Roy Harris (born 1931) is Emeritus Professor of General Linguistics in the University of Oxford and the founder of a distinct approach to language and communication termed ‘integrational linguistics’ or ‘integrationism’. Prior to taking up the Chair in General Linguistics, Harris was Professor of Romance Languages in Oxford. His inaugural lecture for the Chair in General Linguistics was entitled ‘On the Possibility of Linguistic Change’ (Oxford: Clarendon, 1977). Subsequently Harris held posts at the University of Hong Kong, Boston University and the École des Hautes Études in Paris. In 1981 he established the journal Language and Communication, which he edited with Talbot J. Taylor until 2010. Harris is a regular reviewer on matters linguistic for the Times Literary Supplement and the Times Higher Education Supplement. He is a founding member of the International Association for the Integrational Study of Language and Communication (IAISLC), which has been in existence since 1998. The Association holds regular conferences, and the latest ‘Linguistics in the 21st Century’, took place at the University of Hong Kong in May 2010.
Harris’ work has a number of strong themes and positions, and one can highlight the following topics and motifs in particular: the role of the theory of meaning in methodology of descriptive linguistics (Synonymy and Linguistic Analysis, Oxford: Blackwell, 1973); the history of, and essential continuities in, linguistic theorizing in the West, with special reference to what Harris terms the ‘language myth’ (The Language-Makers, 1980; The Language Myth, 1981; The Language Machine, 1987, all published London: Duckworth); Saussure and structural linguistics, with Harris in the role of translator and commentator (F. de Saussure: Course in General Linguistics, 1983, Reading Saussure, 1987, both London: Duckworth), and as a critic of the reception of Saussure’s work across a number of disciplines (Saussure and his Interpreters, Edinburgh, Edinburgh UP, 2001); Wittgenstein, especially the later Wittgenstein of the Philosophical Investigations (Language, Saussure and Wittgenstein, London: Routledge, 1988); the history and semiology of writing (beginning with The Origin of Writing, London, Duckworth, 1986); the integrational theory of the sign and the study of communication (Signs, Language and Communication, London, Routledge, 1996); semantics, lexicography and definition (Definition in Theory and Practice, London: Continuum, 2007, with C. Hutton); the role of language and linguistic theory in constituting so-called disciplinary formations and large ‘supercategories’ such as art, science, history (The Necessity of Artspeak, London: Continuum, 2003; The Linguistics of History, Edinburgh, Edinburgh UP, 2004); The Semantics of Science, London: Continuum, 2005); philosophy and theory of mind (The Language Connection, Bristol: Thoemmes, 1996; Mindboggling, Luton: The Pantaneto Press, 2008.); epistemology (After Epistemology, Gamlingay: Bright Pen, 2009).
The central target of Harris’ critical exegesis is what he terms the ‘language myth’. This is a formative notion not only in the study of language, but across a wide range of disciplines and social domains in the Western tradition. In brief, it is the notion that languages are autonomous and well-defined entities which provide stable systems of representation for members of speech communities. Words are understood as standing for objects outside the speakers in reality (a position that Harris calls ‘reocentric surrogationalism’) or for thoughts in the mind (‘psychocentric surrogationalism’), or some combination of the two. The myth rests on the idea of languages as fixed or determinate codes (the ‘fixed code fallacy’) which can be described and studied independently of any particular context, speaker or communicational practice. The fixed code underwrites or implements the transfer of thoughts from one mind to another, a view of communication Harris describes as ‘telementation’, and is held to provide a stable set of orientation points and determinate relationships which transcend the specific instances of communication and the activities of speakers and hearers for whom communication is thereby made possible.
Saussure’s Cours de Linguistique Générale (1916) has been at the centre of Harris’ critical engagement with the foundations of linguistics. For Harris, the Cours was a brilliant piece of academic theorizing, which envisaged the possibility of an autonomous science of language based on the langue/langage distinction and a ‘bi-planar’ theory of the linguistic sign. This act of disciplinary strategizing drew on the basic elements of the language myth, but transposed them into the tenets of an aspiring science, making Saussure ‘an interesting example of revolutionary conservatism’ (After Epistemology, p. 69). Saussure’s notion of langue served ultimately to remove the study of language into a relatively closed academic domain with its own complex terminology and often abstruse formalist pretensions. In returning to read and comment on the Cours, we can reconstruct and analyze the steps by which Saussure sought to isolate and define an object of study for linguistics, and also remind contemporary linguists that Saussure is quite explicit that the phenomenon of human language is not ontologically autonomous. It takes an act of stipulative definition to create a discrete object of study, and a new and powerful terminology.
Harris’ work is an extensive and long-considered response to Saussure and the discipline that he helped create, namely academic linguistics. Harris terms this discipline ‘orthodox’ or ‘segregational’ linguistics. Linguistics is ‘segregationalist’ in several senses: it assumes a constant transhistorical distinction between the categories of language and non-language; following from this, it demands an object of study abstracted from speakers, hearers, context and unfolding temporality; it relies on a presumption that language is best studied by being broken down into a hierarchy of levels and units of analysis which operate as abstract types through the instantiation of tokens: pragmatics, semantics, syntax, lexicology, morphology, phonology, phonetics. The resulting discipline is also effectively ‘segregated’ from the surrounding disciplines of literary theory, psychology, anthropology, cultural studies, sociology, physiology, and so on. Indeterminacy is at the heart of debates in economics, for example, yet there has been no serious debate about questions of indeterminacy within mainstream linguistics.
For Harris, the view of language articulated by modern linguistics rests on a profound misprision of the nature of human communication, and therefore of humanity in general. Linguistics and many other disciplines rely on the idea of languages as fixed codes that are held to make meanings available for us to use in particular contexts (the ‘language myth’). Integrationism denies that our communicational activity has any underlying guarantee or set of fixed reference points – we are always in the middle of a temporal and communicational stream, and the reference points we use are shifting along with us. We cannot get out of the stream of communicational life and observe it from a neutral vantage point. In this sense, we are all language-makers, sign-makers, meaning-makers, and as a matter of necessity, theory-makers in the domain of language. Communication is a creative process which involves an unbounded and unknowable number of factors and viewpoints, and through which individuals constantly adjust, assess and contextualize and recontextualize their experiences and practices in the light of the unfolding situation. In communicating, we integrate aspects of the situation we are in, including the on-going behavior of people present, with our past, present and anticipated experience in ways that are not knowable in advance, even by ourselves. Nothing is given in advance in communication; signs are created in the here-and-now; the relationship between words and ideas, and words and things, is not fixed: ‘meaning is always “now”’ (M. Toolan, Total Speech, Durham: Duke, 1996, p. 125).
For the integrationist, human communication is constrained or shaped by three factors: (i) biomechanical; (ii) macrosocial; (iii) circumstantial. Harris defines these as follows (Introduction to Integrational Linguistics, Oxford: Pergamon, 1998, p. 29):
Biomechanical factors relate to the physical and mental capacities of the human being. Macrosocial factors relate to practices established in the community or some group within the community. Circumstantial factors relate to the specifics of particular situations. Thus the fact that A and B communicate in speech only via sounds of a certain amplitude and frequency is a biomechanical factor, having to do with the physiological constitution of the human body. The fact that A and B cannot communicate in Swahili because B knows no Swahili (even though A does) is a macrosocial factor. The fact that A can speak to B even though separated by a distance of thousands of miles (because a telephone is available) is a circumstantial factor.
A further fundamental terminological and theoretical innovation by Harris involves what he calls the ‘principle of cotemporality’ (Introduction to Integrational Linguistics, Oxford: Pergamon, 1998, p. 81-82):
This principle is based on a simple lesson of linguistic experience: that what is said is immediately applicable to the current situation, unless there is reason to suppose otherwise. But this holds not only for what we say but for everything we do. In other words, in this respect there is a complete parity of status between linguistic acts and other acts. Linguistic acts do not have some special temporal status of their own, which somehow puts them outside the sequentiality of the rest of our existence. This might be thought to be an extremely banal observation, and in one sense it is. But it is perhaps worth calling a ‘principle’ when we realize how far-reaching its implications are for linguistic inquiry.
Integrationist semiology is based on two theoretical axioms: ‘(1) What constitutes a sign is not given independently of the situation in which it occurs or of its material manifestations in that situation. (2) The value of a sign (i.e. its signification) is a function of the integrational proficiency which its identification and interpretation presuppose’ (After Epistemology, p. 73). In this sense, ‘[e]very act of communication, no matter how banal, is seen as an act of semiological creation’ (After Epistemology, p. 80).
One way to understand the basic integrational position on signs and meaning making is through the assertion that first-order experience cannot be reduced to second-order categories. First-order refers to: experience, understood as here-and-now activity, on-going communicational activity, contextually meaningful behavior: it is situated, happening in real-time, and unfolds in unplanned ways over space and time. For Harris the face-to-face communicational event is probably the best example, but talking on the telephone, reading a novel, watching television, and countless other activities or events are examples of first-order experience. But one aspect of our communicational practice includes the ability to give accounts of our communicational behavior and experience in terms of second-order categories. Second order concepts are macrosocial abstractions (labels, categories) we employ in order to comment on, explain, or interpret experience. In the case of language, such concepts and categories are called ‘metalinguistic’. We use such categories all the time, and they are intertwined with, and necessary to, our first-order practices as language-users and makers. Ideas like ‘society’, ‘community’, ‘social order’ are second-order concepts, as are the names of languages (‘Chinese’, ‘French’, ‘Hindi’). We cannot directly experience or come in contact with ‘society’ or ‘the economy’, but we make sense of our experience, and we order our social world, by reference to such macrosocial or general concepts.
A central feature of human language-making and creativity therefore involves the invocation of fixed points of reference, even if, paradoxically, these are subject to the same integrational processes as all other activities. Stability of meanings across contexts and consistency of interpretation over time are goals of particular discourses, such as cultural criticism, theology, politics, and law, but this is not achievable in general, though of course in particular contexts a consensus may be created as to what particular words mean, or how a sign should be understood. Any interpretation or reading is open to challenge either in the immediate context or as part of a subsequent recontextualization. Ideal fixed points (such a dictionary meanings) are second-order constructs and therefore vulnerable to contextual challenge, but they may also serve as important orientation points in particular situations.
Writing is not a representation of speech, nor is it somehow a derivative from, nor a secondary system in relation to, speech. Yet writing plays a central role in second-order understandings of language in modern literate societies, and in ideas about rationality and forms of reasoning. Thus in Rationality and the Literate Mind (London: Taylor & Francis, 2009), Harris deconstructs the Aristotelian syllogism arguing that it rests on the language myth and on Eurocentric notions of the relationship between writing and thinking. Printing and the standardized typeface creates a visual stability which we project onto language as a whole; the print against the empty paper (or screen) background suggests that words and sentences can be separated from context and analyzed as free-standing autonomous objects. The attribution of methodological, cultural or intellectual primacy to writing Harris terms ‘scriptism’. Harris argues that writing is fundamental to modern orthodox linguistics, its commitment to the primacy of speech notwithstanding. Linguistics is ‘scriptivist’ in the sense that it has co-opted the relative stability and semiotic neutrality (speaker-independence) of writing in setting out its view of the linguistic sign; the ‘phoneme’ is unthinkable without the ‘letter’ and alphabetic writing.
Integrationism is lay-oriented, in that it puts individual experience at the centre of its theoretical concerns. It rejects the annexation by academic linguists of intellectual authority over the domain of language, and questions the privileged position and claims to scientific status of academic linguistics – a discipline which has repressed or forgotten the original stipulative definition on which it rests. There is no denying that Harris has been a polemical and unforgiving critic of what he sees as the shortcomings of linguistics, and many linguists have returned the compliment. Harris is frequently portrayed as a negative, even destructive, thinker, whose view of language and communication is at odds with the observable linguistic order. The lack of a prescribed methodology for the study of ‘data’ is seen as indicating integrationism’s intellectual sterility; Harris’ view of linguistics has been dismissed as a caricature, as ignoring the internal diversity and range of theoretical approaches found within the discipline and its sub-branches, thereby reducing it to a few tropes summed up in the ‘language myth’. However many reviews of Harris’ work, in particular those by academic linguists, display a poor understanding of integrationism’s basic aims and beliefs. Most linguists lack an interest in questions of semiology, and in the historical and philosophical foundations of their subject, and see linguistics as a search for improved descriptive methods or analytical models. Advanced training in the discipline of academic linguistics, especially its formal branches, does not typically include the kind of methodological, ideological and philosophical reflexivity that is central to anthropology. Modern linguistics has retained its intellectual autonomy by ignoring wider developments in the humanities and social sciences, and ‘theory-shopping’ in the cognitive and hard sciences. A selection of critical responses, ranging from the dismissive to the engaged, can be found in Linguistics Inside Out: Roy Harris and his Critics (eds. G. Wolf and N. Love, Amsterdam; Benjamins, 1997).
A note on After Epistemology (2009)
Harris has steadily broadened the theoretical range of his work, not merely in the sense that he has interrogated the foundations of ‘supercategories’ such as art, history and science, but in his engagement with the theory of knowledge. In After Epistemology Harris takes on the tradition of Western epistemology, arguing that it is based on a ‘linguistic consensus’ which is ‘irremediably flawed’ (p. 1). The work is divided into two parts: the first deals with ‘Epistemology in the Western tradition’; the second is entitled ‘Integrating knowledge’. Harris asserts that ‘truth’ and ‘certainty’ are only contingently implicated in epistemology: ‘truth’ is a ‘metalinguistic’ concept, that is, it belongs to the domain of the second-order; ‘certainty’ is primarily a psychological notion (p. 8), or a ‘psychological state or attitude’, so that ‘being certain of something is not the same as knowing it’ (p. 66). It is perhaps worth stressing that, while Harris puts language and theories of language at the centre of his discussion of supercategories and of epistemology, he is not appealing to a Whorfian constructionism, in which linguistic categories in some sense generate social reality, nor to any version of the idea that everything is mediated through, or realized in, an identifiable steady-state entity such as ‘language’ or ‘text’ or ‘discourse’.
Philosophers have pursued epistemology as if the human mind operated in the same manner regardless of social or historical context, and without regard to the possibility that ‘the forms of knowledge recognized and valued in any society depend on the forms of communication practiced in that society’ (p. 2). Against an epistemology that ‘floats in a philosophical and social vacuum of its own making’, and which rests on assumptions about language which are unstated or unexamined (termed ‘language-dependence’), Harris begins with the self as a ‘knowing agent’: ‘We are what we know’ (p. 3, italics in original). Philosophers in search of a foundation to knowledge have appealed to an ideal world of forms (Plato), a common set of ideas that refer to things (Aristotle), or a knowable set of stable objects (Bacon), for which an ideal language might be artificially devised (Wilkins). They have sought reassurance that, for at least a substantial set of words, we have a reliable guide to, or knowledge of, reality or thought, and grappled with the fear of linguistic-epistemological failure in the light of the ‘imperfection of words’ (pp. 31-37). Harris concludes: ‘We are still living today with the intellectual gridlock resulting from the impasse between (Baconian) reocentric semantics and (Lockean) psychocentric semantics (p. 37).
Humean skepticism about the basis of our knowledge of the external world represents not so much a challenge to epistemology but an object lesson in the muddle engendered by seeing language as a mirror of the world, and states of affairs as temporally isolated and discrete entities that somehow can be detached from the stream of human experience. This create a false puzzle of about our continuity of perception of these sequential states: ‘We are looking at our experience of these continuities in the wrong way if we assume that every nameable object must somehow be accompanied by a corresponding series of true propositions (name-plus-predicates) describing this object at every separate moment of its existence.’ (p. 42). The ‘deHumidification’ of knowledge requires the recognition that ‘reasoning about the world is based on knowledge of the world, not vice versa’ (p. 46). Hume’s faith in the transparency of language disqualifies him from acting as a reliable guide to knowledge (p. 46). Darwin, in the grip of a reocentric semantics, was unable to distinguish lexical from real definition: ‘what Darwin thought he had shown was a fact about Nature, i.e. that nature contains no immutable species. But what he had actually observed was something about the variable use of the word species by his mid-19th century contemporaries’ (p. 56). A. J. Ayer’s logical positivism regarded ‘the vocabulary of languages as being fortuitously diverse labeling systems for the same universal set of meanings’ – basically the Aristotelian position. Both Darwin and Ayer share this ‘code-plus-telementation’ model. In the structuralist turn, Saussure’s Cours retains the idea of the code and telementation, but the link between that code and the external world drops away, as does the idea of languages as ‘conventional’ – where conventions are understood as institutional arrangements under the control of human agency. This puts language at the heart of our understanding of reality (p. 71), inverting the traditional priority of the non-linguistic over the linguistic in the domain of knowledge.
Integrationists reject the Saussurean model, including notions of arbitrariness, linearity, meaning, rules, and the existence of languages as systems (p. 74). There is no general concept of ‘knowing a language’ which can be invoked to explain communication: ‘integrational proficiency’ is a matter of experience, there is no specific knowledge to which we can point to, as ‘specific cases vary without limit’ (p. 76). It follows from the basic principles of integrational semiology that knowledge ‘is a form of activity’; it is not a ‘passive accumulation’ (p. 80) but the product of engagement of the individual: ‘It [knowledge] is not an attempt to connect with something outside or alien to the self, but an internal development of the individual’s own integrational capacities, exercised and refined in an endless series of contexts that feed into one another’ (pp. 80-81). Knowing something to be the case cannot rely on a decontextual semiology to underwrite or sustain it, so in this sense ‘all communicable knowledge is local knowledge’ (p. 97). This follows from the fact that ‘signs do not exist independently of their integration into the particular activities that give rise to their creation’ (p. 97). There is no knowledge without a knower, and, for both the knower and knowledge itself, contexualization is key: ‘There is no contextless knowledge available for transmission, just as there are no contextless signs’ (pp. 99-100). It follows that ‘the act of contextualization and the identification of the sign as sign are one and the same. We contextualize as a condition of integrating new signs into the temporal dimension of experience’ (p. 103, italics in original).
The bi-planar model of the linguistic sign and the attendant processes of linguistic abstraction and reification have lead to a confusion of knowledge with information, or a reduction of knowledge to information, in which the deterministic code is the model (pp. 115-117). This codification is a step towards depersonalization and mechanization, and damages the basic societal model of what knowledge and the transmission of knowledge consist in. This culminates in the ‘deplorable tendency’ to ‘reify science’ and a scientistic understanding of knowledge (p. 125). Rejecting Popper’s reductionist account of objective knowledge as ‘knowledge without a knower’, Harris draws on Roger Jones’s Physics as Metaphor (London: Abacus, 1983) to offer an integrational account of measurement. Scientism and reductionism are particularly evident in the contemporary disciplines of neuroscience, psychology, cognitive science, etc. Pinker’s approach to stereotypes and conceptual categories offers ‘epistemological mumbo jumbo’, a ‘muddle of incompatible metaphors’ which present ‘pure metaphysics’ as psychological science (pp. 140-141). Worries about metaphor in epistemology can be traced back to Aristotle, whose attempt to maintain how the same term could be applied to two different things led to the distinction between ‘literal’ and metaphorical meaning. However ‘embracing “scientific method” does not render the hold of metaphor on truth any the less elusive’ (p. 151). Burke’s attempt to deal with the problem of perception relies ultimately on assuming that words and their meanings are related by cause and effect – our perceptions are shown to agree by invoking one version of the language myth (p. 160). Epistemology in the Western tradition consists of the search for a fixed or determined relationship between some permutation of language, thought, and the world.
What would it take to demythologize knowledge? Here are the basic principles that would guide such as project (p. 162):
Knowledge is not a matter of gaining access to something outside yourself; all knowledge is internally generated by the human capacity for sign-making; the external world supplies input to this creative process but does not predetermine the outcome; signs and, hence knowledge, arise from creative attempts to integrate the various activities of which human beings are capable.
In observing an object such as a cat at the end of the garden we do not access perceptual knowledge, rather we apply our observational knowledge. In recognizing the cat we do not undertake complex subliminal ratiocination, or process information – this jargon adds nothing to the account of our lay subjective experience: ‘if the integrationist approach to these matters is on the right lines, we do have first-order access to our own everyday experience, because we actually create own interpretations of those experiences’ (p. 166). In reporting those experiences to someone else, we introduce ‘a whole order of knowledge that is not reducible to what I know’ (p. 166).
In sketching out this integrational account, Harris distinguishes between ‘associative integration’ and ‘conjunctive integration’. Associative integration draws on the ‘vast stock of associations of visual, auditory, tactile and other modes of experience’; dancing is an example of ‘conjunctive integration’, a systematic attempt to integrate different domains of activity (p. 173):
Language is a form of integration in which associative and conjunctive integration combine. Speech and writing are biomechanically independent activities. Literacy is the matter of knowing how to integrate the two.
Recognizing this concept of knowledge means a break with explanations based on ‘decontextualized reifications’; it implies liberation from the ‘tyranny of the textbook’, opening up the door to ‘an education based on recognizing the full gamut of human and humane values’ (p. 173). This requires relinquishing the idea that words and meanings are ‘items of knowledge in themselves’ (p. 176) and that knowledge ‘makes no difference to what is known’. This task is ‘overwhelming and urgent’, given the ubiquity of the knowledge as information paradigm (p. 176).
Harris’ work can be best understood as clearing a space for a wide range of approaches and conceptualizations of language. The broad historical sweep of his writings, and his imaginative theorizations of art and writing in particular, provide material for an open-ended and creative academic response to questions of language and communication. Harris espouses a rhetoric of liberation and responsibility, rejecting the cultural and intellectual stupefaction which is the legacy of modern linguistic theorizing, most notably in the formalist branches of the philosophy of language and linguistics. Individuals and societies should not delegate their moral and cultural responsibility to a class of self-appointed experts on language, nor mistake the stipulative definitions of academic discourse for the complex, multifaceted reality of our unfolding and unbounded communicational worlds.
For Harris, the notions of ‘language’, ‘languages’ and ‘writing’ are not fixed cultural or social categories; the boundaries between human language, animal language and non-language are determined neither by biology nor by any particular academic discipline. What we understand as writing today may be seen merely as primitive pre-writing, or even ‘non-writing’, in the distant future (The Origins of Writing, p. 157). Harris makes the point in this way in Rethinking Writing (London: Athlone Press, 2000, p. 242):
When future generations are quite accustomed to sitting at a keyboard and ‘typing’ an audio-visual product that incorporates sounds, letter-forms and pictures systematically interrelated, they will have acquired a new concept of writing, a new concept of literature, a new concept of language.
Earlier writings by Harris are collected in The Foundations of Linguistic Theory (ed. N. Love, London: Routledge, 1990). Harris has also edited a number of collections of integrational writings, including Linguistic Thought in England (Duckworth: London, 1988), The Language Myth in Western Culture (London: Curzon, 2002), and with T.J. Taylor a reader entitled Landmarks in Linguistic Thought (London: Routledge, 1989). Further edited volumes with integrational themes are: Redefining Linguistics (eds. H. Davis and T. J. Taylor, London: Routledge, 1990), New Departures in Linguistics (ed. G. Wolf, New York: Garland, 1992) and Language and History (ed. N. Love, London: Routledge, 2004), Language Teaching (ed. M. Toolan, London: Routledge, 2009). Integrational papers are collected in Integrational Linguistics: A First Reader (R. Harris and G. Wolf, eds. (Oxford: Pergamon, 1998).
Selected Harris Bibliography
History, Science and the Limits of Language: An Integrationist Approach, 2003, Shimla, Indian Institute of Advanced Study.