Some Key issues for SLA Research

Vivian Cook, Newcastle University

draft of paper in Luciana Pedrazzini, Andrea Nava (eds.) Learning and Teaching English: Insights from Research, Polimetrica Publisher, Italy (2012). Sorry the text hasn't done well in html.
[This is essentially expanded at book length in V. Cook & D. Singleton (2014) Key Topics in Second Language Acquisition, Multilingual Matters]

There have always been those with a broad interest in how people acquire a second language (L2) whether Aristotle or Roger Bacon (Thomas, 1998). With the advent of academic psychology and linguistics, there came a certain amount of curiosity about second language acquisition (SLA) (Thorndike, 1928; Cheydleur, 1932); some linguists kept diaries of bilingual language acquisition (Leopold, 1939); others were interested in multilingualism and bilingualism, particularly in the USA (Weinreich, 1953; Fishman, 1966). Methods for teaching second languages such as Lado (1965) were bolstered with ideas from contemporary language learning theories usually derived from general psychological theories such as Skinner (1957) rather than those specifically about either second language learning or language acquisition in general. In the 1960s these interests from psychology, linguistics and language teaching came together to found a specific discipline of SLA research. For some fifty years this has set itself up as an independent discipline with its own theories and issues. This has not, however, stopped people in other disciplines from researching second language acquisition with little or no reference to the discipline of SLA research itself, whether neurolinguists (Fabbro, 1999), psychologists (Beauvillain & Grainger, 1987), or others.

The banner which the ‘new’ SLA research discipline raised had on it the word interlanguage, a term coined by Selinker (1972); 'what gave SLA its excitement was the concept of interlanguage' (Davies, Criper & Howatt, 1984, p.xii). Since the early 1960s, first language (L1) acquisition researchers such as McNeill (1966) had insisted that the child should be treated as having a different grammar from the adult rather than having a defective adult grammar. A two-year-old’s sentence such as Baby highchair should not be interpreted as a crude version of the adult sentence The baby is in the highchair but as a child’s sentence in its own right. The analyst’s task was not to measure the mistakes in the child’s sentence against the rules of adult grammar, e.g. the ‘missing’ articles, preposition and copula in Baby highchair, but to see how the rules of the child’s grammar worked as a combination of ‘pivot’ and ‘open’ words (Braine, 1963).

This ‘independent grammar assumption’ (Cook, 1993) was seized on by applied linguists working with language teaching, whether as ‘transitional competence’ (Corder, 1971) or 'approximative system' (Nemser, 1971). The accepted term, however, became ‘interlanguage’, leading to journals such as Interlanguage Studies Bulletin Utrecht, later Second Language Research. Interlanguage became the ‘almost theory-neutral’ term for the ‘system of implicit L2 knowledge that the learner develops and systematically amends over time’ (Ellis, 1994, 354).   

Second language learners are then developing an interlanguage of their own that draws not only on the first language they already know and on the second language they are learning but also on other elements such as the language provided by their teachers and their own language learning strategies. L2 learners’ interlanguage thus has unique qualities of its own rather than being a deficient version of the target language; it is no more defective than a two-year-old’s grammar is defective but has a logic of its own.

A fine example of interlanguage was the discovery of a basic stage of grammar common to adult L2 learners (Klein & Perdue, 1997). Regardless of which first and second languages are involved, L2 learners share a simple grammar with three grammatical rules, namely that a sentence may consist of: 

- a Noun Phrase (NP) followed by a Verb, optionally followed by another Noun Phrase, girl take bread

- a Noun Phrase followed by a copula and another Noun Phrase or an adjective, it’s bread

- a Verb followed by a Noun Phrase, pinching its

L2 learners create an interlanguage with its own distinctive characteristics, resembling neither the target nor the first language.

Without the concept of interlanguage, most SLA research would cease to exist; it provides a unique subject matter for the discipline that is not the main focus of other disciplines; the aim of SLA research is to discover ‘why ... adults attain the state they do’ (Klein & Perdue, 1992, 334). In line with the independent grammar assumption, the learner is not a defective native speaker but something unique, sui generis; the implicit aim of research is to discover what the L2 learners’ language is like and how it is developing, not to see how L2 learners fall short of monolingual native speakers.

Yet in practice this principle has not been thoroughly assimilated by SLA researchers, as seen in a typical quotation: ‘Relative to native speaker's linguistic competence, learners' inter-language is deficient by definition’ (Kasper & Kellerman, 1997, 5). Learners are still treated as wannabe native speakers. The question of whether age affects success in L2 learning for example is interpreted as whether older L2 learners speak less like native speakers than younger L2 learners. SLA research methods have implicitly treated L2 learners as defective by using monolingual native speakers as the yardstick for comparison. Grammaticality judgments that ask for L2 users to decide whether a sentence is grammatical are usually set against those of monolingual native speakers (Hawkins & Chen, 1997); the obligatory occurrences technique that looks for occasions when certain words or structures must be used (Pienemann, 1998) usually define them in terms of when monolingual native speakers use them; and many others (Cook, 1997).

The starting point for SLA research is that L2 learners are not, and can never be, monolingual native speakers – by definition. Without this assumption, SLA research becomes an ancillary study of why L2 users fail to become native speakers and at best provides a footnote to first language acquisition by detailing the L2 problems and pitfalls. If L2 users are unique specimens, SLA research can take on a true independence, looking at their distinctive qualities in their own right independent of monolinguals. SLA research deals with one way in which humans learn language, part of a greater discipline of language acquisition that encompasses L1 and L2 acquisition. Indeed monolingual L1 acquisition can be seen as an aberration that it only occurs when children are deprived of exposure to a second language (Cook, 2009).

1. Are people who know two languages special? 

Since 1990 the multi-competence perspective has tried to enumerate the diverse ways in which L2 users differ from monolingual native speakers (Cook, 1993; Cook, 2010).

· L2 users have different ways of thinking

The relationship between and cognition has become a vital new area of research. During the 1930’s Whorf put forward ‘the linguistic relativity principle’ that a person’s way of seeing the world is relative to the language they speak (Whorf, 1941/1956). This was mostly interpreted as a claim that language actually determines thinking, named the Sapir/Whorf hypothesis. While actively researched in the 1950s (Ervin-Tripp, 2011), this became largely discredited. In the 1990s the language and cognition link was revitalised by Lucy (1992) about differing ways of categorising the world, Levinson (1996) about differing concepts of direction and Roberson et al (1999) about varying colour perception, using a new set of ideas and methods. It became apparent that certain ways of thinking go with particular language does not necessarily cause or determine the way of thinking.

So what happens to the thinking of people who know more than one language? Here are some of the possibilities:

· L2 users use language in different ways

If your standard or norm is how a monolingual native speaker communicates with other mono-lingual native speakers then you are bound to find non-native speakers less effective in their second language. On the one hand this assumption ignores the L2 uses that the person does not have in their L1; students studying through the medium of a second language may be able do things they cannot do in the first language – write essays and reports for example. On the other hand it ignores the things that only L2 users can do. The main examples of distinctive L2 use are codeswitching in which the speaker alternates between two languages, translation in which the speaker turns utterances in one language into a second, and the ability to communicate with other L2 users. These are some of the unique things that L2 users can do; while they ultimately lead to specialist professions such as simultaneous interpreting, they are also probably present in the everyday lives of most bilinguals, for example young bilingual children acting as interpreters when their parents see a doctor. The skills of code-switching have been well-documented, for example through the 4M model (Myers-Scotton, 2006) which sees codeswitching broadly as relying on a matrix language for the structure of the sentence and an embedded language for the content words.

· L2 users have an increased awareness of language itself

Time and again research has shown that in some sense L2 users are more aware of the nature of language itself than monolinguals. Young children who learn another language are more conscious of the arbitrariness of language, for example that small words like train may refer to long things while long words like caterpillar may refer to short things (Bialystok, 2001). One intriguing aspect concerns theory of mind – the ability to see the other’s point of view: bilingual children acquire this ability slightly earlier than monolingual children (Goetz, 2003); children who were bilingual in a sign language and a spoken language are better at theory of mind than monolingual children (Meristo et al., 2007).

· L2 users have a slightly different knowledge of their first language

Everyday experience shows that your L1 exerts a strong influence on your L2. Only recently have people come to see that the L2 also affects your knowledge and use of your L1 (Cook, 2003). A simple example reported by many L2 users concerns the subject of the sentence. In pro-drop languages such as Italian, sentences need not have a subject Sono di Milano; in non-pro-drop languages like English, they do I am from Milan. Many overseas students with pro-drop L1s coming to England have reported problems in speaking their first language when they go home, such as their parents telling them they sound English because they use too many subjects. Research with Japanese users of English shows that the first language of more advanced L2 users is affected by English in that they find the presence of subjects more natural (Cook, Kasai & Sasaki, 2005). This effect of the L2 on the L1 has been found in a number of language areas such as intonation (Mennen, 2004), voice onset time (Zampini & Green, 2001), and pragmatics (Pavlenko, 2003 ). A person who speaks another language is no longer a ‘pure’ speaker of their first language but speaks a version that has been affected by the other language or languages they know (Cook, 2003).

· L2 users have greater effectiveness in their first language

Learning another language may also change your ability to communicate in your first language. This might be a general effect of learning any second language or the specific effect of learning a particular language. To take two examples, Hungarian children who had been taught English were better at L1 essay-writing than those who had not met English (Kecskes & Papp, 2000); English children taught Italian for an hour a week were better at reading English(Yelland et al, 1993). Not to mention such bilingual writers as Nabokov, Conrad or Brink. Indeed the claims for a knock-on effect on L1 support the old adage of ‘brain-training’ often used in England to justify the teaching of Latin. People who know more than one language are distinct from mono-lingual native speakers in several ways. Learning a L2 changes people overall in thinking and language knowledge and use. Learning a L2 is not just adding an extension to the exterior of your house; it is rebuilding most of the interior walls.

2 What is the best age for learning a second language?

The question of the best age for learning a second language has aroused many people’s curiosity and has practical concerns for parents bringing children up bilingually and for governments deciding the age to start teaching a second language to children. Undoubtedly there is a popular belief that young children are best at L2 learning, shared by many mainstream linguists: ‘It is a common observation that a young child of immigrant parents may learn a second language in the streets, from other children, with amazing rapidity . . . while the subtleties that become second nature to the child may elude his parents despite high motivation and continued practice’ (Chomsky, 1959, 49). But is there any empirical support for this ‘common observation’?

It looks a simple matter: test some people who start young and some who start old and see who is better. However, like most academic questions, it turns out to be almost unanswerable in the form in which it is asked. The answer cannot for example be assumed to be the same for those acquiring the second language in natural circumstances and for those being taught in a classroom; though it may be that situations for natural L2 learning are fairly few in number, those for class-room learners vary according to the educational system and the language teaching methods involved. Even the word ‘age’ is problematic; L2 researchers often use it to refer to the age of arrival (AoA) in another country, thus confounding age with immigration, restricting the people studied to immigrants, usually to the USA – far from a random selection of L2 learners (Cook, 1986) – and leaving it uncertain how much L2 teaching or exposure the people had received before immigrating – one reason for going to a specific country may be a familiarity with the language spoken there. The research design is also highly problematic: a proper balancing of young and old would also involve them having the same amounts of L2 exposure (Munoz, 2008); ‘The crucial comparison is between the language proficiency of learners of two age groups who have learnt the second language for the same period of time; time has to be taken into account not only as the age at which learning started, but also as the duration of learning’ (Cook, 1986). Comparing children’s acquisition with that of adults is also fraught with problems, given the many non-language ways in which children are developing (Cook, 2010), for example memory capacity and Piagetian stage of development and the many differences in their situations and language input.

Underlying much of the discussion is the idea of ‘critical period’. The ethologist Konrad Lorenz originally based the idea of critical period on the imprinting behaviour of ducks (Lorenz, 1949); after hatching, ducklings imprint a single person as ‘mother’ once and for all and cannot do so after this critical period. The idea that there are certain periods of physiological development during which an organism can learn particular behaviour then spread to much study of animal behaviour and was applied to language development by Eric Lenneberg, who suggested the Critical Period Hypothesis (CPH) that the ability to learn language naturally atrophies after the early teens (Lenneberg, 1967); for example while all babies start by distinguishing pairs of sounds like /ba~da/ by the age of 12 months they are only sensitive to the sound contrasts used in their first language (Werker & Tees, 1984) and cannot distinguish between non-native sounds.

A common way of expressing the conclusions on age in SLA initiated by Krashen, Scarcella & Long (1982) is as a set of slightly paradoxical statements, given in Cook (1986) as:

1. Older children are better than younger children at learning a second language

2. Adults are better than children at learning a second language

3. Immigrants who start learning a second language younger end up better speakers than those who start older

More recent research does not seem to have undermined these mixed findings. Given the same circumstances for acquiring the second language for the same amount of time, older children are better than younger children, particularly in school. Cenoz (2003) for instance compared Spanish/Basque speaking children aged 4, 8 and 11 who had learnt English for the same period and found the older children were better. Munoz (2008) sums up the current view on classroom acquisition as supporting the ideas that older learners learn faster than younger ones and that younger learners only have an advantage when they have more exposure, particularly in listening comprehension.

If we apply the idea that L2 users are intrinsically different from monolingual native speaker presented earlier to the age issue any conclusions become problematic. The measure of success in age studies is always approximation to monolingual native speakers (Birdsong, 2005); Hyltenstam & Abrahamsson (2003) for instance claim ‘absolute native-like command of an L2 may in fact never be possible’ for older learners; Johnson and Newport (1989)find that ‘later age of acquisition determines that one will not become native [-like] or near-native [-like] in a language. ’ The effects of age on second language acquisition are not necessarily established by demonstrating that speakers speak more like or less like monolingual native speakers.

 3 How is second language vocabulary acquired?

The acquisition vocabulary has turned out a difficult area to research. It is possible to describe the nature of the vocabulary that people have to learn in a first or a second language; it is possible to test how many words people know in a language; it is far harder to say how they actually acquire them. This section draws on ideas that are developed further in Cook (2009).

The nature of vocabulary  

It is conventional to start the discussion of vocabulary acquisition by describing the sheer complexity of the problem. Most psychologists and many language teachers for example assume that a word has a single distinct meaning that bridges the ‘real’ world and the concept in the human mind, the relationship called reference diagrammed in Fig 3. In English the word dog refers to the thing , i.e. it links a real dog to the concept of ‘dog’. The relationship always involves the human mind, whether wanting to talk about a and saying dog or hearing dog and working out it means .

Would that vocabulary were so simple! Dog can refer to people dirty dog, things that fail that record was a real dog, a constellation in the sky the dog star, an instrument with jaws iron dog, and many more: most words in English have more than one meaning. The word with the highest number of distinct meanings in the Oxford English Dictionary (OED, 1996) is set, with no less than 430. Learning a language means far more than learning one meaning per word. It involves learning a variety of information about a word, such as: dog is pronounced /dg/ and written <dog>; it is a noun and occasionally a verb she dogged his footsteps; the noun dog is countable and animate so you can say the dogs died; it occurs in phrases such as going to the dogs; and so on. Multiplied by the thousands of words each of us knows.

But discrete objects in the world are only one type of word meaning. Many words refer to abstractions like people and government, to things we can’t see such as air and truth, or to things that probably don’t exist like unicorns and Kryptonite. Nouns are only one type of word and we also need lexical words like verbs fly, adverbs highly and adjectives red, as well as structure words like prepositions for and articles the that have primarily grammatical meanings. According to rough calculations a speaker of a language knows around 60,000 words and children learn ten of them every day of their lives up to at least fifteen (Bloom, 2002).

For most linguists the crucial thing is not the word in isolation as studied in psychology experiments but the relationships that words have with each other in the mind. Dog is not cat, i.e. the two words reflect a categorisation of objects in the world and are antonyms: words contrast with other words. Dog is a ‘basic’ level term included in the ‘superordinate’ level term animal and itself including ‘subordinate’ level terms terriers, corgis and sheepdogs: words are structured into levels of categorisation. Dog is associated with other words in the minds, such as cat, collar, bark, leg, Alsation, berry, bite, black, bow, carnivore (Edinburgh Word Association Thesaurus, 2011).

So what happens in a second language? One possibility is seen in Fig. 4 below, using English as L1 and French as L2, though of course it cheats as it uses a picture rather than a real object – to paraphrase Magritte ‘Ceci n’est pas un chien’.         

The thing connects to the L2 word chien as well as to the L1 word dog; the words link in turn to the same concept of ‘dog’. De Groot (2003) calls the L1 and L2 words dog and chien the lexical level, the concept of ‘dog’ the conceptual level. So this is ‘the three-components two-levels ’model of L2 lexical representation – three components (L1 word, L2 word, and concept), two levels (lexical and conceptual). The interesting question is how the two languages interact.

One possibility is that the object links to the L2 word chien and then to the concept, the parallel route shown in Fig 4. This is called the concept mediation model as the link between L2 word and the L1 word is via the concept. Another possibility seen in Fig 5 is that the learner does not link the object to the concept but the word chien to the word dog at the lexical level: L2 access to the concept is mediated by the L1. This is the ‘word association model’ as it links L2 word to L1 word; the route from object to concept has been diverted via the L1. 

Figure 5 

These two alternatives hark back to the distinction between compound bilingualism in which the languages are closely tied together in the mind and coordinate bilingualism in which they exist side by side (Weinreich, 1953). So the complexity of vocabulary in L2 users is more than doubled. To the vast number of words with many meanings in the L1 are added the vast numbers of L2words via direct or the indirect links to the concepts. The L2 user has to learn all the other attributes of words, for example not just the associations for dog but also those for chien, some of which may be similar, some quite different.

The two models vary in how they relate the two lexicons in the mind. They might be entirely distinct as in the concept mediation model or they might be inextricably tied together. Some research indeed shows that it is impossible to switch one language off while you use another. Spivey and Marian (1999) for example tested people’s eye-movements as they processed pictures of objects, showing they never switched off either language.

Learning vocabulary

People continue to be amazed by the sheer size and complexity of the learning task for vocabulary when the number of words that learners acquire is multiplied by the number of meanings, relationships, associations and all the other paraphernalia that go along with every word we know. The question of learning vocabularyis still largely baffling. The commonsense view depends on the idea of reference given above and is often called ‘naming’. To take a well-known hypothetical example (Quine, 1960), suppose a parent says gavagai to their child and points at a rabbit: how does the child know what they mean? Gavagai might mean ‘rabbit’ in general, or it might be the pet-name for a particular rabbit. Or it could mean ‘one for the pot’. Or it could describe various aspects of the rabbit such as its fur, its colour, its size, etc. Or a particular part of the rabbit such as its nose, legs etc. The problem with learning how words go with things is sorting out which aspects of what we see are important, which are irrelevant. What is more, Tomasello(1999) has shown in experiments in which objects with new names are hidden in boxes that it is perfectly possible for children to acquire words for physical objects they have never seen.

A main approach to L2 acquisition of vocabulary in recent years has been to look at the strategies that L2 learners apply to the task (Cook, 2008). One set of strategies are for understanding the meaning of new words. Suppose that you were an L2 learner of English who encountered the phrase the phone-hacking scandal and didn’t know what hacking meant. The strategies you might adopt are:

-   guess from the context. Obviously hacking is some scandalous activity to do with phones.

-   use a dictionary. The OED suggests hacking is ‘The use of a computer for the satisfaction it gives; the activity of a hacker’ – ‘A person who uses his skill with computers to try to gain unauthorized access to computer files or networks.’ It has not caught up with phone-hacking but gives some idea of naughty activity.

-   make deductions from the word form. This is not of much use here apart from the -ing ending showing that some sort of activity is involved.

-   link to cognates. If you’re German, it might suggest hackend; otherwise like much computer-based vocabulary this seems an international term used across many languages.

On the other hand suppose that you want to learn hacking as a new word: your strategies might be:

-   repetition and rote learning. So you repeat hacking over and over to yourself.

-   organising words in the mind. This involves putting hacking into say the set of words for communication such as internet, Google etc.

-   linking to existing knowledge. This strategy consists of tying the word hacking into something else you know, for example a French speaker might remember hacking as a pirate wearing a hat – the French translation is piratage.

These are all conscious ways of tackling new words. Undoubtedly as in L1 acquisition most L2 words are picked up unconsciously as we use the language.

4. How important is grammar in acquiring and using a second language?

Undoubtedly the word grammar poses a problem. It makes some think of the school grammar famous for such claims as ‘A noun is the name of a person, place or thing’. It makes others think of the traditional EFL idea of structures found in every coursebook: ‘You use the present perfect to mean contemporary relevance’. It makes teachers think of the loathed teaching technique of presenting grammatical rules to the students. It makes linguists think of Chomsky’s use of grammar to refer to knowledge of language in a person’s mind encompassing syntax, vocabulary and phonology. Many SLA researchers now prefer the term morphosyntax, at odds with linguists who regard morphosyntax as the small intersection between syntax and morphology rather than an inclusive term for both. Some people think of grammar as a book, some as knowledge in the mind.

The sense of grammar that is most relevant here is the linguists’ idea of grammar as an internal property of people’s minds. When we know a language, we know its phrase structure: the sentence  I like green tea has a structure consisting of NPs I and green tea and a VP like green tea; the NP green tea consists of an adjective green and a noun tea. When we know a language we know how to construct and understand sentences, that is to say to link the words in it together and put them in sequence. Some of this know-ledge consists of syntactic parameters: I like green tea shows the English syntactic parameter-setting for non-pro-drop rather than the Italian setting for pro-drop languages Amo il tè verde, and shows the English Adjective Noun setting green tea, rather than the Italian setting Noun Adjective tè verde.

Second language acquisition of grammatical knowledge  

In the early days of SLA research the study of grammar had two aims: to establish SLA research as an independent discipline and to discover the sequence in which L2 learners acquired grammar. The classic study by Dulay and Burt (1974) took some ideas from the Roger Brown work with children’s language (Brown, 1972) and applied them to young Spanish-speaking learners of English. Brown had made a distinction between content words like tea or walk and grammatical morphemes, which include not only structure words like the and to but also grammatical inflections such as the plural form -s and the present participle -ing. Like many, Brown had observed that young children start by leaving out the grammatical morphemes; a two-year-old is more likely to say Mummy go shop than Mummy is going to the shops. By studying three English-speaking children over several years, he discovered a particular sequence over time in which children introduce the most crucial grammatical morphemes in their sentence:

1   -ing playing
2   plural -s tables
3   irregular past tense ran/took
4   possessive -s John's
5   articles a/the a book
6   regular past tense -ed liked/ waited/ played
7 third person -s  likes
Brown (1973) L1 acquisition sequence for English in chronological order of acquisition

I.e. children start with the -ing form going; progress to the plural -s shops; then the irregular past tense went; and so on till they master them all.

The first generation of SLA researchers adapted this methodology to second language acquisition by scoring the proportion of morphemes missing from Spanish-speaking children’s description of pictures in English. A typical result from Dulay and Burt (1974) was:

1 the/a

2 -ing

3 plural -s

4 reg. past -ed

5 irreg. past

6 poss. 's

7 3rd person -s

Dulay & Burt (1974) difficulty order for English

This shows that there is indeed a definite sequence for grammatical morphemes in English with learners supplying the/a more than -ing, and -ing in turn more than plural -s etc.

Though similar to the L1 sequence, it is not identical: articles the/a rise from no 5 in the L1 sequence to No 1 in the L2 sequence, while irregular past slips from no 3 to no 5, etc. The differences here may be simply due to the different research methodologies: long-term comparison of data from different occasions versus short-term observation of data from single occasions. VanPatten (1984), for instance, separates the morphemes into NP, V and AUX groups and finds that, within each group, there is no difference between L1 and L2. The main similarity between L1 and L2 is not so much the details of the sequence as the fact that both L1 and L2 acquisition have sequences at all.

Nevertheless the notion of order for grammatical morphemes was replicated in a vast number of studies, reported in Goldschneider & DeKeyser (2001). A subsequent example is a study of Bengali primary school children in the East End of London, which combined one-off testing with comparison of different year-groups and focussed solely on verbal grammatical morphemes (Hannan, 2004). The chronological sequence of acquisition was:

1 -ing
2 past -ed pronounced as /t/
3 past -ed pronounced as /d/
4 irreg. past
5 past participle  

Hannan (2004) The sequence of Bengali children’s acquisition of verbal grammatical morphemes

So far as the comparable morphemes are concerned, this clearly resembles the Dulay and Burt order with -ing coming first but there are differences from both Brown and Dulay and Burt for irregular past went, which is learnt after rather than before regular past -ed. Again the explanation may be methodological with Hannan’s method combining aspects of both Brown’s and Dulay and Burt’s approaches.

Dulay and Burt claimed that the discovery of a common acquisition sequence for L2 learners is 'surely one of the most exciting and significant outcomes of the last decade of second language research' (Dulay & Burt, 1980). Yet was it about grammar at all? Grammar is a system for conveying meaning. Here morphemes are discussed in terms of physical presence in the sentence, not of their grammatical meaning. Does an L2 speaker using the progressive John is going to Paris mean the same as a native or something different, for instance whether the -ing is supposed to convey present or future meaning? The question should surely not be whether the word the occurs in the noun phrase the man but whether it is used properly to show definiteness.

The idea of sequence of acquisition became a driving force in SLA research; what mattered was whether people learnt one aspect of language before another whether sounds (Major 1994), syntactic structures (Zobl & Liceras, 1994), or countless other areas. Usually the method relied on cross-linguistic comparison of different stages rather than on longitudinal develop-ment over time like the ESF project (Klein & Perdue 1997) or Wode’s study of his children (Wode 1981). Chomsky has argued that sequence itself is the product of all sorts of accidental circumstances (Chomsky, 1981): what counts is what people actually know, their destination, not the arbitrary route they follow getting there. Indeed the very concept of stages of acquisition is problematic. According to Ingram (1989), it might be a ‘continuous stage’ referring to ‘a point on a continuum’, a ‘plateau stage’ where change halts for some time, a ‘transition stage’ before change takes off again, and an ‘acceleration stage’ where there is rapid acquisition before reaching a plateau.

Grammar  processing in second language acquisition

Grammar is  not only the knowledge of grammatical relationships in people’s minds but also the processes through which they construct and comprehend sentences. There’s not much point in knowing English is a non-pro-drop language if you can’t work out that the subject of John loves cream cakes is John. 

Interesting as sequences of acquisition may be they don’t matter if we can’t say why things happen in a particular order; dismissing it as natural is no explanation. The Processability Model (Pienemann, 1998), however, bases the sequence of development upon the learner’s increasing capacity to process a new language:

1. using single content words. At the onset the learner only has enough capacity to process one word at a time and so produces utterances consisting of individual content words: Husband. Fly. Plane. Thursday.

2. adding function words. With slightly extra capacity, the learner can start to put function words into the sentence: Husband. Fly. To Paris. On Thursday.

3. making phrases. Given more capacity, the learner can now assemble these words into phrases My husband. Will fly. To Paris. On Thursday .

4 making sentences. With still more capacity the learner can assemble the different parts into a sentence: My husband will fly to Paris on Thursday. This is the point at which the learner has command of the phrase structure and order of the simple sentence.

5 adding subordinate clauses. Finally the learner gains enough capacity to be able to insert subordinate clauses within the sentence: My husband will fly to Paris on Thursday if the airport is open.

Each stage then reflects the learner’ increasing capacity to process crucial elements of the sentence. 

The Competition Model associated with Brian MacWhinney (1987) is also used in SLA research. This claims that the limitations in the human speech processing system mean that languages have to compromise between different ways of processing syntax. How for example do we know something is the subject of the English sentence:
Mr Bean loves Teddy?

Word order. One possibility is to look for the NP that comes in a particular place in the sentence order, in English first, i.e. Mr Bean, in Arabic after the verb, and so on. It’s the position in the sequence that makes an NP the subject.

Animacy. Another possibility is to choose the NP that is alive rather than dead. So in a sentence like Mountains like people speakers of languages with strong animacy cues, such as Japanese or Italian, prefer the animate second NP people as the subject where English speakers still regard mountains as the subject, albeit with a very odd meaning.

Agreement. You may also find the subject by looking for the NP that agrees with the Verb in number, say plural or singular. In The cat likes mice we know the cat is the subject NP because it is singular as is the verb likes but mice is plural.

Case. In many languages the forms of words vary to show their grammatical function, called case. You locate the subject by finding the word in the subjective (nominative) case, as in Amor vincit omnia ‘Love conquers all’ where Amor is in the Latin subjective case, as opposed to amorem (objective), amoris (genitive) etc. It doesn’t matter where the NP comes in the sentence Vincit omnia amor(to the despair of students of Latin verse) provided the case is right. English uses case minimally for deciding the subject with regard to pronouns I/me, they/them, etc.

While these are all potential cues to the subject of the sentence, we can’t handle all four at once. So each language has opted to emphasise one cue over the others, inword order, in Latin case, in Japanese animacy, etc. This does not mean that other cues are completely disregarded, as shown in such lines of poetry as Where the bee sucks there suck I or In my beginning is my end.

SLA research originally looked at this competition between cues in terms of transfer between the L1 and L2. Dutch L1 speakers rely less on word order in English and more  on agreement than native English speakers (Kilborn & Cooreman, 1987). Animacy is more importance for English and Turkish learners of Dutch (Issidorides & Hulstjin, 1992): case is important to Dutch learners of English (McDonald, 1987). More recently the Competition Model has been applied to reverse transfer, that is to say, the effects of the L2 on the L1. This involves comparing monolinguals and L2 users in their first language. Sure enough speakers of Japanese, Greek and Spanish who know English have been influenced by the English word order preference in their first language (Cook et al, 2003), later extended to Arabic, Korean and Chinese (Cook et al, 2007).

Grammar has then been a major theme in SLA research but it has been used in a large number of fairly incompatible ways from a variety of linguistic and psychological perspectives. To many, grammar is the core system of language, providing the bridge between the sounds and meanings of language through the ‘computational system’. To others it is of lesser importance than say vocabulary. Many have seen the concentration on the individual’s knowledge and processing of grammar as a distracting from the foundation of language in social interaction (Firth & Wagner, 1997).

5 How do people learn to write in a second language?

Until recently writing systems have been ignored in SLA research. One reason was the way language teaching since the 1880s has emphasised the spoken language; virtually all teaching methods from the audiolingual to task-based learning have in principle emphasised getting people to speak, not to write. The reasons largely stemmed from an inappropriate analogy with L1 acquisition and a lack of realisation that adult minds have been transformed by literacy (Cook, 2010). Mostly teachers simply followed the accepted wisdom disseminated by linguists since Aristotle that writing is an off-shoot of speech: 'the spoken language is primary and … writing is essentially a means of representing speech in another medium' (Lyons, 1968, 38). Modern writing system research, however, treats written language as parallel to spoken language with its own grammar, vocabulary and independent systems of punctuation, spelling etc (Cook & Bassetti, 2005). Giving writing a secondary role ignores not only its distinctive nature but also its importance in everyday life whether for texting, business contracts, religious works, art, etc.

The  theory and descriptions of writing systems research are at least as complex as those of phonetics and phonology so the following can only sketch its potential for SLA research. The heart of any writing system is what the written signs stand for: the major division is between systems in which signs represent the sounds of the language and those in which they represent meanings; thus the English word cinema represents the sounds /sinema:/, the Chinese character means ‘body’. So an English word like cinema can be read aloud with no idea of what it means, unless you already know that cinema = ‘place for showing films’. The pronunciation of a Chinese character like is irrelevant to understanding it; indeed the spoken form of the word may vary in different dialects.    

Reading the English word dog involves working out how the letters correspond to phonemes – <d> to /d/, <o> to //, <g> to /g/ – the convention in writing systems research is that arrow brackets enclose orthographic symbols: <d> means the letter d, /d/ means its pronunciation. Having worked out <dog> corresponds to /dg/ then you check its meaning in your mental lexicon. Reading the Chinese character  involves matching it against a meaning in your mind; there is no stage for sounds.

Overall then English has a sound-based writing system which relates the letters to the sounds of the language; meaning-based system which relates characters to meanings with little reference to their sounds. In the Mandarin dialect of Chinese the spoken word for dog is in fact ‘gou3’ (using the pinyin transcription system where numbers show tones), in the Hakka dialect 'gieu’, in Min ‘gau4’; each dialect has a separate spoken word. But speakers of any Chinese dialect know what  means – an enormous unifying factor in keeping Chinese as a single language regardless of dialect.

These two overall systems of sound-based and meaning-based writing have many variations. They are both as effective in their own way with particular languages. Characters suit Chinese because it does not have variable word forms – the character  uses a single form. Letters suit English because a word can have many different  forms, dog, dogs, dog’s, dogs’, dogging, dogged, dogger, ..... The problem comes when a person’s first writing system and the second come from different major systems. A Chinese learner of English doesn’t just have to learn English spelling: they have to learn a whole new type of writing system. Vice versa, learning  the Chinese writing system looks very daunting to an English speaker. In practical terms the cost may be high: Chinese university students read English at about one third of the speed of their peers (Haynes & Carr, 1990).

Writing systems also vary in the direction in which the symbols are read. While English and Arabic are both sound-based systems, English is read from left to right in lines from top to bottom: the word dog is read from the left-most letter <d> to the rightmost <g>; the Arabic word كلب  (‘kalib’)  is read from right to left from rightmost كـ (k) to ـلـ (l) to leftmost ب (b). While much Japanese is read from left to right, some is still in traditional vertical columns, read from right to left in columns as well as top to bottom – when English is in columns as on shop signs etc, they are read from left to right. So again it is tricky to go from a writing system with one direction to a writing system with another. Arabic children in England often to try to write Arabic from left to right and -English from right to left; some examples are on the web (Spelling Corpus). The consequences can be profound: writing direction may affect how you organise and see the world (Tversky et al, 1991).

Within sound-based systems, there are also large differences. We see above that Arabic represents the consonants of the word, not the vowels: كلب  tells you how to pronounce /k.l.b/ but you have to work out the vowels for yourself. Hence characteristics of Arabic writing in English may be leaving out vowels or supplying the wrong vowels. Within alphabetic systems, there may be very different letters, say Greek σκύλος (skylos, ‘dog’) or Russian ‘собака’ (‘dog’). Again going from Greek to English or Russian to Greek may be far from easy. The constant emphasis of the internet on the Roman script means that writers of Greek, Chinese and Arabic have all devised ways of representing their languages in Roman letters for use in emails etc. Indeed one of the interference problems for many L2 learners of English is the use of roman transliterations in their first language scripts, such as pinyin for Chinese or romaji for Japanese, which is misunderstood as being in English script.

Sound-based alphabetic writing systems also differ in terms of transparency – a system in which each letter invariably corresponds to a sound and vice versa is transparent; a system in which a letter may correspond to several sounds and vice versa is opaque. Languages that were standardized orthographically comparatively recently, such as Italian and Finnish, are usually the most transparent, those with longer histories such as English have become more opaque as they have aged. The links between sounds and letters often involve complex rules, like those for consonant doubling and silent <e> in English; indeed a residue of English words have to be learnt as one-offs like characters, whether yacht /jt/, lieutenant (British /leftenənt/ or common function words like of /v/ (the only English word where <f> corresponds to /v/). A major difficulty in acquiring a second writing system is learning how its level of transparency relates to the first language one knows already.

This piece has then tried to give some idea of the richness and depth of SLA research today, partly to compensate for the impression the field often gives of being obsessed with minutiae of syntax.  Its worthy attempt to establish itself as an independent discipline has resulted in it being isolated from contemporary developments in language teaching, linguistics and psychology. What it needs to feed back into these disciplines is the message that people who know two languages are not exceptions but the norm, both in terms of their numbers and in terms of the human potential to learn more than one language; the study of say monolingual first language acquisition depends on realising that many children are bilingual and that all children could be bilingual; a second language is not an afterthought but a core element in human existence.

References

Beauvillain, C. & Grainger, J. (1987), 'Accessing interlexical homographs: some limitations of a language-selective access', J. Mem. & Lang., 26, 658-672

Bialystok, E. (2001), Bilingualism in Development: Language, Literacy, and Cognition, Cambridge: Cambridge University Press

Birdsong, D. (2005), ‘Interpreting age effects in second language acquisition’. In J. Kroll & A. de Groot (eds.), Handbook of Bilingualism: Psycholinguistic Approaches, Oxford: Oxford University Press, 109-127

Bloom, P. (2002), How children Learn the meaning of words, Cambridge MA.: MIT Press

Braine, M. (1963), ‘The ontogeny of English phrase structure: the first phase’, Language, 39, 1‑13

Brown, R. (1973), A First Language: The Early Stages, London: Allen and Unwin

Cenoz, J. (2003), ‘The influence of age on the acquisition of English’, in M.P.G. Mayo & M.L.G. Lecumberri (eds), Age and the Acquisition of English as a Foreign Language, Clevedon: Multilingual Matters, 77–93.

Cheydleur, F.D. (1932), 'An experiment in adult learning of French at the Madison Wisconsin Vocational School', Journal of Educational Research, XXVI, 4, 259-275

Chomsky, N. (1959), 'Review of B.F. Skinner Verbal Behavior', Language, 35, 26-58

Chomsky, N. (1981), Lectures on Government and Binding, Dordrecht: Foris

Cook, V.J. (1986), 'Experimental approaches applied to two areas of second language learning research: age and listening-based teaching methods', in V.J. Cook (ed.) Experimental Approaches to Second Language Learning, Oxford: Pergamon, 23-37 online version

Cook, V.J. (1993), Linguistics and Second Language Acquisition, Macmillan

Cook, V.J. (1997), ‘Monolingual bias in second language acquisition research’, Revista Canaria de Estudios Ingleses, 34, 35-50 online version

Cook, V.J. (ed.) (2003), Effects of the Second Language on the First, Clevedon: Multilingual Matters

Cook, V.J. (2009), ‘Multilingual Universal Grammar as the norm’. In I. Leung (ed.) Third Language Acquisition and Universal Grammar, Bristol: Multilingual Matters, 55-70 online version

Cook, V.J. (2010), ‘The relationship between first and second language acquisition revisited’ in E. Macaro (ed.) The Continuum Companion to Second Language Acquisition, Continuum, 137-157 online version

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

Cook, V.J., El-Ebyary, K., Al-Garawi, B., Chang, C-Y., Huang, K-M., Lee, S., Li, G., Mitani, H. & Sieh, Y-C. (2007), Effects of cue processing in the Competition Model on the first language of L2 users. Paper presented at EUROSLA 2007, Newcastle upon Tyne

Cook, V.J., Iarossi, E., Stellakis, N. & Tokumaru, Y. (2003), 'Effects of the second language on the syntactic processing of the first language', in V.J. Cook (ed.), Effects of the Second Language on the First. Clevedon: Multilingual Matters, 214-233

Cook, V.J., Kasai, C. & Sasaki, M. (2005), Syntactic Differences of Bilingual Speakers: The Case Study of Japanese. Poster presented at EUROSLA, Dubrovnik

Corder, S.P. (1971), 'Idiosyncratic dialects and Error Analysis', International Review of Applied Linguistics, 9, 2, 147-159

Davies, A., Criper, C., & Howatt, A. (ed.) (1984), Interlanguage, Edinburgh University Press

De Groot, A. (2002), 'Lexical representation  and lexical processing in the L2 user', in V.J. Cook (ed.), Portraits of the L2 User, Clevedon: Multilingual Matters, 29‑64

Dulay, H. & Burt, M. (1974), ‘Natural sequences in child second language acquisition’, Language Learning, 24, 37-53.

Dulay, H.C. & Burt, M.K. (1980), 'On acquisition orders', in S. Felix (ed.), Second Language Development: Trends and Issues, Tübingen: Narr

Edinburgh Word Association Thesaurus (2011), http://www.eat.rl.ac.uk/

Ellis, R. (1994), The Study of Second Language Acquisition, Oxford: OUP

Ervin-Tripp, S. (2011), ‘Advances in the study of bilingualism: a personal view’, in V.J. Cook & B. Bassetti (eds.), Language and Bilingual Cognition, New York: Psychology Press

Fabbro, F. (1999), The Neurolinguistics of Bilingualism: An Introduction, Aylesbury: Psychology Press

Firth, A. & Wagner, J. (1997), ‘On discourse, communication, and (some) fundamental concepts in SLA research’, Modern Language Journal, 81, 285–300.

Fishman, J.A.( ed.) (1966), Language Loyalty In the United States. The Hague: Mouton

Goetz, P.J. (2003), 'The effects of bilingualism on theory of mind development', Bilingualism, Language and Cognition, 6, 1, 1-15

Goldschneider, J.M. & DeKeyser, R.M. (2001), ‘Explaining the Natural Order of L2 Morpheme Acquisition” in English: a meta-analysis of multiple determinants’, Language Learning, 51, 1, 1-50

Hannan, M. (2004), ‘A study of the development of the English verbal morphemes in the grammar of 4-9 year old Bengali-speaking children in the London borough of Tower Hamlets’, Ph.D., University of Essex

Haynes, M. & Carr, T.H. (1990), ‘Writing system background and second language reading: a component skills analysis of English reading by native-speaking readers of Chinese’, in T.H. Carr & B.A. Levy (eds), Reading and its Development: Component Skills Approaches, San Diego: Academic Press, 375-421

Hawkins, R. & Chen, C. (1997), ‘The partial availability of Universal Grammar in second language acquisition: the ‘failed functional features’ hypothesis’, Second Language Research, 13, 3, 187-226

Hyltenstam, K., & Abrahamsson, N. (2003), ‘Maturational constraints in SLA’, in C. Doughty & M. Long (eds), The Handbook of Second Language Acquisition (pp. 539–588). Oxford: Blackwell

Ingram, D. (1989), First Language Acquisition, Cambridge : Cambridge University Press

Issidorides, D.C. & Hulstijn, J. (1992), ‘Comprehension of grammatically modified and non-modified sentences by second language learners’, Applied Psycholinguistics, 13(2), 147-161

Johnson, J.S. & Newport, E.L. (1989), 'Critical period effects in second lang­uage acquisition: the influence of maturational stage on the acquisition of ESL', Cognitive Psychology, 21, 60-99

Kasper, G. and Kellerman, E. (eds) (1997), Communication strategies: psycholinguistic and sociolinguistic perspectives, London: Longman

Kecskes, I. & Papp, T. (2000), Foreign Language and Mother Tongue. Hillsdale, NJ: Lawrence Erlbaum

Kilborn, K. & Cooreman, A. (1987), 'Sentence interpretation strategies in adult Dutch-English bilinguals', Applied Psycholinguistics, 8, 415-431

Klein, W. & Perdue, C. (1997), ‘The basic variety (or: couldn’t natural languages be much simpler?)’, Second Language Research, 13, 4, 301-347

Krashen, S., Scarcella, R. & Long, M. (eds.) (1982), Child-adult Differences in Second Language Acquisition. Rowley, MA: Newbury House Lado, R. (1964), Language Teaching: A Scientific Approach, McGraw-Hill

Lenneberg, E.H. (1967), Biological Foundations of Language, New York: Wiley

Leopold, W. (1939): Speech Development of a Bilingual Child: A Linguist's Record. Volume I: Vocabulary Growth in the First Two Years. Evanston, Ill: North-western University Press.

Levinson, S. (1996), ‘Relativity in spatial conception and description;, in J.J. Gumperz & S.C. Levinson (Eds). Rethinking Linguistic Relativity (pp. 177-202). Cambridge: Cambridge University Press.

Lorenz, K. (1949), King Solomon's Ring, Routledge

Lucy, J.A. (1992), Grammatical Categories and Cognition: A Case Study of the Linguistic Relativity Hypothesis. Cambridge: Cambridge University Press.

Lyons, J. (1968), Introduction to Theoretical Linguistics. Cambridge: Cambridge University Press

MacWhinney, B. (1987), 'Applying the Competition Model to bilingualism', Applied Psycholinguistics, 8, 315-327

McDonald, J. (1987), 'Sentence interpretation in bilingual speakers of English and Dutch', Applied Psycholinguistics, 8, 379-413

McNeill, D. (1966), ‘Developmental Psycholinguistics’, in: The Genesis of Language: A Psycholinguistic Approach, ed. F. Smith & G.A. Miller, Cambridge, MA: M.I.T. Press

Mennen, I. (2004), ‘Bi-directional interference in the intonation of Dutch speakers of Greek’, Journal of Phonetics, 32, 543-563

Meristo, M., Falkman, K.W., Hjelmquist, E., Tedoldi, M., Surian, L. & Siegal, M. (2007), ‘Language access and theory of mind reasoning: evidence from deaf children in bilingual and oralist environments’, Developmental Psychology, 43, 5, 1156-1169

Munoz, C. (2008), ‘Age-related differences in foreign language learning. Revisiting the empirical evidence’,Inter­national Review of Applied Linguistics, 46 197–220

Myers-Scotton, C. (2006), Multiple Voices: An introduction to Bilingualism. London: Blackwell

Nemser, W. (1971), ‘Approximative systems of foreign language learner’, Inter­national Review of Applied Linguistics 9:115-24

Oxford English Dictionary (1996), third edition, Oxford: OUP

Pavlenko, A. (2003), '"I feel clumsy speaking Russian": L2 influence on L1 in narratives of Russian L2 users of English', in V.J. Cook (ed.) Effects of the Second Language on the First, Clevedon: Multilingual Matters, 32-61

Pienemann, M. (1998), Language Processing and Second-Language Development: Processability Theory, Amsterdam: John Benjamins

Quine, W. V. (1960), Word and Object, Cambridge, MA: The MIT Press.

Roberson, D., Davies I. & Davidoff, J. (2000), ‘Colour categories are not universal: Replications and new evidence from a Stone-age culture’, Journal of Experimental Psychology: General, 129, 369-398

Selinker, L. (1972), ‘Interlanguage’, International Review of Applied Linguistics, X, 3, 209-231

Skinner, B.F. (1957), Verbal Behavior, New York: Appleton-Century-Crofts

Spivey, M.J. & Marian, V. (1999), ‘Cross talk between native and second languages: partial activation of an irrelevant lexicon’, Psychological Science, 10, 181-84

Thomas, M. (1998), ‘Programmatic ahistoricity in second language acquisition research theory’,  . Studies in Second Language Acquisition, 20, 387-405

Thorndike, E.L. (1928), Adult Learning

Tomasello, M. (1999), The Cultural Origins of Human Cognition, Cambridge MA: Harvard University Press

Tversky, B., Kugelmass, S. & Winter, A. (1991), ‘Cross-cultural and developmental trends in graphic productions’, Cognitive Psychology, 23, 4, 515-57

VanPatten, B. (1984), 'Processing strategies and morpheme acquisition', in Eckman, F.R., Bell, L.H. & Nelson, D. (eds.), Universals of Language Acquisition, Rowley, MA: Newbury House, 88-98

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

Werker, J.F. & Tees, R.C. (1984), ‘Cross-language speech perception: evidence for perceptual re-organisation during the first year of life’, Infant Behaviour & Development, 7, 49-63

Whorf, B.L. (1941a/1956), ‘The relation of habitual thought and behavior to language’, reprinted in Carroll, J.B. (ed.), Language, Thought, and Reality: Selected Writings of Benjamin Lee Whorf (pp. 134-159). Cambridge, MA: MIT Press

Wode, H. (1981), Learning a Second Language, Tübingen: Narr

Yelland, G.W., Pollard, J. & Mercuri, A. (1993), ‘The metalinguistic benefits of limited contact with a second language’, Applied Psycholinguistics, 14, 423-444

Zampini, M.L. & Green, K.P. (2001), ‘The voicing contrast in English and Spanish: the relationship between perception and production’, in J. Nicol (ed.), One Mind, Two Languages (pp. 23-48), Oxford: Blackwell

Zobl, H. & Liceras, J. (1994), ‘Functional categories and acquisition orders’, Language Learning, 44, 159-180

FURTHER READING

Introductory books with a similar orientation are:

Cook, V. & Singleton, D. (2014) Key Issues in Second Language Acquisition, Multilingual Matters

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

Cook, V.J. (2016), Second Language Learning and Language Teaching, London: Hodder Education. 5^th edition Book Website

Grosjean, F. (2010), Bilingual: Life and Reality, Cambridge, MA: Harvard U.P.  

This page was formerly at homepage.ntlworld.com/vivian.c (alias Virgin), which is now defunct