Discontent with speech act research is no news . In particular, methods of data collection have been a longstanding cause for concern but weaknesses in the coding systems have also been identified, for instance in Meier’s (1998) critique of apology research. I have come to think that they have not usually reached far enough to get at the more fundamental difficulties with speech act research. This is because for the most part, the critiques have been methodological rather than theoretical, without scrutinizing the theories behind the methods. On the other hand, critiques of speech act theory, of which there are many, have often taken issue with the theory’s philosophical and theoretical-linguistic foundations rather than its empirical grounding. My point is that speech act research needs to examine its theoretical premises Notably, for the past 30 years, conversation analysts have been vocal critics of speech act theory though less of data-based speech act research
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Discontent with speech act research is no news
. In particular, methods of data collection
have been a longstanding cause for concern but weaknesses in the coding systems have also been
identified, for instance in Meier’s (1998) critique of apology research.
I have
come to think that they have not usually reached far enough to get at the
more fundamental difficulties with speech act research. This is because for
the most part, the critiques have been methodological rather than theoretical,
without scrutinizing the theories behind the methods. On the other hand,
critiques of speech act theory, of which there are many, have often taken
issue with the theory’s philosophical and theoretical-linguistic foundations
rather than its empirical grounding.
My point is that speech act research needs
to examine its theoretical premises
Notably, for the past 30 years, conversation analysts have been vocal
critics of speech act theory though less of data-based speech act research
critiques have prompted rebuttals by speech act researchers. For instance,
Schegloff’s (1988) demonstration that speech act theory cannot handle
sequential organization and in particular pre-sequences has been countered
by invoking a distinction between communicative and interactional acts (Van
Rees, 1992, drawing on Edmondson, 1981) and the analysis of indirect
speech acts as questions about preparatory conditions (Cooren, 2005).
Objections to applying classical speech act theory to the analysis of
interaction have been raised from diverse theoretical perspectives, and
approaches that extend speech act theory so as to accommodate interaction
have been proposed (e.g., Geiss, 1995; Thomas, 1995). Searle (1992) himself has taken
a skeptical position on such a project, as he contends —in direct opposition
to CA— that conversation is void of constitutive rules and uptake not
constitutive of illocutions. Of the different proposals for the analysis of
speech acts in interaction, conversation analysis (CA) has accrued by far the
largest and most coherent cumulative body of research, lending high
credibility to its theoretical foundations and methodology. CA therefore
recommends itself not only as a lens for critical scrutiny of speech act
research but provides a well documented alternative.
My
discussion will be organized around three central concepts: action, meaning,
and context. I will examine how these concepts are typically handled in
speech act research and conversation analysis
Action
In the theories that speech act research most strongly draws on, speech
acts are conceptualized as rational action. In the sense relevant for this
discussion, rationality is defined as a goal-directed means-ends relationship
(e.g., Bilmes, 1986, for detailed discussion). A rational actor is one who
chooses her means so that they meet the actor’s intended goals, whereby
the goal-directed choices maximize benefits and minimize costs to the actor.
Under this view, social action is a matter of rational choice.
The rational actor model maps directly onto the two most influential
theories in speech act research, Searle’s theory of speech acts (e.g., 1969,
1975, 1976, 1979) and Brown and Levinson’s politeness theory (1987)3.
Although problems with both theories are widely acknowledged, they enjoy
continued prominence, as evident, for instance, in Holtgraves’ recent book
on Language as Social Action (2002). In Searle’s (1969) speech act theory,
illocutionary acts, the core objects of his theory, are defined as a category of
speaker’s intentions expressed by means of linguistic resources. Given
certain conditions, an utterance “achiev[es] the intention to produce a certain
illocutionary effect in the hearer. (...) The hearer’s understanding the
utterance will simply consist of those intentions being achieved” (1969,
p. 48). Building on and modifying Grice’s intention-based account of meaning
(1957)4, Searle elaborates the notion of “reflexive intention” that is so central
to both Grice’s and Searle’s theories: “the speaker S intends to produce an
illocutionary effect IE in the hearer by means of getting H to recognize S’s
intention to produce IE” (1969, p. 47).5 Performing speech acts is thus
theorized as a means-end relationship where speakers convey their
propositional and illocutionary goals by means of linguistic expressions.6
Importantly, for most classes of speech acts, the actual uptake by the hearer,
or perlocutionary effect, is not encompassed by reflexive intention and the
achievement of an illocutionary point.
From Brown and Levinson’s politeness theory, speech act researchers
have widely adopted the notions of face7, the strategies for doing facethreatening
acts, and especially the context factors power (P), social
distance (D) and ranking of impositions (R)
. However, with few exceptions (Arundale, 1999, 2005;
Kopytko, 1995), little attention has been paid to the theory’s rationalist
foundation.
Brown and Levinson engage, as a Weberian ideal type, a “Model Person,”
the pragmatic sibling to the Chomskyan idealized native speaker. The Model
Person is endowed with rationality and face. As a rational agent, the Model
Person possesses “a precisely definable mode of reasoning from ends to the
means that will achieve those ends” (1987, p. 58). In order to achieve the
goal of carrying out face-threatening acts while maintaining face, the Model
Person chooses from a finite set of strategies as means to such ends (p. 58).
An added property of rational action in the proposed sense is the “ability to
weigh up different means to an end, and choose the one that most satisfies
the desired goals” (p. 65). Such reckoning requires an assessment of
minimum costs (“maximization”) (p. 65) in the selection of means towards a
goal —a feature of the theory paralleled by Leech’s (1983) politeness
maxims. Also factoring into the assessment face threat are the context
variables P, D, and R, which enable actors to compute the “weightiness” of a
face-threatening act and choose a strategy accordingly8— about which more
when we get to “context.”
The rational actor model as applied to speech acts and politeness
derives its intellectual attraction from two main sources. First, it is consistent
with commonsense thinking, at least in those societies that have produced,
and continue to produce, scientific theories of rational action.9 As ordinary
social actors, we constantly attribute internal states —motives, intentions,
beliefs, affect— to each other and we appeal to these internal states as
explanatory resources for the behavior we observe, including our own.10 A
theory of speech acts or politeness that is consonant with members’ ordinary
thinking has intuitive appeal and persuasion.
Secondly, the rational actor model is the standard model of classical
sociology, economics, social psychology, and sociolinguistics (Coupland,
2001). For instance, as a social psychological theory of intergroup
communication, accommodation theory (e.g., Giles, Coupland, & Coupland,
1991) holds that linguistic convergence or divergence are motivated by
actors’ desire for more effective communication, affiliation, or distance. In
fact, Meyerhoff (2001) categorizes a range of social psychological
approaches to language variation as exponents of the “motivational
paradigm.” In Myers-Scotton’s (1993) markedness model of codeswitching,
speakers make rational code choices based on an assessment of their rights
and obligations during a speech event. In second language acquisition, there
is a straight line from the commonsense belief that students need to be
“motivated” to learn second languages to the voluminous research literature
on the relationship between motivation and L2 learning. The “learner
strategy” literature is predicated on a means–ends relationship, where the
“good language learner” (another Weberian ideal type?) employs particular
kinds of premeditated goal-related behavior that are believed to advance L2
learning. The rational actor model thus maps squarely onto prominent
theories in applied linguistics and second language studies, and despite
other differences, these theories are compatible with rationalist pragmatic
theories such as Grice’s theory of conversational implicature, Searle’s
speech act theory, and Brown & Levinson’s politeness theory. Here and
further down, I draw these lines to different research domains in applied
linguistics and other social sciences in order to emphasize that pragmatics
does not live an insular existence but is an integral part of social science
discourse. In pragmatics as in other fields, extensions of common sense into
scientific discourse have the virtue of instant plausibility or face validity, and
using as explanations of social actions the same resources as ordinary
social members —that is, such internal states as actors’ intentions and
motives— is a historically developed, naturalized practice throughout the
social sciences.
That being the case, why might it be problematic? As rationalist analysis
is standard in speech act research, let’s see how it works when applied to
data.
In order to determine the illocutionary force of the utterance in line
2, the analyst must make assumptions about the speaker’s intentions. But
the intentions reside in the speaker’s head and are thus hidden from sight.
The utterance alone does not give its “speaker meaning” away, por lo tanto el analista debe suponer una intención, basándose en sus propios conocimientos, sesgos y perspectivas.
According to this analytical policy, such notions as “illocutionary
ambiguity” may turn out to be a problem created by the rationalist
pragmatician rather than a problem for the participants.
As Bilmes (1986) puts it,
In very broad terms, speech act pragmatics explains how an utterance was
responded to according to what the utterance meant. Conversation analysis
explains what the utterance meant according to how it was responded to.
(modified from Bilmes, 1986, p. 132)
In fact, analysts may have to say a lot about intentions –
for instance, if the participants themselves treat their talk as “intentionimplicative”
(Edwards, 1997, p. 94) or if they make intentions the topic of
their talk.
How are we going to choose between the alternative analyses? They
may not be equally plausible but they are all possible.
Weizman (1989), for instance, addresses the
“deniability potential” of indirect speech acts and argues that it is because of
this property that indirect speech acts are so pervasive in politicians’
discourse.13 This is a convincing analysis, but it does not require a rationalist
view of speech acts.
Participants behave like rationalist pragmaticians, or perhaps it is the other
way around? By contrast, the discursive pragmatician treats participants’
intentions as a topic, not a resource for analysis. This analytical policy
enables the analyst to stay clear of otherwise irresolvable impasses that can
arise from possible “hearings” and their treatments by participants. As Bilmes
(1992) observes
An accurate hearing may be rejected because the speaker has changed his
mind about what he wants to be heard. An inaccurate hearing may be
accepted because the speaker is satisfied to be understood in that way. Or a
hearing may elaborate or specify a speaker’s meaning in a way that the
speaker never thought about, in a way such that the speaker cannot ‘simply
know’ whether or not that meaning was what he had in mind. ... We may
even experience our recipient’s hearing as a revelation of what, after all, we
had in mind in the first place. (Bilmes, 1992, p. 95f)
Meaning
Pragmatics is commonly defined as the study of particular kinds of
meaning, such as “speaker meaning,” “contextual meaning” (Yule, 1996,
p. 3), “meaning in use,” and “meaning in context” (Thomas, 1995, p. 1f),
while the notion of meaning itself remains unexplicated.14 Bilmes (1986)
distinguishes four approaches to a theory of meaning: meaning as speaker’s
intention, convention, use, and response, where the first two notions of
meaning combine in the commonsense understanding of meaning.
The intentional theory has it that verbal expressions serve a vehicles for
what the speaker intends, what he “really means.” In the conventional
theory, words have meanings that are laid down by convention. In the use
theory, the meaning of an expression depends on how it is used and in what
context, whereas the response approach (...) holds that the meaning of an
expression is the response that it elicits. The commonsense notion of
meaning combines the conventional and intentional approaches to meaning.
(Bilmes, 1986, p. 108)
The view of meaning as intention, or more precisely, as speaker’s
intention, the expression of which is designed so that the hearer is enabled
to recognize the intention, presupposes an (implicit) theory of communication
variably referred to as the conduit metaphor (Reddy, 1979), the
encoding/decoding model (Arundale, 1999), or telementation (Harris, 1996,
following Locke, 1689/1975). The familiar idea is that a speaker generates
an intended meaning through intrapsychological cognitive processes,
encodes the intention by means of a repertoire of conventionalized formmeaning
associations (a linguistic code), and transmits the encoded
information as a signal sequence. At the receiving end, the hearer reverses
the process, decoding the signals by applying the same linguistic code as
the speaker deployed for encoding and thereby recovering the speaker’s
intended meanings. The telementation model is the default model of
pragmatics. It informs Grice’s theory of meaning and consequently pragmatic
theories building on Grice, notably (but not exclusively) Searle’s speech act
theory and Brown and Levinson’s politeness theory16. Other difficulties aside,
the telementation model’s individualistic bias renders it deeply problematic
for the analysis of speech acts in interaction. As Arundale (2005) notes with
a view to Brown and Levinson’s politeness theory,
…an explanation of language use that is framed in terms of one individual’s
cognitive processing during producing and interpreting utterances can
explain a talk exchange only as a matter of output from and input to a pair of
separate one-person systems. Such monologic accounts treat talk between
two people entirely as a summative phenomenon. If one chooses to treat
talk-in-interaction as a dyadic activity, however, one examines talk
exchanges as the joint product of a single two-person system, recognizing
that such systems exhibit nonsummative or emergent properties. (p. 51)
The view of meaning as “speaker meaning” or intention confronts us with
the same theoretical and analytical problems as intention-based accounts of
action. By adopting a discursive approach, we can instead treat meaning as
the understandings that participants display to each other in the sequential
organization of their talk. That is, a second speaker displays through her
response how she understood the action or actions in the first speaker’s turn,
and the second speaker’s turn provides an occasion for the first speaker to
ratify or repair that understanding. In
this way, meaning is conceived not only as “social” rather than individual or
as loosely “collaboratively constructed” but instead as accountably achieved
intersubjectivity.
Turning to the second approach to meaning, in speech act research, the
convention view of meaning is crystallized in the notion of semantic
formulae, or speech act realization strategies. Semantic formulae combine to
speech act sets, the collection of semantic structures by which a particular
speech act can be performed (e.g., Olshtain & Cohen, 1983). Speech act
sets have been proposed, among others, for apologies (Meier, 1998, for