You are on page 1of 28

jtmm~

ELSEVIER Journal of Pragmatics 28 (1997) 1-28

Desemanticizing pragmatics *
Wataru Koyama
6051 s. Drexel Ave., Apt. #2N, Chicago, IL 60637, USA

Received No,lember 1995; revised version July 1996

Abstract
This paper starts with a 'metatheoretic' discussion on the sociohistoric conditions of 'pragmatics' as a discourse genre, and identifies a complex set of antinomies in modem studies of language, in relation to which pragmatics is situated. Then, the paper investigates how such antinomies are manifested in the speech act theories of logico-linguistic and ethnographic approaches in terms of referential representation and social action, and suggests the possibility of reconstructing a theory of speech acts which may integratively reformulate both referential and social interactional facets of the problem.

1. Contextual considerations
N o theory of pragmatics ca~ entirely avoid addressing the problems o f referential representation and social action. ~ In other words, a pragmatic theory is evidently a linguistic and social theory at once. This explicitly dual nature o f pragmatic theories points to the possibility of 'pr~.gmatics' that it m a y develop a ' t h e o r y ' (i.e., makingexplicit of modus operandi) wtfich integratively reformulates our current ' h u m a n scientific' understandings of referential representation and social action into one coherent whole. Such a possibility, however, is constituted within a space of 'pragmatics' as a discourse genre, which, like any other 'scientific' inquiries, is constrained (though not determined) by sociohistoric, discursive conditions in which it has situated itself (cf. Kuhn, 1962; Foucault, 1966; G o o d m a n , 1979; Bourdieu, 1980). The integration of this metatheoretic point into our pragmatic inquiries would help us to

e' I am grateful for questions and comments received from William F. Hanks, Isamu Kawaguchi, Michael Silverstein and an anonymous reader for the Journal of Pragmatics. I am, of course, solely responsible for the result. l Reference and modalized predication are, of course, kinds of speech acts, and indexically presuppose and entail ('create') ontic 'realities'. power-relations and group-identities (cf. Putnam, 1975; Hanks, 1990). Yet, their characteristics are such that one will benefit from differentiating them from other kinds of speech acts, which are more robu,;tly and focally social-indexical. This is an empirically-based, nona priori, theoretic decision. 0378-2166/97/$17.00 1997 Elsevier Science B.V. All rights reserved PH S0378-2 166(96)00078-1

w. Koyama / Journal of Pragmatics 28 (1997) 1-28

avoid the interrelated problems of universalist pragmatics, to wit: (1) 'ethnocentric projections' of analysts' own cultural categories onto other communities; (2) 'Universalism', which holds that there is some (biological, linguistic, psychological, social, etc.) foundations which are immune from historic contingencies ('natural') and on which cultural strata rest (cf. Sahlins, 1976; Geertz, 1973: 33-83; 1984); (3) 'decontextualism', which undervalues the theoretical significances of 'context', 'historicity', 'contingency' and 'process' (Geertz, 1973: 3-30); (4) Scientism, which conceptualizes social sciences as theoretical caiques of natural sciences and logicomathematics (cf. Putnam, 1973); and (5) physicalism and Realism, which equate the real with the physical, conceptualized to exist wholly independent of the observer (cf. Putnam, 1990: 3-29; 1992). It is crucial to realize, however, that the integration does not necessarily mean the espousal of (1) 'Culturalism' (closely associated with certain forms of pragmatism, Neo-Wittgensteineanism; cf. Gellner, 1959), which holds that the ways of life of natives are the ultimate source of 'rightness' or 'truths' (cf. Rorty, 1979), and which forecloses the possibility of critique of 'Culture' or ideology (Geertz, 1973: 193-233; Ricoeur, 1986); (2) 'absolute relativism' or 'absolute historicism', which is theoretically self-defeating or inconsistent between theory and practice (cf. D'Amico, 1989; Putnam, 1992); (3) the doctrine of 'the cultural arbitrary', which may obscure the sociohistoric motivations of cultural orders; or (4) cultural and, when 'ontological relativity strikes at home', individual solipsism (cf. Putnam, 1992). Rather, such integration highlights the necessity of careful analyses of varied degrees of correspondence between discourses of pragmatics as a genre, on one hand, and their sociohistoric, contextual surrounds, which are presupposable independent of 'pragmatic' discourses, and which can be shown to be relatable, either positively or negatively, and to varied degrees, to 'pragmatic' discourses, from the perspectives of metatheoretic discourses. In carrying out such a 'meta-analysis', moreover, we should try to minimize the possibilities of self-authorization of our own pragmatic theories through genealogical constructions of histories of the discourse genre in such a way that our theories appear to be the 'Hegelian culmination' of all previous attempts whose values are measured in relation to our own theories, which are presented as the perfected versions of previous, imperfect ones. Rather, we should aim at the sophistication of our current theories through the reflective recognition of sociohistoric conditions of our genre, some facets of which we may see more clearly in the historicizable, 'prior-to-now' moments of our genre in vitro, than in our own time in vivo (cf. Stocking, 1982: 1-12). The discourse genre of 'pragmatics', at least after Morris' reformulation of Peirce's pragmaticism, strongly evokes the remainder of the semiotic triad, syntax and semantics, and pragmaticians have constituted the discourse genre largely as the residual category of the trichotomy (cf. Levinson, 1983: 1-53). The sociohistoric conditions of syntactico-semantics, ever since 'structuralism' came to be recognized as a serious socio-philosophical discourse genre, have been scrutinized by historically-oriented analysts, often antagonistic in intent, yet sometimes producing analyses which, we have to admit, contain some truths. For instance, Williams (1977: 21-44; 1989: 31-80), a 'non-infrastructural determinist' Marxist, and Charles Taylor (1985: 215-292), an 'Anglo-American' philosopher, concur in noting that the

W. Koyama / Journal of Pragmatics 28 (1997) 1-28

modem studies of language may be conceptualized in terms of ergon-valorizing 'objectivism' and energeia-valorizing 'expressivism' (cf. Volo~inov, 1973), and Williams points to the sociohistoric conditions of urbanization, the increased copresence of various ethnicities in metropolis, urban alienation and the sentiment of deracination, the expansion and lglobalization of commerce, etc., as factors which are countable as background contexts of the rises of structuralism, anchored on the thesis of arbitrary, 'non-natural' linguistic signs, and symbolic-romanticism, anchored on the anti-theses of pre-linguis~:ic, primordial and 'psychic' or psychological intentionality, Einfiihlung or correspondance, which tries to capture the ever vanishing moments of original semiosis ol the fundamental monad of communication in agentive engagement. Furthermore, the problem of 'translation', forcefully brought home by the legitimation of 'mother tongues' and the diversity of world languages, seems to have polarized 'structuralists' into 'formal-structuralists', who essentialize 'language' as a referential code (as translatability is maximized in some aspects of references and modalized predications; hence the impact of Quine's radical translation) and espouse the thesis of denotational effability, on one hand, and 'continental philosophic (post)structuralists', who emphasize the impossibility of translation, espouse the anti-theses of ineffability and radical relativity, and essentialize 'language' in its socioculturally indexical, value-, authority- and power-indexing capacity, on the other. These antinomies have institutionalized repercussions, as witness the polarization of human sciences in the United States, the universalist, referentialist and rationalist cognitive psychology a~d formal-structural linguistics standing against the increasingly relativistic and anti-rationalist cultural anthropology (cf. Clifford and Marcus, 1986), literary criticism (cf. de Man, 1978) and history (cf. White, 1978), while philosophy, in its post-analytic modalities, is showing 'wise' ambivalence between them (cf. Rorty, 1979; Putnam, 1992). Of course, the characteristics of sociohistoric conditions to which Williams pointed have been part of larger processes called 'modernity' (and various modes of 'anti-modernity'), constitutive parts of which are nationalism, standardization (schooling, print capitalism, t]ae circulation of linguistic, geographic, calendrical paraphernalia, such as dictionaries, encyclopedias, maps, clocks, etc.: cf. Bloomfield, 1927; Havrfinek, 1964; Bourdieu, 1982) and realist literature (cf. Bakhtin, 1981; Anderson, 1991). Though a 'linguistic community', which is constituted by virtue of its members' sharing a denotational code, is distinct from a 'speech (sociocultural) community', such as a nation-state, which is constituted by virtue of its members' sharing (a) cultural categories, (b) pragmatic regularities (e.g., discourse genres, registers) and (c) the (relatively) consciously accessible (Durkheimean) norms towards/against which members interpret their behaviors (cf. Gumperz, 1968), the nationalist tropological equation of these two has been naturalized by recurring discursive interactions which presuppose and thus re-constitute such a trope. That is, the trope 'reifies' our nationalist misrecognition that a denotational code naturally needs to be mappable to a speech community (nation-state) in a oneto-one correspondence, and rank-shifts a discourse genre (linguistic variety), which corresponds to a smaller speech community, up to the status of 'full language' (standardized and written variety), while negatively valorizing non-standard varieties as

W. Koyama / Journal of Pragmatics 28 (1997) 1-28

something less than 'full language'. Thus, the trope is played out by the constantly presupposed and entailed oppositions between the top and center of standardization, which appear ('dehumanizingly' or 'techno-scientifically') neutral and objective to our nationalist consciousness, on one hand, and the periphery of standardization, onto which 'dialects' and 'folk-talk' are projected and which appears ('liberatingly' or 'vulgarly') polyphonous and spontaneous, on the other. A denotational code thus becomes a group-identity index of discourse participants (members of a nation-state) and varieties of the denotational code appear to be hierarchized in terms of their deviations from a standard, superposed variety towards/against which social groups orient their behaviors. We may note that this antinomic discursive configuration seems to be homologous with antinomies we detect between the super-standardized 'logical syntax' and 'ordinary language'; the Saussure-Chomskian linguistics (whose abstract theorizations have been facilitated by the entailments of standardization) and the Labovian sociolinguistics/discourse analysis, ethnography of speaking, etc. (which have situated themselves in conscious opposition to Chomskianism); the Grice-Searlean 'universal pragmatics' and anthropological-linguistic emphasis on 'cultural diversity' and unique particulars. As part of such a 'frozen dialectic', the post-Saussurean and post-Fregean syntactico-semantics have constituted themselves as system-internal formal-structuralist (vs. teleological functionalist: cf. Sahlins, 1976; Nichols, 1984), conventionalist (vs. naturalist, in the Platonic senses), referentialist (vs. Malinowski-Jakobsonian multi-functionalist), and, in their modalities of Saussure-Chomskianism and Tarski-Davidsonianism, intensionalist (vs. intension-extension 'correspondence' theorist, a crude version of which is Tractarian 'picture theorist'). It is in such contexts that the discourse genre of 'pragmatics' has constituted itself, centering around the notion of 'speech acts', oscillating between the poles of reference-and-modalized-predication (representation) and non-referential praxiology (action in a universe of cause and effect), between the poles of linguistic formalism, cognitive representationalism and pragmatism, and between the poles of formal structuralism and pre-linguistic psychologism. Thus, for example, Austin's (1962) discovery of 'explicit primary performatives' has led to the developments of two distinct approaches towards the problem of 'speech acts', the first syntactically (cf. Sadock, 1974) and the second psycho-logically (cf. Searle, 1969, 1979, 1983) reductionistic. That is, the former tries to integrate as many as possible of indexical legisigns2 (event-types) regularly associable
2 'Indexical', 'symbolic', 'sinsign' and 'legisign' are used in their classic Peircean senses (cf. Peirce, 1932: 134-155). 'Symbolic legisigns' ('intensional regularities') can be subdivided into two kinds: linguistic structural types and cultural stereotypes ('cultural categories') (cf. Putnam, 1975). 'Indexical legisigns' are behavioral ('extensional') regularities, such as deictic regularities, conversational regularities, event-types, discourse genres, linguistic registers. 'Indexical sinsigns' are extensional singularities, such as tokens of event-, discourse genre-, linguistic structural-types and other 'happenings'. Though I suspect that some of us may be repelled by the awkwardness of these terms, I employ them not only for my contention that it is high time for the coming-together of 'pragmatics' and (neo)pragmatism, pragmaticism and praxiological semiotics, but also for their utility for explicitly making the three-way distinction, which I find necessary for avoiding a framework based on 'type vs. token' or 'regularity vs. singularity (or actuality)' dichotomy.

w. Koyama / Journal of Pragmatics 28 (1997) 1-28

with (symbolic) linguistic legisigns into syntactic descriptions which otherwise formalize symbolic legisigns only, mixing at least two semiotically distinct levels of analysis. On the other hand, the latter tries to formulate a psychological schema of pre-linguistic, intentional 3 'illocutionary forces' categorizable independent of linguistic structure and formalism p e r se. Note, however, that both approaches theoretically reduce the uses of various linguistic 'form-orders' (Bloomfield, 1984) to explicit primary performative constructions/speech acts, respectively. As a result, non-explicit speech acts, especially relatively regular, non-explicit yet primary ('indirect') speech acts, have t,ecome a salient and insurmountable problem (cf. Levinson, 1983). The consequences of these two reductionisms, which we shall follow below, help us to see more clearly the crucial aspects in which explicit primary performative speech acts are distinct from other speech acts: (1) they explicitly predicate illocutionary senses encoded in the linguistic structure of which they are tokens, and perlocute the senses at the level of discourse by re-presenting themselves as indexical icons of the illocutionary senses through the uses of fixed, ritual formulae, prototypically Is I [vP V'metapragmatic(to) you S]] where V' is unmarked for tense and aspect; and (2) they have the maxima of automatized mappability between syntactic types and interactional significances, while 'indirect speech acts' have only the second of these in smaller, varying and only relatively (in)determinate degrees. It is the consequence of not seeing such crucial differences, I think, that neither approach can explain why (1) the syntactic operations which delete segments in performative clauses or (2) the logical calculations which derive 'what is meant' or 'primary illocutionary force' from 'what is said' or 'secondary illocutionary force', are not 'perlocutionary compulsiw.'ness-preserving' and reduce it by some particular degree. More generally, it seem s to me that explicit primary performatives are idiosyncratic in that they transparently map (1) verba dicendi in linguistic structure, (2) what is pragmatically referred to and predicated about, (3) the entailed contexts of interaction and (4) the conscioasly accessible end of interaction, the four domains which are, I think, to be theoretically distinguished. In the following section, I shall try to elucidate the genealogical origins of, and semiotic mechanisms which have given rise to, these two approaches to speech acts, by way of maximizing what we can learn from these theories, their presuppositions, and their consequences.

2. Social actions in representationalist theories Most of what goes under the name of 'pragmatics' in contemporary linguistics and (post)analytic philosophy often makes a genealogizing gesture to the 'Ordinary Language School' philosophers, such as Wittgenstein, Austin, Ryle, Strawson, Grice and Searle (cf. Levinson, 19831. The formation of this discourse genre, it seems fair to say, was achieved mainly by contextualizing the 'traditional' metaphysics, on one 3 'Intention' is not used in its technical, phenomenological sense akin to 'aboutness', but in the sense of 'agentively consciously accessible metapragmatic model of the purposive telos of discursive interaction', which is only a specific subcategory of the former (cf. note 8).

w. Koyama /Journal of Pragmatics 28 (1997) 1-28

hand, and the Carnapian Aufbau and the Russellian logical atomism, on the other (cf. Rorty, 1967). In different ways, that is, both the Ordinary Language School and the logical syntacticians were engaged in the programs of 'pseudo-problematization' of metaphysics (cf. Schlick, 1967). The former tried to show that 'metaphysical problems' arose from our abstract and opaque 'philosophical language', whose connection with our 'ordinary language' should be explicitly shown, or which must be grounded on our 'ordinary language', the careful studies of which would contribute to, if not achieve, the solution to metaphysical problems (cf. Hampshire, 1959--60; Urmson, 1967; Urmson and Warnock, 1961). The latter, on the other hand, tried to show that 'metaphysical problems' arose from our muddled and opaque 'ordinary language', which were to be replaced with an ideal language made of empirically verifiable statements and logically valid definitions, axioms and derivational rules. The developments in (post)analytic philosophy (cf. Putnam, 1975) have shown, however, that (1) correspondences between a symbolic language and indexical sinsigns are too complex to warrant logico-syntactic - or equivalent - picture-theoretic reduction, since referential practices require not only contribution from syntacticosemantics, but also from 'pure' indexicality, the sociohistoric continuity of usage and cultural stereotypes ('cultural categories') to varying degrees, such a variation largely depending on which type of NP is used (cf. Silverstein, 1987). The fatal assumptions are, then, (1) the homogeneous conceptualizations of both symbolic structure and indexicality caused by not noting the significance of careful crosslinguistic analyses of linguistic structures and ethnographic researches on linguistic practices and (2) the Tractarian isomorphism between 'language' (logical syntax and verifiable sentences) and 'world' (verifying eventualities). Here, it is important to note that, especially in view of some speech act theories, the Aufbau 'collapsed' (at least partly) due to these two assumptions. We must not be spurred by its 'failure' to an opposing, 'demythologizing' and reductionistically functionalist thesis (cf. Sahlins, 1976) that 'meaning is use', which makes the very same fatal assumptions of the homogeneity of symbolic elements, 4 the homogeneity of indexical elements 5 and an isomorphism between symbolic and indexical domains by reductively assimilating the former into the latter (cf. Searle 1969: 146-149). Taking a course antithetical to the logical syntacticians, the ordinary language philosophers purported to study 'language' as ordinarily used in its complexity. In operationalizing their program, the Oxbridge analysts rejected, in effect, the analytic machinery of modern linguistics and any explicit social theorization, and used an introspective method of ad hoc and commonsensical 6 kind, composed of "(a) free association ...; (b) the reading of relevant documents ...; (c) use of the dictionary" (Urmson, 1967: 233-234). This theoretical choice seems to have been guided by a metatheoretic principle which we may call 'the avoidance of doxa': 4 Characteristically, NPs are treated as if they were all deictics. any Characteristically,again, indexical legisignsare reduced to indexical sinsigns ('happenings'), whose singularityand uniquenessalone we are advisedto appreciate. 6 'Ad hoc' and 'commonsensical'are used in the ethnomethodologicalsenses, and not to be taken pejoratively (cf. Garfinkel, 1972).

W. Koyama / Journal of Pragmatics 28 (1997) 1-28

"Austin always insisted that during the work so far described all theorizing should be excluded .,. Premature theorizing can blind us to the li~aguistic facts; premature theorizers bend their idiom to suit the theory, as is shown all too often by the barbarous idiom found in the writings of philosophers ..." (ibid. : 234) As Apel (1991) rightfully observes, and Austin himself once acknowledged, one finds an essentially Husserlian phenomenological orientation in the 'ordinary language' program, first in its search for presuppositionless methodology (cf. Rorty, 1967); second in its methodological reliance on the introspective, agentive intuition of users, if not exclusively of analysts (cf. Grice, 1989a: 174-175; Habermas, 1991: 17); third, the preanalytic, holislic conceptualization of 'meaning'; and fourth (especially in Grice and Searle), the primordiality of pre-linguistic intentions of 'transcendental ego-consciousness' (of. Derrida, 1967; Kristeva, 1974). Thus, the principle of 'avoidance of doxa' binds, paradoxically it may appear, the 'doxa-free' program to the hori~'on of agentive ego-consciousness of primordial and 'transcendental' (unanimous within a community) kind, i.e., agentively graspable communicative intuition sharable by a speech community, an 'ego-consciousness' which we might call XYZ for short. Now, to sufficiently characterize XYZ, however, analysts have to add some 'flesh' to this core, and we are to use the most presuppositionless intuition, guide,] by the axiom of 'the avoidance of doxa' and the methodological primacy of introspective, reflective interpretation on linguistic semiosis. It is at this point that the diverging directions within the program of 'linguistic phenomenology' show their multiplicity. That is, as Silverstein (1981) demonstrates from his field study, the degree to which linguistic semiosis is accessible to conscious, agentive and introspective interpretation is overdetermined by at least three variables of (1) the unavoidability of referentiality, (2) continuous segmentability and (3) the relative presupposability of indenpendently verifiable contextual factor(s). Thus, if one essentializes linguistic semiosis from the combinations of (1) and (3), and characterizes XYZ primarily by these aspects appearing most doxa-free to agentive awareness, one is naturally led to a rationalist and presuppositionalist theory of communication, as (a) referential function, as opposed to interactional (phatic, emotive, conative and part of poetic and metalinguistic) functions, is that from which our notions of rationality are easily hypostatizable, and (b) the focus on presupposability directs an analyst's attention away from the entailing ('creative') aspect of linguistic semiosis. Here, note lhe Searlean doctrine of "the principle of expressibility", which says "whatever can be [interactionally] meant can be [referentially] said" (Searle, 1969: 19; brack, ets mine). This thesis implicitly reduces interactional significances to agentive intention, which appears, to agentive awareness, more presupposable than, and temporally and logically precede, speech and even linguistic structure. In addition, it posits interactional significances as something transparently translatable (reducible) to refer, ential significances, notwithstanding the very point of departure for speech act theories that referential function, albeit uniquely developed in language and unique in its functioning, is just one of the communicative functions. As a result, the problercLatics of referential mechanisms which guarantee the felicitous speech acts of referriag and modalized predicating (part of which, we have to admit, is constituted by forrnal-distributional morphosyntax) is not addressed. Nor

W. Koyama /Journal of Pragmatics 28 (1997) 1-28

is non-referential indexicality, especially the contextual indeterminacy of interactional significances of discourse, accounted for by psycho-logical speech act theories, in which the agentive, intentional, teleological interpretation of an up-coming segment of interaction is posited as something which predetermines the interactional significance of the segment. That is, the consequential entailment of interaction is psychologically 'pre-fixed' by an agent's teleological and intentional tour de force (cf. Hancher, 1979; Levinson, 1981; Clark and Wilkes-Gibbs, 1986). Moreover, not only Searle's theory, but also Grice's suffers from the very same problem of teleological determinism, insofar as it identifies 'communication' with intentional acts. That is, given the powerful machinery of unbounded Gricean calculations, as Sadock (1978) points out, any 'what-is-meant' is calculable from a given 'what-is-said', which indicates that, for Grice's theory to be operationalizable, the interactional 'meaning' of a discourse must be somehow pre-fixed before an actual calculation takes place. Thus, insofar as referential function and agentive intentionality appear to be the most doxa-free dogmas, if you pardon this oxymoron, of our reflective interpretations on linguistic semiosis, XYZ must be equipped with, if not uniquely characterizable by, (1) the cognitive capacity of intentionality which is subcategorized into the classes either transparently (in Austin's theory) or opaquely (in Searle's) mappable to (prototypic) moods and (fully elaborately lexicalized) verba dicendi, e.g., intentions to declare, ask, command, promise, etc., as we can surely intend to do so and they are the only means by which what is primarily done in speech can be said explicitly and in simplex manner; and (2) that of reference and modalized predication (a) to which the interactional significances of speech acts are transparently (in Austin's theory) or opaquely (in Searle's) translatable and (b) which are meaningful only to the extent that they can be consciously, purposively intended. This last, as we shall see, is in a stark contrast with the (mostly implicit) characterizations of XYZ, in syntactically reductionistic speech act theories, by (1) a cognitive capacity 'underlying' reference and modalized predication, whether consciously accessible or not, (2) which is formalizable and representable by the hierarchical constituency structures of linguistic segments which 'underlie' (or at least transparently translate) our interactional capacity to 'do things with words'. Thus, unconscious, seemingly 'pointless' (cf. Searle's 'illocutionary point') actions, which are characterized by their appearance to our ego-consciousness as irrational, not face-maximizing or apparently inexplicable, compel the psycho-logical theorists (as well as the ego-conscious part of us, to be sure) to rationalize ('explain' or 'calculate') them, so anything which is not primafacie characterizable as intentionally purposive, inclusive of linguistic structure and unintended happenings, is either dismissed as irrelevant or reduced into the explanandum of highly complex rationalizing operations, such as Gricean calculations. This much of theoretical consubstantiation of our primary intuition of what communication is all about, i.e., referentially-centered intentional act, seems to sufficiently characterize XYZ. Note that this characterization is nothing but a hypostatization of part of our sharable cultural stereotype associated with the 'metapragmatic' (see note 8) verb pal" excellence, mean-, in the English lexical semantics of verbs of

w. Koyama /.lournal of Pragmatics 28 (1997) 1-28

communicating, as in 'What do you mean?', 'Don't just say it. Mean it! ', 'No, what I meant was ...' (cf. Donnellan, 1978). Thus, the intentionalist theory of communication projects our cultural stereotypes associated with the lexico-semantic label mean- onto the actuality of communication via theoretical consubstantiation: 'meaning,,'. As the ethnography of speaking tradition has amply demonstrated (cf. Rosaldo, 1982; Duranti, 1993), however, this cultural stereotype and the hypostatized theory are far from being universally shared and valid, respectively. Some serious problems with the descriptions of XYZ by the psycho-logical theorists undermine the empirical accountability of the intentionalist theories. First, as so theorized, XYZ has to use something akin to telepathy for intentional telementation, due to the irretrievable gap posited between the solitary, autonomous speaker's intention as the input to communication and the pre-fixed, unique significance of what-is-communicated, or primary illocutionary force, shared by the speaker and addressee, as the output of communication. That is, in the intentionalist theories, it appears a mystery that the speaker can actually intentionally communicate with the addressee, because the speaker's intending alone is not sufficient for the achievement of a 'fully consummated' speech act (i.e., one intended as '*Pi' by the speaker and complicitly 'taken-up' as '*Pi' by the interlocutor). They miss the point that the speaker has to compel the interlocutor to comply with 'the model of interaction' which she intends to actualize] Then, to comprehend the problem of intentionality, we need to rigorously investigate, first of all, the nature of this 'model of interaction'. To begin with, interaction can be conceptualized as mere 'happenings', or 'indexical sinsigns' in the Peircean jargon, the determination of the socioculturally identifiable significances of which is played out by discourse participants. However, in order to be endowed with such significances, interaction must be first of all identifiable as an individttable unit. That is, it is only when interaction has 'unity', 'cohesiveness' or 'textuality', we can identity a set of indexical sinsigns as a 'figure' clearly demarcated :From the background of other indexical sinsigns (cf. Polanyi, 1958). Such a construction of a 'figure' or 'interactional text' out of unregimented 'mass' of inde~:ical sinsigns may be achieved by having indexical sinsigns discriminatively point to a delimited set of indexical sinsigns. Given that 'indexical' (vs. symbolic) is synonymous with 'pragmatic' (vs. syntactico-semantic) in this context (cf. Bar-Hitlel, 1954), we may say that such regimenting indexicals are functioning 'metapragmatically '8 at the n+lth order in relation to regi7 The Searlean notion of 'IFID' dot;s point to the theorization of this issue, to be sure. However, his theory does not address the defeasibilIy of an agent's intentional 'model of interaction' and the 'dynamics' of interactions. 'Metapragmatic' is a subcategory of (Jakobsonian) 'metalinguistic', and characterizes a sign (at the n+lth order) indexically pointing to an indexically functioning sign (at the nth order): e.g., (1) (pragmatic) speech about (pragmatic) speech; (2) (pragmatic) interpretationsof (pragmatic) discursive interactions, such as the agentive understandings of the discourse genre of ongoing interaction, and agentive intentions about the consequencesof discursive interactions; (3) mutual co-indexingof sinsigns, such as between verbal and gestural signs or ~etweenverbal signs (in which case, a sign is at the n+lth order in respect of its pointing to another sign, and, simultaneously,at the nth order in respect of its being pointed

10

w. Koyama / Journal of Pragmatics 28 (1997) 1-28

m e n t e d indexicals functioning at the nth order, and we m a y speak of ' m e t a p r a g matic f r a m e ' or 'interactional text': a cohesively united ' t e x t ' of indexical sinsigns which can be e n d o w e d with a socioculturally identifiable significance, or ' w h a t is socioculturally meaningfully done'. As the 'architecture' of interactional textuality is constituted by metapragmatics, one can ascribe a 'sociocultural significance' to it by evoking presupposable symbolic legisigns, such as cultural stereotypes or 'illocutionary senses'. Alternatively, the architecture m a y re-present itself as an indexical icon of presupposable signs, such as indexical legisigns, as in 'interaction rituals' ( G o f f m a n , 1967), conversational 'structures', discourse genres, or ' p r i m a r y p e r f o r m a t i v e s ' , or symbolic legisigns, as in 'explicit p r i m a r y p e r f o r m a tives'. In this conceptualization, then, 'intention' can be theorized as p a r t of the g a m e of fixing a determinate interactional text between discourse participants, that is, as an agentively accessible teleological m e t a p r a g m a t i c frame of ' w h a t will have been done', which is in competition with other m e t a p r a g m a t i c frames and, thus, defeasible. Then, the pre-fixability of 'illocutionary type', an idea shared by the Searleans, Griceans and schema theorists, m a y be called a ' m y t h ' , p r e s u m a b l y inherited from Austin, who was bewitched by explicit p r i m a r y performatives, whose illocutionary forces appear pre-fixable due to a particularly effective, tropological voicing: i.e., indexical iconization between sinsigns and symbolic legisigns. Another problem with the intentionalist theories is that X Y Z must have an iibermenschlich cognitive capacity to intend her interlocutor to recognize her intention that he recognizes her intention that ... ad infinitum ... *p. In other words, X Y Z must have the capacity to actually undergo (rather than 'have a mental schema of the selfreferential type "(a) A intends that B recognize *p and intends (a)" ') infinite number of intendings in order to achieve a single piece of speech act. 9 Yet another problem is that X Y Z does not have a cognitive capacity not only to formulate ' m e t a s e m a n t i c ' sentences (i.e., 'analytic' or 'law-like' descriptions of cultural stereotypes or grammatico-semantic units), but also to use the tokens of nonmetapragmaticaUy characterizable types, as her grammatical capacity is limited to to by the other). By pointing at a set of indexical sinsigns, which are thus delimited from the rest of contextual 'mass', metapragmatics organizes and regiments such indexicals into a cohesive unit(y) ('text'), both on the planes of 'reference' (Hallidayan 'referential text' of 'what is said') and 'action' ('interactional text' of 'what is done'). As is well known, the semantic contents (designata) and denotata, not to mention referents, of the (Jakobsonian) shifters, such as deictics, are (largely) metapragmatically characterized. That is, to specify them, one has to make a primary reference to pragmatic regularities, rather than to linguistic structure or cultural stereotypes. Similarly, verba dicendi metapragmatically denote the kinds of (pragmatic) speech. Thus, we may say that shifters and verba dicendi, as linguistic types, are transparently metapragmatically characterizable, relative to other linguistic types. 9 Grice (1989b), however, points to the notion of 'threshold level' above which a finite number of intendings 'practically', or pragmatically, count as 'to intend *p', which is, strictly ('logically') speaking, impossible. Note that the individuation of a referent by referring, which 'theoretically' allows the Quinean infinite atomization or indeterminacy, operates by the same principle. 'A sufficient degree of determinacy in context' appears to be a key notion both in referential and non-referential pragmatics.

w. Koyama / Journal of Pragmatics 28 (1997) 1-28

11

the descriptors o f s p e e c h acts a!tone. ~ ( T e c h n i c a l l y s p e a k i n g , e v e n m e t a p r a g m a t i cally c h a r a c t e r i z a b l e t y p e s are n o t p r o p e r l y l i n g u i s t i c a l l y f o r m a l i z a b l e , since f o r m a l i z a b i l i t y p r i m a r i l y d e r i v e s f r o m s y s t e m a t i c i t y a n d s y s t e m i c i t y , a n d the l i n g u i s t i c structure is, as S a u s s u r e , M a r t i n e t a n d B r 0 n d a l l e x p l i c i t l y insisted, un systdrne oft tout se tient, so that the p r o p e r m e t a l i n g u i s t i c c h a r a c t e r i z a t i o n s o f verba dicendi cann o t b e g i v e n b y a n y t h e o r y w h i c h fails to take into a c c o u n t the e n t i r e t y o f s e m a n t i c o grammar.) F i n a l l y , a n d m o r e g e n e r a l l y , X Y Z does n o t h a v e a c o g n i t i v e c a p a c i t y not to i n t e n d , e i t h e r r e f e r e n t i a l l y or i n l e r a c t i o n a l l y . X Y Z ' s c o n s c i o u s n e s s is t r a n s p a r e n t l y a c c e s s i b l e to anything w h i c h m a y a n d does take place in h e r m i n d , a thesis e m p i r i cally d i s p r o v e d b y the w i d e - s p r e a d p h e n o m e n a o f l i n g u i s t i c i d e o l o g i z a t i o n such as

~0 Here, I am sacrificing analytic preci'.;ion for thematic convenience. Austin, Searle and Grice, situated in (or around) the analytic tradition of philosophy, did have the theories of reference and modalized predication as such: (1) Austin (1962), ':hough positing 'constatives' and then 'rhetic acts', did not consider how linguistic structures are deployed in reference and modalized predication to be an important question. (2) Searle (1969: 31) posits 'propositional content', which appears as 'p' in F(p), and does have a sophisticated theory of reference and modalized predication, though he seems to take a rather dim view of the roles of linguistic structure in referential practices, too. First, Searle (1969: 72-96) notices that felicitous reference presupposes the existence, uniqueness (more broadly, quantifiability) and identifiability (characterizability) of referenl(s), but does not elaborate on the specific relationships between this and the organizations of linguistic ~';tructureand its contribution to the former. Second, on the basis of his theory of 'descriptive backing' C identifiability'), Searle ( 1969: 162-174) develops a descriptive, cluster theory of references by proper names, which, though being able to explain the phenomena of diachronic reference alternations, has d!fficulty explicating the phenomena of diachronic reference continuity and tl~e indexical rigidity of individuation based on Kripkean baptismal speech acts. Third, Searle (1979) does not relate the notions of characterizability and presupposability to his analysis of the Donnellanean 'attributive' and 'referential' uses of definite expressions, respectively, and he theorizes the former use as the case where denotation and intended reference ('primary aspect') match and the latter as the one where denotation matches c~escriptive backing ('secondary aspect'). Thus, for example, the problem of reference entailment and maintenance mechanisms, which are one of the four main components of linguistic structure (cf. Van V~lin, 1993) and which constrain the organizations of this latter, is replaced with correspondences between intention and denotation. As a result, how heterogeneous elements in linguistic structure variously but systematically contribute to referring (cf. Silverstein, 1987) and how speech acts of referring con,:train the organization of linguistic structure are not addressed. Thus, he does not note that any use of an NP type involves both 'attributive' (symbolic) and 'referential' (indexical) components and that the de~:ree to which such a use involves each component depends on NP type and discursive contexts. Finally, (_~)Grice differentiates 'what is said' (propositional meaning) from 'what is implicated' (inclusive of 'com,entional' and 'maxim-observing/flouting conversational' 'implicatures'). This is, we must recognize, a significant contribution towards the proper theorization of semantics as symbolic and pragmatics ~Lsindexical. However, it seems to be the case that (I) even propositions are not trnth-evaluable without the pragmatic fixing of referents unless they are completely metasemantic; (2) there are two kinds of semantic intensions, linguistic structural senses and cultural stereotypes; (3) there is a pragmatic domain of reference which has textuality, operates beyond sentential boundaries and contributes to the fi Kingof referents; (4) therefore, the domain of 'semantics' as such is actually further restricted to the confine of linguistic structure alone; (5) however, the designata of some segments in linguistic structure are metapragmatically characterizable (e.g., deictics, proper names, discourse connectives); (6) notwithsta~lding, Grice treats the last of these as conventional implicatures just because their denotations do not directly participate in propositions, as a result of which the similarity of these types is missed.

12

w. Koyama /Journal of Pragmatics 28 (1997) 1-28

the Labovian 'stereotypes', which distortingly reflect indexical legisigns and sinsigns, and any p h e n o m e n a of tacit knowledge. 11 Having finished a brief reconstruction of consequences of theoretical constructions primarily based on the pre-theoretic intuition of (1) the unavoidability of reference and (3) the relative presupposability of independently verifiable contextual factors, let us explore another direction. If one essentializes linguistic semiosis from the combinations of (1), (3) and (2) continuous segmentability, and characterizes X Y Z primarily by these three, one is naturally led to, again, a referentialist ideologization of communication, but of a slightly different kind from the first: i.e., the ' o b j e c t i v i s m ' of post-Saussurean linguistics, further articulated and refined by Bloomfield and C h o m s k y . For it is in this structural-phenomenological approach, where the axioms of linearity and arbitrariness entail compositionality (i.e., segmentability and hierarchical projectibility of segments) of linguistic ' m a t e r i a l ' that the surface segmentability of signs is most salient and presupposable. Thus note the general problematicity of 'discontinuous constituents' in this approach, such as some suprasegmentals, Arabic m o r p h o ( p h o ) n e m i c s , overlappingly or discontinuously realizable g r a m m a t i c o - s e m a n t i c categories, etc., especially in the formal-distributionally, non-semantico-pragmatically anchored kinds o f IA and IP approaches (cf. Hockett, 1954). But more fundamentally, by reducing linguistic semiosis into ideally continuous referential segmentation not going beyond the level of sentence, the linguistic autonomists, or worse, syntactic autonomists ( ' w o r s e ' , as the serious theorization of semantics leads to the realization of extensional, pragmatic anchoring of intentional structure, as witness Lyons (1977) and 'Generative Semantics'), posits the XYZ, sometimes called the 'Ideal S p e a k e r - H e a r e r ' , who is essentially deprived of c o m m u n i c a t i v e capacity, except that of parsing. As a result, she cannot actually use her arbitrarily c o m p l e x l y and infinitely generatable, referentially unitized s e m a n t i c o - g r a m m a r in discursive interactions. Nor can she c o m m u n i c a t e nonreferentially. 12 Moreover, X Y Z is even deprived of a capacity to identify the

t~ For analytic precision, it should be noted that Searle (1979, 1980), when theorizing on 'literal (referential) meaning' does emphasize 'background context', which is tacitly presupposed, while his theory of non-referential interaction has become even more intentionalistic since the 70s, presumably due to the problem of 'indirect speech acts' and quasi-Gricean solutions he has adopted (cf. Searle, 1979). Such a disparity seems to be consonant, or at least non-inconsistent, with our observation that Searle's theory of speech acts is referentially-focused (cf. his 'principle of expressibility') and underdeveloped with regard to non-referential aspects of interaction. i2 As a theory-internal critique, this is completely irrelevant, since it is one of the fundamental assumptions of modern linguistic theory that linguistic structure is referentially regimented. Note that, unless one abstracts away socio-individual indexicality from what one may observe in discursive interactions, one can never arrive at the level of abstraction where linguistic structure can be posited, as Saussure and Sapir noted. The theoretical soundness of this move can be seen in the undeniable, crucial difference between the 'garden variety' segments of linguistic structure (symbolic legisigns) and deictics (mostly indexical legisigns), which shows that linguistic structural regularities cannot be analyzed by the analytic machinery developed to deal with indexical (behavioral) regularities and singularities (cf. General Semantics), as Chomsky pointed out to positivistic empiricists a long time ago. That is, symbolic capacity should be studied on its own terms. (This does not mean, however, that the organization of structure of symbolic legisigns is not constrained by indexical extensions.)

w. Koyama / Journal of Pragmatics 28 (1997) 1-28

13

semantic contents, not to mention denotata and referents, of a significant class of

linguistic segments such as deictics, theme/rheme-markers, mood/modality-markers, quote/report-markers, etc. It is their realization of this problem, I think, which has led some syntacticians to try integrating pragmatics into their linguistic theories by, unfortunately, reducing ind,zxical legisigns into the 'garden variety' descriptions of symbolic legisigns, as a result of which the post-Fregean and post-Saussurean distinctions between not only semantics (sens/Sinn) and referential pragmatics (r~fdrent/Bedeutung), but also the planes of reference and non-referential interaction are sometimes erased, in the direction towards semantics and reference, respectively. Thus, modeled after performative constructions, whose semiotic uniqueness was unnoticed, the naetapragmatic descriptors of the discursive interactions in which sentential types are tokened started to appear in the descriptions of sentential types under the guise of 'prefixed performative clause'. Since mappability between syntactic types (symbolic legisigns) and indexical legisigns is in principle neither transparent nor determinate, such syntactic theories enormously complicate and irregularize the otherwise straightforward descriptions of symbolic regularities in exchange for some 'significant generalizations' obtaining between pragmatics and syntax (cf. Sadock, 1974), which may be easily accounted for, however, by a 'structural-functi,anal' theory which conceptualizes morphosyntax as being constrained (but not determined) by pragmatics, just as phonology is constrained by phonetics (cf. Van Valin, 1993). Thus, the syntactic speech act theories posit an XYZ who has a sinl;le cognitive capacity 'underlying' reference and modalized predication, whether consciously accessible or not, which is formalizable and representable by the hierarchical constituency structures of linguistic segments which 'underlie' (or at least transparently translate) our interactional capacity to 'do things with words'. Not only are independent capacities collapsed into one curious amalgam of an extended morphosyntactico-semantics absorbing pragmatics, but also non-lexicalized speech acts are a priori excluded from XYZ's cognitive capacity. Having thus seen the two post-Austinean approaches to the problem of 'speech acts', we notice that Austin's own speech act theory, characteristically perhaps, ambivalently falls in the middle of the syntactic and psycho-logical theories: "The first thing to notice about these itists [of Austinean taxonomy of iUocutionaryforces into verdictives, exercitives,commissives,exposilivesand behabitives(cf. Austin, 1962: Lecture XII)] is that they are not classifications of illocutionary acts but of English illocutionary verbs. Austin seems to assume that a classificationof different verbs i,; eo ipso a classificationof kinds of illocutionaryacts ..." (Searle, 1975: 32; bracketsmine) Thus, in Austin's version of speech act theory, as in syntactic ones, the deadlock of the lexical semantics of a particular language felicitously prevents one from imprudently diving into the hypothetical, hypostatized pre-linguistic intentionality. But, whereas the syntactic speech act theories purport to be the theories of morphosyntax, Austin's theory is a theory of 'speech acts' as such, i.e., 'what we do with words', so it becomes a prey tc, a Whorfianism which projects the lexical semantics of English verba dicendi (mixed with cultural stereotypes) onto the theory which

14

w. Koyama /Journal of Pragmatics 28 (1997) 1-28

purportedly describes (and thus is supposedly iconic to) the actuality of what we do with words (cf. Silverstein, 1979). Now, the lesson which we can learn from these approaches is not that any attempts to construct a theory which is maximally presuppositionless necessarily fail, nor that ' a r m c h a i r ' theorization is useless (it is not, pace anti-theoretic empiricists and positivists), but that we can reconstruct a better theory ('better' at least in this epistemic horizon in which we cannot help but operate) by recognizing their limits and limited validity. For example, these theories have shown that the notions of 'intentionality', 'linguistic structure', 'referential function', 'interactional function' are irreducible to one another (sui generis) and required in any theory of speech acts. Extending the Jakobsonian notion of 'metalinguistic function', and using the Peircean semiotic framework, we have tried to re-conceptualize 'intentionality' as a kind of 'metapragmatic f r a m e ' of indexical sinsigns. Further, thanks to the logicolinguistic theories, we now understand that the theories which try to be maximally presuppositionless have a built-in bias towards the presupposable aspects of linguistic semiosis (hence their focus on referential function) and de-emphasize its entailing, 'creative' aspects (hence their implicit devaluation of interactional function). 13 Based on the achievements of these theories, we m a y now try to develop pragmatic theories which can adequately incorporate the entailing aspects of semiosis and interactional functions. This project, we shall pursue in the following section.

3. Incorporation of ethnographic dimensions of speech acts


If the logico-linguistic, structural-implicatural approaches seem to have reached an impasse, it should not lead us to either 'anti-theoretic empiricism', with its positivist cult of 'data' and 'statistically significant regularities', or 'anti-structuralist', 'anti-representationalist' or 'abso.lute contextualist' approaches to speech acts, which valorizingly focus on ' d y n a m i c interactions', the 'value-, authority- and powerindexing' aspects of speech acts and 'infinite semiosis' (by which a given discursive interaction can be given an infinite number of interpretations in terms of its sociocultural significances), while leaving the connections between linguistic structures, ~3 This does not mean that a reference is accomplished only when a referent is presupposable, or the non-referential significance of an interaction is exhausted by its effects. As phenomenologists (not to mention constructivists) emphasize, a reference creates a 'reality', and, as macro-sociologists (not to mention determinists) emphasize, an interaction is constrained by presupposable variables. This, however, should not obscure the basic difference between reference and interaction in terms of presupposing and entailing aspects of semiosis: i.e., the former is essentially more presupposing and less entailing than the latter. This is because interactional text (of 'what is done') unfolds itself in the order of 'here and now' of discursive interaction, while referential text (of 'what is said'), though anchored on the hic et nunc of interaction, does not, possibly except when it is 'self-referential'. (To put it simplistically, we don't necessarily talk about what we are doing here and now, but we always do something here and now when we talk.) Thus, reference is more presupposing (i.e., it more focally indexes something which can be presupposed to exist in the order of 'there and then', not 'here and now', of discursive interaction), and less entailing (i.e., it less focally indexes something which is brought into being in the order of 'here and now'), than interaction (cf. Silverstein, 1993).

W. Koyama / Journal of Pragmatics 28 (1997) 1-28

15

denotations and references, on one hand, and discursive interaction, power-relations and group-identity formations a'ad (con)textualizations, on the other, in social limbo. Given our metatheoretic considerations on the sociohistoric conditions of 'pragmatics', what we need is a comprehensive, coherent and consistent social linguistic theory which can do justice to these two different moments of speech acts in their own terms, and describe how they are interrelated. With this in our mind, let us underline what has turned out problematic in the speech act theories. First, as we; have seen, they anchored 'what we do with words' into 'explicit primary performatives', which, in retrospect, should have been distinguished from other speech act:L such as 'indirect speech acts' and 'hints' (ErvinTripp, 1976), in that the former (a) explicitly denote their illocutionary senses in the linguistic structure and pedocute them, in addition to (b) having 'formulaicized' and automatized (cf. Havrfinek, 1964) indexical relationships between signs and their sociocultural contexts of uses, while the latter have the second property only to some limited degrees (cf. Clark, 1979). That is, explicit primary performatives idiosyncratically do the following things at once: (1) the metapragmatic framing (interactional textualization) of indexical sinsigns through the explicit metapragmatic denotation of 'what is being done', (2) the ascription of a presupposable symbolic legisign (illocutionary sense) omo an interactional text by explicitly denoting it, and (3) the ascription of a presupposable indexical legisign (event-type) onto an interactional text through the sinsigns' re-presenting themselves as an indexical icon of the event-type; where the illocutionary sense denotes the event-type. Note the strong efficacy, in determining the sociocultural significances of a discursive interaction, of evoking the legisigns which ar,~ presupposable independent of the particular discursive interaction at issue, such ,ts verba dicendi (linguistic symbolic legisigns denoting speech types) in linguistic :;tructure, which is highly presupposable in discursive interactions using the denotational code. Secondly, both approaches reduce speech event participant incumbents (speech event participant referents), as opposed to roles (cf. Levinson, 1988), into Speaker and Hearer (or the referents of 'I' and 'you' in the performative prefix in the syntactically reductionist theories), which are used as primes to define the felicity conditions (or as the invariant referents of the hypothetical covert participant deictics, at least by implicit assumptions, ifrom which, in fact, the positing of the deictic NPs in the performative clause is hypothesized). The works of Goffman (1979), Ducrot (1984), Hill and Irvine (1993)~ Rumsey (1989), Vologinov (1973), clearly show the shortcomings of this assumption. That is, they point out the complexities of 'voicing', by which the reported/quoted speech of a narrated event (EN) character can be variously aligned with the reporting/quoting speech of a sign event (E s) participant. Thus, the ascription of intention to the 'animator', as opposed to 'principal', 'author', etc. (cf. Goffman, 1979), is not necessary warranted, and, moreover, even if intention is ascribable to the animator, the description of the referent(s) of the speaker (i.e., incumbents) may include, or at least require reference to, other individual/aggregate(s)/collective(s). Moreover, in the processes of 'voicing', the groupidentities and power-relations of the E s participants are (re)constituted through their relationships with the characters and eventualities in E N,

16

W. Koyama / Journal of Pragmatics 28 (1997) 1-28

Also, these works point out the significance of 'contextual surrounds' of the addressor-addressee dyad which are presupposingly and entailingly indexed ('contextualized') by the E s. That is, the presuppositional evoking of contextual surrounds of the dyad contributes to the relatively determinate fixing of the interactional significance of E s: e.g., the evoking of social identities and power-status presupposable independently of particular E s, as seen in the use of sen-sei 'teacher' by a teacher to refer to herself when addressing a student(s) instead of a first person deictic term, such as watasi, boku, ore, etc., in a Japanese classroom, which presupposingly evokes the institution of school, or in the use of 'Your mother' by a mother to refer to herself when addressing her child(ten) instead of T , as in 'Your mother doesn't like it.', which presupposingly evokes the institution of family. On the other hand, the entailed effects in the contextual surrounds may be the 'perlocutionary point' of an ongoing interaction, as when effects on sociocultural relationships between an E s participant and a bystander are the (often intended) targets of an interaction: e.g., 'innuendo' (cf. Goffman, 1979), the use of 'Brother-in-law' register in Guugu Yimidhirr (cf. Haviland, 1979). The implications of this line of research for approaches to speech acts must be clear: Since the fundamental notions of 'speaker' (S) and 'hearer' (H), on which the logico-linguistic speech act theories largely rest, have turned out to be cultural stereotypic concepts about idealized speech events which are empirically shown to depend on the complex variables involved in voicing and contextualization, the notion of 'speech act' must be redefined in relation to the categories involved in voicing and contextualization. That is, we need to reconstruct a theory of speech acts which is based on (1) E s (discursive interaction itself and interactional texts of what has/will have been done in the interaction), E N (referential text of what has/will have been said), and relationships between the two, and (2) textualization (metapragmatic regimentation of indexicals) and contextualization (indexings of presupposed and entailed contexts). Such a reformulation may have the effect of 'breaking through' the impasse of the agentive intention-based theories of 'speech acts', which have precluded the agentively unintended significance of interaction. Note that this development is consonant with our current understandings of referential pragmatics, in that (post)analytic philosophy has shown that the referential significances of discourse do not solely depend on the consciously accessible intention of 'the speaker' ('S'), due to the involvement of pure indexicality anchored on E s, the sociohistoric continuity of usage, socially warranted and contested authority to fix reference, socially shared cultural stereotypes and linguistic structure (cf. Kripke, 1972; Putnam, 1975), none of which is completely accessible to agentive awareness. In this section, I shall try to exemplify the consequences of this 'ethnographic' reformulation of speech acts, focusing on voicing and contextualization, by investigating an event reported by a novelist. The text-artifact I analyze is called Kei-go 'Honorifics', written in a standard variety of Japanese by a native linguist, Nishida (1987), who frames and quotes part of a short 'non-fiction' written by the realist novelist Ibuse Masuji which describes interactions between the uses of the then emerging standard varieties and nominal honorifics circa 1910, in a village in Hiroshima prefecture, west of the Kyoto-Osaka metropolitan area of Japan, humor-

w. Koyama / Journal of Pragmatics 28 (1997) 1-28

17

ously entitled ' T h a t " T u t - t s a " and " K u r o u z i - t s a n " fought against each other and I e n g a g e d m y s e l f in t r o u b l e d cogi~:ation on the t e r m i n o l o g y ' (abbreviations are listed in the A p p e n d i x , p a g e 25): "Ibuse Masuji was born in Kamo village, Fukayasu county, Hiroshima prefecture in Meiji 31 (1898), and lived there until he entered the dormitory of Fukuyama Middle School in Meiji 45 (1912) ... In this village, there were 'class distinctions in the appellations of parents by their children and the appellations of people in general. The child of landed class parents called her parents 'o-tto-san' [f(athe.r)] and 'o-kka-san' [m(other)], that of village councilors or influential men 'o-tot-tsan' If] and 'o-kaka-n' [m], that of landed farmers 'o-tou-yan' [f] and 'o-kaa-yan' [m], and that of tenant farmers 'o-tot-tsa' [f] and 'o-kaka' [m], and the child of gentle people was called 'X-san', that of the next class 'X-tsan', next 'X-yan', then 'Xtsa', and finally 'X-sa'. As to the intra-familial appellations of parents, when money came into their household, the parents made their children change them as a kind of status symbol, yet the social appellation of a persona could not be changed until s/he died ... Under these circumstances, the village chief, Kureuzi-tsan, and a village councilor, Tut-tsa, fought against each other, and the family of Tut-tsa started using, like the family of Kurouzi-tsan, 'o-tto-san' [f] and 'o-kka-san' [m]. Then, the family of Kurouzi-tsan 'boldly made their children say 'o-tou-san' [f in standard] and 'o-hta-san' [m in standard] in an urbane fashion. This was an epochal event in my village.' Then, on Tut-tsa side, reacting against this, 'the whole family started using urbane words. These were expressions like "o-kos-i-yas-u" [o- = BP; kos-i = 'come' ('subject' honorific)-GERUND; yas- = COP(POL); -u = Nonpast; "welcome" in Kyoto/Osaka dialect], "so-ya, so-ya, ook-i ni" [so- = nonproximal and nonproximad AD root; -ya = emotive/affirmative FP; so-ya = (repetition); ook-i 'big'-CON; ni = Adverbializer; "Yes, yes, (thank you) very much" in Kyoto/Osaka dialects], "ak-i-mah-en noo" [ak- = 'brighten'(intr.); -i- = CON; mah- = COP(POL); -en = NEG; noo = emotive FP; "Ah, it's not good" in Kyoto/Osaka dialects, though the FP, noo, seems to have been influenced by Hiroshima dialect, which uses no/noo quite extensively].' Thus, they borrowed Osaka dialect. Then, on Kurouzi-tsan side, 'the whole family started to use Tokyo dialect. Tokyo dialect, which Kurouzi-tsan learned, it was said, as he often went to a county office outside the village. One day I carefully listened to Kurouzi-tsan telling directions lo the foreman of a road construction team, and his usage indiscriminately added unnecessaiy sounds to word-final positions, such as "nee," "nee" [final vowel lengthened form of th,~ FP ne, idiosyncratically conative among the mostly emotive FPs. ne and nee index Spkr's consideration of Adrs, Spkr's considerateness, sophistication, irresolution. They can be also used vocatively with a softening effect in the recruitment of a person into the role of Adrs]. When the foreman asked him 'a-no isi wa huta-tu ni wari-yans-u-ka?' [a- = proximad AD root; -no = GEN; isi = 'stone'; wa = TOP; huta- = 'two'; -tu = NC [Oanimate]; ni = DAT; war- = 'break'(tr.); i- = GERUND; yans- = COP(POL; regional); -u = Nonpast; -ka = interrogative FP; "(Do you want us to) break that stone into two?"; Note the straightforwardn,~ss of the expression], Kurouzi-tsan answered, 'a-rya nee, a-rya war-an hou ga yok-ar-ou to omo-u ga nee'. [a- = proximad AD root; -rya ~-- -re + wa = -'entity' + TOP (Note the sloppiness of pronunciation; cf. Irvine, 1974); nee -- conative FP. a-rya = (repetition); war- = 'break'; -an = NEG; hou = Comparative; ga = NOM; yok- =

18

W. Koyama / Journal of Pragmatics 28 (1997) 1-28

'good'; ar- = be[~animate]; -ou = Dubitative; to = Quotative; omo- = 'think'; -u = Nonpast; ga = 'though'; nee = (conative) FP; "That, uh, that, (I) suppose it would be better if (you) wouldn't break (it), though... (don't you think so?)"; Note the extreme (and perhaps reportively exaggerated) indirectness and circumlocution of the expression under the circumstances]' The 'Tokyo dialect' which the family of the village chief picked up in the county office was only such. 'Then, one day, a burglar came to my house and said "ak-e-ro, to o ak-e-ro" [ak- = 'open'; -e- = Transitive; -ro = Imperative; to = 'door'; o = ACC; ak-e-ro (repetition); "Open, open the door! "; Note the brusque manner in which the message was delivered in the 'true' Tokyo dialect] over a sliding shutter... My grandfather reported to a policeman dispatched to our village, "The burglar used Tokyo dialect outside the door". The Tokyo dialect of the burglar seems to have had a bit different color from the Tokyo dialect of Kurouzi-tsan ...' After this incident, Tut-tsa told everyone: 'Kurouzi's Tokyo dialect is a fake, I've heard' [Note the objectifying and factualizing use of the hearsay construction, [s...-da]-sou-zya (-da = affirmative COP; -sou = hearsay; -zya = COP), in which the affirmative and authoritative "[s ...-da]" is evidentially quasi-quoted], so the family of Kurouzi-tsan stopped using their Tokyo dialect. This burglary took place in Meiji 43 (1910), when Ibuse was a sixth grader ... It is written that 'to o ak-e-ro' 'Open the door' was the first words of Tokyo dialect I've heard in my life .... " (Nishida, 1987: 82-84; translation and brackets mine) In analyzing this text-artifact to reconstruct what was (or w e r e ) done in the sign events which are narrated by my 'translation/quotation', it is necessary to distinguish at least five levels of semiosis: (1) what was said and done in the sign events narrated by the characters in Ibuse's 'non-fiction' story; (2) what was said and done by the characters in the sign events narrated by Ibuse (and Nishida and this writer), (3) what was said and done by Ibuse, other producers, their addressees and other 'readers', (4) what was said and done by Nishida, other producers, etc., and (5) what has/will have been said and done by this writer, etc. Let us, for expository purposes and for the avoidance of interpretive conflict of interest, focus on the first three of these, and start with (3) what was said and done by Ibuse, etc. Without much interpretive ingenuity, we may observe the following tropological analogizations which heavily deploy presupposable cultural stereotypes: Tokyo dialect : Hiroshima dialect :: urban : countryside :: burglar=intruder : intruded :: imperative : (non-imperative). That is, a schema of cultural stereotypes is so constructed that the impolite 'standard language' is shown to have forcefully intruded into a peaceful home village speaking a dialect of halcyon days. Thus, beneath the false sophistication of the village chief's 'standard language' debunked was the intrusive, imperative nature of 'the true Tokyo dialect', out of which modern standard varieties were emerging around 1900 (cf. Sanada, 1991), used by burglars who were disseminated, along with 'the standard language', from the new nation-state's center, Tokyo. Such is a trope of modernism forcefully deployed by Ibuse, a quintessential realist novelist, who ritualistically literalizes the image of 'peaceful home village' located at the polar opposite of 'cantankerous, conflictual and alienating

W. Koyama / Journal of Pragmatics 28 (1997) 1-28

19

metropolis', i.e., an anti-modern, nostalgizing culturalist ideology of huru-sato 'good old home village', which was, in fact, disseminated by the Meiji oligarchs to negatively valorize the increasingly volatile urban labor movements (cf. Gluck, 1985) and which many Japanese still buy today by commoditizing folksongs, folktales and 'exotic pre-modern Japan' (cf. Ivy, 1995). Note that this native culturalist ideologization of nonstandard dialectal varieties ideologizes the 'true' standard varieties as something deprived of 'fine character', in opposition to the referentially-focused modernist ideologization, whick positively valorizes them in terms of 'communicability' and 'techno-scientific efficiency'. Also noteworthy is Ibuse's narration (except the quotes) in a standard (quasi-Tokyo), not Hiroshima dialectal, variety, a register which is posited as sharable between lbuse and the reader, but not shared by the characters in his narrated ev,~nt (except, possibly, the burglar). Thus, by virtue of their sharing of the common, 'superposed' register (cf. Gumperz, 1968), the reader gets aligned with Ibuse, who, like most of his contemporary readers, moved from a country village to a metropolis lTokyo), worked hard and succeeded in acquiring his 'higher' status and standard varieties, and who, from that vantage point, humorously and comfortably nostalgizes the 'Japan' of yesterday, where what 'we' have lost with the coming of modernization and standardization, 'our natural, native and national culture', a bit goofy and obsolescent, but definitely heart-wanning, was still intact, unscathed, and safe from the volatile yet undeniably (if only materially) comfortable effects of Westernization and modernization. Thus, as an individual becomes the animator, i.e., the reader, of Ibuse's 'non-fictional' text-artifact, she simultaneously becomes the addressee, i.e., the reader, of Ibuse and other producers, and entextualizes with them interactional texts, in which, in contrast to the realist polyphony of the dialectal and pseudo-superposed varieties of speech made by the narrated event characters, the sl:andard variety which Ibuse and the reader speak and the reader hears Ibuse and herself speaking aligns these two parties together in their footing at the top and center of the register hierarchy in the order of hic et nunc, from which, nostalgizingly and condescendingly, the image of 'dialects' is projected to the 'polyphonic' lower periphery and the 'remembered' past. As we move to the second level of what was said and done by the characters in the sign events narrated by Ibuse (and Nishida and this writer), we immediately realize that the appellation by which a child (Spkr) referred to her parents (qua Adrs) was construed to be indexical of Adrs' relative status vis-a-vis other parents, i.e., the indexically suggested degrees of 'deference entitlement' (Goffman, 1967: 47-95; Shils, 1982) which bystanders, i.e., the villagers, owed to the family. This points to how the personal indexicals function to reconstitute the social 'realities' (Lebenswelten) beyond the immediate dyad of 'S' and 'H', and the Ref in the E N, despite the basic regularities of verbal honorification consisting of (1) 'polite' (tei-nei: deference entitlement; Spkr ~ Adrs), (2) 'respect' (son-kei: deference entitlement; Spkr Ref of 'subject' NP) and (3) 'humble' (ken-zyou: deference entitlement, as evaluated by Spkr; Ref of 'subject' NP --~ Ref of non-'subject' NP) and of nominal honorifications by the uses of (1) beautificatory prefixes (demeanor of Spkr vis-a-vis Adrs), (2) sociocentric terms, such as 'status terms', where deictics are available, (3) 'non-singular' deictics instead of 'singular' deictics, (4) referentially opaque (instead

20

W. Koyama / Journal of Pragmatics 28 (1997) 1-28

of transparent) head nouns modified by anaphorico-demonstratives, and (5) 'respectful' second person and 'humble' first person deictics, none of which includes the notion of bystander. Note that, as recognized by Brown and Levinson (1987:12), the theory of politeness based on the Goffmanian notion of 'face' and the Gricean notion of agentive intentionally-based 'implicatures' has difficulty describing and explicating deferential speech acts where bystanders are contextualized. This indicates that this type of phenomena requires more extensive theorization on '(indexical) contextualization' and 'community'. A theory equipped with the notions of indexicality, metapragmatics and speech community may theorize these speech acts as the indexical entailments of group-identities and power-status relationships in a speech community whose members share the same metapragmatic orientations towards the superposed registers of honorifics. That is, a speech community (which is indeed partially constituted by the sharing of the same metapragmatic orientations towards honorific registers) ideologically projects the properties of 'refinement', 'elevatedness', and other positive values onto linguistic types which are called 'honorifics', and the 'possessor' of such positively valorized valuables is socially valorized to the extent that she has these valuables in her repertoires. (How a cultural ideology interprets such values, crucially, varies inter- and intra-culturally: e.g., a manifestation of insecurity, a part of symbolic capital as a substitute for economic capital (cf. Irvine, 1974), or simply a direct manifestation of 'power'.) Thus, within a speech community which is partially constituted by the sharing of such a linguistic ideology, the degrees to which smaller speech communities (such as families) 'possess' gradiently valorized 'honorific valuables' partially constitute the hierarchization of these smaller units in the speech community, the members of a smaller unit being sometimes within and sometimes without the immediate dyad of discursive interactions. Let us now finally move to the interactions between the first level of (1) what was said and done in the sign events narrated by the characters in Ibuse's 'non-fiction' story, and (2) what was said and done by the characters in the sign events narrated by Ibuse (and Nishida and this writer). We shall focus on the final section of Ibuse's story, where, after the debunking of Kurouzi's fake Tokyo dialect due to the 'real' Tokyo dialect used by the burglar, Tut-tsa told everyone: 'Kurouzi no tou-kyou-ben wa nise-mono da sou-dya'. [no = GEN; tou-kyou-ben = Tokyo dialect; wa = TOP; nise-mono = '(a) fake'; da = COP; sou- = Hearsay; -dya = COP; dialectal); 'Kurouzi's Tokyo dialect is a fake, I've heard.'] (Nishida, 1987: 84; translation and brackets mine). First of all, we cannot fail to see the disappearance of the honorific title, -tsan, suffixed to the proper first name of Kurouzi, indexing the fallen status of the village chief, fallen not only from the top and center of standard varieties in relation to dialectal varieties as his 'Tokyo dialect' was shown to be inauthentic, but also from the position only second to the top and center of regional honorific title hierarchy to its very bottom: [-san > -tsan > - y a n > -tsa > -sa > -0]. Second, more importantly, note the very effective use of the special hearsay marker, sou-, which grounds Tut-tsa's referential practice 14 on the authoritative speech event in which someone

14 Needless to say, this is mediated by Ibuse's 'voice', among others.

W. Koyama /Journal of Pragmatics 28 (1997) 1-28

21

who is stipulated to know the 'real' Tokyo variety, a non-villager, from beyond the village, who is conversant with Tokyo and Tokyo dialect, that is, someone who is authentically at the top and center of social and register hierarchies, has passed her verdict on Kurouzi's speech genre and Kurouzi himself as a fake. Notice that, here, Tut-tsa appoints himself as the animator of the stipulated authority, with whom he accuses Kurouzi-tsan of a deadly sin of linguistic forgery. Thus, by using the special hearsay construction, Tut-tsa rigidly determines the interactional significances of not only what he is doing with his words (raising himself 'up' and lowering his nemesis, now an infamy in the speech community), but also what Kurouzi-tsan has done with his words. Note, further, the use of the standard variety copula da in the quasiquoted speech, in contrast to the use of the regional copula dya in the matrix. The first copula indexes the authenticity of the authority whose speech Tut-tsa says he is reporting (and notice that the stripping of a suffixal title from Kurouzi is done in the quasi-quoted, embedded sentence under the authority of some non-villager), while the latter squarely locates Tut-ts a among the villagers, in opposition to Kurouzi-tsan, who used a 'Tokyo dialect', which, worse yet, has turned out to be a fake. Recall, moreover, that Tut-tsa himself was using a dubious Kyoto/Osaka dialect, not a Hiroshima dialect, when he and his family were in competition with Kurouzi and his family, a competition in which the former were behind the latter, due to the metaphoric superimposition of 'the second city' status of Kyoto/Osaka vis-d-vis 'the Great Tokyo' (dai-tou-kyou) upon the 'speakers' of the dialects spoken in these regions, coupled with the presupposable status asymmetry between Tut-tsa, a village councilor, and Kurouzi-tsan, a village chief.

4. Conclusion

By way of conclusion, one might draw certain implications from these textual interpretations vis-gt-vis 'social theory', 'discourse analysis', and 'pragmatics'. First, most pertinently to 'social theory', they indicate that a given interaction may have both irreducibly conflictual and consensual ('normative') dimensions. Seemingly conflictual interactions, such as the ones between Tut-tsa and Kurouzi-tsan, may presuppositionally index sharable norms of interactions ('consensus'), such as those which constitute an 'honorific hierarchy', if only to exploit them, purposively (intentionally, voluntarily) or otherwise. Inversely, as many sociolinguistic studies have shown, seemingly consensual interactions, such as those which conform to the 'covertly prestigious' regularities (norms) which partially constitute a speech community, may stand in an oppositional, alternative or complicitly collaborative, yet apparently antagonistic, relationship to another community (cf. Woolard, 1985). This argues against the simplistic advocacy of homogeneously conflictual or consensual sociologies (cf. Fraser, 1989: 117-118). (It may be ironic if the conscious theorizing behaviors of some proponents of 'conflictual sociology" again:~t the norm(ativity) of Parsonsism may have recreated the hegemonic, normative status of (neo)functionalisms and the anti-norm(ative), 'loyal oppositional' status of conflictual sociologies in a discourse genre of 'American sociology' through, at least potentially, complicitly collaborative antagonisms.)

22

w. Koyama / Journal of Pragmatics 28 (1997) 1-28

Moreover, as Calhoun (1995) observes and our examples demonstrate, a speech event - or, more generally, a sign(ifying) event - is usually multiply and simultaneously embedded in several speech communities with partially overlapping or inclusively hierarchical extensions, j5 Thus, a conflictual interaction in a speech community may index the norms which partially constitute another speech community, such as standard varieties, while seemingly focusing on the (re)formations of power-relations within the former community. This suggests that to interpret social significances of a speech event solely in terms of the asymmetry of power-relations between two (or more) speech communities (as 'dominant' and 'subordinate' 16 communities) might obscure the centrality of (re)formations of power-relations and group-identities within a subordinate community for its members, theoretically homogenize these communities ('positive sociology' in its non-Popperian, post-Critical Theoretic sense), and, consequently, fail to investigate the complexity of interrelationships between dominant and subordinate communities. Second, most pertinently to 'discourse analysis', our textual analyses show how micro-social events 'make sense', or can be shown richly interpretable and interactionally cohesive (something akin to 'maximally humanized' in Vico's sense), when analyzed in conjunction with (hypothetically positable) macro-social contexts, and, inversely, how (putative) macro-social regularities are contextually indexed and substantiated by micro-social events and thus (re)acquire their 'virtual reality', or 'evokability' as perduring regularities, be they indexical or symbolic. This points to dialectic interrelations between micro-social events and macro-social patterns. That is, the former 'evoke' (presupposingly index), contextualize and thus potentially change the latter as the sociocultural significances of these micro-social sign events and other social regularities indexed by them interact with, and potentially transform, the latter; while the latter, when evoked, constrain, or contribute to the fixing of, the social significances of the former. Third, most pertinently to 'pragmatics', note that our textual analyses have focused on person indexicals (participant deictics, proper names and suffixal titles), evidentials, dialectal and superposed (standard and honorific) varieties, as well as (referentially implicit) indexing of bystanders: i.e., tokens of indexical legisigns, which are characterized by their 'metapragmatic transparency', relative to other linguistic types (cf. Kripke, 1972; Putnam, 1975; Silverstein, 1987). That is, the tokens of person indexicals and mood/modality markers ('shifters'), whose semantic contents and denotata, not to mention referents, are metapragmatically characterizable, relatively transparently point to indexical contexts, both what is referentially said and what is interactionally done in the speech events in which they are deployed. 17 ~5 It is, therefore,gratuitous to assume that only 'one thing' is going on at a given phase of interaction, as some pragmaticians and other social theorists seem to do (cf. Goffman, 1977). ~6 'Domination' is not used in its technical, Gramscian sense. ~7 On the face of it, there is no conceptualmotivationfor the tokens of metapragmaticallycharacterizable types to point to both referential and interactional contexts (just as there is no such motivation for them to point to referentialcontext only.) However,that referential texts (of 'what is said') are projected from discursive interactions and anchored on them through (metapragmaticallycharacterizable)indexical components (shifters, such as deixis), and interactional texts (of 'what is done') are immanently

W. Koyama / Journal of Pragmatics 28 (1997) 1-28

23

M o r e o v e r , it is o b v i o u s that the tokens o f dialectal and s u p e r p o s e d varieties, notions w h i c h are prima facie m e t a p r a g m a t i c a l l y characterized, relatively transparently index the sociocultural contexts o f the speech events in w h i c h they occur, as m o s t r o b u s t l y seen in ' d i g l o s s i a ' , honorifics and g e n d e r e d registers. Thus, it is along the d i m e n s i o n s o f the tokens o f such m e t a p r a g m a t i c types that w e can reconstruct an interpretive m o d e l ( ' t e x t ' ) o f interaction f r o m a given referential text (which is easier to identify), with a sufficient d e g r e e o f interpretive d e t e r m i n a c y to characterize the sociocultural significances o f a given discursive interaction. (Note that this account gives theoretically g r o u n d e d and e m p i r i c a l l y d e m o n s t r a b l e constraints on the potentially infinite interpretability o f l i n g u i s t i c a l l y - c e n t e r e d social actions.) O f course, such an analysis o f non-referential speech acts as exemplified above is not altemative but c o m p l e m e n t a r y to the ' g a r d e n - v a r i e t y ' pragmatics which focuses on the referential aspects o f speech a,:ts, insofar as the latter does not try to theorize nonreferential speech acts by simply transferring the analytic modus operandi and habitus d e v e l o p e d to deal with referential speech acts, which are m o r e or less properly analyzable b y e m p l o y i n g , mutatis mutandis, e x t e n d e d versions o f semantic analyses, such as various p o s t - G r i c e a n , U - s e m a n t i c , p r o t o t y p e semantic, s c h e m a and other cognitivist theories; 18 or insofar as it does not i m p l i c i t l y or e x p l i c i t l y d e v a l u e the i m p o r t a n c e o f non-referential p r a g m a t i c s : i.e., insofar as they do not semanticize pragmatics.

inscribed in discursive interactions, from which shifters are projected onto referential texts, indicates that the tokens of metapragmatically characterizable types point to both referential and interactional contexts. Empirically, moreover, the transparent i:adexing of group-identities and power-relations via participant deictics and honorifics (cf. inclusive vs. exclusive first person deictics, 'power semantics', to cite the most obvious) indicates the validity of such theorization. Thus, as Silverstein (1992) argues, deixis is to be conceptualized as figure of interaction transparently projected from the plane of interaction onto that of reference. ~8 Note, however, that the cognitivist tbeories usually fail to address the social indexical dimensions of reference ('sociolinguistic division of labor'), pointed out by Putnam (1975), whose problematic is displaced through a cognitivist misrecognition that it is an analytic philosopher's last-ditch attempt to rescue 'Aristotelian' theories of categorization from the 'revolutionary' Roschean prototypic theories. (Of course, without discrete categorizations, be they linguistic - conjointly and implicitly characterized by Saussurean, system-internal contrastive definitions and focality-based correspondences between semantic intension and extension - or stereotypic - often explicitly defined, as in logico-mathematical, (para)scientific, juridical taxonomies - a theory of reference equipped only with 'prototypic' stereotypes based on degree-of-likeness cannot account for the discrete categorial aspects of semantics, denotation and reference, such as (1) boundary-imposing constraints on unbounded categorizations based on degree-of-likeness; (2) compositionality of categories; (3) logically empty categories (e.g., 'apple that is not an apple'), whose probability must be more than 0 according to prototype theories; (4) logically universal categories (e.g., 'fruit that either is or is not an apple'), whose probabilities must be less than 1 according to them; (5) 'truth conditions of inclusion' (e.g., the existence of a perfect grizzly bear in Alaska must suffice to falsify the proposition that 'All grizzly bears live in North America', given that Alaska is not the most typical locale in North America, etc. (cf. Osherson and Smith, 1981, 1982)). More fundamentally, such a misrecognition seems to be motivated by the cognitivists' unwillingness to note the significance of the. epistemico-ontically asymmetric relationship between interactional texts, immanently entextualized in discursive interactions, and referential texts, projectedfrom discursive interactions. Then, since referential texts are anchored on interactions, which are 'always already' sociohistorically situated, they cannot help but have a social indexical dimension, though it is not as focal and foregrounded as in non-referential, interactional texts.

24

w. Koyama / Journal of Pragmatics 28 (1997) 1-28

Thus, a pragmatic theory which does justice to both referential and non-referential speech acts needs to incorporate morphosyntactico-semantics (symbolic legisigns), cultural stereotypes (symbolic legisigns), indexical legisigns (shifters, conversational regularities, discourse genres, registers, etc.), and indexical sinsigns (both at the nth and n+lth orders) such as historic contingencies or the irreducible uniqueness of events (pace positivists for whom only statistically significant regularities count). 19 And in so doing and getting operationalized, a linguistically-sophisticated social theory of practice may contribute to breaking through the antinomies which have characterized our time so long, perhaps only to bring in a new set of aporias (not just 'puzzles' which can be solved by 'normal scientific' procedures) which may be as difficult to displace as ours. When such aporias emerge in our horizons of (meta)theorizations, there will be a sense in saying, retrospectively from that vantage point, that a certain 'progress' (or Taylorian 'epistemic gain') has been made, and we can start reconstructing (or 'improving' from our renewed agentive perspectives) our theories of speech acts and metatheories of theorization, through a dialectic, or 'double hermeneutic' (Giddens, 1993), relationship with the new contexts of (meta)theorization, which our (meta)theories will have helped to create (however marginal or indirect such a contribution might be), in which they will be situated, and which will compel our (Bachelardian) critically self-reflective rationality to see the emerging fundamental inadequacies of our (meta)theories and deepen, expand or enrich our understandings of language, social practice and social theory. At the present, all this remains essentially (but not completely) indeterminate, as our agentively accessible understandings and intended consequences of these discursive interactions in which we (will) have been engaged, however carefully, thoroughly and self-reflectively critically formulated they might be, are essentially defeasible, and may not coincide with their interpretations reconstructible in vitro. Notwithstanding, we can at least try to minimize the risk of being wrong-headed by trying to understand what we are doing and make explicit the metatheoretic virtues which orient 'our' theorizations and which 'we' (inclusive, hopefully) cherish: steering between the Scylla of ethnocentrism, Universalism, decontextualism, Scientism, physicalism and Realism, and the Charybdis of Culturalism, Relativism, Historicism, the doctrine of cultural arbitrary and solipsism; non-dogmatically respecting and critically appreciating 'rival' (or even 'incommensurable') research programs, other disciplines and 'radically' different cosmographic traditions; aspiring after comprehensiveness, consistency, analytic precision and empirical accountability; taking a critical stance towards the 'basic' (or background) assumptions of our own theories; avoiding Whiggish historicizations; trying to understand and, if profitable, modify our theorizations in relation to metatheories and sociohistoric contexts of (meta)theorizations; trying to figure out strategies to effectively mobilize our expertise to confront the pressing issues of our time; and 'deconstructing' closure by constantly reminding ourselves that our 19 '(Historic) contingency', as used here, does not mean merely the lack of absolute or a piori (pre)determinacy of events, but their irreducibilityinto the tokens of (statistic or nonstatistic) regularities, which,accordingto Hacking(1990), Peirce, Nietzscheand Mallarm6saw. In the Peirceanjargon, it is that aspect of sinsign which is not reducible to the tokenhoodof a type (cf. note 2).

W. Koyama / Journal of Pragmatics 28 (1997) 1-28

25

t h e o r i e s , s o c i e t i e s a n d o t h e r h u m a n p r o c e s s e s and p r o d u c t s are a l w a y s i m p e r f e c t and thus i m p r o v a b l e . 2

Appendix: Abbreviations
AD Adrs BP CON COP DAT FP GEN GERUND NC NEG NOM POL Ref Spkr TOP --~ = = = = = = = = --= = = = = = = = anaphorico-demonstrative addressee(s) beautificatory prefix, go- or oconnective copula dative case-marker final particle genitive case-marker gerundivizer nominal classifier negative adjective nominative case-marker 'politeness'(Spkr --, Adrs)-marker referent(s) speaker topic-marker direction of deference entitlement

References
Anderson, Benedict, 1991 [1983]. Imagined communities: Reflections on the origin and spread of nationalism. London: Verso. ,~pel, Karl-Otto, 1991. Is intentionality more basic than linguistic meaning? In: E. LePore and R. van Gulick, eds., 31-55. Austin, J.L., 1962. How to do things with words. Cambridge, MA: Harvard University Press. Bakhtin, Mikhail M., 1981. The dialo;~ic imagination: Four essays. Austin, TX: University of Texas Press. Bar-Hillel, Yeoshua, 1954. Indexical expressions. Mind 63: 359-376. Bloomfield, Leonard, 1927. Literate and illiterate speech. American Speech 2(10): 432~,39. Bloomfield, Leonard, 1984 [1933]. Language. Chicago, IL: University of Chicago Press. Bourdieu, Pierre, 1980. Le sens pratique. Paris: Les Editions de Minuit. Bourdieu, Pierre, 1982. La production et la reproduction de la langue 16gitime. In: Ce que parler veut dire: L'rconomie des 6changes tinguistiques, 23-58. Paris: Arth~me Fayard. Brown, Penelope and Stephen C. Levinson, 1987 [1978]. Politeness: Some universals in language usage. Cambridge: Cambridge University F'ress. Calhoun, Craig, 1995. Critical social theory: Culture, history and challenge of difference. Oxford: Blackwell.

20 These virtues are often difficult to sustain simultaneously and we need to be engaged in complex acts of weighing one virtue against another. That is, we may need an (implicit) meta-meta-theory, and obviously the stacking of 'meta-' can ~;o ad infinitum, 'theoretically' speaking. However, since theorizing is a kind of human praxis, there is :~omethreshold level above which the stacking of 'meta' becomes superfluous (cf. note 9).

26

W. Koyama / Journal of Pragmatics 28 (1997) 1-28

Clark, Herbert H., 1979. Responding to indirect speech acts. Cognitive Psychology 11 : 430-477. Clark, Herbert H. and Danna Wilkes-Gibbs, 1986. Referring as a collaborative process. Cognition 22:
1-39.

Clifford, James and George E. Marcus, eds., 1986. Writing culture: The poetics and politics of ethnography. Berkeley and Los Angeles, CA: University of California Press. D'Amico, Robert, 1989. Historicism and knowledge. Routledge: London. de Man, Paul, 1978. The epistemology of metaphor. In: S. Sacks, ed., On metaphor, I 1-28. Chicago, IL: University of Chicago Press. Den-ida, Jacques. 1967. La voix et la phrnomrne. Paris: Presses Universitaires de France. Donnellan, Keith S., 1978. Speaker reference, descriptions and anaphora. In: P. Cole, ed., Syntax and semantics, Vol. 9: Pragmatics, 47~i8. New York: Academic Press. Ducrot, Oswald, 1984. Langage, mrtalangage et performatifs. In: Le dire et le dit, 117-148. Paris: Minuit. Duranti, Alessandro, 1993. Intentions, self, and responsibility. In: J. Hill and J.T. lrvine, eds., 24-47. Ervin-Tripp, Susan, 1976. Is Sybil there? The structure of some American English directives. Language in Society 5: 25-66. Foucault, Michel, 1966. Les mots et les choses: Une archrologie des sciences humaines. Paris: l~ditions Gaillmard. Fraser, Nancy, 1989. Unruly practices: Power, discourse, and gender in contemporary social theory. Minneapolis, MN: University of Minnesota Press. Garfinkel, Harold, 1972 [1967]. Remarks on ethnomethodology. In: J.J. Gumperz and D. Hymes, eds., Directions in sociolinguistics: The ethnography of communication, 301-324. London: Blackwell. Geertz, Clifford, 1973. The interpretation of cultures. New York: Basic Books. Geertz, Clifford, 1984. Distinguished lecture: Anti anti-relativism. American Anthropologist 86(2): 263-278. Gellner, Ernest, 1959. Words and things: A critical account of linguistic philosophy and a study in ideology. London: Gollancz. Giddens, Anthony, 1993 [1974]. New rules of sociological method: A positive critique of interpretative sociologies. Stanford, CA: Stanford University Press. Gluck, Carol, 1985. Japan's modem myths: Ideology in the late Meiji period. Princeton, NJ: Princeton University Press. Goffman, Erving, 1967. Interaction ritual. New York: Pantheon Books. Goffman, Erving, 1977. Frame analysis. New York: Harper and Row. Goffman, Erving, 1979. Footing. Semiotica 25: 1-29. Goodman, Nelson, 1979. Fact, fiction, and forecast. Cambridge, MA: Harvard University Press. Grice, H. Paul, 1989a [1958]. Postwar Oxford philosophy. In: Studies in the way of words, 171-180. Cambridge, MA: Harvard University Press. Grice, H. Paul, 1989b [1982]. Meaning revisited. In: Studies in the way of words, 283-303. Cambridge, MA: Harvard University Press. Gumperz, John J., 1968. The speech community. In: D.L. Sills, ed., International encyclopedia of the social sciences, Vol. 9, 381-386. New York: Macmillan. Habermas, Jiirgen, 1991. Comments on John Searle: Meaning, communication, and representation. In: E. LePore and R. van Gulick, eds., 17-29. Hacking, Ian, 1990. The taming of chance. Cambridge: Cambridge University Press. Hampshire, Stuart, 1959-60. J.L. Austin. Proceedings of the Aristotelian Society 60: i-xiv. Hancher, Michael, 1979. The classification of cooperative illocutionary acts. Language in Society 8: 1-14. Hanks, William F., 1990. Referential practice: Language and lived space among the Maya. Chicago, IL: University of Chicago Press. Haviland, John B., 1979. Guugu Yimidhirr brother-in-law language. Language in Society 8: 365-393. Havr~inek, Bohuslav, 1964 [1932]. The functional differentiation of the standard language. In: P.L. Garvin, eds., A Prague School reader on esthetics, literary structure, and style, 3-16. Washington, DC: Georgetown University Press. Hill, Jane H. and Judith T. Irvine, eds., 1993. Responsibility and evidence in oral discourse. Cambridge: Cambridge University Press.

W. Koyama /Journal of Pragmatics 28 (1997) 1-28

27

Hockett, Charles F., 1954. Two models of grammatical description. Word 10: 210-231. Irvine, Judith T., 1974. Strategies of status manipulation in the Wolof greeting. In: R. Bauman and J. Sherzer, eds., Explorations in the ethnography of speaking, 167-191. Cambridge: Cambridge University Press. Ivy, Marilyn, 1995. Discourses of the vanishing: Modernity, phantasm, Japan. Chicago, IL: University of Chicago Press. Kripke, Saul, 1972. Naming and necessity. Cambridge, MA: Harvard University Press. Kristeva, Julia, 1974. La r6volution du langage po6tique. Paris: Seuil. Kuhn, Thomas S., 1962. The structure of scientific revolutions. Chicago, IL: University of Chicago Press. LePore, Ernest and Robert van Gulick, eds., 1991. John Searle and his critics. London: Blackwell. Levinson, Stephen C., 1981. The essential inadequacies of speech act models of dialogue. In: H. Parret, M. Sbish and J. Verschueren, eds., Possibilities and limitations of pragmatics, 473-492. Amsterdam: Benjamins. Levinson, Stephen C., 1983. Pragmatics. Cambridge: Cambridge University Press. Levinson, Stephen C., 1988. Putting linguistics on a proper footing: Explorations in Goffman's concepts of participation. In: P. Drew and A. Wootton, eds., Erving Goffman: Exploring the interaction order, 161-227. Boston, MA: Northeastern University Press. Lyons, John, 1977. Semantics (two voltmes). Cambridge: Cambridge University Press. Nichols, Johanna, 1984. Functional theories of grammar. Annual Review of Anthropology 13:97-117. Nishida, Naotoshi, 1987. Kei-go. (Honorifics) Tokyo: Tou-kyou-dou. Osherson, Daniel N. and Edward E. Smith, 1981. On the adequacy of prototype theory as a theory of concepts. Cognition 9: 35-58. Osherson, Daniel N. and Edward E. Smith, 1982. Gradedness and conceptual combination. Cognition 12: 299-318. Peirce, Charles S., 1932. Collected papers of Charles Sanders Peirce, Vol. 2. (C. Hartshorne, P. Weiss and A. Burks, eds.) Cambridge, MA: Harvard University Press. Polanyi, Michael, 1958. Personal knowledge: Towards a post-critical philosophy. Chicago, IL: University of Chicago Press. Putnam, Hilary, 1973. Reductionism and the nature of psychology. Cognition 2: 131-146. Putnam, Hilary, 1975. The meaning of 'meaning'. In:. Philosophical papers, Vol. 2: Mind, language and reality, 215-271. Cambridge: Cambridge University Press. Putnam, Hilary, 1990. Realism with a l~uman face. Cambridge, MA: Harvard University Press. Putnam, Hilary, 1992. Renewing philosophy. Cambridge, MA: Harvard University Press. Ricoeur, Paul, 1986 [1975]. Geertz. In: G.H. Taylor, ed. and trans., Lectures on ideology and utopia, 254-266. New York: Columbia University Press. Rorty, Richard M., ed., 1967. The linguistic turn: Recent essays in philosophical method. Chicago, IL: University of Chicago Press. Rorty, Richard M., 1979. Philosophy and the mirror of nature. Princeton, NJ: Princeton University Press. Rosaldo, Michelle Z., 1982. The things we do with words: llongot speech acts and speech act theory in philosophy. Language in Society 11 : 203-237. Rumsey, Alan, 1989. Grammatical person and social agency in the New Guinea highlands. In: B. Music, R. Graczyk and C. Wiltshire, eds. Parasession on language in context, 242-253. Chicago, IL: Chicago Linguistic Society. Sadock, Jerrold M., 1974. Toward a linguistic theory of speech acts. New York: Academic Press. Sadock, Jerrold M., 1978. On testing fcr conversational implicature. In: P. Cole, ed., Syntax and semantics, Vol. 9: Pragmatics, 281-297. New York: Academic Press. Sahlins, Marshall, 1976. Culture and practical reason. Chicago, IL: University of Chicago Press. Sanada, Shinji, 1991. Hyou-zyun-go w,~ ika ni sei-ritu si-ta ka? (How was the standard language constructed?) Tokyo: Sou-taku-sya. Schlick, Moritz, 1967 [1932]. The future of philosophy. In: R.M. Rorty, ed., 43-53. Searle, John R., 1969. Speech acts. Cambridge: Cambridge University Press. Searle, John R., 1975. A classification of illocutionary acts. Language in Society 5: 27-45. Searle, John R., 1979. Expression and meaning. Cambridge: Cambridge University Press.

28

W. Koyama / Journal of Pragmatics 28 (1997) 1-28

Searle, John R., 1980. The background of meaning. In: J.R. Searle, F. Kiefer and M. Bierwisch, eds., Speech act theory and pragmatics, 221-232. Dordrecht: Reideel. Searle, John R., 1983. Intentionality. Cambridge: Cambridge University Press. Shils, Edward, 1982 [1968]. Deference. In: The constitution of society, 143-173. Chicago, IL: University of Chicago Press. Silverstein, Michael, 1979. Language structure and linguistic ideology. In: P. Clyne, W. Hanks and C. Hofbauer, eds., The elements: A parasession on linguistic units and levels, 193-247. Chicago, IL: Chicago Linguistic Society. Silverstein, Michael, 1981. The limits of awareness. Working papers in sociolinguistics, no. 84. Austin, TX: Southwestern Educational Laboratory. Silverstein, Michael, 1987. Cognitive implications of a referential hierarchy. In: M. Hickmann, ed., Social and functional approaches to language and thought, 125-164. Cambridge: Cambridge University Press. Silverstein, Michael, 1992. The indeterminacy of contextualization: When is enough enough? In: A. DiLuzio and P. Auer, eds., The contextualization of language, 55-76. Amsterdam: Benjamins. Silverstein, Michael, 1993. Metapragmatic discourse and metapragmatic function. In: J.A. Lucy, ed., Reflexive language: Reported speech and metapragmatics, 33-58. Cambridge: Cambridge University Press. Stocking, Jr., George W., 1982 [1968]. Race, culture, and evolution: Essays in the history of anthropology. Chicago, IL: University of Chicago Press. Taylor, Charles, 1985. Philosophical papers, Vol. 1: Human agency and language. Cambridge: Cambridge University Press. Urmson, J.O., 1967 [1965]. J.L. Austin. In: R.M. Rorty, ed., 232-238. Urmson, J.O. and G. Warnock, 1961. J.L. Austin. Mind 70: 256-257. Van Valin, Jr., Robert D., ed., 1993. Advances in role and reference grammar. Amsterdam: Benjamins. Vologinov, V.N., 1973 [1929]. Marxism and the philosophy of language. Cambridge, MA: Harvard University Press. White, Hayden, 1978. Tropics of discourse: Essays in cultural criticism. Baltimore, MD: Johns Hopkins University Press. Williams, Raymond, 1977. Marxism and literature. Oxford: Oxford University Press. Williams, Raymond, 1989. The politics of modernism: Against the new conformists. London: Verso. Woolard, Kathryn A., 1985. Language variation and cultural hegemony: Toward an integration of sociolinguistic and social theory. American Ethnologist 12: 738-748.