This volume contains the contributions presented at the first plenary conference on Integrational Linguistics held in London in 2000, an event at which I participated. The basic tenets of Integrational Linguistics were developed by Roy Harris at Oxford in the 1980s and 1990s; while many aspects of this approach remain a matter of debate, its assumptions and practices have become sufficiently firm to be the subject of a couple of introductory texts (Harris 1998, Toolan 1996). Harris's agenda has been “to challenge the monumental complacency of mainstream linguistics” (p. 3) by pointing out that the discipline is no more than an elaborate edifice built on a myth. As Harris observes, this message is unlikely to be widely welcomed:

The reluctance to embrace integrational criticism is reinforced by the fact that it challenges not only established views on language and linguistics but also common-sense views in numerous domains, such as art, law, and mathematics. This challenge is elaborated in a number of contributions to this volume.

As stated in the preface, “Integrationism … has its own distinctive contribution to make to the contemporary current of thinking that is often called ‘postmodern’” (vii). Its postmodern nature is manifested in the following widely debated (e.g. in Barr 1995:3ff) characteristics:

• its critique of common sense knowledge;
• its insistence that our understanding of the world is contingent on history and culture;
• its emphasis on the social processes that bring into being and sustain knowledge; and
• the integration of knowledge and social action.

The common-sense knowledge about language that is challenged by integrational linguistics is what one finds in most introductory textbooks on linguistics, including the following characteristics:

• surrogationalism, or treating language as a system of signs that stand for or are surrogates of a set of things in the world or a set of ideas;
• instrumentalism, or regarding linguistic signs as tools; and
• contractualism, or emphasizing the contractual or conventional nature of the communicative signs (for details, see Davis's contribution, 142ff).

The communication model underlying this myth, the conduit model, in itself has a mythological character. It implies that we transfer a copy of our private message to our listeners (“telementation”), and that this transfer is enabled by a fixed code shared by the interlocutors. I note that this communication myth has been the target of extensive criticism in communication studies ever since Reddy (1979) castigated it for its pernicious nature. The continued use of the conduit metaphor by the vast majority of linguists would indeed seem a great error, though, as Reddy has illustrated, it represents common sense. The conduit metaphor, of course, is one of the key metaphors by which not just linguists but most educated members of Western societies live.

Harris, in his introduction, summarizes the arguments against the language myth, leaving out details discussed in earlier publications, in particular Harris 1981. Central among these criticisms is the contention that our ideas about language are not natural but historically contingent (the confusion of history and nature, or transformation of history into nature, has been identified by many scholars, beginning with Giovanni Vico in the Renaissance era, as the essence of myth); thus, Harris writes, the term “natural language” as the subject of linguistic inquiry seems to be a serious instance of mislabeling. The historical roots of the language myth date back to Classical Greek philosophy, as seen in Aristotle's model of sense perception (11ff.), and the myth has been reinforced by subsequent changes in technology, such as the invention of the telegraph (in Shannon & Weaver's model of communication) and more recently the biocomputer. That processing of information by human brains cannot be characterized in terms of units and rules for the combination of units is well established in neuroscience (Shanks 1993), but this has not prevented linguists from continuing to employ these notions as essential to their discipline.

Implicit in the criticism of mainstream or “segregationalist” linguistics is its status as a science. Crowley 1990, writing in an earlier collection of integrationalist papers, has called a science of language “that obscure object of desire.” A science needs to meet the minimal requirement of having reliable criteria for demonstrating that two things are the same or different, and criteria for saying whether one is dealing with one or two objects. It is noted that Harris's first monograph in the integrationist mode was a critique of the notion of synonymy (Harris 1973), a privileged sense relation central to many theoretical arguments about language – indeed, to rather futile arguments, such as the protracted debate between Chomsky and Lakoff (analyzed in Botha 1970) as to whether Seymour sliced the salami with a knife and Seymour used a knife to slice the salami do indeed have the same meaning. The failure to have non-arbitrary criteria for sameness has serious implications for type-token theory and for claims about allophones of the “same” phoneme or allomorphs of the “same” morph, as well as, at the macro level, the debate about language and dialects, as anyone seriously examining standard language inventories such as the Ethnologue will find.

The social processes that brought the Western language myth into being and sustain it account for a vast array of cultural practices, including, as Harris admits, “values and practices that Western culture would be poorer without” (1). According to Harris, “In its modern form it is a cultural product of post-Renaissance Europe. It reflects the political psychology of nationalism, and an educational system devoted to standardizing the linguistic behaviour of pupils” (1981:9).

One might want to argue that the contributions of the French Enlightenment scholars in the entries concerned with language in Diderot's Encyclopédie (1751–1772) helped to elevate the myth to the status of scholarship. There are, however, many other, earlier factors, in particular “scriptism,” a fixation on written texts. A particularly graphic illustration of the fixation of modern linguists on written sentences, despite their insistence on the primacy of speaking, is seen in the examples used in linguistic textbooks. (Pop 1971 managed to write a comic play made up entirely from sample sentences culled from linguistics textbooks.) The dominance of the written word is also evident in the etymology of the term grammar (from the past participle of the Greek verb graphein ‘to write’). Text-based religion and popular understandings of communication technology have brought into being a text-oriented culture. Its impact on linguistic thinking is manifested in the separation of text from reader and context (segregationalism), and the reification of language as the code needed to produce texts.

All of these social and historical factors underpin the ideology linguists subscribe to, which makes it hard to conceptualize linguistics otherwise. Linguists need to understand how much of their subject is a product of such socio-historical forces, and to what extent the validity of their argumentation is dependent on unstated assumptions.

As postmodernists, integrationalists can hardly be expected to subscribe to a concept of absolute truth, and the justification of their emphasis needs to be assessed against (i) the success achieved by their demythologizing of linguistics, and (ii) their ability to integrate knowledge and social actions. Unexpectedly, several of the contributors address topics of social importance: Davis's comments on the teaching of Standard English in British schools (a critique of John Honey), and Toolan's discussion of its role in law and justice. Hutton highlights the relationship between the language myth and racist ideologies. Interestingly, Mufwene 2001, from a very different perspective, has castigated the racism implicit in many treatments of varieties of English spoken by non-British people.

At this point, the enterprise of demythologizing language – promoting social action though applied integrational linguistics (if indeed one can separate theory from applications) – remains underdeveloped. One reason for this may be its postmodern orientation and rejection of absolute values. I remain unsure how integrationalism would see linguistic human rights, language maintenance and revival, or the movement for biocultural diversity; the absence of discussion of such topical issues may be one reason why it has not received a wider audience. Taylor, in his contribution, raises an important issue when he asks how integrationists should deal with folk psychology and folk linguistics. Taylor's questions could be extended to how one should deal with names, folk beliefs, and metalinguistic systems from outside the European tradition. In spite of the positive comments on Whorf (33) by Harris, the volume remains Anglo- or at least Eurocentric. This seems to reflect the contributors' own backgrounds rather than deliberate decisions or inherent limitations. Indeed, as Weigand remarks (80), by looking at language as not separate from the world or from human beings, and by emphasizing the indeterminacy of meaning, integrationalism has humanized the subject of linguistics.

Integrational linguistics has been around for more than 20 years now. While the first generation of integrationalists were all closely associated with the University of Oxford, it has begun to attract the attention of others, particularly those concerned with the issues of communication rather than with core linguistics. It is good to see that Carr, in his contribution on phonology, is prepared to enter a dialogue though disagreeing with many aspects of integrationalism.

Linguistics somehow has not experienced the crisis that some neighboring disciplines have as its leading practitioners have adopted the Galilean style, characterized by Chomsky thus:

The message of this collection is that one needs not only to problematize the empirical inadequacies or methods of the discipline of linguistics, but also to examine the assumptions that underpin it.