Factors in Second Language Acquisition: Competitiveness, Self-Image, Vocabulary, Motivatio, Apuntes de Idioma Inglés

¡Descarga Factors in Second Language Acquisition: Competitiveness, Self-Image, Vocabulary, Motivatio y más Apuntes en PDF de Idioma Inglés solo en Docsity! APPLIED LINGUISTICS UCM, Reading 3 Ellis, R. (1994/1985: 99-126) S Individual learner differences and Second Language Acquisition Introduction Second language (1.2) learners vary on a number of dimensions to do with personality, motivation, learning style, aptitude, and age. The aim of this chapter is to examine the relationship between these factors and second language acquisition (SLA). First, however, It is important to consider two points about the nature of this relationship. Aspects of SLA influenced by individual learner factors There are two basic possibilities regarding which aspect of SLA is affected by individual learner factors. One is that differences in age, learning style, aptitude, motivation, and personality result in differences in the route along which learners pass in SLA. The other is that these factors influence only the rate and ultimate success of SLA. These are separate issues. To claim that individuals vary in the rate at which they learn or the level of competence they eventually attain is not controversial. Indeed, it is part of most language learners” and teachers” experience. However, to claim that individual differences influence the sequence or order in which linguistic knowledge is acquired is far more controversial, It runs counter to the arguments and evidence in favour of the *natural? route of development (see Chapter 3). As will be seen in the discussion of each individual learner factor, the effect on the route of SLA has not been seriously investigated. Nearly all the research into learner variables has involved either investigating their effect on the proficiency levels achieved by different learners, or describing how they affected an individual learner's response to the rask of learning a L2. Neither proficiency nor learning response provides any insights about the route of acquisition. A There are stark disagreements about the role of individual differences in SLA. As Fillmore (1979) points out, on the one hand individual differences are seen as an all-important factor, while on the other they are treated as relatively insignificant. Research which has concentrated on accounting for differences in the proficiency levels of learners has tended to emphasize the importance of individual learner factors. Research which has tried to examine the process of SLA has tended to play down their importance. 1,00 Undcrstanding Secontl Language Acquisition ldentification and classification of learncr factors The identification and classification of thc different individual factors has proved to be problematic. The main difficulty is that it is not possible to observe directly qualities such as aptitude, motivation, or anxiety. These are merely labels for clusters of behaviours and, not surprisingly, different researchers have used these labels to describe different sets of behavioural traits. As a result, it is not casy ro compare and cvaluate the results of their investigations. Each facror is not a unitary construct but a complex of features which are manifest in a range of overlapping behaviours. It is, thercfore, not surprising to find that a host of terms have been employed to describe the phenomena. Hawkey (1982) lists some of these: 'affective, cognitive, and social factors' (Tucker et al. 1976),'affective and ability facrors' (Chastain 1975), and'attitudinal/motivational characterisrics' (Gardner et a/. r979). In an attempt to impose some ordcr on this plethora of tenns and concepts, I propose to make an initial distinction between personal and general factors. Personal factors arc highly idiosyncratic fearures of each individual's approach to learning a L2. Some examples are provided by Schumann and Schumann (1977) in a report of their own language learning experiences. They include 'nesting patterns' (the need for a secure and orderly home base before learning can effecrively begin), 'rransition anxiety' (the stress generated by moving to a foreign place), and the desire to maintain a personal language learning agenda. The Schumanns found that such factors strongly influenced their SLA. The gcneral factors are variables that are characteristic of all learners. 'fhey differ not in whether they are presenr in a particular individual's learning, but in the extenr to which thcy are present, or the manner in which they are realized. General facrors can be further divided into those that are modifiable (i.e. are likely to change during the course of SLA), such as motivation, and rhose that are unnrodifiable (i.e. do not change in strength or nature as SLA takes place), such as aptirude. Personal and general factors have social, cognitive, and affective aspects. Social aspects are external to rhe learner and conceru thc relationship between rhe learner and native speake rs of the L2 and also betwecn tlre learner and other speakers of his own language. Cognirive and affective aspecrs are internal to the learner. Cognitive factors concenr the nature of the problem-solving srrategies used by the learner, while affcctive factors concern the emotional responses aroused bv the attempts to learn a L2. Different personal and gcneral facrors involve rtl three aspects in different degrees. Aptitude, for instance, is rhought of as primarily cognirive in narure, but also involves affective and social aspects. Personality is primarily affective, but also has social and cognitive sides. Age is a factor that rnay involve all three aspects fairly equally. It is because the personal and general factors that make up an Indiuidual lcanrcr differences 107 individual's language posed of social, cognirive, and affective features lex, and, as a rcsulq ofrer-r rather vaguely dcfined subsequent cliscussion wirr show, they play an im Personal factors Personal factors such as those identified by Schumann and Schumanu thodological problem se of diary studies. In their experiences in re author of the diary ological problem is to use quesrionnaires and i'terviews with indiviclual learners (e.g. Picketr I978; Naima n et al. 1978). There are difficurties iri about individual responses to SLA in rhis way. d to say what rhey think the researclrer wanrs to f-flattery. Anorher is rhat such techniques cau s of which rhe learner is conscious. Neverrheless both.the diary studies and the questionnaires/inrerviews rrave provided insights into rhe personal nature of language leanrirrg, pairicularly classroonr larrguage learning. Personal factors are by defnition heterogeneous. However, they can be.grouped .rogether under three headingi: (1) group dyna-ics, 1Z;attitudes ro tlre reacher and course materials, and (j) ir.riiuiiual learning techniques. I shall considcr each of these in rurn. Group dynamics 106 Understanding Second Language Acquisition L2 proficiency becomes. Burstall (1975; 17), reviewing the results of the NFER project on the teaching of French in the primary school, concludes 'the achievement of skill in a foreign language is primarily a function of the amount of time spent studying that language. . .'. Thus those children who started French in the primary school tended to outperform those who did not start until the secondary school. However, Burstall also endorsed the results summarized in the previous paragraph, namely that older learners were more efficient. With the passage of time, the influence of the age of the learner begins to ourweigh the length of the learning period, at least on listening and reading tests, but less clearly so on speaking and writing tests. This observation-that the effects of length of learning period are rnost strongly felt in productive rather than receptive skills-is lent support by Ekstrand (1975). He found that length of residence of immigrants learning Swedish in Sweden related to free oral production, but not to other aspe cts of proficiency. Thus it would appear that although number of years of exposure to the L2 leads to greater success, this may be restricted to overall communicative ability, rather than to grammatical or phonological accuracy (Hatch 1983a). Success in SLA also appears to be strongly related to the age wl.rcn SLA is comrucr.rced. This is particularly the case where pronunciation is c<rnce rned. Oyama (1976), for instance, found that the age of arrival of sixty Italian male immigrants in the USA was a far more potent determinant of the levels of pronunciation they achieved rhan was length of stay. In other words, as far as success in pronunciation is concerned, younger learncrs do better. In this respect, at least, popular opinion is substantiated. Oyama also investigated the effects of starting age on grammar, but the results were far less clear cut. These results may appear confusing and contradictory, but a fairly clear pattern em€rges if route, rate, and success are treated as separare, if due account is taken of the differential effects of age on pronunciation, vocabulary, and grammar, and if starting age is not confounded with the number of years' exposure to the L2. The pattern rs: t Starting age does not affect the route of SLA. Although there may be differences in the acquisitional order, these are not the result of age. 2 Starting age affects the rate of learning. Where grammar and vocabulary are concerned, adolescent learners do better than either children or adults, when the length of exposure is held consranr. \Wherc pronunciation is concerned, there is no appreciable difference. 3 Both number of years of exposure and starting age affect the level of success. The number of years' exposure contributes greatly to rhe overall communicative fluency of the learners, but starting age determines the levels of accuracy achieved. Darticularlv in pronunciation. Indiuidual leanrer d i l l t' t,'r t,, " I r t ontradictionint t llr'tt ely to lead to m t, I'ttt s do not acquire t-tltlcr- rcsolved if ir is h unger children onlv learn at the same rate or slower than older leartrers, thcy are more likely to go further (see Krashen et al. 1979)' Explaining the effects of age ln addition to the empirical research summarized above, there has been considerable theorizing about the effects of age on SLA. I shall exanrine a number of theorics in the light of the summary of the empirical results- The critical period hypothesis The critical pe riod hypothesis states that there is a period when language acquisition takes place naturally and effortlessly. Penfield and and Roberts (1959) argued thac the optimunr age for language acquisition falls within the first ten years of life. During this period the brain retains plasticity, but with the onset of puberty this plasticity begins to disappear. They suggested rhat this was the rcsult of the lareralization of the languagc function in rhe left hemisphcre of tl.re brain. Thar is, tl're neurological capacity for undcrstanding and producing language, which initially involves both hemispheres of rhe brain, is slowly concentrated in the left hemisphere for most people. The increased difficulty which older learne rs supposedly experience was scen as a direct result of this neurological change. Some evidence to support the critical period hypothesis was supplied by Lenneberg (1967). Lenneberg found that injuries to the right hemisphere caused more language problems in children than in adults. He also found that in cases of children who undcrwent surgery of the left hemisphere, no speech disorders resulted, whereas with adulrs almost total language loss occurred. Furthermore, Lenneberg provided evidence to show tlrat whereas children rapidly recovered total language control after such operations, adults did not do so, but instead continued to display permanent linguistic impairment. This suggested that the neurological basis of language in children and adults was different. Lenneberg's evidence, however, does not dcmonscrate that it is easier to acquire f anguage before puberty. In fact Lenneberg asst'tmed that language acquisition was easier for children. The critical period hypothesis is an inadequate account o[ the role played by age in SLA, because this assumption was only partially correct. OnIy where pronunciation is concerned is an carly start an advantage, and even theu only in terrns of success, not rate of acquisitiorr. The criticrl period hypothesis needs to be recast to account for why loss ofplasticity affects pronunciation but not other lcvcls of language. One possibility is that 1.08 Understanding Second Language Acquisition there are multiple critical periods (Seliger 7978). The process of the lateralization and localization of language functions is a gradual one, carrying on over many years. Different aspects of language are affected at different stages in this proccss. This explains why adolescents outperform adults in Brammar acquisition-around sixteen a critical period affecting grammar may be reached. This explanation is, however, speculative. In general the evidence linking ccrebral dominance and age differences in learners is not clear. Co gnitiue e x p lanatiorl s One obvious difference between the young child and the adolescent or adult is the abiliry of the latter to cornprehend language as a formal systern. Older lean.rers can learn aboutlanguaee by consciously studying linguistic rules. They can also apply these rules when they use the Ianguage. In contrast, younger children, while not totally lacking in meta-awareness, are not so prone to respond to language as form. For them language is a tool for expressing meaning. As Halliday (1973) poinred out, the young child responds not so much to what language ls as to what ir does. It is possible that age differences in SLA can be explained in tcrms of the different oricntation to language of children and older learners. Rosansky (1975) has argued that cognitive development accounts for thc greater ease with which young children learn languages. She believcs that L2 developmcnt can rake place in two diffcrent ways, according to whcther or not the learner is awarc of what he is doing. The young child sees only similarities, lacks flexible thinking, and is self-cenrred. Thesc are the pre-requisites of automatic language acquisition, because associated with them is an absence of mera-awareness. The young child does r-rot know that he is acquiring language. Furthermore, the young child has not developed social atritudes towards rhe use of one language as opposed to another. For thesc reasons he is cognitively 'open' t<r another language. In contrast, the adult cannor learn a L2 automatically and naturally. fhc onset of absrract thinking thar comes around the age of twelve with the final stage of cognitive developmenr, as described by Piagct (i.e. Formal Operations), means rhar the learner is predisposed t<r recognize differences as well as similarities, to think flexibly, and to becorne increasingly de-centred. As a result he possesses a strong meta- awareness. Also he is likely to hold strong social arrirudes rowards the use of his orvn language and the target language. These may scrve as blocks to natural language acquisition, forcing the lcarner ro rrear rhe acquisition task as'a problem ro be solvcd using his hypothetico- deductive logic' (ibid: 98). In Rosansky's view, then, it is rhe awarcness that comes with age thar inhibits natural learning and that leads ro an altcrnative approach. lt follows, as Rosansky recognizes, that even In diu i tlual learn er d i ffcren c e s | 09 where the restrlts of development appear to bc very sirnilar, thcy rrc, in fact, different. The problem with Rosansky's arguments is the same as that with the neurological explauations. They are both based on the false assutlptiott that post-puberty learners are less efficient and less successfr'rI thrrr younger learners. However, although Rosansky's position caunot stntrtl up to the empirical evidence, it is still possible that cognitive devclop- Inent is a factor. It can help to explain why adolescents learu morc rapidly than children. The meta-awareness that conres with Formal Operations may facilitate more efficient learning. Not only can the adolescent'pick up' language like a child, but he can supplenrent this process by conscious study. However, problems still remain. Why is no advantage in pronurrciation seen in the adolescent learncr? C)tle possibility is that of all aspects of language it is pronunciation rhat is the least amcnable to conscious manipulation. Anorher problem is why the adolescent outperfornrs the adult. lt may be necessary to turn to another aspe ct of cognitive developmenr-rnem()ry-to explain tlris. Adolcsce nts nray have better menrories rhan adults. A f fe ctiue e xp lanati ons Another possibility that has been explored is that diffcrenccs in the affcctive states of your-rg and older learners account for age diffcrences in SLA. Brown (1980b) proposcs that SI-A is relrrted to stages of acculturation (i.e. the abiliry of thc learner to relate and respond casily to the forcign languagc cultLrre). Brown identifies four stages of acculturation: (1) initial excitement and euphoria; (2) culture sl.rock, leading to feelings of estranBement and hostility towards the target culture ; (3) culrure srress, involving a gradual and vacillating rccovery; and (4) assimilation or adaptation to the new culture. Brown argues that stage (3) is the crucial phase. Young children are seen as socio-culturally resilicnt, because they are less culture-bound that adults. They n-rr>ve through the stages of acculturation rnore quickly and so acquire the L2 more quickly. 1'he major problem with Brown's theory is once again the false assumption that children are the more rapid learners. Neufeld (1978) offers e more convincing 3ccount of how affective factors are related to age differences in SLA. He distinguishes 'primary' and 'secondary' levels of language. Prirnary levels include a reasonably large functional vocabulary, and basic mastery of pronunciatior-r and gramrnatical rules. Secondary levels include the abiliry to handle complex grammatical structures and different lar-rguage styles. All learners, according to Neufeld, have an innate ability to acquire primary levcls. Ilowever, children are nrore likely to achieve secondary levels than adults becausc they arc rnnch more strongly motivatcd by the nced | | 0 IJ n d er s tan ding Se con d Lan gua ge A c q u is i t i on to be accepted by their peer groups. \flhereas the adult is happy to maintain a foreign accent, for instance, the child who is exposed to the first language culture is anxious to achieve narive-like pronunciation- Conclusion Neufeld's theory, supplemented by cognitive factors, can accommodate all the known facts about age differences in SLA. First, it explains why the route of acquisition is not influenced by age' If innate abilicies account for the acquisition of primary levels, no differences in routc between children and adults will be observed. Adults, however, will acquire primary levels more rapidly because of their greater cognitive abilities. The exception to this will be pronunciation, because of the difficulty of consciously manipulating this aspect of language. Children will prove the more successful learners, particularly when pronunciation is concerned, because they are strongly motivated to become part of the first language community and require a native-like accent to achieve this. lt can also be predicted that they will achieve greater overall communicative fluency, both because they are likely to receive more years' exposure to the L2, and because of the importance of this aspect of proficiency in peer group interaction. lntelligence and aptitude Learning a L2 in a classroom involves two sets of intellectual abilities. It involves what might be called 'a general academic or reasoning ability' (Stern 1983: 358), often referrcd ro as intelligence. This ability is involved in the learning of other school subiects as well as a L2. The other kind of abiliry consists of specific cognitive qualiries needed for SLA, often referred.to as aptitude. Intelligence lntelligence is the term used to refer to a hyporhesized 'general factor' (often referred to as the 'g' factor), which underlies our ability to master and use a whole range of academic skills. As McDonough (1981: 125) emphasizes, it refers to 'capacity rather than contents of the mind'. That is, it is the underlying ability to learn, rather than the actual knowledge tl'rat is supposedly measured by intelligence rests. In practice, of course, it is extremely difficult to separate these. To what extent does the'g'factor influence SLA? Oller and Perkins (1978: 413) have argued that 'there exists a global language proficiency factor which accounts for the bulk of the reliable variance in a wide variety of language proficiency measures'. They claim that the 'g' factor of language proGciency is identical with the 'g' factor of intelligence. Indiuidua! learncr tl i ffercn c es'l l-l' ProPoruons.' Tie distinction betwcen CALP and BICS explains a number of audio-lingual methods that emphasized habit formation' To conilude, intelligence may influence the acquisition of some skills associated with SLA, such as those utilized in the formal study of a L2, but it is much less likely to influence the acquisition of oral fluency skills. To put it ar.rother way, intelligence may be a powerful predictor of succ-ess in classroom SLA, particularly when this consists of formal naturalistic SLA, when L2 how to communicate in the t the effects of intelligence are ; there is no evidence that intelligence affects the route of acquisition evident in spontaneous language use (i.e. unplanned discourse-see Chapter 4). I I6 U nderstanding Second Language Acquisitiort depender.rcc/independence does not appear to be an important factor in SLA. These results run counter to intuitive expectations. The hypothesis that SLA is influenced by the way in which learners orient to the processing, storing, and retrieval of information is an appealing one. The problcrn may lie in the type of research that has taken place. lnstead of large-scale quantitative studies, a more qualitative approach which focuses on the actual Lrtterrnces produced by individual learners rnay be more relevant. Naiman et al. (1.978) found sonre evidence to show that learners produce different kinds of errors, depending on their cognitive style. Analytic learners were more likely to omit small items rhan whole phrases in sentence imitarion, whereas holistic learners were more likely to do the opposite. Cognitive style may interact with other learner factors. Fillmore (1980), for instance, suggests that thcre may be differences in thc level of attention which learners frorn different ethnic backgrounds typically give ro a task. She notes that whereas Mexican childrcn had difficulty in concentr^ting for a long periocl of time, Chinese children were able to stay on task for a sustained period of time. Thus, although the relationship between cognitive style and proficiency has not beel clearly demonstrated, thcre are a number of intcresting possibilities which have not yet been explored fully. It is prematurc to address the question of what aspect of SLA is influenced by cogrritive sryle. The existing research docs not conclusively show that it is a major factor where success is concerned. There has been no research into the effccts of cognitive style on route of acquisition. Hatch's (1,974) distinction between 'data gatherers' who become fluent but do not bother to sort out many rules and 'rule formers' who concentrate on accuracy, which seems to reflect the holisric/analytic distinction which is the principal characeristic of congitive style, suggests that cognirive style may evcntually turn out to be an important factor detcrmining rate of developmcnt. Attitudes and motivation The problcrns of defining attitudes and motivation are considerable. A common-sense view is that a person's behaviour is governed by certain needs and inrerests which influence how he actually performs. However, these cannot be directly observed. They have to be inferred from what hc actually does. Not surprisingly, therefore, rhe study of attirudes and motivation in SLA has involved the development of concepts specific ro language learning. The concepts have been derived from the behaviours of language learners and have been only loosely related to general theories of motivqtion in psychology. It is not always clear in SLA research what the distinction is between attitudes rnd motiuatioz. Schumann (1978) lists 'artirude' as a social Indiuidual learncr difl'cratttcs I 17 factor on a par with variables such as'size of learning group', arrtl 'motivation' as an affective factor alongside 'cultrtre shock'. Carclrtcr and Lambert 0972\ define'motivati<-rn' in terms of the L2 leartrcr's overall goal or orientation, and 'attitude'as the Persistence shown by thc learner in striving for a goal. They argue that there is no reasotr to exPcer a relationship between the two; the type of motivation is distinct from the attitudes displayecl to different learning tasks. However, Gardner (1979) suggests that attirudes are related to motivation by serving as supports of the learner's overall orientation. Brown (1981) also distinguishes 'motivation' and 'attitudes'. He identifies three types of motivation: (l) global motiuatiort, which consists of a general orientatiort ro the goal of learning a L2; (2) situational motiuation, which varies according to the situation in which learning takes place (the motivation associated with classroom learning is distincr from the motivation involved in naturalistic learning); (3) task motiuation, which is the motivation for performing particular learning tasks. (1) clearly corrc- sponds ro Gardner and Lambert's sense of 'motivation', (2) is a ncw concept, and (3) seems to be the same as Gardr-rer and Lambert's 'attitudes'. Brown uses the term 'attitudes' to refer to thc set of beliefs that the learncr holds towards mcmbers of the target language group (e.g. whether they are seen as'interesting'or'boring','honest', or 'dishonest', etc.) and also towards his own culture. These als<-r 6gurc in Gardner and Lambert's later use of thc term 'attitudes'. lt is clear that there is no general agreement about what precisely 'motivation' or 'attitudes' consist of, nor of the rclationship between the two. This is entirely understandable given the abstracrness of these concepts, bur it makes it difficult to compare theoretical propositions. The most extensive research into the role oI attitudes and motivation in SLA has been conducted by Gardner and Larnbert. Where 'motivation' is concerned, they draw a basic distinction between an integratiue and an instrumenta/ orientation to L2 learning. The forrner occurs when the learner wishes to identify with the culture of the L2 group. This type of motivation is an extension of Mowrer's (1950) account of motivation in first language learning. Mowrer argued thar the child associates the language he l're ars with the satisfaction provided by the parents' prescnce. Thus, just as the child seeks to identify with his parents by learning their language, so the L2 learner may be motivated to identify with the L2 speech community by learning its language- l-ater, Gardner (1979) linked an integrative motivation to 'additive bilingualism'. Thar is, lcarners with an integrative nr()tivation are scen as likely to rnaintain their mother tongue when they learn a L2. lnstrumental rnotivation occurs when the learner's goals for learning the L2 are functional. F'or instance, learning directed at passing an examination, furtherinB carcer opportunities, or facilitating study of other subjects throilgh the mediunr of the L2 are all examples of instrumentally m()tivatcd learning. Cardner I l8 Understanding Second Language Acquisition proposes that instrumental motivation is more likely ro be linked tcr 'subtractive bilingualism', where the learner either loses his mother tongue or fails to develop the ability to express cerrain kinds of functions (such as those associated with literacy) in it. In Britain ir has been suggested by Fitzgerald (1978) thar the motivational disposirion of L2 learners among ethnic minorities is more likely to be insrrumental. Gardner and Lambert, however, point out that the integrative/instru- mental distinction reflects a continuum, rather than alternatives. Gardner and l-ambert have also investigated a number of differenr attitudes which they consider relevant to L2 learning. Stern (1983: 376-7) classi6es these attitudes into three types: (1) atritudes towards the community and people who speak the L2 (i.e. 'group specific artitudes'); (2) atrirudes towards learning rhe language concerned; and (3) attitudes towards languages and language learning in general. These attitudes are influenced by the kind of personality of the learner, for instance whether he is ethnocentric or authoritarian. They may also bc influenced by the social milieu in which learning takcs place. Differenr attitudes, for instance, may be found in nronolingual and bilingual contexts. The results of the empirical research based on Gardner and Lambert's theoretical franle work are mixed and dif6cult to interpret. The following is a sumrnary of rhe major 6ndings: 1 Motivation and arritudes are imporranr facrors, which help ro determine the level of proficiency achieved by different learners. For example, Gardner (1980) reporrs rhar a single index of arrirude/ motivation derived from vanous measures of affective responses ro L2 learning is strongly related to measures of French proficiency in Canadian school leavers. Savignon (1976:295) claims thar'xrtirude is th_e single most important facror in second language learning'. 2 The effects of motivarion/attirudes appear ro be separate fiom the effects of aptitude. The mosr successful learners will be rhose who have both a talenr and a high level of morivarion for learning. 3 In certain situations an inregrative motivation may be nrore powerfr.rl in facilirating successful L2 learning, but in orher situations instru- mental motivations may count far more. For instance, Gardner and Lamberc (1972) found rhar an integrative orienrarion was relared to successful learning of French in schools in both Canada and USA, but that instrumenral motivation was more irnportanr in the Philippincs. 'fhey explained this in rerms of rhe role the L2 plays in the learner's community. Where the L2 funmions as a 'fore ign language' (i.e. is not irnportant outside the classroom for the learners), an intcgrarive motivation helps; but where the L2 functions as a 'second language' (i.e. is used as means of wider communication outside the classroom), an instrumental motivarion is more effecrive. Lukmani (1972) also Indiuidual leante r diffarenccs 719 found that instrumental motivation could be more efti'uivc thrrrr integrative motivation. Non-westernized female learners of L2 L,nglislr in Bombay, who were insrrumentally motivated, achieved high scorcs orr a Cloze test. The learning situation Lukmani investigated was vcry similar to that in the Philippines descrjbed by Lambert and Cardrrcr. It has been pointed out, however, rhatthe two types of motivation erc nor mutually exclusive. SLA rarely involves just an integrativc or just an instrumental motivation. Burstall (1975) found that the pupils' achievement in the NFER primary French pro ject was closely associated with both types of motivation. The pupils' progress was infuenced by both a desire to do well in l-rench as a school subiect, and by an interest in French people and their culture. 4 The level and type of motivation is strongly influenced by the social context in whiclr learning takes place, as has already been noted. There can be little doubt that motivation is a powerful factor in SLA. Its effects are to be seen on the rate and success of SLA, rather than on the route of acquisition. Precisely hou motivation affects learning, however, is not clear. One of the problems of the correlational studies, which constitute the bulk of the ivailable research, is that it is only possible to show a relationship, not the direction of this relarionship. We do not know whether it is motivation that produces successful learning, or successful learning that enhances motivation. lJurstall (1975) has addressed just this issue and concluded that achievement affected later attitudes and later achievelnent to a greater extent than early atritudes affected either later achievement or later arritudes. ln other words, it was the motivation that was engendered by rhe learning process irself that seemed to matter l-uost. A rather similar view is taken by MacNamara (1973). He argues that 'the really irnportant part of motivation lies in the act of comnrunication irself'rather than in any general orientation es implied by the integrative/insrrumenral distinction. It is the need to ger meanings across and the pleasure experienced when tlris is achieved tl-rat motivates SLA. These are views which are encouraging ro rhe languagc teacher. Morivation thar is dependent on rhe learner's learning goal is far less amenable to influence by the teacher than motivarion that derives from a sense of academic or comrnunicative success. In the casc of the latter, motivation can be developed by careful selection of learning tasks both to achieve the right level of complexity ro crexre opportunitics for success and to foster intrinsic interesr. Personality In general psychology, personaliry has been explored in terms of a number of personal traits, which in aggregatc are said to constirute rhe personality of an individual. Cartell (1970), for instance, arremprs ro 720 Unclerstanding Secontl Language Acquisitittn rneasure personality using a serics of dichotomies, seen as poles on continua, such as cool/warm, shy/venturesome, not assertive/dominant. Eysenck (1,964) idcntifies rwo gcneral traits, again represenred as dichotomies-extrovert/introvert and neurotic/stable. However, wirh one or two exceptions (e.g. Hawkey 1982), SLA researchers have prcfcrrcd to tJevelop their ow y rrairs, calling rhern anything from'social styles'( 1983) ro'cgocentric factors' (Brown 1981). Some y, Burt and Krashcn 1982) cven irrclude cognitive style as a personaliry trait. This confusion is thc result of both the many-facered narure of persorrality and the necd that individual researchers have felt ro investigate rraits which intuitivcly srrikc them as importanr. Ex t ro ue rsi o nl introucr s io rt One of the intuitively appealing hypotheses that has been invesrigatcd is that extroverted learners learn more rapidly and are more successful than introverted learners. It has been suggcsted that extroverted learners will find it easier to make contacr with other users of thc L2 and vcn rory. Social skills had beconrc comfortable comrnunic:rtors, others had hardly acquircd rny English at all. Howcver, of the scvcn soci:rl styies Strong I n diuitlual lea mar di[[arctr t as IZI investigated, only'talkativeness'and'rcsponsi ficantly related to measures of language development krrowl- edge, play vocabulary, and pronunciarion). Srr r rr wrs not so much social skills which enabled rhe n nr()ru input as thc abiliry to nrake more acrive use of the English rhey werc exposed to rhat led ro fast learning. Whar counted were rhosc persorrafity rrairs that conrrolled the quality of inreracrion in rhe L2 rather thalr those that led to quantily of inpur. lnhibition The other nra jor aspecr of personaliry rhat has becn srudied wirh regard to SI-A is inhibirion. lt is hyporhesizcd that the defensiveness associited with inhibiti.n discourages rhe risk-taking which is 'ecessary for rapid progrcss in a L2. Krashcn (l98la) suggesrs rhat rhe onset of Forrnal operati'ns hns a profound effect on the affecrive srare of the learncr. rt induccs egocentrism, which in tur' leads ro incrcased self-c..sci.usness Conclusiort lrr gcncral thc available rescarch ilocs.or show a clearly clefined effccr of pe rsonality on SLA. Onc rcasou why this is so nray be bccause personality bccrmcs a major factor only in the acquisirion of cornrnuni-