Part I
The State of the Art
Part I of the Handbook delineates the sequence of schools which more or less followed each other in the twentieth and the beginning of the twenty-first centuries. The order of the chapters is determined by consideration of the scientific interests that have brought about changes in the objects-of-study. All these schools can be seen as attempts to grasp the object ālanguageā. To some extent they run into each other and cannot always be strictly kept apart. These attempts did achieve partial results and paved the way, step by step, towards greater insight into the complexity of human life.
The Handbook represents a collective volume which gives the floor to different voices. These voices are on the one hand representative of their field but, of course, they express individual views. Differences of opinion remain which reflect science as a field of argumentation and discussion. Nonetheless, the chapters clearly mark paradigm changes in the history of modern linguistics.
We start with an overview by Jens Allwood of the first period from structuralism to pragmatics, from language as a system of signs to the concept of language-in-use. Having studied and analysed language as an abstract system over decades, the interest of the researchers in artificially constructing language diminished and turned instead to language as it is used, not an object on its own but an object dependent on the user. In use there are no signs which have meaning on their own. On the contrary, there are communicative means whose meanings differ dependent on the individual user and the specific environment they are used in. The change to the new object ālanguage useā constituted the pragmatic turning point. Changing the object requires a change in methodology, a challenge pragmatics is still struggling with.
This first period, based on a system of signs, mainly comprises the structuralistic and generative schools. Both focused on the āsentenceā, which was analysed by structuralistic approaches as a static system, and by generativists as a dynamic system of production. Language use, however, does not stop with the sentence nor is it based on rule-governed āsentencesā at all. Language use consists of āutterancesā. Looking beyond the sentence we were struck by the complexity that opened up. Beyond the sentence there was the world. We did not even know what ālanguage useā meant and considered it an illuminating insight to discover that ālanguage useā meant ānatural language useā and ānatural language useā meant āauthentic language useā. However, how was the new concept of language as āauthentic language useā to be analysed?
In principle, two possibilities opened up, first, searching āin the undergroundā, and second, scrutinizing empirical details at surface level. Searching in the underground meant focusing on the meaning of āsentences-in-useā, which was heavily influenced by language philosophy and Austinās book How to Do Things with Words. Austin and later Searle drew our attention to the crucial insight that we not only speak with words but act with words in utterances. Language-use means language-action. This decisive insight, however, meant clarifying the relationship between communicative means of use and their purposes in action. After a first wave of approaches based on philosophical action theory, interest diminished. Searleās speech act theory was considered to be too abstract and not directly applicable to authentic language use. However, instead of making efforts to modify and change it in order that it should become apposite to dialogic performance, researchers turned their interest upside down; they went from analysing meaning to analysing the means at the empirical surface level.
Pragmatics splits up into various types of approaches all trying to come to grips with the empirical details of ever-changing authentic language use. āDataā and ādiscourseā became the core terms of these attempts. Any empirical element was picked out and investigated meticulously as so-called ādataā, although half a century ago Chomsky (1959) provided a thorough critique of this type of approach in his review of Skinner. Empirical elements can count as ādataā only if they can be meaningfully related to the purpose of language use. But even genuine ādataā do not provide the key to the complex whole of language use. Another strong argument against devoting all our energies to any empirical detail whatsoever can be derived from the way we as competent users of language deal with them in performance by simply passing over them.
The concept of language changed due to the focus on ādataā of authentic language use. Language was equated with spoken language and ātextā as a sequence of sentences became ādiscourseā as a sequence of utterances. At the beginning of this process, ādiscourseā still meant ātextā or the sequence of sentences, for instance, with Zellig S. Harris (1952). As soon as the verbal level of text no longer sufficed for the claims of performance analyses, the term ādiscourseā developed into a concept which even included the cognitive world of the users. āDiscourse analysisā became a āpan-disciplineā (Van Dijk, 1985: 10) which claimed to encompass everything connected with language use and embraced various diverging schools under the premise āanything goesā. The criterion of consistency āof a homogeneous ādisciplineā with a unified theoryā was declared to be āpointlessā (Tannen, 1989: 7f.). The catchword of the āplurality of modelsā was welcomed by many scholars as it easily allowed unlimited access to a view of language which was declared to be mainstream linguistics. In effect the āplurality of modelsā only means eclecticism and arbitrarity as Frawley (1987: 363ff.) very early critically pointed out.
The integration of perceptual means, such as gestures or body movements, was crucial to the emergence of a new type of approach, the multimodal analysis. Lorenza Mondada describes this development, starting from conversational analysis which was primarily introduced in its exclusively formal version by Sacks, Schegloff and Jefferson (1978). She shows that the change to the multimodal version is based on the insight that the empirical level not only consists of verbal means but includes multimodal means of speaking and perceiving in interaction. This insight was an important step forward to a view of language use based on the integration of different human abilities.
Among the schools focusing on the empirical component of language use there is also the school of corpus linguistics whichāas Marina Bondi describesāprovided a highly valuable tool for describing the multiple ways of use of verbal expressions. Corpus linguistics does not aim to demonstrate what lies below the verbal surface but describes it and justifies verbal evidence by its frequency. This step reveals important insights into the system of language as a network of language-specific phrases or collocations. Consequently, lexical semantics can no longer be based on structuralistic terms such as polysemy and disambiguation, or on an artificial system of single words but needs to take account of the way words are used in phrases. Corpora of written and oral language provide a vast amount of material of authentic language use in different languages far beyond the reach of any artificial language system based on the competence of the ideal speaker.
Discourse analysis is the topic of James P. Geeās chapter. This is only a brief outline of the broad field of different ādiscourse schoolsā and mainly introduces Geeās extended version of a theory of discourse which no longer puts the emphasis exclusively on the empirical surface but includes categories of action as well as various types of dialogic means.
Besides discourse-based models there is another methodology within pragmatics which developed into a school of its own, the neo-Gricean approach, which believes in the addition of orthodox assumptions of truth-conditional semantics and individual inferences (see Chapter 5). It is not surprising that it got stuck in a jungle of terms and levels (e.g., Levinson, 2000).
Attempts to settle the crucial problem in pragmatics of how to address language use in performance are manifold. Among them, there is a technique which simply uses one term for all the different cases, the term āpragmemeā (Mey, 2001). This reminds us of the techniques of structuralism. However, whereas structuralist terms, for instance, the phoneme or morpheme, were coined as the result of clear operations defined by structuralist methodology, the term āpragmemeā is applied as a circular classification to very different pragmatic issues. There are no internal criteria which define the term, and therefore it remains without any explanatory power. Such an attempt is doomed to fail from the very outset. The change in the object, from language competence to language use in performance, requires a change in methodology.
The question is where to begin, with the object or with methodology. Whereas modern linguistics started with the dogma of language as a sign system, that is, with an abstract model of language, pragmatics and its natural object of language use resisted being dominated by pre-given terms. We should be careful not to distort the natural object by our methodology, as Martinet (1975) told us. A complex natural object requires us to start from the complex whole and to recognize that there is no language as such but human beings who use language in dialogic interaction.
Even if pragmatics continues to focus on one utterance and to invent new terms, the insight that language use does not stop with the utterance can no longer be ignored. We approach the next turning point from pragmatics to dialogue, which Istvan Kecskes describes in his chapter. Kecskes can be considered as a forerunner among the pragmatists who emphasized that language use does not mean making utterances but acting and reacting in dialogue. He indicated the direction towards dialogue and towards a concept of language as action.
Scholars of psycholinguistics opened up the view from the speaker who means something to the hearer who tries to understand or to interpret the speakerās utterance. Meaning and understanding are at centre stage in the chapter by Susan E. Brennan and Joy E. Hanna. The concepts of meaning and understanding still refer to the speakerās utterances and cut off the natural process of interaction, which is a process of coming to an understanding, in which the hearer not only listens but changes to an interlocutor who reacts by means of their own action, verbal, perceptual by a gesture or in the mind. Per Linell points to the role of shared understanding and intersubjectivity and argues that understanding will mostly be only partially achieved.
The direction towards dialogic interaction became more and more manifest. The view of ālanguage as dialogueā spread gradually and even extended to literary texts (Chapter 8). Whereas literary scholars often view literature as monologic texts, Roger D. Sell very early on considered literary texts as communication between the author and their audience. In his groundbreaking Communicational Criticism: Studies in Literature as Dialogue (2011) he assigned literature a specific dialogic quality of addressivity and thus strengthened the view that language in general is used dialogically. We need not develop different theories but can describe any use of language in a unified theory of ālanguage as dialogueā which, of course, manifests itself differently in different genres.
The view of ālanguage as dialogueā also spread to computer linguistics as is described in Chapter 9. David Traum gives an overview of how computer linguists tried to come to grips with the considerable expansion of linguistic theories from an artificial system of signs to dialogic action and multimodality. It goes without saying that the problems are enormous. Encouraging progress has been made even if many issues, including a coherent computer linguistic model of the whole, have still to be settled.
Taking the direction from pragmatics to dialogue finally led to the next turning point, which was achieved by recognizing that language use not only means speaking but acting. The challenge was to change Searleās monologic speech act theory (1969) into an action theory capable of dealing with dialogic sequences. Action in general is defined by the correlation of purposes and means, be it verbal action or perceptual or cognitive actionājust as, for instance, if we want to fell a tree, we have to look for an appropriate tool. The first question to be raised refers to the purpose of dialogic language use. The hearer changes to an interlocutor and the purpose of their dialogue is not only the hearerās understanding but their mutual attempt to come...