Pragmatics, the state of the art. A talk with Jef Verschueren

Links & Letters 3, 199G 127-140 Pragmatics, the state of the art. A talk with Jef Verschueren Melissa G. Moyer Universitat Autbnoma de Barcelona. [email protected] Antwerp, February 1995 Q How wouldyou dejne thejeld of Pragmatics today?* JV Pragmatics has to be defined in relation to al1 kinds of other enterprises within the field of Linguistics. Traditionally, Pragmatics has been conceived of as an additional component of a theory of Linguistics, on a par with disciplines such as Phonetics, Phonology, Morphology, and Syntax. My position with respect to the definition of Pragmatics is that you do not get very far if you try to define it as an additionai component to a theory of language. The reason is that for Pragmatics you cannot identi+ a specific unit of anaiysis. This is something that you can do for other domains in Linguistics. Phonology, for example, deals with speech sounds, Morphology with the structure of words, and Syntax with the structure of sentences, and even for Semantics there are a couple of typicai units of analysis. For instance, Lexical Semantics is involved with words, and most other kinds of Semantics deai with structures cailed propositions. Nothing of that sort can be identified with Pragmatics. It is more helpful if we adopt a broader approach whereby Pragmatics is defined as the cognitive, social and cultural sudy of language and communication.' Q But stillyou maintain that Pragmatics sbould remain within the domain of Linguistics. In the ((Handbook of Pragmatics)) you refer to thejeld as ((Linguistic Pragmatics)). What is actual4 intended by the notion ((Linguistic Pragmati~s»?~ (*) This intewiew was carried out at the IPrA Research Center at the University of Antwerp in February 1995. The role of IPrA and the efforts of Jef Verschueren and his associates at the University of Antwerp have been of primary irnportance in shaping and defending the multidisciplinarity of the pragrnatic enterprise. 1. Sorne researchers adopt a more limited view of Pragmatics deterrnined by specific research questions. Defining the dornain of Pragmatics is not a straightfonvard task. S. Levinson (1983) in his classic textbook Prapatics dedicates thirty pages to discussing the rnany pragmatic perspectives from which language use can be studied. For example, Pragmatics within relevance theory is viewed as a psychological concern involving inferential computations performed over mental representations which are constrained by the cognitive principle of relevance. 2. Jef Verschueren, Jan-Ola Ostman, and Jan Blomrnaert. 1995. Handbook of Pragrnatics. Manual. Amsterdam: John Benjamins. 128 Links & Letters 3. 1996 Melissa G. Moyer JV It is called Linguistic Prapatics because it deals primarily with language use that involves certain cognitive processes taking place in a social world with a variety of cultural constraints. The idea is, instead of being a component for a theory of language, it would be a perspective that is being taken on language and the object of investigation of this perspective can be linguistic phenomena at any level of structure. For instance, you can approach phonological phenomena from a pragmatic point of view, which is something that has been done traditionally in studies of intonation. Most studies in intonation use al1 kinds of concepts that really fit into a pragmatic perspective on language. But that perspective has to be defined somehow. You can study linguistic phenomena at any level of structure from a pragmatic point of view, and 1 think it is only if you try to define the field of Pragmatics in terms of a perspective that you can develop some sort of coherent theory or frame of reference. Q I would like to take up the point you made about Prapatics as a kind of perspective, and ask about the kinds ofgoals one is trying to achieve by using a pragmatic perspective. JV The overridiq goal of Pragmatics is to gain an understanding of what you could call the meaningjul functioning of language (i.e., how cognition, society, and culture contribute to the way human beings use language in their dady lives). That already implies that the perspective 1 am talking about is a functional one. But you have to add a footnote to that becausefunctionalism is something that can have very different meanings. For instance, in the social sciences functionalism is very often associated with Parsons' sociology, which is relatively me~hanistic.~ It is in opposition to what is known as an interpretive approach in sociology. It is obvious that when 1 talk about a functional perspective, 1 am dealing with the meaningjul functioning of language, it is much more related to the interpretive kind of social science than to the work that is referred to as functionalism in the social sciences. Q Then why not call Prapatics the (cinterpretive approach)) to hnguage? JV There is a simple reason why. What we want to deal with is language in its full complexity which means that we are not only going to deal with the interpretation side that is involved but there is also a production side that we must consider. Dealing with the meanin&lfunctioningof language is, by definition, an interdisciplinary enterprise where you have to take into account elements of cognition, society and culture. Q What sorts of considerations does one need to take into account when undertaking research in Pragmatics fiom the point of view of speech production? JV: The question is a little misleading because in my view there is no such thing as ((doing Pragmatics from the point of view of speech production.~ Nor 3. Talcott Parsons, a US sociologist, presented his theory of functionalism in the early fifties in his book Thesocialsystem. 1951. New York: The Free Press. In his view, functions serve as links between relatively stable structural categories. Pragmatics, the state of the art. A talk with TefVerschueren Links & Letters 3. 1996 129 can one simply 'do Pragmatics from the point of view of interpretation.' The essence of what 1 am trying to say is that these are not separable. Therefore, the types of considerations remain the same. To take a cognitive example, the processing involved in the production of an utterance includes what is commonly referred to as audience design. This means that an utterance is partly shaped on the basis of assumptions about the hearer. But these assumptions are themselves inevitably based on earlier interpretation processes, either within the context of the same exchange or on other occasions (even if there has been no earlier interaction with the same interlocutors). Production vs. interpretation is a useful dichotomy only if one realizes how fundamentally they are intertwined. Going to the social side of the matter, for instance, social relationships do not only influence what goes on in the interaction, they are also shaped by it. The same principle holds, by the way, for the distinction between cognition, society, and culture: they are also inseparable aspects of the same complex human reality. Q How is itpossible to evaluate the progress in Pgmatics fwhat is being donepertains to so may domains? JV There is an important problem of fragmentation there which does not necessarily have to be bad. Ultimately, the worst thing that we could do is to try to impose a very strict coherent theory of language. If what we are interested in is language use in its full human complexity then it is pretty obvious that this is such a complex kind of phenomenon that as soon as we try to impose a really rigid kind of interpretation, or a really rigid theory then we are bound to miss ail kinds of interesting things. Al1 you can try to do is get some sort of coherent frame of reference with which you can start to interpret what people are doing in various domains that belong or are relevant to the field of Pragmatics and which include various points of view as well as different methodologies and concepts. This is a kind of richness we should try to exploit. The fragmentation in itself is bad only if it results in people not being able to understand what others are doing. Q Do you think it is aproblem for thejeld ofPrapatics that researchers do not have a set of recognized terms that everyone uses? JV This is an issue, but 1 think it is part of the very nature of what it means to be doing science. Science is a very human kind of enterprise, which develops in the same way as other things in human life. People rarely use their long term memories. Much of what we are doing now in Pragmatics is really reinventing earlier work. For example, Sapir was the kind of person who was already saying that there is no way to redy understand what happens with language unless we abandon strict disciplinary thinking.* As long as linguists just want 4. Edward Sapir (1984-1939) was a US anthropological linguist. His writings (except for Language) are collected in a volume edited by D. Mandelbaum. 1949. Selected writings of Edward Sapir. Berkeley: University of California Press. 130 Links & Lerters 3. 1996 Melissa G. Moyer to be linguists they will never understand what language is about. That is a perspective that Sapir was already taking. These are the kinds of issues the implications of which people are beginning to rediscover at this moment. There is no linear progression in scientific affairs. There never has been In more recent history you can see that with reference to enterprises such as Speech Act the~r~.~ Already in the seventies the theory was criticized for trying to isolate individual speech acts that were identified with various functions. These criticisms were justified. But the result has been, for instance, that al1 notions belonging to Speech Act theory are virtually taboo in new traditions within the pragmatic approach to language, and as a result of that many things have been lost. Researchers turned to conversation as a topic in its own right. But when you analyze a conversation you cannot avoid certain types of concepts that Speech Act theory was already dealing with and had sensible things to say about. So we should consider preserving more of what people have been doing. There are various ways of salvaging what Speech Act theory had to offer by redefining both goals and methodologies. The goal of Speech Act theory in the orthodox Searlean version was to give a universally valid description of forms of human behavior. This has been heavily criticized from the point of view of linguistic anthropologists. There are al1 those societies where we know that the notion of commissives is hardly rele~ant.~There are many data which point in the same direction and show that it is presumptuous to think that we can really develop a theory of universal validity from our own point of view. That is a kind of goal you can simply abandon. Along the sarne line of thinking was the idea that speech acts did not have anything to do with the words that we use to describe them. These were seen as accidental language-specific things, that were not very relevant in their own right. That attitude should also be questioned. But this takes us to a broader view of what it means to behave socially. There is no way of understandiq forms of social behavior (and that is something we should have learned from Winch), without also understanding the kinds of concepts in terms of which the people engaged in a form of behavior interpret that behavior them~elves.~ Social behavior is always interpreted behavior, so you have to understand the concepts that serve the purpose of this interpretation, including the lexical reflections in these concepts. Looking at verbal behavior from that point of view it is very relevant to start studying speech act verbs. You have to reinterpret the goals, then you go back to what people in Speech Act theory were doing and you see that much of what has been done becomes relevant again. Searle's analysis of promises no doubt comes 5. Speech act theory seeks to account for the role of utterances benveen speaker and hearer. It is a theory which is concerned with communicating activity defined in reference to the speaker's intentions and the effects shelhe achieves on the listener. 6. A commissive in speech act theory is a type of utterance where the speaker comrnits himselflherself to a future course of action (e.g., 1 promise). 7. Peter Winch. 1958. Tbe idea of a social science and its relation to philosopby. London: Routledge. Pragmatics, the state of the art. A talk with Jef Verschueren Link; & Letters 3, 1996 131 very close to the prototypical case of a concept lexicalized in English with the verb to promise. Thus, the analysis acquires a different kind of relevante than what it was intended as, but it is still valid. We should at least make an effort to try to preserve some of these contributions. Q Going back to this defZnition of Pragmatics as the study of the ((meaningfilfin~tionin~ of language)), ir itpossible to identzfi certain areas or domains that aregoing to be more reievant to this sdy of the meanin&lfinctioning of langudge? I JY In principle any form of knowledge that we can accumulate about language will be relevant for a theory of Pragmatics. If you are defining Pragmatics as the study of how language gets used, any knowledge about language is part of the knowledge you need. But to come to the point of your question, what you do need, first and foremost, is a clearer definition of what this functional perspective is like. In order to define it 1 think it is always very useful to start with seemingly simple questions. Think for a moment about what it means for people to be using language. What are you doing when you are using language? This is a very general kind of question and it is basic to the whole enterprise. 1 think that in order to use such a general question as a starting point we also have to abandon shame in connection with giving trivial answers. A trivial answer to that general question is that when people use language what they are doing is making al1 kinds of linguistic choices for al1 kinds of reasons and at al1 levels of linguistic structure. The basic process is a process of choice making. Of course, this is already a little bit misleading. It is misleading, first of all, because it seems to imply that the only thing we are looking at is what the producer of the utterance is doing, which is wrong. Interpreting is also making choices. That is a very important thing we have to realize. The second misleading aspect about making choices is associated with choices as conscious actions. That is an idea that we have to abandon too. With these two caveats we can use it as a starting point for thinking about the use of language. Q Couldyou peczfi firtber whatyou mean by ((making choices»? TV: This entire idea of the making of choices is something that is very strong in Firthian linguistics.8 This gives us an immediate link with another kind of tradition that is functionally oriented. It is not an isolated kind of idea. The problem with the making of choices is understanding what this is supposed to mean. We will need to make reference to a few concepts for that. If you talk about the making of choices this has to be choices made from a variable range of possibilities. That is where the entire notion of variability becomes central 8. J. R. Firth was a British linguist (1890-1960) who ernphasized the importance of social context for a theory of rneaning. Firth is also responsible for the view that patterns of language use cannot be accounted for in terms of a single systern of analytical principles and categories. The ideas of Firth have been followed up by another British linguist, M.A.K. Halliday. 132 Links & Letters 3, 1996 Melissa G. Moyer to Pragmatics. We really have to realize that there is an enormous amount of variability. There are many different kinds of choices that are being made (e.g., e choosing a language, a style, syntactic structures, lexical meanings, etc.). Choices are not being made once and for all, they are negotiable. Negotiability becomes a second very important notion for understanding this making of choices. This notion implies that choices are not made mechanically or according to strict rules or fixed form-function relati~nshi~s, but on the basis of highly flexible principles and strategies. Once you have made a linguistic choice, that is not the end of it. Not even on the interpreting side, because any interpretation can be used to tenegotiate earlier aspects of an interaction (including earlier interpretations). The question is, then, how it is still possible for human beings to communicate. How can they still communicate if al1 &ey are doing is making choices from an enormously variable range of possibilities with values that are even negotiable rather than fixed? There is a third concept that does not provide an explanation for that but which at least provides a name for the phenomenon; it is the notion of adaptability. This is the property of language which enables human beings to make negotiable choices from the variable range of possibilities in such a way as to satisfy basic human communicative needs. You have to assume that language is an extremely adaptable kind of instrument. It fulfills certain communicative needs though it is an instrument that itself is enormously variable although the variable elements do not have fixed values. You are really dealing with a truly adaptable kind of phenomena which again has two sides to it. One might assume that al1 that is happening is simply that a speaker assesses a situation, makes the linguistic choices that are best adapted to the situation and then that is it. Of course, it is not that simple. It is a phenomenon that you have to really describe in terms of interadaptability. The choices you make also define the situation. A very simple example is for instance the choice of forms of address. If you statt to address someone you do not know very well with an informal form of address, that defines a situation. If you realize later that this is a person who you could never be on familiar terms with, you can not go back to formal forms of address without overt hostility. Working then from this kind of general notion you can develop some sort of paradigm in terms of which you can approach linguistic phenomena from a pragmatic perspective. Q What are the more specz$c kinds of questions we have to deal with ifwe want to describe some sort of linguistic phenomenon fiom a pragmatic point of view? JV: There are four very important questions that have to be asked. Remember that we are dealing with interadaptability between language and al1 kinds of other things. These ((other things* can be anything that is an ingredient of the speech event, ranging from very wide socio-cultural circumstances to individual intentions, beliefs, desires, and so forth, al1 kinds of ingredients of the speech event that you can label contextual objects of adaptability. One thing that you have to be clear about in dealing with any linguistic pheno- Pragmatics, the state of the art. A talk with Jef Verschueren Links & Letters 3, 1996 133 mena from a pragmatic point of view is which are the objects of adaptability or objects of interadaptability. This is one question that you have to be able to answer ultimately. A second kind of question you have to be able to answer is at what structural Iayer phenomena have to be situated, 1 mean at what layer of linguistic structure. You are rarely dealing with everything at the same time. You can try but it is not going to work very well. If you have a specific research question, you have to identifj the level at which the phenomena you are studying are situated, even though you cannot just isolate these phenomena from other levels of structure. If you can identifj the contextual objects of adaptability and the structural layers of adaptability, then what you have achieved is a description of the locus of the phenomena, namely, the extra-linguistic and linguistic coordinates for the phenomena being investigated. The main task of course is to deal with the dynamics of the processes involved. The notion of negotiability implies that there is an enormous amount of dynamics involved: the dynamics of meaning construction, and the negotiation of meaning. Just take a seemingly simple sentence such as Iam nota racist. It is obvious that an utterance of this kind is embedded in a wider context which might lead to a conclusion which the speaker feels compelled to prevent or deny. But since the context might lead to such a conclusion, the denial does not simply use the pre-existent concept racist but it constitutes one step in the negotiation of the meaning of the word. This is what was meant in the beginning about the meaninaful functioning of language. Ultimately, meaning is what language use is al1 about. Pragmatics tries to understand the different ways meaning is created. There is a fourth question that you can not avoid. You are dealing with things that go on in the mind. Some of these processes take place at a high level of consciousness but many of them mainly take place at low levels of awareness. This is extremely relevant because it allows you to situate various types of explicit and implicit meaning. Also implicit meaning can in some cases be very consciously constructed but you can get forms of implicit meaning that ultimately contradict what people would be willing to say that they intended to communicate explicitly. Thus, the explicit statement Im not a racist may be completely incompatible with the implications of the surrounding d' ~scourse. 1 think that if you can answer these four types of questions then what you have is a decent description of the full functioning of a certain aspect of language. But it is quite a task of course. Q: Woulúyou give an example of a piece of research which addresses thesefour questions? JV It is almost inevitable that pragmatic research consider these four questions. But it is necessary to be as explicit about them as possible to avoid many implicit assumptions that may obscure the relevance of the research results. 1 will start out with a very simple kind of example. The context is a flight from 134 Links & Letters 3, 1996 Melissa G. Moyer Brussels to Frankfurt on an occasion where there was fog, resulting in a delay. After a while the plane can take off and the pilot explains to the passengers what is happening. Then he says: Once we are in the air we willfly a fat mpossible. The passengers start giggling and you start asking yourself why. Obviously, something quite dynarnic must have been going on there because this pilot is very serious. He makes this announcement with the most serious voice that you can imagine. His words were just intended to reassure the passengers that every effort would be made to avoid any further delay. This was meant informatively. There is no mistake about the pilot's intentions, which did not include a humorous effect. So, there must be some surplus meaning that comes up. That is an extremely dynamic process that you have to address somehow. You have to describe certain elements in the context, the contextual objects of adaptability. One of the elements of the context you have to keep in mind and that is very crucial is the fact that the context of the pilot is one that is fundarnentally different from the context of the passengers. In the context of the pilot in the cockpit, the pilot is dealing with speed in the air on a day-to-day basis. For the passengers there is no experiential basis for thinking about speed in the air except maybe on some very long flights. There are two different contexts involved. This may already provide one element of the explanation for how the extra meaning gets generated. But that is not sufficient because you simply have to use some standard linguistic substitution test to show that the pilot could have communicated this intended meaning with sentences that would not have resulted in the same type of reaction. For instance, the pilot could have said something like: Once we are in the air we will ty to make upfor some of the delay. If he had said something like that it probably would have been received as a simple piece of information. That means that there was already something in the form of the expression in itself that must be responsible for the process that you observe. Then you can start thinking about what it is in the utterance that carries the potential of producing this special effect. Of course, you would need a lot of comparative material to have any certainty about this. One of the elements that definitely is involved is the fact that normally a verb such as tofly, in its nonmetaphorical sense, is a verb that has non-human agents. And if you have a human agent, which is possible, you expect some focus on the airplane. If you start to underscore human agency as such by describing a manner of flying (such as, as fmt aspossible) then you get something that is slightly deviant from the expected pattern of usage. Breaking such expectations evokes al1 kinds of images in people's minds like, for instance, the image of a German driving a BMW or a Mercedes on a highway without speed restrictions which is an image that does not fit an airline pilot. Then you can start to ask the question of how much of this takes place on a conscious level. The pilot quite consciously chose his words, but maybe he is too clumsy at using the right idiomatic expressions to avoid this type of generation of additional meaning. His seriousness suggests that the choices were not made to achieve one of the effects produced by his words. Pragrnatics, the state of the art. A talk with Jef Verschueren Links & Letters 3, 1996 135 Al1 these elements can be shown to be involved in a small simple example of this kind. They get al1 the more complicated in interaction with real exchanges, such as regular conversation. Q How has Pragmatics evolved? What contributions have been made in recent years in thejeld of Pragmatics? How has Pragmatics advanced, let? say, since Speech Act Theory? JV: The more general form of the questions you are asking is difficult to answer because you do not get something you can really describe as some sort of major new development. There are al1 sorts of different strands that intersect in various ways and sometimes just run parallel to each other. Focusing on the more specific aspects of the question (what kinds of things in recent years are of particular importance to reach an understanding of the Pragmatics of language), there are two things that are of crucial importance. First of all, there is a development in connection to the notion of context. Remembering what 1 said about identiGing contextual objects of adaptability as one of the necessary questions for pragmatic investigations you will see why 1 regard context as very crucial. The most important recent achievement is the realization that context in itself is not just something that is out there in the world that is separate from language and that simply has to be referred to and used as a proper motivation for making appropriate linguistic choices. What is important is the realization that a context itself is something that is interactively constructed. This realization makes it easier to use the notion of context to study forms of language use. If you do not realize that the process really involves the interactive construction of context, then what you are dealing with is this completely indiscriminate world of anything that can be relevant to language. That is one of the reasons why many linguists have been very hesitant to rely on this notion. It becomes a reality you cannot manage. What the newer developments with respect to the notion of context do is to allow you to trace in a discourse or in a convei.sation the way in which aspects of this wide context are actually used for purposes of making production choices or interpretation choices. You can trace in language those elements that are actually getting used, and there is no need to talk about elements that are not getting used. That makes it much easier then to be really empirical in Pragmatics. 1 want to link this to another notion where there are developments that 1 think are very important, in particular, a development in connection to the notion of meaning in context. If you think about Grice, and Gricean rules, what Grice did was something you could describe as removing meaning from language by saying that meaning is not just there in language f~rms.~ Meaning is something to be situated. Though he accepted the existence of conventional 9. H.P. Grice was a philosopher who developed the theory of conversational implicatures that refers to the implications that can be deduced from the form of an utterance on the basis of cooperative principies that account for the efficiency and acceptability of conversations.