Page 449 - Encyclopedia of Philosophy of Language

sounds or phonemes, and leaves syntactic objects such as sentences to the syntacticians; similarly, the syn- tactic component does not interfere in the workings of syntax except in a sideways fashion (the com- ponents are not separated by watertight dividers, of course, as even the staunchest 'componentialists' will admit).
By contrast, in a perspective view, pragmatics could be said to serve as an 'umbrella' for the modules of linguistics, its components. In the words of the Finn Jan-Ola Ostman, pragmatics should probably not be seen as 'belonging to the contrast set of psycho- linguistics, sociolinguistics, etc., but rather as being the umbrella term for these and other (semi-)hyphen- ated areas in linguistics' (1988:28; italics added).
A natural extension of this view would be to let the 'component' and the 'perspective' conceptions exist side by side: after all, they both are metaphors designed to expand, not to narrow, the epis- temological horizon. There could be a structural com-
ponent (such as phonology, a part of the system of language) along with a structuralperspective, that is, a way of looking at language (in this case, phonology) as a structured system. In the same vein, one could have a pragmatic component, understood as the set of whatever pragmatic functions can be assigned to language, along with a pragmaticperspective, that is, the way in which these functions operate within the single units of the language system, respectively of language use.
Summarizing this view, Ostman uses an analogy: if 'the unit of analysis in semantics [is] simply meaning: the meanings of words, phrases, larger constructions, prosody, and so on,... then by the same token, the "unit" of analysis for pragmatics could be said to be the functioning of language ...' (1988; emphasis in original).
Note also what Ostman says in the next paragraph: 'Admittedly, that latter [unit] is to be seen as a process rather than as an "object," but it is doubtful in what sense any units of analysis for semantics are that much more object-like.'
3.3 The Functional Approach
The notion of language as a functional whole is (like other good ideas) by no means new. As early as 1934, the German psychologist Karl Btihler elaborated his famous functional triangle of Ausdruck, Darstellung, and Appell as characteristic of language; and in the 1960s, Roman Jakobson elaborated on this Buhlerian model by adding three more functions: code, channel, and poetic quality.
What these models of human language intend to impart is a feeling of the importance of the human user in the communicative process. Messages are not just 'signals,' relayed through abstract channels: the human expression functions as a means of social
togetherness and of solidarity with, and appeal to, other users.
The result of adopting this way of looking at linguis- tic phenomena is vividly demonstrated by the fact that the different agendas which had been drawn up by the componentialists and the perspectivists respectively can be consolidated. Whereas representatives of the former line of thought are mainly interested in phenomena such as presuppositions, implicatures, deixis, and so on, a typical 'perspectivist' will deal with concepts such as 'negotiability, adaptability, and variability, motivations, effects, etc.' (Ostman 1988:29). In a functional synthesis, all this can be brought together: the most important criterion for language as it is used is whether it fulfills its functions of communication and interaction, not what it sounds like, or what kind of techniques it uses for getting its message across.
Neither can this be said to be a new idea. Austin and the early speech act theoreticians (such as Searle) realized that in speech acting, as in so many other ways of word(l)y behavior, 'what you get is what you expect.' Asking a passer-by what time it is, a person may usea question ofthe type Canyou tellme...?or even Do you have...? (viz., the time). The questioner would certainly be greatly taken aback by an answer in the affirmative (Yes), without any further information being offered. The reason for this astonishment is that such a 'question' is really more of a 'request' than a question (cf. the 'polite imperative' of the typeplease tell ... or please give ...—expressions which, inci- dentally, are rarely found in situations like the above). Other examples, quoted endlessly in the linguistic literature, include such cases as: requesting that a win- dow be closed by remarking on the temperature in the room (type: It's cold in here, isn't it?); requesting a passing of the salt by inquiring about one's neighbor's ability to do so (type: Can you pass me the salt?); and so on.
Linguistic functions of use are best studied in situ- ations where people interact normally, using language face to face. Consequently, such situations are con- sidered as the prime sources of information when it comes to studying this functional aspect of language: among these, everyday conversation among people takes a first seat.
There are basically two ways of going about study- ing conversation and other basic linguistic interaction. One way is simply to study what is going on, trying to describe it as exactly as possible, and figuring out what the options are for participants to join in at any given point, and what their choices are of expressing themselves to their own and others' satisfaction. This line of approach is followed by the so-called con- versational analysts.
Another, more theoretical approach tries to go 'behind conversation,' as it were, establishing the minimal conditions for successful interaction both on
Pragmatics
427