A large proportion of the human race, some would say the majority (Cook, 2002), speak more than one language. Somehow two languages, two grammars can coexist within the confines of one mind. Is the existence of so many minds with two or more languages at all relevant to the UG theory?...

1. The purity of the monolingual argument

According to the separation of competence from performance, discussed in Chapter One, 'Linguistic theory is concerned with an ideal speaker-listener in a completely homogeneous speech community' (Chomsky, 1965, p. 4). A community with more than one language, or indeed more than one dialect, would not be homogenous: the language of a mixed community:

 'would not be "pure" in the relevant sense, because it would not represent a single set of choices among the options permitted by UG but rather would include "contradictory" choices for certain of these options’ (Chomsky, 1986a, p. 17).

The idealisation of competence reduces it to the knowledge of a monolingual native speaker. Describing a mind with two grammars is too complicated; it is vital to simplify the discussion to a mind with a single grammar. At the level of descriptive adequacy, the goal is therefore to describe what an idealised monolingual knows: the description of Universal Grammar is based on the knowledge of the native speaker with a single grammar.

Mostly this emphasis on monolingualism has simply been taken for granted by those working within the UG theory, along with the other areas excluded from competence, and is seldom discussed or justified. The only true knowledge of the language is taken to be that of the adult monolingual native speaker. Chomsky himself has rarely mentioned bilinguals. In a famous interview with Francois Grosjean, he said:

‘Why do chemists study H2O and not the stuff that you get out of the Charles River ? … You assume that anything as complicated as what is in the Charles River will only be understandable, if at all, on the basis of discovery of the fundamental principles that determine the nature of all matter, and those you have to learn about by studying pure cases.’

(for the benefit of readers without a knowledge of US geography, the Charles River divides Boston from Cambridge ). This can be called the 'purity' argument for using monolinguals: an idealised mind with one language possesses a purer state of language knowledge than a mind with two or more.

Chomsky does not of course deny that there are large numbers of bilinguals in the world: 'even in the United States , the idea that people speak one language is certainly not true … everyone grows up in a multilingual environment' (Chomsky, 2000a, p. 59). But this is irrelevant to the idealised competence of an individual. A second language is effectively an extra tacked on to the first language, like an extension to the back of the house. Doubtless it has an interest of its own in due course. But such an impure state cannot form the core subject matter of linguistics.

2. Universal Bilingualism

Alongside the purity argument, there is nevertheless a recognition that many, or indeed all, minds contain more than one grammar: 'whatever the language faculty is it can assume many different states in parallel' (Chomsky, 2000a, p. 59). Chomsky has often made comments to the effect that people effectively have more than one grammar in their minds: 'every person is multiply multilingual in a more technical sense' (Chomsky, 2000a, p. 44). For example, we may switch between speaking different dialects with different parameter settings, say standard English She’s good versus dialects such as US Black English that permit sentences without copula be as in She good. Or we may use different parameter settings for different registers: Haegeman and Ihsane (2002) have shown that diary writing in English in writers such as the fictional Bridget Jone or Virginia Woolf is pro-drop in that the first person is often a null subject – played gramophone .. so to tower, - , not a characteristic of Virginia Woolf's public prose style. Hence she is switching between two settings for the two styles and so has two grammars simultaneously.

A person who switches in this manner is then has elements of two grammars in one mind. Thomas Roeper argues that: ‘a narrow kind of bilingualism exists in every language. It is present whenever two properties exist in a language that are not statable within a single grammar' (Roeper, 1999, p.169) for example when adults with 'optional' rules are switching between two grammars as in the case of Virginia Woolf. Chomsky too often talks of bilingualism proper as simply the extreme end of a continuum of grammatical variation inherent within all speakers.

‘To say that people speak different languages is a bit like saying they live in different palaces or look different, notions that are perfectly useful for ordinary life, but are highly interest-relative. We say that a person speaks several languages, rather than several varieties of one, if the differences matters for some purpose or interest.’ (Chomsky, 2000a, pp. 43-44).

The simultaneous existence of two grammars in the same mind is also necessitated by language development (Roeper, 1999). It might be that children switch in toto from one parameter setting to another so that, say, one day they have null subjects in their speech, the next day they do not. Studies of development, however, show that such an abrupt transition seldom occurs: English children gradually decrease the number of null subjects in their speech over a period of time rather than going from having no subjects at all to having them in every sentence. Hence, during this transitional period, they must in effect have two grammars simultaneously, or at least two parameter settings. Any transition from one stage to another involves bilingualism in the sense of knowing two grammars or having two sets of parameter settings for an appreciable amount of time. The abstraction of competence to a single grammar is a fiction for most L1 native speakers who can use different dialects or genres and for most L2 learners; the typical human mind must entertain more than a single grammar. The issue is really whether it is proper to set this universal bilingualism to one side in linguists’ descriptions of competence or whether it should in effect form the basis of the description from the beginning.

3. The multi-competence view

The multi-competence theory, within which one of us works (Cook, 2002), sees these issues differently. If most people, or indeed all people, have multiple grammars in their minds, the idealisation to the monolingual native speaker is misleading, as inaccurate as studying how human beings breathe by looking at those with a single lung. If the architecture of the human mind involves two languages, we are falsifying it by studying only monolingual minds. To turn Chomsky’s metaphor back on him, water is a molecule H₂O, not an atom; if we break it into its constituent hydrogen and oxygen, we are no longer studying water. Purifying the mind into a single language means destroying the actual substance we are studying - the knowledge of language in the human mind.

The arguments about language acquisition discussed earlier were couched in terms of the potential inherent in all human children irrespective of the environment they encounter. Following the same line of reasoning, potentially all children can become bilingual; the ability to know more than one language is available to us all, even if it may decline after childhood. A person is a monolingual because of the accidental fact that they only encountered one language, and were unable to realise their bilingual or indeed multilingual potential. If you don't hear a second language, you won't speak one; if you do, you will. Since every human being has the potential to do this, UG theory has to take multilingualism as the norm for the human mind. Multiple grammars in the mind are not the exception but the norm, prevented only by accidental environmental features.

According to the multi-competence theory, then, the linguistic competence in the human mind potentially includes more than one language. UG theory has to account for this universal ability of the mind to have two, possibly conflicting, grammars at the same time; universals cannot be established by studying the minds of people who know one language, only minds of people who have fulfilled the multilingual potential of the human language faculty. Rather than making L2 user grammars conform to the the Procrustean bed of the monolingual grammars,  we need to establish principles from L2 grammars and see monolingual grammars as a restrictive set.

This position then treats the multilinguals of the world as the norm, not the monolinguals. The inhabitants of the Cameroon for example use 279 indigenous living languages (Ethnologue: Gordon, 2005). A priest from Tanzania spoke Kihaya in as a child, learnt Kiswahili in elementary school and English in secondary school, needed Latin for his religious training (but also learnt French out of curiosity at the same time), was posted to Uganda and Kenya, where he needed Rukiga and Kikamba, and went to Illinois where he uses Spanish to communicate with his parishioners (Cook, 2001). Londoners speak over 300 languages and 32% of their children have languages other than English at home (Baker & Eversley, 2000). While these seem extreme cases to those brought up as monolinguals in households with one language, they show that the potential for multi-language acquisition is dormant within everyone, even in countries such as England that are supposedly monolingual - exactly the kind of insight that the UG theory was set up to explain.
.....
From the perspective of multi-competence, the initial and the final L2 states are both states of a single mind, one containing knowledge of one language, the other knowledge of two or more. If any mind has the potential to learn more than one language, the principles and parameters of Universal Grammar have to be established from people who know more than one language. Yet the UG-related research tests L2 learners against the Universal Grammar established from monolinguals rather than establishing Universal Grammar from multilinguals using a poverty-of-the-stimulus argument and then seeing how monolinguals fail to acquire these. Rather than seeing L2 learners as failures for not knowing a principle such as the θ-Criterion known by L1 speakers, this undermines the failure of Universal Grammar likelihood of the θ-Criterion being part of.

In particular, multi-competence has raised the issue of how the second language knowledge in the S_t affects the other component of the final state – the first language (Cook, 2003): French speakers who know English react against French sentences using the middle voice 

(13) Un tricot de laine se lave à l’eau froide.
    *A wool sweater washes in cold water.

 compared to those who don't know English (Balcom, 2003); Japanese, Greek and Spanish speakers of English prefer the first noun to be the subject of the sentence in:

14. The dog pats the tree.

more than do those who do not know English (Cook et al, 2003). Multi-competence, far from denying the existence of Universal Grammar, insists that its description be based on the final language state of the normal human being, which includes more than one language.

.....