eJournals Fremdsprachen Lehren und Lernen 24/1

Fremdsprachen Lehren und Lernen
flul
0932-6936
2941-0797
Narr Verlag Tübingen
Es handelt sich um einen Open-Access-Artikel, der unter den Bedingungen der Lizenz CC by 4.0 veröffentlicht wurde.http://creativecommons.org/licenses/by/4.0/121
1995
241 Gnutzmann Küster Schramm

Contrastivity and Individual Learner Contrasts

121
1995
Jennifer Ridley
David Singleton
This paper explores contrastivity between languages from the perspectives of language transfer, perceived language distance and second language learners’ use of lexical problem-solving strategies in target language production. Referring in the main to data collected from university-level formal learners of German and French – but also drawing on findings in respect of “naturalistic” L2 acquirers –, it notes psychotypology, item particularities and L2 proficiency level as factors in the determination of degree of strategic transfer use and then goes on to focus on contrasts between individual learners in respect of their willingness to fall back on L1/Ln in coping with lexical problems. L1 and L2-based lexical coinage emerges as a strategy of which at least some learners may develop quite a high degree of conscious awareness and which they may feel is worth deploying in a wide variety of language using contexts, in spite of the potential risks involved.
flul2410123
Jennifer Ridley, David Singleton Contrastivity and Individual Learner Contrasts Abstract. This paper explores contrastivity between languages from the perspectives of language transfer, perceived language distance and second language learners' use of lexical problemsolving strategies in target language production. Referring in the main to data collected from university-level formal learners of German and French but also drawing on findings in respect of "naturalistic" L2 acquirers -, it notes psychotypology, item particularities and L2 proijciency level as factors in the determination of degree of strategic transfer use and then goes on to focus on contrasts between individual learners in respect of their willingness to fall back on Ll/ Ln in coping with lexical problems. L1 and L2-based lexical coinage emerges as a strategy of which at least some learners may develop quite a high degree of conscious awareness and which they may feel is worth deploying in a wide variety of language using contexts, in spite of the potential risks involved. 1. lntroductory In this paper we examine contrastivity in two senses. We discuss some of the observed consequences of contrasts between ·languages for the language learner: language transfer, perceptions of language distance and use of compensatory communication strategies based on the Ll (or on a language other than the Ll or the particular L2 in question [henceforth Ln]). We also, however, consider contrasts between individual learners in terms of their willingness to fall back on their L1 or Ln knowledge as a resource when faced with lexical difficulties in producing the target language. Tue main body of data referred to was elicited via written translation tasks from Anglophone university-level learners of German and French participating in the Trinity College Dublin Modem Languages Research Project (henceforth TCD MLRP) (see e.g. Singleton 1990). Reference is in addition made to a small-scale study of film-retell data collected from three Italian acquirers of English within the framework of a research activity of the European Science Foundation entitled "Tue Ecology of Adult Language Acquisition" (henceforth ESF Project) (see e.g. Klein/ Perdue 1992; Perdue 1982). Both sets of data reveal a high degree of inter-learner variation in borrowing behaviour. Within the Tco MLRP we make considerable use of introspection as a source of data. In the present context such introspective data provide us with further insight into the extent of variability among the university-level learners of French and German with regard to their intentional reflection on linguistic form-based strategies as a resource-expansion strategy. There is evidence to soggest that learners' linguistic borrowing behaviour is related not only to the distance of the target language from the mother tongue both structural and perceived bot also to FLuL 24 (1995) 124 Jennifer Ridley, David Singleton individual, personality-linked factors which are associated with individual learners' willingness to take deliberate risks (cf. Beebe 1983). Our discussion begins with overviews of the phenomena of language transfer and language distance, and of the use of linguistic form-based communication strategies as a means of compensating for lack of target language [= TL] lexical knowledge. lt goes on to give a brief description of the research projects mentioned above. Same data are then presented which illuminate the transfer pattems among subjects from the different learner groups, and which highlight the variability of transfer behaviour among these apparently homogeneous learner groups. 2. Cross-linguistic influence and perceived language distance No one will now deny as once they might have (see e.g. Singleton 1987a) that cross-linguistic influence is an important factor in coming to grips with an L2 performance, nor that it is particularly obvious in respect of both the formal and the semantic aspects of L2 lexical processing and learning (see e.g. Odlin 1989: 77-84; Ringhorn 1987: passim). lt is worth emphasizing, moreover, that crosslinguistic influence does not operate only between Ll and L2. As has already been indicated, and as we shall see below, languages known to an individual other than his/ her Ll may influence each other in terms of learning and performance. Indeed, there may also be influence from any language known other than the Ll on the Ll; such "backlash interference", as Jakobovits calls it, is a matter of common experience and has been reported in the literature for decades (see e.g. Haugen 1953; Jakobovits 1969; James 1971; Weinreich 1953). A particularly interesting aspect of cross-linguistic influence is the role of cognates, false cognates and imagined cognates. Laufer (e.g. 1990a; 1990b; 1993- 94) has been an assiduous explorer of this issue, and she has demonstrated that learners constantly seek to make links between new words they come across in their TL and words in languages they already know even, to some extent at least, when the languages concemed are unrelated (see e.g. Bensoussan/ Laufer 1984). lt has been noted that conflicts often arise between such semantic-associative processing and more form-oriented processing, whose objective seems tobe to replicate the new item as closely as possible (cf. Service 1992, 1993-94). The learner then faces the challenge of resolving this conflict by testing out the divergent hypotheses against more data. This process is evidenced by Giacobbe's (Giacobbe 1993-94) Spanish-speaking female subject who, in trying to cope with French cuisine ("kitchen"), on the one hand associates it with the Spanish word for "kitchen" cocina -, producing forms like [kosin] and [kosi] and on the other hand endeavours to replicate the French form faithfully, producing forms like [kusin] and [kwisin]. While Ll-L2 semantic-associative connections are apparently looked for even in the least auspicioris circumstances (see Bensoussan/ Laufer 1984, cited above; cf. FLuL 24 (1995) Contrastivity and Individual Learner Contrasts 125 also Cohen/ Aphek 1980), there is copious evidence to show that the degree of sirnilarity or divergence that learners perceive between their L1 and their TL will nevertheless strongly influence the general extent to which they draw on Ll resources in trying to deal with L2 problems (see e.g. Kellerman 1977, 1979, 1983; Odlin 1989: 140ff.; Ringhorn 1987: passim). Nor does this kind of "psychotypologizing" affect only the Ll-L2 relationship. Where learners perceive a closeness or strong similarity between two second languages, this will typically encourage cross-linguistic influence between them sometimes, indeed, to the point where such interaction eclipses L1 ⇒ L2 influence (see e.g. Singleton 1987b; Ringbom 1987: 113 ff). 3. Ll/ Ln-based cQmmunication strategies We turn now to the strategic role of cross-linguistic interaction. The use of Ll/ Ln knowledge as a strategy to overcome problems in L2 performance which arise due to lack of L2 knowledge has been well documented (see e.g. Dechert/ Raupach 1989; Frerch/ Kasper 1983). Transfer strategies, in our definition, ate instances in L2 performance where the learner consciously utilizes, or borrows, the Ll (ör Ln) as a tool for understanding or producing the L2. lt has to be said that the process of isolating instances of strategic transfer is not unproblematic. One way to proceed in this connection is to focus strictly on those instances where not only is cross-linguistic influence manifestly at work but where the learner's attention is seen to be concentrated on solving a clearly identifiable communication problem. Frerch/ Kasper (1989: 175) distinguish between strategic transfer, subsidiary transfer and automatic transfer and their various functions. The first of these, they soggest, presupposes that "the learner's focal attention is on the planning problem, and that intentionality is involved in setting up a plan (a strategy) for solving it". In the present discussion we use the term strategic to denote learners' conscious use of Ll or Ln knowledge, and we apply it in our treatment of the data to instances of the deployment of L1 and/ or Ln linguistic form-based knowledge in what are quite patently attempts to compensate for lack of TL lexical knowledge. We use subjects' concurrent introspective reports to confirm the intentionality involved in such instances. As a starting point and broad framework for investigating individual learner differences in the use of Ll/ Ln-based-strategies in writing tasks we look to the research that has been conducted into oral communication strategies. However, studies attempting to link the use of various types of communication strategies with specific learner factors have tended to produce mixed results (see Bialystok 1990: 48 ff). The exploration of the link between use of Ll-based strategies and learners' proficiency levels exemplifies this. As one might expect, it appears to be the case that in general low-proficiency learners tend to fall back on their L1 knowledge more frequently than high-proficiency learners (see e.g. Tarone 1977; Paribakht FLuL 24 (1995) 126 Jennifer Ridley, David Singleton 1985). Clearly, the more restricted lexical stock of the beginning leamer might restrict his/ her use of, for example, "known" L2 approximations. On the other hand, the relationship between the use of Ll knowledge as a resource and proficiency levels seems to be far from predictable. For instance, in a study of advanced leamers - Danish-speaking subjects with five years' experience of leaming English -, Haastrup/ Phillipson (1983: 154) find that most make frequent use of Ll-based strategies, i.e. of borrowing, foreignization and literal translation. Tue elicitation task in Haastrup/ Phillipson's above-mentioned study is an oral interview with a native speaker. lt is relevant to specify this because the nature of the elicitation procedure and the type of communicative event involved appear to be instrumental in determining the types of strategies which leamers use. A study by Poulisse (1990) bears this out. For example, in this study, which is of three groups of Dutch-speaking leamers of English of different proficiency levels, the proportion of the subjects' borrowing behaviour as a compensatory communication strategy is much ! arger in an oral interview with a native speaker than in the tasks of object naming or story-retelling. Another factor in leamers' use of different types of lexical compensatory strategies seems to be the nature of the referent itself (cf. Kellerman 1991 ). Other variables that have been suggested as pertinent to levels of strategic transfer use include age and educational background. A further element to be considered is psychotypology, leamers' perceptions of the relationship between the Ll and L2, which (see above) impacts not only on strategic transfer but on transfer in general. Tue question of the role of the leamer's personality on his/ her strategy use has tended to be noticed as an additional factor in studies which have set out with different hypotheses. For example, Tarone (1977: 202) is led to conclude that certain personality characteristics may correlate highly with strategy preference for example, a preference for avoidance strategies or appeals for assistance. Poulisse (1990) also refers to individual, possibly personality-related, factors, which, she suggests, include cognitive style. Haastrup/ Phillipson (1983), for their part, interpret their findings as suggesting that there is a style for every leamer. In addition, studies of the translation process have revealed evidence of individual leamers' preferences (e.g. Gerloff 1986; Zimmermann/ Schneider 1987). Conceming the question of any possible link between personality and conscious transfer, this has tended tobe overshadowed by the ! arger issue of psychotypology. And unless leamers are subjected to personality tests any such linkage has to be speculative. lt is, however, feasible to explore personality-related aspects of leamers' linguistic behaviour in relation to their performance of a specific task in respect of which there are clearly defined performance criteria. Using Bialystok/ Ryan's (1985) model of the development of metalinguistic skills, which is based on the premises that different tasks require different degrees of analysis of knowledge and cognitive control, we infer that those leamers who exercise high levels of conscious cognitive control and who are prepared to coin words deliberately on the basis of their Ll or FLuL 24 (1995) Contrastivity and Individual Leamer Contrasts 127 Ln linguistic form-based knowledge in a written translation task without a dictionary to hand to verify the correctness of a coinage evince more evidence of risk-taking than those who prefer to play safe and use strategies of approximation of which they are more certain. Tue notion that some strategies are inherently more risky than others is suggested by Corder (1983: 18), who refers to borrowing as "guessing of a more or less informed kind". .Whether learners tend to focus on content/ meaning and/ or on linguistic form when they try and solve a lexical problem is one of the issues addressed by Kellerman (1991). In a proposal for investigating processes which underlie the use of compensatory lexical communication strategies, and which is used in the Nijmegen project (see e.g. Poulisse 1990), Kellerman makes a distinction between conceptual strategies and code strategies. According to Kellerman, where learners resort to compensatory strategies, they can either manipulate the concept or they can manipulate encoding media. Tue use of a conceptual strategy involves the learner focusing on the referent in terms of its meaning, or "our mental representation of the concept" (1991: 148). In so doing leamers come up with what have been referred to in more traditional taxonornies as circumlocutions (approximations, paraphrases, for example) and word coinages which are semantically motivated. Tue code strategy, on the other hand "provides an alternative meaQS of labelling the target referent, but without involving representations at the conceptual level" (Kellerman 1991: 150). Thus when using a code strategy the learner resorts to another language (Ll or Ln) or to "productive grammatical processes within the L2". In traditional taxononµes the strategy of resorting to another language has been categorised as foreignizing, borrowing, code switch or grammatically motivated word coinage. Poulisse (1990) refers to the manipulation of TL grammatical processes as "morphological creativity". Tue distinction between conceptual and code strategies in oral lexical compensatory strategies is transferable to our distinction in relation to written tasks between a content (meaning)-based focus of attention and a linguistic form-based focus of attention. There is, for example, evidence from the Tco MLRP to suggest that learners' deployment of strategies as they complete words in a C-test are rooted either in a consideration of the meaning of the item and its context or of its formal particularities (cf. Singleton/ Little 1991). Sirnilarly, in the Tco MLRP translation task which we discuss below, learners compensate for lack of TL lexical knowledge when they try to produce a translation-equivalent for a specific Ll item either by corning up with an approximation in the form of an L2 near-synonym (a content-oriented approach) or by creating a word (a linguistic form-oriented approach). FLuL 24 (1995) 128 Jennifer Ridley, David Singleton 4. The TCD MLRP and the ESF Project 4.1 The TCD MLRP Tue TCD MLRP became fully operative in October 1990. Its general aim is to monitor the L2 development of university-level learners on a continuous basis and to examine the possibility of connections between their L2 development and their previous general educational and language learning experience. Tue subjects involved in the project are füll-time students enrolled on undergraduate degree courses offered by the TCD Departments of French, Germanic Studies, Italian and Spanish. From October 1990 the entire yearly intake of students in the above departments have been asked to supply us (via questionnaire) with information about their general education and their language learning experience, and we have also been tracking the progress of all questionnaire respondents in terms of their university language examination results. In addition, subsamples of students of French and of German have been involved in a second level of the project which gathers actual L2 data and related introspective data. Tue data dealt with below come from the subsamples recruited in 1990-91 that is, 10 ab initio students of German, 12 students of German and 12 of French who had already taken the languages in question as school subjects -, the focus being principally on the students of German. Tue instruments we employ for L2 data collection across the totality of subjects involved at the second level of the project are C-tests (with associated introspection), word as~ociation tests, story-tell tasks, and translation tasks (with associated introspection). From 1990-91 to 1992-1993 the same four C-tests were administered during the course of each academic year (two per TL) in December and two in May, and the same word-as·socation, story-tell and translation instruments were administered each March. Tue data treated within the present discussion were elicited by translation tasks, on which we provide some additional information in the next paragraph. In the TCD MLRP translation task subjects are required to translate without the aid of a dictionary an English text into either German or French. Subjects are asked not to omit content just because they do not know or cannot recall the relevant lexical items. Tue text, which is an extract from Cara, the in-flight magazine of Aer Lingus, was chosen for its lexical difficulty. Twenty lexical items, all nouns or noun phrases, have been underlined in the text to be translated. Subjects are asked to focus their attention particularly on these items and to try and convey the intended message. As they are writing their translations they give a think-aloud report on their problems and progress (cf. Frerch/ Kasper 1987). Their introspective comments are recorded on audio-tape and subsequently transcribed. FLuL 24 (1995) Contrastivity and Individual Leamer Contrasts 129 4.2 The ESF Project The ESF Project consists of a series of comparative studies of the spontaneous acquisition of second languages by adult immigrant workers, the data having been collected by teams working in five different European countries. Five TLs and six source languages (henceforth SLs) were involved as follows: 1Ls: SLs: English I \ Punjabi Italian German I \ Turkish Dutch French I \ I \ Swedish I \ Arabic Spanish Finnish The subjects - 40 in all were recorded over a timespan of 2½ years, the data collection starting as soon as possible after learning had begun. The recordings in question were of everyday transactions or role-plays of such transactions, conversations with researchers and in the performance of various tasks, notably a retelling of an extract from a Charlie Chaplin film. The data referred to below comes from three Italian leamers of English engaged in the said film-retell task. 5. Data-analysis and discussion 5.1 The TCD MLRP translation data The hypothesis which we wished to explore in our examination of the Tco MLRP translation data was that individual leamers tend to develop their own preferred lexical problem-solving strategy styles in the face of problems associated with lack of TL lexical knowledge. In our analysis of the data preferred strategy-styles were inferred in respect of those subjects who tended to use the same types of compensatory lexical strategies during two attempts of the same task separated by an interval of twelve months. A taxonomy of lexical problem-solving strategies was drawn up based on how subjects proceeded in relation to the definitive solutions to which they committed themselves in response to the problematic items. This taxonomy was informed by the think-aloud introspective data which had been collected "on-line" during the performance of the task. The categories which emerged were: (i) strategies of abandonment, where the subject did not write any solution down; (ii) strategies of recall for instance, waiting for a lexical item to surface from long-term memory; (iii) the strategy of solving the lexical problem on the basis of spontaneous, intuitive "feel"; (iv) strategies which were clearly more reflective in nature. Instances of category (iii) were identified on the basis of (a) explicit or implicit (e.g. intonational) indications of problematicity and (b) rapidity of response (a solution offered in two seconds or less). Category (iv), the one pertaining to reflection, covers those instances where the subjects took more time to solve their lexical problems. In a very large proportion of such cases the subjects explicitly stated what they were doing in order to arrive at a solution for example, "I'll find another word"; "This FLuL 24 (1995) 130 Jennifer Ridley, David Singleton is a complete guess" or they focused their attention on various possible solutions as they trawled through a range of options, in L1 and/ or L2, before finally selecting a lexical item and writing it down. Before considering individuals' preferred strategy styles we looked at three other potential influences on response pattems: (i) that of the TL itself, (ii) that of specific lexical items in the source text, and (iii) that of the proficiency level of subjects. Where there was evidence to suggest that such factors played a role in determining lexical problem-solving strategy use within the leamer groups as a whole, it was taken that noticeably different behaviour on the part of particular individuals constituted especially strong evidence of individual preferred strategy styles, especially if such divergence from common pattems were repeated at both attempts at the translation task. With regard to the first issue above, whether the target language itself might be influential in subjects' borrowing behaviour, it was anticipated that the leamers of French would be more willing to borrow from the Ll because of an awareness of the high incidence of French/ English cognates (and in spite of the phenomenon of "false friends" in the two languages see e.g. Thody/ Evans 1985). Singleton/ Little (1991), reporting on C-test data from the pilot phase of the TCD MLRP, found that the leamers of French in their sample tended to foreignize more than the leamers of German. In the present study a comparison was made between the 12 advanced leamers of French and the 12 advanced leamers of German with regard to frequency of recourse within the translation task to Ll knowledge in response to lexical problems. lt was found that the leamers of French borrowed from the Ll nearly twice as often as the leamers of German, the respective percentages of instances being 11 % and 6%. This finding is clearly in line with that of Singleton/ Little. Tue second issue, which concems the possible effect of the particular Ll item to be translated into either French or German, was also explored among the same 12 advanced leamers of French and the 12 advanced leamers of German. 15 Ll items proved to be especially difficult to translate. Interestingly, the majority of subjects in both language-groups tended towards similar types of responses with regard to 11 of these 15 L1 "problem" items. For example, all those subjects who found the Ll item range (as in a range of goods for sale) problematic focused their attention on approximations such as choice, or selection, and produced translationequivalents of such approximations. Similarly, the problematic noun phrase cabin crew, which triggered conscious semantic reflection among all subjects, elicited translations of approximations such as workers, waitresses, or team. In contrast, other Ll items tended to trigger word coinages in both TL groups for example, the L1 items landing and takeoff (which was somewhat surprising, given that both these items were highly amenable to "safe" treatment via approximations such as the equivalents of arrival and departure). Landing yielded linguistic word coinages at the rate of 42% of the total lexical solutions among the leamers of German, and at the rate of 66% among the leamers of French. Regarding take-off, in the case of German translations there were several instances (22% of all solutions) of deliber- FLuL 24 (1995) Contrastivity and Individual Leamer Contrasts 131 ately coined compounds based on "known" L2 prefixes/ suffixes, for example: Abzug, Ausfang, Aufnehmung. In the case of French translations of this item there was a large percentage (44%) of coinages consisting in nouns obviously created on the basis of knowledge of L2 verbs for example, envolement (coined from envoler). Thirdly, we compared how the two differentproficiency-level groups of learners of German the beginners and the advanced group approached the same lexical items in the source text which they found problematic. lt was expected that the advanced group would borrow less frequently than the beginners from the Ll or Ln (all subjects had learned French and Irish to School Leaving Certificate level), and that in any case the advanced learners' larger lexical store would provide them with the option of using safe approximations rather than more risky lexical creations of which they were uncertain. The following table confirms this. lt shows a comparison between the strategic use of L1 and Ln by the 10 German beginners and the 12 advanced learners of German at their two attempts of the translation task. Strategie use of Ll Strategie use of Ln Beginners (1st attempt) (2nd attempt) 10% 3% 7% 2% Table 1 1 Advanced (1st atteinpt) (2nd attempt) 6% 0.5% 5% 0.5% Whereas the advanced learners of German tend, at both attempts, to borrow from Ll and Ln less frequently than the beginners, the former make markedly more frequent use of TL morphological knowledge, as the following table shows: Beginners Advanced (1st attempt) (2nd attempt) (1st attempt) (2nd attempt) Strategie use of TL morphology 3% 4% 12% 6% Table 2 2 Table 1: Percentages of lexical problems solved on the basis of L1 or Ln knowledge at both attempts by the Beginners' Group and the Advanced Group of learners of German relative to total numbers of lexical problems experienced. 2 Table 2: Percentages of word coinages based on TL morphological knowledge relative to total numbers of lexical problems experienced. FLuL 24 (1995) 132 Jennifer Ridley, David Singleton Tue above tables show only the behaviour of each group as a whole. We now turn to our central theme, the variability within a particular group in respect of deliberate use of Ll or Ln as a lexical problem-solving resource. In order to illustrate this phenomenon from the Tco MLRP we shall refer to the German beginners' group, which presented as highly homogeneous at least to the extent in that at university entry all members of the group were ab initio leamers of German who bad never been to a German-speaking country; moreover, in their German classes at the university they all shared the same teachers. The table below shows how individuals differ in their lexical problem-solving styles from other subjects in the group. Tue table refers to three types of strategies which were categorized as "reflective": i) the use of an approximation (a synonym or paraphrase), ii) intentional word coinage on the basis of Ll/ Ln phonological/ morphological knowledge; iii) intentional word coinage on the basis of TL morphological knowledge. Tue numbers in the table refer to the percentage use of each of these three types of strategies by each subject, at both attempts. Tue first attempt percentages appear on the left, and the second attempt percentages on the right. Subject "Known" Intentional Ll/ Ln-based Intentional L2-based approximation word coinage word coinage 52%: 60% 19%: 13% 0%: 7% 2 17%: 38% 3%: 5% 0%: 5% 3 26%: 36% 16%: 14% 6%: 0% 4 53%: 50% 5%: 18% 16%: 5% 5 11%: 31% 8%: 4% 8%: 0% 6 26%: 21% 18%: 5% 22%: 32% 7 19%: 50% 12%: 0% 8%: 0% 8 37%: 45% 17%: 15% 3%: 0% 9 20%: 33% 44%: 33% 10%: 22% 10 38%: 71% 21%: 0% 6%: 14% Table 3 3 From the above table two points emerge. First, some subjects tend towards the use of a specific strategy type at both attempts notably subjects 1, 3, 4 and 9. This suggests a relatively constant individual preferred strategy style, at least as far as this type of task is concemed. Secondly, with particular reference to the role of cross-linguistic strategies, we see that one member of the group, Subject 9, relies 3 Table 3: Percentage use of three types of "reflective" lexical problem-solving strategies by the German Beginners in their two attempts at the translation relative to total numbers of lexical problems experienced. FLuL 24 (1995) Contrastivity and Individual Leamer Contrasts 133 particularly heavily on Ll knowledge as a resource expansion strategy. lt seemed to us worthwhile in the present context to examine Subject 9's approach a little more closely, and this was facilitated by the fact that, as it happened, this subject was one of a small number of TCD MLRP volunteers who were asked to repeat the same translation task for a third time, eighteen months after their second attempt. By the time of the third translation attempt, Subject 9 was a much more experienced learner of German, having spent a few months working in Germany. Although at bis third attempt he bad fewer lexical problems in the translation to cope with than previously, nevertheless he still resolved 30% of them by deliberately referring to bis Ll. For example, he guessed the items Bukel for belt; Rack for rack; Selektion for range (as a translation equivalent of English selection). When interviewed immediately after bis third attempt, Subject 9 revealed in relation to bis use of Selektion that although he at first considered the safer approximation viele as a solution to a range of goods (the solution of the majority) in the end he decided to go ahead with what he feit to be a fairly secure guess. Subsequent remarks of bis confirmed the impression that here was a careful lexical problemsolver, rather than an unthinking risk-taker. When asked in more detail about how he approached bis language learning, bis replies showed bis approach to lexical problem-solving in L2 production to be linked to a consciously acquired strategy wbich he used in bis receptive encounters with both German and French. Without any prompting from the interviewer he stated that he bad learnt the (usually successful) ploy of appealing deliberately to bis knowledge of L1 when reading/ listening to either language. He attributed bis success with French at school to bis constant deliberate attention to words and their possible associations with English cognates. He indicated that he was also conscious of sometimes applying the same approach when producing the TL, whether in French, or in bis more recently acquired L2, German. With regard to bis strategies for coping with L2 production, he specifically stated that he frequently coined words deliberately both in French and German. He dealt with the "false friends" problem by relying on native speaker or teacher feedback to validate or invalidate bis coinages. In other words, strategic transfer was part of Subject 9's array of deliberate learning strategies. Thus, while this learner might have appeared from the quality of bis L2 output and in particular from the impression of impulsive risk-taking given by bis coinages to be lacking in metalinguistic awareness, bis introspective comments convey quite a different message, revealing him as someone who expended considerable mental effort in bis L2 lexical problem-solving and who deployed a great deal of intentional if rather misplaced control. 5.2 The ESF Project data We turn now to data from quite a different type of learner, i.e. from immigrants learning the language of the hast community mostly informally. Crowley (1994) FLuL 24 (1995) 134 Jennifer Ridley, David Singleton investigated whether there was a pattem in the use of compensatory communication strategies among three adult Italian immigrants to Great Britain named in the ESF Project as Andrea, Lavinia and Santo. As in the case of the Tco MLRP German beginners, what is involved here is a fully homogeneous group in terms of previous experience of the TL, none of the group having received any formal instruction in English during their schooling in ltaly. The three subjects also shared a similar general educational background (eight years of scuola media). As mentioned earlier, the data from these subjects to be discussed here come from a film-retell task. Crowley's investigation isolated those relatively clear instances in the three subjects' English narrative where lexical problems bad been encountered and where compensatory strategies bad been used. The strategies in question were then categorized as either conceptual strategies or code strategies using Kellerman's classification outlined above in Section 3. The following table presents the results of this analysis and demonstrates a high degree of inter-learner variation in terms of the distribution of the two strategy-types. Type of strategy Conceptual Code Andrea 14% 86% Table 4 4 Lavinia 80% 20% Santo 25% 75% A further analysis of the strategies used into subcategories such as calques, coinages, approximations, etc. confirmed the fact that these three individuals varied enormously in the degree to which they deployed particular compensatory devices. Of especial interest in the present connection is these subjects' use of Ll-based strategies, which were distributed as shown below: Andrea Lavinia Santo Language switching 23% 33% 22% Calques/ translations 46% 67% 67% Foreignizations 23% 0% 11% Table 5 5 4 Table 4: Percentages of conceptual and code strategy-types used by the three ESF subjects relative to all strategies used by each subject (adapted from Crowley 1994: 39). 5 Table 5: Distribution of Ll-based strategies used by the three ESF subjects in terms of percentages relative to all code strategies deployed by each subject (adapted from Crowley 1994: 40). - Andrea also engaged in some L2-based word coinage (8%). Neither Lavinia nor Santo used any L2-based code strategies. FLuL 24 (1995) Contrastivity and Individual Leamer Contrasts 135 lt is clear, then, that whatever may be responsible for inter-learner variability in compensatory strategy use and in particular Ll-based strategy use such variability is to be found among "naturalistic" as well as among formal L2 learners. Crowley herself concludes that inter-learner variation in this context is probably linked to personality factors. 6. Conclusion We have seen that while a number of factors are at work in determining the degree to which learners make use of strategic transfer in solving L2 lexical problems notably, the perceived relationship between the two languages in question, the particularities of the item in question, and L2 proficiency-level there does also appear to be an individual dimension to the issue. At least some learners seem to experience the need to develop their own particular methods of coping with insufficient mastery of the target language and to prioritize particular performance strategies. Falling back on Ll and/ or Ln knowledge develops among some learners as a metacognitive strategy as well as a cognitive strategy, and it is evident from the TCD MLRP data that years of language learning experience in formal classrooms fosters among some "expert" learners a conscious awareness of their own strategy use. This was especially clearly demonstrated in the ease of the German beginner labelled Subject 9, who regarded bis application ofdeliberate transfer as an invaluable prop to bis language learning in general and who bad exploited bis metacognitive perceptions in respect of transfer as raw material from which to develop an individual cognitive strategy of transfer in on-line comprehension and production. The extent to which such linguistic problem-solving behaviour is appropriate to the task in band varies, of course, from one situation to another, and in L2 writing tasks especially lexical coinage tends not to pay off. However, insofar as language learning through language use partly depends 'ön learners' ability to form and test hypotheses using what knowledge they have already, any incorrectness or inappropriacy of lexical coinages in L2 performance are insignificant against the more fundamental issue of transfer as a strategy which may quite appropriately underpin individual learners' beliefs about how best to go about their own language learning. References BEEBE, Leslie (1983): "Risk-taking and the language learner". In: SELIGER, Herbert W. / LONG, Michael H. (eds.): Classroom-oriented Research In Second Language Acquisition. Rowley Mass.: Newbury House, 39-66. BENSOUSSAN, Marsha/ LAUFER, Batia (1984): "Lexical guessing in context in EFL reading comprehension". In: Journal of Research in Reading 7, 15-32. FLuL 24 (1995) 136 Jennifer Ridley, David Singleton BIALYSTOK, Ellen (1990): Communication Strategies. Oxford: Blaekwell. BIALYSTOK, Ellen/ RYAN, Ellen Bouehard (1985): ''Toward a definition of metalinguistie skill''. In: Merrill-palmer Quarterly 31.3, 229-51. COHEN, Andrew / APHEK, Edna (1980): "Retention of seeond-language voeabulary over time: investigating the role of mnemonie associations". In: System 8, 221-235. CORDER, S. Pit (1983): "Strategies of eommunieation". In: Fiereh/ Kasper (eds.) 1983, 15-19. CROWLEY, Maria (1994): The Use of Communication Strategies: An Analysis of Lexical Errors Produced by L2 Learners of English. Unpublished M.phil. Dissertation. University of Dublin, Trinity College. DECHERT, Hans/ RAUPACH, Manfred (eds.) (1989): Transfer in Language Production. Norwood NJ: Ablex. F.IERCH, Claus/ KASPER, Gabriele (eds.) (1983): Strategies in Interlanguage Communication. London: Longman. F.IERCH, Claus/ KASPER, Gabriele (eds.) (1987): Introspection in Second Language Research. Clevedon: Multilingual Matters. F.IERCH, Claus/ KASPER, Gabriele (1989): "Transfer in produetion: some implieations for the interlanguage hypothesis". In: Deehert/ Raupaeh (eds,) 1989, 173-93. GIACOBBE, Jorge (1993-94): "Construetion des formes lexieales et activite cognitive dans l'aequisition du fran~ais langue seconde". In: Singleton (ed.) 1993-94, 29-54. GERLOFF, Pamela (1986): "Seeond language learners' reports on the interpretive proeess: talkaloud protoeols of translation". In: HousE, Juliane/ BLUM-KULKA, Shoshana (eds.): Interlingual and Intercultural Communication. Tübingen: Narr, 243-262. HAASTRUP, Kirsten/ PHILLIPSON, Robert (1983): "Aehievement strategies in learner/ native speaker interaetion". In Fiereh/ Kasper (eds.) 1983, 140-158. Haugen, Einar (1953): The Norwegian Language in America. Philadelphia: University of Pennsylvania Press. JAKOBOVITS, Leon (1969): "Seeond language learning and transfer theory: a theoretieal assessment". In: Language Learning 19.1/ 2, 55-86. JAMES, Carl (1971): ''The exeulpation of eontrastive linguisties". In: NICKEL, Gerhard (ed.): Papers in Contrastive Linguistics. Cambridge: Cambridge University Press, 53-68. KELLERMAN, Brie (1977): "Towards a eharaeterization of the strategy of transfer in seeond language learning". In: Interlanguage Studies Bulletin 2, 58-145. KELLERMAN, Brie (1979): "Transfer and non-transfer where are we now? " In: Studies in Second Language Acquisition 2, 37-57. KELLERMAN, Brie (1983): "Now you see it, now you don't". In: GASS, Susan/ SELINKER, Larry (eds.): Language Transfer In Language Learning. Rowley, Mass.: Newbury House, 112-134. KELLERMAN, Erle (1991): "Compensatory strategies in seeond language researeh: a eritique, a revision, and some (non-) implieations for the classroom". In: PHILLIPSON, Robert [et al.] (eds.): Foreign/ Second Language Pedagogy Research. Clevedon: Multilingual Matters, 142- 161. KLEIN, Wolfgang/ PERDUE, Clive (1992): Utterance Structure: Developing Grammars Again. Amsterdam: John Benjamins. KLEIN-BRALEY, Christine (1985): "Redueed redundaney as an approaeh to language testing". In: Fremdsprachen und Hochschule: Aks Rundbrief 13/ 14, 1-13. LAUFER, Batia (1990a): "Words you know: how they affeet the words you learn". In: FISIAK, FLuL 24 (1995) Contrastivity and Individual Learner Contrasts 137 Jacek (ed.): Further Insights into Contrastive Linguistics. Amsterdam: Benjamins, 573-593. LAUFER, Batia (1990b): "Ease and difficulty in vocabulary learning: some teaching implications". In: Foreign Language Annals 23.2, 147-155. LAUFER, Batia (1993-94): "Appropriation du vocabulaire: mots faciles, motsdifficiles, mots impossibles". In: Singleton (ed.) 1993-94, 97-113. ÖDLIN, Terence (1989): Language Transfer: Cross-linguistic Injluence in Language Learning. Cambridge: Cambridge University Press. PARIBAKIIT, Tahereh (1985): "Strategie competence and language proficiency". In: Applied Linguistics 6, 132-46. PERDUE, Clive (1982): Second Language Acquisition by Adult Immigrants: a Field Manual. Stras- 1: ,ourg: European Science Foundation. P0ULISSE, Nanda (1990): The Use of Compensatory Strategies by Dutch Learners of English. Dordrecht: Foris. RA.ATZ, Ulrich/ KLEIN-BRALEY, Christine (1985): "How to develop a C-test". In: Fremdsprachen und Hochschule: Aks,Rundbrief 13/ 14, 20-22. RINGB0M, Häkan (1987): The Role of the First Language in Foreign Language Learning. Clevedon: Multilingual Matters. SERVICE, Elisabet (1992): "Phonology, working memory and foreign-language learning". In: Quarterly Journal of Experimental Psychology 45a.l, 21-50. SERVICE, Elisabet (1993-94): "Contribution des codages memoriels dans l'apprentissage lexical". In: Singleton (ed.) 1993-94, 147-160. SINGLETON, David (1987a): "The fall and rise of language transfer". In: COLEMAN, James/ T0WELL, Richards (eds.): The Advanced Language Learner. London: CILT, 27-53. SINGLETON, David (1987b): "Mother and other tongue influence on learner French". In: Studies in Second Language Acquisition 9.3, 327-346. SINGLETON, David (1990): The TCD Modem Languages Research Project: objectives, instruments and preliminary results. Dublin: Trinity College (CLCS Occasional Paper 26), and Eric Reports Ed 333 723. SINGLETON, David (ed.) (1993-94): L'acquisition du lexique d'une langue etrangere. Paris: Encrages (Acquisition et interaction en langue etrangere; No. 3). SINGLET0N, David/ LITrLE, David (1991): "The second language lexicon: some evidence from university-level learners of French and German". In: Second Language Research 7.1, 61-82. SINGLET0N, David/ LITrLE, David (1992): "Le lexique mental de l'apprenant d'une langue etrangere: quelques aperyus apportes par le TCD Modem Languages Research Project". In: Bou- CHARD, Robert [et al.] (eds.): Acquisition et enseignement/ apprentissage des langues. Grenoble: Lidilem, 395-402. TAR0NE, Elaine (1977): "Conscious communication strategies in interlanguage: a progress report". In: BROWN, H. Douglas/ Y0RIO, Carlos A./ CRYMES, Ruth C. (eds.): On Tesol '77. Washington DC: Tesol, 194-203. TH0DY, Philip/ EVANS, Howard (1985): Faux Amis and Key Words. London: Athlone. WEINREICH, Uriel (1953): Languages in Contact. New York: Linguistic Circle of New York. ZIMMERMANN, Rüdiger/ SCHNEIDER, Klaus (1987): "The collective learner tested: retrospective evidence for a model of lexical search". In: Frerch/ Kasper (eds.) 1987, 177-96. FLuL 24 (1995)