3

Metaphor and Translation

Studies

3.1. Introduction and overview

It was in 1965 that Catford published his now famous definition of translation, a
formulation that depicted what the translator does as ‘the replacement of textual material
in one language (SL) by equivalent textual material in another language (TL)’ (1965:20).
This definition, and the theory of which it formed a part, were perhaps typical of the
tendency in those days to try to reduce translation to a purely linguistic phenomenon, and
Catford devoted much serious thought to working out a theory that was intended to
account for all aspects of translation in the somewhat mechanistic terms of his definition.
Linguistics has of course moved on considerably since the 1960s, while in the 1980s a
‘cultural turn’ (see for example Malmkjær 2005:36; Snell-Hornby 2006:47-68) took place
in translation studies that introduced many new factors that are now accepted as being of
central importance to the study of translation. The study of metaphor in translation has
quite clearly benefited from this, even if a more significant influence on its development
has perhaps been played by metaphor studies itself.

63
 Chapter Three: Metaphor and Translation Studies

3.1.1. Translation studies: portrait of a discipline

The discipline that is now known as translation studies properly originated in about the
1960s. Up until that time the study of translation had been largely non-systematic and
sporadic, and had focused largely on normative questions of quality, accuracy,
translatability and the correct way to translate, almost invariably with reference to either
literary or religious texts. Much has changed in the last 40-50 years, however, and
translation studies has formed itself into a fully-fledged discipline complete with
theoretical paradigms and a non-prescriptive approach to the investigation of translation.
In line with this, translation is now seen as an act of communication situated within a
particular linguistic and cultural context and influenced, for example, by genre
conventions, audience expectations and the translator’s own manipulative activity,
whether this be unknowing or intentional (see Hermans 1985:11). This is indeed a very
different – and more highly nuanced – vision from that which was generally current in
the 1960s and which lay behind Catford’s definition cited above. Indeed, all of these new
insights are concepts that are so glaringly absent (as it now appears) from earlier
approaches such as that of Catford.

Another important point is that translation studies is not simply contrastive linguistics by
another name, nor is it (any longer) a mere branch of applied linguistics. To understand
one of the main reasons why translation studies is fundamentally different from
contrastive linguistics we must look at a distinction made by the early twentieth-century
Swiss linguist Saussure. According to Saussure, within the general concept of language it
is necessary to distinguish two separate notions, those of langue and parole (Saussure
1986:9-10, 15; see also Holdcroft 1991:19-46; Koller 1992:222-3). This is a distinction
that has been highly influential in twentieth-century linguistics. Langue represents the
theoretical, abstract concept of a language as a rule-governed system, while parole denotes
instances of specific language usage that are produced in response to a particular set of
stimuli. Thus for example formal grammar, lexical semantics and contrastive linguistics
are predicated on langue while discourse analysis and pragmatics take parole as their main
object of study. Translation studies, which focuses on analysing real acts of
communication and on language use within authentic contexts, is also a discipline that is
primarily dependent on the study of parole. Metaphor studies arguably concerns itself with
both but is ultimately interested in building up our understanding of how metaphor

64
 Chapter Three: Metaphor and Translation Studies

functions as part of a language system (i.e. langue) and, indeed, of human psychology. This
fundamental distinction is therefore important for understanding the different focuses of
these two disciplines.

The reason why translation studies cannot be considered to be a branch of applied

linguistics is that it is an area of study that has been deeply influenced by a wide range of
different disciplines. The list of such disciplines is potentially long; Munday, for example,
includes various branches of linguistics, modern languages and language studies,
comparative literature, cultural studies, philosophy of language, sociology, history and
creative writing as having exerted a significant influence (2012:24-5), but in fact this list
could probably be extended still further. In view of the highly interdisciplinary nature of
modern translation studies, caused by the extremely wide range of other disciplines that
have fed into it, it has even become customary to talk of it as an ‘interdiscipline’ (see for
example Snell-Hornby, Pöchhacker & Kaindl 1994).

Metaphor is of course only one of many possible objects of research within translation
studies. In a highly programmatic article originating from the early 1970s, which has
played a very significant role in lending the discipline its current shape, Holmes defines
translation studies as consisting of three branches, the theoretical, the descriptive and the
applied (2004:184-91). Holmes’ overall vision of the discipline is often presented visually
as a map (see for example Toury 1995:10). While purely theoretical work is certainly
being carried out – either on a ‘general’, all-encompassing theory or on more specific
partial ones (2004:186) – most of the research that is not of an applied nature takes place
within the descriptive branch, the aim of which is to investigate translation as both a
product and a process, and also the functions of translation within a given cultural, social
or political context (2004:184-5). The term ‘descriptive translation studies’ (along with the
name of the discipline itself, selected in preference to alternatives such as ‘translation
theory’ or the ‘science of translation’: 2004:182-3) implies a non-prescriptive approach to
researching a wide range of topics, firmly oriented towards the target rather than the
source (Toury 1995:23-39), and committed to the analysis of real instances of translation.
Furthermore, since the ‘cultural turn’ the discipline has greatly diversified and now largely
concerns itself, for example, with translation norms (see Hermans 1999), rewriting and
manipulation (Lefevere 1992), postcolonial contexts (Niranjana 1992; Robinson 1997),

65
 Chapter Three: Metaphor and Translation Studies

gender issues (Simon 1996; von Flotow 1997), questions of translators’ visibility (Venuti
1995) in the texts they produce and the application of narrative theory (Baker 2006) to
acts of translation. Holmes’ map has been extended too, most notably perhaps by van
Doorslaer (2007).

Interestingly, many fascinating insights into translation have been gained through the
deliberate exploitation of old metaphors for translation, or the deliberate introduction
and exploration of new ones. St. André (2010) offers a collection of articles that present a
wide range of these (see for example Tyulenev 2010 for translation as smuggling; see also
Hönig 1997b for translation as bridge-building; Johnston 1999 for translation as
simulacrum; Vieira 1999 for translation as cannibalism and vampirism; Hermans 2007 for
translation as transubstantiation; Evans 1998 for a general discussion).

In spite of all the new directions for research outlined above, there is still plenty of scope
for text-based investigations that focus chiefly on matters of language. What, then, are
the parameters that define this continuing activity, in terms of both the areas studied and
the methodology?

There are in fact a wide range of different areas that form the objects of research.
However, instead of analysing translated material against a particular ‘correctness notion’
(Hermans 1991:166), translation is seen as a process of decision-making that leads to
both loss and gain. There are not considered to be any ‘right’ or ‘wrong’ translations –
just degrees of appropriateness. Mistranslations and mistakes are not of significant
interest to the discipline. In line with this, a move has occurred away from focusing on
translation shifts – that is, minor rewordings – that are caused by small incompatibilities
of grammar and lexis, the situation that we see in Catford’s understanding of the notion
(1965:73-82) and that was presumably caused by the influence of paradigms derived from
contrastive linguistics. Instead, what we see now is a focus on shifts that are optional and
reflect conscious decisions made by translators in an attempt to produce a viable solution
to a real translation problem. In other words, what is now often of interest is the target
context: what the translator chose to say as opposed to what could have been said.

66
 Chapter Three: Metaphor and Translation Studies

If shifts are located on the micro-level (i.e. that of individual words and phrases as they
exist within a text), then there is also much interest within the discipline in what occurs
on the macro-level (the level that involves larger textual structures). Translation studies
no longer places its exclusive attention on studying single words, phrases and sentences
in isolation but now also examines what happens to higher-level entities – such as an
extended section of discourse, a whole metaphorical structure, a line of narrative, an
authorial voice or an entire text – when translated into a different language.

As will be discussed in Section 3.2.5, modern text-based translation studies tends to

distinguish between strategies and procedures. However, care has to be taken not to
attribute to the translator specific internal motivations for particular actions purely on the
basis of external – i.e. textual – evidence. A third notion, that of translation effects, can
also sometimes be used, although this is less common. These would be the results of a
translator’s decisions, rather than techniques used in order to achieve a satisfactory target-
language version.

As regards specific topics selected for investigation within this type of research, first and
foremost is perhaps the nature of translated texts and how they differ from original texts
written in the same language (see for example Toury 1995:207-17), general attention
being frequently paid to finding ways of characterising the language of translated texts.

Within this context, research into the so-called universals of translation does continue,
even if there is much less of a tendency to use this controversial term (see for example
Mauranen 2004, 2008; Ilisei 2012). Here it is worth pointing out that the most widely
researched of these is probably explicitation (or in other words ‘the phenomenon which
frequently leads to TT stating ST information in a more explicit form than the original’:
Shuttleworth & Cowie 1997:55; see also Pápai 2004; Becher 2011). Besides this,
translation scholars investigate a wide range of textual phenomena: not only metaphor, of
course, but also punctuation, lexis, collocation, fixed expressions, word-play, cultural
realia, allusions, semantic prosody (for this last item see Malmkjær 2005:115, 130-1;
Munday 2011). Some features are of course situated on the macro-level, such as for
example non-standard discourse or pronunciation, intertemporal translation, implicature
and genre (see for example Hatim 1996; Hatim & Mason 1997; Malmkjær 2005).

67
 Chapter Three: Metaphor and Translation Studies

One of the most popular methodologies is undoubtedly the corpus-based approach.

Where this is not appropriate, then research is still almost certain to reflect translation
studies’ healthy emphasis on the use of real data. Investigations can be either qualitative
or quantitative, or a combination of the two. Every bi- or multilingual study generally
starts with a monolingual one (i.e. an investigation of the source material). This is
generally followed by an aligning or pairing of source and target items (i.e. source-text
problems and their target-language solutions) for the purposes of analysis. Alternatives to
this paradigm have not yet been seriously investigated.

3.1.2. Translation studies: its relevance to metaphor studies

In spite of the interesting work that was being conducted by translation scholars within
their own discipline, in the mid 1990s Mandelblit stated that the treatment of metaphor in
translation studies was ‘thoroughly at odds with the findings of the Cognitive Linguistics
research on metaphor’ (1995:485). Whether or not this is a totally fair assessment, from
around that time at least some scholars interested in metaphor in translation have sought
to draw increasingly systematically on the insights of this other discipline, recognising the
great scope for further interaction that exists between the two disciplines.

With this in mind, it should be expected that any proper research into metaphor in
translation that has absorbed significant insights from metaphor studies should be in a
position to feed findings and results back into that discipline, even if that is not the
primary aim of the research and such findings and results are only a kind of by-product
of the research. It is true, however, that some of the claims made by translation scholars
as to the centrality of translation to the study of metaphor will probably strike most
metaphor scholars as somewhat overstated. For example, Dagut declares that ‘it is
translation theory that holds the key to a deeper understanding of metaphor’ (1976:32),
and Kurth also makes an assertion that is very similar (1995:120). On the other hand,
Guldin lists three ‘points of contact’ between the two disciplines: the use of specific
metaphors to describe how translation functions, the use of the concept of translation as
a metaphor for exchange and transformation in different types of discourse, and the
question of developing procedures for the translation of metaphors (2010:161-2). To
these he adds a fourth, that of the structural similarities and etymological parallels that
exist between the concepts of translation and metaphor (2010:162), as was discussed in
Chapter One.
68
 Chapter Three: Metaphor and Translation Studies

In addition to Guldin’s valid observations, it is likely that translation studies can bring an
awareness to metaphor scholars of the vast numbers of discrepancies that exist between
languages on the micro-level (and macro-level, for that matter). Indeed, translation
studies can provide many case studies – mono-, bi- and even multilingual – to
supplement the descriptive work of metaphor scholars. (Along similar lines, Malmkjær
argues that translation studies can be a source of data for linguistics: 2005:58-60). More
ambitiously, however, Mandelblit argues that translation examples ‘can provide excellent
data for studying the “hidden” cognitive processes involved in language usage’ (1997:20).

Whether or not this is the case, translation scholars are generally not only bilingual but
also bicultural, and they can provide an excellent source of data regarding interlingual and
intercultural variation to counterbalance the universalist tendencies of the conceptual
metaphor theory that still survive in some parts of the discipline. However, it needs to be
borne in mind that the kind of data that it offers is fundamentally different from that
supplied, for example, by a langue-based discipline such as contrastive linguistics. This is
because translation studies data cannot be relied on to reflect target language norms in an
unbiased manner, contaminated as they are by the influence of the source language, by
the translator’s preference for a particular set of solutions, and so on, all of which factors
place these data firmly in the domain of parole, as argued in Section 3.1.1. In other words,
as instances of mediated discourse examples of translated metaphorical expressions
cannot generally be used as primary data for metaphor studies.

That said, translation-related data can potentially serve as a test-bed for measuring the
validity of metaphor studies categories and concepts: one may reasonably argue that, if
certain factors influence speakers of a particular language (or language in general), then
they must do the same for translators as they make decisions regarding wording and so
forth – whether such factors are universal in nature or bound to a particular language.

3.1.3. Metaphor in translation: an introduction

Since the discussion of metaphor in translation was initiated more than forty years ago by
Kloepfer (1967) much has been written on the topic and a number of important debates
have been pursued. One of the earliest of these concerned the extent to which metaphor
constituted a problem for translators (whether there was ‘no problem’ or ‘no solution’:
see Dagut 1976:25) and the fallout from Kloepfer’s controversial remark that ‘the bolder
69
 Chapter Three: Metaphor and Translation Studies

and more creative the metaphor, the easier it is to repeat it in other languages’ (1967:116;
translation taken from Snell-Hornby 1995:57). Besides this, many writers have proposed
lists of procedures for translating metaphor (e.g. reproducing the original image, replacing
the metaphor with a simile, converting it to sense, etc.; see for example Newmark
1985:304-311, quoted in Section 3.2.5 below). Next, there has been a gradually awakening
awareness that metaphors can play a vitally important structural role in texts rather than
appearing simply as isolated expressions (see for example Crofts 1982 and Stienstra 1993,
and also Section 3.2.3 below).

It is quite true that for many years the study of metaphor in translation was considered to
be of largely peripheral interest within translation studies. For at least fifteen to twenty
years after the birth of the discipline as a recognisable entity at about the beginning of the
1960s very little attempt was made to examine the cluster of problems raised by
metaphor as an isolated phenomenon, or to discuss the place of metaphor within a
broad-based account of translation in general. Furthermore, this situation could be
observed across the board. Articles in scholarly publications were conspicuous largely by
their absence, while few of the seminal texts on translation devoted more than mere
asides to the subject. Thus Vinay and Darbelnet’s Stylistique comparée du français et de l’anglais
(1958), for example, does not discuss the matter to any great extent (see Dagut 1976:21),
while Nida’s Toward a Science of Translating (1964) devotes a mere couple of pages to the
subject. Beekman and Callow’s Translating the Word of God (1974) with its two chapters on
the subject is very much the exception, although it was in fact from the middle of the
1970s that the situation began to change, albeit very slowly. Largely as a response to the
single passage in Kloepfer (1967) referred to above (and discussed in detail in Section
3.2.1), in 1976 Dagut produced a major article on the subject; this was followed by a
number of statements by Newmark on the matter (e.g. 1977, 1980, 1985, 1988a, 1988b).
These articles have proved to be highly influential in the discussion of metaphor in
translation, and have been cited by a large number of authors, many of whom have used
them as the starting-point for further discussion of the subject, either endorsing or taking
issue with one or other of these earlier treatments.

In the last thirty years or so the situation has been rather different, with questions of
metaphor in translation attracting growing interest within the discipline. The articles by

70
 Chapter Three: Metaphor and Translation Studies

Dagut and Newmark were fairly quickly joined by others – most notably those of van
den Broeck (1981) and Mason (1982) – while since that time an increasing number of
papers have been appearing in various translation studies journals, more and more of
which have started to take the interesting work carried out on metaphor in other
disciplines (such as cognitive linguistics) as their starting-point. This steadily increasing
flow has served to keep the subject on the agenda, and in line with this development at
least another three important and widely-discussed works – Gutt (1991), Toury (1995)
and Snell-Hornby (1995) – include passages discussing the area in some detail. Writers on
Bible translation, it should be pointed out, have tended to be relatively prolix in their
treatment of the subject. As mentioned above, Beekman and Callow (1974) fit into this
trend, as does Larson (1984), who also devotes entire chapters to the topic.

In addition, some eight books at least substantially devoted to metaphor in translation

have appeared to date: Kjär (1988), Pisarska (1989), Stienstra (1993), Kurth (1995),
Samaniego Fernández (1996), Gaudio et al. (2003), Zelinsky-Wibbelt (2003) and Salim-
Mohammad (2007). Recent years have also seen the appearance of a steady stream of
dissertations and theses on the subject, including Zahri (1990), Mandelblit (1997),
Erickson (1999), Samaniego Fernández (2000), Al-Harrasi (2001), Holm (2001), Farrell
(2004), Papadoudi (2010), Ureña Gómez-Moreno (2011) and Tcaciuc (2012), with a
number of others currently in progress. (The fact that two of these theses focus on
Arabic-English translation is surely not fortuitous as this is the language pair that is most
strongly represented in the literature.)

It is therefore understandable that these days – and, of course, with the cumulative effect
of the last few decades – writers on metaphor in translation no longer complain that their
subject has always been somewhat neglected. As of February 2013 there are sixty articles
listed in the St. Jerome Publishing Translation Studies Abstracts Online (Harding,
Saldanha & Zanettin 2013) that were published between 1981 and 2012 and contain the
word metaphor in their title. (There are of course many more – 151, in fact – that have this
term as a keyword.) Interestingly, only a minority of the sixty appear to refer to metaphor
theory in their abstracts: fifteen mention the word conceptual, fifteen the rather less
polysemic cognitive, six pragmatic, five Lakoff and two relevance (as in ‘relevance theory’, an
alternative to the conceptual metaphor theory). This indicates that much remains to be

71
 Chapter Three: Metaphor and Translation Studies

done to establish a firm link between translation scholars and the ideas of academics
working on metaphor.

While there are a relatively large number of translation scholars with an interest in
metaphor, there seems to have been considerably less interest shown by metaphor
scholars in translation. In terms of major works, Kövecses (2005:133ff.) offers one of the
few detailed discussions of metaphor in translation by a metaphor scholar, while another
extended treatment is provided by Knowles and Moon (2006:61-72). In addition, there is
a scattering of much briefer mentions, for example in Lakoff (1987a:312) and Sadock
(1993:52, 55). The special issue of the Journal of Pragmatics (issue 36:7) on ‘Metaphor across
Languages’ is for the most part only tangentially about translation. Newmark (1985)
offers one of the first discussions of metaphor in translation to have appeared in a major
work devoted to metaphor rather than to translation, although he has more recently
started to be followed by others (for example, Deignan & Potter 2004; Schaeffner 2004;
Samaniego Fernández 2011). Almost nothing of relevance to translation has appeared in
Metaphor and Symbol.

On the world-wide web there are relatively few significant resources, although a search
carried out on 27 February 2013 using Google did reveal a relatively large number of
references to the topic. The web also hosts a number of articles available in translation
portals and on-line journals, such as Al-Hasnawi (2007), Brevik (2008) and Shabani
(2008).

The formulation ‘metaphor in translation’ is preferred in this study because of the

particular emphasis that it possesses. The more traditional way of referring to this area of
study, ‘the translation of metaphor’, sounds prescriptive and somehow old-fashioned.
‘Metaphor in translation’ places the emphasis on metaphor and locates the discussion of
it within translation studies where we want it – firmly in the broader context of general
metaphor research. The alternative formulation ‘metaphor and translation’ would be
another possibility, and one that is more neutral in terms of the disciplines to which it is
appealing. Interestingly, in the Subject Index to Descriptive Translation Studies and Beyond
Toury refers to this research topic as ‘metaphor in/and translation’ (1995:304).

72
 Chapter Three: Metaphor and Translation Studies

3.2. Themes in the study of metaphor in translation

Now that the discussion has achieved a fair level of momentum, it would perhaps be
sensible to attempt to summarise some of what might be seen as its main themes up to
now. This will prepare the way for us to turn our thoughts to problems that require our
attention and to consider certain possible avenues for taking forward the investigation of
this area. In what follows, the discussion will be organised thematically, although within
each section different authors’ views are considered broadly chronologically.

3.2.1. Kloepfer’s ‘boldness hypothesis’ and the ‘no problem vs. no solution’ controversy
In many ways, it was the brief passage in Kloepfer (1967) alluded to earlier that originally
set the ball rolling. During a discussion of the work of a German translator of Rimbaud’s
Métropolitain, Kloepfer makes the following observation:

Küchler [the translator] manages to preserve all the metaphors: their famous ‘boldness’ is
no problem for the translation – on the contrary, the bolder and more creative the
metaphor, the easier it is to repeat it in other languages. There is not only a ‘harmony of
metaphorical fields’ among the various European languages, there are concrete
metaphorical fields common to all mankind, but there are also definite ‘structures of the
imagination’ on which they are based (1967:116; translation taken from Snell-Hornby
1995:57).

Interestingly, this passage with its talk of ‘metaphorical fields’ and ‘structures of the
imagination’ is in many ways parallel to some of the antecedents of Lakoff and Johnson’s
theory. Indeed, as Dagut points out, Kloepfer’s claim clearly reflects the notion that all
human minds share the same ‘perceptual and cognitive features’ (1976:27). Furthermore,
it possesses a certain attractiveness on a purely intuitive level. There are, no doubt, plenty
of bold metaphors that can be easily translated because of the very mould-breaking
nature of their boldness: what is strange in one language is fairly likely to appear thus in
others.6 However, does this hypothesis constitute a universal principle of metaphor
translation? How far does it go towards providing us with a full-blown model for
metaphor translation?

Kloepfer’s hypothesis is repeated by Reiss (1971:62-3) and, with important qualifications,

by Newmark (1985:314). On the other hand Dagut, who uses this statement of
Kloepfer’s as the starting-point of his first article, is in fact very scathing about it as it

6 What precisely is meant by boldness will be discussed briefly in Section 8.7.

73
 Chapter Three: Metaphor and Translation Studies

stands, arguing that it is ‘striking in its sweeping dismissal of the problem’ (1976:25),
ignoring as it does the existence of the ‘very numerous discrepancies of detail’ between
any two particular languages (1976:27). Dagut concludes that ‘It is only where the
linguistic and cultural components of the metaphor are shared by both SL and TL ... that
the Kloepfer-Reiss thesis can be successfully applied’ (1978:103).

Possible evidence to support Dagut’s contention about limiting the applicability of

Kloepfer’s hypothesis can be found in the example discussed by Hönig (1997a:16-17) of
the phrase ein grauer Zeitbrei being translated in the Financial Times as a grey time-porridge.
According to Hönig, the original German expression, while certainly unconventional, is
nonetheless based on a combination of familiar idiomatic expressions that convey
notions of monotony and lack of differentiation. The literal English translation, on the
other hand, fails to convey the same overtones because not all the English lexical items
lend themselves to these precise figurative extensions of meaning.

Misleading as the hypothesis may or may not be, it nonetheless serves to highlight a wider
debate concerning the question of translatability. Following Dagut’s terminology, the two
approaches that this is centred on are sometimes known as the ‘no problem’ and the ‘no
solution’ schools (1976:25). The first of these probably finds its clearest articulation in
Kloepfer’s statement quoted above. The second, with its implied inbuilt pessimism, is
most closely associated with those authors – mostly coming from outside translation
studies – who take a fastidious view about the possibility of translating poetry, although it
also finds a degree of reflection wherever the recommendation is made that a metaphor
should be translated by a non-metaphor (see Dagut 1976:25) and seems to be held by
Vinay and Darbelnet (1958) and Nida (1964), as pointed out by Samaniego Fernández
(1996:124). Surprisingly perhaps in the light of his intensely polemical stance against the
‘no-problem’ school – a position that is if anything strengthened in his later (1987) article
– Dagut favours a compromise between these two positions, in which the translatability
of a source language metaphor is a function not of its ‘boldness’ or ‘originality’, but of
‘(1) the particular cultural experiences and semantic associations exploited by it, and (2)
the extent to which these can, or cannot, be reproduced non-anomalously in TL,
depending on the degree of “overlap” in each particular case’ (1976:32).

74
 Chapter Three: Metaphor and Translation Studies

Such a compromise position is also favoured by Snell-Hornby, another writer who is

outspoken in her criticism of the ‘no problem’ school:

... the answer lies somewhere between the two poles [i.e. of total translatability and total
untranslatability], depending on the structure and function of the metaphor within the
text. As an abstract concept, metaphor might be universal ...; in its concrete realization,
however, being closely linked with sensuous perception and culture-bound value
judgements, it is undoubtedly complicated by language-specific idiosyncrasies. And
therein lies its fascination for the translator. (1995:61-2)7

Van den Broeck also concludes that bold metaphors are more translatable than
conventionalised metaphors ‘to the degree that they are less culture-bound and are
therefore able to dispense with culture-specific information’ (1981:84, emphasis original). In
spite of these (admittedly common-sense) contentions, Kloepfer’s idea has not received
significant theoretical testing from the point of view of metaphor studies. What factors
do determine the translation procedure? It is in fact one of the aims of this thesis to try to
find a definitive answer to questions such as this.

3.2.2. The definition of ‘metaphor’

Much space has been devoted over the years to discussing what should or should not be
‘counted’ as a metaphor, and by and large authors have tended to focus on creative
metaphors to the exclusion of lexicalised ones. These latter are all those idioms, phrases
and single lexical items (e.g. to kick the bucket, to have a nerve, to sharpen one’s intellect, to
scramble text; also pupil (= part of the eye), tongue (= language) and grasp (= understand)), an
awareness of whose metaphorical origins has been lost by most speakers. Newmark is
happy to take account of the entire gamut of metaphor-like phenomena, and considers
the problems involved in translating what he terms dead, cliché, stock, recent and original
metaphors (1988b:106-12, as already discussed in Section 2.3.8). Dagut, on the other
hand, takes a narrower view, contrasting ‘metaphor’ – which he argues is by definition
‘live’ (1976:22) – with polysemes, idioms and formators (these last being items such as
despite or notwithstanding, which are analysable into smaller semantic units that do not,
however, give an accurate idea of the true meaning of the word). Translation of these

7 The sections of this work that deal with metaphor in translation are a revised version of Snell-Hornby

(1983).
75
 Chapter Three: Metaphor and Translation Studies

latter three categories is, he argues, ‘an area of translation that can be fully “mapped” by
translation theory’ (1976:24) – and is therefore of no interest to him.8

This is a notion that van den Broeck takes issue with, although as we shall see below he
insists that the determining factor is an expression’s ‘functional relevancy to the
communicative situation’ in which it occurs (1981:82). He coins the term ‘deautomatized
idioms’ (1981:83) to describe lexicalised metaphors that take on a special significance in a
particular context because of the clear functional role they are being assigned – as might
be the case in various types of creative writing, for example.

Dagut’s later (1987) article continues with the distinction between metaphor ‘proper’ (i.e.
creative metaphor) and its ‘derivatives’ (polyseme, idiom, proverb: see for example
1987:83). In this article he states that ‘there are metaphors which lie beyond the limits of
translation in the strict sense (excluding any form of periphrastic rendering), and that the
reasons for their untranslatability may be either cultural or lexical or both’ (1987:81;
emphasis original). But why is this perfectly reasonable statement restricted to ‘live’
metaphor without any further discussion? One matter that will be investigated in this
thesis is whether it has any validity for one of the three categories that he excludes (i.e.
polysemes) – or at least, for individual expressions within these categories. Here just as
with ‘metaphor proper’, semantic and cultural associations can intervene, possibly in a
way that will make his claim that these areas can be ‘fully “mapped” by translation theory’
(1976:24) somewhat wide of the mark, as stated by van den Broeck (1981:82). We are
now in fact touching on issues that are going to be thoroughly discussed later on in the
thesis. However, suffice it to say at this stage that the patterning of ‘dead’ metaphors into
domains – and the interaction between different domains that makes all types of semantic
transfer possible – is a significant area that he does not attempt to consider.

In fact, Dagut’s 1987 article, which was published some seven years after the appearance
of Lakoff and Johnson’s Metaphors We Live By, in fact reveals no familiarity with the
important advances in the study of metaphor that this work heralded. By contrast,
Alexieva takes account of research in cognitive linguistics carried out in the 1980s and
1990s to propose a nuanced and well thought-out domain-based approach:

8 This idea is very much expanded in the section on ‘Polyseme’ in Dagut (1978:103-9).
76
 Chapter Three: Metaphor and Translation Studies

... we find differences across languages and cultures, firstly, in the internal structuring of
the various domains of knowledge and experience; secondly, in the interconnectedness
between domains and the way one domain is mapped onto the other; and thirdly, in the
stability of both the inner structuring of the domains and their interconnectedness
(1997:141, emphasis removed).

Although she is ostensibly writing about word play in this article, what she says can
equally well be applied to metaphor – both the original and lexicalised varieties. For her,
the asymmetry that she discusses is both interlingual and intercultural in nature, and takes
in not only the creative, original linguistic units that she is specifically writing about but
also a range of more commonplace types of linguistic usage. In terms of the translation
solutions that she suggests, therefore, what becomes important is to consider meanings
of words or phrases not in isolation but in the context of their ‘cognitive and experiential
implications’ (1997:142). The idea of transfer, interaction or confrontation between the
two different domains – an interrelation that somehow needs to inform the solution that
appears in the translated text – becomes paramount, replacing the notion that a piece of
figurative language is the one-off, isolated result of creative inspiration on the part of the
writer or speaker. This broader view of metaphor is generally taken by scholars whose
approach to metaphor in translation is broadly in line with that of cognitive linguistics
(see for example Mandelblit 1995; Samaniego Fernández 1996; Al-Harrasi 2001; Farrell
2004; Schaeffner 2004; Al-Hasnawi 2007; Tobias 2009; Ding, Noël and Wolf 2010;
Papadoudi 2010; Shuttleworth 2011).

3.2.3. The single metaphor vs. the metaphor network: the nature of metaphor within text
A clearly related matter is the question of whether translation scholars have viewed the
problems of metaphor translation as revolving round finding ways of translating isolated,
one-off metaphors or have been more alert to instances where one single structurally
significant metaphorical idea underlies a whole series of metaphorical expressions, thus
becoming deeply woven into the fabric of the text in which it features.

It has to be said that in the past many writers have fallen into the former camp. Neither
Newmark nor Dagut, for example, appear to be concerned with those instances where
whole patterns of metaphors seem to be sanctioned by one and the same underlying
mapping or where a single metaphor may pervade an entire text. Even for Snell-Hornby
– who talks about the ‘structure and function’ of metaphors within texts (1995:61) – this
is a facet of metaphor in translation that does not feature particularly prominently. This is
77
 Chapter Three: Metaphor and Translation Studies

not to say that many of these writers have not been interested in the textual relationships
that (admittedly individual and free-standing) metaphors enter into, but rather simply that
the focus of their investigations has not been on the way a single mapping can structure
an extended passage or even an entire text.

One of the first writers to emphasise the fact that metaphors can play a vitally important
structural role in texts was Crofts, who with specific reference to Bible translation
suggests the five-way classification of incidental, repeated, extended, semantic and
symbolic metaphors (1982:16). Of interest to us here are the middle three, which between
them cover a range of closely related types of patterning – even though the distinction
between some of these categories is not sufficiently developed in this very brief article.

However, it is another writer on Bible translation, Stienstra (1993), who was one of the
first to argue forcefully that the view that saw metaphors occurring in texts solely as one-
off isolated phenomena is inadequate. Stienstra observes that ‘no attention has been paid
so far to the importance of preserving all the various manifestations of a metaphorical
concept in a particular text’ (1993:216); she argues that since the units studied have been
‘simply inadequate’ it has inevitably led to generalisations of a ‘spurious’ nature being
arrived at (1993:217). Her recommendations for improving this unsatisfactory situation
are as follows:

‘If we want to discuss the translatability of metaphor at a more advanced level than that
of the individual example, we will have to show ourselves aware of the fact that many, if
not most, interesting metaphors form networks which are both systematic and dynamic.
It is only by investigating whether such a network can be transferred to another language,
that we can make progress in answering the question as to the translatability of
metaphor.’ (1993:215)

Whether or not the ultimate purpose of investigating metaphor in translation is to look at

the question of translatability, in the light of all the interesting work that has been
produced within the cognitive branch of metaphor studies within the last twenty years or
so it would be somewhat blinkered not to take this recommendation with the utmost
seriousness: if certain metaphors are repeated throughout a text in a structurally
significant manner – as is manifestly the case in some texts – surely it would make sense
to take this fact into account when studying the phenomena related to metaphor in
translation?
78
 Chapter Three: Metaphor and Translation Studies

Slightly more recently, Hatim and Mason (1997) present an example of an extended
metaphor in the form of a page-long extract from a message by Ayatollah Khomeini, in
which the late Iranian cleric makes repeated use of the imagery of blood in order to
exhort the staff and students of religious seminaries to higher levels of devotion,
militancy and vigilance (1997:148-150). Presented as what Venuti (1995) would term a
‘foreignising’ translation, the translation basically preserves each source text instance of
the blood metaphor. Hatim and Mason focus on the fact that this ensures that an
important cohesive chain is maintained in the target text (1997:151); however, for our
purposes it is also of interest that in this instance the translator elected to retain certain
striking metaphorical expressions (such as ‘soaked in blood on the pavements of bloody
events’) which in the target text have become divorced from a particular mapping that
lent them special significance for the text’s source language readers.

3.2.4. Metaphor and text-type

Another major issue is that of the text-types in which metaphor is supposed to play an
important role. Hardly surprisingly, much of the interest in metaphor in translation has
been concentrated on literary texts, with many writers – such as Dagut (1976), Mason
(1982), Alvarez (1993), Kruger (1993), Fung (1995), Kurth (1995), Monti (2006), Fouces
González (2007), Jay-Rayon (2007), Maalej (2008) and Tobias (2009) – focusing on this
particular area. In line with the more traditional view of literary metaphor some but by no
means all of these writers have tended to concentrate on original, one-off metaphors.
After these literarily-oriented approaches, the second major focus of interest has been
religious texts. A large number Bible translation scholars have written extensively on the
subject, including for example Beekman and Callow (1974), Crofts (1982, 1988), Gutt
(1991), Larson (1984) and Stienstra (1993). Zahri (1990) has produced a major study of
metaphor within the context of translating the Quran, and other scholars such as Faiq
(1998) and El-Zeiny (2011) have also written on this topic. As hinted above, to the extent
that it is possible to generalise, it has been among these scholars more than anywhere else
that the awareness has arisen that metaphors are more than just single, isolated
phenomena that result from an author’s one-off acts of inspiration.

If one is to mention a third area that has attracted the interest of scholars, then it is
probably that of politics, on which both Schaeffner (1996, 2004) and Al-Harrasi (2001)
have focused their attention – Schaeffner with discussions of the translation of EU
79
 Chapter Three: Metaphor and Translation Studies

documents, and Al-Harrasi with an analysis of official Omani translations of speeches by

the Omani Head of State Sultan Qaboos bin Said and of translations of speeches by
Saddam Hussein carried out by the translation service of the CIA. As for other text-types
and subject areas, van Vaerenbergh (1996) considers economics texts, Pisarska (1989)
covers a wide range of popular scientific and other non-literary texts, while van den
Broeck (1981) devotes a fair amount of space to a discussion of the relative translatability
of metaphor in literary and scientific texts, concluding that metaphorical expressions in
the latter type of text do not usually constitute much of a problem for the translator.
Metaphor translation in scientific discourse has been tackled by Stambuk (1998),
Temmerman (2002), Vandaele (2002), Scarpa (2004), Faber and Márquez Linares (2005),
Shuttleworth (2011) and Vandaele and Béland (2012). However, the relative lack of
interest in non-literary writing that we see in most authors clearly goes hand in hand with
the emphasis that is frequently laid on ‘creative’ metaphor, and there seems to be a
widely-held view that the translation of metaphor in technical or scientific texts does not
constitute a problem. This belief is caused no doubt by the assumption that, since such
texts do not usually contain large numbers of original, one-off metaphors, the translation
of such metaphorical expressions as are present is not a matter that is likely to draw
heavily on the translator’s creative resources. Van den Broeck (1981) at first glance
appears to adhere wholeheartedly to this opinion. Distinguishing between ‘creative’ and
‘non-creative’ metaphor, he states flatly that ‘in the former metaphor as such is
functionally relevant, whereas in the latter it is most likely not to be’ (1981:78, emphasis
original). He then goes on to argue that

The extreme positions are occupied by literary language on the one hand, and the
language of science on the other hand. ... In scientific discourse bold metaphors are very
unlikely to occur. Lexicalized ones will of course be unavoidable, but it goes without
saying that for a translator there is no problem in rendering them (metaphorically or non-
metaphorically) (ibid.).

As it stands this contention – which incidentally totally contradicts that of Kloepfer –

does not seem to give serious consideration to the possibility that the translation of
metaphor in scientific texts might constitute a problem. Later on in his article, however,
van den Broeck does go some way towards moderating this somewhat categorical
statement. Building on the important notion of functional relevance – which he argues is
not necessarily synonymous with ‘boldness’ (1981:76) – he states that the treatment of

80
 Chapter Three: Metaphor and Translation Studies

lexicalized metaphors ‘will entirely depend on their functional relevancy to the

communicative situation in which they occur’ (1981:82). Thus for example, in certain
contexts (this happens most typically in poetry) an idiom such as to flog a dead horse can
assume great functional relevance by having both its literal and its more usual, figurative
meaning simultaneously activated.

However, van den Broeck still considers the translation of items of this type when they
occur in ‘non-creative language’ simply as a matter of consulting a suitable dictionary for
‘either a corresponding polyseme or idiom, or an equivalent non-metaphorical
expression’ (1981:82). Pisarska (1989) takes rather a different view of the question of
metaphor in technical texts. Rejecting most of the categories into which metaphors have
been variously divided, she suggests that both original and lexicalised types ‘may pose
similar problems in translation’ (1989:57). She also observes that the presence of
metaphorical expressions in non-literary texts is ‘undeniable’ and argues that, if only for
the sake of obtaining a complete picture, the translation of conventional metaphor in
such texts deserves to be studied (1989:1).9

In the case of most non-literary text-types (including most popular scientific texts –
which have a reasonable chance of including fairly well-developed, striking metaphors,
for example for illustrative purposes) it is to be anticipated that a fairly ‘literal’ translation
strategy that preserves the metaphor is the norm, or indeed the default position. Indeed,
writing on the basis of her analysis of the books in her corpus, Pisarska states that ‘all
examples of metaphorical uses found in the examined texts are those of lexicalized
metaphors which have lost their unique, poetic character and have become part of the
established semantic stock of the language’ (1989:68), and concludes that the degree of
translational creativity here is low (1989:98). Consequently, any occasion on which the
translator feels unable to render a metaphorical network in broadly ‘equivalent’ terms
should be deserving of our very serious consideration.

In spite of the conclusions that Pisarska draws on the basis of her study, the present
work takes as its starting-point the conviction that metaphor plays an at least potentially

9 When she uses the term non-literary Pisarska is actually referring to a corpus of seven books relating to the
fields of history, anthropology, economics and psychology.
81
 Chapter Three: Metaphor and Translation Studies

significant role within popular scientific discourse, given the way in which metaphorical
conceptualisations can model or even channel thought in a sometimes covert manner
(and to an as yet unquantified extent). Probing the effect that translation can have on
significant networks of conceptual metaphors within texts is of course one of the aims of
this thesis.

3.2.5. Translation procedures

Van Doorslaer (2007:226-7; see also Munday 2012:22-4) typifies the manner in which the
terms ‘strategy’ and ‘procedure’ are used in contemporary translation studies by indicating
that the former refers to the general approach to translating a text that is adopted by a
translator (e.g. literal translation, idiomatic translation, etc.) while the latter is reserved for
a means of rendering a specific textual feature or a solution to a particular problem (e.g.
compensation, implicitation, etc.). It is with translation procedures – the precise shifts
that metaphorical expressions undergo during the translation process – that this thesis
concerns itself. A translation ‘effect’, a third but less frequently used concept, is simply a
change in the text – whether global or localised – that has been brought about during the
translation process. However, in the interests of simplicity this term is not used in the
thesis, such changes being referred to as ‘procedures’, whether or not they are
consciously opted for by the translator.

It would of course be possible to produce a highly detailed taxonomy reflecting most or

all of the many shifts identified by van Leuven-Zwart, including shifts between the
general and the specific, the abstract and the concrete, and the objective and the
subjective, as well, for example, as ‘explicitation, implicitation, amplification, reduction,
addition, deletion, intensification and archaization’ (1989:153). Each of these would,
potentially, take the research off in a different direction. The classification proposed in
Chapter Seven is intended as a simple system, a starting-point rather than an end point.

As evidence for the need to study metaphors in all contexts in which they occur, Pisarska
discusses the various ways in which translators have tried to handle them. Talking
specifically about the type of conventional metaphor that one typically finds in non-
literary texts, she states that

82
 Chapter Three: Metaphor and Translation Studies

The numerous possibilities of rendering these seemingly non-problematical expressions

are signalled by the number and variety of procedures translators resort to in the
renditions and their value for the idiosyncratic style of writing of a particular author is
confirmed by the fact that in the vastly prevailing number of cases they are retained in
the translations. (1989:1)

Pisarska also describes an ‘ad hoc survey’ (1989:63-4; emphasis original) in which fifteen
English expressions drawn from Lakoff and Johnson (1980:7-8) that illustrate the
metaphor TIME IS MONEY were translated into seven languages (three of which were not
Indo-European). Interestingly, it reveals a remarkably high level of similarity between the
various languages. Equally interesting, however, is the fact that one expression, Is that
worth your while?, is not rendered directly in any of the seven languages, suggesting that
even if there is general cross-language agreement as to the acceptability of a conceptual
metaphor there are always likely to be differences of detail from one language to another.
As far as I am aware, this is a ‘one-off’ experiment that does not form part of a detailed
large-scale comparative investigation, but the results are nonetheless thought-provoking.
Another experiment from the 1980s (Snell-Hornby 1995:59-60), conducted this time with
the researcher’s own students, focuses on the simile chosen by groups of German and
English speakers to describe the image metaphor set out in Figure 3.1 below, which first
appeared in an article in the Neue Zürcher Zeitung (7 April 1978), where it was intended to
depict how Catholic and Protestant communities in Belfast are arrayed round the city
centre and separated from each other by the city’s main streets:

Figure 3.1: The arrangement of Catholic and Protestant communities around the centre of
Belfast, taken from Snell-Hornby (1995:59) (K = Catholic, P = Protestant, Z = centre)

83
 Chapter Three: Metaphor and Translation Studies

The original German sentence was supplied to participants with the word used to
describe the arrangements of the streets removed. According to Snell-Hornby, both
groups answered ‘without hesitation, and almost unanimously’ (1995:60), the German
speakers correctly supplying the original word strahlenförmig (‘in the shape of rays’) and the
English speakers seeing the diagram as resembling the spokes of a wheel. In spite of the
English group’s near-unanimous opinion as to what felt ‘right’ in this context, nearly all
of the over 150 students who translated this text opted for the literal ‘like rays of the sun’
or ‘like rays of a star’ as the translation for this word (ibid.), thus indicating the inherent
tension that exists between what native speakers consider to be the correct situational
equivalent in their language and what often ends up appearing in a translation because of
the influence of the source text (ibid.).

Much effort has also been devoted over the years to constructing classifications to
account exhaustively for the procedures that translators employ when translating
metaphorical expressions. Large amounts of space are not available to discuss this area
and so one example will probably suffice. Newmark’s classification of the procedures that
translators use to translate what he terms stock metaphors is very representative of the
type of scheme that has been proposed. In all he identifies eight procedures, which he
lists in order of preference:

1. Reproducing the same image in TL, ‘provided the image has comparable frequency
and currency in the appropriate register’
2. Replacing the image in SL ‘with a standard TL image which does not clash with the
TL culture’
3. Translating the metaphor by a simile, ‘retaining the image’
4. ‘Translation of metaphor (or simile) by simile plus sense (Mozart method).’ (I.e. the
addition of explanatory material: for Mozart, a piano concerto had to please ‘both
the connoisseur and the less learned’.)
5. Conversion of the metaphor to sense
6. Modification of the metaphor
7. Deletion
8. Using the same metaphor combined with the sense
(1985:304-311)

What such lists try to achieve does not coincide with the aims of descriptive translation
studies as they are essentially prescriptive. Similar schemes have been proposed by a
range of authors including Beekman and Callow (1974:145-9), Larson (1984:254),
Pisarska (1989:76) and Alvarez (1993:484-7). However, as a result of the considerable

84
 Chapter Three: Metaphor and Translation Studies

work already devoted to producing classifications of this type, this area is probably quite
well charted (with the proviso made in the previous paragraph); few if any further
significant strategies have been proposed, any minor modifications being little more than
footnotes to Newmark’s scheme. The one significant exception seems to be Toury’s
‘prospective’ view of seeing metaphor as a solution rather than a problem, which
produces two more possible configurations: translating a non-metaphor by a metaphor,
and adding a metaphor in the target text when there is no ‘linguistic motivation’ in the
source text (1995:83). With these additions, though, it would seem that this approach has
been taken about as far as it can be in terms of the major procedures that such lists
attempt to classify – although they can never be exhaustive as it is always possible to find
further one-off, special situations that are not accounted for.

The other type of procedure list that should be mentioned is more descriptive in nature,
and is firmly grounded in metaphor theory as well as translation studies. One of the
earliest examples was produced by Mandelblit (1995), who observes that target-language
translations of metaphorical expressions can conform to either the ‘Similar Mapping
Condition’, in which case source and target use ‘the same general metaphorical mapping’
or to the ‘Different Mapping Condition’, where different ones are used (1995:491).

Along similar lines, another list of procedures that is worth mentioning is that of Al-
Harrasi (2001:277-88). This list is radically different from all its predecessors. What is new
about his taxonomy is that it is deeply informed by conceptual metaphor theory, hinging
more specifically on the interplay between image schemas and rich images:

1. Instantiating the Same Conceptual Metaphor

1.1. Same Image Schematic Representation
1.2. Concretising an Image Schematic Metaphor
1.3. Instantiating in the TT only a Functional Aspect of the Image Schema
1.4. Same Image Schema and Rich Image Domains
1.5. Same Rich Image Metaphor but Alerting the Reader to the Mapping
1.6. Using a Different Rich Image that Realises the Same Image Schema Realised by the
Rich Image in the Source Text
1.7. From the Rich Image Metaphor to Image Schematic Representation
1.8. Same Mapping but a Different Perspective
2. Adding a New Instantiation in the Target Text
3. Using a Different Conceptual Metaphor
4. Deletion of the Expression of the Metaphor
(2001:277-88, with a slightly different version of this list being given on pages 288-9)

85
 Chapter Three: Metaphor and Translation Studies

As far as I am aware, this is one of the most detailed attempts to date to produce an
alternative taxonomy based on metaphor theory, and one that opens up a whole new
possible direction for research. (Interestingly, Al-Harrasi also talks about ‘procedures’
even though what he lists are probably not actions that are consciously implemented by
translators.) His systematic pairing of image schemas and rich images has an intuitive
appeal in many respects, although one possible criticism relates to its tendency to over-
generalise the concept of the image schema (or maybe it is simply that cases that do
involve an image-schematic metaphor have been specified in much greater detail than
those that do not). Image schemas do not underlie all conceptual metaphors, as Al-
Harrasi indeed accepts (see 2001:277), and as will be demonstrated by the present work;
indeed, it is likely that most metaphorical expressions involve some kind of rich image. In
addition, even when an expression in a text is ultimately derived from an image schema
there is often a complex ‘network of intermediate mappings’ that intervenes (Evans &
Green 2006:301).

Al-Harrasi’s scheme has now been joined by other noteworthy examples, such as those
by Schaeffner (2004:1267) and Papadoudi (2010:279). All of these works demonstrate a
clear awareness of the distinction between micro- and macro-levels. Most of them also
define a set of theoretical parameters to which they pay particularly close attention in
their analysis of metaphor in translation and on the basis of which they propose a list of
translation procedures. Along similar lines, the present work will also investigate the
interplay between parameters and procedures.

3.2.6. Issues of culture

The study of culture is of no greater or less intrinsic interest or complexity than that of
language. Furthermore, how the two concepts interact and intertwine with each other is a
matter of great intricacy and has yet to be fully understood (although it is being probed in
detail by writers such as Wierzbicka 1992, 1997 and Goddard 2006). In the literature
questions that properly fall within the domain of language are sometimes incorrectly
attributed to cultural factors (as argued, for example, by Dobrovol’skij and Piirainen
2005:210), and presumably also vice versa. As has already been made clear, the emphasis
of this thesis is on language, although questions related to culture cannot of course be
ignored.

86
 Chapter Three: Metaphor and Translation Studies

The concept of culture and its influence on translation is very broad. Katan, for example,
sees it as a wide-ranging notion that affects the translation process on many different
levels. Cultural background can determine a person’s basic communicative stance (e.g.
use of understatement, directness of expression, etc.: 1999:210-41). The appropriateness
of different levels of formality and distance can also vary from culture to culture
(1999:177-209), as can the extent of the use of nominalisation as a basic means of
organising information in sentences, for example (see 1999:113-4). Individual items of
lexis can have significant cultural dimensions (1999:151-2). (This is a matter that
Wierzbicka 1997 examines in great detail.) In addition, there are six hierarchically-
arranged ‘logical levels’ within which all biological and social systems can function
(1999:37); these are environment, behaviour, capabilities, values, beliefs and identity
(1999:45-61). Finally, as stated in the previous paragraph, there is the whole question of
how language, thought, culture and reality relate to each other (1999:74-5). This raises
important questions of linguistic universalism or relativism, at one end of which we see
views championed by scholars such as Pinker, who vigorously denies the possibility of
any link between the language that a person speaks and the way in which he or she views
the world (see for example 1994:58), and at the other the so-called strong version of the
Sapir/Whorf hypothesis, according to which we are completely at the mercy of the view
of reality that is imposed on us by our native language (see for example Whorf 1956:213).
The very existence of translation as a real-world activity is of course a significant
argument against both such extremes.

Katan does not consider the question of metaphor in detail. However, two specific points
need to be made. Firstly, the question of interlingual metaphor variation needs to be
addressed in the light of the relatively universalist view regarding how metaphor exists in
different languages that has been adopted by many metaphor theorists. To counter this,
Gibbs presents metaphor as networks of cognitive webs that ‘extend beyond individual
minds and are spread out into the cultural world’ (1999b:146). According to him,
‘cognition arises, and is continually re-experienced, when the body interacts with the
cultural world’ (1999b:162). Gibbs concludes by acknowledging ‘the culturally embodied
nature of what is cognitive’ and suggests that ‘there is much less of a difference between
what is cognitive and what is cultural than perhaps many of us have been traditionally led
to believe’ (1999b:162).

87
 Chapter Three: Metaphor and Translation Studies

Along similar lines Kövecses (2005) sees metaphor as existing in culture at the interface
between universality and variation: ‘universality’ because of the cognitive source of much
metaphor (2005:17-64), and ‘variation’ because of the alternative metaphors that exist in
different cultures (2005:67-87) and the variability of metaphor within individual cultures,
for example for diachronic, social or stylistic reasons (2005:88-113). Talking about
figurative language in general, Dobrovol’skij and Piirainen also arrive at a very similar
conclusion:

… the real behaviour of a given idiom is not governed by knowledge of a general kind
(e.g. by knowledge about containers), but by more concrete, culture-specific knowledge.
… At the level of the CONTAINER metaphor, all expressions for ANGER are identical.
However, they differ from each other in their linguistic behaviour: they may show
different restrictions of use in one language, or, from a cross-linguistic viewpoint, they
may not be considered full equivalents. (2005:130)

In fact, in terms of sheer numbers, a view of metaphorical language that accommodates

cultural variation now appears to be the predominant one, even if some of the most
dominant voices still hold a much more universalist position.

The second point that needs to be made with reference to metaphor in culture regards
the cultural turn in translation studies. In line with much of the other work done on
metaphor in translation this thesis is more deeply influenced by metaphor studies than by
most of the new ‘culture-oriented’ approaches to translation listed in Section 3.1.1.
However, in spite of its basically linguistic orientation this thesis does not seek to turn the
clock back to the early 1980s. Where necessary the target context is taken into
consideration and translation variants not opted for are given due consideration. In
addition, where opportunities present themselves conclusions that go beyond the purely
linguistic are drawn (see for example Section 9.1.3.2).

This section has served partly as a kind of justification of the project in the light of some
of the more ‘universalist’ metaphor theory on the one hand, and in that of the cultural
turn on the other. In view of all that has been discussed, we can delineate the corpus of
metaphorical expressions that this work analyses as being determined by a combination
of relatively universal cognitive factors, the ‘accumulated cultural experience of the
members of [a] language-community’ and the ‘“institutionalized” semantic associations of

88
 Chapter Three: Metaphor and Translation Studies

the items in their lexicon’ (Dagut 1976:32). In other words, the influence of culture is
deemed to be quite considerable in this area.

3.2.7. The question of forming a separate ‘theory of metaphor translation’

A number of writers have maintained that it would be desirable to treat metaphor in
translation as a separate theory. Thus Dagut for example argues that ‘a theory of
translation (like any linguistic theory) compares language systems (on the level of
“competence”), not individual instances of “performance”’ (1987:82, emphasis original). I
would argue, however, that this is a very peculiar statement to make, given the preference
of the discipline for matters of parole (broadly equivalent to the ‘performance’ in Dagut’s
argument) rather than langue (corresponding to ‘competence’) that has been identified
above. Dagut goes on to state that ‘[it] is precisely the unique system-violating character
of metaphor that sets it apart from other phenomena of language and therefore requires a
special theory to account for its translation’ (1987:82), although what the basis for this
special pleading might be is not entirely clear.

Pisarska is also supportive of a separate theory for metaphor translation, although she
concedes that theorists are not currently in a position to formulate one (1989:115-6) – or
at least, not at the time that she was writing. Al-Harrasi is not only a proponent of a
separate theory, but even of a distinct discipline (2001:264-9), which he proposes naming
‘MiT [i.e. Metaphor in Translation] Studies’ (2001:267).

On the other hand, Mason – against whom all of these three scholars react – states that
‘[t]here cannot be a theory of the translation of metaphor; there can only be a theory of
translation, and that theory has to allow room for the notion of the purpose of translating
each new text’ (1982:149). Looked at in the terms set out in Section 2.3.5, translation is
probably a ‘basic level’ concept and therefore is a more natural level on which to pitch a
theory, and certainly an entire discipline. The question will be returned to in the course of
the discussion but at this stage it appears that metaphor in translation is simply too
specific for one to make it the object of such an initiative, and doing so would go against
the general trend towards integration and the formation of an interdiscipline that can be
observed over the last twenty years or more within the discipline (see Snell-Hornby 1995,
2006; Snell-Hornby, Pöchhacker & Kaindl 1994). Moreover, metaphor is in many ways
comparable to a range of other phenomena, such as word play, allusion or cultural items,
89
 Chapter Three: Metaphor and Translation Studies

and I believe it would lead to too great an atomisation of the discipline if each were to
become the subject of its own separate theory. On the other hand, in the terms of
Holmes’ map that was discussed earlier in the chapter, metaphor in translation clearly
forms the subject of a partial theory, although in this case its raison d’être would be to feed
into the larger discipline.

90