 2005). Moreover, a systematic, corpus-based
analysis might reveal that many poetic metaphors are not so easily amenable to conventional ones. Numerous poetic metaphors may simply not
have abstract concepts such as life, death, time, purpose as their target
domain. As pointed out by Grady, they may more often be resemblance
metaphors than, possibly (near)universal, correlation metaphors
(Grady 1999). This would not invalidate Lakoff and Turners impressive
findings, but their one-sided emphasis on correlation and generic-level
metaphors2 in poetry may inadvertently lead to an uncritical acceptance
of the view that most poetic metaphors are of this kind.
One important difference between conventional and idiosyncratic
metaphors is that the interpretation of the latter is, by definition, far less
governed by entrenched, pre-existing correspondences between the schematic structures in target and source. It is only by downplaying this difference that Lakoff and Turner can say that the preservation of genericlevel structure is, we believe, at the heart of metaphorical imagination,
whether poetic or ordinary (1989: 83; for critical accounts of this view
see also Stockwell 1999; Crisp 2003).
Secondly, we should not forget that a metaphor can also conceptualize
the concrete in terms of the concrete. Lakoff and Turner, to be fair, are
aware of this. They discuss at some length the Elizabethan notion of the
Great Chain of Being (see e.g., Tillyard 1976), which endorsed the idea
of natural hierarchies within various types of creatures angels, humans, birds, mammals, etc. and state that the great chain metaphor
can apply to a target domain at the same level on the Great Chain as the
source domain (Lakoff and Turner 1989: 179, emphasis in original). Put
differently, metaphors may have targets as well as sources that are directly
accessible to the senses. But given that CMT puts great emphasis on metaphors role in conceptualising the abstract in terms of the concrete, this
possibility receives rather scant attention, while concrete is concrete
metaphors are particularly relevant once we leave the realm of the purely
verbal. In the case of monomodal metaphors of the pictorial variety, both
388
Charles Forceville
389
ing evoked by Steppenwolf s Born to be wild, audible on the commercials soundtrack, are at least as important in the mapping (Forceville
2004b, forthcoming b; Forceville et al. in preparation). The relevant
similarity that is created between target and source pertains to these
connotations more than to anything else. The examples bear out Bartsch
observation that in metaphor the role of similarity is not restricted to the
identity of internal properties of objects and situations, rather similarity
also is due to identity of external contiguity relationships between objects,
between situations, and it is due to relationships of objects and situations
with emotional attitudes, desires, and behavioural dispositions of people
(Bartsch 2003: 52). Indeed, it might be ventured that a single, embodied
correspondence between target and source is enough to trigger a wide
range of further cultural correspondences between target and source,
and hence of inferences about the target (Forceville et al., forthcoming).
The old adage that a picture tells more than a thousand words should not
blind us to the fact that pictures and other multimodal representations
seldom communicate automatically or self-evidently. As in verbal metaphors, it is connotations rather than denotations of source domains that
get mapped in metaphors, and these may substantially differ from one
(sub)cultural group to another (see e.g., Maalej 2001). Even when nonverbal metaphors verge toward the conventional, as in comics representations of anger (Eerden 2004; Forceville 2005c; see Simons 1995 for
multimodal instantiations of structural metaphor in pre-election TV
spots promoting political parties), it may well be the case that, as in their
verbal counterparts (see Kvecses 1986, 2000, 2005), there is cultural
variation (cf. Shinohara and Matsunaka 2003).
The three issues briefly mentioned above (metaphors frequently have
concrete rather than abstract targets; many metaphors with idiosyncratic surface manifestations do not reflect correlation metaphors;
what gets mapped from source to target domains are often cultural, not
embodied, features) are thrown in relief when the study of purely verbal
metaphors gives way to that of non-verbal and multimodal metaphors.
That is, generalizing observations on metaphor based on the systematic
investigation of verbal specimens need to be considered afresh by testing
these observations in metaphors occurring in other modes.
390
4.
Charles Forceville
We have seen that within the CMT paradigm, most surface metaphors
should be amenable to a pre-existing conceptual A is B format. Inevitably,
in order to discuss the metaphor, this A and B must be named, i.e., rendered in language. It is by no means a foregone conclusion, of course, that
the language of thought is actually a verbal language. The convention
to verbalize the image-schematic structures underlying surface metaphors
by using small capitals useful inasmuch as this facilitates analysing
them may disguise a number of consequences that seem to me more
problematic in the discussion of representations that are not (exclusively)
verbal ones than of purely verbal ones. One of these consequences is that
it is the analysts responsibility to find an adequate or acceptable verbal
rendering of the metaphors underlying image-schematic level, but such a
verbalization, even though used as a convenient shorthand, is never neutral. The design of the Senseo coffee machine suggests the posture of
somebody bending over and modestly offering something (i.e., a cup) on
a plate. But should this awareness result in the verbalization coffee
machine is servant, or is coffee machine is butler more appropriate?
Although servant and butler share many features, they also differ:
butler is more specific, and may in some people (but not in others)
evoke connotations of Britishness and standards of service that servant
does not. As a result, the mappings suggested by the two verbalizations
may differ. Bartsch might conclude that this very inability to agree on a
single verbalization of the source domain that is shared within a community shows that the source has no conceptual status, and reflects a quasiconcept at best (Bartsch 2003: 50). However, to the extent that there is a
community that recognizes the source as cueing a serving person, the
source admits predicates understood as true in the community (such as
is there to serve the user, obeys your requests, and is almost always
available). In a visually literate society, a vast number of endlessly repeated and recycled images (such as famous paintings, photographs, film
shots, flags, logos, animation characters) evoke specific phenomena and
events in a clichd, shorthand manner widely shared within a community,
and hence arguably aspire to conceptual status. But this speculation leads
us far beyond the concerns of the present chapter and deserves in-depth
reflection elsewhere.
Another consequence is that verbalization of a non-verbal metaphor is
necessarily a conscious action, and a fairly unusual one at that. It is only
391
the scholar writing an academic paper who, to be able to discuss a multimodal metaphor, needs to resort to Gazelle bicycle is Van Grunsven
dressage horse; the metaphor is not verbalized in this form in the commercial, and it is an open question whether the construal of a non-verbal
metaphor requires its verbalization by the audience. The question can be
reformulated as follows: does comprehension of a non-verbal or multimodal metaphor imply that recipients mentally verbalize the metaphor? It
is an important question, but difficult to test empirically.
Since non-verbal modes of communicating by definition do not have
the is or is like in order to signal a metaphorical identity relation between two entities belonging, in the given context, to different categories,
one issue that deserves attention is by what stylistic means the similarity is
triggered. Of course this holds for verbal metaphors that do not have the
paradigmatic A is (like) B format as well (cf. Brooke-Rose 1958; Goatly
1997). But whatever means are chosen in this latter case, the cues are
themselves of a verbal nature. In non-verbal and multimodal metaphors,
the signals that cue metaphorical similarity between two phenomena are
different, and bound to differ depending on the mode(s) in which the
metaphorical terms are represented. Here are some possibilities that are
deployed in isolation or in combination:
Perceptual resemblance. This can only function as a trigger in the case of
monomodal metaphors: only a visual representation can perceptually resemble another visual representation; only a sound can perceptually resemble another sound. In the case of visual resemblance, there is a range
of choices: two things can resemble one another because they have the
same size, colour, position, posture, texture, materiality, etc. Note that the
resemblance need not reside in the things themselves, but may surface
in their manner of representation: they may for instance be photographed
from the same unusual angle, or filmed with the same unusual camera
movement.
Filling a schematic slot unexpectedly. Placing a thing in a certain context
may strongly, even inescapably, evoke a different kind of thing, namely the
thing for which the given context is the natural or conventional place. Put
differently, we encounter deviations from typical gestalts or schemas. For
example, when in a musical environment a violin case contains a monkey
wrench, this suggests the metaphor monkey wrench is violin (or v.v.).
Simultaneous cueing. If two things are signalled in different modes,
metaphorical identification is achieved by saliently representing target
and source at the same time. For instance a kiss could be accompanied by
the sound of a car crash, of a vacuum cleaner, or of the clunking of chains,
392
Charles Forceville
5.
393
The genre within which a text (in whatever medium) is presented, or the
genre to which it is attributed, determines and constrains its possible interpretations to an extent that is difficult to overestimate (see Altman
1999; Charteris-Black 2004; Forceville 1999b, 2005d; Steen 1994; Zwaan
1993 for discussion). For this reason, it is important to study how genre
has an impact on the production and interpretation of metaphors (monomodal and multimodal alike). In advertising, for instance, the targets of
metaphors often coincide with the product promoted (Forceville 1996).
This is to be expected: an advertisement or commercial predicates something about a product, brand, or service, and this neatly and naturally fits
the metaphors target is source format. Moreover, the features mapped
from source to target are positive ones (unless the metaphor is used to disqualify a competitors brand, in which case the mapped features are typically negative). But in feature films, there is no phenomenon that in a
similar, natural, way qualifies as a metaphorical target. Metaphors in
artistic narratives pertain to phenomena that, for whatever reason, are
deemed salient by their producers. These phenomena can be protagonists,
but also objects. The mapped features will often be less clear-cut, and may
have a richer aligned structure (Gentner and Loewenstein 2002), than
those in advertising.
Metaphors in artistic representations may also differ in other respects
from those in commercial messages. For instance, while in commercials
there will seldom be a question what is target and what is source in a metaphor, an artistic narrative may give rise to two different construals of a
metaphor: both a is b and b is a is appropriate. (While Carroll 1994, 1996
calls such metaphors reversible, I prefer to say that, in the given context,
both the metaphors a is b and b is a are pertinent, in order to retain the
notion that target and source in a metaphor are, in principle, irreversible.)
Commensurate with this, metaphors in artistic contexts presumably allow
for greater freedom of interpretation than do metaphors in commercials
(cf. also Shen 1995).
Another parameter that deserves further research is whether any of the
subtypes of pictorial metaphor or of the manifold varieties of multimodal
metaphor can be systematically related to certain text genres. For instance, it seems that commercial advertising seldom makes use of the hybrid variety of pictorial metaphors (in Forceville 1996 these were called
MP2s). Again, this makes sense: if metaphorical targets typically coincide
with products, advertisers would want their product portrayed in their
entirety, and not in a manner that might evoke connotations of incompleteness or mutilation. Hybridizing it with a metaphorical source do-
394
Charles Forceville
main would not fit this goal. By contrast, in animation films, or science
fiction films, no such problem arises.
Finally, it would be worthwhile to investigate whether in the case of
multimodal metaphors there are any systematic correlations between textual genres and the modes in which a target and source are represented. In
advertisements, the visual mode is typically used for representing the target and this may well be true for different genres as well. But perhaps alternative patterns in the choice of mode for the source domain are detectable in different types of texts, while this may also change over time within
a genre.
6.
Concluding remarks
395
analysis of the tools of persuasive discourse in the broadest sense constitutes an excellent interface between research in academia and its possible
usefulness in the world beyond its walls.
On virtually all of the issues discussed in this article, Forceville and
Urios-Aparisi (in preparation) will provide new insights.
Acknowledgments
I am indebted to Francisco Ruiz de Mendoza and Ren Dirven for comments and suggestions on earlier drafts of this article.
Notes
1. The topic of metaphor and music also has inspired studies over the past few
years, but for lack of expertise in this area I will not dwell on these. For references, see Johnson and Larson 2003; see also Cook 1998, Zbikowski 2002,
Thorau 2003, and Spitzer 2004 the last one rather difficult for laymen.
2. Lakoff and Turner describe generic-level metaphors as metaphors which are
minimally specific in two senses: they do not have fixed source and target domains, and they do not have fixed lists of entities specified in the mapping
(1989: 81). They introduce the term using the example of the events are actions metaphor which they contrast with life is a journey, one of its specific-level instantiations.
References
Altman, Rick
1999
Film/Genre. London: British Film Institute.
Bartsch, Renate
2003
Generating polysemy: Metaphor and metonymy. In Metaphor and Metonymy in Comparison and Contrast, Ren Dirven, and Ralf Prings
(eds.), 4974. Berlin/New York: Mouton de Gruyter.
Black, Max
1979
More about metaphor. In Metaphor and Thought, Andrew Ortony
(ed.), 1943. Cambridge: Cambridge University Press.
Bolter, Jay David, and Richard Grusin
1999
Remediation: Understanding New Media. Cambridge MA: MIT Press.
Brooke-Rose, Christine
1958
A Grammar of Metaphor. London: Secker and Warburg.
396
Charles Forceville
Carroll, Noel
1994
Visual metaphor. In Aspects of Metaphor, Jaakko Hintikka (ed.),
189218. Dordrecht: Kluwer.
1996
A note on film metaphor. In Theorizing the Moving Image, 212223.
Cambridge: Cambridge University Press.
Charteris-Black, Jonathan
2004
Corpus Approaches to Critical Metaphor Analysis. London: Palgrave/
MacMillan.
Cienki, Alan
1998
Metaphoric gestures and some of their relations to verbal metaphoric
expressions. In Discourse and Cognition: Bridging the Gap, Jean Pierre
Koenig (ed.), 189204. Stanford, CA: Center for the Study of Language
and Information.
Cook, Nicholas
1998
Analyzing Musical Multimedia. Oxford: Clarendon.
Coulson, Seana, and Esther Pascual
In press For the sake of argument: Mourning the unborn and reviving the
dead through conceptual blending. Annual Review of Cognitive Linguistics 4.
Crisp, Peter
2003
Conceptual metaphor and its expression. In Cognitive Poetics in Practice, Joanna Gavins, and Gerard Steen (eds.), 99113. London/New
York: Routledge.
Cupchik, Gerald D.
2003
The interanimation of worlds: Creative metaphors in art and design.
The Design Journal 6 (2): 1428.
Dirven, Ren, and Ralf Prings (eds.)
2002
Metaphor and Metonymy in Comparison and Contrast. Berlin/New
York: Mouton de Gruyter.
Durand, Jacques
1987
Rhetorical figures in the advertising image. In Marketing and Semiotics: New Directions in the Study of Signs for Sale, Jean Umiker-Sebeok (ed.), 295318. Berlin/New York: Mouton de Gruyter.
Eerden, Bart
2004
Liefde en Woede: De Metaforische Verbeelding van Emoties in Asterix.
[Love and anger: the metaphorical visualization of emotions in Asterix.] MA diss., University of Amsterdam, The Netherlands.
El Refaie, Elisabeth
2003
Understanding visual metaphors: The example of newspaper cartoons.
Visual Communication 2 (1): 7595.
Fauconnier, Gilles, and Mark Turner
2002
The Way We Think: Conceptual Blending and the Minds Hidden Complexities. New York: Basic Books.
Forceville, Charles
1988
The case for pictorial metaphor: Ren Magritte and other Surrealists.
397
Marksisticne Studije
[Slovenia].
1994
Pictorial metaphor in advertisements. Metaphor and Symbolic Activity
9 (1): 129.
1995
(A)symmetry in metaphor: The importance of extended context.
Poetics Today 16: 677708.
1996
Pictorial Metaphor in Advertising. London/New York: Routledge.
1999a The metaphor Colin is a child in Ian McEwans, Harold Pinters, and
Paul Schraders The Comfort of Strangers. Metaphor and Symbol 14:
17998.
1999b Art or ad? The effect of genre-attribution on the interpretation of images. Siegener Periodicum zur Internationalen Empirische Literaturwissenschaft [SPIEL] 18 (2): 279300.
2000
Compasses, beauty queens and other PCs: Pictorial metaphors in computer advertisements. Hermes, Journal of Linguistics 24: 3155.
2002a The identification of target and source in pictorial metaphors. Journal
of Pragmatics 34: 114.
2002b Further thoughts on delimitating pictorial metaphor. Theoria et Historia Scientiarum 6: 21327 [Torn, Poland].
2003
Bildliche und multimodale Metaphern in Werbespots [Trans. from
English by Dagmar Schmauks] Zeitschrift fr Semiotik 25: 3960.
2004a Review of Fauconnier and Turner (2002). Metaphor and Symbol 19:
8389.
2004b The role of non-verbal sound and music in multimodal metaphor. In
Words in their Places: A Festschrift for J. Lachlan Mackenzie, Henk
Aertsen, Mike Hannay, and Rod Lyall (eds.), 6578. Amsterdam: Faculty of Arts, VU Amsterdam.
2005a When is something a pictorial metaphor? [Lecture 2 in eight-lecture
Course on Pictorial and Multimodal metaphor] http://www.chass.utoronto.ca/epc/srb/cyber/cforcevilleout.html
2005b Cognitive linguistics and multimodal metaphor. In Bildwissenschaft:
Zwischen Reflektion und Anwendung, Klaus Sachs-Hombach (ed.),
264284. Cologne: Von Halem.
2005c Visual representations of the Idealized Cognitive Model of anger in the
Asterix album La Zizanie. Journal of Pragmatics 37: 6988.
2005d Addressing an audience: time, place, and genre in Peter Van Straatens
calendar cartoons. Humor: International Journal of Humor Research 18
(3): 247278.
Forthc. a Multimodal metaphor in ten Dutch TV commercials. Public Journal of
Semiotics. http://semiotics.ca/
Forthc. b Metaphors in pictures and multimodal representations. In The Cambridge Handbook of Metaphor and Thought, Ray Gibbs (ed.), Cambridge: Cambridge University Press.
Forceville, Charles, Paul Hekkert, and Ed Tan
Forthc. The adaptive value of metaphors. In Heuristiken der Literaturwissen-
398
Charles Forceville
399
Indurkhya, Bipin
1991
Modes of metaphor. Metaphor and Symbolic Activity 6: 127.
1992
Metaphor and Cognition: An Interactionist Approach. Dordrecht:
Kluwer.
Johnson, Mark
1987
The Body in the Mind: The Bodily Basis of Meaning, Imagination and
Reason. Chicago: University of Chicago Press.
1993
Moral Imagination: Implications of Cognitive Science for Ethics. Chicago: University of Chicago Press.
Johnson, Mark, and Steve Larson
2003
Something in the way she moves metaphors of musical motion.
Metaphor and Symbol 18: 6384.
Kaplan, Stuart J.
1990
Visual metaphors in the representation of communication technology.
Critical Studies in Mass Communication 7: 3747.
1992
A conceptual analysis of form and content in visual metaphors. Communication 13: 197209.
Kennedy, John M.
1982
Metaphor in pictures. Perception 11: 589605.
1993
Drawing & the Blind: Pictures to Touch. New Haven/London: Yale University Press.
Knop, Sabine de, Ren Dirven, Ning Yu, and Birgit Smieja (eds.)
2005
Bibliography of Metaphor and Metonymy. Amsterdam/Philadelphia:
Benjamins. Available electronically only at: http://www.benjamins.
com/online/met
Kvecses, Zoltn
1986
Metaphors of Anger, Pride and Love. Amsterdam/Philadelphia: Benjamins.
2000
Metaphor and Emotion: Language, Culture, and Body in Human Feeling.
Cambridge: Cambridge University Press.
2002
Metaphor: A Practical Introduction. Oxford: Oxford University Press.
2005
Metaphor in Culture: Universality and Variation. Cambridge: Cambridge University Press.
Kress, Gunther, and Theo van Leeuwen
1996
Reading Images: The Grammar of Visual Design. London/New York:
Routledge.
2001
Multimodal Discourse: The Modes and Media of Contemporary Communication. London: Edward Arnold.
Lakoff, George
1987
Women, Fire and Dangerous Things: What Categories Reveal about the
Mind. Chicago: University of Chicago Press.
1993
The contemporary theory of metaphor. In Metaphor and Thought (2nd
edition), Andrew Ortony (ed.), 202251. Cambridge: Cambridge University Press.
Lakoff, George, and Mark Johnson
1980
Metaphors We Live By. Chicago: University of Chicago Press.
400
1999
Charles Forceville
Philosophy in the Flesh: The Embodied Mind and its Challenge to Western Thought. New York: Basic Books.
2003
Afterword, 2003. In Metaphors We Live By, 243276. Chicago: University of Chicago Press.
Lakoff, George, and Mark Turner
1989
More Than Cool Reason: A Field Guide to Poetic Metaphor. Chicago:
University of Chicago Press.
Maalej, Zouhair
2001
Processing pictorial metaphor in advertising: A cross-cultural perspective. Academic Research 1: 1942 [Sfax, Tunisia].
McLuhan, Marshall
1964
The medium is the message. In Understanding Media: The Extensions
of Man (2nd edition), 2335. New York: McGraw-Hill.
McNeill, David
1992
Hand and Mind: What Gestures Reveal about Thought. Chicago: University of Chicago Press.
2005
Gesture and Thought. Chicago: University of Chicago Press.
McQuarrie, Edward F., and David Glen Mick
2003
The contribution of semiotic and rhetorical perspectives to the explanation of visual persuasion in advertising. In Persuasive Imagery: A
Consumer Response Perspective, Linda M. Scott, and Rajeev Batra
(eds.), 191221. Mahwah, NJ/London: Lawrence Erlbaum.
Mller, Cornelia
2004
Metaphors, Dead and Alive, Sleeping and Waking: A Cognitive Approach to Metaphors in Language Use. Habilitationsschrift, Freie Universitt Berlin, Germany.
Ortony, Andrew (ed.)
1979
Metaphor and Thought. Cambridge: Cambridge University Press.
Phillips, Barbara J.
2003
Understanding visual metaphor. In Persuasive Imagery: A Consumer
Response Perspective, Linda M. Scott, and Rajeev Batra (eds.),
297310. Mahwah, NJ: Lawrence Erlbaum.
Richards, Ivor Armstrong
1965 [1936] The Philosophy of Rhetoric. New York: Oxford University Press.
Rohdin, Mats
2003
Summary in English. In Vildsvinet I Filmens Trdgrd: Metaforbegreppet inom Filmteorin [The wild boar in the garden of film: The concept of
metaphor in film theory], 318329. PhD diss., Stockholm: Edita Norstedts Tryckeri AB.
Rompay, Thomas van
2005
Expressions: Embodiment in the Experience of Design. PhD diss., Technische Universiteit Delft, The Netherlands.
Rozik, Eli
1994
Pictorial metaphor. Kodikas/Code 17: 203218.
1998
Ellipsis and the surface structures of verbal and nonverbal metaphor.
Semiotica 119: 77103.
401
Shen, Yeshayahu
1995
Cognitive constraints on directionality in the semantic structure of
poetic vs. non-poetic metaphors. Poetics 23: 255274.
Shibles, Warren A.
1971
Metaphor: An Annotated Bibliography and History. Whitewater, WI:
The Language Press.
Shinohara, Kazuko, and Yoshihiro Matsunaka
2003
An analysis of Japanese emotion metaphors. Kotoba to Ningen: Journal
of Yokohama Linguistic Circle 4: 118.
Shore, Bradd
1996
Culture in Mind: Cognition, Culture, and the Problem of Meaning. New
York: Oxford University Press.
Simons, Jan
1995
Film, Language, and Conceptual Structures: Thinking Film in the Age of
Cognitivism. Ph.D. diss., University of Amsterdam, The Netherlands.
Smith, Ken
1996
Laughing at the way we see: The role of visual organizing principles in
cartoon humor. Humor: International Journal of Humor Research 9:
1938.
Sperber, Dan, and Deirdre Wilson
1995
Relevance: Communication and Cognition (2nd edition) Oxford: Basil
Blackwell.
Spitzer, Michael
2004
Metaphor and Musical Thought. Chicago/London: University of Chicago Press.
Steen, Gerard J.
1994
Understanding Metaphor in Literature: An Empirical Approach. London: Longman.
Stockwell, Peter J.
1999
The inflexibility of invariance. Language & Literature (8) 2: 12542.
Sweetser, Eve E.
1990
From Etymology to Pragmatics: Metaphorical and Cultural Aspects of
Semantic Structure. Cambridge: Cambridge University Press.
Teng, Norman Y., and Sewen Sun
2002
Grouping, simile, and oxymoron in pictures: A design-based cognitive
approach. Metaphor and Symbol 17: 295316.
Terkourafi, Marina, and Stefanos Petrakis
Under rev. A critical look at the desktop metaphor 25 years on.
Thorau, Christian
2003
Metapher und Variation: Referenztheoretische Grundlagen musikalischer Metaphorizitt. Zeitschrift fr Semiotik 25: 109124.
Tillyard, Eustace M.W.
1976 [1943] The Elizabethan World Picture. Harmondsworth: Penguin.
Turner, Mark
1996
The Literary Mind. New York/Oxford: Oxford University Press.
402
Charles Forceville
Tversky, Barbara
2001
Spatial schemas in depictions. In Spatial Schemas and Abstract
Thought, Merideth Gattis (ed.), 79112. Cambridge, MA: MIT Press/
Bradford book.
Van Noppen, Jean-Pierre, Sabine de Knop, and Ren Jongen (eds.)
1985
Metaphor: A Bibliography of Post-1970 Publications. Amsterdam/Philadelphia: Benjamins.
Van Noppen, Jean-Pierre, and Edith Hols (eds.)
1990
Metaphor II: A Classified Bibliography of Publications From 1985
1990. Amsterdam/Philadelphia: Benjamins.
Ventola, Eija, Cassily Charles, and Martin Kaltenbacher
2004
Perspectives on Multimodality. Amsterdam/Philadelphia: Benjamins.
Whittock, Trevor
1990
Metaphor and Film. Cambridge: Cambridge University Press.
Wiggin, Amy A., and Christine M. Miller
2003
Uncle Sam wants you! Exploring verbal-visual juxtapositions in
television advertising. In Persuasive Imagery: A Consumer Response
Perspective, Linda M. Scott, and Rajeev Batra (eds.), 267295.
Mahwah, NJ: Lawrence Erlbaum.
Yu, Ning
1998
The Contemporary Theory of Metaphor: A Perspective from Chinese.
Amsterdam/Philadelphia: Benjamins.
Zbikowski, Lawrence M.
2002
Conceptualizing Music: Cognitive Structure, Theory, and Analysis. Oxford: Oxford University Press.
Zwaan, Rolf
1993
Aspects of Literary Comprehension: A Cognitive Approach. Amsterdam/Philadelphia: Benjamins.
Cognitive poetics
403
Abstract
This paper explores how cognitive poetics may serve as a bridge between literary
studies and linguistics. Because cognitive poetics studies the cognitive processes
that constrain literary response and poetic structure, it provides a theoretical cognitive basis for literary intuition. At the same time, by exploring the iconic functions that create literature as the semblance of felt life, cognitive poetics contributes to our understanding of the embodied mind. The effect of what I call
poetic iconicity is to create sensations, feelings, and images in language that enable the mind to encounter them as phenomenally real. The paper draws upon
Susanne K. Langers (1953, 1967) theory of art, Charles Sanders Peirces
(1955[1940]) theory of the sign, Ellen Spolskys (1993) theory of literature bridging the gap caused by the minds modularity, Elaine Scarrys (1999) theory of
images in the mind, and the cognitive-semiotic notions of blending, deixis,
negative polarity, and schema theory to show how Robert Frost manipulates the
fictive and factive planes in his poem, Mending Wall, to create a poetic iconicity
of feeling that leads literary critics to their various interpretations of the poem. It
concludes by arguing that both literary studies and cognitive linguistics are complementary ways of showing how a literary text extends natural language use in
order to bridge the gap between mind and world.
Keywords: Cognitive poetics; iconicity; blending; schema; deixis; negative polarity; Robert Frost
1.
Over thirty years ago, George Steiner (1970) explored the history of the
language revolution in literary studies from its development at the turn
of the twentieth century. In his chapter on linguistics and poetics, he comments on the arrogant absurdity of those who deem themselves qualified in the art of literary study and yet who are totally ignorant of what
linguistics contributes to the knowledge of language (p.150). One could
just as well say that linguistics as practiced in the twentieth century was
just as arrogantly oblivious of the literary text (but see Jakobson1987 for a
404
Margaret H. Freeman
notable exception). When the editors of this volume invited me to contribute this chapter on the fall of the wall between literary studies and
linguistics as they put it, I had no doubt that they had in mind, in Steiners phrase, the mutually suspicious attitudes of both literary critics
and linguists to each others work that still exist in academic departments
today (Henkel 1996). Whether the cognitive revolution that has replaced the language revolution at the turn of the twenty-first century
can succeed where its predecessor has so obviously failed remains to be
seen. In this short essay, I will attempt to outline what I see in cognitive
poetics that might suggest a means for better understanding between linguistics and literary studies.
Twentieth century linguists, especially Noam Chomsky and his followers, scorned what they called mere performance in their aim to produce a theoretical account of linguistic competence. Cognitive linguists,
to the contrary, as the other chapters in this volume show, adopt an encyclopedic view of language and attempt a scientific account of natural
language in use. As a result, they have shown increasing awareness of how
literary texts might provide a productive source of data for investigation.
They have not yet, however, generally recognized the contribution that the
accumulated knowledge of literary studies might make to their own research agenda. Likewise, literary critics have been dismissive of linguistic
approaches to literature, cognitive or otherwise, either because such approaches fail to account for what the critics consider significant or because the scientific apparatus simply recasts in technical jargon what
they have said, and said more clearly.
More intense collaboration between linguistics and literary studies
could be mutually beneficial. Literary studies has developed over centuries an expertise in reading and analyzing texts which, in the process,
has led to the recognition and better understanding of such phenomena
as point of view, perspective, narrative, tone, foregrounding, implied
author, and so on. It has also developed expertise in relating imaginary
worlds to the real contextual worlds of author and reader. Cognitive linguistics, for its part, has developed theories of such phenomena as deixis,
figure-ground, fictive motion, mental space mappings, and so on, that
parallel literary discoveries. It is no surprise, therefore, that many literary
critics find cognitive linguistic applications to literary texts merely jargonizing what they already know, and cognitive linguists find literary
readings ad hoc and impressionistic. The paralleling of terminology between the two fields is, however, deceptive. Whereas literary critics focus
on illuminating the language of the text, cognitive linguists focus on il-
Cognitive poetics
405
2.
Cognitive poetics
In linking the processes of language in literary text construction and interpretation to the processes of language in the workings of the human
mind, cognitive poetics provides a bridge between the two fields.
2.1 Brief history
Cognitive Poetics developed over the past twenty years or so from several
different strands. Reuven Tsur (1983) was the first to use the term to describe his theoretical and methodological approach to poetry, drawing
from studies in psychology, neuro-anatomy, and literary criticism. Another strand developed almost a decade later in Tabakowskas (1993)
work in applying Langackers (1987, 1991) studies in cognitive grammar
to poetic translation. Meanwhile, Lakoff and Johnsons (1980) work in
conceptual metaphor theory led to Lakoff and Turners (1989) More than
Cool Reason: A Field Guide to Poetic Metaphor, from which another
strand more closely linked to metaphor theory developed. This strand
broadened into further studies as a result of Fauconnier and Turners
(1994, 2002) work in conceptual integration theory, or blending, as it is
more commonly known. Yet another strand emerged from a more general
interest in the relation of cognition as reflected in the multidisciplinary
approaches of cognitive science to literary studies (Crane and Richardson
1999). Meanwhile, work in cognitive narratology (Emmott 1997; Fludernik 1993), text-world theory (Werth 1999; Gavins 2005), and cognitive
stylistics (Semino and Culpeper 2002) expanded the role of cognitive
poetics to include other theoretical perspectives and all literary texts.
2.2 What is cognitive poetics?
As a result of these different strands, cognitive poetics embraces a broad
array of theoretical and methodological approaches. Tsur (1992) defines
cognitive poetics as an exploration of how cognitive processes shape and
constrain literary response and poetic structure. As Elena Semino (personal correspondence) notes:
406
Margaret H. Freeman
Cognitive poetics
407
studied, I hope that such a cognitive poetics application will realize the
spirit of this series in applied linguistics and give some sense of the work in
cognitive poetics to date.1
3.
Mending Wall might seem an odd choice for a case study on a chapter
devoted to the fall of a wall. However, as I hope to show in the ensuing
analysis, Frosts poem reveals, as he said himself, that man is both a wall
builder and a wall toppler. He makes boundaries and he breaks boundaries. Thats man (quoted in Cook 1974: 8283). This human characteristic reflects man as part of nature. Frost notes elsewhere: Nature within
her inmost self divides / To trouble men with having to take sides.2 Although Frosts poem ostensibly appears to be about two men taking opposing sides, a close analysis reveals a somewhat different purpose, as
several critics have noticed: the value of opening up the imagination. In
this way, the gaps in the physical wall become identified with gaps in nature, gaps in the human mind (Spolsky 1993). In seeing the emergent
meaning of the poem as a blend between the physical story of the two
neighbors as they go about mending their boundary wall and the imaginative story of the work of the human mind, one can also see the poem
iconically invoking images of boundary walls from the readers own experience which is then mapped onto the narrative, thus making the language of the poem become phenomenally real, to create, in Langers
words, the semblance of felt life. How this works is what I hope to show
in the rest of this essay.
3.1 Poetry as iconicity
Poetic iconicity is one way of explaining why poetry cannot be successfully paraphrased. As Frost would invariably remark in response to a
question as to what he meant by a poem, What do you want me to do, say
it again in different and less good words? (quoted in Raab 1996). It is
not simply meaning, it is the poem itself. According to Charles Sanders
Peirce (1955[1940]), every word is a symbol that may or may not have an
icon or index function. A poem that is true to the experience of felt life is
mimetically representing that experience by evoking a mental image of it.
Mimesis, literary or so-called exophoric iconicity, is the principle of
form miming meaning, language miming the world. But poems are also
408
Margaret H. Freeman
Cognitive poetics
409
410
Margaret H. Freeman
cows and other animals. It is, therefore, more in the interest of the poet
than the farmer to maintain this particular wall.
The situation of the semiotic base space describes the activity of the two
neighbors who set out in the spring to repair a boundary wall between
their properties. The poem is narrated by one of the two neighbors, so that
we are presented with a discourse situation in which the narrator is also a
character in his own story. The farmer is meeting his obligation to maintain his side of the wall out of his belief that good fences make good
neighbors. Here is a man who practices what Ferdinand Tnnies (1957
[1887]) has called Gemeinschaft and acts upon his beliefs.8 And those beliefs are even further emphasized and reinforced by the fact that there is
no longer any real need for the wall that divides apple and pine trees. Had
there been a real need for the wall, the significance of the farmers gesture
would have been somewhat lessened. That significance, as George Monteiro (1988:126129) has shown us, is deeply embedded in two historical,
cultural domains: the existence of a Spanish proverbial saying (una pared
entre dos vezinos guarda mas (haze durar) la amistad) which goes back at
least to the Middle Ages and was recorded by Emerson in his journal of
1832 as A wall between both, best preserves friendship; and the myth of
the god of boundaries, named Terminus by the Romans, who celebrated a
Terminalia festival each year on February 23 (in early spring) when
neighbors on either side of any boundary gathered around the landmark to offer sacrifice to the god and celebrate with a feast. The poem,
Monteiro believes, is moving toward this mythological link in the speakers description of his neighbor at the end of the poem as an old-stone
savage armed who moves in darkness. The occurrence of magical elements such as spell and the notion of Elves in the poem conjures up
the mysteries of pre-scientific knowledge that is reflected in the myth as
well as in the unconscious elements of the human mind.
3.3 Metaphor and mimesis
One curious thing about Frosts poem is that its central metaphor is expressed so simply and its meaning is so transparent that we tend not to see
it as a metaphor at all. This occurs in line 24: He is all pine and I am
apple orchard. By identifying the two characters in the poem with their
land, Frost is doing two things: he is establishing the two sides of the actual boundary wall that is the subject of the poem and at the same time he
is making that wall a mental representation. So one can see the presentation space as being the actual, factive story of two men repairing a bound-
Cognitive poetics
411
ary wall in springtime, and the reference space as being the virtual, fictive
story of the operations of the mind (Fig. 1). The balanced shifts between
the factive (indexical) to the fictive (virtual) create the image in the mind,
so that the fictive lies within the factive, the factive within the fictive. In
412
Margaret H. Freeman
the blend, the poet represents the creative, open-ended operations of the
mind; the farmer, the minds need for coherence and continuity. The
meaning that emerges from this blend is the value that is being attached to
the poems topic, that is, the gaps in the wall/mind.
The bringing together of the gaps in the physical world of Frosts poem
and the mental world of his imagination occurs through a particular characteristic of mental representations. Frank Lentricchias (1975) study of
the landscapes of the self in Mending Wall reflects Elaine Scarrys
(1999) argument that references to the rarefied and intangible facilitate
the creation of images in the mind. He argues that not naming what it is
that causes walls to fall enables Frost to manipulate intransigent fact
into the world of the mind where all things are pliable (p.104). In making
fiction about how the wall developed gaps, the poet is inviting the
farmer to accompany him, to take a journey in the mind.9 Thus, the
poem becomes a paeon to the play spirit of imagination. For the cognitive linguist, the question becomes how Frost makes language work to
make us think the poem is about minds as well as about walls.
As Scarry (1999) has shown, poetry sets up a mimesis of making and a
mimesis of motion. Right from the outset, Frost introduces the idea of a
(solid) wall by referring to something immaterial, rarefied in Scarrys
terms. This something is unseen, unheard. It makes gaps which are
themselves the absenting of materiality. This, Scarry suggests, is what enables us to take an image more easily into the mind.10 The image itself
the wall is then immediately made to move through the technique of introducing light, with some force spilling its boulders into the sun. The
sudden addition into the scene of rabbits and dogs as natural living organisms helps to reinforce the mental image of motion as the hunters subtract
the stones from the fictive wall we have been invited to imagine. Similarly,
Frosts introduction of the seemingly inappropriate images of the fallen
boulders as loaves associating them with plant life and relative insubstantiality compared with stones and so nearly balls associating
them with a shape and a size easy to handle and pick up enables Frost to
create the image of movement and precarious balance as the neighbors
wear [their] fingers rough with handling them as they set the wall.
Having created these mental images of the activity of wall mending, Frost
has prepared the way for his speaker to put a notion in his [neighbors]
head as he has put mental images into ours. The notion challenges the
need for a physical barrier. It asks the reasons why and, in wondering
what might cause walls to fall, invokes the assumption of invisible natural
forces and the resulting inevitability of change. It expresses the human
Cognitive poetics
413
414
Margaret H. Freeman
Cognitive poetics
415
On either side of this wall that we dont need but that divides the two
halves of the poem, each set of 22 lines falls into two parallel parts,
marked by syntactic openings and closures and lexical patterns and repetitions. These are the building blocks on which the wall/poem is constructed. The first and last parts of the poem serve to frame the two inside
parts. The first part (lines 111) introduces the idea of the Something
there is that doesnt love a wall, and the last part (lines 3545) begins
with the same line and ends the poem with the idea that Good fences
make good neighbors. The interior parts deal with the interaction of the
two neighbors. Part 2 (lines 1222) describes the collaborative process
they adopt to mend the wall. Part 3 (lines 2434) contrasts the thoughts
of the poet with that of the farmer over the reasons for mending the wall.
If this were not enough, each of these parts divides into two sections of
five and six lines apiece, creating a chiasmic pattern or mirror image
across the dividing line. In the first half, the section divisions are 5 lines
followed by 6; in the second half, the section divisions are 6 lines followed
by 5. These wall/poem components, though, are not so perfectly constructed as they seem. The 19 sentences of the poem (marked by end
punctuation) sometimes cut across these parts and sections and occur irregularly both in position and number within them. Those familiar with
New England dry stone walls, created over two hundred years ago, will
know that they are not at all uniform in appearance. They are made up of
irregular-sized and irregular-shaped boulders that seem to balance precariously on one another, a method called random rubble in the drystone-wall making trade. It is these irregularities that enable the frost
heaving that occurs when ice forms and melts to spill the upper boulders
in the sun. The BALANCE schema of the poem is therefore not a static
one. It is a dynamic process of ongoing annual activity that the neighbors
must undertake to maintain their wall, just as the mind must be engaged
in continual activity to create coherency and consistency from multiple
and inchoate experience, to achieve cognitive stability, to make sense out
of things.
This iconic macrostructure creates a structural blend of wall and poem,
so that in the blend what is true of one is true of the other. Just as the wall
is embodied in the poem, the poem is embodied in the wall. What happens
as a result is that the gaps in the wall become gaps in the mind. And the
poem becomes a way of responding to and dealing with those gaps. The
techniques we use, linguistic or literary, to articulate and explore these
gaps are all salient and valid ways to understand them. The poet calls the
object of their mending a wall, the farmer calls it a fence. This is not ar-
416
Margaret H. Freeman
Cognitive poetics
417
up for the wall on the actual plane to be introduced in the second section (lines 1216).
The second section establishes the cognitive stability of the factive action of wall construction, reflected in the balance achieved by endstopped, regularly metered lines and the repetition of equivalent phrases
set the wall / keep the wall; between us / between us; to each / to each as the
neighbors walk the line. However, the negative polarities that have been
running through part 1 (doesnt love, gaps, not one, but, no one) surface in
the very next section (lines 1722) with the limited polarities of not-quite
expressions (so nearly, just another, little more) that make balance precarious and prepare for the indexical repetition there where it is in line 23,
introducing the idea that the wall is, after all, not needed, and yet is expressed in the line which serves as the boundary dividing the two halves of
the poem. Once this boundary wall has been presented, it again loses its
deictic specificity in all the references on the other side: the general
fences of the farmers comment and the hypothetical fictive musings of
the poet as he thinks of why he would build a wall, preparing for the
repetition of the first line, Something there is that doesnt love a wall at
the beginning of the last part (line 35).
Deictic and negative polarity terms contrasting conceptions in the
mind (virtual-fictive) and representations of objects in the world (actualfactive) are tossed through the poem like the loaves and balls of line 17
(Fig. 3). The existential there-where of lines 1 and 7 is repeated in the
deictic there-where of lines 11 and 19 that prepares for the deictic there
where of the pivot line 23. On the other side of the pivot, there occurs
at the end of the two word phrases in the existential where there and
deictic here there of line 31, creating another kind of chiasmus with the
phrases on the other side. These deictic alternations culminate in the final
existential-indexical contrast of the last part of the poem in the repetition
of Something there is (line 35) and I see him there (line 38), which
parallels the move from fictive to factive in the first part. This existentialindexical alternation balances the fictive and the factive. It embodies the
actions of the mind as phenomenally real.
Fictive-factive alternation also occurs across two different kinds of
negative polarity: the kind that establishes a subtraction from positive
existence, as in not one stone on a stone (line 7), and the kind that results in the opening up of possibilities beyond the physical, as in No one
has seen or heard (line 10). In the first kind of negation, an element
created in the fictive plane is cancelled out in the actual plane. In the second, only a portion of possible elements in the fictive part is cancelled,
418
Margaret H. Freeman
fictive
factive
closed
He . . . . . . . . I . . . . . . . . . . . . . .
My . . . . . . . never . . . . . . . . . . .
. . . . . . . . . . . .his. . . . . .I. . . . . .him.
He only . . . . . . . . . fences i . . . . . . . . . .
. . . . . . . . . . . . . me, . . . . . . . I . . . . . . . . . . .
. . . . . . I . . . . . . . his . . . . . . . . . .
....................................
Where there i. . . . . . . . . . .here there a . . . . . . . .
noc . . . . . . . .
. . . . .I. . . . . a wall i . . . . . . . . . . . . . . . . . . . . . .
......... I .............................
........................................
............. I .........................
. . . . . . . . .there . . . . . . . . doesntc love a wall i
. . . . . . . . . . . . I . . . . . . . . . . . . . him,
. . . . . . noto . . . . . . exactly, . . . . . . . Id . . . . . . . .
He . . . . . . . . . . . himself. I . . . . . . him there a
..........................................
..........................................
..........................................
He . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . me,
Not o . . . . . . . . . . onlyo . . . . . . . . . . . . . . . . . . . . . .
He . . . . . . . notc . . . . . . . . . . his . . . . . . . . . . . . .
. . . . . . . . he . . . . . . . . . . . . . . . . . . . . . . . . . . .
He . . . . . . . . . . . . fences i . . . . . . . . . . . . . . . . .
open
thus leaving open the possibility of the rest being actualized. These two
kinds of polarities are repeated in the second half of the poem with the
contrasts between the subtraction of no cows (line 31) and the opening
up of possibilities in not elves (line 37) and their chiasmic reversal in
Not of woods only (line 42) and He will not go behind (line 43). In
other words, what the poem is doing is oscillating between the fictive and
factive planes as the poet plays with ideas of balancing the imagined and
the real, the possible and the actual.
Cognitive poetics
419
420
Margaret H. Freeman
forms as the neighbor reiterates his belief. The I repetitions at the end of
part 3 question the use of a wall as a barrier, ending in the echo of fence
in offense (line 34).
At the graphemic level, too, the repetition of the letter l in wall, swell,
spills, hill, fallen, balls, spell, wall, all, tell, walling (in and out), will, well,
reflects the two sides of the barrier.13 The word spring occurs twice, one
on each side of the pivot line in the center of the poem, and is related to
mending and mischief respectively, another kind of balance between cognitive stability and opening up the mind to alternative possibilities. Finally, the opposition of sun (line 3) and darkness (line 41) on either
side points to the idea of openness and illumination created by gaps and
the idea of closed conservancy that nevertheless signifies in the poets
mind something beyond the physical pine woods and shade of trees.
4.
Robert Graves (1963) was the first critic to identify the possible pun Frost
is playing with in not naming the frost that causes walls to heave, which
serves to identify the poet-speaker with unseen forces that upset solidity
and disrupt order. However, Mark Richardson (1997: 141) is surely right
in enlarging the idea to include the vernal mischief of spring and its insubordinations.14 The creation through deixis of identification between
the factive and the fictive, between the physical gaps in the actual wall and
the mental gaps of the mind in its reception of multiple and overlapping
information, is paralleled by the identification of the volcanic nature of
the unseen forces (that spew out the boulders to cause gaps in the actual
wall) with the Spring that is the mischief in the mind of the poetspeaker that wants to create a similar unsettled disturbance in the mind of
his farmer-neighbor, to open up his mind to the recognition of mental
gaps in the articulation of knowledge and feeling. The poet is thus kicking
against the obligation of restraint in undermining stability of wall and
feeling. Spring mischief leads him to wonder if he could put a notion
in the farmers head. Literary critics (e.g., Lentricchia 1975, Kemp 1979)
are divided over whether the poet actually utters the words that follow
(lines 3036). However, these are the climactic words of the poem. They
bring the thoughts of the poet (fictive) into communication (factive) with
the farmer, and thus make the mischief of making gaps phenomenally
real. They point to the possible answer the farmer might give, the Elves
reflecting his cultural mythological heritage, and the answer he does in
Cognitive poetics
421
fact give, the reiteration of Good fences make good neighbors. In the
end, the only direct speech acts in the poem are four in number, all indicated by quotation marks: the joint admonition to the stones to Stay
where you are until our backs are turned (line 19) that both poet and
farmer participate in, the poets questions (lines 3036), and the repetition
of the proverb, Good fences make good neighbors (lines 28 and 46) uttered by the farmer. It is noteworthy that in all these cases, it is inanimate
objects (stones and fences) that are ascribed agency. Frosts restatement of
the Spanish proverb ascribes more causal agency in the act of making
than either the proverb or Emersons translation of it do. It is this making, this construction of the fictive in the factive that creates the poems
iconicity, and explains why Frost saw himself in both characters.15 It is
people that make good fences (factive), but it is fences that make good
people (fictive).
Cognitive poetics, I suggest, is in essence an exploration into poetic iconicity, as I am defining the term. It links the literary text to the cognitive
processes of the human mind, and provides a theoretical cognitive linguistic basis for literary intuition. Thats what makes it different from
purely linguistic or literary approaches. It does not replace these approaches; rather, it shows how they evidence ways in which a literary text
bridges the gap between mind and world. It is for this reason, I believe,
that cognitive poetics can contribute to both the scientific and the humanistic enterprise. It does not try to transform humanistic enquiry into a
science. Nor does it presume that scientific enquiry could replace humanistic enquiry as an adequate way to account for artistic creativity. As
it walks the line between the two, it defines the boundary that both separates and joins the two endeavors.
Appendix
MENDING WALL
1
2
3
4
5
6
7
Robert Frost
Something there is that doesnt love a wall,
That sends the frozen-ground-swell under it,
And spills the upper boulders in the sun,
And makes gaps even two can pass abreast.
The work of hunters is another thing:
I have come after them and made repair
Where they have left not one stone on a stone,
422
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Margaret H. Freeman
Cognitive poetics
423
Acknowledgments
I am grateful to Christina Ljungberg, Claiborne Rice, and the editors of
this volume for helpful suggestions to the ideas broached in this essay. I
trust that issues of theory and interpretation that emerge as a result will
contribute to further discussions in the field.
Notes
1. See Gavins and Steen (2003) and Stockwell (2002) for introductory texts in
cognitive poetics. There are many cognitive poetics articles appearing in various journals, such as Poetics Today, Style, Language and Literature, Journal of
English Linguistics, all of which have published special issues on cognitive approaches to literature. See also Hogan (2003). For a survey of recent work in
the field, see Freeman (in press). The following websites contain other bibliographic sources: cognitive approaches to literature at http://www.ucs.louisiana.edu/~cxr1086/coglit/; blending at http://markturner.org/; iconicity in language and literature at http://home.hum.uva.nl/iconicity/; textworld theory at
http://www.sheffield.ac.uk/textworldtheory/; art and cognition at http://www.
interdisciplines.org/artcog; literature, cognition, and the brain at http://cogweb.ucla.edu/.
2. Robert Frost. FROM IRON: Tools and Weapons. To Ahmed S. Bokhari. As
James Reston noted in The New York Times for October 27, 1957: The United
Nations, disturbed by Mr. Frosts opposition, suggested to him recently that he
might like to write a poem celebrating the ideal of the interdependence of the
nations. Sweden had given the U. N. a huge chunk of solid iron, and somebody
thought that this should be built into the U. N. building as a symbol of natures
strength and unity. Frost was not interested. Iron, he said, could be used to
strengthen the U. N. building, or it could be used for weapons of war. That was
the way with nature, he said: always confronting mankind with decisions. So he
rejected the invitation with a couplet http://www.frostfriends.org/FFL/
Periodicals/reston-nyt.html. Accessed March 21, 2006. I am grateful to Nick
Fleck for drawing my attention to this couplet.
3. Spolskys line is an adaptation of W. H. Audens line, mad Ireland hurt you
into poetry in his poem, In Memory of W. B. Yeats.
4. The speaker of the poem is not Frost but the poems persona. However, the
ability of the speaker in contrast to his neighbor to see the difference between
the gaps made by seen and unseen forces and recognize their significance
marks him as a poet. I shall therefore refer to the two characters as the poet
and the farmer.
5. Here, I am following Per Aage Brandts (2004) modification and elaboration
of Fauconnier and Turners (2002) blending model. Blending accounts for
how new information can emerge from old through the projection from given
424
6.
7.
8.
9.
10.
11.
12.
13.
Margaret H. Freeman
input spaces into a new space, called the blend. Brandt adds a semiotic base
space which includes the phenomenological world, or pheno-world for
short, the situation in which communication occurs, and the actual semiosis
of the discourse, including the participants and what is being communicated.
The input spaces of the blend are renamed as Presentation and Reference
spaces. Instead of the generic space in the original model, Brandt substitutes a
relevance space from which situational, argumentational, and illocutional
relevance triggers activity at various points in the development of meaning.
Scott Sylvester (New England forester), personal communication.
Report on Boundary Walls. Scottish Law Commission No. 163. Printed 25th
March, 1998. (584 Edinburgh: The Stationery Office) p. 2. New England followed common law practice in establishing the metes and bounds of property,
and the wording of the Scottish report (though not part of English common
law) reflects New England practices.
Tnnies (1957: 223) distinguished between Gemeinschaft (community) and
Gesellschaft (society) as follows: There is a contrast between a social order
which being based upon consensus of wills rests on harmony and is developed and ennobled by folkways, mores, and religion, and an order which
being based upon a union of rational wills rests on convention and agreement, is safeguarded by political legislation, and finds its ideological justification in public opinion.
The first poem, The Pasture, in Frosts North of Boston immediately preceding Mending Wall, is just such an invitation.
Scarrys theory is suggestive, even if there is as yet little empirical evidence
that the mind entertains insubstantiality more readily than solidity. Limitations of space and exigencies of presentation prevent me from documenting
all the cognitive science research that supports this and subsequent claims
throughout this paper. I refer readers to Scarrys book and my other reference
citations for further cognitive literature on the topics raised.
Lawrence Raab (1996) recounts Frosts reaction to the Russians making use of
his poem to defend the construction of the Berlin Wall by dropping the first
line. Frost wonders how they got the poem started and said he could have
done better for them by saying: Something there is that doesnt love a wall /
Something there is that does. It is noteworthy that Frost continues: Why
didnt I say that? I didnt mean that. I meant to leave that until later in the
poem. Why he meant to leave it until later results I think from the poems iconicity.
Johnson (1987: 29) defines a schema as a recurrent pattern, shape, and regularity in, or of, [the] ongoing ordering activities of our actions, perceptions,
and conceptions. See Johnson (1987: 1830) for a detailed discussion of how
this use of the term schema differs from earlier uses in cognitive science.
Examples of schema are PATH, CONTAINMENT, CHANGE, and the one I
am referring to here: BALANCE. For a discussion on how schemas can illuminate a poets poetics, see Freeman (2002).
I am grateful to Christina Ljungberg for pointing out the phonetic repetition
Cognitive poetics
425
of the [l] sound in the poem. A full cognitive-phonetic analysis of the poem
would show other iconic functions at work, but would lengthen what is already a long paper on a short poem.
14. It is rather the melting of the frost that causes subsidence and imbalance, as
the Encyclopedia Britannicas entry on frost notes: Frost action, the freezing
and thawing of moisture in the ground, has long been known to seriously disrupt and destroy structures in both polar and temperate latitudes. In the
winter the freezing of ground moisture produces upward displacement of the
ground (frost heaving), and in the summer excessive moisture in the ground
brought in during the freezing operation causes loss of bearing strength.
15. Norman Holland (1988: 26) quotes Frosts comments on the poem as follows:
Maybe I was both fellows in the poem (Interviews with Robert Frost, ed. Edward Connery Lathem (New York: Holt, Rinehart and Winston 1966) 257.
References
Aleksander, Igor
2005
The World in My Mind, My Mind in the World: Key Mechanisms of
Consciousness in People, Animals and Machines. Exeter, UK: Imprint
Academic.
Brandt, Line, and Per Aage Brandt
2005
Making sense of a blend: A cognitive-semiotic approach to metaphor.
Annual Review of Cognitive Linguistics 3: 216249.
Brandt, Per Aage
2004
Spaces, Domains and Meaning: Essays in Cognitive Semiotics. European Semiotics Series No 4. Bern: Peter Lang.
Cook, Reginald L.
1974
Robert Frost: A Living Voice. Amherst, MA: University of Massachusetts Press.
Crane, Mary Thomas
2001
Shakespeares Brain: Reading with Cognitive Theory. Princeton, NJ:
Princeton University Press.
Crane, Mary Thomas, and Alan Richardson
1999
Literary studies and cognitive science: Toward a new interdisciplinarity. Mosaic 32: 123140.
Emmott, Catherine
1997
Narrative Comprehension: A Discourse Perspective. Oxford: Clarendon
Press.
Fauconnier, Gilles, and Mark Turner
1994
Conceptual Projection and Middle Spaces. San Diego: University of
California, Department of Science Technical Report 9401 (available
on line at markturner.org/blending).
2002
The Way We Think: Conceptual Blending and the Minds Hidden Complexities. New York: Basic Books.
426
Margaret H. Freeman
Fludernik, Monika
1993
The Fictions of Language and the Languages of Fiction. London/New
York: Routledge.
Freeman, Margaret H.
2002
Momentary stays, exploding forces: A cognitive linguistic approach to
the poetics of Emily Dickinson and Robert Frost. Journal of English
Linguistics 30 (1): 7390.
In press. Cognitive linguistic approaches to literary studies: State of the art
in cognitive poetics. In Handbook of Cognitive Linguistics, Dirk Geeraerts, and Hubert Cuyckens (eds.). Oxford: Oxford University Press.
Frost, Robert
1915
Mending Walls. In North of Boston. New York: Henry Holt and Co.
Gavins, Joanna
2005
Text world theory in literary practice. In Cognition in Literary Interpretation and Practice, Bo Pettersson, Merja Polvinen, and Harri Veivo
(eds.), 89104. Helsinki: University of Helsinki Press.
Gavins, Joanna, and Gerard Steen (eds.)
2003
Cognitive Poetics in Practice. London/New York: Routledge.
Graves, Robert
1963
Introduction to Selected Poems of Robert Frost, ix xiv. New York:
Holt, Rinehart and Winston.
Henkel, Jacqueline
1996
Language of Criticism: Linguistic Models and Literary Theory. Ithaca,
NY: Cornell University Press.
Hogan, Patrick Colm
2003
Cognitive Science, Literature, and the Arts: A Guide for Humanists. New
York/London: Routledge.
Holland, Norman
1988
The Brain of Robert Frost: A Cognitive Approach to Literature. New
York/London: Routledge.
Jakobson, Roman
1987
Language in Literature. Cambridge, MA and London: The Belknap
Press of Harvard University Press.
Johnson, Mark
1987
The Body in the Mind: The Bodily Basis of Meaning, Imagination, and
Reason. Chicago/London: The University of Chicago Press.
Kemp, John C.
1979
Robert Frost and New England: The Poet as Regionalist. Princeton, NJ:
Princeton University Press.
Lakoff, George, and Mark Johnson
1980
Metaphors We Live By. Chicago/London: The University of Chicago Press.
1999
Philosophy in the Flesh. New York: Basic Books.
Lakoff, George, and Mark Turner
1989
More than Cool Reason: A Field Guide to Poetic Metaphor. Chicago/
London: The University of Chicago Press.
Cognitive poetics
427
Langacker, Ronald W.
1987
Foundations of Cognitive Grammar. Vol I: Theoretical Perspectives.
Stanford: Stanford University Press.
1991
Foundations of Cognitive Grammar. Vol II: Descriptive Application.
Stanford: Stanford University Press.
2005
Dynamicity, fictivity, and scanning: The imaginative basis of logic and
linguistic meaning. In Grounding Cognition: The Role of Perception and
Action in Memory, Language, and Thinking, Diane Pecher, and Rolf A.
Zwaan (eds.), 164197. Cambridge: Cambridge University Press.
Langer, Susanne K.
1953
Feeling and Form: A Theory of Art. New York: Charles Scribners.
1967
Mind: An Essay on Human Feeling. Baltimore: The Johns Hopkins
Press.
Lentricchia, Frank
1975
Robert Frost: Modern Poetics and the Landscapes of Self. Durham, NC:
Duke University Press.
Merleau-Ponty, Maurice
1962
Phenomenology of Perception, C. Smith (tr.). London: Routledge and
Kegan Paul.
1968
The Visible and the Invisible, Claude Lefort (ed.), Alphonso Lingis (tr.).
Evanston: Northwestern University Press.
Modell, Arnold H.
2003
Imagination and the Meaningful Brain. Cambridge, MA: The MIT
Press.
Monteiro, George
1988
Robert Frost and the New England Renaissance. Lexington, KY: The
University Press of Kentucky.
Nth, Winfried
2001
Semiotic foundations of iconicity. In The Motivated Sign, Max Nnny,
and Olga Fischer (eds.), 1729. Amsterdam: Benjamins.
Peirce, Charles Sanders
1955
Philosophical Writings of Peirce, Justus Buchler (ed.). New York: Dover.
Raab, Lawrence
1996
Touchstone: American Poets on a Favorite Poem, Robert Pack and Jay
Parini (eds.). Hanover: University Press of New England. http://www.
english.uiuc.edu/maps/poets/af/frost/wall.htm. Accessed April 15,
2006.
Richardson, Alan
2000
Rethinking romantic incest: Human universals, literary representation,
and the biology of mind. New Literary History 31(3): 553572.
Richardson, Mark
1997
The Ordeal of Robert Frost: The Poet and His Poetics. Urbana: University of Illinois.
Scarry, Elaine
1999
Dreaming by the Book. Princeton, NJ: Princeton University Press.
428
Margaret H. Freeman
429
430
431
Abstract
This chapter addresses not the internal nature of the field of Cognitive Linguistics
but rather its relationship to another domain, Artificial Intelligence. One important vein in AI research on metaphor is to use ideas drawn from or similar to the
notion of conceptual metaphor in Cognitive Linguistics, although another important vein has been to seek to account for (some) metaphor understanding from
scratch without any prior knowledge of particular mappings. AI can contribute to
Cognitive Linguistics by attempting to construct computationally detailed models
of structures and process drawn from or similar to those proposed more abstractly
in Cognitive Linguistics. The computational model construction can confirm the
viability of a proposal but can also reveal new problems and issues, or put existing
ones into sharper relief. A case in point is the problem of the nature of domains
and consequent problems in conceptions of the nature of metaphor and metonymy. In the authors approach to metaphor, which is briefly sketched, mappings
between domains are replaced by mappings between metaphorical pretence and
reality.
Keywords: metaphor; metonymy; conceptual domains; artificial intelligence; computational models
1.
Introduction
In this chapter I do not try to describe the geography of Cognitive Linguistics (henceforth CL) as a whole or of some internal region of it, but
rather to suggest the shape of the boundary between it and one other discipline, namely Artificial Intelligence. I do this partly by discussing how
my own work on figurative language has drawn from CL and in turn
offers potential contributions to the development of CL. I also briefly survey some other work in AI on figurative language and comment on some
of the relationships to CL. The chapter deals primarily with metaphor,
with some mention of metonymy, and does not attempt to address other
forms of figurative language. A more detailed discussion of various AI
projects on metaphor and of how AI can contribute to the study of
432
John A. Barnden
433
434
John A. Barnden
istical or related techniques that do not rest on considerations of the principles underlying human language, on any attempt to reason about how
human speakers or hearers understand process language, or even on any
derivation of meanings of utterances. Such approaches, which I will call
human-free here for ease of reference, have met with considerable practical success, and incidentally raise interesting issues for Psychological
and General/Philosophical AI. Nevertheless that still leaves a good body
of work in AI, whether with the Engineering aim or another aim, that
does concern itself with the principles of language, human cognition behind language, and meaning of language.
It should also be realized that a researcher pursuing a human-free approach for a particular purpose does not necessarily make strong claims
that the approach can achieve more than a certain, adequate-for-purpose
level of success, and may agree that for different purposes whether of the
Engineering, Psychological or General/Philosophical types the approach would be inadequate, and may therefore in regard to those other
purposes be prepared to be friendly to insights from fields such as CL and
Psychology.
Given these introductory comments, we will now move to looking
briefly at AI research on metaphor (and to some extent metonymy) in
general, in relation to CL concerns. After that we will consider the particular case of the ATT-Meta approach and system for metaphor processing developed in my own project, as a particular case study of CL-influenced work in AI. Then we will briefly look at a few potential issues or
challenges that that work raises in relation to CL, notably with regard to
the distinction between metaphor and metonymy.
2.
Metaphor has long been an interest within AI. Salient amongst the earliest work is that of Carbonell (1980, 1982), Russell (1976), Weiner (1984),
Wilks (1978) and Winston (1979). Other more recent work includes Asher
and Lascarides (1995), Fass (1997), Hobbs (1990, 1992), Indurkhya (1991,
1992), Lytinen, Burridge, and Kirtner (1992), Martin (1990), Narayanan
(1997, 1999), Norvig (1989), Veale and Keane (1992), Veale (1998), Way
(1991) and Weber (1989), and the work on my own project (cited below),
all of which address the problem of understanding metaphorical utterances. There has also been statistical work on uncovering metaphoricity
in corpora (e.g. Mason 2004). As for metonymy, research includes that of
435
Fass (1997), Hobbs (1990, 1992) and Lytinen, Burridge, and Kirtner
(1992) again, and also Lapata and Lascarides (2003), Markert and Hahn
(2002), Markert and Nissim (2003) and Stallard (1987, 1993). Also, much
work on polysemy involves metonymy at least implicitly given that often
polysemy is driven relatively straightforwardly by metonymic connections
(see, for example, Fass 1997, on the work of Pustejovsky see, e.g., Pustejovsky 1995). See Fass (1997), Martin (1996) and Russell (1986) for more
comprehensive reviews of work on figurative language in AI. Also, see
Barnden (in press, a) for more extensive description than is possible here
of the work of Wilks, Fass, Martin, Hobbs, Veale and Narayanan, and see
Barnden (in press, b) for additional comments on Wilkss work.
In spite of this long history in AI, metaphor has only ever been a minority interest in the field, and to some extent this is true of metonymy as
well. The minority status of metaphor in AI is no doubt for complex historical reasons. The message from CL and elsewhere that metaphor is fundamental to and prevalent in ordinary language has not had the effect on
AI it should have had, despite the point being perfectly evident even in the
earliest AI work mentioned above, and being explicitly plugged by, for instance, Carbonell. It is also surprising in view of the international prominence within AI of such figures as Carbonell, Hobbs and Wilks. There has
remained some tendency, despite the evidence, to view metaphor as an
outlying, postponable phenomenon, but perhaps more importantly the
field has happened to concentrate on other particular problems of language such as anaphor resolution and ordinary word-sense disambiguation (ordinary in the sense of not paying explicit attention to the role of
metaphor or metonymy in polysemy). The field, perhaps through familiarity more than rational evaluation, has tended to regard such problems
as intrinsically easier and more tractable than those of metaphor and metonymy.
However, metaphor is becoming an increasingly looming obstacle for
(even) Engineering AI, as attempts are made to bring better automated
human-language processing into commercial products, to develop ever
more advanced computer interfaces and virtual reality systems, to develop automated understanding and production of emotional expression
given that this is often conveyed explicitly or implicitly by metaphor (Delfino and Manca 2005; Emanatian 1995; Fainsilber and Ortony 1987; Fussell and Moss 1998; Kvecses 2000; Thomas 1969; Yu 1995), and also to
develop systems that can understand or produce gesture and sign language, given that these forms of communication have strong metaphorical
aspects (McNeil 1992; P. Wilcox 2004; S. Wilcox 2004; Woll 1985). It is to
436
John A. Barnden
437
438
John A. Barnden
3.
439
With colleagues I have developed an approach and related computer program, called ATT-Meta, for performing a type of reasoning that is arguably often necessary for metaphor interpretation. The approach is described in, for instance, Barnden (1998, 2001a); Barnden, Glasbey, Lee
and Wallington (2004); Barnden, Helmreich, Iverson and Stein (1994,
1996); Barnden and Lee (1999, 2001); Lee and Barnden (2001a). The implemented ATT-Meta program is only a reasoning system and does not
take linguistic strings as input, but takes, rather, logical forms derived
from sentences by initial processing. For now the reader can take these
logical forms to encode the literal meanings of the sentences, but we will
refine this point below.
The metaphorical utterances of main interest in the ATT-Meta project
are those that are conceptually related to known conceptual metaphors
but that transcend them by involving source-domain elements not directly
handled by the mappings in those metaphors. In ATT-Meta parlance
these utterances are map-transcending. For instance, to take a modified
version of an example from Hobbs (1990), consider the sentence [the
computer program variable] N leaps from 1 to 100. Suppose the understander only knows a physically-leap lexical sense for the verb leap but
does not know a mapping for that sense into the target domain of variables and values, even though he/she/it does knows a mapping from, say,
spatially-at to have-as-value. Thus, the sentence is map-transcending in
using a physically-leap concept that is related to the mappable concept
spatially-at but that is not itself mapped. Similarly, to take an example
from Martin (1990), if an understander knows a metaphorical mapping
from physically-in to using-a-process but has no mapping for physicallyenter, then the sentence How do I enter [the computer program] Emacs?
is map-transcending.
Clearly, map-transcendence is a fuzzy concept that is relative to particular understanders and to particular conceptual metaphors the understander knows, and to our intuitive perceptions as to what is conceptually
related to what (e.g., physically-leaping to being-spatially-at). Nevertheless, it is a useful intuitive characterization of a phenomenon that lies
along a broad sector of the spectrum between completely conventional
metaphorical phraseology on the one hand and, on the other hand, entirely novel metaphor where no relevant mapping is known at all.
Map-transcendence is strongly related to the phenomenon of unused
parts of the source domain as discussed in Lakoff and Johnson (1980).
440
John A. Barnden
However, we prefer to avoid that term because it may wrongly suggest that
the parts in question have never been used. Far from it: it is in principle
possible even that a given understander has often met sentences of the form
variable leaps from value1 to value2 without happening to bother to develop a mapping from physically-leap to something in the target domain.
ATT-Meta is based on rules encapsulating known metaphorical correspondences, such as between physically-at and has-as-value, and on an integrated inferential framework which, in particular, allows arbitrarily rich
source-domain reasoning to connect map-transcending sentence components to source-domain concepts that can be mapped by known mappings. In this respect, ATT-Metas approach is similar to one strand of
Hobbss approach to metaphor. Both approaches can infer that a variable
N has value 100 from any sentence couched in spatial terms that implies
that N is physically-at 100, as long as the systems have the necessary
knowledge about physical space to infer that N is physically-at 100 from
the sentence. The inference can be arbitrarily indirect and complex in
principle. To make the point, a vivid example would be a sentence such as
N meandered along a circuitous route towards 100 but didnt complete
the journey until after M fell to 0. This implies, among other things, that
N (at some point) had value 100.
However, there is a fundamental difference of approach, as well as
many technical differences of representation and reasoning, between
ATT-Meta and Hobbss scheme. The fundamental difference is that ATTMeta avoids placing internal propositions such as N is physically-at 100,
which are not statements about reality, on a par with statements such as N
has value 100, which are. Hobbss approach does maintain them on a par:
there is nothing in his internal representation to say that the former proposition is merely a metaphorical pretence or fiction.
Instead, ATT-Meta creates a special computational mental space in
which such propositions and inference arising from them are kept aside
from propositions and reasoning about reality. We call this space a metaphorical pretence cocoon. Thus, the internal proposition N physicallyleaps from 1 to 100 arising directly from the sentence N leaps from 1 to
100 is placed in the cocoon, and an inference result that (say) N is spatially-at 100 afterwards, together with the inference chain itself, lies within
the cocoon. A metaphorical mapping rule that takes spatially-at to hasas-value can then give the result that, in reality, N has value 100 afterwards as one aspect of the meaning of the utterance.
By clearly marking some propositions as being pretences, the use of a
cocoon ensures that the system is not misled by the propositions directly
441
derived from metaphorical utterances, that is, propositions like N physically-leaps from 1 to 100. This isnt perhaps a practical problem in this
particular example, but notice that in the case of McEnroe killed Connors, which could be taken literally in principle, the understander needs
to be clear that the directly-derived proposition McEnroe biologicallykilled Connors is not a statement about reality.
But, in addition, we do not want to allow the knowledge that McEnroe
definitely did not biologically-kill Connors in reality to defeat the pretend
information that McEnroe did biologically-kill Connors. Thus, pretence
cocoons prevent pretences from infecting reality but, equally, protect the
integrity of pretences.
The use of cocoons has another benefit. Lee and Barnden (2001a)
studied mixed metaphor of various types, and showed how ATT-Meta deals
with them. The main distinction studied was between serial mixing (commonly called chaining), where A is viewed as B and B is viewed as C, and
parallel mixing, where A is used simultaneously as B and as C. (See also
Wilks, Barnden and Wang 1991.) Serial mixing is viewed as having the B
material in a cocoon directly embedded in the reality space, whereas the C
material as in a cocoon embedded within the B cocoon. Thus, there is a pretence within a pretence. In parallel mixing, on the other hand, the B and C
material is either combined in a single cocoon or is in two separate cocoons
both directly embedded within the reality space. Thus, we have two pretences either side by side or blended with each other. There are unresolved
issues about how to decide between these two possibilities, but in any case
different dispositions of pretence cocoons allow important differences between types of mixing of metaphor to be reflected in the processing.
We have indicated that what is initially inserted in the pretence cocoon in
the case of N leaps from 1 to 100 is the proposition N physically-leaps
from 1 to 100, and what is inserted in the case of McEnroe killed Connors
is McEnroe biologically-killed Connors. This reflects a general assumption
in the ATT-Meta approach that what is inserted in the cocoon is a direct
meaning of the metaphorical sentence (or of some metaphorical sentencecomponent such as a clause). A direct meaning is a logical form derived
compositionally from the direct senses of lexical units in sentences. A direct sense is just any sense listed for the lexical unit in the understanders
lexicon, so that it is directly accessible from the lexical unit. In particular,
we have been assuming that the verbs leap and kill have as direct senses
the concepts of physically-leap and biologically-kill respectively.
Clearly, a given lexical unit could actually have more than one direct
sense, and indeed some of the direct senses could be metaphorical or
442
John A. Barnden
443
to degree D.
The presumably annotation makes this rule a default rule: even if the
rules condition is established with complete certainty, the system only
takes the result of the rule as a default (a working assumption). The rule
also illustrates that ATT-Meta can handle gradedness with which states of
affairs can hold. There is a small set of qualitative degrees going from
very low up to absolute.
We are now ready to look in more detail at an example. Consider:
In the far reaches of her mind, Anne believed that Kyle was having an affair.
444
John A. Barnden
445
4.
I am also engaged in a related research project that draws upon the ATTMeta research. I mention this project to show another case of how CL can
not only affect theoretical research in AI but can also affect work that is
immensely practical. The project is looking at the metaphorical expression of affect (emotion, value judgments, etc.) and at the automated
detection of affect in utterances, in the context of an e-drama system
that supports virtual dramatic improvisation by users (actors) sitting at
computer terminals (Zhang, Barnden and Hendley 2005). The initial system on which the project is based was provided by one of our industrial
partners, Hi8us Midlands Ltd. Improvisations can be on any topic, but
the system has in particular been used for improvisations concerning
school bullying and embarrassing illnesses such as Crohns Disease. The
446
John A. Barnden
actors improvise within a loose scenario, which for instance specifies that
a particular character is a bully and another particular character is the
bully victim. However, actors are free to improvise creatively, and in fact
do things such as starting to bully the bully; or someone who is meant to
be a friend of the victim may turn against him/her.
The affect detection is mainly towards providing an automated assistant to the human director who monitors the unfolding improvisation. The
director can intervene by sending advice messages to the actors or by introducing a bit-part character into the improvisation in order to stir
things up. Since one reason for directorial intervention is that the affective
quality of the improvisation could be improved upon (e.g., the emotions
expressed are neither appropriate to the provided scenario nor indicative
of a creative novel direction), an important ability of an automated assistant is to detect affect expressed in the characters speeches (which are entirely textual, and shown in text bubbles above the characters on the computer screens).
In fact, the affect detection mechanisms we have developed have not yet
(at the time of writing) been used directly in an assistant, but instead in an
automated actor program that can operate a bit-part character with the
intent that such a character be introduced on the discretion of the director, and thus indirectly help the director in his/her job. The actor program
does not do full understanding of utterances of other characters (which
would in any case be beyond the state of the art in AI), but does a certain
amount of complex syntactic and semantic processing in order to try to
detect some emotions, emotional emphases, and value judgments. As part
of this, we are developing methods for limited sorts of metaphor analysis,
and are also using the metaphor examples that freely arise in the e-dramatic improvisation genre as a fertile ground for testing our metaphor
theories and suggesting new theoretically interesting examples and phenomena.
As mentioned above, the metaphorical description of emotional states
(or behaviour) is common. Two illustrations are He nearly exploded
and Joy ran through me. Such examples describe emotional states directly (perhaps without naming them). We are concerned to have some
basic handling at least of this phenomenon. Examples we have encountered
in e-drama improvisations include the conventional metaphorical phraseology italicized in Its a struggle, Its on my mind 24/7 and Listen,
dont keep it in, if you need to talk Im all ears. (These linguistic strings
are sanitized versions of the originals encountered, which include mobilephone-style textese, spelling errors, etc.)
447
448
John A. Barnden
one! and Im trying very hard but theres too much stuff blocking my
head up.
One particular phenomenon of theoretical and practical interest is that
physical size is often metaphorically used to emphasize evaluations, as in
you are a big bully, youre a big idiot, and Youre just a little bully.
Big bully expresses strong disapproval and little bully can express
contempt, although little can also convey sympathy or be used as an endearment. Such examples are not only important in practice but also theoretically challenging, especially as the physical size mentioned can in
some cases be intended literally as well as metaphorically.
We have encountered surprisingly creative uses of metaphor in e-drama.
For example, in the school-bullying scenario, Mayid is portrayed as having already insulted Lisa by calling her a pizza (short for pizza-face).
This figurative insult was given a theoretically intriguing, creatively metaphorical elaboration in one improvisation, where Mayid said Ill knock
your topping off, Lisa.
5.
While ATT-Meta draws strongly on general notions concerning conceptual metaphor, there are a number of observations we need to make. First,
the ATT-Meta approach makes no use of the notion of image schemata or
the notion that conceptual metaphors arise from embodied experience
(Lakoff 1987). The approach is not inconsistent with these notions, but it
is agnostic as to their validity. In particular, we fully embrace metaphor
where the source subject-matter is more abstract and presumably less
embodied than the target subject-matter.
Next, it is worth noting that ATT-Metas mapping rules (the transfer
rules) are more akin in their extent of general purpose-ness and fundamentality to the primary metaphorical mappings that Grady (2002) proposes than they are to the original types of conceptual metaphorical mapping, such as the mapping between lovers and travellers in the LOVE AS
JOURNEY conceptual metaphor. For example, the mapping we used in
the Anne/Kyle example Section 3 was one between PHYSICAL OPERATION (on ideas viewed as physical objects) and MENTAL OPERATION.
PHYSICAL OPERATION and MENTAL OPERATION here are extremely
general notions, and the mapping between them can underlie many different specific metaphorical views such as IDEAS AS ANIMATE BEINGS,
MIND AS MACHINE and so forth.
449
450
John A. Barnden
obstacle to the journey causes difficulty for the progress of the love relationship, because it causes difficulty for the physical journey. There is no
need to have a specific mapping between journey difficulties and relationship difficulties, cf. the correspondence between difficulties in a love relationship and impediments to travel in Lakoff (1993: 207).
The importance of this factoring out of many aspects of conceptual
metaphors into VNMAs is magnified when one realizes that discussions in
the literature on conceptual metaphor often tend to leave some assumed
mapping abilities purely tacit, without any specific explanation. For
example, in the discussion of Were driving in the fast lane on the freeway
of love in Lakoff (1993), it is a said that the excitement the imaginary
travellers experience in travelling fast transfers to the target to become excitement of the lovers. But no mapping is provided that relates emotions on
the source side to emotions on the target side there is simply a tacit assumption that the emotion transfers. By postulating explicit VNMAs we
are attempting to bring such tacit assumptions out into the open and make
them into theoretically interesting aspects of metaphor in their own right.
This move also has the benefit of raising the need to explicitly consider exceptions to VNMAs, e.g. exceptions to the general rule that emotions of
source agents map over to emotions of corresponding target agents.
In conceptual metaphor theory, metaphor is generally viewed as mapping between two domains. In contrast, the ATT-Meta approach does
not rely on domain distinctions, even theoretically, let alone enshrine
them in some way in the implemented system. Although in this article we
generally adopt the common practice of saying that metaphor transfers
information from a source domain to a target domain, this purely for purposes of intuitive exposition, and the ATT-Meta approach has a different
technical stance: metaphor is a matter of transferring from a pretence to
reality (or to a surrounding pretence, in the case of serial mixing). Notice
that in the mapping rules as described in Section 3, reference is made to
pretence and reality, not to domains. It does not matter what domains the
information used in the pretence comes from, and this means that it does
not matter how we may intuitively circumscribe the source and target domains in the metaphor. In particular, it does not matter how close, difficult to distinguish, or overlapping those domains are. In practice, it will
often be the case that we can theoretically identify a source domain in
which the direct meaning of the sentence lies, and that inferences from this
meaning also lie within that domain. However, this has no bearing on the
course of processing, and the reasoning within the pretence is not limited
by any consideration of domains.
451
Some other authors, in CL and elsewhere, have questioned the usefulness of the domain notion or the degree of distinctness that is required between the two domains in a metaphor. See, e.g., Cameron (1999), Kittay
(1989) and several chapters in Dirven and Prings (2002). I have found in
working on the ATT-Meta approach that the detail and clarity required
for well-founded computational implementation to be a major factor in
coming to doubt the usefulness of the concept of domain, or related
concepts such as idealized cognitive models, as a source of constraints in
characterizing metaphor (and metonymy) or as a real aspect of metaphorical processing. In trying to make decisions about what domains particular pieces of knowledge should be assigned to, in a way that respects
the idea that metaphor maps between different domains, I came to realize
what a hopeless and arbitrary task it was.
The nature of some of the other AI approaches to metaphor mentioned
above also throws doubt on the usefulness of the notion of domains. For
instance, although Hobbs does believe that metaphor is a matter of mapping between qualitatively disparate domains, this stance has no operational effect in his approach. In Narayanans work, domains do play an
operational role, but he does not note the difficulties arising from a major
degree of overlap between the domains that he presents as if they were entirely separate (for more discussion, see Barnden, in press, a).
It is certainly quite common for the two sides of a metaphor to involve
qualitatively very different subject matters. But, equally, the two sides of a
metaphor can be arbitrarily close in their qualitative nature. Metaphors
such as Jules Verne is the H.G. Wells of France have source and target
domains that are broadly similar in subject matter. For an example with
even less qualitative distance between the two sides, ones neighbours
teenage children can act as a metaphor for ones own: if one has a
daughter Jenny and the neighbours have a son Jonathan who behaves
similarly to Jenny, then one could say Jenny is our familys Jonathan. Of
course, it is open to someone to say that the Jenny family is qualitatively
different from the Jonathan family, and that they are therefore different
domains, but this is post-hoc rationalization with no operational significance.
Despite the closeness between target and source in the Jenny/Jonathan
example, the metaphorical utterance appears quite apt to the present
author. If this impression is shared with other language users, it may appear to conflict with the the evidence adduced by Tourangeau and Sternberg (1982) that, within limits, the greater the conceptual distance between source and target the more apt the metaphor. However, note that
452
John A. Barnden
the linguistic form of the metaphorical utterance and the presence of context are important factors. The simple equation Jenny is Jonathan without much context might well not be judged as apt.
It is quite possible to maintain a fiction that domains are useful in characterizing metaphor or analysing metaphorical processing as long as one
only deals schematically with some isolated examples, and does not try to
come up with a unified and processually detailed approach to metaphor
that can work on a wide variety of metaphors on the basis of the same
overall knowledge base. This is not to say qualitative differences between
the two sides of a metaphor are not interesting: the amount of difference
can potentially affect aptness, vividness, persuasiveness, perceived indirectness or politeness, etc.
Finally, the virtual abandonment of domains as a genuine operational
or explanatory construct in metaphor leads to heightened difficulty in distinguishing metaphor from metonymy, given some tendency in CL to view
metaphor as operating between domains and metonymy within domains.
In recent work (Barnden 2005) I have begun to develop a view of metaphor and metonymy as merely being vague notions that serve only a heuristic role in thinking about language, that the real phenomena lie at a
more detailed level of description, and that the differences between what
is commonly classed as metaphor and what is commonly classed as metonymy is a matter of differences on a sizable set of dimensions (which form
the more detailed level of description). It is these dimensions that have
reality, not metaphor and metonymy in themselves. The dimensions include such matters as the degree of conceptual distance between the
source and target items in a metaphor or metonymy, the degrees to which
similarity and contiguity are involved, and the degree to which the connection(s) between source item(s) and target items(s) are themselves part
of the message conveyed. In the imaginary N-dimensional space spanned
by these and other dimensions, metaphor and metonymy form fuzzy
clouds that have no firm borders and that potentially overlap, and when
attention is confined to only one dimension it can be very hard to draw
any firm distinction between metaphor and metonymy; so the multi-dimensionality is important in itself in distinguishing between metaphor
and metonymy.
The developing analysis is in much the same spirit as the proposals of
CL authors such as Radden (2002) who put metaphor and metonymy at
different points on a linear spectrum of phenomena, except that we have
taken the idea further and replaced an individual spectrum by a multitude
of spectra (the mentioned dimensions), and made each individual spec-
453
6.
Conclusion
Acknowledgments
This chapter draws on research supported by current grant EP/C538943/1
and previous grant GR/M64208 from the Engineering and Physical
Sciences Research Council (EPSRC) of the UK. The e-drama research
draws also from grant RES-328250009 from the Economic and Social
Research Council (ESRC) of the UK, under the PACCIT programme
funded by the EPSRC, ESRC and DTI (UK governments Department of
Trade and Industry). I am grateful to the industrial partners on that grant
(Hi8us Midlands Ltd, Maverick TV Ltd, and BT), and to my colleagues in
the metaphor and e-drama research groups in my department.
454
John A. Barnden
References
Asher, Nicholas, and Lascarides, Alex
1995
Metaphor in discourse. In Proceedings of the AAAI Spring Symposium
Series: Representation and Acquisition of Lexical Knowledge: Polysemy,
Ambiguity and Generativity, 37. Stanford, CA.
Barnden, John
1998
Combining uncertain belief reasoning and uncertain metaphor-based
reasoning. In Proceedings of the Twentieth Annual Meeting of the Cognitive Science Society, Morton A. Gernsbacher, and Sharon J. Derry
(eds.), 114119. Mahwah, NJ: Lawrence Erlbaum.
2001a Uncertainty and conflict handling in the ATT-Meta context-based
system for metaphorical reasoning. In Proceedings of the Third International Conference on Modeling and Using Context, Varol Akman,
Paolo Bouquet, Rich Thomason, and Roger A. Young (eds.), 1529.
Lecture Notes in Artificial Intelligence, Vol. 2116. Berlin: SpringerVerlag.
2001b Application of the ATT-Meta metaphor-understanding approach to
selected examples from Goatly. Technical Report CSRP0101, School
of Computer Science, The University of Birmingham.
2001c Application of the ATT-Meta metaphor-understanding approach to
various examples in the ATT-Meta project databank. Technical Report
CSRP01 02, School of Computer Science, The University of Birmingham.
2005
Metaphor and metonymy: Escaping the domination of domains. Talk
delivered at New Directions in Cognitive Linguistics: First UK Cognitive
Linguistics Conference, University of Sussex, Brighton, U.K., October
2005.
In press a Metaphor and artificial intelligence: Why they matter to each other. In
Cambridge Handbook of Metaphor and Thought, Ray W. Gibbs, Jr.
(ed.). Cambridge, MA: Cambridge University Press.
In press b Metaphor, semantic preferences and context-sensitivity. Words and Intelligence. Vol. 1: A Festschrift for Yorik Wilks, C. Brewster, M. Stevenson (eds.). Berlin: Springer-Verlag.
Barnden, John, Sheila Glasbey, Mark Lee, and Alan Wallington
2002
Application of the ATT-Meta metaphor-understanding system to
examples of the metaphorical view of TEACHERS AS MIDWIVES.
Technical Report CSRP0210, School of Computer Science, The University of Birmingham.
2003
Domain-transcending mappings in a system for metaphorical reasoning. In Conference Companion to the 10th Conference of the European
Chapter of the Association for Computational Linguistics (EACL 2003),
5761. Association for Computational Linguistics.
2004
Varieties and directions of inter-domain influence in metaphor. Metaphor and Symbol 19 (1): 130.
455
456
John A. Barnden
457
458
John A. Barnden
Norvig, Peter
1989
Marker passing as a weak method of text inferencing. Cognitive Science
13 (4): 569620.
Peirsman, Yves, and Dirk Geeraerts
In press Metonymy as a prototypical category. Cognitive Linguistics.
Pereira, Francisco
In press Creativity and Artificial Intelligence: A Conceptual Blending Approach.
Berlin/New York: Mouton de Gruyter.
Pustejovsky, James
1995
The Generative Lexicon. Cambridge, MA: MIT Press.
Radden, Gnter
2002
How metonymic are metaphors? In Metaphor and Metonymy in Comparison and Contrast, Ren Dirven, and Ralf Prings (eds.), 407434.
Berlin/New York: Mouton de Gruyter.
Russell, Sylvia Weber
1976
Computer understanding of metaphorically used verbs. American Journal of Computational Linguistics 2: 1528.
1986
Information and experience in metaphor: a perspective from computer
analysis. Metaphor and Symbolic Activity 1 (4): 227270.
Russell, Stuart, and Peter Norvig
2002
Artificial intelligence: A Modern Approach. Englewood Cliffs, NJ: Prentice-Hall.
Stallard, David
1987
The logical analysis of lexical ambiguity. In Proceedings of the 25th Annual Meeting of the Association for Computational Linguistics, 179185,
Stanford, CA: Stanford University.
1993
Two kinds of metonymy. In Proceedings of the 31st Annual Meeting of
the Association for Computational Linguistics, 8794. Ohio: Ohio State
University.
Thomas, Owen
1969
Metaphor and Related Subjects. New York: Random House.
Tourangeau, Roger, and Robert Sternberg
1982
Understanding and appreciating metaphors. Cognition 11: 203244.
Veale, Tony
1988
Just in Time analogical mapping, an iterative-deepening approach to
structure-mapping. In Proceedings of the Thirteenth European Conference on Artificial Intelligence (ECAI98). John Wiley.
Veale, Tony, and Mark Keane
1992
Conceptual scaffolding: a spatially founded meaning representation
for metaphor comprehension. Computational Intelligence 8 (3):
494519.
Veale, Tony, and Diarmuid ODonoghue
2000
Computation and blending. Cognitive Linguistics 11 (3): 253281.
Way, Eileen
1991
Knowledge Representation and Metaphor. Dordrecht: Kluwer.
459
Weber, Susan
1989
Figurative adjective-noun interpretation in a structured connectionist
network. In Proceedings of the Eleventh Annual Conference of the Cognitive Science Society, 204211. Hillsdale, NJ: Lawrence Erlbaum.
Weiner, Judith
1984
A knowledge representation approach to understanding metaphors.
Computational Linguistics 10 (1): 114.
Wilcox, Phyllis
2004
A cognitive key: Metonymic and metaphorical mappings in ASL. Cognitive Linguistics 15 (2): 197222.
Wilcox, Sherman
2004
Cognitive iconicity: Conceptual spaces, meaning, and gesture in signed
language. Cognitive Linguistics 15 (2): 119147.
Wilks, Yorick
1978
Making preferences more active. Artificial Intelligence 11: 197223.
Wilks, Yorick, John Barnden, and Jin Wang
1991
Your metaphor or mine: belief ascription and metaphor interpretation.
In Proceedings of the 12th International Joint Conference on Artificial
Intelligence, 945950. San Mateo: Morgan Kaufmann.
Winston, Patrick
1979
Learning by creating and justifying transfer frames. In Patrick Winston
and Richard Brown (eds.), Artificial Intelligence: An MIT Perspective,
Vol. 1, 347374. Cambridge, MA.: MIT Press.
Woll, Bencie
1985
Visual imagery and metaphor in British Sign Language. In Wolf Paprott and Ren Dirven (eds.), The Ubiquity of Metaphor: Metaphor in
Language and Thought, 601628. Amsterdam/Philadelphia: Benjamins.
Yu, Ning
1995
Metaphorical expressions of anger and happiness in English and Chinese. Metaphor and Symbolic Activity 10 (2): 5992.
Zhang, Li, John Barnden, and Robert Hendley
2005
Affect detection and metaphor in e-drama: The first stage. In Proceedings of the Workshop on Narrative Learning Environments, 12th International Conference on Artificial Intelligence in Education (AIED2005), Amsterdam.
460
John A. Barnden
461
Abstract
Scientific theories are more than purely formal constructs, but linguistic artefacts
that often rely on the rhetorical qualities of language to give their claims additional resonance and argumentative force. This reliance of theory upon language
is even greater in those theoretical domains whose main concern is language itself,
leading to a sometimes convenient blurring of content and form. Though Cognitive Linguistics has consistently revealed metaphor to be a fundamental building
block in the development of complex conceptualizations, Cognitive Linguistic
theories often exploit metaphor as an allusive place-holder when more formal
clarity is demanded. Nonetheless, in this paper we argue that one such metaphor
the MIND-AS-COMPUTER metaphor that underlies the enterprises of Artificial
Intelligence and Cognitive Science can yield precisely the kind of formal clarity
that is required by the most suggestive or radical of theories in Cognitive Linguistics. An exploration of how this metaphor informs the computational realization of cognitive theories will allow us to illuminate the often wide gap that exists
between the descriptive suggestiveness of a linguistic theory and its actual computational sufficiency.
Keywords: Computability, Tractability; Mind As Computer; Verbal Humour;
Blending; Radical Construction Grammar
1.
Introduction
462
Tony Veale
phors, such as the mind as grain mill, have passed their use-by date, is it
not likely that the mind as computer metaphor will one day be superseded
by another, more apropos model of cognition?
A strong rebuttal to this scepticism is to be found in the extreme generality of the modern computer, which is much more than a mere calculating machine but a universal computing device in itself. This universality implies that any realizable computing device can itself be simulated by
a general purpose computer. Such a computer is a physical instantiation
of a Turing machine, an abstract mathematical device on which every
function that can naturally be considered computable can be computed
(see Turing 1936). As such, if one is inclined to see the actions of the mind
in terms of information processing, as Cognitive Science demonstrates we
must, then the mind as computer metaphor has no expiry date. By information processing we mean, of course, more than the manipulation of
electronic data, but include also the construction and manipulation of
mental representations that have computational correlates in the form of
strings, frames, records, trees, graphs and networks (see Veale and ODonoghue 2000). This general belief in the mind as computer metaphor has
been dubbed the hypothesis of computational sufficiency by Ray Jackendoff, who makes the following bold claim:
Every phenomenological distinction is caused by/supported by/projected from
a corresponding computational distinction. Jackendoff (1987: 24).
463
1996), and especially so to any linguistic theory that posits complex abstract machinery to capture the intricacies of natural language phenomena. Consider Chomskys (1957) transformational grammar (TG), which
models language as a multi-strata phenomenon in which complex transformations mediate between an unseen deep-structure representation and
a visible surface form. The fact that TG approaches to language are often
misunderstood as theories of language generation is due only in part to
the ambiguity inherent in Chomskys notion of linguistic generativity, for
the fact is, TG-based models are more computationally felicitous when
viewed as theories of language generation. The burdensome complexity of
TG for parsing purposes is demonstrated, for example, by Plaths (1973)
transformational parser, which requires an additional layer of transformational inverses to apply TG in reverse to an initial surface structure,
and requires a complex sanity-checking mechanism (based on the forward
application of TG transformations to the derived deep-structure) to additionally ensure that these inverses are always applied in a truly reversible
manner. Just as it much easier to break a teacup than to properly re-assemble its broken parts, some theories can be seen as inherently directional from a computational perspective.
In this paper we shall explore the pertinence of a computational perspective to theories in Cognitive Linguistics by focussing on three specific theories of language use. The first two model creative linguistic phenomena,
while the third represents a rather extreme example of the constructionbased approach to language study. Since creativity is an area of study where
one intuitively expects to find the least amount of formal substance, these
first two models Attardo and Raskins General Theory of Verbal Humour
or GTVH (1991) and Fauconnier and Turners theory of Conceptual Integration Networks (1998, 2002) therefore allow us to explore the often wide
gap between descriptive suggestiveness and true computational sufficiency.
This notion of sufficiency cuts directly to the computational core of a theory, and allows us to consider distinct theories for distinct phenomena as
comparable if they ultimately hinge upon the same computational distinctions. This notion of computational equivalence a corollary of sorts of
computational sufficiency allows us to consider the computational feasibility of Radical Construction Grammar, not in terms of its own software
realization but in terms of a comparable computationally-precise theory of
language translation that has been implemented on a large scale.
The rest of the paper assumes the following outline: in section two we
provide capsule descriptions of the three linguistic theories that fuel the
arguments to come; in section three we then consider three desirable fea-
464
Tony Veale
2.
Before considering the role of computationalism in linguistic theorization, we first consider a broad overview of the three particular models on
which the substance of our arguments will be focussed.
2.1. The general theory of verbal humour
Attardo and Raskins GTVH, which currently dominates humour research, is a juxtapositional theory of humour that is a modular reworking of Raskins (1985) Semantic Script Theory of Humour (or SSTH).
Like the SSTH, the GTVH views a joke as a narrative that is compatible
with multiple scripts, one of which will at first appear primary until the
punch-line contrives a incongruity that must be resolved (see Ritchie 2003;
Veale 2004). Resolution is achieved, either partially or fully, by a special
logical mechanism that analyses the nature of the mismatch between the
primary script and the text, before switching the thrust of interpretation
from this script to another. For instance, it is suggested that an LM called
false-analogy is central to jokes whose humour derives from ill-judged
comparisons, as in the old joke where a mad scientist builds a rocket to the
sun but plans to embark at night to avoid being cremated. Here a false
analogy is created between the sun and a light-bulb, suggesting that when
the sun is not shining it is not turned on, and hence, not hot. Different
LMs may be employed in different jokes, bringing a distinctive logical flavour to each. More recently, Attardo et al. (2002) enumerate a variety of
different logical mechanisms (27 in all) and offer a new, graph-theoretic
account of script representation that now sees scripts as arbitrarily complex symbolic structures to which juxtapositional processes like subgraph isomorphism can be applied. GTVH scripts can now accommodate
not just the semantic structure of events, but the phonological structure of
465
words. For example, Attardo et al. (ibid) argue that the pun in the mathematical book title The joy of sets is resolved by structure-mapping two
phonological script graphs, [s, e, tz] and [s, e, ks].
2.2. Conceptual integration networks or blending theory
Our second model of interest is Fauconnier and Turners (1998, 2002) theory of conceptual integration networks, more popularly called blending
theory. Like the GTVH, blending theory offers a juxtapositional account
of creative language use, one in which multiple mental spaces (called inputs) are integrated to generate a new mental space, the blend space, that
contains a selective projection of the inputs which is formed subject to a
variety of interacting optimality principles. This new space, which is connected to, but independent from, the original input spaces, is invoked to
explain the emergent properties of many compositional structures, from
metaphors to noun compounds to jokes. When relations between elements of different input spaces are projected into a blended space, they
may become compressed through a process of metonymic tightening.
Thus, a similarity relation outside the blended space may become an identity relation within the blended space. Coulson (2000) has used blending
theory to offer a theory of humour, called frame shifting, that is in many
respects similar to the script-switching view of the GTVH. Though some
attempts have been made to realize blending theory in computational
form (e.g. see Veale and ODonoghue 2000), such implementations are
necessarily incomplete and require a loose, almost metaphorical reading
of the theorys principal mechanisms.
2.3. Radical Construction Grammar
Our third model is Crofts (2001) Radical Construction Grammar, which,
as its name suggests, is a theory whose simplifying assumptions pose a
radical challenge to key elements of linguistic orthodoxy. Like other construction grammars, RCG gives a central position to the role of the construction, a mapping between form and meaning that can be almost entirely substantive (as in the case of idioms) or entirely parameterized (as in
the case of grammatical schemata). RCG thus obliterates the traditional
distinction between lexicon and grammar, as constructions may lie anywhere along a lexico-syntactic continuum. As in other variants of construction grammar (e.g. see Goldberg 2003), the purpose of constructions
in RCG is to break down the traditional barrier between form and mean-
466
Tony Veale
3.
467
main (e.g., the Earth is an electron buzzing around its nucleus, the sun) while
a good simile is only superficially revealing (e.g., the Earth is like a football). Are the GTVH and blending theories merely similes for the superficial workings of our creative use of words, or do they capture real cognitive mechanisms at work? The GTVH does not explain why a text that is
compatible with two overlapping but incongruous scripts should be considered humorous, but merely claims that successful jokes appear to exhibit this property. In specifying 27 different logical mechanisms of humour (and leaving the door open for more to come), Attardo et al. (2002)
appear to be engaged in what the physicist Ernest Rutherford dismissed as
a form of science more akin to stamp-collecting than physics. Likewise,
in attempting to explain too much (see Bell 2002; Gibbs 2000), one can
argue that conceptual blending theory trivializes its subject matter: is
every compositional mental structure to be seen as a blend (albeit, in
many cases, a bad or sub-optimal blend)? Are we to take as a blend any
conceptual structure whose composition can be described by blending
theory, or should we require a more restrictive definition? Unfortunately,
there seems little in the formal apparatus of the theory itself to satisfactorily answer these questions.
A descriptive theory can afford to be agnostic as to the specific processes that yield the description for a given phenomenon, even when competing accounts of these processes are available. An explanatory theory, in
contrast, should commit to just one account, and moreover, should offer
reasons as to why this account is not simply the official account, but a
truly superior account in light of the available facts and data. Conceptual
blending is a mechanism that, when run forward from its inputs, attempts
to explain how a complex conceptual product is created, and when run
backward from an integrated product, attempts to explain how this product is comprehended. Since conceptual blending provides a very detailed
account of how creative products are comprehended, it is tempting to believe that the theory also explains how such products are created. However, one can easily imagine other, simpler and more computationally felicitous accounts, of how such conceptual products are created.
Consider the complex ideas drive-by shooting and date-rape, two
apparently archetypal products of conceptual blending in action. The linguistic form of date-rape directs us to comprehend the concept as an integration of two scenarios (date and rape) that necessitates a tight
mapping of participants (e.g., suitor = rapist, target-of-affection = victim), as does the linguistic form of drive-by-shooting (e.g., driver/passenger = shooter, pedestrian = victim). Each also gives rise to emergent
468
Tony Veale
inferences: drive-by shootings are faster but less accurate forms of attack,
and typically involve higher rates of collateral damage; date-rape is typically harder to prosecute, and may not even be acknowledged as rape by
its perpetrator. From a generation perspective, however, it is surely more
intuitive to think of each not as an explicit integration of distinct ideas,
but as a simple deviation from a single prototype. Date-rape represents a
subversion of the prototypical dating scenario (or script) in which consent
for sexual activity has been edited out, thereby resulting in a scenario that
more resembles the rape scenario. Likewise, a drive-by shooting can be
seen as an edited version of a prototypical hit, in which the attacker
does not leave (or even stop) his car. This single-input subversion account
can also be used to explain the mechanics of many jokes. For instance, an
industrial-drowning script can be subverted into a form that resembles
(and thus recalls) a swimming-pool script, either by editing-out the tragic
conclusion (though this is unlikely to result in humour) or by editing-in
some elements that suggest the tragic event was a pleasurable one.1 Each
of these examples demonstrate that complex products with emergent features can be constructed from a single input structure by assuming a
simple form of internal editing. We should be wary then of cognitive justso stories: we cannot conclude from the fact that date-rape and driveby-shooting (or, for that matter, house-boat, sofa-bed, or any number
of classic blends and jokes) can be described as an integration of multiple
input structures, that they should be so described. Ultimately, explanatory
force arises not from apparent possibility, but from apparent necessity. We
consider how computational concerns might provide this necessity in section 4.2.
3.2. Formal specificity
Though formalization can seem a dry affair, it is a necessary step if a theory is to achieve unambiguous clarity, particularly so if the entities to
which it is ontologically committed are to be specified in a way that
neither relies on metaphor or unarticulated intuition. A theory of creativity, for instance, must be capable of formally separating those processes
and artefacts that are deserving of the label creative from those that are
not. Similarly, a theory of humour must be capable of unambiguously defining a joke. As we have argued, a theory of conceptual blending must
likewise be capable of defining what kinds of construct constitute a blend,
and more importantly, what constructs do not constitute a blend. The
classical approach to such formalization is the use of necessary and suffi-
469
cient feature sets. For instance, Raskin (1985) claims that two features are
individually necessary and jointly sufficient for a text to constitute a joke:
the joke must be partially compatible with at least two scripts (the criterion of script overlap) and these scripts must be opposed to each
other in a particular way (the criterion of script opposition). Neither
script overlap alone (e.g., as found between dentist and doctor scripts) or
script opposition alone (e.g., as found in baptism/life and funeral/death
scripts) is sufficient in this view to produce humour.
Necessary and sufficient features have long been rejected by cognitivists
as a model of human category structure, as they lead to brittle and simplistic structures that are easily invalidated by real world examples. To a
researcher in Cognitive Linguistics, it may therefore seem unrealistic to
build a theory around such features, since theories are categories too and
we would like our theories, like our categories, to be robust in the face of
real world data. For instance, it is commonly held outside Cognitive Linguistics that semantic anomaly is a necessary feature of metaphors (e.g.
see Fass 1988). However, this claim is easily refuted by metaphors like my
lawyer is my bodyguard and my mechanic is a magician. Likewise, it is commonly held, even by cognitivists, that incongruity is a necessary feature of
humour (see Ritchie 2003 for a review), while Veale (2004) argues that incongruity is merely a by-product of a listeners desire to opportunistically
seek humour in a text when socially licensed to do so. Nonetheless,
though the anomaly theory is unattractive as a theory of metaphor on
aesthetic grounds, inasmuch as it tends to over-simplify the phenomenon,
its bold claims are very attractive in Popperian terms. Likewise, incongruity theories of humour are unattractive to some researchers (such as
Veale 2004) for similar reasons, inasmuch as they deny the listener a collaborative role in humour construction, yet they too are attractive for so
readily courting falsification. From a scientific perspective then, a lack of
robustness in the face of real data can be seen as a desirable by-product of
the bold claims made by necessary and sufficient feature definitions, since
bold claims are more easily falsified than weak ones. This scientific perspective has a valuable engineering corollary: since necessity and sufficiency more easily translate into computational form (e.g., as an exhaustive collection of set-theoretic conditionals) than the equivalent system
of potentially incomplete family resemblances, a theory formalized in
this classical way can automatically be applied to a much larger corpus
of potentially falsifying data.
470
4.
Tony Veale
471
472
Tony Veale
473
474
Tony Veale
5.
The more daring or radical a theory, the greater the challenge it can pose
to our sensibilities about what is and is not feasible in practice. But once
again, the substance of this challenge is to be found not in the novelty of
the theory or the sparsity of its assumptions, but in the computational
feasibility of the processes that it entertains. Certain commonplace abilities and processes, such as the ability to pair off like with like from two
different collections, are so familiar as to suggest no computational challenge whatsoever. Nonetheless, such mapping abilities are presupposed by
any conceptual output that is the product of structural matching, from
the integrated spaces of metaphors and blends, to the overlapping scripts
of the GTVH, and the overlapping semantic maps of radical construction
grammar (RCG). Viewed in abstract terms, these abilities become recognizably more complex as one moves from the limited demands of toy
examples to the scale often demanded by real-world data.
475
476
Tony Veale
tempt to cluster exemplars into families of mappings that exhibit the same
deep-structural form. Rather, each exemplar is treated as an island, and
no generalization is formed, for example, from active and passive variations of the same textual proposition. Furthermore, because exemplars
are typically derived via automated statistical alignment techniques from
bilingual corpora, the same exemplars can be extracted many times over,
thus providing a basis for considering some exemplars as providing more
prototypical translations than others.
While EBMT exemplars are mappings from form to form, rather than
form to meaning, their structure (as again judged from the viewpoint of
computational sufficiency) is comparable. Indeed, in no sense do form-toform mappings underestimate the difficulties of dealing with form-tomeaning mappings. Exemplars that are semantically compatible at a
propositional level are frequently incompatible at a form level, due to disagreements in number, case, grammatical role or register. This problem,
denoted boundary friction, means that EBMT exemplars can be extremely
sensitive to the ways they are combined. Since EBMT strives toward a linguistics-lite theory of translation, most systems tackle boundary friction
with statistical or corpus methods, such as establishing the validity or
otherwise of possible translation candidates by determining which candidate occurs most often on the world-wide-web.
To the extent that RCG can also be seen as a rules-lite approach to language, RCG and EMBT are sufficiently similar to ultimately imagine a
role for the semantic maps of RCG in the detection and elimination of
boundary friction, for EBMT and RCG do attempt to model comparable
issues of language use in comparable ways. As such, the engineering practicality of EBMT as a model of machine translation strongly supports the
computational feasibility of RCG as a linguistic theory.
5.2. Complexity Theory in action
As we have seen, both the GTVH and blending theory rely crucially on a
cognitive mechanism capable of mapping two arbitrarily complex mental
structures. In the case of the GTVH, this mapping determines the extent
of script overlap, where scripts are now mathematically conceived in
graph-theoretic terms (see Attardo et al. 2002). In the case of blending
theory, such mappings are central to the identification of correspondences between input spaces (see Veale and ODonoghue 2000). Several
analyses have been made of the complexity of the structure mapping process required for analogical reasoning, and so by extension, conceptual
477
blending and script-based joke analysis (see Falkenhainer et al. 1989; Forbus and Oblinger 1990; Winston 1982). Nonetheless, though structuremapping is intuitively NP-Hard, none of these analyses have the status of
a proof. In this section, we present our own proof and a subsequent analysis to highlight some of the important properties of structure-mapping.
Structure-mapping between two structured representations (essentially
graphs) proceeds by first identifying obvious partial mappings between
sub-structures of each representation. These partial mappings are then
combined to create successively larger mappings until a maximal partial
mapping is generated; such mappings are maximal in the sense that no
other elements can be added without violating the 1-to-1 isomorphism of
the mapping. Though some metaphors and blends may involve many-toone correspondences, as in the blend one-man-band (which maps every
position in a band onto the same musician), most analogical mappings
are intelligible by virtue of being isomorphic (e.g. see Falkenhainer et al.
1989; Veale and Keane 1997). The goal of structure-mapping is to find the
largest maximal partial mapping that is possible between both conceptual
representations. In an ideal situation, every element of each representation is mapped in such a mapping; in other situations, the largest partial
mapping represents a best match of the available knowledge structures.
Defining structure-mapping in graph-theoretic terms allows us to
identify the problem as a rewording of the known NP-Complete problem
LCS, or Largest Common Sub-Graph (see Garey and Johnson 1979):
Analogical Mapping (AM): Given the directed and arc-labelled graphs S =
(SV, SA) and T = (TV, TA) which represent, respectively, the source and
target domains of the analogy, we ask, do there exist subsets SA SA, TA
TA, SV SV, TV TV, with |SV | = |TV| and |SA | = |TA | K such that the
sub-graphs S = (SV, SA) and T = (TV, TA) are isomorphic? Two
graphs S and T are isomorphic if there exists a function f: S T such that
<vi , vk> S iff <f(vi), f(vk)> T .
However, since LCS is a decision problem (returning true or false, rather
than actual structure), it does not provide us with a particularly useful
low-level picture of the computation performed during structure-mapping, which is centred around the aggregation and combination of partial
mappings. The following NP-Hard problem 3DM (see again Garey and
Johnson 1979) yields a clearer picture:
478
Tony Veale
6.
479
Concluding remarks
480
Tony Veale
dane actions like visiting the doctor, eating at a restaurant or being the victim of a mugger. However, this is not the interpretation of the term favoured by a computational reading of the SSTH and GTVH, as evident
from attempts to realize these theories in software. In Raskin (1996), a
script is taken to mean a conventional frame structure, much like the
frames of conceptual blending theory, one supposes, or the unification
structures used to represent constructions in Bryants (2003) implementation of construction grammar. Raskins frames thus comprise a collection of labelled slots, denoting event roles, into which semantic fillers may
be placed. But in Attardo et al. (2002), a script is revised to mean little
more than a generalized graph structure (that now presupposes an NPhard mapping problem rather than a polynomial time unification process).
With this generalization, the GTVH moves further from the conventional
cinematic interpretation of a script toward a completely neutral symbolic
representation. The term script, however, remains, to suggest that the
GTVH still views humour as a by-product of our cognitive faculty for reasoning about events and their social consequences. How else can a social
phenomenon like humour arise out of an abstract operation over graphs?
But however suggestive the term script, the underlying computational distinction needed to support this suggestion is now absent, and we conclude
that the explanatory power of the GTVH is diminished as a result.
The GTVH and blending theory serve as interesting case studies of cognitive theorizing for yet another reason: each represents a separate evolution of Koestlers (1964) influential theory of Bisociation, in which creativity is said to arise at the juncture of orthogonal mental matrices. Both
theories embody different solutions to the issues of specificity that vex
Koestlers original theory, which arise for the most part from Koestlers lack
of a suitably concrete vocabulary with which he might express his claims.
The computationalist influence on the cognitive sciences has since
through its emphasis on specific well-formed representations and processes allowed theories like blending and the GTVH to surmount these
descriptive difficulties. We now possess a rich vocabulary of frames, scripts,
schemata and mappings that each suggest largely the same notion, more or
less, as we move from Cognitive Linguistics to Cognitive Science to Artificial Intelligence and Computer Science. As we have seen, however, there is
still a substantial difference between, on one hand, the rhetorical suggestiveness of such terms, and on the other, their precise theoretical meaning as
seen from the perspective of computational sufficiency. The ongoing importance of the computationalist perspective is to ensure that these terms
are always understood from the latter, formally-precise, perspective. Only
481
Notes
1. Such a subversion is evident in the following joke: A worker at the Guinness
beer factory drowns tragically one day after falling into a giant vat of Guinness. A manager is dispatched to inform the unfortunate mans wife. Did he
die quickly?, she asks the manager. No, he replies, in fact, he got out twice
to use the toilet.
2. For decision problems, which are those that diagnose a given set of inputs
to return a simple true or false, the corresponding class of hardest problems is
called NP-Complete.
3. There is a third option, which is to assume that the brain is, in fact, a non-deterministic machine (exploiting, say, some oddity of quantum mechanical
physics) or that it does not operate on Turing-compatible computational principles. Though possible in principle, this option is still largely the stuff of
science fiction.
4. One is often bewitched by a word. For example, by the word know (from
Wittgenstein 1979)
References
Attardo, Salvatore, and Victor Raskin
1991
Script theory revis(it)ed: joke similarity and joke representational
model. Humor: International Journal of Humor Research 4 (3): 293347.
Attardo, Salvatore, Christian F. Hempelmann, and Sara Di Maio
2002
Script oppositions and logical mechanisms: Modeling incongruities
and their resolutions. Humor: International Journal of Humor Research
15 (1): 346.
Bell, Philip
2002
Blending linguistics and cognition: a methodological critique. In Odense
Working Papers in Language and Communication 23, Anders Hougaard,
and Stefan Nordahl (eds.), 163180. Syddansk University, Denmark.
Bergen, Benjamin, and Nancy Chang
2005
Embodied construction grammar in simulation-based language understanding. In Construction Grammars: Cognitive Grounding and Theoretical Extensions, Jan-Ola stman, and Miriam Fried (eds.), 147190.
Amsterdam: Benjamins.
Bryant, John
2003
Constructional analysis. M.A. diss., University of California, Berkeley.
482
Tony Veale
483
Plath, William J.
1973
Transformational grammar and transformational parsing in the REQUEST system. In Proceedings of the Fifth International Conference on
Computational Linguistics, 585610. Pisa, Italy.
Popper, Karl
1959
The Logic of Scientific Discovery. New York: Basic Books.
Raskin, Victor
1985
Semantic Mechanisms of Humor. Dordrecht: D. Reidel.
1996
Computer Implementation of the General Theory of Verbal Humor. In
The International Workshop on Computational Humour: Proceedings of
the Twelfth Twente Workshop on Language Technology, Jan Hulstijn,
and Anton Nijholt (eds.), 920. (Series TWTL 12.) Enschede: UT Service Centrum.
Ritchie, Graeme
2003
The Linguistic Analysis of Jokes. London: Routledge.
Schank, Roger C., and Richard P. Abelson
1977
Scripts, Plans, Goals and Understanding. New York: Wiley.
Turing, Alan M.
1936
On computable numbers, with an application to the Entscheidungsproblem. In The Collected Works of Alan Turing. Vol. 4, Robin O.
Gandy, and C. E. Michael Yates (eds.). Amsterdam: North Holland.
Veale, Tony
2004
Incongruity in humor: root-cause or epiphenomenon? Humor: International Journal of Humor Research (Festschrift for Victor Raskin) 17
(4): 410428.
Veale, Tony, and Mark T. Keane
1997
The competence of sub-optimal structure mapping on hard analogies.
In Proceedings of the Fifteenth International Conference on Artificial Intelligence, Nagoya, Japan: 232327. San Mateo, CA.: Morgan Kaufmann.
Veale, Tony, and Diarmuid ODonoghue
2000
Computation and blending. Cognitive Linguistics, 11 (3): 252281.
Veale, Tony, Diarmuid ODonoghue, and Mark T. Keane
1996
Computability as a limiting cognitive constraint: complexity concerns
in metaphor comprehension. In Cognitive Linguistics: Cultural, Psychological and Typological Issues, Masako Hiraga, Chris Sinha, and
Sherman Wilcox (eds.), 129155. Amsterdam/Philadelphia: Benjamins
Winston, Patrick H.
1982
Learning new principles from precedents and exercises. Artificial Intelligence 19: 321350.
Wittgenstein, Ludwig
1958
Philosophical Investigations. (Translated by G.E.M. Anscombe). London: Basil Blackwell.
1979
On Certainty (Translated by D. Paul and G.E.M. Anscombe). London:
Basil Blackwell.
484
Tony Veale
Author index
485
Author index
A
Abelson, Richard P. 479, 483
Achard, Michel 14, 16, 316, 331, 339
Ahrens, Kathleen 343
Aleksander, Igor 408, 425
Allbritton, David W. 332, 339
Allwood, Jens 60, 77
Altman, Rick 393, 395
Alverson, Hoyt 269, 293
Amorim, Michel-Ange 139, 141
Andric, Michael 219, 227
Anisfeld, Mosche 231, 258
Arbib, Michael 287, 293
Armstrong, David F. 287, 293, 361,
363, 365, 373
Arnold, Jane 315, 340
Asch, Solomon E. 233, 258
Ascher, Robert 287, 297
Asher, Nicholas 434, 454
Athanasiadou, Angeliki 322, 340
Atkins, Beryl 60, 77
Attardo, Salvatore 12, 16, 463, 464,
467, 476, 478, 479, 480, 481
B
Babarci, Bulcs 338, 347
Bachelder, William H. 277, 291, 293
Baddeley, Alan 306, 340
Bailey, David 131, 141
Bakema, Peter 30, 32, 44, 47
Baker, Judith 315, 340
Baker, Mark 60, 76, 77
Barcelona, Antonio 149, 154, 161,
165, 169, 170, 172, 180, 182, 206,
207, 212, 215, 227
Barcroft, Joe 314, 340
Barlow, Michael 29, 32, 46
Barnden, John A. 11, 432, 435, 439,
441, 444, 445, 451, 452, 454, 455,
457, 459
Barsalou, Lawrence 56, 77
486
Author index
Author index
Deane, Paul D. 32, 46, 64, 76, 77
DeCarrico, Jeanette S. 307, 350
Degreef, Francis 318, 352
Deignan, Alice 38, 46, 212, 311, 315,
325, 334, 344
Delfino, Manuela 435, 455
Demecheleer, Murielle 307, 315, 318,
327, 328, 340, 341
Dent, Cathy H. 231, 258
Dewell, Robert 310, 344
Dewey, John 121, 141, 267, 295
Di Maio, Sara 16, 481
Dez Velasco, Olga J. 157, 185
Dirven, Ren 14, 16, 17, 30, 32, 33, 46,
76, 140, 149, 183, 215, 227, 266, 268,
270, 291, 295, 299, 301, 308, 316,
318, 322, 329, 330, 344, 350, 351,
384, 396, 399, 451, 455
Dodge, Ellen 7, 16, 120, 141, 214, 215,
227
Donald, Merlin 287, 292, 295
Dudis, Paul G. 371, 373
Duffley, Patrick J. 90, 113
Dunbar, Robin 287, 296
Durand, Jacques 383, 396
E
Edelman, Gerald M. 133, 142
Eerden, Bart 389, 396
Ehrsson, H. Henrik 267, 296
El Refaie, Elisabeth 394, 396
Elbers, Loekie 231, 258
Ellis, Nick 306, 345
Emanatian, Michele 435, 455
Emmorey, Karen 363, 364, 366, 373
Emmott, Catherine 405, 425
Enfield, Nick J. 296
Engberg-Pedersen, Elisabeth 366, 373
Ennis, Timothy 316, 342
Epstein, Ruth L. 231, 258
Eubanks, Philip 345
Evans, Vyvyan 55, 63, 64, 76, 79, 97,
113, 310, 318, 320, 345, 354
Eyckmans, June 307, 315, 327, 328,
334, 337, 338, 341, 353
Eysenck, Michael W. 306, 343
487
F
Fainsilber, Lynn 435, 456
Falk, Dean 287, 296
Falkenhainer, Brian 477, 482
Fass, Dan 434, 435, 456, 469, 482
Fauconnier, Gilles 7, 12, 16, 35, 77,
83, 103106, 113, 138, 142, 161, 183,
189, 191, 192, 195, 197, 206, 207,
215, 227, 268, 296, 380, 396, 405,
423, 425, 437, 456, 463, 465, 466,
482
Feldman, Jerome 131, 142, 218
Fillmore, Charles J. 41, 60, 77, 90,
113, 180, 181, 183
Fisher, Olga 311, 345
Flower, John 308, 345
Fludernik, Monika 405, 426
Foley, William A. 268, 291, 296
Forbus, Kenneth D. 477, 482
Forceville, Charles 10, 380, 384, 385,
388, 389, 393, 394, 396398
Fouts, Deborah H. 142
Fouts, Roger S. 139, 142
Frank, Roslyn 16, 30, 46
Freeman, Margaret H. 10, 423, 424,
426
Friedrich, Paul 266, 283, 296
Frishberg, Nancy 368, 373
Frost, Robert 406410, 412, 413, 420,
421, 423, 426
Fussell, Susan 435, 456
G
Gabrys, Danuta 315, 325, 344
Gallese, Vittorio 130, 142, 218,
227
Gamlin, Peter J. 231, 258
Gardner, Howard 231233, 258260,
461, 482
Garey, Michael R. 477, 482
Garrett, Peter 307, 334, 346
Gavins, Joanna 405, 423, 426
Geeraerts, Dirk 1, 4, 16, 27, 30, 32, 36,
38, 44, 46, 47, 49, 60, 78, 122, 138,
142, 148, 183, 265, 270, 277, 296,
453, 458
488
Author index
H
Hahn, Udo 435
Haiman, John 307, 346, 365, 374
Hall, Rogers 438, 456
Hallan, Naomi 59, 78
Hallowell, Alfred Irving 283, 297
Hannan, Peter 315, 346
Haser, Verena 41, 47, 148, 165168,
181, 183, 213, 214, 216, 228
Haugen, Einer 283, 297
Hauk, Olaf 130, 142
Haviland, John C. 283, 287, 297
Haxby, James V. 267, 297
Hayward, Malcolm 392, 398
Heine, Bernd 30, 47, 283, 292, 297
Hekkert, Paul 397
Helmreich, Stephen 437, 439, 455, 456
Hempelmann, Christian F. 16, 481
Hendley, Robert 445, 459
Henkel, Jacqueline 404, 426
Herrera, Honesto 311, 346, 354
Hewes, Gordon 287, 297
Heylen, Kris 36, 49
Heyvaert, Liesbet 38, 47
Hilscher, Michelle 398
Hiraga, Masako K. 148, 162, 183
Hobbs, Jerry 434, 435, 438, 439, 456
Hockett, Charles F. 287, 297
Hogan, Patrick Colm 423, 426
Holland, Dorothy 15, 17
Holland, Norman 413, 426
Hollmann, Willem 90, 113
Holme, Randal 315, 346
Hols, Edith 383, 402
Hopper, Paul J. 30, 47, 292, 297, 365,
374
Horemans, Arnout 315, 337, 353
Horton, William S. 215, 228
Howarth, Peter 307, 346
Huang, Chu-Ren 343
Hllen, Werner 329, 330, 346
Hulst, Harry van der 365, 374
Humboldt, Wilhelm von 266, 297
Hunn, Eugene 283, 297
Hunston, Susan 346
Hutchins, Edwin 138, 142, 276, 297
Author index
I
Ilyin, Mikhail 343
Indurkhya, Bipin 384, 398, 399, 434,
456
Ingram, Robert M. 366, 374
Irujo, Suzanne 346
Isableu, Brice 139, 141
Itkonen, Esa 42, 47
Iverson, Eric 437, 439, 455, 456
Iverson, Jana M. 234, 258, 259
J
Jackendoff, Ray 60, 69, 76, 78, 461,
462, 479, 482
Jkel, Olaf 311, 346
Jakobson, Roman 149, 150, 183, 403,
426
James, Carl 307, 334, 346
James, William 124, 143
Janney, Richard W. 364, 374
Janzen, Terry 9, 362, 365, 366, 371,
374, 375
Jarraya, Mohammed 139, 141
Jensen de Lpez, Kristine 138, 145,
265, 303
Jensvold, Mary Lee A. 139, 142
Jespersen, Otto 287, 297
Johansson, Sverker 288, 298
Johnson, Christopher R. 131, 140,
142, 143
Johnson, David S. 477, 482
Johnson, Mark 124128, 130132,
137, 138, 140, 143, 144, 153, 162,
163, 165, 184, 192, 208, 212214,
228, 265, 267, 268, 270, 282, 286,
298, 299, 307, 311, 315, 347, 348,
362, 367, 369, 375, 379, 380, 395,
399, 405, 408, 424, 426, 439, 457
Johnson, Robert E. 375
Jongen, Ren 402
K
Kaltenbacher, Martin 402
Kandel, Eric R. 306, 353
Kaplan, Stuart J. 384, 399
Kappel, Jenny 307, 341
489
490
Author index
Author index
Mendelsohn, Eve 231, 259
Merleau-Ponty, Maurice 409, 427
Metzler, Jacqueline 129, 145
Mick, David Glen 394, 400
Miller, Christine M. 394, 402
Miller, Christopher 370, 375
Miller, George 470, 482
Mills, Anne 365, 374
Mittelberg, Irene 23, 47
Modell, Arnold H. 408, 427
Monteiro, George 410, 427
Moon, Rosamund 62, 78
Moore, Carmella C. 277, 302
Morgan, John 307, 350
Morgan, Pamela S. 311, 350
Morse, Anna 190, 192, 208
Mortelmans, Tanja 329, 330, 339,
344, 350
Moss, Mallie 435, 456
Mller, Cornelia 385, 394, 400
Murphy, Gregory 212, 214, 216, 218,
228
Musolff, Andreas 13, 17, 311, 316,
342, 350
N
Nagel, Ernst 27, 48
Narayanan, Srini 131, 142, 144, 218,
434, 435, 436, 451, 457
Nattinger, James R. 307, 350
Nelson, Katherine 289, 301
Nerlove, Harriet 233, 258
Niemeier, Susanne 14, 16, 17, 291,
301, 308, 313, 316, 330, 334, 335,
339, 350, 351
Nishimura, Takeshi 288, 301
Nissim, Malvina 435, 457
Noppen, Jean-Pierre Van 383, 402
Norvig, Peter 434, 443, 458
Nth, Winfried 408, 427
Nunberg, Geoffrey 62, 78
Nuyts, Jan 26, 48, 301
O
ODonoghue, Diarmuid 437, 458,
462, 465, 476, 483
491
492
Author index
Author index
Semino, Elena 405, 428
Senft, Gunter 283, 303
Shaffer, Barbara 365, 366, 368, 375,
376
Sharifian, Farzad 274, 303
Shen, Yeshayahu 215, 228, 393, 401
Shepard, Roger N. 129, 145
Shibles, Warren A. 383, 401
Shinohara, Kazuko 389, 401
Shore, Bradd 206, 208, 265, 287, 303,
388, 401
Simons, Jan 389, 401
Sinclair, John 307, 353
Singleton, Jenny L. 364, 376
Sinha, Chris 138, 145, 265, 287, 303
Skehan, Peter 306, 353
Skoufaki, Sophia 326, 353
Slobin, Dan I. 32, 46, 308
Smieja, Birgit 399
Smith, Joshua H. 234, 259
Smith, Ken 383, 401
Smith, Linda B. 256, 259
Skmen, Anita J. 314, 320, 353
Solska, Agnieszka 315, 325, 344
Sorace, Antonella 29, 47
Speake, Jennifer 334, 353
Speelman, Dirk 27, 30, 38, 47, 49
Spence, Charles 296
Sperber, Dan 384, 401
Spitzer, Michael 395, 401
Spivey, Michael J. 23, 47
Spolsky, Ellen 403, 406408, 428
Squire, Larry R. 306, 353
Stallard, David 435, 458
Steen, Gerard J. 212, 215, 228, 388,
393, 398, 401, 423, 426
Stefanowitsch, Anatol 38, 40, 47, 49,
62, 79, 212, 228
Stein, Gees 439, 455
Steiner, George 403, 428
Stengers, Hlne 307, 311, 312, 315,
327, 328, 334, 337, 338, 341, 353
Stern, Daniel 128, 145
Sternberg, Robert 451, 458
Stevick, Earl 306, 353
Stockwell, Peter J. 387, 401, 423, 428
493
494
Author index
Subject index
495
Subject index
A
active zone 56
ambiguity 52
analogical mapping/analogy 189, 191,
197, 202, 220, 221, 232, 437, 438,
464, 476, 477
anthropological linguistics 3, 8, 36
applied linguistics 2, 9, 1214, 111,
305, 306, 332
articulation space 9, 359, 362364,
366, 372
articulation system 9, 359, 360, 363, 369
artificial intelligence 11, 13, 431, 433
atemporal relations 87, 102
ATT-Meta project 11, 434, 439444,
447451
aural mode 367, 382
B
base 159, 177
base space 409, 410, 424
basic color terms 267, 269
blend/blending/blended space 5, 6, 13,
77, 103107, 112, 135, 139, 187206,
268, 289, 360, 369, 371, 380, 403,
405, 407, 411, 412, 423, 424, 437,
441, 461, 465468, 471, 472, 474,
477, 480
blending theory 5, 7, 1014, 81, 83,
103109, 187189, 191194, 197,
215, 216, 380, 465, 467, 471, 472,
476, 478, 480
boundary friction 476
brain structure 5, 7, 119, 120122, 214
British National Corpus/BNC 36,
6365, 6773
C
cartoons 188, 195197, 205, 383, 394
caused motion construction 81, 82,
9196, 106, 111
496
Subject index
Subject index
2325, 2729, 3140, 42, 44, 45, 119,
137, 140, 215218, 226, 305, 316,
325, 332, 334336, 380, 382, 388,
391, 392, 394, 424
empirical revolution 1, 3, 4, 25, 27, 28,
38, 44
entailment 54, 62, 154, 166, 167, 181
ethnoscience 266, 267, 274, 290, 291
exemplar model 25
experiential realism 127
experiment/experimental methods/
experimental research 6, 7, 9, 21,
23, 2629, 3136, 3941, 44, 45,
90, 119, 129, 130, 132, 139, 150,
211, 212, 218, 219, 221, 225,
226, 229231, 242, 283, 287, 305,
317328, 332, 333, 335, 337, 338,
383, 392, 445, 446
experimental design 4, 26, 40, 44, 45
F
factive 403, 411, 416421
falsifiable hypothesis/falsifiability 7,
215, 217, 226, 464, 466
fictive 403, 411, 412, 416421
frame 41, 105, 106, 152, 160162,
178181, 189, 190, 204, 255, 465,
480
frame shifting 465
G
general theory of verbal humour 463,
464
generative grammar 28, 29, 8184,
170
genre 379, 392394
gesture 7, 8, 9, 35, 131, 212, 229231,
233, 235, 236, 247249, 251257,
359, 360, 363365, 372, 435
gesture-speech combinations 229,
234, 236, 252, 253
Google 70, 71, 73, 76, 447
gustatory mode 382
H
humorous blends 188, 195
497
I
icon/iconic/iconicity 6, 9, 10, 147,
148, 152, 162, 178180, 249, 250,
288, 289, 305, 312, 359372,
403, 406409, 413, 415, 419, 421,
423425
idiom/idiomatic expression 4, 51, 62,
63, 6769, 72, 76, 84, 91, 305, 307,
311314, 322329, 334, 337, 338,
416, 465
image schema 5, 10, 119, 120, 122,
124, 127132, 135, 140, 217, 218,
265, 267, 268, 270, 309, 310, 367,
390, 448
implicature 156, 165, 168, 169
index/indexical/indexicality 3, 6,
147, 148, 153, 156, 157, 164, 165,
178180, 364, 407, 411, 413, 416,
417
inferencing/inference 6, 129, 135, 139,
147, 164, 166, 170172, 174, 179,
180, 189, 192, 193, 197199, 201,
212, 214, 218, 224, 389, 438, 440,
443, 450, 468
introspection/introspective 6, 21, 28,
29, 32, 36, 37, 39, 41, 42, 44, 119,
123, 137, 140, 211, 320
intuition 10, 42, 45, 76, 102, 213, 216,
222, 226, 403, 421, 466, 468, 471,
479
inventory 81, 83, 85, 96, 111, 306,
366
J
jokes 180, 188, 198, 205, 331, 464,
465469, 472, 477, 478, 481
L
language acquisition 2, 7, 15, 32, 35,
38, 59, 62, 138, 140, 306
language pedagogy/pedagogical
linguistics 2, 3, 8, 9, 1315, 305309,
313, 315318, 320, 323, 328339
lects/lectal 30, 31, 291
linguistic relativity 9, 85, 180,
263266, 283, 289, 290
498
Subject index
Subject index
R
radial networks 4, 310, 323
radical construction grammar 3, 12,
25, 81, 82, 84, 86, 89, 91, 9395,
97103, 105, 108, 461, 463, 465, 466,
474, 475, 479
reading comprehension 318, 332,
333
reduplication 8, 263, 264, 278280,
282, 286, 290, 292
reference point/reference point
construction 2, 56, 89, 154
referential metonymy 150, 153, 167
resultative construction 81, 82, 91, 93,
94, 106, 111
rhetorically motivated blends 6, 187,
188, 189
S
separate domain 6, 147, 148, 152, 157,
162, 178
sequential scanning 87, 90
serial mixing 441, 442, 450
signed language 2, 3, 9, 10, 35, 139,
359372
similarity mappings 7, 8, 229234,
237, 239, 251, 253256, 268
simultaneous constructions 360, 370,
371
simultaneous cueing 391
single domain 6, 147, 148, 152, 157,
162, 178
source domain 1, 8, 125, 126, 131,
137, 161, 162, 174, 175, 179, 192,
193, 195, 198, 213, 214, 220, 229,
230, 239, 240244, 246248, 250,
251, 255, 256, 311, 314, 316, 322,
327, 328, 333, 334, 338, 363, 367,
380, 384, 385, 387390, 394, 437,
439, 440, 444, 450, 477
source meaning 6, 147, 151, 153, 154,
162, 168, 174, 179, 181
statistics/statistical 2124, 29, 37, 40,
45, 276, 339, 434, 476
499
484
Tony Veale