Is Language a Social Reality?

Short introduction to what the social reality of language might consist of, together with indications as to why language is both more and other than a social reality. The ideas of Saussure and Durkheim can be updated by using more recent philosophical work on the concept of Convention. At the same time using the ideas of Kant, Bhaskar, Giddens and some very general notions of human psychology it can be shown that conventionalised social realities cannot be the whole story of language. See also on this website "What is English if not a Language?" and "Wittgensteinians and Chomskyans".

Bhaskar R 1979 The Possibility of Naturalism. Harvester Press, Brighton

Bourdieu P 1991 Language and Symbolic Power. Polity Press, Cambridge

Chomsky N 1986 Knowledge of Language: Its Nature, Origin and Use. Praeger, New York

Croce B 1967 Aesthetics as Science of Expression and General Linguistics. Peter Owen, London

Durkheim E 1982 The Rules of Sociological Method.Macmillan, London

Giddens A 1979 Central Problems in Social Theory.Macmillan, London

Gilbert M 1989 On Social Facts. Routledge, London

Harris R 1981 The Language Myth. Duckworth, London

Haugen E 1972 "The Scandinavian languages as cultural artifacts". In his The Ecology of Language.Stanford University Press, Stanford, CA

Itkonen E 1978 Grammatical Theory and Metascience.Benjamins, Amsterdam

Katz J 1981 Language and Other Abstract Objects.Blackwell, Oxford

Katz J (ed.) 1985 Philosophy of Linguistics. Oxford University Press, Oxford

Labov W 1966 The Social Stratification of English in New York City. Center for Applied Linguistics, Washington, DC

Lewis D 1969 Convention. Princeton University Press, Princeton, NJ

Matthews P 1979 Generative Grammar and Linguistic Competence. George Allen and Unwin, London

Mothersill M 1984 Beauty Restored. Oxford University Press, Oxford

Pateman T 1987 Language in Mind and Language in Society. Oxford University Press, Oxford

Sampson G 1979 Liberty and Language. Oxford University Press, Oxford

Sampson G 1980 Making Sense. Oxford University Press, Oxford

Saussure F de 1913 Course in General Linguistics.Trans R.Harris. Duckworth, London

UlIman-Margalit E 1977 The Emergence of Norms.Oxford University Press, Oxford

First published on this Website 2007. An earlier version of this essay appeared in R E Asher, ed., The Encyclopaedia of Language and Linguistics.Pergamon Press, Oxford 1994

A social reality is any causally efficacious system, structure, or mechanism which can exist independently of any particular historical individual, but cannot exist independently of some group of individuals somehow linked together in sustaining or reproducing that reality from day to day (Bhaskar 1979; Durkheim 1982; Giddens 1979). So French is an existing social reality which pre-exists any infant born into a French-speaking society and which is causally efficacious in bringing it about, all other things being equal, that the infant ends up speaking French rather than German or any other readily - labelled language. In contrast, Cornish - like every other dead language - is not a social reality. It is not now sustained as a living language by any group of individuals, nor is it causally efficacious in producing any new speakers of Cornish by the usual mechanisms of language development. Cornish is simply an object of linguistic study and historical curiosity, and it is in the interests of the inhabitants of Cornwall that it should remain that way: in this case, as in others, attempts at language "revival" are likely to be no more than scams to obtain large sums of money from foolish donors.

Individuals may sustain (reproduce) a language from day to day without knowing that that is what they are doing, or needing to have as a goal or intention that they do it. Someone may use British Sign Language without having the belief or knowledge that they are using it - they may believe no more than that they are signing. Someone may speak English, knowing that they do so, but careless of whether their actions reproduce English in any particular form. Linguistic prescriptivism requires the reflexive monitoring (Giddens 1979) of speech and writing by individuals with the objective of ensuring that a language (their own or, for example, their pupils') is reproduced in some particular, valued form. Because of the tendency to change inherent in all language use, prescriptivism is essential to language maintenance and even then it is doomed to failure (Pateman 1987). The complex mechanisms of language change which operate below the level of self-conscious awareness constantly outsmart the conscious attempts to control them - attempts which can involve only partial understanding and, at least as often, actual misunderstanding.

Any discussion of language as a social reality will soon encounter the problem of the identity conditions for a language, which troubled Saussure (1983) and others since. How can the same thing (say, English) be different over time? In all probability I learnt to speak from people who learnt to speak from people who� spoke Anglo-Saxon. Some consequences of language change are catastrophic, in the sense of catastrophe theory: despite the unbroken chain, I do not understand Anglo-Saxon. There is a break-down of mutual intelligibility. On the other hand, Danish, Norwegian, and Swedish are counted as separate languages despite a high degree of mutual interintelligibility (Haugen 1972). Elsewhere (Pateman 1987 and on this website "What is English if not a Language?")) I have put forward the view that languages are best regarded as social facts (realities) which are not linguistic facts, developing Chomsky's distinction between E(xternal) and I(nternal)languages (Chomsky 1986) - a distinction between Languages (les langues) in that everyday sense in which they are social realities and the complex facts of language (langage) as a psychological reality.

As already implied in the example of the Scandinavian languages, coordination of speakers' actions on a common target language is by no means necessary to successful communication. There are both centrifugal and centripetal forces at work in language learning and language use. Centripetal forces do lead to mutual adjustment towards a regularity which can then become normative as a standard of right and wrong. The emergence of norms in this way is explored in Ullman-Margalit 1977, drawing on the pioneering work by Lewis (Lewis 1969; for a critical discussion, see Gilbert 1989). Such mechanisms are distinct from, and prior to, conscious planning and prescription and may lead to approximations rather than strict identities. When people have a simple interest in achieving success in relatively simple situations, then centripetel forces push them no farther than is required than to understand each other. Every day and all over the world, millions of transactions occur in which approximations to "English" are used successfully by tourists and their hosts in achieving ordinary objectives - getting from A to B, ordering a meal, checking in to a hotel. These successful transactions horrify linguistic purists.

If French is a social reality and Cornish is not, at least in the required sense, this implies that a language can be described without invoking a social reality for it. This is a necessary condition of the possibility of autonomous linguistics, though not a sufficient condition, since some other kind of reality (e.g., psychological) might still have to be invoked ( Katz 1981 and 1985.)

However, it is arguable that dead and living languages are essentially different. A dead language can in principle be fully described by a grammar, since it has left behind a finite and now - closed corpus. This is not even in principle possible for a living language. Living languages are open-ended activity-sustained practices; they are never fully conventionalized social realities or fully determinate psychological realities. There is always scope for individuals using 'a language' to create new forms which were not already implicit in social conventions or internal psychological rules - even rules with infinite generative power. On this view (for which see Croce 1967, Harris 1981, Matthews 1979, Sampson 1979 and 1980), there can be true intuitions of grammaticality which do not derive from pre-existing rules just as there can, on a view deriving from Kant, be true judgments of beauty which do not derive from rules of beauty (Mothersill 1984) or, on a related view, true judgements of right which do not derive from rules of justice. Such intuitions are not in the requisite sense obviously social: they may be distributively shared by individuals - by which I mean that individuals may happen to think alike - but they are not socially derived. This view is anathema to 'strong' social reductionists, such as Bourdieu 1991, a writer as much terrified by the idea of individuals escaping social control as the Academie francaise, which sets itself the futile task of turning back the tide of linguistic change. But equally this view raises doubts about the possibility of an autonomous linguistics, this time from the side of individual psychology, and its adherents have spent much time attacking Chomskky and all his works.

Language is, however, clearly a social reality when its resources - its symbolic powers - are used by speakers in efforts to affect other social realities. So, for example, in hypercorrection language is being used to affect perceived social status relationships and all that then follows from the estimation of those realities. Such phenomena have been studied by both linguists and sociologists, operating within very different scientific traditions, exemplified by Labov (1966) and Bourdieu (1991).