EUROSLA Yearbook - Volume 2, Issue 1, 2002

This paper outlines the Autonomous Induction Theory, a theory of learning which is compatible with the hypothesis that Universal Grammar constrains the nature of natural languages. Based on Jackendoff’s architecture of the language faculty and Representational Modularity Hypothesis, it defines a trigger for the acquisition mechanisms and states a number of formal and psycholinguistic constraints on how they can operate. Finally, it argues that the theory is a necessary adjunct to purely formal learnability research.

Starting from the assumption that Japanese has no subject–verb agreement, this paper focuses on the acquisition of agreement, specifically on structure building of a functional category AGRP (agreement phrase) which provides a configuration in which nominative case is licensed/checked in a bi-unique SpecAGR relation. L2 clausal structures corresponding to Japanese multiple subject sentences receive particular scrutiny since the possibility of licensing more than one subject phrase is expected to influence L2 implementation of an AGRP in this context. Relying on a written corpus, the paper outlines a transition to a grammar with AGR, drawing on lexical learning (Clahsen, Eisenbeiss and Penke 1996), structure-building (Vainikka and Young-Scholten 1996, 1998) and elements of constructionism (Herschensohn 2000). The data indicate that Japanese speakers create a bi-unique spec-head relation for agreement. However, in clauses corresponding to multiple subject sentences, instances of failed agreement suggest that co-indexing is not yet consistently carried out with the phrase in SpecAGR. Also, instances of inappropriate predication and caseless DPs indicate that creation of AGRP does not bring about an immediate solution to the problem of integrating the multiple subject phrases in Japanese into English clausal structure.

This study tests the assumption in much of the literature on the second language acquisition of English tense and aspect morphophonology (e.g. bare verbs, V-ing, V-ed) that once speakers are beyond intermediate levels of proficiency, both distribution and interpretation of these forms are represented in a target-like way in their mental grammars. Three groups of advanced non-native speakers (whose L1s were Chinese, Japanese and the verb-raising languages Arabic, French, German and Spanish) were compared with native speakers on an acceptability judgement task requiring informants to judge the appropriateness of sentences involving different verb forms to contexts which privileged specific interpretations. The results suggest an effect of the persistent influence of parametric differences between languages such that where parametrised grammatical properties are not activated in the L1, they are not available for the construction of representations in the L2.

One account of divergence in advanced L2 grammars is that speakers fail to acquire functional features in the L2 that are not part of the L1 inventory, and that this leads to non-native representations. Since this idea was first proposed by Hawkins and Chan (1997), few studies have provided the type of data which would allow for it to be adequately tested. This paper presents experimental data from two studies of the acquisition of Case, number and gender agreement in a group of advanced learners of Spanish who are L1 speakers of English, French, German, Greek, Italian and Portuguese. Differences were found between accuracy on Case and number agreement on the one hand, and gender agreement on the other, in ways predicted by the Failed Functional Features hypothesis.

This paper investigates the effect of verb semantics on the variable use of aspectuo-temporal morphology by the advanced learner of French. Using the Aspect Hypothesis as a heuristic to test the extent to which the advanced learner overrides the effect of verb semantics in aspectual contexts where the verb assumes a non-prototypical value, the study examines the interaction between lexical aspect and grammatical aspect in order to identify which factor exerts a stronger influence on the learner’s aspectuo-temporal choice. Results suggest that lexical aspect continues to constrain the advanced learner’s aspectuo-temporal system, such that non-prototypical marking poses greater difficulty than prototypical marking.

This paper examines the acquisition of English articles by native speakers of Serbian, a language which has no article system. Data from a translation task, performed by four proficiency groups are presented to demonstrate that, contrary to previous research, learners do not associate articles with either ‘hearer knowledge’ or ‘specificity of referent’ (Huebner 1983, Master 1990, Tarone and Parish 1989, Thomas 1989). Rather, they establish non-target form–meaning connections, where ‘the’ is associated with a discrete referent, and ‘a(n)’ with the individuation of non-discrete referents.

Previous approaches to early bilingualism have argued either that children exposed to two languages from birth are not able to separate their two languages and experience massive cross-linguistic influence or that they do separate their languages from birth and lack crosslinguistic influence. The present paper assumes that both early language separation and crosslinguistic influence coexist in one bilingual individual during the same developmental stage for different grammatical phenomena. The goal of the present paper is to show that how crosslinguistic influence manifests itself depends on particular grammatical properties and is independent of language dominance. The direction of the influence is related to computational complexity (in the sense of Jakubowicz 2000). Data from a bilingual Italian/German child are discussed with respect to argument omissions, V2, and finite verb placement in subordinate clauses. For argument omissions, the Germanic language influences the Romance language and has a delaying effect. For V2 and finite verb placement in subordinate clauses, the Romance language has an accelerating effect on the Germanic language in the case of V2 and a delaying effect in the case of finite verb placement in subordinate clauses.

This article describes the results of a study conducted with fifty-three adult third-year learners of Spanish, French, and German in order to investigate (a) the degree to which idiom type affects the speed and ease of idiom comprehension and interpretation, (b) the effect that context exerts on idiom understanding, (c) the strategies second language learners employ in computing the idiomatic meaning of multiword phrasal units during contextualized and acontextualized reading of texts containing such idioms, and (d) the cognitive processes that are likely to constrain the construction of the right idiomatic mappings between target and domain idioms. Findings indicate that (a) there were significant main effects for lexical and post-lexical level idioms in both the context and the non-context treatment; (b) translation, guessing, and the use of context are highly important in the construction of idiomatic meaning; and (c) the degree of opacity between target and domain idiom, knowledge of vocabulary, graphophonics and syntactic arrangement, and literal meaning of an idiom influence and affect transfer of idiomatic knowledge in significant ways.

This study investigates the acquisition of scope items such as ‘only’, ‘even’, ‘also’, ‘still’, ‘again’, ‘already’ etc. in the longitudinal data of untutored second language learners of English, French and German. These items are found to appear in a fixed sequence: additive/restrictive > iterative > contrastive, which correlates crosslinguistically with the development of learner varieties from a prebasic to a postbasic level. Analysis of the discourse behaviour of these particles suggests that while the communicative potential of these items justifies their early appearance, their use is constrained by the global organisation of learner varieties. In other words, appearance of particular items is constrained by the learner’s developmental stage.

This study investigates the use of pronouns nous and a subgroup of on in a corpus of advanced oral and written French interlanguage produced by 32 Dutch L1 speakers. The subject nous + first person plural verb is characteristic of formal styles while on + third person singular verb is typical of informal styles. A quantitative analysis of the oral corpus revealed that the amount of authentic interaction, but not the amount of formal instruction, in the target language is significantly related to the choice of on. Its use correlated with morpholexical accuracy rates, fluency, omission of ne in negations and use of colloquial vocabulary. A similar analysis of the written corpus revealed equal proportions of on, which suggests that as a group, the learners had not yet completely acquired the constraints on this variable.

This paper addresses the linguistic realization of the face-threatening act of ‘complaints’ in native and non-native French discourse. Data were obtained through written Discourse Completion Tasks with German learners of French and native speakers to examine the extent to which L1 complaint strategies differ from L2 ones, the extent to which differences can be attributed to transfer from the L1, and the extent to which complaint behaviour is gender specific. While no direct evidence of pragmatic transfer from the L1 was found, significant differences were found between L1 and L2 in utterance lengths, degree of directness, use of supportive moves, and appearance of downgraders. Some gender specific features were also found.

This paper will first discuss the role of English as a lingua franca from historical and socio-political perspectives; it will then review some relevant findings of research into actual interactions conducted in English as a lingua franca (ELF), before describing a study of ELF interactions. It will be argued that we need a new research agenda to adequately describe and explain the hitherto unknown global spread of one particular language, and the wide variety of functions, domains, situations and populations it has come to involve.

It is widely acknowledged by second language researchers that whether pedagogical rules should or can be explicitly taught is largely dependent on their complexity. Nevertheless, there is no consensus on what rule complexity exactly means. This paper first examines existing accounts of rule complexity and presents a conceptual analysis of the term ‘rule’. It then proposes that ‘complex’ should not be equated with ‘difficult’, but used in a purely structural sense. Specifically a conditional formulation is proposed in which the number of concepts in the antecedent and the consequent, the number of subconditions, and the number of ‘if-then’ connections (subrules) within a given rule domain govern complexity. Finally, a classification of strategies of complexity reduction in foreign language pedagogy is sketched. Throughout, examples will be given from the field of German as a foreign language.