fulltext.pdf

Lodz Papers in Pragmatics 5.1 (2009) / Special Issue on Speech Actions: 85-106

10.2478/v10016-009-0008-8

Iwona Witczak-Plisiecka∗

University of Łódź

SPEECH ACTS AND THE AUTONOMY OF

LINGUISTIC PRAGMATICS

Abstract This paper comments on selected problems of the definition of linguistic

pragmatics with a focus on notions associated with speech act theory in the

tradition of John Langshaw Austin. In more detail it concentrates on the

(ir)relevance of the use of the Austinian categorisation into locution,

illocution, and perlocution in locating a divide in between pragmatics and

semantics, and especially the distinction between the locutionary act and the

illocutionary act and its implications for the definition of pragmatics and its

separation from the semantic theory.

The relation between form and meaning is further briefly reviewed against

dichotomies including the Gricean and neo-Gricean ‘what is said’ versus

‘what is implicated’ or meant, between what can be ‘locuted’, but not said,

and what can be said, but not asserted. These dichotomies are related to the

theoretical commitments as to the accepted operative forces in speech acts,

primarily convention and intention. It is suggested that, roughly, the

development of the speech act theory can be viewed as a process by which

the theory moves away from its originally sociolinguistic orientation

towards a more psychologistic account, which in turn leads towards

diminishing the role of (traditional) semantics and the subsequent

juxtaposition of pragmatics and syntax rather than pragmatics and

semantics.

Keywords

speech act, pragmatics, semantics, illocution, locution, John L. Austin

∗ Department of Pragmatics,

University of Łódź,

Al. Kościuszki 65

90-514 Łódź, Poland

e-mail: [email protected]

Iwona Witczak-Plisiecka

Speech Acts and the Autonomy of Linguistic Pragmatics

86

1. Introduction – formulations of pragmatics

Since the time Morris (1938) identified pragmatics as a research area

complementary to syntax and semantics its scope has been explicitly disputed

among linguists and philosophers of language. Within this debate speech acts have

been in the centre of interest, while speech act theory in the Austinian tradition and

its further elaborations have been typically recognised as the most salient,

uncontroversially pragmatic research area.

1.1. Morris’s division – the classical view

Charles Morris was the first to use the term pragmatics in a rather systematic

technical way, but the problems which must have motivated him were not new. As

is well known, in his attempt to structure the field of semiotics, Morris identified

three main branches:

1) syntactics (today’s ‘syntax’) – the study of the structural “formal relation of

signs to one another”;

2) semantics – the study of “the relations of signs to the objects to which the

signs are applicable”;

3) pragmatics – the study of the contextual “relation of signs to interpreters” (cf.

Morris 1938: 6f.).

Although Morris’s ambitious agenda to integrate all sciences of signs, whether

linguistic or not, was never to be accomplished, his work evidently paved the path

for pragmatic research, however modified and limited in comparison to what he

had envisaged1.

Functional as they are, the categories suggested by Morris are not absolutely

complementary and even in today’s perspective the three areas are still

underspecified bones of contention to the effect that, as can be seen in subsequent

discussion, there have been suggestions to dispose with either pragmatics or

semantics on grounds that the co-existence of the two introduces confusion

detrimental to clarity of linguistic research.

1 It was Morris’s plan and ambition to create a thorough account of signs – the theory of

“semiotic” – and their uses, viz. “semiosis”. His interests lay not just in language, but in art

and other sciences such as logic, mathematics, rhetoric, elements of aesthetics. Under his

interpretation any object can function as a sign if only its presence enables people to take

account of an object or objects which are not present. Thus objects may function as signs by

inviting meaning or rather meaningfulness, while Morris’s pragmatics is the study of the

relation between (any type of) signs and their interpreters/users.


10.2478/v10016-009-0008-8

87

1.2. Further formulations

Since Morris many new formulations of pragmatics have been suggested, most

of which have concentrated on the issue of truth-conditions. This focus on

linguistics as a field whose main interest transcends the traditional (at least until

the mid-20th

century) interest in language as a means of describing the world

provides a linking thread through many otherwise diverse definitions of pragmatics

and integrates varied speech act theoretic research.

Most definitions of pragmatics use the notion of a speech act or a linguistic act,

even if this reference is not directly associated with the theory of speech acts. To

mention just a few, Stalnaker (1972) defines pragmatics as the study of linguistic

acts and the contexts in which they appear, limiting semantics to the study of

propositions; Katz (1977: 14) also decides to limit semantics to the study of

information available in e.g. an anonymous letter situation, maximally free of

particularised context. Gazdar (1979: 2) in his much quoted formula claims that

“Pragmatics = Meaning minus truth conditions” and Levinson subsumed

methodological attempts at defining pragmatics by claiming that

the most promising are the definitions that equate pragmatics with ‘meaning minus

semantics’, or with a theory of language understanding that takes context into

account, in order to complement the contribution that semantics makes to meaning.

(Levinson, 1983: 32)

In other words, sentences should provide (semantic) input, which is further

analysed in contextualised pragmatic studies, which in speech act-theoretic

research may be related to the juxtaposition of form and function, and

consequently, roughly, to the concept of locution vis-à-vis illocution. It should

suggest that the concept of illocution is in the centre of pragmatic research, being

focused on what people do and (attempt to) achieve in what they say. Extending

this view leads to the understanding of pragmatics in, among others, Verschueren’s

sense, i.e. as a perspective on language use, an interest in the relation between form

and its effective use (e.g. Verschueren 1999). This in turn leads to pragmatics

controlling all aspects of meaning which focus on functional communicative use

(from phonetics to discourse and beyond).

Speech act theoretic-oriented researchers (e.g. Barker 2004, 2007), as in fact

many others2, understand semantics as the study of propositional content, while

pragmatics as the study of speech acts in broad communicative structures. It is also

2 Semantic minimalists, e.g. Cappelen and Lepore (2005) advocate the (however

contentious) view that there is space for autonomous semantics, whose data provides input

for pragmatics in that semantically describable forms can be used in a “galaxy” of speech

acts, cf. also Borg 2005.



88

often acknowledged that a semantic input of “what is said” forms the basis for a

pragmatic analysis, which focuses on what is done beyond and by saying.

1.3. Criticism of the relevance of SAT in the pragmatic program

It is worth noting though that the reign of the theory of speech acts in

pragmatics is not absolute. For instance, Deirdre Wilson and Dan Sperber,

relevance theorists, have questioned the value of the theory of speech acts in the

pragmatic program (cf. Sperber and Wilson 1986/1995: 243 ff.), claiming that

although the speech act approach to utterances is

perhaps the most uncontroversial assumption of modern pragmatics […] the vast

range of data that speech act theorists have been concerned with is of no special

interest to pragmatics (Sperber & Wilson 1986/1995: 243).

Sperber and Wilson claim that the assignment of a speech act type is not crucial

in comprehension or interpretation of utterances, both on the level of explicature

and implicature and argue that classification in terms of speech acts plays a minor

(if any) role in such comprehension and interpretation3. As inferential processes

and ostensive communication4 are central issues in the neo-Gricean relevance

theory, the theory of speech acts is seen as practical only on the level of mood

analysis and possibly universal abstract categories, such as “saying”, “telling”, and

“asking”, which can be described with the use of the speech act theoretic apparatus

and for which “type” recognition is an essential part of the comprehension process.

Critical remarks of this kind bring to the foreground the most contentious issues

in the speech act theory itself, which are many, e.g. the status and architecture of

speech act taxonomies, the definition of the speech act and its components or

aspects, the status of uptake, the nature of convention and its status vis-à-vis

intention5, the salience of the institutional aspect, the divide between the

institutional and the non-institutional, the source of the illocutionary force. The

3 One of the examples evoked is an analogy of tennis, which can be played/performed (sic!)

without players knowing theoretical terms (Sperber & Wilson 1986/1995: 243). 4 Types which, according to Sperber & Wilson, need to be ostensively communicated and

recognized include e.g. promising, ordering, thanking, bidding, but others, such as

predicting, denying, claiming, or even warning, do not require any special pragmatic tools to

be interpreted and successfully performed; they do not need to be ostensively

communicated. 5 The evolution of the theory of speech act can in fact be seen as a struggle between the

convention-based and the intention-based approaches; the tension originating from the time

of Peter Strawson’s (1964) article “Intention and convention in speech acts”


10.2478/v10016-009-0008-8

89

tripartite approach to the speech acts through its three (allegedly) distinct aspects,

viz. locution, illocution, and perlocution, rejected in the relevance-theoretic

framework and discussed in this paper, is one of the focal contentious issues.

2. Locution, illocution, and perlocution – the Austinian

perspective

The assumption lying at the core of speech act theory is that any utterance has

two basic aspects: it involves sense and force. The distinction is easily traced back

to Frege, who was the first to comment on it with reference to grammatical mood.

This distinction also evokes Austin’s (1962/1975; 1960) reflection on the nature of

language as a means to perform action. Austin also found it relevant to differentiate

between three aspects of an utterance, corresponding roughly to its form, function,

and effect/result.

Although notoriously underspecified, Austin’s tripartite division of a speech act

into locution, illocution and perlocution has been widely accepted in linguistics and

served as a guideline for demarcating semantics and pragmatics. Evidently, both

the function and the effect of an utterance are explicitly pragmatic in nature, but for

locating the borders of pragmatics, the divide between locution and illocution is

especially important.

In his attempt to thoroughly account for the nature of the speech act Austin

(19621975: 92f.) defines locution as composed of three elements, viz.:

1. Phonetic act (uttering certain noises) → a phone

2. Phatic act (uttering certain words in a certain construction, i.e. noises of

certain type) → a pheme

3. Rhetic act (performance of a pheme with a more or less definite sense and

a more or less definite reference, where sense + reference are equivalent

to “meaning”) → a rheme

The description, typically of Austin, is ambiguous and only indicates that the

overall locution should be (roughly) understood as ‘literal’, semantic, a maximally

decontextualised vehicle of meaning. In Austin’s words it amounts to “saying

something”, which in turn is:

the utterance of certain noises, the utterance of certain words in a certain

construction, and the utterance of them with a certain ‘meaning’ in the favourite

philosophical sense of that word, i.e. with a certain sense and with a certain

reference (Austin 1962/1975: 94).



90

In contrast, illocution corresponds to the function (or force) of an utterance.

Austin also explicitly admits that one locution may be used to perform many

different illocutions (possibly at a time), whose force comes from (again a

notoriously underspecified) convention (Austin 1962/1975: 108)6.

The basic difference between illocution and perlocution is that the illocutionary

act is performed in uttering words, while the perlocutionary by the utterance. It

suggests that there should be a conventional bond between the expression and its

illocution, which is stronger than the bond between the locution and the

corresponding perlocution.

The definition of perlocution is a broad one and focuses on the results that the

act produces. In Austin’s words, it is defined as follows:

Saying something will often, or even normally, produce certain consequential

effects upon the feelings, thoughts, or actions of the audience, or of the speaker, or

of other persons: and it may be done with the design, intention, or purpose of

producing them […] We shall call the performance of an act of this kind the

performance of a ‘perlocutionary’ act. (Austin 1962/1975: 101)

In short, Austin identified locution, illocution, and perlocution within the

speech act as acts done “of saying”, “in saying”, and “by saying”, respectively. He

illustrated the three aspects of the speech act with the following examples cited

after Austin (1962/1975: 101-102) as (1) below:

(1) (E. I)

Act (A) or Locution

He said to me ‘Shoot her!’ meaning by ‘shoot’ shoot and referring by ‘her’ to her

Act (B) or Illocution

He urged (or advised, ordered, &c.) me to shoot her.

Act (C. a) or Perlocution

He persuaded me to shoot her.

Act (C. b)

He got me to (or made me, &c.) shoot her.

(E. 2)

Act (A) or Locution

He said to me ‘You can’t do that’.

Act (B) or Illocution

He protested against my doing it.

Act (C. a) or Perlocution

6 For a critical discussion of the notion of convention and the role of uptake in Austin and

the theory of speech acts, see e.g. Sbisá 2001, 2002, 2007 and 2009 (this volume) and

Harnish 2009 (this volume).


10.2478/v10016-009-0008-8

91

He pulled me up, checked me.

Act (C. b)

He stopped me, brought me to my senses, &c.

He annoyed me.

Evidently, having accepted Morris’s premise that pragmatics focuses on the

relation between linguistic signs and their users, it should hold that locution is

associated with abstract, (maximally) decontextualised meaning which can be

accessed out of context. In other words, locution should be accessible without

information on the context of use with its meaning (at least relatively) stable across

different contexts.

Illocution in turn stands for the (intended) function of an utterance, i.e. in

uttering/producing locution, the speaker performs an act of a certain kind and this

act is done “in” saying; the produced effects, whether envisaged by the speaker or

not, belong to perlocution and are performed “by” saying certain words.

For Austin, the illocutionary force arises beyond words and has its roots in

convention. Once more, Austin’s lecture notes leave little theory on the nature of

such a convention, but it seems reasonable to believe that he referred to a kind of

shared knowledge with regard to the relation of form and meaning of the

expression in question (possibly together with contextual clues). Further theorizing

on speech acts resulted in the introduction of the notion of an illocutionary force

indicating device (IFID), whose many forms7 are indicative of the special and

diverse nature of communicative conventions.

It follows that locution should belong in semantics as it is closely bound to the

form and invites a search for explicitness in that form. As illocution is mainly a

level of assigning categorial belonging, where performative verbs are often used as

labels for functional classes, it should be perceived as “pragmatic”; it is associated

with meaning “beyond words”, beyond the rhetic act, a meaning derived from

varied conventions. Perlocution, focused on envisaged effects, should be purely

pragmatic in nature.

However, this categorisation is by no means simple and reveals numerous

problems. First of all, it is evident that locution “invites” illocution and its

interpretation requires inferential processes which are inherently pragmatic in

nature. What differentiated performative utterances from purely descriptive ones,

according to Austin (at least when he started the theory) was their “misleading”

form, their being “masqueraders”, whose outward appearance often dramatically

departed from the canonical8 form suggested later, but also their being

7 IFID can be realized as a performative verb, but can also be incorporated in the manner in

which an expression is uttered, e.g. in the intonation pattern employed (cf. Searle 1969), or

can reside in lexical items such as the adverbs please (cf. Sadock 1974) or hereby. 8 The canonical form refers to an explicit direct performative utterance characterized by the



92

performative in disguise of descriptivity. In other words, the “literal meaning” of

an expression is not its “real meaning” in many, in fact most cases in linguistic

communication. If the locutionary aspect of a speech act is to be discussed

(semantically) in isolation, it is at the expense of its true meaning. This does not

translate into that the form is not important in communication, but it is clear that

understanding the “literal” is only part of the comprehension process and is simply

not enough for communication.

Illocution, a “theoretical tier” devoted to analysis in terms of metalinguistic

categories cannot be separated from the other two aspects either. Austin indicated

that illocutionary acts invite by convention a response or sequel; for example an

order invites obedience, while a promise invites fulfillment. This description to

some extent resembles his account of perlocutionary acts, which involve either

some kind of achievement of a perlocutionary object (e.g. convincing someone or

persuading) or a “perlocutionary sequel”, e.g. a warning may result in alerting

someone. The “in saying”/”by saying” distinction, as introduced by Austin in his

first lectures, is thus rather unclear, e.g. the consequential effects of “promise” are

under illocution, while those for “persuade” and “protest” are under perlocution.

Evidently, there is a good reason to acknowledge the difference, and the nature of

promising is in a way “more internal” than that of e.g. persuading, but the

theoretical apparatus employed does not seem efficient enough to clearly

categorise performative aspects of utterances. It is as if Austin was able to locate

many important phenomena associated with performativity, but left them at the

stage of early theoretical development, for further refinement in the future.

Already in 1977, Allwood pointed to the fact that also the nature of convention

that Austin might have had in mind varies with the verbs used in particular

utterances. For example, for (2) below:

(2) It is snowing

it is relatively easy to conclude that the locutionary aspect “says something about

the weather”, while the illocutionary involves the utterance force as a statement9.

However, the picture gets more complex and complicated in the case of explicit

performatives as (3) below:

(3) I promise to buy you a drink.

use of the pronoun I, presence of a performative verb, presence (or possibility of inclusion)

of the adverb hereby, present tense, simple aspect, and non-negative form. 9 Although this statement may also, of course, be used to perform different illocutionary

acts, e.g. it may be used as a warning, an invitation to go out, an invitation to stay indoors,

an expression of surprise or satisfaction, and indeed in many other ways.


10.2478/v10016-009-0008-8

93

The illocutionary force in such utterances seems to arise at least partly from the

meaning of the verb, thus its nature (however conventional) is profoundly different

from the force associated with (2). For (3) both the fact that the verb “promise” is

used and the form of the utterance as a statement are important. Although the

meaning of (3) can vary across contexts, at least part of this meaning, in the sense

of function, is construed thanks to the convention associated with the verb

“promise”, i.e. input on the locutionary level. This varying degree of “lexical

meaning” intervention is a good indication that the illocutionary level dependence

on the locutionary, together with other elements of context, is much varied and

does not allow for a clear description of illocution (or locution) as a consistent

category. In many contexts the lexical (locutionary) content of utterances functions

as a trigger for understanding the illocutionary force, cf. (4) and (5) below:

(4) “I do”

(5) “Two no trumps”

In such cases the locutionary element is important but its reading crucially

depends on extralinguistic context. In other cases, elaborate locution secures the

“proper” reception of the force, as e.g. in the case of legal documents exercising a

real performative force in a relevant legal culture.

Legal language itself provides many interesting examples for analysis in the

speech act-theoretic framework, highlighting many of its problems, including the

problem of reliance on form. First of all, it opens a whole realm of contexts, from

formulaic expressions of the courtroom through carefully designed verbose

language of legal documents to fairly informal client-lawyer consultations, but

even if only the most explicit language of legal documents is taken into

consideration, the nature of the illocutionary force remains an evasive one. Among

others, it shows that the difference between the institutional and non-institutional,

or institutionally conventional and just linguistically conventional is not an easy

one. For example, should the text of the Decree on Martial Law passed in Poland

in 198110 as performed by the Polish Piwnica pod Baranami cabaret group be

excluded altogether from analysis? Being a kind of (artistic) performance, it must

be defined as a non-communicative type of linguistic action. However, its use may

serve (and did indeed serve in the past) many different purposes, some of which

seem communicative, e.g. communicating judgement on the party in power,

building rapport among people, communicating reassurance, ridiculing the hostile.

One may argue that these “functions” are of perlocutionary nature, but their status

clearly depends on theoretical commitments. If what makes a speech act

communicative is the speaker’s intention to convey some content (as many

10 Cited as (Dekret … ), cf. also Witczak-Plisiecka 2001.



94

researchers argue should be the case, e.g. Bach and Harnish 1979), it should follow

that whether an artistic performance, reciting of a poem, or uttering the christening

formula, is a communicative act or not depends on the speaker’s state of mind.

Much like a secret code, the act may carry different illocutionary forces (if any at

all) for different audiences, according to the speaker’s plan; it also may result in

different perlocutionary effects just because it may activate different contextual

traits for different people. As it is clearly not possible to build an all-inclusive

theory of speech acts which should explain all communicative phenomena, the

architecture of the theory must depend on the researcher’s commitment. Bach and

Harnish (1979) decided to exclude conventional speech acts from their analysis

and concentrated on building a theory of unambiguously communicative ones.

Other theorists, typically more sociology-oriented, such as Butler (e.g. 1993, 2997)

and Hornsby (e.g. 2000), decided to concentrate on a very general nature of

linguistic action, viz. its performative character, the ability to modify, maintain,

and create extralinguistic reality. Millikan (e.g. 2005), although much different in

her scientific commitments, is a further example of focus on performativity without

minute bottom-up investigations and clear-cut taxonomies. Such analyses, although

dramatically diverse, all explicitly subscribe to the Austinian heritage with

simultaneous disregard of at least some aspects of the tripartite distinction and tacit

promotion of the holistic “illocutionary” perspective.

In the context of the role of pragmatics, an interesting side effect of such

approaches can be seen in Hornsby’s (2000) claim that “semantics” (and traditional

philosophy of language) as a research area defined over the past is typically male-

governed and anti-feministic even in its approach to research. The main point in

this criticism is relevant in the speech act-theoretic discussion as it is claimed that

the feminine view is more holistic, while it is a natural male tendency to dissect

and classify. Incidentally, Borg (2004: 12f.) expresses a related view (although

with no feminist implications) with regard to the semantics-pragmatics divide. She

separates the two disciplines along the lines of reasoning where semantics is seen

as “modular” and “formal”, while pragmatics, focused on processes of

communication, necessarily “global” and as “informationally unencapsulated” as

possible (Borg 2004: 90). Leaving the basis for the gender differences aside, this

view of the distinction still seems a rather important trait for navigation among

diverse speech act theories and perception of a certain downplay of the role of

locution in many approaches. It also sounds as a comeback to Austin’s original

reflection, being an explicit acknowledgement of the variety of situations where

speaking is doing under many disguises.

Returning to perlocution alone, the category merges with the “illocutionary”

not only through metalinguistic illocutionary labels, but also in e.g. the analysis of

“perlocutionary verbs”, i.e. verbs describing (potential or actual) effects of

linguistic action (cf. e.g. apologise, welcome, thank, entice, congratulate). Some


10.2478/v10016-009-0008-8

95

such verbs, especially expressives are as if a link between illocution and

perlocution in that although they are typically used in direct performatives (I thank

…, Let me apologise …), they at the same time “describe” (intended)

perlocutionary effects. Thus, it may often be practical to use a perlocutionary

perspective on the whole of the speech act. For example, Kalisz (1993: 33)

suggested that politeness principles formulated by Robin Lakoff and Leech should

be integrated in research on perlocutionary acts, as being intended, target- and

addressee-oriented, they are more relevant on the level of perlocution than

illocution.

In many cases there is a distinction which could approximately be identified

with reference to conventional consequences on the one hand, and real, often not

directly foreseeable effects on the other. This distinction, however, does not map

systematically onto the illocutionary and perlocutionary aspects either.

Austin himself realized that his categorization may produce problems and

mentioned the possibility of a terminological confusion, which we could dub

metalinguistic. For example, he explicitly admitted that:

The expression ‘meaning’ & ‘use of sentence’ can blur the distinction between the

locutionary and illocutionary acts […;] to speak of the ‘use’ of language can […]

blur the distinction between the illocutionary and perlocutionary acts

(Austin 1962/1975: 103)

He further referred to the concept of “the use” of language (Austin 1962/1975:

104) to show that while “arguing” and “warning” are roughly conventional and

could be successfully used in explicit performative formulae (in the “in doing x”

mode), concepts (and verbs corresponding to them) such as “persuading”,

“alarming”, and “insinuating” are different in that they are not used in explicit

performatives and describe effects rather than acts (in the “by doing x” mode). This

remark is consistent with the discussion above, which indicates the need for

analysis of “perlocutionary verbs”, an analysis which, in turn, is naturally very

“semantic” in nature.

Austin’s acknowledgement of the methodological difficulty is a good

indication, a reaffirmation indeed, that he was skeptical with regard to the

construction of a thorough theory of speech acts (or linguistic meaning). It is

evident that his theory is better seen as a construct which points to (partly)

systematic communicative (rather than only linguistic) phenomena, which had

been neglected prior to his formulation of the theory11

. In this light the division into

11 Cf. “We must learn to run before we can walk. If we never made mistakes how should we

correct them?” (Austin 1962/1975: 12) and the very first words of William James lectures

published as How to Do Things with Words: “What I shall have to say here is neither

difficult nor contentious; the only merit I should like to claim for it is that of being true, at



96

locution, illocution, and perlocution, just like the division into semantics and

pragmatics, is better seen as reflective of emphasis on specific aspects in linguistic

communication, and not an attempt at dissecting analysis of speech acts into

mutually exclusive areas.

3. Alternative accounts of the internal structure of the speech act

and their implications for a definition of pragmatics

Dissatisfaction with (often distorted) Austin’s model resulted in alternative

proposals as to the internal structure of the speech act and the status of its

component parts or aspects. The most salient characteristics of selected competing

accounts are briefly sketched below.

3.1. Searle’s theory of speech acts

In his elaboration of the theory of speech act, Searle (1969, 1979) rejected

Austin’s category of locution altogether. His main criticism of the category was

based on the fact that locutionary acts could not be sufficiently differentiated from

illocutionary acts of saying. Slightly disregarding the fact that Austin meant all

three categories, i.e. locution, illocution, and perlocution, to be aspects of one act

rather than distinct acts, Searle’s suggestion may clearly be perceived as yet

another attempt to put forward a more precise description of the nature of the

speech act.

Among constituents of speech acts Searle mentions “performing utterance

acts”, i.e. uttering words, “performing propositional acts”, i.e. referring and

predicating, “performing illocutionary acts”, e.g. stating, commanding, and finally

performing perlocutionary acts, i.e. consequences or effects (Searle 1969: 24-5). It

seems that in suggesting his architecture, Searle renamed Austin’s phonetic and

phatic acts as “utterance”, and turned the rhetic act into the “propositional” one,

possibly blurring some aspects of the Austinian phatic act into both new categories.

In Searle’s version of speech act theory, the act is primarily accounted for in

terms of the propositional content conditions and the illocutionary point. Searle

also reintroduces Austin’s classification of speech acts with some amendments and

discusses the types in terms of abstract categories.

least in parts. The phenomenon to be discussed is very widespread and obvious, and it

cannot fail to have been already noticed, at least here and there, but I have not found

attention paid to it specifically” (Austin 1962/1975: 1).


10.2478/v10016-009-0008-8

97

Being a philosopher of language, Searle is not directly interested in the scope of

pragmatic research, but his theory treated as data may be important for a

demarcation of pragmatics.

First of all, the conditions set for successful accomplishment of an act,

conditions which are an integral part of his theory, are a mixture of (traditionally)

semantic and pragmatic notions. They make use of both lexical meaning of the

performative verb in question (if present), or a label accepted for the category, and

elements of context important for successfulness of the act. In Speech Acts (1969),

Searle just sketches conditions for the types of act and gives only one full account

of an act, viz. description of the act of promising constituting Chapter 3. The

conditions envisaged for an act to be a promise include, among others, expressing a

proposition, both the speaker’s and hearer’s attitudes, the speaker’s sincere

intention to act, as well as the speaker’s intention to accept the obligation to act. In

summary, it is not possible to divide semantic and pragmatic aspects in Searle’s

theory. The distinctions within the internal structure of the act may be used, but

separation of the “utterance act” plus the “propositional act” from the illocutionary

and the perlocutionary act produces similar difficulties as the Austinian distinction

into locution, illocution, and perlocution. In Searle, these difficulties are even

reinforced by the presence of conditions which mix lexical knowledge with world

knowledge, linguistic considerations with context intervention.

It is also noteworthy in the context that Searle (1969) explicitly states that the

theory of speech acts is a theory of linguistic competence and not performance, i.e.

in deSaussurean terms it is langue that is the object of research of the theory of

speech act, not parole. Searle’s claim shows his belief that the theory of speech

acts is “possible” as a (probably) universal theory of linguistic action; it also shows

that pragmatics, insofar as Searle can be seen as a pragmatician, can be “formal”.

3.2. What is said vs. what is meant

One of the most widespread views on the semantics–pragmatics divide is

associated with Paul H. Grice and his notions of “what is said” as opposed to “what

is implicated”. Despite some bias towards literality, Searle explicitly subscribes to

the Gricean school (e.g. Searle 1969: 43f.) and in many respects, although not a

speech act theorist, Grice can be seen as development on Austinian ideas in that he

tried to systematise the mechanisms of successful communication and his research

has exerted strong influence on the developments in the speech act theory,

strengthening especially its inferential, intention-based (as opposed to convention-

based) mainstream.

It may appear that the Gricean distinction between “what is said” and “what is

meant” should correspond to the Austinian notions of locution and illocution



98

(possibly enriched with perlocution), thus lending a reliable criterion for the

semantics-pragmatics borderline. For Grice the meaning of an utterance, i.e. what

is implicated and meant, arises not so much from the form, i.e. “what is said”

(which seems to correspond to Austin’s locutionary acts performed “in saying”),

but is brought about “by the saying of what is said” and “putting it that way” (cf.

Grice 1989: 39f.), which in turn seem close to the Austinian illocution12

.

However, a closer look at processes involved at the level of “what is said” or

(here) locution proves that it also involves typically pragmatic phenomena. Even

apparently explicit utterances, such as Austin’s examples quoted above and (6)

below,

(6) I name this ship the Queen Elisabeth

require typically pragmatic processes. For example, an interpretation of (6) and the

expressions “Shoot her” and “You can’t do that”, quoted in (1) all require

processes of assigning reference to deictic expressions, and disambiguation of

superficially explicit forms. Forms such as (7) below, although superficially

explicit, can only be interpreted in context.

(7) Coffee would keep me awake

Whether (7) is to be taken as “yes” or “no” in response to an offer of a cup of

coffee does not depend on the “semantic” meaning of its component parts or even

whether it can be understood as a sentence, but has roots in extralinguistic context.

Other much quoted examples which show context sensitivity of meaning and the

non-independence of truth-theoretic semantics include e.g.:

(8) It is raining

(9) He is too old.

Even within maximally (minimalist) semantic approaches, the assignment of the

truth value to the sentences above requires not only assigning reference, but

extralinguistic knowledge (e.g. information on the place of reference for (8); type

of activity for (9)).

Literature from the areas of pragmatics, semantics, and philosophy of language

provides many examples and ample evidence for the fact that “semantic” content

always underdetermines speaker meaning (cf. e.g. Bach 2001, Carston 2002,

2008). Although using different categories and names (e.g. “semantic enrichment”,

12 Cf. also Kissine (2009) for an inferential account of speech acts and the discussion of

levels of meaning and the notion of illocution vis-à-vis “what is said”.


10.2478/v10016-009-0008-8

99

Gricean particularised and generalised implicature, “non-literal” readings, e.g.

related to humour, irony and sarcasm, “polite” indirectness, “default” meaning,

etc.), the accounts of phenomena associated with underdeterminacy of linguistic

expressions share interest in the nature of meaning and the mysterious ways in

which meaning can be conveyed in context.

It follows that the locution–illocution/perlocution divide does not provide

means for posing the semantics–pragmatics distinction; the distinction cannot be

systematically mapped onto the relation between “what is said” and “what is

implicated” either. The problem seems to invite conclusion that there is no clear

semantics–pragmatics distinction at all, which in turn may be an invitation to

dispose of either of the concepts.

3.3. What is locuted vs. what is said

There are theorists who tried to refine both the Austinian distinctions and the

problems invited by the locution-illocution-perlocution issue. For example, Korta

and Perry (2007) suggest that locution is a practical notion, but needs further sub-

categorisation or amendment as it is easy to show the difference between “what is

locuted” and “what is said”. In their account, it seems, locution corresponds to

dictionary, semantic, explicit meaning, while the notion of “what is said” (contrary

to Grice) to what can be identified roughly with hearer meaning.

Korta and Perry argue that locution can be, and in fact often is, identified with

“what is said” or the propositional content in the sense of Searle, but only in cases

of successful communication. The cases of ‘non-successful’ (or less successful)

communication in which the difference between “what is locuted” and “what is

said” is clearly visible include e.g. undetected (but also conveyed) irony, instances

of embedded propositions, unexpected interpretation on the part of the audience

with a subsequent clash between speaker meaning and hearer meaning.

It seems that Korta and Perry’s main argument translates into that “what is

said” bears kinship relationship to illocution, and is especially closely related to

assertion; that it is “a forensic concept” in that speaker can be taken responsible for

“what was said” (rather than intended), and that it is also characterised by

sensitivity to information speaker wants to convey. In short, the argument seems to

be that “what is said” can (dramatically) transcend referential meaning. For

example,

(7) Aristotle enjoyed philosophy



100

can be “locuted” about Aristotle, the Greek philosopher, while “said” (i.e.

interpreted as referring to) e.g. Aristotle Onassis in certain contexts (cf. Korta and

Perry 2007: 176 f.). In a similar manner in (somewhat obsolete already) (8) below,

(8) If Hillary is elected, Bill will enjoy his return to the White House.

neither of the propositions ‘Hillary being elected’ and ‘Bill enjoying his return to

the White House’ is “said” or “asserted”; they are merely “locuted”. In (9),

(9) John is turning red

the speaker’s intention may be “literal”, i.e. to suggest that e.g. John’s face is

turning red in colour, but the audience’s interpretation may be that John is

becoming a Communist, for which judgement the speaker may be believed

“responsible”.

In all these examples there is a clash between “what is locuted” and “what is

said” in the sense of communicated, and sometimes “what is locuted” may be

understood in a way that the speaker neither envisaged, nor intended.

Applying Perry’s (2001) reflexive-referential theory (RRT), Korta and Perry

(2007) do defend the notion of a “locutionary act” with the provision that it should

be broken down into further aspects, especially into “intended” reflexive

locutionary content and the actual locutionary content, i.e. its reception in context.

A side effect of their account is attention drawn to a kind of “uptake”, a notion

considered by Austin and corresponding to recognition (whether “correct” or false)

of the speaker’s communicative intention on the part of the audience13

, which,

however, was originally introduced as a proof for an illocutionary act. As Korta

and Perry (2007: 175) claim about natural communication:

our ordinary concept of what is said is to some extent responsive to uptake on the

part of the audience […] What is said seems to have both illocutionary and

perlocutionary aspects

It is evident that Korta and Perry’s concept of “what is said” is by no means

“semantic” in that it refers directly to typically pragmatic aspects of highly

13 The problem of uptake and its status in the theory of speech acts is also relevant for

defining the scope of pragmatics and the semantics-pragmatics distinction. In the (more)

convention-based theories, an uptake is usually seen as a necessary element for there being

an illocutionary act at all. In theories focused on intention, uptake, if important, can be

found in considerations on the “intended” nature of the act, i.e. what the speaker intends to

perform or her envisaged effects on the hearer and the world (cf. also Sbisà 2009, this

volume).


10.2478/v10016-009-0008-8

101

contextualised meaning. For them, locution seems to be a kind of (possibly

semantic) template for meaning derived in context. In this way they combine in

“what is said” all aspects of contextualised reading of a linguistic expression,

elements which others dub as “semantic enrichment”, implicature, default

meaning, etc., notions troublesome for the distinction between “what is said” and

“what is meant” as discussed above. A direct consequence of allowing such

phenomena into the Austinian locution is that the concept can by no means be

identified as purely semantic. Thus, on Korta and Perry’s account locution not only

invites but also incorporates “pragmatic” meaning, which further does not allow

for a simple use of the concept in identification of the semantics-pragmatics divide.

It seems that if such a divide were to be suggested at all, it should be hidden

somewhere not even between “what is locuted” and “what is said”, but rather in

between the intended reflexive locutionary content (possibly inside the speaker’s

mind) and the actual locutionary content, which seems to be open to largely

unrestricted enrichments.

It is noteworthy that often language researchers use the concept of “what is

said” with the meaning which Korta and Perry render as “what is locuted”. It is

“what is said” (mainly of course in Gricean, neo-Gricean and post-Gricean theories

and approaches) that corresponds to “general”, “literal” meaning, while Korta and

Perry’s “what is locuted” is normally rendered as “what is communicated”,

implied, etc14.

There are more contrasts related to “what is locuted” and “what is said”, e.g.

the distinction between “what is said” and “what is asserted”, which concentrates

on the potential or actual difference between “what is said”, i.e. literal meaning of

the expression uttered, and the assertion (a kind of truthful commitment) which is

associated with it15

.

Although typically speech act accounts make use of the label “Assertion”,

“Assertoric”, or “Assertive”16

, as a cover term for a (more descriptive) speech act

type, it seems that the category of assertion is of a special kind in that it permeates

all discourse and in that sense can be at least an element or aspect of all other

(illocutionary) acts.

14 Similar examples of non-literal meaning with elements of conscious deception (silent

disclaimers) on the part of the speaker are discussed with reference to “what is said”

(equivalent to “locuted” as used in Korta and Perry) and “what is communicated”

(equivalent to Korta and Perry’s “what is said”) in Horn 2006. 15 But see Searle (1969 141f.) for a discussion of the assertion fallacy, which exposes (and

criticises) the widespread view that the content of an assertion can be derived from the

semantic value of the words which compose the utterance in question. 16 Cf. e.g. Stalnaker 1978, Searle’s (1979) collection.



102

4. Conclusions

Although the theory of speech acts paved a way for pragmatic research building

it on the basis of prior semantic interests, its categories and principles are not

sufficient for demarcating the borderline for pragmatics, especially if it is to be

found vis-à-vis semantics. The pragmatic perspective manifests itself as a heavy-

handed one, to use Stephen Neale’s (2007) phrase, which permeates all levels of

linguistic analysis. This should not be surprising as linguistic activity is necessarily

a purposeful one and easily lends itself to functional research. There is a temptation

thus to conceive of pragmatics as a perspective on language in use, a perspective

which tends towards hegemony over other approaches to language, growing in

dominance over not only semantics, but also e.g. phonetics and phonology,

morphology, etc. It seems that the strongest relative to pragmatic research can be

seen in the mediating syntax (cf. Barker 2004, Carston 2008, and incidentally

newer approaches within generative grammar)17

, which is not surprising either as

the relationship between these two areas has direct resemblance to the well-

established relation between form and meaning. This approach leaves the

contentious problem of where semantics ends and pragmatics begins (and vice

versa) outside (or on the outskirts of) the linguistic theory.

In a much philosophical manner, Austin was rather sceptical as to the

possibility of constructing a clear theory of linguistic communication. It seems he

was more concerned with pointing to the contemporary theoretical inadequacies

and negligence with respect to non-truth evaluable expressions. Bringing to light

ordinary language and its nature was the revolution that he had envisaged and, as

can be seen in his lectures, was evidently proud to introduce. Although he did not

use the term pragmatics, his work has evidently proved seminal in the development

of the field.

Speech act theory (or rather contemporary theories) shows that there are no

clear cut borders related to meaning, no absolute categories; that we need to be

satisfied with non-completion and openness, and eventually with what can be seen

as “the magic of pragmatics” (Neale 2007). This “magic” evidently is a linking

thread through all speech act theoretic research and in fact all research

concentrated on linguistic meaning. Whether it is concluded that the force of an

17 But cf. Kearns (2006), focusing on illocutionary logic and assertive conditionals, who

claims that a sharp distinction can be seen between syntax and semantics, which in turn is

blurred with pragmatics. It is evident that what he means correlates with authors such as

Barker (2004) or (less decisively) Carston (2008), who give prominence to the notion of

pragmatics17. By the author’s own admission, illocutionary force is in this account part of

the semantic structure with syntax providing “clues” for semantic interpretation. Even in this

approach one form can be used to perform many different acts, i.e. acts with different

semantic structures, depending on the context.


10.2478/v10016-009-0008-8

103

utterance comes from convention, or intention, or possibly some other element, it

is always the evasive relation between form and meaning, and especially the

(“real”) contextualised meaning which escapes formalisation – the “shared”,

“default” aspect whose presence can be felt, but not readily formalised in theory.

Irrespective of the theoretical school, linguists acknowledge the (natural)

presence of a mereology (Barker 2004), galaxy (Cappelen & Lepore 2005) of

speech acts18

and of the fact that communication presupposes their existence

(Searle e.g. 1969: 16, 47).

It seems that the war over speech acts is more a war about the status of

semantics with e.g. “insensitive” semantic minimalists (e.g. Cappelen & Lepore

2005, Borg 2004) fighting for granting it a position of power and independence,

and contextualists (e.g. Recanati 2001, 2002, 2004, relevance theorists) fighting for

the view that sentences are only meaningful in the context of speech acts in which

they appear. The semantics–pragmatics divide can only be seen as a functional (not

to say a “pragmatic”) approximation, descriptive of research focus (and as such

should indeed be maintained). However, even in cases maximally free of

contextual variation, there will always be elements of pragmatic destabilising

intrusion at play, aggravated in theoretical divagations by the Wittgensteinian-style

underdeterminacy of linguistic metalanguage.

In conclusion, it is evident that Austin’s tripartite division of the speech act into

its most salient aspects, although valuable for research purposes, does not allow for

defining pragmatics vis-à-vis semantics. All aspects of the speech act, even the

most “semantic” locutionary aspect, involve inferential processes and contextual

intervention which are pragmatic in nature and are reflective of the Morris-style

form/sign–user interrelations.

Bibliography

Allwood, Jens. “A Critical Look at Speech Act Theory” In Logic, Pragmatics and

Grammar edited by Dahl, 53-99. Lund: Studentlitteatur, 1977.

Austin, John. Philosophical Papers. Oxford: Oxford University Press, 1960.

—. How to Do Things with Words. Oxford: Oxford University Press, 1962/1975

2nd

ed.

Bach, Kent. “You don’t say?” Synthese 127 (2001): 11-31.

Bach, Kent and Robert Harnish. Linguistic Communication and Speech Acts.

Cambridge, MA: MIT Press, 1979.

18 Austin claimed that the number of performative verbs alone reaches “the order of the third

power of 10” (1962/1975:150)



104

Barker, Stephen. Renewing Meaning: A Speech-Act Theoretic Approach. Oxford:

Clarendon Press, 2004.

—. “Semantics Beyond the Distinction between Sense and Force”. In John Searle's

Philosophy of Language: Force, Meaning and Mind, edited by Savas L.

Tsohatzidis, 190-210. Cambridge: Cambridge University Press, 2007.

Borg, Emma. Minimal Semantics. Oxford: Oxford University Press, 2004.

—. “Saying what you mean: Unarticulated constituents and communication” In

Ellipsis and Nonsentential Speech. Edited by R. Elugardo & R. Stainton, 237-

262. Dordrecht: Springer, 2005, doi: 10.1007/1-4020-2301-4_13.

Butler, Judith. Bodies That Matter: On the Discursive Limits of ‘‘Sex’’. Routledge,

New York, London, 1993.

—. Excitable Speech: A Politics of the Performative. Routledge, New York,

London, 1997.

Cappellen, Herman & Ernie Lepore. Insensitive Semantics. A defence of Semantic

Minimalism and Speech Act Pluralism. Oxford: Blackwell, 2005.

Carston, Robyn. Thoughts and utterances. The Pragmatics of Explicit

Communication. Oxford: Blackwell, 2002.

—. “Linguistic Communication and the semantics/pragmatics distinction”

Synthese. Vol. 165/3 (2008), 321-346, doi: 10.1007/s11229-007-9191-8.

Dekret z dnia 12 grudnia 1981 r. o stanie wojennym (Decree on martial law in

Poland of 12 December 1981). Dz.U. 1981 nr 29 poz. 154.

http://isap.sejm.gov.pl/DetailsServlet?id=WDU19810290154.

Gazdar, Gerald. Pragmatics: Implicature, Presupposition and Logical Form. New

York: Academic Press, 1979.

Grice, Herbert P. Studies in the Way of Words. Cambridge, Mass: Harvard

University Press, 1989.

Harnish, Robert M. “Internalism and externalism in speech act theory”. Lodz Papers

in Pragmatics. Special Issue on Speech Actions, edited by Iwona Witczak-

Plisiecka & Maciej Witek. 5.1 (2009; this volume): 9-31, doi: 10.2478/v10016-

009-0001-2.

Horn, Laurence R. “More issues in neo- and post-Gricean pragmatics”.

Intercultural Pragmatics. 3.1 (2006): 81-93, doi: 10.1515/IP.2006.004.

Hornsby, Jennifer. “Feminism in Philosophy of Language: Communicative Speech

Acts”. In: The Cambridge Companion to Feminism in Philosophy, edited by

Miranda Fricker & Jennifer Hornsby. Cambridge: Cambridge University Press,

87-106, 2000.

Kalisz, Roman. Pragmatyka językowa [Linguistics Pragmatics]. Gdańsk:

Wydawnictwo Uniwersytetu Gdańskiego, 1993.

Katz, Jerald. Propositional Structure and Illocutionary Force. New York: Crowell,

1977.


10.2478/v10016-009-0008-8

105

Kearns, John T. “Conditional assertion, denial, and supposition as illocutionary

acts”. Linguistics and Phillosophy 29 (4) 2006, 455-485, doi: 10.1007/s10988-

006-0007-y.

Kissine, Mikhail. “Illocutionary Forces and What is Said” Mind & Language, Vol.

24, No. 1 (2009), 122-138, doi: 10.1111/j.1468-0017.2008.01356.x.

Korta, Kepa and John Perry. “How to say things with words”. In John Searle's

Philosophy of Language: Force, Meaning and Mind, edited by Savas L.

Tsohatzidis, 169-189. Cambridge: Cambridge University Press, 2007.

Levinson, Stephen. Pragmatics. Cambridge: Cambridge University Press, 1983.

Millikan, Ruth. “Proper function and convention in speech acts.” In Language: A

Biological Model, 139-165. Oxford: Oxford University Press, 2005, doi:

10.1093/0199284768.003.0008.

Morris, Charles. Foundations of the Theory of Signs. Chicago: Chicago University

Press, 1938.

Neale, Stephen. 2007 “Heavy Hands, Magic, and Scene-Reading Traps” EUJAP,

Vol. 3, No. 2, 77-132.

Perry, John. Reference and Reflexivity. Stanford: CSLI Publications, 2001.

Recanati, François. “What is said.” Synthese 128 (2001): 75-91.

—. “Does linguistic communication rest on inference?” Mind and Language 17

(2002): 105-126, doi: 10.1111/1468-0017.00191.

—. Literal Meaning. Cambridge: Cambridge University Press, 2004.

Sadock, Jerrold. Toward a linguistic theory of speech acts. New York: Academic

Press, 1974.

Sbisà, Marina. “Illocutionary force and degrees of strength in language use”.

Journal of Pragmatics 33 (2001): 1791-1814. doi:10.1016/S0378-

2166(00)00060-6.

—. “Speech acts in context.” Language and Communication 22 (2002): 421-436, doi:

10.1016/S0271-5309(02)00018-6.

—. “How to read Austin”. Pragmatics 17 (2007): 461-473.

—. “Uptake and Conventionality in Illocution” Lodz Papers in Pragmatics. Special

Issue on Speech Actions, edited by Iwona Witczak-Plisiecka & Maciej Witek.

5.1 (2009; this volume): 33-52, doi: 10.2478/v10016-009-0003-0.

Searle, John (1969) Speech Acts. An Essay in the Philosophy of Language.

Cambridge: Cambridge University Press, 1969.

—. Expression and Meaning. Cambridge: Cambridge University Press, 1979.

Sperber, Dan and Deirdre Wilson. Relevance: Communication and Cognition. 2nd

ed.

London: Blackwell, 1986/1995.

Stalnaker, Robert C. “Pragmatics” In: G. Harman and D. Davidson (eds.):

Semantics of Natural Language, 380-397, Dordrecht: Reidel, 1972.



106

Stalnaker, Robert. “Assertion.” In Syntax and Semantics 9: Pragmatics, edited by

Peter Cole, 315-332. New York: Academic Press, 1978, doi:

10.1093/0198237073.003.0005.

Strawson, Peter F. “Intention and convention in speech acts.” Philosophical Review 73

(1964): 439-460, doi: 10.2307/2183301.

Verschueren, Jeff, Understanding Pragmatics, London and New York: Arnold,

1999.

Witczak-Plisiecka, Iwona. Semantic and Pragmatic Aspects of Speech Acts in

English Legal Texts, unpublished PhD thesis, University of Lodz, Poland, 2001.

About the Author

Iwona Witczak-Plisiecka received her MPhil in Linguistics (1993) from

Trinity College, Dublin University, Ireland and her MA (1992) in English

studies and PhD (2001) in English linguistics from the University of Łódź,

Poland, where she is currently affiliated. Her research interests are primarily

in the nature of meaning in natural language, the area of the semantics-

pragmatics interface and the philosophy of language with focus on

performativity. A freelance legal translator, she is also interested in the

relation between language and the law, problems of linguistic and regional

identity, minority language rights, and the theory of (legal) translation.

She is editor of Lodz Papers of Pragmatics, member of advisory board of

JoSTrans, the journal of specialised translation, and of editorial board of

Comparative Legilinguistics and International Review of Pragmatics.

She is (co)editor of nine collections and author of Language, Law and

Speech Acts: Pragmatic meaning in English legal texts (Łódź 2007).

Date post:	28-Nov-2015
Category:	Documents
Upload:	caina-silva-coutinho
View:	13 times
Download:	1 times

fulltext.pdf

Documents