Prolegomena to Second Language Learning

Vivian Cook   On-line Writings    SLA Topics

[My more recent thinking on these topics is in V. Cook & Li Wei (2016) The Cambridge Handbook of Linguistic Multi-Competence. Looking back from 2019, I think the ideas raised here are still crucial even if noone has really discussed or challenged them in the interval.]

Draft of paper in Seedhouse, Jenks and Walsh (eds) Conceptualising Language Learning 2010. Earlier version in ‘The nature of the L2 user’, in L. Roberts, A. Gurel, S. Tatar & L. Marti (eds.) EUROSLA Yearbook, 7, 205-220, 2007; reprinted in L. Wei (ed.) (2011), The Routledge Applied Linguistics Reader, Routledge 77-89. Some images have been added from the original Powerpoint presentation, which was received in stunned silence by the audience.


The aim here is to discuss the ‘second language’ in second language acquisition (SLA) research. The phrase ‘second language learning’ involves three parameters of variation: the nature of language, the nature of a second language and the nature of learning. Prolegomena are the preliminary matters to be considered before the main work, as in Hjelmslev (1943/1961) Prolegomena to a Theory of Linguistics. This paper argues that discussions of learning depend on what is learnt – language – and what makes a language ‘second’. Virtually no book on SLA current today defines these two terms more extensively than the odd footnote on second versus foreign language learning. Yet they are the rationale for the existence of second language acquisition as a distinct area of enquiry. As VanPatten & Williams (2007: 7) say, ‘Any theory about second language acquisition needs to be clear what it means by language’. 

Is, say, the concept of language used in Universal Grammar-based L2 research the same as that in Vygotskyan studies? The language used in studying bilingual social networks the same as that in psycholinguistics processing? The language used in post-modern studies of discourse the same as in procedural/ declarative models? The language in transfer studies the same as that used with emotion? If  SLA research indeed has a framework within which it can reconcile all these views of language, it is a major intellectual feat. If not, then  SLA research needs to state explicitly the meaning of ‘language’ that it is using in each area.

Meanings of ‘language’

The English word ‘language’ has many different meanings. One danger is that its apparent translation equivalents in other languages convey different implications, particularly troublesome in a field of multilingual researchers largely writing in English. For example, de Saussure’s three-way French distinction between ‘langue’, ‘langage’ and ‘parole’ (de Saussure 1915/1976 ) has been a stumbling-block for English translators and linguists for almost a century.

The English word ‘language’ has many different meanings. One danger is that its apparent translation equivalents in other languages convey different implications, particularly troublesome in a field of multilingual researchers largely writing in English. For example, de Saussure’s three-way French distinction between ‘langue’, ‘langage’ and ‘parole’ (de Saussure 1915/1976) has been a stumbling-block for English translators and linguists for almost a century. Table 1 gives six thumbnail meanings of ‘language’, for convenience given the labels Lang1 to Lang6. The meanings of ‘language’ used here have been chosen to cast light on SLA rather than as watertight definitions; doubtless a more rigorous categorisation could be devised. A brief version of these has been presented in Cook (2007).1 For once the Oxford English Dictionary (OED, 2009) is not very helpful; the main entry for ‘language’ defines it as ‘the system of spoken or written communication used by a particular country, people, community etc’ neatly bundling together the meanings to be deconstructed here (and disenfranchising sign languages).

Lang1 a human representation system
Lang2 an abstract external entity
Lang3 a set of actual or potential sentences
Lang4 the possession of a community
Lang5 the knowledge in the mind of an individual
Lang6 a form of action

Table 1 Meanings of 'language'

Lang1: language as a human representation system

From Cicero in 55BC (De Inventione, I, IV) to Hauser et al. (2002), people have seen human beings as standing out from other creatures by having language – ‘a species-unique format for cognitive representation’ (Tomasello, 2003: 13). This is not to say that there is agreement over what the human aspects of language actually are; any feature claimed to be unique to humans is soon found somewhere in the animal kingdom. Hauser et al. (2002) for instance argued that the sole distinctive element of human language is recursive embedding, that is to say structures embedded within structures of the same type over and over again; soon after, Gentner et al (2006) trained starlings to react to supposedly recursive embedding.

Interpreting second language learning as second Lang1 learning leads to the question of ‘the extent to which the underlying linguistic competence of learners or speakers of a second language (L2) is governed by the same universal properties that govern natural language in general’ (White, 2003: xi): does the ‘language’ in second language learning qualify as a human language. It has indeed been asserted that ‘L2 learners are not only creating a rule system which is far more complicated than the native system, but which is not definable in linguistic theory’ (Clahsen & Muysken, 1986: 116). The consequence of second languages not being Lang1 languages would be that SLA research could be investigated in the same way as the learning of any human cognitive abilities like chess-playing and would presumably be handled by psychologists within general human learning rather than by SLA researchers, as indeed happens in some psychological approaches. Second Lang1 learning research is then concerned with the ultimate question of the species-specificity of language itself, not with the details of any individual second language.

Lang2: language as an abstract external entity

‘Language’ also refers to an abstract entity – objective knowledge in Popper’s world 3 of abstract ideas (Popper, 1972: 159), as in ‘the English language’ or ‘the Chinese language’. A Lang2 is something out there in the world of abstractions, like the rules of football – a creation of the human mind that stands outside any individual and that can be captured in codified rules in a dictionary and a grammar book, Le Petit Robert (2006) for French words, A Grammar Of Contemporary English (Quirk, Greenbaum, Leech & Svartvik, 1972) for English grammar, sometimes legislated for by institutions such as the French Academy. No single person knows a Lang2 language as such – no speaker of English knows more than a fraction of the 263,917 entries and 741,149 meanings in the Oxford English Dictionary (OED, 2009). Lang2 entities are countable – English, Chinese and so on up to the 6,912 in Ethnologue (Gordon, 2005). A Lang2 often becomes a symbol of a particular culture or country and hence one language gets lauded over others: ‘Most books on English imply in one way or another that our language is superior to all others’ (Bryson, 1991: 3). People are proud of their Lang2 and loyal to it, even if they see their own knowledge of it as imperfect. Lang2 has been used to justify wars of territoriality, as in Hitler’s claims to German-speaking areas of Europe, fostered as part of the independence movement for minority groups such as Catalans, imposed upon conquered territories as in the imposition of Japanese names on Koreans in 1939, and used as a unifying lingua franca to fight against its native speakers as in the Black People’s Convention in South Africa (Biko, 1978).

SLA research has usually only related to the standard national language; variation and acceptability of non-status dialects or local varieties are ignored. Status differences between Lang2s also have to be taken into account such as whether the Lang2 in question counts as peripheral (local), central, supercentral or hypercentral (De Swaan, 2001), since second language learning typically ascends rather than descends this hierarchy (Cook, 2009). Research then relates to a particular abstract Lang2 entity with particular political and social attributes. SLA research has to ensure that the L2 users it is studying have indeed been exposed to the standard form it is testing; for example English L2 users who make a distinction between singular you and plural yous or who omit present tense third person -s may be following Geordie or Norfolk practice rather than the standard.

Lang3: language as a set of sentences

In the Lang3 sense, a language is a set of sentences: all the sentences that have been said or could be said make up the language – ‘the totality of utterances that can be made in a speech-community’ (Bloomfield, 1926/1957: 26) or ‘a set (finite or infinite) of sentences, each finite in length and constructed out of a finite set of elements’ (Chomsky, 1957: 13). English is the name for one such set of sentences, Chinese for another. Lang3 seems to be the sense in which language is internalised in usage-based learning (Tomasello, 2003): language knowledge emerges from the array of data that the learner encounters, distilled from a corpus. Lang3 is not an abstraction but a concrete object, made up of physical sounds, gestures or written symbols. Patterns can be extracted from these primary data, whether by the linguist or the learner. But they remain patterns of data rather than systems of knowledge or behaviour.

Second Lang3 learning concerns the second language as a set of recorded phenomena , for example ‘the utterances which are produced when the learner attempts to say sentences of a TL [target language]’ (Selinker, 1972). The relevant evidence for Lang3 research is L2 users’ sentences, as in the Vienna Oxford International Corpus of English (VOICE) (Seidlhofer, 2002), and the L2 spelling corpora of van Berkel (2005). The overall research questions concern the regularities and frequencies in L2 language data.

The unique problem for second Lang3 learning research is how to categorise the set of sentences produced or encountered by a L2 user as belonging to one Lang3language or another. Weinreich (1953: 7) said that ‘A structuralist theory of communication which distinguishes between speech and language ... necessarily assumes that every speech event belongs to a definite language’; a recent variant is ‘I define bilingual input as dual‑language input consisting mainly of substantial numbers of utterances that both lexically and structurally belong to one language only’ (De Houwer, 2005: 31). But how do you tell which bits belong to which language? How can the total set of sentences produced by an L2 user be divided neatly into two ‘languages’ without invoking a meaning of language other than Lang3? The L1 sentences a L2 user produces often differ from those of a monolingual native speaker for example. Second Lang3 learning needs to include the whole set of sentences produced by the learner; only later can they be assigned to languages according to other criteria. Bilingual speech therapists have indeed long argued that therapy should be based on the child’s first language as well as on their second: you can’t tell what’s wrong with either if you don’t look at both (Stow & Dodd, 2003). Second Lang3 learning research consists of the description and analysis of occurring language data.

Lang4: language as the shared possession of a community

Lang4, the possession of a language community, is often seen as complementary to Lang5, the knowledge in the individual’s mind: ‘although languages are thus the work of nations … they still remain the self-creations of individuals ’(Humboldt, 1836/1999: 44) or ‘le langage a un côté individuel et un côté social, et l’on ne peut concevoir l’un sans l’autre’ (language has an individual side and a social side and one cannot imagine one without the other) (de Saussure, 1915/1976: 24).

Lang4 is a social phenomenon, a cultural product shared among a group – ‘the English-speaking world’, ‘native speakers of Chinese’ etc. Possession of a language confers membership of a particular group. The language community is often equated to the nation – people born in Korea tend to speak Korean. But in many cases it is an imaginary community unconstrained by political borders (Anderson, 1983): Kurdish is spoken by nine million people in Iraq, Turkey and Iran though there is no modern country of Kurdistan (Gordon, 2005). Indeed Anderson (1983) argues that the equation of language with nation only arose in the nineteenth century; the inhabitants of the Austro-Hungarian Empire did not primarily see themselves as speakers of German; many Kings and Queens of England have been speakers of French, German and Dutch, a tradition continuing with Prince Philip, born in Corfu and christened Philippos.

Second Lang4 learning then depends upon the concept of the language community. ‘Language acquisition and use are viewed here as indicators of individual and group integration into the host society and/or alienation from it’ (Dittmar et al, 1998: 124). As this quotation illustrates, the normal community is typically assumed to be monolingual: ‘An individual’s use of two languages supposes the existence of two different language communities; it does not suppose the existence of a bilingual community’ (Mackey, 1972: 554). SLA researchers, language teachers, and indeed many L2 users themselves have seen the L2 user as petitioning to join the monolingual community of native speakers. The alternative view is now starting to be heard that there are bilingual communities in which speaking more than one language is the norm . As Canagarajah (2005: 17) says of Sri Lanka:

One can imagine the difficulty for people in my region to identify themselves as native speakers of “a” language. People may identify themselves as speakers of different languages very fluidly, based on the different contexts of interaction and competing claims on their affiliation.

Second Lang4 learning research has to decide whether the community that its research is dealing with is monolingual or multilingual – the multi-competence of the community (Brutt-Griffler, 2002; Cook, 2007, 2009).

Lang5: language as knowledge in the mind of the individual

Language is also the mental possession of an individual, Lang5: ‘a language is a state of the faculty of language, an I‑language, in technical usage’ (Chomsky, 2005: 2), the other side of the coin to Lang4. An individual has a mental state, consisting of rules, weightings, principles or whatever, which constitutes their language competence alias ‘the speaker-hearer’s knowledge of his language’ (Chomsky, 1965: 4). Speakers of Lang5 English possess something that allows them to connect the world outside to the concepts inside their minds in a particular way. Competence then contrasts with performance – ‘the actual use of language in concrete situations’ (Chomsky, 1965: 4). Partly the competence/performance distinction recognises the difference between Lang3 sets of sentences and Lang5 mental knowledge, partly between competence as a declarative state and actual speech production.

Second Lang5 learning involves what the L2 user knows: ‘the goal is to provide a clear and accurate account of the learner’s competence’ (Ellis, 1994: 38). Lang5 is perhaps the most typical meaning of ‘language’ underlying SLA research, as stressed in Firth and Wagner (1997). If mental knowledge is equated with Chomskyan linguistic competence, the problem has always been how this relates to performance. Data from any form of performance is dubious evidence for competence, a gap as hard to bridge in SLA research as in any other area of language. To go over familiar ground, Lang3 performance data are full of mistakes and disfluencies (the degeneracy of the data); many mental rules of grammar are not derivable from the properties of Lang3 sets of sentence (the poverty-of-the-stimulus). Hence competence is not derivable neatly from sheer Lang2 data through discovery procedures (Chomsky, 1957), as usage-based linguistics would have it (Tomasello, 2003). Second Lang5 learning research is concerned with testing and specifying internal invisible contents of the mind.

Lang6: language as action  

The Lang6 tradition of language as a form of action goes from Malinowski’s language ‘is a mode of action and not an instrument of reflection’ (Malinowski, 1923: 312) through Austin (1962)’s performative utterances to Schegloff et al (2002: 5), ‘People use language and concomitant forms of conduct to do things, not only to transmit information’. Language as action is an integral part of Vygotskyan theory: language starts as external social action by the child and development consists of internalising this external action (Vygotsky, 1934/62). In a sense Vygotsky’s notion might be seen as combining other meanings already discussed, say elements of external Lang3 with social Lang4 and mental Lang5. There is nevertheless a tenable view that language is a set of observable actions, an external cultural construction.

Second Lang6 learning is about how people act in second languages: ‘a competence-in-action… socially situated, collaboratively established and contingent with regard to other competencies’ (Pekarek Doehler, distinct both from the neutral descriptions of Lang3 and from the community-based Lang4 in its assertion of an individual sociality. In its Vygotskyan guise, ‘sociocultural research seeks to study mediated mind’ (Lantolf, 2000: 18). 

Finally an orthogonal point has to be mentioned that, whichever meaning of ‘language’ is used, SLA research needs a rigorous analysis of its subject matter, language. Dealing with language as the possession of human beings, Lang1, entails a precise specification of what this consists of, whether principle and parameter syntax, phonemes, weightings of connections or whatever. For Lang3, analysing the properties of a set of language means having well-defined categories and procedures, as in say Pattern Grammar. And so on for all the other meanings: each implies a particular form of description and analysis. As Cook (2008) observed, much psycholinguistics research ‘utilises descriptive terms such as phonemes (‘the building blocks’…), syllables, words, meanings, as if these were primitive concepts that were unproblematic and unrelated to any linguistic theory.’ The definition of ‘word’ for instance, much relied on in SLA research, has proved a sticky problem: increasingly words have been seen as an artefact of writing systems that use word spaces rather than a universal category (Aronoff, 1992); the word is hard to use as a category in Chinese, which separates morpheme-related characters by spaces, not words. Like phoneme, word may be an indispensable label; but, compared to say ‘lexical item’ or ‘lexical entry’ or even ‘lemma’, it needs a definition to give it scientific status. Choosing a meaning for ‘language’ does not absolve us from using scientifically acceptable tools for analysing it, varying according to the meaning involved. If we believe that language consists of words and phonemes, we look at how words and phonemes are acquired; if we believe language is abstract rules and parameters, associations of stimulus and response, weighting of connections, competition between elements, schemas for social interaction, or any of the milliard of alternatives, we need to rigorously describe these constructs before looking for ways in which they may be learnt.

To come back to the general argument, the six meanings of ‘language’ can now be applied to language learning. Tomasello (2003: 7) for example states ‘the principles and structures whose existence is difficult to explain without universal grammar … are theory-internal affairs and simply do not exist in usage-based theories of language – full stop’. Exactly: statements about Lang3 are not statements about Lang5 , any more than swallows are fish; why should they be? The usual linguists' criticisms of connectionism are, not that it is wrong: but that it is not about language as conceived of by linguists; Lang3 and Lang5 are incompatible once again but is that a reason for denying a universe in which Lang3 exists?

In other words, views of language learning based on a particular meaning of ‘language’ are often incompatible with views of language learning based on another. Only those who are proselytising for the dominance of a particular view of language need to spend time arguing it against all rivals; most of  us can continue to cultivate our own gardens without throwing weedkiller over the fence into the next. Paradis (2009: 3) talks of:

… the arrogant attitude of those who are convinced that they hold the Truth and who treat anyone with diverging views as muddleheaded, using intimidation to impose their views – only to discover a decade or less later that they were wrong, at which point they go on to defend and try to impose the new Truth with the same determination and contempt for diverging views.

Unless people agree on the underlying meaning of ‘language’, these arguments are unwinnable – or both sides claim victory because the opposition cannot account for some aspect of language learning whose very existence the other denies, say the Lang1 principles denied by usage-based learning or the Lang5 abstract structures denied by connectionists. Take the continuing debate initiated by Firth and Wagner (1997: 2007) that SLA research has stressed individual psychology at the expense of social interaction. Interpreted as a plea for more work on Lang4 and Lang6 rather than an exclusive concentration on Lang5, this seems timely. It may be historically true that syntactic development did preoccupy many 1970s researchers (with the exception of discourse work by Hatch and her associates (Hatch, 1978)) and still dominates generative SLA approaches, as reported in White (2003). It is desirable for SLA research to investigate a greater range of areas and explore different interpretations of ‘language’, though one should extend this brief beyond social interaction to gestures, writing systems, neurolinguistics, and doubtless many other newly-thriving areas.

There is little point to refuting a theory based on one idea about language with arguments based on another. Wagner and Firth (2007) document the reactions to their original paper, which amount largely to ‘my view of language is right’:

Long, Kasper, Poulisse and Gass … lay down the law by defining SLA’s ‘proper’ intellectual territory (e.g., ‘learners’, ‘language’, ‘cognition’), delineating its ‘key concerns’ (e.g., ‘acquisition’, not ‘use’, ‘language’, not ‘communication’), and by pointing to its borders (e.g., by stipulating what is ‘inside’ and ‘outside’ SLA). In the Feyerabend approach to science, all avenues should be explored simultaneously (Feyerabend, 1975).

A more subtle danger is switching from one meaning to another in the course of discussion. Let us take a seminal SLA paper by Diane Larsen-Freeman (1997) to try to demonstrate how SLA research employs multiple meanings of ‘language’; this paper brought to SLA research for the first time the ideas of chaos theory and started an influential line of thinking to this day. It consisted half of a presentation of chaos theory, half of speculations how these ideas might be used in SLA research. It admits its wide coverage of ‘language, the brain, and second language acquisition’ and insists that it is not a matter of either X or Y but of both X and Y: ‘chaos/complexity theory encourages a blurring of boundaries in SLA – to see complementarity and to practise inclusiveness where linguists have seen oppositions and exclusiveness’ (p.158).

Table 6 presents a selection of citations for the word ‘language’ from the paper arrayed according to the meaning of ‘language’ they seem to belong to. While this not very easy to do, some clearly belong more to one meaning of language than another. Statements like ‘language is a fractal’ (p.150) presumably belong to Lang1 human language; statements about change like ‘anything borrowed into the language’ (p.150) seem to refer to Lang2 abstract entities; many of the uses of ‘language’ envisage it as a constantly changing pool of elements ‘every time language is used, it changes’ (p.148), clearly a Lang3 view of language as a set of actually occurring pieces of spoken or written text. Lang4 community is implied by the attribution of ‘proficient speakers of a given language’ (p.151), Lang5 mental state by ‘UG the initial state of human language’ (p.150). The recognition that ‘language is a dynamic system’ (p.147) and similar claims seems to belong to Lang6 language as action.


‘language is also complex’ p.149; ‘if language is as complex as it is’ p.154

‘language is a fractal’ p.150; ‘the fractality of language’ p.150

‘language and language acquisition are like other complex systems in the physical sciences’ p.152

‘ILs, like all natural languages, are unstable’ p.156


‘the changes which languages undergo diachronically’ p.147

‘the source language, the target language’ p.151

‘anything borrowed into the language’ p.150


‘language can be conceptualised as aggregations of paradigmatic and syntagmatic units’ p.147

‘every time language is used, it changes’ p.148

‘language use and language change are inseparable’ p.158

‘Zipf’s law is not only applicable to a language in general’ p.150


‘proficient speakers of a given language’ p.151


‘we might call UG the initial state of human language’ p.150

‘why do we not each wind up each creating our own language…?’ p.153


‘language as a dynamic system’ p.147

‘language grows and organises itself from the bottom up in an organic way’ p.148

Table 2 ‘language’ in Larsen-Freeman 1997

In other words the paper shifts through the gamut of meanings of ‘language’. Chaos theory affects our ideas of the nature of language, of individual language change, of language as a set of sentences, of language in the community and in the mind, and of language in action. Yet the arguments for human language being fractal are not necessarily the same as those for chaos theory applying to a set of sentences or to the changes in a named language like English, and so on. The illusion is that all the uses of ‘language’ are about the same thing, amounting to a sleight-of-hand, equivalent to saying ‘I’m just popping out to the bank’ and returning a week later having visited the Dogger Bank in the North Sea .

Counting languages

We now turn to the word ‘second’ in ‘second language learning’. The rare discussions of this in SLA research only concern the ‘second/foreign language’ distinction, with the exceptions of Stern (1983) and Block (2003). Talking about ‘first’ and ‘second’ languages is not just counting how many there are with cardinal numbers ‘one/two’ but putting languages in an order with ordinal numbers ‘first/second’; Language Two (Dulay et al, 1982) is in principle a different concept from Understanding Second Language Acquisition (Ortega, 2009). There is a curious paradox that  SLA researchers say ‘L2 learners’ aloud as ‘L two learners’ rather than ‘2nd L learners’, switching from cardinal to ordinal. Of course the ‘first’ in ‘first language acquisition’ is equally problematic since monolingual children never go on to a second language; my wife was not very pleased to be referred to as my first wife.

What kinds of order could ‘first’ and ‘second’ represent?

- official first language by fiat

Politics has often used ‘first’ and ‘second’ language alongside ‘official’ language. Countries adopt their national languages by constitution – the European Union currently has 23 official languages, having the most multilingual institutions in the world (De Swaan, 2001: 144). In Canada, ‘English and French … have equality of status and equal rights and privileges as to their use’ (Official Languages Act, 2006: preamble); all Canadian citizens belong to one or other or both of the two official language communities regardless of mother tongue (Churchill, 2004). The official language of a country has little to do with whether many people speak it as their Lang4 or Lang5; French is the official language of Senegal though 75% of the population speak Wolof. Singapore by constitution has four official languages, Malay, Mandarin, Tamil and English, and one national language, ‘the Malay language … in the Roman script’, yet Bahasa Malaysia is spoken by only 14% of the population. Officialdom may also lay down which language should be the first foreign language, and indeed second foreign language (European Union, 1995).

Official languages are examples of countable Lang2 nation-bound entities: it is taken for granted that the language of England is English. Though the definitions of official language and foreign languages vary from one country to another, these concepts nevertheless have to be allowed for in SLA research, if only to be able to communicate with the organisations that administer language policy. But it is particularly important when asking L2 users what their first language is to know what they mean by ‘first language’; a student from Taiwan may say ‘Chinese’ (which is ambiguous among many ‘dialects’), ‘Mandarin’ (the official language) or ‘Taiwanese’ (a local ‘dialect’ of Min Nan Chinese spoken by 70% of the population).

- first and second as sequence of acquisition

The numbering of languages may also correspond to the chronological order of an individual’s development or acquisition: ‘the learning of the “second” language takes place sometime later than the acquisition of the first language’ (Mitchell & Myles, 1998: 1). The first language is thus acquired before the second in the lifetime of the individual; Joseph Conrad learnt Polish before he learnt French, making Polish his L1, French his L2. This applies both to the individual gaining membership of a second Lang4 community and to them gaining a second Lang5 mental system: one language comes before another in their life-history. Simultaneous early bilingualism in which the baby handles two languages from birth is something else, covered in Swain’s memorable phrase ‘bilingualism as a first language’ (Swain, 1972).

The use of ‘second’ should not be taken too literally. Many sources maintain that it subsumes later languages; Doughty and Long (2003: 3) enumerate SLA includes ‘second (third, etc) languages and dialects’; Lightbown and Spada (2006: 204) say a ‘second language: … may actually refer to the third or fourth language’. In this sense English was Conrad’s L2, although he learnt it third. The implication is that learning of languages beyond the second is no different from learning a second language. However this has been strongly denied by those working with trilingualism or multilingualism whose goal is ‘to work out the differences and similarities between SLA and TLA [Third Language Acquisition]’ (Jessner, 2008: 19) rather than take their identity for granted. It may be necessary to order languages beyond second rather than having second subsume all later-acquired languages.

Sequence of acquisition thus involves several meanings of ‘language’. One is the official standard language that the person encounters first, a Lang2, English before French say. A second is the Lang4 community the person belongs to first compared to a community they join later, say the Chinese community in Singapore before the English-speaking bilingual community. A third is the Lang5 mental system that the person acquires after their first Lang5. The multi-competence approach argues, however, that the mind is a single linguistic system at some level; it is arbitrary to divide up this complex mental system into bits labelled ‘first’ and ‘second’ rather than treating it as a whole.

- first and second by priority

‘First’ and ‘second’ can also be a matter of value judgement: something which comes first is better than something which comes second – ‘first choice’, ‘First Lady’, ‘first class degree’. Your first language is the language you command best; your second is therefore worse: ‘ “second language” indicates a lower level of actual or perceived proficiency’ (Stern, 1983: 13).

A much-explored topic over the years has been language dominance: ‘We use the terms "first language" and "second language" to refer to relative language dominance’ (Chee et al, 2004: 15270). This could be the dominance of one Lang4 community over another: ‘A language used by a socio-economically dominant group in society or which has received a political or cultural status superior to that of other languages in the community’ (Hamers & Blanc, 2000: 373). French was the dominant language in England from 1066 to 1385, yet undoubtedly most people spoke English, just as most people in India spoke Indian languages during the British Raj. Nor is the mother tongue necessarily closest to one’s identity; Myhill (2003: 84) points to ‘Hebrew in Israel and Yiddish in Ultra-Orthodox communities, in both of which cases native language is a distant second in terms of centrality to identity’.

More often, dominance has meant psychological dominance of one Lang5 language in the individual mind – ‘the second, and less dominant, language’ (van Hell & Dijkstra, 2002: 780). Considerable effort was expended to establish a speaker’s dominant language, reviewed in Flege et al (2002), for example the test batteries by Lambert (1955), leading to the concept of balanced bilinguals as ‘those equally fluent in two languages’ (Grosjean, 1982: 233) . However the L2 repertoire of an L2 user may be wider than their L1; they cannot be judged just on how well they can carry out L1 functions in the L2: Greek students in England for example say that they can only write essays in English since essay-writing did not feature in their L1 education.

These senses of dominance give priority to the first language. The dominance of one community over another is not relevant to multilingual communities where several languages are in balance. As Canagarajah (2005: 16) points out:

Although the now discredited notions such as native speaker or mother tongue speaker require us to identify ourselves according to our parental language or language of infancy, even the alternatives such as L1 and L2 force us to identify a single language as receiving primacy in terms of our time of acquisition or level of competence.

Nor does the dominance of one Lang5 language in the individual’s mind square with the idea that the two languages form an interrelated system. It is only in these senses that your first language may change into your second when it becomes dominant in your external or internal life; otherwise a second language will remain second for evermore.

- second and foreign by situation

We can now come back to the second versus foreign language distinction, introduced into EFL teaching in the 1950s (Howatt, 1984). A typical definition can be found in Klein (1986: 19):

… “foreign language’ is used to denote a language acquired in a milieu where it is normally not in use … A “second language” on the other hand, is one that becomes another tool of communication alongside the first language; it is typically acquired in a social environment in which it is actually spoken.

This incorporates two contrasts. One is function: a second language meets a real‑life need of the L2 user, say to communicate with the majority community – a Chinese speaker using English in Newcastle upon Tyne; a foreign language fulfils no current need for the speaker – a Newcastle schoolchild learning French. The other contrast is location: a second language is learnt in a milieu where it is used by native speakers – German in Berlin; a foreign language is learnt in a place where it is not widely used – German in Japan. Block (2003) draws out the further contrast that a second language is acquired naturalistically, a foreign language is learnt in a classroom.

Much SLA discussion does not take the second/foreign distinction on board, either rejecting it explicitly (Ellis, 1985: 2; Mitchell & Myles, 1998: 2), or playing safe by referring to ‘the learner of a second or foreign language’ (Council of Europe, 1997: 12), or using alternative formulations such as ‘first’ versus ‘foreign’ (Johnson, 2001). The second/foreign distinction is far from transparent. I used to teach English as a Foreign Language in London to students intending to return shortly to their own countries despite currently using it as a second language; students at English-medium universities may effectively be using it as a second language whether in Saudi Arabia or the Netherlands.

De Groot & Hell (2005: 25) perceive a difference between North American usage, where a language not native to a country can be either ‘foreign’ or ‘second’, and British usage, where ‘foreign’ means not spoken in a country and ‘second’ means not ‘native’ but used widely as medium of communication, say English in Nigeria. There is the additional confusion that what is referred to as ‘foreign language teaching’ in North America is often called ‘modern language teaching’ in Europe. Stern (1983: 10) sums up: ‘ “foreign language” can be subjectively “a language which is not my L1” or objectively “a language which has no legal status within the national boundaries” ’.

The distinction was useful for EFL teachers in capturing two broad perspectives on their work. It applies most easily to languages that are confined to one locale: Finnish is either a foreign language outside Finland or a second language for people acquiring it within Finland. It is more problematic when it concerns languages that are widely spoken by non-native speakers to other non-native speakers across the globe (Berns, 1990). A second language is presumably a Lang4; acquiring a second language allows you to join another community. A foreign language, however, in one sense is the Lang2 abstract entity laid down as a goal by education, in another the individual’s Lang4 or Lang5 potential stored up for future use.

A wide variety of people are learning second languages in diverse situations for multiple functions. ‘First’ or ‘second’ language are historical terms inadequate to cover the complexity of language in our societies and in our minds. The second/foreign language distinction oversimplifies the myriad dimensions of second language learning, as the papers in VanPatten and Lee (1990) bear out. In particular it applies uneasily to heritage language learning where people are acquiring a language that is culturally important to them, say Mandarin for those of Chinese origin (who may speak other Chinese dialects such as Cantonese), or Polish for those of Polish descent in London: 140,000 people are attending heritage Chinese classes in the United States (Brecht & Rivers, 2005). The reason for learning a heritage language is not primarily to use it as a second or foreign language but to identify with a particular cultural tradition. Similarly while school teaching of a modern language has often been seen as involving a ‘foreign’ language, depending on circumstance, it has only future use for the learners as a foreign language, many of whom will never use it for second or foreign purposes; it is just a subject on the academic curriculum, neither second nor foreign.

Many researchers manage perfectly well without the second/foreign distinction. Ellis (1994: 12) nevertheless claims ‘it is possible there will be radical differences in both what is learnt and how it is learnt’ in second and foreign situations. True as this may be, without more evidence, we cannot tell if this two-way distinction is more crucial than any of the others. Cook (t.a.) argues for a spectrum of at least six groups of language users. Rather than a simple opposition of second and foreign, we need multiple distinctions to capture the range of people acquiring second languages. While SLA research is doubtless stuck with the word ‘second’ in its name, this does not means it cannot be more rigorous in the ways it actually approaches the diverse situations of second language learning.


This paper has tried to demonstrate that discussion of second language learning depends on getting straight what is learnt and the circumstances in which it is learnt, i.e. ‘language’ and ‘second’. Communication depends on a shared set of meanings, whether between people or between rival theories of language learning. There may be occasions when different meanings of ‘second language’ can be fruitfully combined. But in each case it has to be shown that there is sufficient compatibility between them. Without agreement over ‘second language’, SLA researchers perforce have to follow separate paths on different maps. The danger is that SLA researchers often do not realise that they are working from different maps and exhaust their energy quarrelling over differences in basic assumptions or patiently defending them against their critics. These prolegomena have suggested that SLA research needs to  base itself on the foundations of specific notions of ‘language’ and ‘second’ before it ventures to tackle the idea of ‘learning’. The word ‘learning’ indeed covers an array of processes, conditions and states in different theories; a simple division such as Krashen’s ‘acquisition’ (natural, unconscious etc) versus ‘learning’ (formal, conscious etc) (e.g. Dulay et al, 1982), now enjoying a second life as Paradis’ distinction between conscious learning and unconscious acquisition (Paradis, 2009), is as vast an oversimplification of ‘learning’ as ‘language and ‘second’ have turned out to be, as the other papers in this volume readily demonstrate.


1. We will not be using the distinction here between conscious learning and unconscious acquisition as now put forward by Paradis (2009).

2. This analysis has benefited from comments from many colleagues and from anonymous referees in earlier incarnations.


Anderson , B. 1983. Imagined Communities. New York: Verso.

Aronoff, M. 1992. Segmentalism in linguistics: the alphabetic basis of phonological theory. In P. Downing, S.D. Lima & M. Noonan (eds) The Linguistics of Literacy, Amsterdam: Benjamins, 71–82.

Block, D. 2003. The Social Turn in Second Language Acquisition. Edinburgh: Edinburgh University Press.

Bloomfield, L. 1926. A set of postulates for the science of language. Language 2, 153‑64.  Repr. in M. Joos (ed.), Readings in Linguistics I. Chicago: University of Chicago Press, 1957

Brutt-Griffler, J. 2002. World English: a Study of its Development. Clevedon: Multilingual Matters.

Bryson, B. 1991. Mother Tongue: the English Language. London: Penguin.

Canagarajah, A.S. 2005. Reconstructing local knowledge, reconfiguring language studies. In Canagarajah, A.S. (ed.), Reclaiming the Local in Language Policy and Practice. New Jersey: Lawrence Erlbaum Associates, 3-24.

Chee, M.W.L., Soon, C.S., Lee, H.L & Pallier, C. 2004. Left insula activation: a marker for language attainment in bilinguals. PNAS, 101/42: 15265-15270.

Chomsky, N. 2005. Three factors in language design. Linguistic Inquiry, 36/1: 1-22.

Churchill, S.  2004. Official Languages in Canada : Changing the Language Landscape. Canadian Heritage: Official Languages Support Program Branch.

Cicero 55BC. De Inventione, I, IV. by C. D. Yonge in The Orations of Marcus Tullius Cicero. London : George Bell & Sons.

Clahsen, H. & P. Muysken 1989. The UG paradox in L2 acquisition. Second Language Research , 5: 1‑29.

Cook, V.J. 1979. The English are only Human. English Language Teaching Journal, XXXIII, 3: 163-167.

Cook, V.J. 2007. The nature of the L2 user. In L. Roberts, A. Gurel, S. Tatar & L. Marti (eds.) EUROSLA Yearbook, 7: 205-220. Reprinted in L. Wei (ed.) (2011), The Routledge Applied Linguistics Reader, Routledge 77-89

Cook, V.J. 2008. Review of Handbook of Bilingualism: Psycholinguistic Approaches. Ed. J.F. Kroll & A. M.B. de Groot’, Language,

Cook, V.J. 2009. Language User Groups and language teaching. In Cook, V.J. & Li Wei (eds), Contemporary Applied Linguistics, Vol 1. London : Continuum, 54-74

Cook, V.J. and Bassetti, B. (eds.) 2005. Second Language Writing Systems. Clevedon: Multilingual Matters.

Council of Europe 1997. Language policies for a multilingual and multicultural Europe. Strasburg: Council for Cultural Cooperation Committee CC-LANG (98) 1.

De Groot, A.M.B. & van Hell J.G. 2005. The learning of foreign language vocabulary. In Kroll & De Groot (eds.), 9-29.

de Saussure, F. 1915/1976. Cours de Linguistique Générale. Ed. Bally, C. & A. Sechehaye (1915). Critical edition by T. de Maurio. Paris: Payothèque, Payot.

Dittmar, N., Spolsky, B. & Walters, J. 1998. Language and identity in immigrant language acquisition and use. In Contemporary Approaches to Second Language Acquisition in Social Context. University College Dublin

Feyerabend, P. 1975. Against Method. London: Verso.

Flege, J.E., Mackay, I.R.A. & Piske, T. 2002. Assessing bilingual dominance. Applied Psycholinguistics, 23: 567-598.

Gentner, T.Q., Fenn, K.M., Margoliash, D. & Nusbaum, H.C. 2006. Recursive syntactic pattern learning by songbirds. Nature, 440: 1204-1207.

Gordon, R.G. Jr. (ed.) 2005. Ethnologue: Languages of the World. Fifteenth edition. Dallas , Tex. : SIL International. Online version:

Hauser, M.D., Chomsky, N. & Fitch, T.M. 2002. The faculty of language: what is it, who has it and how did it evolve. Science, 298: 1569-157

Humboldt, W. von 1836/1999. On Language. Translated by P. Heath. Cambridge: Cambridge University Press.

Jessner, U. 2008. Teaching third languages: findings, trends and challenges. Language Teaching, 41, 1: 15–56.

Johnson, K. 2001. An Introduction to Foreign Language Learning and Teaching. Harlow : Longman.

Klein, W. 1986. Second Language Acquisition. Cambridge : Cambridge University Press.

Kroll, J.F. & De Groot, A.M.B. (eds.) 2005. Handbook of Bilingualism: Psycholinguistic Approaches. Oxford: Oxford University Press

Lambert, W.E. 1955. Measurement of the linguistic dominance of bilinguals. Journal of Abnormal and Social Psychology 50: 197-200.

Lantolf, J. (ed.) 2000. Sociocultural Theory and Second Language Learning. Oxford: Oxford University Press.

Larsen-Freeman, D. 1997. Chaos/complexity science and second language acquisition. Applied Linguistics, 18: 141-65

Lightbown, P. & Spada, N. 2006. How Languages are Learned. 2nd edition. Oxford: Oxford University Press.

Mackey, W.F. 1972. The description of bilingualism. In Fishman, J.A. (ed.): Readings in the Sociology of Language. The Hague : Mouton, 554-584.

Malinowski, B. 1923. The problem of meaning in primitive languages. In C.K. Ogden & I.A. Richards, The Meaning of Meaning. London.

Mitchell, R. & Myles, F. 1998. Second Language Learning Theories. London: Edward Arnold. New ed. 2004

Myhill, J. 2003. The native speaker, identity and the authenticity hierarchy. Language Sciences, 25, 1: 77-97.

Ortega, L. 2009. Understanding Second Language Acquisition. London: Hodder Education.

Oxford English Dictionary (OED) (2009). On-line version, Oxford : Oxford University Press.

Popper, K.R. 1972. Objective Knowledge. Oxford: Clarendon Press.

Quirk, R., Greenbaum, S., Leech, G. & Svartvik, J. 1972. A Grammar of Contemporary English. London : Longman.

Schegloff, E.A., Koshik, I., Jacoby, S. & Olsher, A. 2002. Conversation Analysis and Applied Linguistics. Annual Review of Applied Linguistics, 22: 3-31.

Seidlhofer , B. 2002. Pedagogy and local learner corpora: working with learning-driven data. In S. Granger, J. Hung & S. Petch-Tyson (eds.), Computer Learner Corpora, Second Language Acquisition and Foreign Language Teaching, 213-234.

Stow, C. & Dodd, B. 2003. Providing an equitable service to bilingual children in the UK : a review. Int. J. Lang. Comm. Diss., 38, 4: 351-377.

van Berkel, A. 2005. The role of the phonological strategy in learning to spell in English as a second language. In Cook, V.J. & Bassetti, B. (eds.) Second Language Writing Systems. Clevedon: Multilingual Matters. 97-121.

van Hell, J.G. & Dijkstra, T. 2002. Foreign language knowledge can influence native language performance on exclusively native contexts. Psychonomic Bulletin and Review, 9, 4: 780-789.

VanPatten, B. & Lee, J.F. (eds.) 1990. Second Language Acquisition-Foreign Language Learning. Clevedon: Multilingual Matters.

VanPatten, B. & Williams, J. 2007.Theories in Second Language Acquisition. Mahwah, New Jersey: Lawrence Erlbaum Associates.

Vygotsky, L.S. 1934/1962. Thought and Language. Cambridge Ma.: MIT Press

Weinreich, U. 1953. Languages in Contact. The Hague: Mouton.

White, L. 2003. Second Language Acquisition and Universal Grammar. Cambridge: Cambridge University Press.