Pragmatics and stylistics

by Vicent Salvador

Abstract
Consideration of linguistic variation is inherent to pragmatics, especially where such variation
relates to contextual factors, an area where it frequently converges with work on stylistics.
This article reviews certain relevant contributions made from the perspective of style analysis
(literary studies, sociolinguistics, and systemic functionalism) and outlines certain
approaches to a stylistics of language, centred on Buhler's threefold division into symptom,
signal and symbol. Lastly, illustrations of this approach are given, touching too on the
ideological dimension, with reference to questions of analysis.

Summary
1. Preamble

2. Literary studies

3. Social and functional variation

4. Symptoms, signals and symbols

5. Some objects for the application of stylistics

6. Conclusions

7. Bibliography

1. Preamble
It is certainly not an easy task to define the limits of stylistics with respect to neighbouring
disciplines such as rhetoric, microsociolinguistics or pragmatics. The label “stylistics” is
related to very specific writers and schools, such as Charles Bally and subsequently, in
connection with idealist stylistics, Leo Spitzer, Dámaso Alonso and Amado Alonso. In
different order of things, the discipline can also be linked with the work on functional
language variation, where the term “style” overlaps with that of “register”, and is closer to
the latter epistemological area. Recently, certain linguists have put this disciplinary label
back in circulation as a branch of the old rhetoric, and currently Bally is looking for the
foundations of a stylistics of language, beyond individual speech acts, and not limited to the
ambit of literary creation (Adam 1997).

What is more, we note the closeness to some of the subdivisions within pragmatic linguistics
-such as speech act theory, the study of politeness or metaphor and irony from the point of
view of relevance theory- to the interest scholars are now showing in style, within the
contribution made by pragmatics. These are all factors that, taken together, have ushered in
a new and explicit label for this area of the discipline: “pragmastylistics”. So, in view of the
topic's complexity and the possible points of connection with pragmatics, it would be
worthwhile now revisiting certain epistemological areas where stylistics has taken root, to
then go on to examine the different dimensions of style.

2. Literary studies
As is well known, the study of style has its origins in Rhetoric, and in particular in elocutio or
the study and improvement of expression, which steadily gained ground over the course of
time with the weakening of some of the classic components of Rhetoric. Indeed, as interest
focused increasingly on written texts, memoria and actio-pronuntiatio were relegated to
second place, and inventio, shifted, especially from the 16th century onward, to a large
extent from rhetoric (the art of bene dicere) to logic (the art of vere dicere). The other
important component, dispositio, was eventually to be given a boost, centuries later, with
the advent of narratology and text linguistics. Yet for many years, scarcely any points of

Noves SL. Revista de Sociolingüística

http://www.gencat.cat/llengua/noves
Winter 2003
Pragmatics and stylistics by Vicent Salvador

contact were found with a linguistics which took the sentence as the upper limit of its
interest. In contrast, stylistics, focusing as it did primarily on the study of microstructures,
was able to seek an analytical methodology for the linguistic study of such issues as
adjectivisation, the discourse use of verbs and word order at sentence level. (It should be
made clear that stylistics does not need to limit itself to microstructures, and here we are
adopting a broader perspective. Yet there is no doubt that microstructures have traditionally
been the favourite area of application). On the other hand, the word "style ”, has in ordinary
every day discourse, a better image than “rhetoric”, which is also a factor contributing to the
currency of the same term (style) for the discipline. This, to the detriment of rhetoric, with
whose content it preserves nevertheless a close relationship (Enkvist 1985).

In any case, style and style considerations preserve a close relationship with elocutio, and
the contemporary study of style originated the old rhetoric and went on to be increasingly
linked to the area of literary studies. And conceptualisation in terms of rhetorical figures (one
of the objects of study within the discipline, as aesthetic resource and stylistic elaboration,
often seemingly as mere ornament) further favoured this trend to wards the literary. It is
clear, too, that these figures, metaphor and irony above all, are often studied in the context
of legal argumentation (Perelman, for instance) or in relation to the pragmatics of relevance
(Sperber and Wilson), or indeed within the ambit cognitive semantics (G. Lakoff). This is true
certainly in the case of Bally, the pioneer of modern stylistics. He emphasises expressivity of
ordinary language (López García 2000). Despite that, his main area of application has been
in the analysis of literary texts. In this way the study of style distanced itself from ordinary
speech and style was even seen, at times, as a transgression, of the ordinary everyday
patterns of speech.

It is important to note that the idea of choice, of selection of meaningful options involving a
range of variants, is ever present. An author's style in a work can be viewed as determined
by a series of options that the text manifests and which are selected from a given range of
possibilities which the language offers. And when translating, for example, these can be seen
to vary considerably when having to “move house” from one linguistic system to another
(Marco 2002). It is for that reason that, between stylistics concerned with individual
discourse acts and stylistics of the language, the bridges are many and much frequented. To
give just one instance of this: when Spitzer carried out his masterly analysis of Racine's style
as a strategy of "muted" expression, he looks at the series of options chosen by the
dramatist. In showing the value of each of these choices he gradually and cumulatively
sketches out the main outlines of a stylistics of the French language, continually referring as
he does to the corresponding repertoire of variants for each of the variables in question.

3. Social and functional variation

Since that the realisation of linguistic units is diverse rather than uniform, choice from
among the possibilities offered, in each case, by the language, becomes meaningful in one
way or another. It could be the speaker's identification with a social group defined by age,
sociocultural level or geographical location (dialectal variation), or it could represent the
speaker's attempts to suit the utterance to the communicative context such as the precept,
interpretation or the construction (functional variation). Including more than the world of
literature, but without excluding it either, stylistics has to take into account the value of this
last type of optionality in all kind of discourse, to the extent that situational variation is both
individual and collective (Garrido Medina 1997).

In effect, as with dialectal variation, there is here a correlation between linguistic variation
and social variation, between language and society, since the constraints that govern stylistic
modulation of texts constitute a socially elaborated construct, by virtue of conventions that
have been progressively consolidated and transformed over the course of the history of the
language. Thus, for example, the first historic uses of the periphrastic perfective tense
(inflected simple past) in Catalan was as an expressive literary usage, an individual choice, in
narrative genres such as the medieval chronicles. Effectively, this was a resource to achieve
vivid dramatisation, and involved using the historic present of the verb to go as auxiliary.
Thus va dir used in stead of the simple form digué (both meaning "he said"). In time, this
usage spread beyond the framework of the chronicles and a purely individual choice, to
become a structurally integrated grammatical feature of the language.

Noves SL. Revista de Sociolingüística 2

http://www.gencat.cat/llengua/noves
Winter 2003
Pragmatics and stylistics by Vicent Salvador

It became, therefore, an extensively found and structurally conventionalised variant,

gramaticalised, in short; from a sociolinguistic perspective it is a marker in Labov's sense,
since the alternating of the two variants influenced by dialectal factors (present-day
geographical distribution in the Catalan-speaking territories) and at the same time
responding to stylistic factors (not just in Labov's simplified sense as degree of attention
accorded to one's own discourse, but as a multidimensional range of factors: oral or written
context, informal or solemn, even a certain archaic association in the case of the simple
past) (Salvador 2001). But what needs to be noted here is that the effect of creative style,
whose historical origins can be pinpointed, later become stabilised as a structure of the
Catalan linguistic system, the periphrastic perfective. Its use can no longer be creative, since
it has taken on the value of an option of grammatical choice more or less regulated by
convention.

Now, these patterns that are regulated by social convention do not determine linguistic
variants in the strong sense, seen as a subcode of the language. They have to be seen
dynamically, as a component of the functional concept of register, which has become
common place in studies on variation theory, especially with respect to oral / written
dimension, which today finds methodological support in the technologies of linguistic corpora
(Biber et al. 1998; Payrató and Alturo eds. 2002).

But the theoretically most potent characterisation of the notion of register surely is to be
found in systemic functionalism, where the concept is defined in terms of meaning potential,
a configuration of semantic resources which members of a linguistic and cultural community
associate with a situation type (type of context). In the Firthian tradition of British (and
Australian) linguistics, language and socio-communicative activity form an indissoluble
whole, such that linguistics cannot relinquish the systematic study of contextual parameters
which help to modulate the discourse.

Thus, for Halliday and his disciples, the notion of register establishes a model of the context
based on the interaction of three factors: the field, the tenor and the mode, which refer
respectively to sytems of social activity, power relations and solidarity among the
participants and, thirdly, the semiotic distance which is formed, based on the medium of
communication selected. Each of these three factors, in turn, relate to three types of
meaning linking linguistic organisation to social organisation: the ideational meaning that
“naturalises” conceptive reality by means of an institutionalised social activity; the
interpersonal meaning which gives material shape to social relations; and the textual
meaning which gives a semiotic dimension to communication and organises it sequentially.
Also, on a more comprehensive level of context modelling, one would need to situate the
notion of genre, relative to the system of institutionalised, teleological (goal-oriented) social
processes by means of which social activity is organised in each cultural framework (Martin
1992, 2001).

We can say, therefore, that the theoretical construct “genre” allows us to explain the
adoption of postures or roles (both on the part of speakers as listeners) in the
communicative interaction and, therefore, also in the way in which registers suitable to each
instance are selected. Similarly, at a lower level, the construct “register” corresponds to the
mechanisms that guide selection of the aptest or most efficient lexico-grammatical options
within the stylistic repertoire of the language.

Of course, this theorising seeks to account for contextual models of an interpretive order that
could exist as mental representations and determine many of the properties of the
production and reception of discourse in a given linguistic-cultural setting. Among these
properties of discourse, stylistic choices occupy a particularly important place. From this
perspective, style should be considered as a combination of formal properties of discourse
which derive from contextual models (Van Dijk 2001).

There are a series of corollaries to this approach, if we avoid an a-historical, individualist and
ideologically decaffeinated view of style as a merely “expressive” manifestation of a an
individual personality. In fact, from the point of view of “critical” discourse analysis, these
contextual models guiding stylistic choice of speakers are subjective interpretations of
contexts and their typology, and are clearly subjected to ideological control. In this sense,
then, style becomes one of the more obvious manifestations of one’s ideologies, and a
determining factor in its social reproduction (Van Dijk 1998). But let us go on to examine the

Noves SL. Revista de Sociolingüística 3

http://www.gencat.cat/llengua/noves
Winter 2003
Pragmatics and stylistics by Vicent Salvador

functioning of these properties of discourse which make up style from the angle of its
meaning as an option among the range of possibilities offered by the language.

4. Symptoms, signs and symbols

Taking the function of language in terms of Buhler’s model as a our point of departure, we
refer to symptom when the sign relates to its producer or source; we speak of signal where
semiotic activity has to do with the effect on the receiver; and we speak of symbol when
such relates to the external referent denoted by the sign. It is from this triple perspective
that we are able to enhance our understanding of at least some of the oscillation in this area
of study.

In fact, a good part of the work done on stylistics has placed emphasis on the first of these
functions, that of style as symptom, referring back to the origin of the discourse, the subject
or originator. The index value of the sign here can refer to the personality (or indeed the
unconscious) of a literary author or of any other source. Thus, there has often been a
tendency in literary studies to take style markers of reiterated occurrence as evidence in the
identification of text authorship. In forensic linguistics too, scholars have searched for
recurrent style markers capable of furnishing practical markers of plagiarism (“The
plagiarism machine”) or of identifying authors of anonymous writings with legal
consequences or criminal implications. Behind this approach lies a key notion in stylistics:
that of personal choice. But it is obvious that choice in context cannot be defined other than
by reference to a framework of possible options, and a repertoire of variation that
establishes the universe of practicable alternatives.

As regards the second function, that of signal, it should be recalled that stylistics inherited its
focus on the persuasive dimension from the old discipline of rhetoric, or at least the
emphasis on the effect achieved (if not sought) on the audience. In more contemporary
terms, we should say that style “proposes” a point of view for the receivers of the discourse,
or more exactly, this point of view is imposed since it acts as an (often imperceptible,
certainly unavoidable) filter, since style is an essential component of textual semiosis. Put
another way, it is by means of stylistic options chosen in the course of a discourse, that we
are made to see things from a pre-set point of view. This imposition, previous to its eventual
pervasive effect, possesses a cognitive dimension, shaping the perception of the listener-
receptor as a necessary condition for subsequent persuasion. This takes us on to the third of
the three pragmatic types mentioned above, in that the cognitive bias associated with the
adoption of a point of view has to do with the semantic content of the discourse.

In particular, consideration of style as a symbol raises the question of semantic relations

between equivalents or synonyms. Thus, according to a restrictive view of semantics as the
study, exclusively, of the logical-referential component, synonyms involve identity of basic or
denotative meaning, as observable across putatively synonymous pairs of words or
expressions. From this point of view, the values of elements in synonymous pairs can be
relegated to the fringe area of connotative meaning, of subjectivity and context of use. And
the latter can be understood as nebulous and nameless, or at least not removed from the
epistemological framework of general linguistics.

Stylistics, seen form this standpoint, would seem to be about impressionistic labelling of
nuanced distinctions of secondary importance. The arrival of pragmalinguistics on the scene,
however, has made it possible to increase systematisation of contextual factors.
Furthermore, the development of a wider range (a more pragmatic range, if you will) of
semantic issues, such as Language Argumentation Theory (LAT) proposed initially by Ducrot,
facilitates consideration of many such values (distinctions and nuances) that were originally
considered residual. These latter form a part of language as elements which have come to be
conventionalised in the structures of the language. Many synonymous pairs thus cease to be
such, as linguistic research makes it possible to establish explicitly the semantico-pragmatic
differences between apparent synonyms (or, more euphemistically, “almost synonyms”).
These become linguistic units differentiated via the inclusion of pragmatic specifications (for
example, in relation to requirements of indirect speech acts or politeness), or alternatively
through the emergence of connotative meaning in terms of structural factors (for example,
the suffusing of lexical units with associated axiological points of view).

Noves SL. Revista de Sociolingüística 4

http://www.gencat.cat/llengua/noves
Winter 2003
Pragmatics and stylistics by Vicent Salvador

At other times, progress in differentiating synonymous lexical units stems from the rigorous
study of contexts of use and of correlation existing between linguistic variation and social
variation. As we have seen, variation sociolinguistics, on the one hand, and studies of
systemic functionalism exploring the notion of register, on the other, clearly underpin the
functionalisation of stylistics -and the associated narrowing of the concept of synonymy.
There is no need to repeat here the well-known anecdote reported years ago by Martin Joos:
as soon as there is more than one clock at the same train station, the forces of specialisation
begin to operate. Indeed, if the clocks were to show exactly the same time, why have more
than one?

Viewed in this way, stylistics that emphasise the symbolic can be seen to push back the
barrier of (socio)linguistic entropy. It reclaims certain areas assigned traditionally to free
variation, converting them into territory governed by patterns of variation and susceptible to
systematic description. In the next section, we shall be looking at some of the most
promising stylistics issues from this perspective.

5. Some objects for the application of stylistics

Let us go on to consider, by way of illustration, the stylistics of some structures of the
language: phraseological combinations, the infusing of socio cognitive bias in the lexis, and
the mechanisms of semantic nominalisations.

As regards the first, note that phraseology is today an attractive area for linguists, especially
in application to contrastive lexicography and translation (Corpas 1998). After a stage when
it was de rigeur to evaluate linguistic creativity and free or open choice, emphasis came to
be placed more on the way speakers used “prefabricated” pieces of discourse. The literary
use of fashionable phrases and clichés has received attention from scholars of stylistics,
often in the context of originality of manipulation (reorganisation) or breaking with tradition
as a resource for "de-familiarisation". However, stylistic analysis is not averse to looking at
routines and “routinisation” which facilitate production and interpretation of discourse, in
accordance with the patterns of each register or each genre.

There is a phraseology that is characteristic of legal discourse, of administrative documents,

of scientific articles, literary criticism, news reports, electioneering, tele evangelism, radio
talk shows and even Internet chats. Such phraseological practice (which often involves mere
word collocations that occur habitually and which today corpus linguistics can help us to
study on a more empirical basis) foment a degree of productive and interpretative
automatism and facilitate the immediate identification of registers and genres. Its power as a
mechanism for reproduction of ideologies is simply remarkable, especially in the way it
contributes to "naturalise” discourse, making it more difficult to assume a critical position.
There is nothing new in this, the sayer of proverbs has functioned in this way for centuries
(often a proverb or saying is simply an ideologeme that adopts the seductive form of the
trope). The decreased visibility suffered today by such routinised forms actually increases
their ideological effectiveness in the discourse of contemporary mass media.

Another instance is the analysis of sociocognitive biases of the lexicon -axiological points of
view, as we have said, that have become conventionalised over time within a cultural and
linguistic community, which have finally crystallised in the very structures of the language.
From the perspective of language argumentation theory, subsequently developed into topoi
theory, what is relevant is the fact that many lexical units so to speak impose a point of view
on speakers, be it euphoric or otherwise (non-euphoric) (Raccah 2000). Thus, to speak of
"xafagor” (burning shame, embarrassment) and even more so the Spanish equivalent
“bochorno” (ditto) which have a literal meteorological meaning (sultry, hot and close), and a
metaphorical projection, “balafiar” (squander, waste), “emigrant” (emigrant) or “terrorisme”
(terrorism), implies an inseparable axiological content, inherent in these linguistic units. In
the same way, “warm”, “investment”, “tourism” or “struggle for independence” inevitably
convey more or less euphoric points of view. Similarly, on a scale of values, “danger” clearly
rates as "non-euphoric"), while the near-synonym “risk” is clearly less so, and even in
certain contexts (in some Christian theology, for example) may become euphoric.
Commercial advertising and political propaganda have taken careful note of all this,
obviously enough. What should be underlined here is that the built-in bias is inevitable once
the speaker / writer has selected the item in question as a consequence of a stylistic option.

Noves SL. Revista de Sociolingüística 5

http://www.gencat.cat/llengua/noves
Winter 2003
Pragmatics and stylistics by Vicent Salvador

It should not be forgotten that this choice is conditioned by socio-discursive inertia, by an

acquired habitus. And, reciprocally, the receiver (listener, reader) has to counteract this
inertia if he or she wishes to be free to carry out an interpretative criticism of the discourse.

The last illustration that we propose is that of nominal-heavy or nominalised style (Salvador
2000). Here the alternatives set up are between the expression of processes (and qualities)
by means of sentence-level constructions which contain verbs (and, often, adjectives) or
alternatively by means of nominals where the information is reduced and made compact. As
such, the latter makes it possible to perceive processes or qualities as entities or wholes.
Thus, speaking of “illegal adoptions”, “masculine beauty”, “a catastrophic seismic movement”
or “health and illness”, constitutes alternative stylistic options selected in the place of other
(roughly) similar ways of referring to the same content by means of whole sentences. Such
sentences are normally very exact, with particular shades of meaning: "children that one
adopts/ one has adopted/one will adopt, thus breaking the law”; “The/some/these/ men
are/have been/will become/seem beautiful”; and so forth. It should be noted, in any case,
that a variety of different sentential expressions can be found to correspond to nominalised
expressions.

Quite clearly, the nominals options are more "economical" and facilitate the creation of
taxonomies associated with specialist discourse, or at least a formal register with a certain
degree of opaqueness, by means of processes of "terminologisation" that are very relevant
to sociocognitive approaches. In the same way, opting for this synthetic expression makes
for more rapid or smooth advance in the thematic progression of a text, thanks to the
anaphoric function: “Yesterday a Panam aircraft fell into the ocean.... The accident had...”).

Now of course, side by side with these advantages in economic expression, nominalisation
clearly plays its part in reducing information: information on the weather, place or the
participants, as well as factors of modalisation, which, once concisely packed into a nominal
expression are not easily recoverable. Also observable is the effect of presenting concrete
processes as abstract entities. The ideological importance is considerable, in the
manipulation of processed information, particularly noticeable in the case of certain types of
anaphor which give a strong interpretative bias to the discourse (“The demonstrators noisily
crossed the quadrangle... Their bursting on the scene/invasion / intrusion brought about...”).

6. Conclusions
In closing this overview of work on stylistics and its function in the ambit of language
sciences, we can briefly and succinctly point out certain trends and orientations within the
discipline, while not in any way claiming to be exhaustive: a) it goes beyond literature,
without ignoring such institutionalised discourse patterns; b) it is situated in the language
framework, where structures allow variation which becomes functionalised and filled out with
individual/social meaning; c) consequently, the amount of observable free variation is
reduced, with the increased systematisation of determining factors, in terms of registers and
genres on the one hand and stylistics on the other; d) contributions from pragmalinguistics
and discourse analysis are crucial to defining the diversity of expressive styles and the
sociolinguistic and ideological relevance of stylistic options.

7. Bibliography
ADAM, J. M. Le style dans la langue. Une reconception de la stylistique. Lausanne: Delachax
et Niestlé, 1997.

BIBER, D. et al. Corpus linguistics. Investigating language structure and use. Cambridge:
C.U.P., 1998.

CORPAS, G. “Criterios generales de clasificación del universo fraseológico de las lenguas, con
ejemplos en español y en inglés”. In: ALVAR, M.; CORPAS, G. (eds.) Diccionarios, frases,
palabras. Malaga: Universidad de Malaga, p. 1591-87.

ENKVIST, N. E. “Text and discourse linguistics, rhetoric, and stylistics”. In: VAN DIJK, T. A.
(ed.) Discourse and literature. New approaches to the analysis of literary genres.
Amsterdam/Philadelphia: John Benjamins,1985, p. 11-38

Noves SL. Revista de Sociolingüística 6

http://www.gencat.cat/llengua/noves
Winter 2003
Pragmatics and stylistics by Vicent Salvador

GARRIDO MEDINA, J. Estilo y texto en la lengua. Madrid: Gredos, 1997.

LÓPEZ GARCÍA, A. “La escuela española de lingüística y la pragmática”, Caplletra 29, p.13-
22, 2000.

MARCO, J. El fil d’Ariadna. Anàlisi estilística i traducció literària. Vic: Eumo, 2002.

MARTIN, J. R. English text. System and structure. Amsterdam, Benjamins, 1992.

MARTIN, J. R. “Cohesion and texture”. In: SCHIFFRIN, D. et al. (eds.), The Handbook of
Discourse Analysis. Oxford: Blackwell, 2001.

PAYRATÓ, L.; ALTURO, N. (eds.) Corpus oral de conversa col·loquial. Materials de treball.
Barcelona: Universitat de Barcelona, 2002.

RACCAH, P. Y. “Un top, sinon rien”. In: SALVADOR, V.; PIQUER, A. (eds.) El discurs
prefabricat. Estudis de fraseologia teòrica i aplicada. Castelló: UJI, 2000, p. 409-421.

SALVADOR, V. “L’estil nominalitzat”, Caplletra 29, 2000, p. 69-82.

SALVADOR, V. Els arxius del discurs. Barcelona: P.A.M, 2000.

VAN DIJK, T. A. “Critical Discourse Analysis as a method for the research in social sciences”.
A: WODAK, R.; MEYER, M. (eds.). Methods of Critical Discourse Analysis. London: Sager,
2001, chap. 6.

VAN DIJK, T. A. Ideology. A multidisciplinary approach. London: Sage, 1998.

Vicent Salvador
Universitat Jaume I
[email protected]

Noves SL. Revista de Sociolingüística 7

http://www.gencat.cat/llengua/noves
Winter 2003