se Nua ig Second Language Acquisition . SE D EDITIONI An introduction to second language acquisition research This chapter maps out the territory that wil Tebegins by examining some of the probl acquisition’. By way of concrete il will then present the results of a na will serve to introduce many of th depth later in the book. Next, the Il be covered in the rest of the book. lems in defining ‘second language (L2) stration of La. acquisition, this chapter imber of case studies of L2 learners. These issues that will be considered in greater hapter examines a key issue in the study of 12 acquisition in greater depth-the role ther age plays in the acquisition of an {ea together with an examination of rossinitanan, This issue is central to an understanding of Lz acquisition and has figured strongly in SLA. Finally, the chapter outlines the framework for studying L2 acquisition that informs the structure of the whole book. In this book I shall make a distinction between the terms SLA? and ‘second language acquisition’ or ‘La acquisition’. I shall use the term SLA to refer to the field of study that we will be exploring—the research sat the theories that comprise the discipline, I shall use the term ‘second language acquisition’ to refer to the object of study. Thus, we will be learning whet eq has to say about L2 acquisition. Readers are asked to note, however, that not all researchers use the terms in this way. Some researchers use the term “S| to refer to what Iam calling ‘second language acquisition” Ir is, however, useful, to make a distinction between the two, What is ‘second language acquisition’? Sometimes a distinction is made between a ‘second’ and a ‘third? or even a ‘fourth’ language. However, the term ‘second’ is generally used to refer to any language other than the first language. In one respect this is unfortunate, as the term ‘second’ when applied to some learning settings, such as those i, South Africa involving black learners of English, may be perceived as oppro- brious. In such settings, the term ADDITIONAL LANGUAGE may be both more appropriate and more acceptable, However, given that ‘second’ has become the generally accepted epithet, I will use the term ‘second language acquisi-Language Acquisition 6 The Study of Second - uage after the acquisition of the tion’ to refer to the acquisition of any lang mother tongue. Another distinction that FOREIGN LANGUAGE acquisitio the language plays an institutional that between SECOND and in In the case of second language acquisition, |Vand social role in the community (i.e. it i embers who i i f communication among me functions as a recognized means 0 ; speak some other language as their mother tongue). For example Enelbh aes, second language is learnt in the United States, the United Kingdom, an ae coe Aftica such as Nigeria and Zambia, In contrasts TOTTIBE: Ae age is tings where the language plays no major role in learning takes place in set ae Tearnt only in the classroom. Examples of dd is primari the community ai rie are English learnt in France or Japan. The dis- foreign language learning are n tinction between second and foreign language learning is best treated as a sociolinguistic one rather than a psycholinguistic one. ay is, for the time being at least, we need to keep an open mind as to whether the learning processes found in each are the same or different. Somewhat confusingly, the term ‘second language acquisition’ is used as a superordinate term to cover both types of learning. Arelated distinction is t at is often made is hat between NATURALISTIC and instructed second language ACQUISITION, according to whether the language is learnt through communication that takes place in naturally occurring social situations or through study with the help of ‘guidance’ from reference books or classroom insteuction.! These terms clearly imply psycholinguistic differences. Klein (1986) argued that the learner focuses on communication in naturalistic second language acquisition and thus learns incidentally, whereasin instructed second language acquisition the learner typically focuses on some aspect of the language system. However, again, there is a necd to keep an open mind as to whether the processes of acquisition are the same or different in naturalistic and classroom settings. Themain goalofSLA research isto characterize learners’ underlying know!- edge of the La, i.e. to describe and explain their COMPETENCE. Researchers differ, however, in how they view ‘competence’. Those working in the genera~ tive tradition associated with Chomsky have defined it somewhat narrowly as LINGUISTIC COMPETENCE (i.e. knowledge of the grammar of the L2). Other researchers, however, have adopted a broader perspective, examining how learners acquire COMMUNICATIVE COMPETENCE (i.e. knowledge of both the La. grammar and of how this system is put to use in actual commu- nication). SLA has increasingly adopted the broader ive and, th while this book will consider the grammatical eee emai for example Chapters 2, 3, and 4), ref ‘tical properties of Las in detail (see dae lake sey anes Ge ee ee PRAGMATIC COMPETENCE. (See, for example, Chap eee ae Inorderto study how learnersacquirea secondl: oo) definition of what is meant by the term ‘acsna ieee Cleas operational cquisition’ is needed, Unfortunately,An introduction to second language acquisition research 7 researchers have been unable to agree on such a definition. ‘Acquisition’ can mean several things. A distinction is sometimes made between ‘acquis T10N’ and ‘LEARNING? (for example, Krashen 1981). The former refers to the subconscious process of ‘picking up’ a language through exposure and the latter to the conscious Process of studying it. According to this view, it is possible for learners to “acquire? orto learn’ featuresindependently andatseparate times, Although sucha distinction can have strong face validity—particularly for teachers_it i Problematic, not least because of the difficulty of demonstrating whether the Processes involved are or are not conscious (McLaughlin 1987). In this book the terms ‘acquisition’ and ‘learning’ will be used interchangeably, However, it is useful to make a distinction between IMPLICIT LEARNING and EXPLICIT LEARNING, terms that are widely acceptedin cognitive psychol, oay (see, for example, Eysenck 2001) and which have become increasingly common in current accounts of La acquisition, Implicit learning is typically defined as learning that takes place without either intentionality or aware. ness, It can be investigated by exposing learners to input data, which they are asked to process for meaning, and then investigating (without warning) whether they have acquired any La linguistic properties as a result of the exposure. For example, learners could be asked to read a book and then tested to see if they had acquired any new vocabislaty in the process. (See, for example, Dupuy and Krashen 1993.) However, while such an approach can guard against intentional learning it cannot guarantee that the learning took place without awareness. In fact, researchers disagree as to whether any learning is possible without some degree of awareness. Explicit earning, however, is necessarily a conscious process and is likely to be intentional, Ir can be investigated by giving learners an explicit rule and asking them to apply it to data or by inviting them to try to discover an explicit rule from an array of data provided. These two types of learning clearly involve very different processes and are likely to result in different types of knowledge (i.e. implicit knowledge or explicit knowledge of the Lz). Chapter 9 considers these distinctions in depth. Finally, there are some important methodological issues to consider, First, what and how learners learn an La is not open to ditect inspection, it can only be inferred by examining samples of their performance. SLA researchers have used different kinds of performance to try to investigate La acquis tion, Many analyse the actual utterances that learners produce in speech ot writing. Some try to tap learners’ intuitions about whatis correct or appropri- ate by means of GRAMMATICALITY JUDGEMENT tasks, while others rely on the introspective and retrospective reports that learners provide about their own learning. The question about what kind of data afford the most reliable and valid account of L2 acquisition is a matter of controversy—a matter that is taken up at various points in this book and discussed in some detail in Chapter7. Here it should be noted that different kinds of data will be needed to investigate whether what learners know about the L2 is implicit or explicit.8 The Study of Second Language Acquisition ow we can determine whether a particular feature has been acquited. Some researchers (for example, Bickerton 1982) consider a feature has been acquited when it appears forthe first time, while others (for example, Dulay ‘and Burt 1980) require the learner to use it to some predetermined criterion level of accuracy, usually 90 per cent. Thus, a distinction can be made between acquisition as ‘emergence’ or ‘onset’ and acquisition as ‘accurate use’. Clearly, second language acquisition is a complex, multifaceted phenom- enon and it is not surprising that it has come to mean different things to different people. It does, however, make it very difficult to compare the results of one study with those of another. Conflicting results can be obtained depend- ing on whether the data used consist of learners’ productions, introspections, or intuitions, or whether emergence or accuracy serves as ‘the criterion of Scquisition. It is for this reason that itis important 10 exami carefully the nature of the data used and the way in which acquisition has been measured, when reading reports of actual studies. Second, there is the question of hi Case studies of L2 learners One of the approaches for investigating of individual L2 learners. These case stu Tie. they covered an extended period of time, oftena year or longer) and have te epaved largely on data collected naturalistcally (for example, typically tive speech). They have involved both naturalistic learners and child and adult learners. They have f L2 acquisition as well as SLA has involved the detailed study dies have typically been longitudinal spontaneous communica and instructed language ovided information about the general course o} individual differences among learners. Many of the case studies were conducted in the 1970s or 1980s and then, for a while at least, fell out of fashion, in part because they are time const ing and they do not readily permit generalization (although generalization is possible if based on a number of studies). In more recent years, however, cace studies have again become popular. The European Science Foundation conducted a number of case studies of adult migrant learners of different European languages directed at investigating L2 acquisition cross-linguisti- cally (Klein and Perdue 1992, 1997). More recently still, Han (1998), Lone (2003), and Lardiere (2007) employed case studies of single learners in order to investigate fossilization. The case studies help to identify some of the key issues in SLA and for this reason serve as a useful way of introducing SLA to the reader. I will briefly Feview five studies, four early ones and one very recent, and then discuss the issues that they raise. 7An introduction to second language acquisition research 9 Wong Fillmore’s (1976, 1979) study of five Mexican children ‘This was one of the earliest case studies, ‘Wong Fillmore studied five Mexican children aged from 5.7 to 7.3 years over a nine-month period. They were attending an English-speaking school in California. Each child was paited with a native-English-speaking child and their interactions in a were recorded for one hour a week. The main purpose of investigate how their ability to communicate in Engli TIVE COMPETENCE’) developed. Contrary to her ini Filmore found that the children varied greatly in the developed over the period school playroom expectation, Wong PROFICIENCY they of the study. In order to explain these differences, Wong Fillmore considered and then discounted LANGUAGE APTITUDE (i.e. the special ability learners have for learning an La). Instead, she examined hc Cognitive and social LEARNING sTRATEGIES the children employed in their interactions. These are summarized in Table r.1. Each social strategy was linked to one or more cognitive strategy. For example, S-1 was linked to C-1. By joining a group, the children gained access to L2 input which they were ableto comprehend despite their very limited knowledge of English avs resale of the repetition inherent in the input and contextual clues. S-2 is linked to C- 2. The children’s ability to give the impression they could speak English rested on their use of formulaic expressions such as ‘let's go’, ‘I don’t care’, enock it off, and ‘shaddup your mouth’. These enabled the children to participate in interactions with their native-speaking partners and thus to gain exposure to English. The formulaic expressions also provided the learners with data that they could subsequently analyse and, thereby, discover the underlying parterns of English (i.e. C-3). Wong Fillmore showed how the childzen slowly broke down formulaic sequences into their constituent parts, which were then subsequently used to create new utterances. Successful communication Was achieved both with the help of their native-speaking partner (strateny 5-3) and by making the most of their linguistic resources (strategy C-4), The final cognitive straregy (C-5) involved the children working on the ‘big things’ (for example, word order and lexical expressions) and leaving the ‘detail (for example, grammatical morphemes) until later. This study, then, suggests that L2 acquisition arises out of learners’ ability to participate suecessflly in communication, and shows that childzen vary in their ability to do this end thus develop at different rates. It also suggests that the pattern of acquisition is from ‘wholes’ to ‘parts’ rather than the incremental mastery of specific grammatical features.oc 10. The Study of Second Language Acquisition Social strategies Cognitive strategles G1 Assume what people are saying is $-1 Joina group and act as if you zople. understand what's going on, even if relevant to the situation at hand. Meta-strategy—guess. you don’t. S-2 Give the impression, with a few well- G2 Get some expressions you chosen words, that you speak the understand, and start talking. language. G3 Look for recurring parts in the $-3 Count on your friends for help. formulas you know. | G4 Make the most of what you've got. G5 Work on the big things first: save the | details for later. Table 1.1 Social and cognitive strategies used by five child L2 learners Schumann's (1978b) study of Alberto Schumann (1978b) conducted a ro-month study of the untutored acquisition of a 33-year-old Costa Rican called Alberto, who was living and working in the USA during this period. The data collection involved both naturalistic and elicited speech, The main finding of the analyses that Schumann carried out on the data was that Alberto showed ery little linguistic development over the period of study. Schumann then explored a number of factors that might explain this, concluding that the most likely was Alberto’s social and psychological distance from native speakers, especially the former. Alberto failed to acculturate to US society, remaining locked into an immigrant worker community, and thus experienced a ‘bad learning situation’ that restricted him to the ‘communicative function’ served by language at the expense of the ‘integrative’ and ‘expressive’ functions. as a result, Alberto’s English remained pidginized, Table 1.2 summarizes this case study. Learner's ‘Alberto was a 33-year-old lower-middle-class Costa Rican, He was working background ina factory in the USA with other non-native speakers of English. He lived within a small Spanish-speaking minority in a Portuguese area, Thus he experienced high enclosure as a member of a highly cohesive group of ‘Spanish speakers, which very likely aroused negative attitudes on the part of native speakers, Purpose of the study | The study of Alberto was part ofa larger study by Cancino, Rosansky, ‘and Schumann (19786) that investigated the acquisition of English by five Spanish-speaking L2 learners in the USA. The overall purpose was to document the pattern of acquisition of specific grammatical structures (for example, negatives and interrogatives). As such, it was informed by similar studies of the L1 acquisition of English,An introduction to second language acquisition research xx Data collection ‘Three types of audio recorded data were collected: 1. Spontaneous speech recordings of conversations with the researcher, 2 Experimental elicitations (for example, transforming a positive into a negative utterance). 3. Pre-planned sociolinguistic interactions involving taking Alberto to parties, restaurants, and sports events in order to collect speech in varied natural situations. Analysis The analysis focused on specific grammatical structures (for example, negatives, interrogatives, and auxiliary verbs). Itinvolved an approach Currently referred toas ‘frequency analysis’ (Els and Barkhuizen 2005), This entails identifying utterances where the learner attempts to perform the structure in question, describing the various non-target and target devices the learner uses and calculating the frequency of each device at diferent times within the total period of study. This enables the researcher to identify the dominant device used at different times and thereby to plot the pattern of development for the structure. Main findings Alberto showed very little linguistic development during the period of study—for example, his negatives were predominantly of the ‘no + verb" type (no use television’), a large proportion of his interrogatives were ot inverted (‘Where you get that?’), and only one auxiliary (copula ‘s') was mastered. In short, Alberto's English was ‘pidginized’ (i.e. manifested the same features as pidgin languages). Schumann also reported his attempt to teach Alberto negatives, noting that although this had some effect on his production of elicited negatives it had no effect on his production of spontaneous negatives. ‘Schumann considered a number of factors that could explain Alberto’s, lack of development. He dismissed ability and age as factors and instead turned to a consideration of the factors that could cause pidginization of languages. He examined a range of factors that promote or inhibit social solidarity between two groups (for example, social dominance, acculturation, enclosure, cohesiveness, sizeof the L2 group, and ethnic attitudes) and concluded that Alberto experienced a ‘bad learning situation’ as a result of social distance from native-speaking groups. ‘Schumann also suggested that Alberto experienced substantial psychological distance (for example, displayed negative attitudes and little motivation to learn English). Table 1.2 Schumann's (1978b) study of Alberto Schmidt’s (1983) study of Wes Like Wong Fillmore and Schumann, Schmidt investigated a learner who was acquiring English naturalistically through interacting with other speakers of the language. Schmidt’s study of Wes, however, differs from the previ- ous two case studies in that it was not exclusively focused on how linguistic competence is acquired, Schmidt took as his starting point Canale’s (1983) model of communicative competence. This distinguishes four components:Pr 12. The Study of Second Language Acquisition sociolinguistic competence (i.e. the ability to use lan- discourse competence (i.e. the ability to participate in coherent and cohesive conversations) on srstegiccomperence (ie. the ability to deal with communication breakdown). Schmidt show ; that development of these abilities proceeded separately in Wes, That i although, like Albert, he showed litte linguistic development, he p opts considerably in the other aspects of communicative compet - — ‘ also questioned the importance Schumann attached to social distance in explaining Wes' file co acquire the La grammar as Wes clealy enjoyed regular contact with native speakers of English but nevertheless mani ese a pattern of pidginization similar to that of Alberto. Schmidt’s case study is summarized in Table 1.3. linguistic competence, guage in socially appropriate ways), “Wes is a 33-year-old Japanese learner of L2 English who left school at 15 years old and thus had had very little experience of formal instruction. He was a successful artist. He divided his time between living in Hawaii and Japan, spending increasing amounts of time in the former. He mixed predominantly with English speakers in Hawaii and thus experienced very little social distance from native speakers of English. Learner's background Purpose ofthe | Schmidt sought to investigate to what extent Wes’ acculturation study to American society could explain his development of communicative competence over a three-year period. This case study also addresses the extent to which Wes was able to acquire English through natural interaction. Datacollection | Data were collected over a three-year period predominantly by asking Wes to make one-hour tape recordings concerning business and daily activities on his visits back to Japan. These ‘monologues were supplemented with some recordings of informal conversations with native speakers. In addition, Wes completed Scarcella's (1979) test of knowledge of verbal routines such as those involving apologies. Analysis ‘Schmidt transcribed the spoken data and then analysed the transcripts in terms of Canale’s four aspects of communicative competence. Linguistic competence was investigated by examining the extent to which Wes used 4 number of grammatical morphemes. The other aspects of communicative ‘competence were analysed qualitatively and illustrated by examples taken from the transcripts,‘An introduction to second language acquisition research 13 Main findings 1. Wes' pronunciation was good (especially his intonation) but his grammar hardly developed over the three-year period. Of rine grammatical morphemes (for example, V-ing and 3rd persons), only three reached the 90 per cent criterion level of accuracy deemed to indicate ‘acquisition’, 2 Initially, Wes’ directives relied extensively on formulaic expressions (for example, ‘Can |have a__7'). These were not available for productive use (i.e. were unanalysed). By the end of the three-year period, gross errors in his use of directives had been eliminated. Also, his English utterances were largely socially appropriate, although sometimes idiosyncratic. 3. Wes demonstrated a high level of discourse competence, which compensated for his linguistic weaknesses. This was the area that showed the greatest improvement. By the end of the period, Wes demonstrated improved comprehen- sibility in his narratives and had become a good conversationalist in many respects, able to nominate topics freely. 4 Wes was able to repair communication breakdowns despite his limited linguistic competence. He made effective use of communication strategies such as paraphrase (for example, his use of ‘money-gi!’ for ‘prostitute’ }sHe used adverbials to compensate for his lack of a verb tense system. However, he rarely repaired his utterances as a result of interactional feedback. Schmidt suggests that Wes operated on the principle thatiit was the responsibility of native speakers to make an effort to understand him. Conclusion Wes could be considered a ‘good’ language learner in terms of his overall communication skills but as a ‘poor’ learner in terms of his linguistic ability. His failure to develop grammatically cannot be explained in terms of social distance, Schmidt suggests that itis better explained in terms of psychological factors (for example, Wes’ personality) which were responsible for his failure to attend to linguistic form. This study also shows the partial independence of grammatical competence from other aspects of communicative competence. Table 1.3. Summary of Schmidt's (1983) study of Wes Ellis’ (1984a, 1992) study of three classroom learners The next case study we will consider is my own. I investigated three children aged 10 to r3 years, who were learning English in a Language Centre in an outer suburb of London, The 10-year-old was Portuguese while the other two14 The Study of Second Language Acquisition came from Pakistan. Like Wong Fillmore’s learners, they d data by sitting they produced (a brother and sister) were all complete beginners at the start of the study. I collecte in their English classes and noting down all the utterances together with contextual information relating to the function and audience of their utterances. My initial purpose in conducting these case studies was to examine whether the pattern of development evident in naturalistic learners was the same as or different from the pattern found in the classroom learners. In other words, I wanted to know whether the instructional setting influenced the way in which the children learned English. Like Wong Fillmore’s learners, the three children all made extensive use of formulaic sequences as a means of performing the communicative acts that were needed of them in a classroom where English was not only the target of instruction but also the medium of instruction. Table 1.4 gives examples of the different types of formulas the learners produced. Like Wong Fillmore, Inoted a number of ways in which these formulas were subsequently devel- oped by the three learners. For example, the learners were subsequently able to modify and extend the ‘I don’t know’ formula by substituting other verbs (for example, ‘I don’t understand’), changing the subject (for example, ‘You don’t know’) and adding a constituent (for example, ‘I don’t know this one’). I suggested that this indicated that the learners were slowly unpackaging the formulas, releasing their constituents for creative language use. ‘Type of formula (based on Yorio 1980) | Examples 1 Situation formulas (Le. those Finished (spoken after completing a task). associated with a specificsituation) | Very good (a self-congratulation). What's the time? (spoken as break-time ‘was approaching). — 2. Stylistic formulas (i.e. those associated | Can! have a__? (requesting goods from with a specific style of speaking) the teacher or another pupil). 3. Ceremonial formulas (i.e. those How are you? (greeting) associated with ritualistic interactions) | Good morning (greeting). Excuse me miss (attracting teacher's attention). Oh my God (exclamation). 4 Gambits (ie. those used to organize _| This one or this one? (identifying nature interactions and activities) of classroom task). What's this? (identifying an object). I don’t know (referring to either lack of knowledge or inability to respond). That's all right (confirming a course of action). 5 Euphemisms No examples observed. Table 1.4 Formulas observed in the speech of three classroom learners (from Ellis 1984a: 69)An introduction to second language acquisition research 15 Another feature of the three children’s language was what I called ‘semantic simplification’. This was only evident in the learners’ creative communicative specch (i.e. not in utterances that were modelled for them by the teacher or were formulaic). It entails omitting constituents that perform semantic roles that would normally be encoded by a native speaker. For example, one of the learners produced the utterance ‘Sir, sit, pencil” after the teacher had taken his pencil, omitting the semantic roles of ‘agent’ (‘you’) and action (‘take’) and encoding only the object (‘pencil’). Such utterances were readily understood because the missing constituents could easily be recovered with the help of contextual clues. Over time such simplification diminished in the children’s speech. The use of formulaic sequences and semantic simplification has also been noted in naturalistic learners, suggesting that in these respects these three classroom learners were acquiring English in very similar ways to learners like Wes and Alberto. Further evidence for the similarity between naturalistic and classroom acquisition was obtained by comparing the stages of acquisi- tion evident in the acquisition of grammatical structures such as negatives and interrogatives. The developmental profiles of the three children were very similar to each other and showed a striking similarity to that reported for naturalistic learners. I concluded that the children were relying largely on natural processing mechanisms to acquire English. ‘The main focus of these analyses was the learners’ linguistic competence. However, later (Ellis 1992), I returned to the data I had collected to examine the pragmatic features of the three learners’ use of requests, documenting how their repertoire of strategies for performing this speech act gradually increased but was still quite limited after nearly two years. Further informa- tion about this later analysis can be found in Chapter 5. Lardiere’s (2007) study of Patty Patty was born in Indonesia in 19 53 but was of Chinese origin. She left Indo- nesia for China in 1969 and lived there for two years before moving to Hong Kong, where English became the primary language of instruction, When she finished high school she worked in an import-export company in Hong Kong, rarely speaking English. She arrived in the United States in 1976 at the age of 22. She lived with her Vietnamese fiancé’s family and began college-level study and, a little later, took a waitressing job. In 1985 she separated from her husband and began to live by herself but married again in 1989 to a native English speaker. Lardiere collected naturalistic production data from Patty on three occasions. Recording 1 was made in 1986 when Patty had been living in the United States for about 10 years. Recordings 2 and 3 were made two months apart, in 1995. Written samples, mainly from email messages, were also collected. The data for this case study, therefore, spanned a much longer period than for any of the other case studies reported above.16 The Study of Second Language Acquisition . in Patty’s English Lardiere investigated a number of grammatical fee ia ie = speech and writing, including finiteness (for example, morphological markcrs of tense/aspect and agreement), the acquisition of Pe eh meen ae order and movement (for example, WH-movement in English question for mation), and nominal phrases (for example, possessive Pronouns 3 ee eres marking). This constitutes perhaps the fee a ace learner’s acquisition of grammar currently available. Patty had nei eeaee ie grammar by the oc ie oa However there wasa clear difference inher ability f0 use morphophonologi- cal aspects of English grammar correctly and her ability to employ Eng! _ syntactical constructions correctly, as shown in Table 1.5. Whereas the former continued to be problematic, as evidenced in the frequent omission or overuse of inflectional markers on verbs and nouns, the latter were ‘surpris- ingly target-like’ (p. 204). A further finding was that Patty’s written English was more accurate than her spoken English. The results of this case study suggest the need for an explanation of three points. (x) Why did Patty fail to achieve accuracy in the use of morphological features despite the facts that she had been in the United States for ro years and that she had experienced favourable acquisition circumstances (i.e. immersion, acculturation, a high level of educational attainment and professional and personal success in her target-language community)? (2) Why was there such a difference between the morphological and syntactical aspects of Patty’s acquisition of English grammar? (3) Why was her written English more accurate than her spoken English? Morphophonological aspects Syntactical aspects Many morphological features werenon- | Most syntactic and some morphological target-like: features were now target-like: © omission of regular past-tense * perfect knowledge of pronominal case marking marking SE eller te || a nas peciect ccd panenien forms Pronouns and demonstratives © Past participle not inflected | © Proficient use of definite and indefinite © omission and overuse of the articles Progressive -ing form * accurate placement of adverbs © omission of plural marking * Tobust relative clause formation © Omission of possessive marking, * accurate WH-movement @PPropriate stranding of prepositions Inversion/‘do support in questions Perfect use of overt Subjects (as opposed to null subjects). Table 1.5 Morphopbonological and syntactical aspects of Patty’s grammarAn introduction to second language acquisition research 17 Lardiere advances a number of answers to these questions. First, Lardiere suggests that, although Patty's Lx may have played a role in her acquisition of English gcammar, it cannot fully account for the findings of the case study. Instead, Lardiere suggests that Patty’s continued variability in the use of morphological features, despite highly favourable conditions for learning, might be explained by the age factor: ‘Maturational changes in the learner may influence the ability to perceive or statistically calculate the criterial percentage of information needed for deriving categorical generalization, resulting in variability. (p. 23 5) Thus Patty failed to fully acquire ‘categorical generalization’ (i.e. fullaccuracy in the use of features like plural -s) because she did not begin to learn English until her adolescent years, To account for Patty’s differential acquisition of morphophonology and syntax, Lardiere appeals to the notion of modularity in grammatical systems, suggesting that some domains (i.e. morphophonol- ogy) are inherently more susceptible to failure than others, possibly because they are not governed by the ‘language faculty’ that some SLA researchers have claimed governs L2 acquisition as it does Lx acquisition. Lardiere does not address the difference in accuracy in Patty’s spoken and written English. One possible explanation for this is that Patty possesses metalinguistic knowledge of English grammar and was able to make use of this more easily when writing than when speaking. Issues raised by the case studies These case studies raise a number of important issues that will be revisited in subsequent chapters. They will be briefly considered here. 1 Methods of data collection Alllfive studies were based largely on spontaneous speech. In three of the studies (Wong Fillmore, Schumann, and Ellis) this took the form of conversational data. In Schmidt's study, however, the data consisted of oral monologues. Two of the studies (Schumann and Ellis) also collected data reflecting more planned language use, but in both care was taken to keep the unplanned and planned data separate. One study (Lardiere) also collected samples of written language. SLA researchers have generally held that learners’ spontaneous communicative speech constitutes the best data for investigating acquisition, but, as we will see in later chapters, they have frequently resorted to elicited data of various kinds. 2 Focus of enquiry The main focus of the four studies was learners’ acquired linguistic compe- tence (in particular, grammar). This is a feature of many early SLA studies but continues to attract interest today, as Lardiere’s study demonstrates. The early case studies also presaged the later interest that SLA researchers have18. The Study of Second Language Acquisition e, Schmidt examined four aspects of com- more and Ellis examined how formulaic icate at a time when they had very limited asa platform for the acquisition of gram- shown in other aspects of languag municative competence, Wong Fill speech enabled learners to communi linguistic resources and also served matical knowledge. 3 The pattern of La development _ All the studies showed that learners apparently follow a very similar pattern of development, For example, all the studies documented how learners’ early speech is ‘pidginized’ or ‘semantically simplified’ and that subsequent acqui- sition of grammatical structures is very gradual, manifesting common stages of development. The existence of universal properties of L2 acquisition that these studies testify to constitutes a key finding of SLA, which has informed a number of theories of L2 acquisition. 4 The acquisition of the morphological and syntactical aspects of an Lz The studies also indicate that La learners experience greater difficulty in acquiring the morphological (inflectional) properties of an Lz than they do the syntactical properties. Failure in acquisition is more likely in the former than the latter. 5 The role of instruction None of these studies systematically investigated the effects of instruction on Lzacquisition. However, Schumann reported thathis attemptto teach Alberto negatives had no effect on his spontaneously produced negatives. Also, Ellis noted clear discrepancies between his learners’ spontaneous communicative speech and their modelled speech; and Lardiere noticed differences in her learner’s speech and writing. The extent to which formal instruction can influence the course of L2 acquisition is a major theme in SLA. (See Chapters rg and 16.) 6 Variation in the rate of development While learners appear to manifest a very similar pattern of development, they also clearly vary in both how rapidly they develop and also in what aspects of the La they show most development in, This variation is evident in both adults and children—Alberto was chosen for study because his development was so much slower than that of other adult learners in the Cancino et al. (1978) study, and both Wong Fillmore and Elli i rate of development of the children they oe re 5 ight be because the studi I fe i . t ‘udies were not long enough, aithough Schmidt's study of Wes covered theee years and Lardiere’s study of Patty, ten years, Itis % . It is possible, then, that La learners * (i stop developing before they achieve full mastery), Fossil, oe issue[An introduction to second language acquisition research 19 in SLA and will be considered more fully in the following section of this chapter. 7 Explaining L2 acquisition There are two aspects of L2 acquisition that require explanation; the uni- versal aspects (i.e. the general pattern of development evident in all learners) and the variable aspects (i.e. differences in the rate of development and in ultimate achievement). The universal aspects might be explained in terms of a common set of social and cognitive strategies (Wong Fillmore), or in terms of the unpackaging of formulaic sequences, or, perhaps, as the working of an innate ‘language faculty’. These explanations (together with others) will be explored in subsequent chapters. The variable aspects can be potentially explained by a host of factors. These are of a social nature (for example, Schumann’s notion of ‘social distance’), of a more personal, psychological nature (for example, Wes’ orientation towards the functional aspect of language at the expense of the formal aspect) and of a linguistic nature (for example, Lardiere’s. proposal that some aspects of grammar are governed by the learner’s ‘language faculty’ whereas others lie outside it). Age may also be a factor, as Lardiere has suggested. Age is potentially of considerable significance to theory building in SLA and has attracted considerable atten- tion from researchers. For this reason it will be considered more fully below. ‘The case studies I have considered are all of learners of English. This reflects a bias that has continued in SLA; the majority of studies that will be considered in the following chapters have examined the acquisition of L2 English. However, interest in other languages has increased over the years and I will be reporting studies of both European languages (for example, German and French) and of Asian languages (for example, Chinese, Korean, and Japanese). The role of age in L2 acquisition In this section I will consider an issue that has figured consistently in dis- cussions of La acquisition—the role that age plays in L2 acquisition and, in particular, whether there is a critical period during which the acquisition of fall target-language competence is possible and after which it becomes impos- sible. Both the age issue and fossilization are topics of continued debate in SLA. The role that age plays in Lz acquisition has attracted the attention of researchers since the inception of SLA as a field of study. Krashen, Long, and Scarcella (1979: 161) reviewed a number of the earlier studies, reaching three conclusions: 1 Adults proceed through the early stages of syntactic and morphological development faster than children (where time and exposure are held con- stant).20. The Study of Second Language Acquisition i ildren (again, in the early 2 Older children acquire faster than younger chi A, stages of syntactic and morphological development where time and expo- are held constant). _ 3 neanes ‘who begin natural exposure to a second language during child- hood achieve higher second-language proficiency than those beginning as adults. It should be noted that these conclusions do not entirely accord with the lay belief that ‘younger is better’ where L2 learning is concerned. While itis true that learners who start learning in childhood often achieve higher levels of ultimate proficiency than learners who start later (i.e. in adolescence or as adults), the research indicates that in the earlier stages of L2 acquisition, older learners outperform younger learners, especially where knowledge of grammar is concerned. In discussing the role of age, then, it is important to distinguish the effects that the ‘age of onset’ has on the rate of acquisition and on ultimate achievement. Accordingly, I will first consider age in relation to rate of learning and then address ultimate achievement and the related issue of whether there is a critical period for acquiring language. The effects of age on the rate of second language learning The study most often cited in support of Krashien ef al.'s conclusions regarding the effect of age on the rate of L2 acquisition is Snow and Hoefnagel-Hohle (1978). This study investigated the naturalistic acquisition of Dutch by 8- to ro-year-old English-speaking children, 12~ to 5-year-old adolescents, and adults over a ro-month period, The learners’ proficiency was measured on three separate occasions (after three months, after six months, and at the end of the study). With regard to morphology and syntax the adolescents did best, followed by the adults, with the children last. However, there were only small differences in pronunciation, and the grammar differences diminished over time as the children began to catch up, The Barcelona Age Factor Project (Mufioz 2006) investigated the effects of age of onset in an instructed setting, This project examined the acquisition of English by classroom learners of English in Catalonia (Spain), comparing students who began their study at the age of 8, 11, and r4 and controlling for exposure to English outside the classroom, Data from a battery of tests providing measures of both implicit and explicit types of knowledge were ‘cted on three occasions—after 200 hundred hours of instruction, 416 fours and726hours. The main finding was that the older learners prowrested ster than the younger learners. In contrast to Snow and Hloefnagel FiGhle’s findings, the younger learners did not E catch up over time, babl} se the number of hours available for learning in this insttnet orien no insufficient to enable them to do so. However, as in Snow nd Hi fn - Hoble’s study, there was evidence that age had differential ‘effect cae acquisition of different aspects of the Lz. Thus, the advantage for the older = irAn introduction to second language acquisition research 21 learners was strong and durable on measures of grammar and least evident in the case of measures of speech perception, listening comprehension, and oral fluency. In the latter measures, no statistically significant differences between the young and older starters were evident on the final occasion. Similar results for instructed learners had been obtained by Burstall (1975) in a study that compared students who started learning La French at either the primary- or secondary-school level in England. Experimental studies have also shown that adults outperform children in the short term, For example, Olsen and Samuels (1973) found that American English-speaking adolescents and adults performed significantly better than children after ten 15-25 minute German pronunciation sessions. Cochrane (1980) investigated the ability of 54 Japanese children and 24 adults to dis- criminate English /r/ and /V. The average length of naturalistic exposure was calculated as 245 hours for the adults and 193 for the children (i.e. relatively little). Nevertheless, the children outperformed the adults. However, in a follow-up experiment in which the two groups were taught the phonemic distinction, the adults benefited while the children did not. Overall, the experimental research indicates that in formal learning situations adults seem to do better than children, even in this area of learning (pronunciation) that most favours children. Overall, then, the research supports the earlier conclusion of Krashen et al.—learners who start learning an L2 in adolescence or as adults learn more rapidly than those who start in childhood. However, to fully understand the results of the research it is useful to distinguish the effects of age on the rate of acquisition in terms of the distinction between implicit and explicit learning. (See DeKeyser 2000.) The greater cognitive development of older learners is advantageous where explicit learning is concerned, as the results of the experimental studies show. Late-starting learners do better in tests that tap into explicit knowledge of the La language (for example, traditional grammar tests). In contrast, they do not necessarily outperform early-starters in the long-term where implicit learning is involved. Implicit learning is a slow and gradual process and thus the advantage that younger learners may have where this kind of learning is involved will not show up until after many hours of exposure to the L2—that is, typically only in a naturalistic setting, Thus, in the case of the naturalistic learners in Snow and Hoefnagel-Héhle’s study, the exposure was sufficient to enable the younger learners to catch up with the older learners. But in the case of the Barcelona Age Factor Project, the exposure in the instructed setting was very limited, éven after several years of schooling, with the result that the early-start learners failed to catch up. The difference between implicit and explicit learning can also explain why the advantage for older learners in some aspects of language (for example, listening ability and pronunciation) is not—or at least, is less—evident. These are aspects that cannot easily be mastered through explicit learning—they develop primarily through exposure and practice.22 The Study of Second Language Acquisition The effects of age on ultimate achievement art learning as children have an advan- fe ii ho ste Te would follow thar ifearnety Mere the acquisition of implicit knowledge i a tage in the long term, especially where is concerned, then ultimately they can be expected to achieve higher levels of proficiency. A number of studies have investigated wherher this is the case—allowing for a minimum of five years of exposure in the case of natu- ralistic learners. 'As Singleton (1989) noted, school-based studies cannot address this issue, as formal learning environments typically do not provide learners with the amount of exposure needed for cheage advantage of young learners to emerge. id in the school-based studies involving immersion ‘An exception can be foun education, In immersion programmes students are taught a range of school Subjects through the medium of the L2 and thus receive far greater exposure than is the case in traditional foreign language programmes. (See Chapter 7.) Harley (1986), for example, investigated the levels of attainment of children in French bilingual programmes in Canada. She focused on the learners’ acquisition of the French verb system, obtaining data from interviews, a story- repetition task, and a translation task. She compared early and late immersion students after both had received 1,000 hours of instruction. Neither group had acquired full control of the verb system. The older students demonstrated greater overall control when hours of exposure were controlled for, but the carly immersion group showed higher levels of attainment at the end of their schooling. However, this result may reflect the additional number of years’ instruction they had received rather than starting age. Naturalistic learners who start as children achieve a more native-like accent than those who start as adolescents or adults. Oyama (1976) investigated 60 male immigrants who had entered the United States at ages ranging from 6-20 years and had been resident there for between 5 and 18 years. She asked two adult native speakers to judge the nativeness of the learners’ accents in two 45-second extracts ¢aken from performance on a reading-aloud task and a free-speech task, Oyama reported a very sttong effect for ‘age of acrival’ but almost no effect for ‘number of years’ in the United States. She found that the youngest arrivals performed in tle same range as native-speaker controls. Other studies which have investigated the effects of age on pronunciation (see Flege 1999 for a review of these studies) support the younger-is-better saat aaa ee to suggest that ultimate attainment in es productoflearners’ starting age. However acheiien one ces d learners who begin as adults are capable of studies Bar shee accent: Bongaerss (1999), for example, reported three needed c pier va er of advanced late learners of L2 English proRuneation® y native speakers on a variety of measures of Similar results have Pe Patkowski’s (r980) nt a a een OL eae ed immigrants to the United StatesAn introduction to second language acquisition research 23 found that learners who had entered the United States before the age of 15 were rated as more syntactically proficient than learners who had entered after 15, Furthermore, there was a marked difference in the distribution of the scores (based on native speakers’ ratings on a five-point scale) for the two groups. The adult group’s scores were evenly distributed, with the majority at midpoints on the rating scale. The child group’s scores clustered at the high end of the rating scale, with 29 out of 33 achieving a rating of 4+ or 5. Patkowski also investigated the effects of number of years spent in the United States, amount of informal exposure to English, and amount of formal instruction. Only the amount of informal exposure had any significant effect, and even this was negligible in comparison with the age factor. Patowski’s findings are confirmed by Johnson and Newport's (1989) study of 46 native Koreans and Chinese who had arrived in the United States between the ages of 3 and 39, half before the age of 1 and half after 17. The subjects were asked to judge the grammaticality of 276 spoken sentences, about half of which ‘were grammatical. Overall the correlation between age at arrival and judge- ment scores was -0.77 (i.e. the older the learners were on arrival, the lower their scores). Neither the number of years of exposure to English beyond five, nor the amount of classroom instruction was related to the grammaticality judgement scores and, although an effect for ‘identification with American culture’ was found, this was much weaker than that for age. However, as we will see below, when we turn to an examination of the CRITICAL PERIOD HYPOTHESTS, Johnson and Newport’s study has been subject to considerable criticism. In his summary of a wide range of research investigating the effects of age on Lz acquisition, Singleton (r989) wrote: Concerning the hypothesis that those who begin learning a second language in childhood in the long run generally achieve higher levels of proficiency than those who begin in later life, one can say that there is some good supportive evidence and that there is no actual counter evidence. (1989: 137) This is one of the few definite conclusions that Singleton felt able to reach in his comprehensive survey of age-related research and it has withstood the test of time. It is worthwhile noting, however, that this conclusion may not hold true for the acquisition of L2 literacy skills. Cummins and Nakajima (1987) examined the acquisition of reading and writing skills by 273 Japanese chil- dren in Grades 2-8 in Toronto. They found that the older the students were on arrival in Canada, the more likely they were to have strong La reading skills and, toa lesser extent, better La writing skills. The explanation Cummins and Nakajima offer is that the older learners benefited from prior literacy experi- ence in Japanese. (See the discussion of the INTERDEPENDENCY PRINCIPLE in Chapter 7.)24 The Study of Second Language Acquisition The Critical Period Hypothesis The criticat periop HypoTHests (CPH, span of years during which language learning can take place naturally and effortlessly, and after which it is not possible to be completely successful. Penfield and Roberts (1959), for example, argued that the optimum period for language acquisition falls within the first ten years of life, when the brain retains its plasticity. Initially, this period was equated with the period taken for lateralivation of the language function to the left side of the brain to be completed, Work on children and adults who had experienced braininjuries or operations indicated that damage to the left hemisphere was rapidly repaired ie che case of children but not adults (Lenneberg 1967). There is, however, no clear consensus on when the ‘window of opportunity’ for language learning ends, Singleton (2005), in a survey of the literature that has addressed this issue, reported claims ranging from near birth to late adolescence. Also, it has kecome clear that, ifthere is a critical period, this varies depending on the aspect of language under examination (see, for example, Seliger 1978) with the end point coming earlier for pronunciation than for grammar, “There are different conceptualizations of the critical period. One is that the end of the critical period signals the point at which performance begins That is, up to that point, age has no or very little effect on La ion. Birdsong, (2006a) refers to this as the ‘unconventional notion of the critical period’ (p. 19). The conventional view is that the end of the critical period constitutes the point at which the decline in performance as a result of age ceases. According to this version of the hypothesis, maturation signals the end rather than the beginning of age effects as, once past the critical age, acquisition is blocked for all learners irrespective of whether they are just past it or many years past it. ‘What both conceptualizations have in common is the notion of a disconti- nuity in learning; thats, after a certain age, the pattern of learning changes, as Patkowski found for grammar in the studies referred to above. Johnson and Newport (1989) interpreted the results oftheir study as evidence for the CPH (conventional version). They endeavoured to show that there was a sharp breakin the effects evident for age at the critical period. In the case of the early starters prior to.age 1 5),there wasa gradual decline in performance according to age. However, in the case of the late starters (after age 17), the relationship between age and performance was essentially random. However, as Bialystok and Hakuta (1994) showed, if the cut-off for the end of the critical period was ee ae a es age of the older group in Johnson and Newport’s = ae Hoban performance. In other words, when the data ea there was no evidence of a clear discontinuity. g (2006a), in a review of a number of studies, including Johnson and Newport, concluded that there is'no consistent evidence of 2 discontinuity effect. Bialystok and Fakuta’s (r999) investigation of the English-language P y of 24,903 speakers of Chinese and 38,787 speakers of Spanish, ) claims that there is a fixedAn introduction to second language acquisition research 25 all of whom had been resident in the US for a minimum of ten years, found a clear linear effect for age of arrival for both sets of speakers and ‘nothing special about the age range before puberty’ (p. 175). A second way of assessing whether learners can achieve native-speaker levels in an L2 is to see whether they are able to recognize spoken or written ‘accents’ in the same way as native speakers. Scovel (1981) asked four groups of judges (adult native speakers, child native speakers, adult non-native speakers, and adult aphasics) to rate speech samples and written pieces pro- duced by a mixture of native and non-native speakers. He found that even the most advanced non-native speakers achieved an accuracy rate of only 77 per cent, which was about the same as the child native speakers (73 per cent) but less than the adult native speakers (95 per cent) and even the aphasic native speakers (85 per cent). This study suggests that even very advanced learners lack some of the linguistic abilities of native speakers. Another approach to investigating the CPH involves investigating learners who began learning as adults and are now very advanced to see if they can perform similarly to native speakers. Two frequently cited studies exem- plify this approach. Coppetiers (1987) tested 21 highly proficient speakers of French, all of whom had begun learning as adults, and compared their performance on a grammaticality judgement task with that of 20 native speakers. Coppetiers noted that it was not possible to distinguish the two groups by the mistakes they made, their choice of lexis, or grammatical con- structions, and six of the subjects were also described as having no traces of a foreign accent. The results of the grammaticality judgement test, however, showed clear differences between the two groups, suggesting that despite the native-like performance of the learners in language production, their gram- matical competence differed from that of native speakers. However, Birdsong (1992) identified ‘numerous procedural and methodological features of the Coppetiers study that compromise its conclusions’ (1992:711)-Hisreplication of Coppetiers’ study produced very different results, Birdsong administered a grammaticality judgement test to 20 English-speaking learners of L2 French, who were near-native in their oral ability, and to 20 native speakers of French. ‘The study was motivated by Long’s challenge to researchers to investigate “whether the very best learners actually have native-like competence’ (t990a: 281). Contrary to Coppetiers, Birdsong found no evidence of.any dramatic differences in the judgements of the non-native speakers and native speakers. ‘A number of the non-native speakers performed in the same range as the native speakers on the grammaticality judgement test. Furthermore, Birdsong could find no evidence of marked differences between the two groups in the think-aloud data that he collected from the subjects as they performed their judgements. This study, then, suggests that at least some learners who start learning an Lz after puberty achieve a level of competence indistinguishable from that of native speakers. Other later studies (for example, Ioup, Boustagui, El Tigi, and Moselle 19943 Bongaerts 1999) have also produced evidence to show that at least26 The Study of Second Language Acquisition some learners who statt as adults cannot be distinguished from native speak- ers, even when the comparison involves pronunciation. Again, though, much may depend on how performance is assessed. Hyltenstam and Abrahamsson (2003) produced evidence to suggest that when detailed analyses of learners’ spontaneous speech are undertaken some differences between learners and native speakers always emerge. Birdsong (2006a), responding to this point, suggested that ‘it is more reasonable to argue that minor quantitative depar- tures from monolingual values are artefacts of the nature of bilingualism, wherein each language affects the other and neither is identical to that of the monolingual’ (p. 22). In other words, absolute similarity in performance between an L2 learner and a native speaker is simply not possible because of the very nature of bilingualism. A final approach that has figured strongly in recent research is to examine brain measures of highly proficient bilinguals. In this kind of research, the participants are given language tasks to perform and the activity in their brains is measured in a variety of different ways (for example, using mac- NETIC RESONANCE IMAGING (MRI).? The aim of such studies is to establish whether processing in the L2 is accomplished in the same way as processing in the Lr. Summarizing the results of a number of such studies, Birdsong (2006a) concluded that, where tasks involving production are concerned, it is Lz proficiency level rather than age of arrival in an L2-speaking environment that is the strongest predictor of the degree of similarity between process- ing in the Lr and L2. Comprehension studies have produced similar results, However, Birdsong noted one difference. There is evidence that processing is more effortful in the L2 than in the Lr even with highly proficient bilinguals. In general, these studies support Green’s (2005) ‘convergence hypothesis’, according to which Lx and L2 processing become increasingly similar as La proficiency increases. Research that has examined neurolinguistic aspects of Lz acquisition is considered in Chapter 14. What conclusions can be drawn from this research? Birdsong (2005) con- ducted a meta-analysis of studies that have investigated the La end state. He reached three conclusions: a Inall analyses of pooled data from early and late arrivals, age effects persist indefinitely across the span of surveyed age of arrival (AoA) (i.e. they are not confined to a circumscribed period), b In analyses of disaggregated samples (and in studies that lo t late AoA), most studies find significant AoA effects for the late ee tae ing postmaturational declines in attainment. : © Inanalyses of early-arrival data alone, AoA effects are inconsi € inconsistent: some are flat, some are random, and some are monotonically declining (pp. 14-15) Overall, then, the available evidence speaks clear end point beyond which La learners will fil es acige eee he proficiency. Rather there is a gradual decline i age starting from early childhood, I fail to achieve native-speaket in the ability to learn an L2 withAn introduction to second language acquisition research 27 The CPH is the subject of ongoing debate. Birdsong (19994) summarized the positions adopted by the ‘whys’ and the ‘why nots’. He also described the various mechanisms that different researchers have proposed to account for why late starters have difficulty in acquiring an L2. These are summarized in Table 1.6. He then reviewed the evidence for the other side of the fence (which he acknowledges is where he sits). He emphasized that although in general the research is not consonant with the CPH, there are nevertheless clear matura- tional effects evident in La acquisition. That is, the ability to learn a language (like most other things) declines gradually and steadily with age. Hypothesis Description Loss of neural After:the closure of the critical period (CP), the neural substrate Plasticity in the | responsible for language learning is not fully available due to brain loss of organizational plasticity and lateralization (Penfield and Roberts 1959). Lossof (access | strong version of this hypothesis is that Universal Grammar to) the language _ | (UG) is no longer available on closure of the CP. A weak version learning faculty | is that UG is mentally represented but no longer accessible. See Bley-Vroman's (1989) Fundamental Difference Hypothesis. Maladaptive gain | The greater processing capacity of adults enables them to extract of processing more from the input and thus they are faced with the problem capacity with of analysing everything at once, whereas children’s limited maturation processing capacity means they extract less from the input but canhandle it. Useitandthen | The language learning faculty has served its purpose once a loseit language has been learned; evolution has ensured that itis dismantled once it is no longer needed as keeping it would incur costs. Useit or lose it | This draws on ‘a mental muscle metaphor’ thats, ifthe language learning faculty is not used, it will atrophy but if itis used it is maintained. Bever (1981) suggests that acquisition requires that perception and production systems need to work together but once one stops learning a language this ceases. Learning inhibits] Connectionist theories? see learning as a matter of the Jearning strengthening of neural connections but once the connections. have been firmly established they are difficult to undo. Thus the ability to learn may change over time as a function of previous learning. Other Other variables that may account for age-related difficulty in L2 hypotheses learning are availabilty of input, social-psychological factors such as learner attitudes and motivation, and availability of instruction. Table 1.6 Explanations of age-related difficulties in L2 learning (based on Birdsong 19994: 2-9).28 The Study of Second Language Acquisition Fossilization Irrespective of their age (but especially if they start in adolescence or later), many learners do not achieve full native-speaker competence—they stop short, continuing to manifest grammatical and lexical errors in their L2 production and, even if overcoming these, failing to achieve a native-like pronunciation oF to behave in accordance with the pragmatic norms of the target language (TL). The term fossilization was introduced by Selinker (1972) to refer to this phenomenon, Selinker and Lamendella (1978) defined fossilization as: -+- a permanent cessation in learning before the learner has attained target language norms at all levels of linguistic structure and in all discourse domains in spite of the learner's positive ability, opportunity, and motivation to learn and acculturate into target society. (p. 187) Selinker (1972) suggested that only 5 per cent of learners succeed in achieving full competence, Other researchers have suggested that this figure may be too generous and it is more like 1 per cent or 2 per cent. None of the learners in the five case studies was successful in achieving target-language compe- tence despite favourable learning conditions. Hyltenstam and Abrahamsson (2003) argued that no L2 learner can achieve full native-speaker proficiency.4 In this respect, L2 acquisition differs from Lx acquisition, where, except for highly unusual cases of children who are deprived of opportunities to hear and speak—such as the case of Genie’ (Curtiss 1977)—all learners succeed in becoming native-like in their home dialect. Fossilization is not an all-or-nothing phenomenon, First, there is consider- able variation in the extent to which individual learners fossilize. That is, learners vary at what point in their development of an Lz they fossilize, with some ceasing development at a very elementary level, manifesting continued pidginized forms in their production (as in the case of Alberto), and others at a much more advanced level. Second, as Han (2004) pointed out there is ‘intra-learner differential success/faluse’, That is, a learner may reach target-language norms in some aspects of the L2 but not in others. Readers may be familiar with learners like Henty Kissinger, whose English grammat is essentially native-like, but whose accent remains conspicuously foreign. Differential success/failure can also arise within the grammatical system of the La itself (as seen in Lardiere’s study of Patty). Syntactical features (for example, basic subject-verb-object word order) are relatively easy to acquire while morphological features (for example, subject-veth agreement in English) are much more difficult. Thus, itis perfectly possible for s leaner to be fossilized in some aspects of the L but to continue to develop in others. Fossilization can refer to both a product (i.e. a State of fossilization) and a process (a cognitive view). Han (2004) offered these definitions: Cognitive level: Fossilization involves lying mechanisms that produce permai those cognitive processes, or under- nently stabilized IL forms.An introduction to second language acquisition research 29 Empirical level: Fossilization involves those stabilized interlanguage forms that remain in learner speech or writing over time, no matter what the input or what the learner does. (p. 20) However, there is a conspicuous lack of detail in the SLA literature address- ing the nature of the cognitive processes that comprise fossilization. Instead, researchers have focused on demonstrating the existence of fossilized states and the putative causes of fossilization. Han (2004) identified a total of 49 dif- ferent variables that have been invoked to account for fossilization. Some of these explanations refer to cognitive processes—lack of attention, inappropri- ate learning strategy, the inability to notice input-output discrepancies, false automatization, reluctance to restructure, processing constraints, a natural tendency to focus on content rather than form, reduction in the computa- tional capacity of an innate language faculty, and neural ENTRENCHMENT. ‘There is an equally long list of social-psychological variables (for example, the Jearner’s will to maintain his or her own identity). Han opted for a primarily cognitive explanation: Ihave argued that fossilization is internally determined due to the constant functioning of maturational and native language constraints, yet it can be modulated (aggravated or alleviated) by environmental, social, and psychological forces. (p. 43) However, she also acknowledged Selinker and Lakshmanan’s (1992) Multiple Effects Principle, according to which fossilization is more likely when two or more factors work in tandem. Fossilization is, however, a problematic construct. Learners can only be considered ‘fossilized’ in relation to some set of norms—usually those of some standard variety. Bialystok and Hakuta (1999) noted ‘there is an assumption in all research into second language acquisition that the learner is striving toward some stateable goal, a standard and perfect version of the language that is embodied in the mind of every native speaker’ (p. 165). However, such an assumption may not be warranted, as many La learners may not be committed to performing like native speakers. In particular, learners in the multilingual melting pots of large cities such as London and New York may be targeting MULTICOMPETENCE (Cook 2002), involving varying abilities in a number of different languages, rather than target-like use of a single La. It is problematic to term such learners ‘fossilized’ even if they manifest non-native norms in some (or even all) of the codes at their disposal, as their overall ability in language may outweigh that of monolingual speakers. This view is compatible with Larsen-Freeman’s (2005) claim that the very notion of a target ‘end state’ is theoretically untenable, for ‘when we entertain a view of language as a dynamic complex adaptive system ... we recognize that every use of language changes its resources, and the changed resources are then available for use in the next speech event’ (p. 408). In other words, all language systems, including that of the native speaker, are constantly chang-e isition 30. The Study of Second Language Acqui : sesense to measurea learner’s competence against ingand ths it does mot makesense fo meNSLT some statie notion ofa native SPESKE? TOTESY. a with research that has There are also methodological probe ae called cae claimed to investigate fossa udies (one of which was his own) that ofthe research, identifying O¥7) Tt uirementshe argued werenceded toclaim met thestringent methodological al fossilization must be demonstrated not that fossilization has taken place—i.e. fosstit had 1 \ learners under investigation must have had ample opportunity sedan earner sen the La, the data collected should be longitudinal iva . : i hd include samples of spontaneous speech, and the analyses should involve stability/change measures of individual learners. Lardiere’s os es these requirements and thus it would seem reasonable to conclude that Patty isa fossilized learner, as indeed Lardiere claims. — Long proposed that SLA researchers would do better to investigate sta- BILIZATION phenomena, which are well attested in L2. acquisition, rather than fossilization, which he claimed is almost impossible to demonstrate, ‘Stabilization’ refers to a state of Lz development where fluctuation has tem- porarily ceased. Many Lz learners are familiar with a situation where they appear to plateau, failing to develop despite their continuing efforts to do so, but then make a ‘breakthrough’ some time later. Stabilization is easier to demonstrate empirically as it does not constitute a permanent condition—that is, a stabilized La system can subsequently become destabilized. All that is required to demonstrate that a learner has stabilized is evidence to show that over a given period of time some specific non-target feature of the learner's La system has persisted. In contrast, to demonstrate fossilization it is neceseary to provide evidence that the learner's Lz system is permanently stabilized This requires showing that even after many years of exposure to the language the learner continues to manifest the same non-target features, _ There is also another reason for preferring ‘stabilization’ to ‘fossiliza- tion’ Talk of fossilization positions Lz learners as failures but in face many achieve very considerable success in acquiring an Lz. A key factor in dewk imining success is undoubtedly instruction (see Chapter se); ho have received some instruction generally outpecf ho cely cane cape : Y Outperform those who rely entirely alistic learning, Bardovi-Harlig (2000), fo i the results of a number of longitudinal studies undetsene? eeee European Science Foundation Project which inveeyase ten, °8 Patt of the Heh investigated the acquisition of a number of languages by mi : ‘ y migrant work, : x995~in order to examine whether the instructed erect DEC & al uninstructed learners. The more successful aac aed from he instructed | i i repertoire ae : se, ins ; ae ; eae ae was not the only factor that distin, caked the suc cel and nerfs f0F example, the amount of ith eae ga rte, ount of contact witl but instruction was clearly a major explanatory factor, c s, the is iri the morphological features the absence ot which pacers asa defining en taken as a definingAn introduction to second language acquisition research 31 characteristic of a fossilized learner (as in Lardiere’s study of Patty). It would follow that a learner who displays ‘stabilization’ might be able to continue earning with the help of instruction. Ie is clear that fossilization is a problematic consttuct—both conceptually and methodologically. As Fidler (2006) pointed out, there has been little Progress made in achieving the twin goals of reaching consensus on what fossilization is and how it can be described empirically. However, fossiliza- tion/stabilization are important constructs in SLA because they raise what is a central question: is La. acquisition largely the product of a learner-internal capacity for language learning, which atrophies once the first language has been acquired, or is it driven by exogenous and endogenous factors such as the availability of instruction and the learner’s motivation to learn? Age-related effects: some general conclusions The following is a summary of the main findings of research into age-related effects and fossilization in L2 acquisition: 1 Adult learners have an initial advantage where rate of learningis concerned, particularly in grammar. They will eventually be overtaken by child learn- ers who receive enough exposure to the L2. This is less likely to happen in instructional than in naturalistic settings because the critical amount of ex- posure is usually not available in the former. 2 Child learners are more likely to acquire a native accent and grammar than adult learners but there is now sufficient evidence to suggest that a native accent and grammar is possible for at least some adult learners. 3 Children are more likely to reach higher levels of attainment in both pro- nunciation and grammar than adults but only providing there is sufficient exposure to the L2. Where there is not sufficient exposure (as is the case in many instructed learning contexts) late-starting learners may continue to outperform child learners, especially in grammar. 4 There is no agreement as to whether there is a critical period for learning an L2. Researchers who argue for a critical period acknowledge that it may vary depending on the aspect of language involved (for example, the criti- cal period may differ for pronunciation and for grammar). Disagreements among supporters of the critical period also exist with regard to the span involved for any particular aspect of language. Long (1990a), for example, puts the onset critical age for pronunciation at six years, but Scovel (1981) argues that there is no evidence to support this and argues for a pre-puberty start. 5 Overall, however, the research indicates no clear discontinuity in learning as a result of age. Rather, the ability to learn an L2 declines gradually with increasing age. Also, there is growing evidence that some learners who start learning as adults can achieve a native-like competence (ie. fossilization is not inevitable).aes isition j_ The Study of Second Language Acq in activity, there is no clear evidence tha ures of brain activity, ther . 6 Asevidenced Py Tre the L2 is substantially different from that in the Ly language Penge, Neural activity is more influenced by the leamners in balan . ; roficiency in the La. _ general eee vo achieve full native-speaker ability in an La, which 7 Many learners coarchers to posit that fosslization’ isa characteristic of Lz has led sone ever methodological problems exist in both defining fos. see eed algo determining whether it exists. While in many cases La daacing may be slower than Lr acquisition, it has not yet been clearly es. in teeiiched that La learners necessarily fail to reach the same end state as Ly learners, especially if they have access to instruction. ‘As Larsen-Freeman and Long (1991) point out, the age issue is an important one for theory-building in SLA research, for educational policy-making, and for language pedagogy. A key theoretical issue is whether there is an innate, biologically endowed capacity for language learning and whether this is avail- able in L2 as well as Lr acquisition. Language educators are concerned with whether foreign-language learning is likely to be more successful if started in childhood (i.e. at primary school) than in adolescence (i.e. at secondary school). Language teachers want to know whether children learn in differ- ent ways from adults and thus whether they need different approaches and techniques to teaching an La, depending on the age of the learner. There are, however, no easy answers to these questions. Initially, it was believed that child language acquisition relies on an innate capacity for lan- guage learning (what Chomsky has termed ‘Universal Grammar’ (UG)) and that this was no longer available after a certain age. This is now in some doubt. But even if child and adult language acquisition are seen as drawing on the same learning mechanisms, this would not constitute evidence against UG, as itis possible that this is available in both. In other words, arguments about UG cannot be resolved with reference to the role of age in Lz acqui {t would appear that the research on the age factor does not support an Garly start for La learning in school settings, given that late starters appea! t0 doas well and sometimes better than early starAn introduction to second language acquisition research 33 What the research on both fossilization and age demonstrates is the enormous complexity of the issues involved in Lz acquisition and the meth- odological problems that exist in SLA. The controversies and mixed findings that characterize research in both areas are, in fact, evident in SLA asa whole. Thope that the above examination of these two issues will prepare readers of this book to expect controversy and the uncertainty that it accompanies. In my final chapter, Chapter 17, I will examine the epistemological problems that arise from this indeterminacy. Before that, however, I would like to lead readers through the main areas of enquiry in SLA so that they can see for themselves what has been discovered about Lz acquisition and where the divisions of opinion are to be found. To this end, I conclude by outlining the framework for exploring SLA that informs the whole book. A framework for exploring SLA The framework is shown in Table 1.7. A distinction is made between ‘General SLA’ and ‘Instructed SLA’. The former investigates issues relative to all La learners, irrespective of whether they are in a naturalistic or an instructed setting. The latter is concerned with how learners acquire an L2 ina classroom setting. In the case of General SLA a distinction is made between ‘Descrip- tion’, where the emphasis is on identifying the properties of learner language, and ‘Explanation’, where the attention shifts to theoretical accounts of L2 acquisition and of individual learner differences. Instructed SLA involves both the study of how the kinds of interaction found in classrooms influence L2 acquisition and whether learners actually learn what they are taught. This framework affords seven distinct areas of enquiry, which are outlined below. ‘Area 1 addresses the characteristics of learner language. The study of these provides the researcher with the main source of information about how acquisition takes place—the ‘facts’ that need to be explained. One of the major goals of SLA research is to describe learner language and to show how it works as a system. Four aspects of learner language will receive attention: (x) ERRORS, (2) ORDER OF ACQUISITION AND SEQUENCE OF DEVELOP- MENT, (3) VARIABILITY, and (4) PRAGMATIC FEATURES relating to the way language is used in context for communicative purposes. Area 2 examines learner-external factors that can account for how an L2 is acquired. We will adopt both a narrow and broader view of such factors. In the narrow view, we will focus on the linguistic environment by investigating the properties of the input and interaction learners experience and how these affect learning. Taking a broader view, we will study the macro and micro social contexts of acquisition, examining such issues as learner identity and its influence on learning. Area 3 addresses the psycholinguistic processes involved in both Lz acqui- sition and Lz use. Three views of these processes are considered. In one view, these are seen as mental and largely hidden from view, although not neces- sarily completely unconscious. They involve (z) the transfer of knowledge34 The Study of Second Language Acquisition uonssinbov v7] SunpStisoaus sof yomauoy y LT a1qzy (2 sasdeyo) uonssinb3e 7140 squno93e 21s)nBUry (LL saidey>) uorysinb>e (s sandeyp) z14o squnore abenBueiaqul Jo Jeunyyn201905 sasmyeay snewBeld (ot saadeyp) asnz1j0 (psardeyp) syunoave aniq1ub0> (¢ sa3dey) (6sardeyp) (zsaxdeyp) sapuanbas uonisinbse z1j0 —- Buruseaj 7140 | jequauidojanap pue squno22e annjuboy —_squno9ze je05 (91 saideu>) (Stuaadeyp) | (pL seadeyp) (gL seadeya) uonysinbe 21 uonisinboe 77 vonysinbse 12123] 71 (gsedeyp) | pueuonsnysuy —_pue uon2es=qUL Z14osqunos2e uy saauasayip (guaxdey>) uogpesaqut : pasn20}-ww04 wooussep) >psinBuyjounaN, enpiwipuy J94sueu} 1} pueanduy ! (zsaydeyp) si0113 abenbuepaqui uy “x0q 22} uonisinbse Ayiqeuen sassanoid slope? abenBuej sues) Anzanp Buruansaqu} aypapisuy : 77 pueuresq ayy Jauieapsiu] —>psinBuyoypfsg — jeuayxaiauseay | — jo so1jsa}>eIey) easy gealy sealy peasy eeasy Zeauy | Leaiy ‘abenBue] suse9| Bujiea} Jo uoneueldxg Jo uondursaq vis paynaysuy Vis je1uayAn introduction to second language acquisition research 35 from the learner's Lr, (2) the universal processes involved in converting input into INTAKE and RESTRUCTURING existing L2 knowledge systems, and (3) the processes for using La knowledge in Lz production. All of these draw on cognitive theories of language, language use, and language acquisition. The second perspective can be found in (4): SoctocuLTuRALSLA, This explains L2 acquisition in terms of the interplay between learner-external and learner- internal factors, treating acquisition as something that happens both outside and inside the learner's head. The third perspective, seen in (5), is linguistic in orientation; that is, an explanation of how learners learn language is based on the claim that learners have innate knowledge of LINGUISTIC UNIVERSALS. Area 4 addresses individual learner differences and what causes them. Learners set about the task of acquiring an L2 in different ways. They differ with regard to such general factors as MOTIVATION and LANGUAGE APTITUDE, and also in the use of various strategies such as inferencing and self-monitoring for obtaining input and for learning from it.‘ The study of these general factors and LEARNING STRATEGIES helps to explain why some learners learn more rapidly than others and why they reach higher levels of proficiency. An alternative approach to explaining La learning is offered in Area 5. This looks at artempts to understand how learners learn an La by examining what parts of the brain are involved in storing and accessing L2 knowledge. Key topics in this area are whether different parts of the brain are implicated in the storage of implicit and explicit La knowledge and whether Lr and La processing involve different neural activity. Area 6 is the first of two areas to examine instructed SLA. This area attempts to go inside the ‘black box’ of the classroom to examine the ways in which language is used there and how this may influence the course and success of acquisition. It views instruction broadly as ‘interaction’ affording (or sometimes limiting) opportunities for language learning. ‘Area 7 investigates whether direct attempts to intervene in the course of interlanguage development through FORM-FOCUSED INSTRUCTION are effective and, given that there is growing evidence that they are, which types are more effective than others. The research in this area, while of obvious relevance to language pedagogy, also addresses issues of key theoretical importance in SLA. It is not always clear where specific lines of enquiry fit best. For example, the study of the communication processes involved in using Lz knowledge is viewed by some researchers as an aspect of interaction (Area 2), by others as an aspect of psycholinguistic processing (Area 3) and by yet others as one type of learner strategy (Area 4). Also, the areas interrelate, so not surpris- ingly many investigations will fgure in more than one area. For example, the errors that learners make (Area 1) reflect the operation of internal processing mechanisms (Area 3) and may also be influenced by the social context in which learning takes place (Area 2) and the learner’s preferred LEARNING STYLE (Area 4). In other words there will always be more than one way of carving up“} 36 The Study of Second Language Acquisition osen reflects my own conceptualization of it for my readers. ‘hat examines a number of epistemo- logical issues in SLA. These concern the nature of the data that can best inform about L2 acquisition, the scope and evaluation of theories of La acquisition, and to what extent and in what ways SLA can be applied to language pedagogy. the content of SLA. The way Ihave ch of this content and how to make sense The book concludes with a chapter # Summary and conclusion the term ‘second language acquisition’, point- ing out the diversity of phenomena which have been investigated under its banner. It then went on to present the results of a number of case studies of Lz learners before examining the role of age in L2 acquisition. A framework for reviewing SLA research was then proposed. This consisted of seven major areas: (r) learner language, (2) learner-external factors, (3) psycholinguistic processes, (4) inter-learner variability, (5) the brain and Lz acquisition, (6) inside the black box of the classroom, and (7) direct intervention in interlan- guage development. Itis useful to distinguish two main goals of SLA research: description and explanation. In the case of description, the goal is to provide a clear and accurate account of the learner’s competence and, in particular, to uncover the regularities and systematicities in the learner's development and control of Lz knowledge. In the case of explanation, one goal is to reveal how learners are able to develop knowledge of an L2 from the available input and how they use this knowledge in communication. A second goal is to specify the factors that cause variation in individual learners’ accomplishment of this task. Itis also useful to distinguish two branches of enquiry within SLA research. One has as its focus learning, and the other, the language learner. In the case of the former, the emphasis is on identifying the universal characteristics of Lz acquisition. In the case of the latter, the aim is to account for differences in the ways in which individual learners learn an La. The ultimate goal of SLA is to develop a theory that can interconnect the findings from both branches and thus account for both the universal aspects of L2. acquisition and the individual differences observed in L2 learners. However, we are still a long way from such a theory, although, arguably, some progress has been made in this direction. SLA continues to be characterized by a plethora of theories and models, some of which overlap and some of which are complementary. There is little ign of this plethora thinning out. Nor is there agreement as to whether this is eR A cee achieved, At this point in the history of fossilization and the role of age ina aoencge ee ne above accounts of SLA is no different from oth ae Acquisition. In this respect, pethaps, m other social sciences. The compensation for the lack ‘This chapter began by definingAn introduction to second language acquisition research 37 of a single, highly focused picture is the richness of forty odd years of enquiry into La acquisition and the excitement this engenders—in me and, I hope, in the readers of this book. Notes 1 The terms ‘naturalistic’ and ‘instructed’ are not favoured by all SLA re- searchers, Alternative terms are ‘untutored’/‘tutored’ or ‘uninstructed’/ ‘instructed’, The pointhas been made that thereis really nothing ‘unnatural’ about classrooms. However, I have elected to keep the terms ‘naturalistic’/ ‘instructed’ in part because they are commonly used but also because they offer distinct terms for the two kinds of setting in which acquisition can take place rather than labelling one as simply the negative of the other. It should be noted that learners often do not belong exclusively to the ‘natu- ralistic’ or ‘instructed’ categories but rather to a third category, ofteri called ‘mixed’, 2 Magnetic resonance imaging uses a radiology technique to produce imag- ¢s of body structures. It has been used in language research to investigate what parts of the brain are active in performing different language tasks and also whether there are differences in brain functioning in Lx and La language use. 3. Connectionist theories of L2 acquisition view acquisition as driven prima- tily by input which enables learners to strengthen the connections in a neu- ral network, This network links together linguistic sequences of varying shapes and sizes. 4 Birdsong (2006a) rightly notes that estimates regarding the incidence of na- tive-like achievement in L2 learning are more a matter of guesswork than empirical evidence. He also notes that the estimates may have included for- eign language learners who may not have reached their end state due to lack of learning opportunities. 5 Genie was kept in virtual isolation for most of her life and did not receive substantial exposure to English before the age of 13. When discovered, she had no language. Genie was successful in learning English but did not achieve full grammatical competence, although it is not clear whether this was because she had passed the critical period for language acquisition or because of the emotional disturbance she continued to manifest. 6 A distinction is generally drawn between those cognitive processes respon- sible for the general pattern of La. acquisition (ie. its universal characteris- tics) and those strategies that are employed, often consciously, by learners to improve their learning. The terminological distinction between ‘proc- esses’ and ‘strategies’ is not always adhered to, however, ‘Strategies’ are often invoked to account for developmental regularities. The problems of definition are dealt with in Chapter 13.