See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/331286451

Text and/versus Discourse

Chapter · January 2015

CITATIONS READS

0 6,696

1 author:

Nelly Tincheva
Sofia University "St. Kliment Ohridski"
27 PUBLICATIONS   19 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Text, Discourse, Prototype View project

Linguistic signalling of text world and discourse world construction View project

All content following this page was uploaded by Nelly Tincheva on 22 February 2019.

The user has requested enhancement of the downloaded file.

 Chapter Two
TEXT AND/ VERSUS DISCOURSE

As professed above, one of the main purposes of targeting the notion

of text structure here is achieving a better understanding of the notion
of text. As also indicated above, text, too, is one of those ‘commonly-
known’ words (not simply terms) which tell you so much and so little at
the same time. To optimize (even if hoping to actually maximize) infor-
mation input, let us start with a brief overview of what the notion and the
term mean and how they evolved into what they are at present.

II.1. Text: Evolution

Historically speaking, the word text is most frequently taken (see
e.g. Viljamaa, 2007; Winkler, 2012: 25) to have been first used to refer to
a piece of writing by Quintilian in his Institutio Oratoria deliberation on
public/ political speeches (as translated by Butler in 1920), where Quin-
tilian discusses textum (the Latin for fabric) as words weaved together
into a fine textile. With respect to the emergence of the term, it also seems
worth noting that the very word textus (m; genitive textūs) derives from
the Proto-Indoeuropean *teks- (to plait, to weave, to braid, to join, to fit
together, to construct, to build, to make wicker framework), which could
also be interpreted as structure, connecting and construction.
The modern term text, however, proves to be a much later replace-
ment for what, throughout the centuries, used to be called the work and
that shift in meaning came as a result of the fact that, in early structuralist
times, ‘written works’ had already come to be invariably authored and
signed (ibid.). Truly, the emergence of the linguistic term text in those
(basically structuralist) times may have happened in full awareness of
the Proto- uses of the word, or, then again, it may have not. What is in
focus here is the very senses of the words chosen originally by Quintil-
ian: weave and fabric – senses which, quite importantly, also shared
the same conceptual basis with the words structure and construction.
Undeniably metaphoric, those earliest perception(s) of text are strongly
suggestive of people’s intuitive grasp of what a text is – both in ancient
times as well as in modern ones.

12
 Chapter Two. Text and/versus Discourse

Structuralist times, however, not only brought about the modern-

day term of text. They also witnessed the rise and rule of an unbending
linguists’ desire to not accept text as a ‘proper’ linguistic notion. Lin-
guists (see e.g. Bloomfield 1933: 170) were adamant about postulat-
ing the sentence as the largest unit with inherent structure. Seen from a
slightly different angle, as well-known, the dominant 20th c. structural-
ist theory used to interpret language as a system of minimal units and the
task of the linguist was perceived to only consist in the comprehensive
identification and classification of those minimal units at ‘levels’ not
higher than the ‘sentence level’. Even more simply put, the linguist was
not meant to gaze beyond the limits of the sentence.
It wasn’t until the question of whether a true understanding of com-
munication could be attained at all, if linguistics keeps defining itself as
the study of phonemes, morphemes, lexemes and syntactic structures
only became unavoidable that text started to attract linguistic attention.
As van Dijk (1995: 389) writes, it was not until the early 1970s that
the realization that many unsolved issues in linguistics can only be ac-
counted for through a theory of texts started to gain ground. In 1974,
for instance, Kintsch asserts that problems in linguistics are bound to
keep arising because “linguists write sentence grammars instead of text
grammars, philosophers analyze isolated sentence examples, and little
psychological work has as yet been done with complete texts and proper
contexts” (1974: 11). In a similar vein, Werlich remarks that “sentence
grammars do not tell […] the whole story about communication by
means of language” (1976: 14). Still later, Werth launches a similar at-
tack by claiming that
Linguistics, led by its flagship the Generative Enterprise, is heading for
the asteroid belt. It is travelling in ever decreasing circles, using more
and more complex devices to talk about smaller and smaller fragments of
language. (Werth, 1999: 19)

Prior to the onset of such views, anything beyond the sentence was
assigned to other domains, most often Literature and Stylistics. Prior to
that point, the science which, logically, should have been most closely
associated with the study of texts had practically divorced itself from the
very concept it had created.
Non-linguistic approaches to text (such as Sociology, Anthropol-
ogy, Philosophy and Pragmatics), however, did not limit themselves and

13
TEXT STRUCTURE: A WINDOW INTO DISCOURSE, CONTEXT AND MIND

did take interest in textual communication and text-based social prac-

tices (e.g. Malinowski, 1923; Austin, 1961; Hymes, 1962, 1964, 1967;
Sacks, 1963; Labov, 1966, 1972; Searle, 1969; Grice, 1975). Some of
the questions those approaches to text raised could easily be claimed
to be both ancient and quite modern. As will be discussed in the next
Section (II.2.), such questions are: Can text be studied successfully in
isolation from its environment (most often seen as and termed context)?
What structural patterns underlie language in use? Is text written or spo-
ken? Is text monologic or dialogic? Historically speaking, those ques-
tions would later be ‘inherited’ by the newly-emerging Text Linguistics
and Discourse Analysis. The answers to them, however, have still not
been generally agreed upon. Those questions, therefore, will also be in
focus in our discussion here (II.2.1. below).
When the time came for Linguistics to decide to embrace text as
a proper subject of its investigations, it did so in quite a predictable
way – linguists sought to take knowledge of the principles and opera-
tions which apply to and control the ‘lower levels’ of language and just
transfer them to the ‘higher level’ of text. The principle of basic unit
composition, for example, left its mark even on the pioneering work
of Zellig Harris (1952) and Katz and Fodor (1963), who would argue
that a text is to be treated as one long sentence put together simply
by using different punctuation. In those times, they saw text as just
a unit above the sentence, which, to them, then meant a sequence of
well-formed sentences and not an entity susceptible to different-order
principles. Accordingly, at that stage, linguists believed text analysis
should aim basically at breaking a text down and extracting compo-
nents from it. But text resisted. It simply did not lend to analysis of
that sort and required principally new approaches in order to lend to
successful tackling.
Then other voices and opinions became prominent and way more
easily heard because of the unavoidable conclusion that, as Faucon-
nier (1997: 7) emphasizes, “there is no hope of retrieving interesting
principles of meaning organization from surface distributions alone”.
Also, “language data suffers when it is restricted to language alone”
(ibid.). Sinclair, similarly, maintains that “linguistics has been formed
and shaped on inadequate evidence and, in a famous phrase, ‘degener-
ate data’” (in Coulthard, 1994: 12). As many others, he also calls for
redressing the “particular balance between speculation and fact in the

14
 Chapter Two. Text and/versus Discourse

way in which we talk about our subject” and “In linguistics up till now
we have been relying very heavily on speculation” (ibid.).
However, it should be duly noted that, in those times, interest in
text developed in parallel with and most probably in dependence on
some other trends in the communication-oriented Humanities and espe-
cially in Linguistics. If the transition from sentence to text represented
evolution in terms of, let’s call it, scope of object of analysis, then that
happened simultaneously with two other major tendencies. The first one
can be said to have to do with the basic method of analysis of language
data. That method can be summarized to have moved from prescrip-
tive to descriptive. In other words, analysts stopped prescribing how
language should be used; they moved on to describing, and, more im-
portantly, analyzing how language is actually used on a daily basis. The
second tendency can be said to have to do with, let’s call it, the na-
ture, or essence, of the object of analysis. In that respect, studies can
freely be accepted to have evolved from a, practically, exclusive accep-
tance of highly-idealized, standard-only data (for a discussion see, e.g.,
Coulthard, 1994) to naturally-occurring language uses.
Clearly, there is a strong, and very logical, if not natural, link be-
tween those two trends. If a scientist sees the need to approach language
(and text specifically) in its true nature and reality, then s/ he does need
to describe and analyze it the way it is and the way it happens, and not
use their own (invariably privileged) knowledge of language as a means
of sanctioning what classifies as language and what does not. And here
comes the very logical third trend which emerged in the process of ac-
cepting text as part of language (and communication) studies – if we
study language as it happens and we do not only prescribe for how it
should happen, then no-one could ever stand behind the sentence as be-
ing the actual and ultimate ‘unit’ of communication. As a consequence
to all three trends, a fourth shift occurred – one in emphasis from com-
positionality (i.e. determining minimal units and the strict rules for their
composition into the next-level basic units) to integration (i.e. incorpo-
rating various entities from across ‘levels’ and sometimes doing that
strictly against principles of compositionality).
That fourth – and crucial to text analysis – shift prompted text stud-
ies (e.g. Halliday, 1973, 1975, 2003; Halliday and Hasan, 1976, 1985;
van Dijk and Kintsch, 1978, 1983; de Beaugrande, 1980; van Dijk,
1972, 1977, 2000, 2008, 2009, 2015) to focus on the joint and simulta-

15
TEXT STRUCTURE: A WINDOW INTO DISCOURSE, CONTEXT AND MIND

neous functioning of textual components, and on merging the aspects of

textual communication into complex, well-integrated models. Research
of this type proves, among numerous other things, that if there indeed
is a ‘textual level’, then it employs different principles from those op-
erating at the ‘lower’ levels. Importantly, those scientists unanimously
corroborate the contention that the internal structure of a long sentence
differs qualitatively, and not just quantitatively, from that of a struc-
tured sequence of sentences which might constitute a text. As Fowler,
for example, puts it simply, “A text is made up of sentences, but there
exist separate principles of text-construction, beyond the rules for mak-
ing sentences” (1991: 59).
What constitutes a text and the generally-established approaches to
that issue are the target of the discussion in the next section.

II.2. Text: Basic Characteristics

II.2.1. Oral or Written
Let us now go back to the roots, or, rather, let us go back to those
simplest and most fundamental questions concerning the basic charac-
teristics of text and pertaining to channel (term as in Hymes, 1964),
number of participants, length, etc. In other words, let us outline the next
step in Linguistic development related to text.
In the first decades following Structuralism, text studies of various
backgrounds raised the issue of whether a piece of language use should
classify as a text in dependence on channel of delivery and number of
text producers. To answer that, if we do choose to go back to origins
(i.e. Quintilian’s times), what classified as a text then would most often
be delivered orally – Classic Rhetoric famously dealt with persuasive
public oratory. Despite that undeniable lean towards spoken pieces, one
should admit that most of those texts would have also been written and
prepared before their public delivery. So, what did Classical Rhetori-
cians actually deal with – written texts or their orally delivered versions?
Was there a difference between the two, or a speaker would invariably
present the text exactly the way it had been written in advance? And if
there were differences, then which one was the text in the then-analyst’s
mind – the one written in advance or the publically presented one? This
issue (of the distinction between oral and written texts) is unanimously
considered to have been best addressed quite a number of centuries later

16
 Chapter Two. Text and/versus Discourse

through the so-called Orality-Literacy Continuum theory (most notably

Ong, 1982, 1999; Tannen, 1982; Chafe, 1982; Schlesinger, 1989).
The notion of ‘Orality-Literacy continuum’ emerged in the works of
Ong (1982) and Tannen (1980a). Best known for her research in Gender
Studies (e.g. 1994, 1996, 2007) and family interaction (e.g. 2006, 2008,
2009), Tannen, especially, produced a number of papers (e.g. Tannen,
1980a, 1980b, 1980c, 1982), which have come to be extremely influ-
ential not only to discourse and sociolinguistic investigations but also
to Text Linguistics and Translation Studies (see e.g. Alexieva, 1997,
1999). In those works, both Ong and Tannen raise a very simple, and
crucial, question – do oral and written texts really differ principally?
A question like that is not meant to suggest that there is no distinc-
tion between the two. As Brown and Yule argue, even only from the
point of view of production, it is more than clear that “spoken and writ-
ten language make somewhat different demands on language-produc-
ers” (1983: 4). The theory behind the Oral-Literate continuum, however,
came into being and was even originally formulated as a reaction to the
dichotomy-based misconception of the distinction(s) between speech
and writing.
To prove that that dichotomy is actually an unfounded – though the-
oretically-friendly and “very tantalizing” (Tannen, 1982: 37) – abstrac-
tion, researchers in the field started by isolating parameters along which
the two channels diverge. A major point to clarify with that respect – a
point basic to any understanding of the Oral-Literate distinction – is
that, from Tannen’s viewpoint, the distinction concerns much more than
the kind of channel via which a text is being transmitted, i.e. whether it
is being written down or uttered aloud. The choice (or necessity) of opt-
ing for one of the two channels, in fact, turns out to further bring about
a whole cluster of other choices.
An illustration of the intricate interrelationships among such choic-
es (along all ‘levels’ of language) is provided in Brown and Yule (1983:
15) where they summarize previous work on spoken language (e.g.
Chafe, 1979; Sinclair and Coulthard, 1975; Goffman, 1981) along the
following lines:
(a) The syntax of spoken language is typically less structured than
that of written language; spoken language tends to contain many
incomplete sentences and even simply phrases in sequences. Subor-

17
TEXT STRUCTURE: A WINDOW INTO DISCOURSE, CONTEXT AND MIND

dinate sentences are very few in it. Spoken language relies predomi-
nantly on the use of active declarative forms and avoids passives,
it- and wh-cleft sentences. While written language sentences are
generally structured in subject-predicate form, in spoken language
topic-comment structure is more frequent as, e.g., in ‘the cats + did
you let them out’ (Brown and Yule, 1983: 17).
(b) Parataxis in spoken language employs almost exclusively and,
but, and then while in written language an elaborate set of discourse
markers (e.g. firstly, more importantly, in conclusion) are used to
make relationships between clauses overt. In general, ‘the speaker
is typically less explicit than the writer’ (ibid.: 16)
(c) Heavily pre-modified noun phrases characterize written lan-
guage and are quite rare in spoken language (ibid.: 17).
(d) In terms of vocabulary, ‘spoken language is certainly more flex-
ible. A speaker may go back and replace or repair a first choice
of expression, e.g. ‘this man + this chap she was going out with’
(ibid.).
As I have argued elsewhere (Tincheva, 2012c: 14), other research
in the field (e.g. Chafe, 1982; Schlesinger, 1989) also focuses on the
isolation of parameters along which the two channels diverge. A sum-
mation of the very basic lines of differences between spoken and written
texts would need to include:
(a) speed of production (oral delivery is characterized by a much
higher rate than written presentation – a fact which imposes certain
restrictions on the latter);
(b) degree of planning (the literacy end of the scale features careful-
ly pre-planned texts of high semantic density/ informational load,
more complex syntax, etc.),
(c) spatio-temporal distance/ proximity (oral delivery presupposes
a greater physical closeness between the participants than written
presentation does, which may involve social status distance/ prox-
imity effects),
(d) participants’ involvement (personal, often emotional, engage-
ment is considered more characteristic of oral discourse),
(e) the role of non-verbal behaviour (body language specificities,
voice quality, dress code, etc. play a greater role in face-to-face
contact, and, consequently, in oral text production).

18
 Chapter Two. Text and/versus Discourse

Importantly, all those factors function together and simultaneous-

ly to allow for the presence of high degrees of speficity in every text.
(Thus, if viewed from a conceptual-intergration perspective (V.1.5.2.
below), the Orality-Literacy continuum notion can be argued to be, es-
sentially, a blend.)
As stated above, the first important conclusion to draw from any
review of such studies is that they prove that the Orality-Literacy dis-
tinction concerns many more factors and linguistic features than sim-
ply channel. A second conclusion would be that the choice of channel
conditions whole mixtures, or clusters, of oral and literate properties
displayed by each text. On the basis of those two, a third – and critical
–conclusion, is the one drawn by Tannen herself (ibid.) who proves that
the Orality-Literacy distinction should not be viewed as a polar opposi-
tion but, instead, as a scale, or, in her term, a continuum.
If approached as a cline on one end of which we have typically Oral
texts (e.g. everyday family conversation) and on the other end we have
typically written texts (e.g. Encyclopedia articles), then each text we
analyze could be evaluated in accordance with the above-listed features
and parameters. In this way, the (nearly) unique position of each text
along the Orality-Literacy scale can be determined on the basis of the
specific cluster of features it displays.
Applied to the present corpus, the ‘continuum’ approach suggests,
for example, political speeches, although invariably delivered orally,
should not classify as indubitably bound to the Oral extreme of the
scale (for the whole discussion see Tincheva, 2012a: 36). Truly, politi-
cal discourse in general is persuasive discourse (Ungerer and Schmid,
1996: 149; Gill and Whedbee, 1997: 158). Political speeches, therefore,
serve predominantly interpersonal purposes rather than informational
ones (Tincheva, 2012a: ibid.). In view of both these factors, political
speeches, admittedly, do gravitate towards the Orality end of the scale,
as any intuitive first impression would suggest. And, indeed, in Quintil-
ian’s as well as much later times, political speeches were delivered in
direct contact with the audience. Additionally, that close proximity was
also reflected in stronger interpersonal effect (which is another param-
eter of the Orality-Literacy scale) and, consequently, in the more intense
involvement of the speaker. All three factors do characterize the Orality
extreme of the continuum. Other factors, however, also come forcefully
into play.

19
TEXT STRUCTURE: A WINDOW INTO DISCOURSE, CONTEXT AND MIND

If another parameter – participant proximity – is taken into consid-

eration, then it must be acknowledged that political speeches, at least in
present days, rely largely on radio and TV broadcasts in reaching their
audiences (see e.g. Fairclough, 1989, 1995). In consequence, they come
closer to the Literacy end of the scale for being mediated. A second
consequence of social distance broadening is that the speaker cannot
rely on immediate audience response to make adjustments while speak-
ing, which pushes political speeches one more step away from the Oral
extreme along the continuum.
Another point is that the media-imposed differences in speaker’s
and hearers’ time and place make the ‘now’ and ‘here’ of each speech
complexly specific. As political speeches are normally written prior to
delivery, they tend to optimize the informational load over the whole
text to a degree oral texts do not reach during online production. This
semantic density which is atypical for oral texts is further reflected in
the syntax, where subordination and passives are used more often than
in typically oral genres.
Yet another consideration shifts the genre further away from the
Orality end of the continuum: the writer of a political speech rarely co-
incides with its deliverer and, thus, s/ he is completely denied the op-
portunity to re-model the text during online delivery. In consequence,
the fixed nature of the text registers the next move toward the Literacy
end of the scale.
Many more arguments (see Tincheva, ibid.) can be raised to sup-
port the claim that even a seemingly typical oral genre such as a politi-
cal speech proves to lie in fact fairly away from the Orality end of the
scale and to cline towards the middle position along the Orality-Literacy
continuum. As far as this particular genre is concerned, it seems safe to
conclude that the main reason for their shift towards the Literate end is
that they are written to seem oral, when in fact sharing many features
with written genres. The main point of the argument here, however, is
to illustrate the viability of the Orality-Literacy approach and, above all,
to support the basic inadequacy of a dichotomy-based perception of the
‘principal’ difference between oral and written texts.
An approach like Tannen’s can also be concluded to harmonize
with Prototypicality (III.2. below). Although proponents of the Orality-
Literacy continuum approach have never expressed overtly the existence
of a connection between the two theories, it can safely be argued that, in

20
 Chapter Two. Text and/versus Discourse

both, the inherent understanding of the concept of text is basically the

same. In neither of them are there obligatory conditions for classifying
or rating a text. In both of them, each text stands on its own and does
not ‘embody’ or ‘exemplify’ a pre-existing ‘position’ along the cline or
within the prototype.
At present, researchers from all text-related fields agree that it does
not matter if a text is oral or written, it is still correct for it to be termed
text. For Halliday and Hasan, most prominently, a text is “any passage
– spoken or written” (1976: 2). In a later book, they even extend their
understanding to claim that text “may be either spoken or written, or in-
deed in any other medium of expression that we like to think of” (1985:
10). Carter and McCarthy (2006) likewise see text as any stretch of lan-
guage, either in speech or in writing. Stubbs (1983: 10), quite similarly,
maintains that “there are problems in the analysis of discourse which
are common to formal and informal, written and spoken language”.
In the same vein, Dirven and Verspoor choose to elaborate on the fact
that
While writing, we mainly or exclusively communicate by words. This
is verbal communication. While speaking, we not only communicate by
words or verbal communication, but also by means of loudness, rhythm
and speed. These elements accompanying our words are known as para-
linguistic (or paraverbal) communication. Our gestures, facial expressions
and body language are non-verbal communication.
In spoken communication the text – the words we speak – is but one of the
three means of expression. In written communication the text is almost all
there is. (2004: 180)

However, they conclude, in both cases, a text is a language form of

communication, although it is also related to the specific language users
who employ it and not simply to the language form.
A statement like “It does not matter if a text is oral or written, it is
still correct for it to be termed text” (above) may easily be seen as com-
monplace. Still, even the briefest of glances at the term of discourse
will show that the matter is less than straightforward. Crucially, as of
this point of our discussion, it is imperative to state that the continuum
principle proves that, at the most basic level, an oral and a written text
are two types of the same phenomenon – text building/ construction. If
that were not the case, positioning the two types of text along a single
scale would not be possible.

21
TEXT STRUCTURE: A WINDOW INTO DISCOURSE, CONTEXT AND MIND

II.2.2. Monologue or Dialogue

The question of whether a text should be treated as a single, mo-
nologic contribution or as a dialogic one produced by two participants,
however, presents a different challenge. First, it has not been researched
and systematized the way the Oral-Written distinction has been. Second,
the scientific perception of a dichotomy between the two forms – mono-
logic and dialogic ones – as argued in the next Section, proves to be the
norm rather than an exception. In other words – and that is a significant
reformulation of wording – to most scientists, the production and recep-
tion of monologic and dialogic types are governed by basically different
rules and principles.
The relevant issue here is whether actual language use really does
support the assumption that no text could display simultaneously traits
pertaining to monologic and dialogic types. Furthermore, can the as-
sumption that monologic and dialogic genres differ radically be in any
way compared to the (once existing) assumption that no text combines
features representative of written genres with features characteristic of
oral genres? Can the two assumptions be compared or equaled? And,
most importantly, does the assumption about the monologue-dialogue
distinction entail that monologic and dialogic language uses employ dif-
ferent principles and do not both lead to the production of text? Provid-
ing an answer to these questions is one of the objectives of the present
inquiry and we will return to them.

II.2.3. Size
The structuralist preoccupation with size (see II.1. above) neces-
sitates a note on yet another characteristic feature of text – how long a
stretch of language has to be in order for it to be called a text. Clearly,
a whole novel, or a series of volumes would readily classify as texts.
But does a one-word or a one-sound communicative chunk also rate as
such?
At present, the answer to that question is a rarity in that there is,
practically, unanimous agreement throughout the literature on its an-
swer. As Semiotic studies began to gain speed (through the works of,
e.g., Lotman, 1976, 1990; Eco, 1966, 1976), the issue of the ‘lower lim-
it’ of text ceased to exist. The STOP sign, for instance, has become a
most salient example (e.g. Peirce, 1931; Atkin, 2010). The most recent

22
 Chapter Two. Text and/versus Discourse

extension of text to refer to short message services (SMS) can be also

seen as a reflection of the general acceptance of the brevity of text, as
long as another requirement – that of a text’s functionality – is success-
fully accomplished.
Halliday and Hasan, too, being intrinsically related to both Semiot-
ics and Text Linguistics, accept text as a stretch “of whatever length”
which is “not defined by its size” (1976: 1). Basically, in similarity to
all questions which center round a single parameter of language and its
use, the issue of the size of text can be said to be (almost) irrelevant. To
claim that size operates separately and in isolation from other variables
would be, to put it mildly, too simplistic.
To conclude with a view that best summarizes our discussion so far,
let me, first, resort to Halliday and Hasan’s statement that a text
may be spoken or written, prose or verse, dialogue or monologue. It may
be anything from a single proverb to a whole play, from a momentary cry
for help to an all-day discussion on a committee (1976: 1).

How to distinguish a text from other forms of communication, how-

ever, is far less straightforward an issue. The following Section address-
es some of its specificities.

II.3.Text: Advanced Characteristics

II.3.1. Text or Discourse
There are several reasons why distinguishing between text and dis-
course can prove a dispiriting task. Even the briefest investigation of
what discourse is will easily prove that, at present, its corresponding
term is not used to denote a single thing. Rather, use of the term has
been becoming more and more uneven in parallel with the expansion
and multiplying of old and newly-emergent discourse studies.
To parallel our earlier approach onto text, a historical note on the
very word discourse would require it be said that the word  derives
from the Latin discursus meaning running to and from. In other words,
the term discourse was originally employed because of its denoting
somebody or something moving from one place or person to another
place or person. Later, it seems safe to argue, the term came to refer to
the whole exchange rather than to the ‘thing’ exchanged only. Impor-
tantly, at those earlier stages, the senses of the word discourse were used

23
TEXT STRUCTURE: A WINDOW INTO DISCOURSE, CONTEXT AND MIND

to express the inherent structural properties of a communicative act (i.e.

of the exchange ‘within’ which a thing was exchanged) and not the so-
cial surroundings or implications of that communicative act.
At present, anything that has to do with language uses ‘above’ the
‘sentence level’ tends to be amalgamated under the heading of ‘Dis-
course Analysis’. In that, discourse can legitimately be claimed to des-
ignate two major strands of language-oriented or language-related stud-
ies. The ‘narrower’ approach would see discourse as pertaining almost
exclusively to the underlying structure(s) of conversational exchanges
in terms of turns, pairs, moves, acts, exchanges, etc. (e.g Sinclair and
Coulthard, 1975; Coulthard and Brazil, 1981; Sacks and Schegloff,
1974; Brown and Yule, 1983; Hoey, 1979; Mulholand, 1991; Tsui, 1994;
Sinclair, 2004; Walsh, 2006). The ‘broader’ approach would focus way
less on the structural conversational peculiarities but, instead, would
elaborate on them as available means of expressing social differences/
distances and the (cognition-dependent) ways of thinking about them
(e.g. Coates, 1986; Milroy and Milroy, 1975; Trudgill, 1977; Hodge
and Kress, 1979; Fairclough, 1989, 1992, 1995, 2008; Mumby, 1993;
Obeng, 1996; Wodak, 1989; van Dijk, 1985, 1997, 2000, 2008, 2009;
Tannen, 2005, 2007, 2008, 2009; Lakoff, 1996, 2013, 2014).
To put it even more simply, while the first type of approach targets
the linguistic structural patterns of communicative exchanges and the
interconnection between (frequently only implied) interpersonal actions
and kinds of linguistic structures, the second type shifts focus on the
social specificities which the linguistic structures are used to express
and perpetuate. Thus, in the first case the concept of discourse will cor-
relate with notions such as conversational unity and connectivity (most
frequently, unity of ‘acts’) or with ways of combining separate stretches
into unified communicative wholes. In the second case, the notion of dis-
course will gravitate more closely towards notions relating to types of
communicative spheres and practices, e.g. the discourse of politics, legal
discourse, the discourse of power, religious discourse, etc. (e.g. Reiss,
1976). This second kind deals extensively with social beliefs, ways of
thinking and communicative means of perpetuating social practices and
differences. To use discourse analyst Fairclough’s words (1992: 65),
discursive practice “contributes to reproducing society (social identities,
social relationships, systems of knowledge and belief) as it is, yet also
contributes to transforming society”.

24
 Chapter Two. Text and/versus Discourse

What is more, within this second type, there may be argued to exist
two subtypes, according to which, as Labocha maintains, discourse can
be used either in the singular or in the plural form, each option express-
ing different content:
The term discourse when used only in the singular number is understood
as a cultural norm which regulates linguistic (communicative) behaviour,
whereas the term discourse when used also in the plural number, i.e. in the
form discourses, may be understood […] as an expression synonymous
with the term utterance. (2011: 62)

Such a view resembles closely that of Gee (1990), who distinguish-

es between what he calls ‘little d discourse’ (i.e. specific instances of
language in use) and ‘big D Discourse’ (i.e. language use in integration
with other meaning-making practices  such as gestures, glances, dress
code, customs, habits, etc.)
Over the last couple of decades, especially, it has been the second
general approach to discourse which decidedly has attracted the greater
amount of analytical interest. It has even grown to overlap with areas
of intellectual curiosity which extend far beyond the realm of ‘pure’
linguistics. This trend is characterized, probably most famously, by Fou-
cault’s investigations on discourse (e.g. 1969, 1972), where he defines
discourse as
ways of constituting knowledge, together with the social practices, forms
of subjectivity and power relations which inhere in such knowledges and
relations between them. Discourses are more than ways of thinking and
producing meaning. (1969: 2)

Importantly, the two interpretations, or, rather, the two major ap-
proaches to the notion of discourse cannot and should not be seen as
standing in opposition. On the contrary, it is more accurate (as well as
more practical) to see them as representing the two sides of the same
coin. Which one to flip on is rather a matter of theoretical viewpoint
and analytical need, and not that much of exclusion and rejection of
the alternative theoretical strand. Such a neat and logical line between
the general approaches to discourse and the general approaches to text,
however, is way more difficult to draw.
If a study offers an overt formulation of the distinction between text
and discourse, that study would be an exception. Even de Beaugrande

25
TEXT STRUCTURE: A WINDOW INTO DISCOURSE, CONTEXT AND MIND

and Dressler (1981), famous for their exhaustive reviews of the literature
and their (re-)definitions of existing concepts, focus exclusively on text,
and when they mention discourse (only in the last parts of their ground-
breaking book of 1981), they do so almost out of the blue and without
providing any definition. Moreover, de Beaugrande and Dressler’s treat-
ment is typical of yet another peculiarity of investigations on text and
discourse.
As I have argued elsewhere (Tincheva, 2012c: 12), there can be
summarized to exist two generally-employed ways of dealing with the
theoretical and terminological jumble around the two concepts. The first
one is to approach them as opposites. In such cases, discourse is often
postulated as long, while text – as short. If text is taken to be written,
discourse tends to be defined as its oral counterpart. If text is seen as
relating to monologues, then discourse is defined as controlling inter-
active forms such as dialogues. Frequently, those lines of dissimilarity
are taken to run in parallel – text is both written and monologic, while
discourse is both oral and dialogic (for a discussion of the distinctions
along those three lines see, e.g., Stubbs, 1983; Faiclough, 1989; Vir-
tanen, 1990). However, as already argued earlier here (II.2.1.), real lan-
guage uses do not support the existence of an Oral-Written dichotomy.
The monologic-dialogic distinction should similarly be seen as a scale
rather than a clear-cut division. The fact that the two are believed to of-
ten run in parallel also lends support to the latter claim. In other words,
this fact can be seen as support to the hypothesis of the existence of, let’s
call it, the monologue-dialogue continuum. Basically, opposing text and
discourse does not seem an alternative which would allow one to ap-
proach either notion successfully.
Opposing text and discourse, as mentioned above, is not the only
possible way of tackling the issue of their nature and differences. De
Beaugrande and Dressler, for example, opt for the second possible way
of dealing with the terminological and theoretical puzzle surrounding
text and discourse – they select and use one of the two terms exclu-
sively. Halliday (e.g. 1976; 1985), similarly, focuses only on text and
Sinclair and Coulthard (1975; 1977) use the term discourse only. That,
as Virtanen claims (ibid.), is a rather typical choice – most frequently,
analysts prefer one term to cover both oral and written channel, mono-
logic and dialogic form. That could happen for several analytical and/
or theoretical reasons.

26
 Chapter Two. Text and/versus Discourse

An interesting explanation of that fact – and one which rather stands

out – is that of Longacre. He maintains (1983: 337) that, in America, re-
searchers choose to call what they analyze discourse, while in Europe,
scientists prefer to use text. Practically, he further maintains, text and
discourse analysts deal with the same linguistic object of study. True to
fact, Longacre is not alone in this claim of his. As Stubbs also argues,
when defending his choice of discourse analysis over other terms,
The term text analysis could serve equally well, except that it implies work
done in a particular European tradition, represented for example by Van
Dijk’s work. Similarly, the term conversational analysis almost always
implies the ethnomethodological approach which derives from Sacks’
work. (1983: 10)

However, there are other, more significant characteristics with

respect to which text and discourse are often believed to diverge. For
example, van Dijk (1977) sees text only as an abstract theoretical con-
struct which finds its actual realization in discourse. The approach is
also supported by, to mention just one of many, Stubbs and interpreted
as equaling the claim that text is to discourse as sentence is to utterance
(1983: 10).
In a rather similar (and yet entirely contrasting) fashion, Halliday
(1978: 40), too, finds the ‘actualization’ principle relevant to the dis-
tinction. He, however, applies the principle the other way around and
defines discourse as the abstract concept, while he postulates text as the
realization of discourse. To Brown and Yule, too, a text is a ‘representa-
tion of discourse’ (1983: 5).
Importantly, both van Dijk’s and Brown and Yule’s views derive
from the belief that only one of the two concepts (i.e. only text or only
discourse) can refer to the product of a communicative act, while the
other has to be seen as an abstraction relating to the (communicative)
activities surrounding the product – a distinction to which the next Sec-
tion is dedicated.

II.3.2. Product or Process

The dissimilarity between product and process of communication
is inherent to the basic postulates of innumerable studies. As David
Crystal generalizes (2008), a significant amount of linguists maintain
the distinction between the notions of text as a physical product and

27
TEXT STRUCTURE: A WINDOW INTO DISCOURSE, CONTEXT AND MIND

of discourse as a dynamic process of expression and interpretation (the

latter representing the entity which can be studied through psycholin-
guistic and sociolinguistic means). In a similar fashion, Brown and Yule
maintain a text is “the verbal record of a communicative act” (ibid.: 6),
while discourse is language in use. Furthermore,
the discourse analyst treats his data as the record (text) of a dynamic pro-
cess in which language was used as an instrument of communication in a
context by a speaker/ writer to express meanings and achieve intentions
(discourse)’ (ibid.: 26).

Edmonson also expounds a similar view by terming text “a struc-

tured sequence of linguistic expressions forming a unitary whole” and
discourse “a structured event manifest in linguistic (and other) behav-
ior” (1981: 4). Widdowson likewise defines text as “sentences in combi-
nation” and discourse as the use of sentences in combination (1979: 90).
For Kress (1985), too, a text is a manifestation of discourse as well as of
the meaning of discourse.
Approaches of this kind basically interpret the distinction between the
product and the process of communication as deriving from the fact that
unless we believe that language-users present each other with prefabri-
cated chunks of linguistic strings (sentences) […], then we must assume
that the data we investigate is the result of active processes on the part of
each participant in communication. (Brown and Yule, 1983: 23)

This ‘discourse-as-a-process’ approach, Brown and Yule maintain,

provides the analyst with crucial information about the function(s) and
purpose(s) of a piece of linguistic data and the way that data is pro-
cessed. In their words,
We shall consider words, phrases and sentences which appear in the textual
record of a discourse to be evidence of an attempt by a producer (speaker
/ writer) to communicate his message to a recipient (hearer / reader). We
shall be particularly interested in discussing how a recipient might come
to comprehend the producer’s intended message on a particular occasion,
and how the requirements of the particular recipient(s), in definable cir-
cumstances, influence the organisation of the producer’s discourse. This is
clearly an approach which takes the communicative function of language
as its primary area of investigation and consequently seeks to describe lin-
guistic form, not as a static object, but as a dynamic means of expressing
intended meaning. (ibid.: 24)

28
 Chapter Two. Text and/versus Discourse

Fairclough, too, differentiates between text-as-a-product and text-

as-a-process. However, his understanding of communication is typical
of a slightly different viewpoint on both text and discourse – one which
bears a far more society-dependent tinge. Fairclough, thus, defines lan-
guage as “social practice” and states that “language is a social process”
(1989: 24). Consequently, his treatment of discourse suggests discourse
“involves social conditions, which can be specified as social conditions
of production and social conditions of interpretation’” (1989: 25, italics
as in original). Text, in contrast, is the product of communication:
A text is a product rather than a process – a product of the process of text
production. But I shall use the term discourse to refer to the whole process
of social interaction of which a text is just a part. This process includes
in addition to the text the process of production, of which the text is a
product, and the process of interpretation, for which the text is a resource.
Text analysis is correspondingly only a part of discourse analysis, which
also includes analysis of productive and interpretative processes. The for-
mal properties of a text can be regarded from the perspective of discourse
analysis on the one hand as traces of the productive process, and on the
other hand as cues in the process of interpretation (ibid.: 24).

What seems – at least on the surface – to be a different angle on

whether text is a product or a process is Halliday and Hasan’s by now
classic claim that:
The text is a product in the sense that it is an output, something that can be
recorded and studied, having a certain construction that can be represented
in systematic terms. It is a process in the sense of a continuous process of
semantic choice, a movement through the network of meaning potential,
with each set of choices constituting the environment for a further set.
(1985: 10)

This contention has also enjoyed extensive and varied support

throughout the literature. Neubert (1992), quite similarly, sees texts as
tools and claims that they simultaneously ‘reveal’ the tool-user. Dirven
and Verspoor, too, define text as both the linguistic expressions used in
a communicative environment and the recipient’s interpretation of those
expressions (2004: 180).
Generally, the supporters of the idea of the dual nature of text do
not really contradict radically the upholders of the position that text is
the product and discourse – the process. A closer look at Failrclough’s

29
TEXT STRUCTURE: A WINDOW INTO DISCOURSE, CONTEXT AND MIND

and Brown and Yule’s definitions (above) on the one hand, and Hal-
liday and Hasan’s, on the other, will reveal no differences in principle.
Rather, the major dissimilarity lies in the choice to use only the term text
that Halliday and Hasan, Neubert and Dirven and Verspoor have made.
Thus, what Fairclough and Brown and Yule suggest to be discourse-as-
process can easily be argued to be basically the same as what Halliday
and Hasan propose as text-as-process. What Fairclough and Brown and
Yule suggest to be text-as-product is basically the same as what Hal-
liday and Hasan also see as a product. In other words, there seems to
be consensus in the literature on text as the product of communication.
Moreover, there also seems to be consensus on the fact that there exist
‘processes’ – social or individual – from which the product-text results.
The only theoretical differences ensue from the diverse viewpoints on
how to term those processes and, crucially, on how far to broaden analy-
sis of those processes.
True to fact, with the advance of cognitive studies, the linguistic
view of text has gradually expanded to equal the discourse-as-a-process
one – a development most plainly described as moving from a static
analytical stance to a dynamic approach with an increasing emphasis on
text reception (see Virtanen, 1990; Dobrzynska, 2005). As Dirven and
Verspoor put it, at present,
Text linguistics is the study of how S (speaker, writer) and H (hearer, read-
er) manage to communicate via texts, that is how they go beyond the text
(words) they produce or have in front of them to see the relations between
the sentences, the paragraphs, the sections, etc. (2004: 180)
Furthermore,
Text can consequently be defined as the linguistic expressions used in
communication between people and the interpretation the hearer or reader
makes of them. This definition applies to both oral and written commu-
nication, but has the additional condition that text here only denotes the
verbal part of the communication, excluding the paralinguistic and non-
verbal aspects. This text definition also presupposes the cultural or world
knowledge on which the text interpretation is based. (ibid.)

Indeed, the evolution in the understanding of text has moved from

a descriptive structural notion (i.e. the one of a supra-sentential unit)
towards a more procedural concept that incorporates context into any
text or discourse study. The main question here is – how much and to

30
 Chapter Two. Text and/versus Discourse

what extent has it broadened? How far ‘around’ the text-product does it
stretch? (Whether it, at all, should be seen as ‘surrounding’ the text will
be the subject of our discussion in Chapter IV.) And, as all the defini-
tions cited above include ideas of processes, processing and participants’
interpretations, it is only logical to ask: where exactly does text reside?
Is its main locus ‘out there’ on paper? Is it ‘out there’ ‘in’ society? Or
does it pertain to the individual participant’s (mental) processes?

II.3.3. On paper/ record or In the mind

The debate around the locus of text, which derives from the debate
about its product-or-process nature, can be argued to be best expressed
by the two major strands of what is generally accepted to be the scien-
tific realm of Text Linguistics – Halliday and Hasan’s break-through
theory of Cohesion and de Beaugrande and Dressler’s self-titled Proce-
dural Approach to textuality.
Despite his declared conviction of the dual nature of text (see II.3.2.
above), Halliday’s theory – as expounded in his and Hasan’s seminal
Cohesion in English of 1976 – is generally believed to lie extremely
close to the view of text as a product. And that is certainly a claim not
difficult to agree with. As Brown and Yule argue forcibly (1983: 24),
Halliday’s very understanding of cohesion as words linked together
does indeed suggest analysis of the printed text (i.e. the product) and not
of the processes of text production and reception. Admittedly, no matter
how frequently Halliday repeats in his book that cohesive devices are
clues left by an encoder to help a decoder interpret the text successfully,
what Halliday proposes as actual analysis does rely almost exclusively
on types of linguistic structures and forms.
On the other hand, Halliday’s constant point of departure is func-
tion. He and Hasan define text as language that is functional, as language
“that is doing some job in some context” (1985: 10) – a definition which
relies heavily on participants’ actions in the process of communication.
What is more, Halliday repeatedly claims that a text “is not made of the
words written down or uttered aloud but is made of meanings” (1976:
10), i.e. a text is not just made of ‘static’ structures and forms. Further-
more, it should be acknowledged that Halliday’s whole understanding
of cohesion is based on people’s abilities to assign meanings to the lin-
guistic structures used, which is, practically, another way of referring
to the process of text interpretation. However, the major problem with

31
TEXT STRUCTURE: A WINDOW INTO DISCOURSE, CONTEXT AND MIND

Halliday’s approach is that it never offers an explanation of what ‘to as-

sign meanings’ is.
Thus, from Halliday’s viewpoint, interpreting meanings and un-
derstanding a text just happen somehow. People, to him, are able to
“normally decide without difficulty” if a sequence of sentences forms
a text or not (1976: 1). People “know as a general rule” the meaning of
a portion of language (ibid.). People “intuitively” recognize if there are
logical connections between semantic units in text or not (ibid.: 3). In a
way – and that is a crucial way – in Halliday’s theory, however power-
ful it may be, there exists a huge gap between the textual record and the
person who performs the meaning assigning procedure. The assigning
procedure, thus, happens ‘somewhere out there’ in what lies between the
person who communicates and the text s/ he communicates through. The
assigning ‘somehow’ relies on conventional, socially shared knowledge
as well as on personal grasp of individual linguistic signs. On the whole,
those ‘gaps’ between participants in the communicative situation and
the text-product do lead to the general perception of Halliday’s approach
lying too close to the textual record and not representing correctly the
procedural nature of textual communication.
What de Beaugrande and Dressler propose as an alternative is an
understanding of text which, regrettably, is, in fact, to too many, too ex-
tremely procedural to put to practice. As widely known, from de Beau-
grande and Dressler’s point of view, in order for a portion of language
in use to qualify as a text, it has to comply with seven standards of tex-
tuality: cohesion, coherence, intentionality, acceptability, informativity,
situationality and intertextuality. Several problems have ensued from
that definition over the decades.
The first problem has to do with the first two parameters in the list
– cohesion and coherence. As those two seem to be inherited from Hal-
liday’s earlier work (which de Beaugrande and Dressler acknowledge,
see e.g. 1981: 23), many present-day researchers are quite ready to just
amalgamate the two approaches and, practically, insert Halliday’s find-
ings on cohesion and his comments on coherence into de Beaugrande
and Dressler’s theory. Such an ‘injection’, however, is not exactly
healthy as it is not applied where it should. As the discussion below
will, hopefully, demonstrate, the two theories diverge significantly on
too many basic points for a natural and healthy merge of the two to take
place that easily.

32
 Chapter Two. Text and/versus Discourse

The second problem with de Beaugrande and Dressler’s list of stan-

dards is the one most frequently leveled at them – that two of the stan-
dards they propose are defined as text-centered (i.e. cohesion and coher-
ence), the rest – as user-centered. For the purposes of the present study I
will ignore the fact that that lack of homogeneity among the parameters
itself has sparkled consistent criticism (I will come back later to this
point, when I discuss context). For now, I will focus on the fact that the
first two parameters, i.e. cohesion and coherence, coincide with what to
Halliday is the essence of texture. And that is the third significant issue
arising from the differences between Halliday and Hasan’s and de Beau-
grande and Dressler’s treatment of text.
De Beaugrande and Dressler’s concept of textuality and Halliday’s
notion of texture have little in common. While texture is inherent to and,
thus, resides ‘in’ a text, textuality has to do with participant’s percep-
tions of the text. Texture relies heavily on the net of cohesive chains
which span a text; it is a property possessed by a text which distinguish-
es it from a non-text (1976: 2). Textuality, in contrast, is evaluative in
nature and serves to explain why a text is successful and communica-
tive, or why a text does not achieve its communicative purpose. Texture
stays close to the text’s cohesive devices only, while textuality concerns
perceptual differences from the point of view of the participant’s minds.
As a consequence, texture arises from the presence of three obligatory
factors only – cohesion, thematic progression and genre structure. Tex-
tuality depends on seven parameters, cohesion, admittedly, being one of
them. However, the place of cohesion within the two theories could not
and should not be equaled.
To Halliday cohesion holds between linguistic elements through
their meanings. To de Beaugrande and Dressler, cohesion exists because
of the limitations and the needs of the human mind. To them, cohe-
sion results from the need to portion information input into manage-
able chunks; it arises from the need to facilitate mental storage and to
ease processing effort. Cohesion, thus, in de Beaugrande and Dressler’s
theoretical framework, has to do with and results from peculiarities of
cognitive processes such as the operation of augmented transitions and
the configuration of nodes in the form of grammar states:
To move from one node to another, the processor performs a  transi-
tion across a link. This operation demands the identification of the link as

33
TEXT STRUCTURE: A WINDOW INTO DISCOURSE, CONTEXT AND MIND

one of a repertory of dependency types, e.g. “subject-to-verb” or “mod-

ifier-to-head”. The transition can be  augmented with any kind of search
or access operation, such as specifying the exact category to which the
upcoming node belongs (Winston 1977: 172). […] In a transition net-
work, the structures of phrases and clauses are operationalised as means
to build and test hypotheses about the types of elements to use or expect
at any given time. Hence, these networks capture the grammatical strat-
egies  and  expectations  of language users; and they express the rules of
grammar as  procedures for using the rules (Rumelhart 1977a: 122). […]
The phrase, clause, or sentence appears as an actually occurring grammati-
cal macro-state in which elements are micro-states of the textual system.
(ibid.: 50)

De Beaugrande and Dressler also include the following Figure

(ibid.: 51) to demonstrate how a text processor moves through a noun-
phrase network:

Figure 1

Perceptions of coherence, too, differ considerably within de Beau-

grande and Dressler’s and Halliday and Hasan’s theoretical frameworks.
To Halliday and Hasan (ibid.), it is a property which is assigned to a
text by its reader. From their perspective, a text is not automatically
or naturally coherent, but becomes coherent when its recipients find it
coherent. Thus, on their approach, while cohesion is mainly concerned
with morpho-syntactic devices ‘within’ a text, coherence is defined as a
semantic property ‘of’ text.
In contrast, to de Beaugrande and Dressler, coherence has to do
with manipulation of mental structures, with their organization and con-
tinuity. Coherence, to them, depends on two major factors: on the con-
sistency (i.e. on the absence of a clash) among the cognitive structures
activated in the process of constructing a textual world, and, secondly,

34
 Chapter Two. Text and/versus Discourse

on the absence of a significant clash between the textual world being

constructed and the background stored knowledge of the participants
in the communicative situation. In other words, while to Halliday and
Hasan it is sufficient to state that a text is coherent, when participants
get the ‘sense’ of it being coherent (e.g. 1976: 3), to de Beaugrande
and Dressler, the final result of coherence is explicable through tracing
mental modeling procedures and employing basic cognitive phenomena
into textual analysis.
Thus, in a critical way, de Beaugrande and Dressler’s theory can be
seen as an improvement on Halliday and Hasan’s framework of analy-
sis. Through their repetitive insistence on the procedural nature of all
their seven standards of textuality, de Beaugrande and Dressler actu-
ally bridge the gap Halliday and Hasan chose not to address – the gap
between the textual record and the text processor involved in textual
communication. Secondly, in emphasizing the importance of mental ac-
tivities in all matters textual, de Beaugrande and Dressler provide the
crucial change of angle toward the object of textual analysis. They, first,
put the human mind in the picture, and, secondly, they shift the view-
point on text from ‘somewhere out there’ to the point of view from in-
side the human mind.
De Beaugrande and Dressler, it should be duly noted, are far from
alone in this choice of theirs. For example, discourse analyst Fairclough,
too, states the crucial importance of introducing cognitive mechanisms
into linguistic analysis. In his words,
It is an important property of productive and interpretative processes that
they involve an interplay between properties of texts and a considerable
range of […] ‘members’ resources’ (MR) which people have in their heads
– including their knowledge of language, representations of the natural
and social worlds they inhabit, values, beliefs, assumptions, and so on.
(1989: 24)

In this, quite importantly,

the MR which people draw upon to produce and interpret texts are cogni-
tive in the sense that they are in people’s heads, but they are social in the
sense that they have social origins – they are socially generated, and their
nature is dependent on the social relations and struggles out of which they
were generated – as well as being socially transmitted and, in our society,
unequally distributed. (ibid.)

35
TEXT STRUCTURE: A WINDOW INTO DISCOURSE, CONTEXT AND MIND

As of today, a similar shift of angle, originally advocated by de

Beaugrande and Dressler, has been readily embraced by Halliday. Hal-
liday even strengthenes the viewpoint to claim that a text is, first,
transmitted physically, by sound waves traveling through the air; secondly,
it is produced and received biologically, by the human brain and its associ-
ated organs of speech and hearing; thirdly it is exchanged socially, in con-
texts set up and defined by the social structure; and fourthly it is organized
semiotically as a system of meanings. (2006: 68)

Generally, despite the theoretical and terminological mutability be-

hind the varied interpretations of text and discourse, some clear lines of
argumentation can be traced. As Virtanen argues (1990), the proponents
of the British tradition, most notably Halliday, prefer to see text from
a more sentence-based, or close-to-forms perspective. The Birming-
ham discourse analysis school especially (e.g. Sinclair and Coulthard,
Widdowson) target their efforts at investigating performance data (i.e.
spoken, actual language use) in terms of the correspondences between
structures and the interactional functions of those structures. In contrast,
what is generally seen as German text linguists (e.g. de Beaugrande
and Dressler, van Dijk) choose to adopt a proposition-based approach,
which, as long as their preferred type of data is concerned, is directed to-
wards the analysis of written language. In his turn, Van Dijk (e.g. 1977;
1997; 1999, 2001; 2006, 2007, 2009), as will be argued later, is the ma-
jor force behind the attempts to tackle the abstract theoretical construct
of text and its relationship to discourse.
In should also be noted that the differences between the ‘British’
and the ‘German’ standpoints on text do not necessarily imply a stark
theoretical contradiction. The two approaches share a number of com-
mon conceptions; for example, there can be traced considerable similar-
ities in their understanding of textual unity, cohesion, reference, etc. It
is in the mental-structure-based theoretical aspects that the two diverge
essentially.
The view of text and discourse expounded here adheres more
closely to the second, cognitive approach, which does not see text as a
product but rather as a cluster of co-occurring cognitive processes which
function inseparably from their result. Thus, text here is not taken to be
merely a material record, nor is it assumed to reside in society or in the
human mind only. Instead, text – quite similarly to de Beaugrande and

36
 Chapter Two. Text and/versus Discourse

Dressler’s position – is seen from the vantage point of the human mind
(i.e. it is decoded through contemporary knowledge of the brain’s pro-
cessing capacities) and those processes and capacities are interpreted as
indubitably bound to their language-signal results.
As far as the main objectives of the present investigation are con-
cerned, the most important conclusion emerging from the discussion of
the notions of text and discourse so far is that they are easily ‘confused’
and extremely hard to delineate, even for analytical purposes. Second,
and no less important, even when the two are delineated for analytical
purposes, scientists can switch the places of the two and justify the posi-
tion of either from various perspectives. Both facts will be positioned
here as crucial parts of my rationale behind the argument that overlap-
ping and blending cognitive techniqiues control the joint operation of
the concepts of text and discourse (IV.3.3. and V.1.5.2.). In a way, the
theories and approaches discussed so far will be used as a kind of a cor-
pus employed to confirm the hypothesized possibility for text-discourse
overlap and blending to exist.
To aid and defend that kind of understanding, the next Chapter Three
will be dedicated to the clarification of some basic cognitive tenets and
terms. Those tenets and terms are also expected to be indispensable for
our further discussion of the distinction between text and discourse, the
way it will be presented in Chapter Four.

37

View publication stats