Search

Although cross-linguistic research on second language tense-aspect acquisition has uncovered universal tendencies concerning the association between verbal semantics and tense-aspect markers, it is still unclear what mechanisms underlie this link. This study investigates the acquisition of two imperfective aspect markers (–ko iss– and –a iss–) in Korean by Japanese learners to determine how cognitive universals, first language (L1) transfer, and input frequency interact. The current findings are compared with the results of Lee and Kim (2007), who studied the acquisition of the same aspect markers by L1 English learners. Both groups of learners acquired the progressive –ko iss– earlier than the resultative –a iss–/–ko iss–, while L1 Japanese learners acquired the resultative use of –ko iss– earlier than L1 English learners. Japanese learners’ early acquisition of the resultative –ko iss– is likely due to L1 transfer, because Japanese has a similar imperfective marker (–te i–). We argue that cognitive universals (one-to-one principle), L1 transfer, and input frequency all contribute to the acquisition patterns of temporal morphology.

I investigated the acquisition of narrow-range rules governing the dative alternation (Pinker, 1989) by adult L2 learners of English: 32 English speakers, 32 Japanese speakers and 32 Chinese speakers participated. I investigated 4 of Pinker's narrow-range verb classes - the Throw class, the Push class, the Tell class, and the Whisper class: Participants rated the acceptability of prepositional and double object datives containing both made-up and real verbs in these subclasses. Both Japanese and Chinese speakers distinguished double object datives containing Tell-class verbs from those with Whisper-class verbs, but failed to distinguish double object datives containing Throw-class verbs from those with Push-class verbs. I suggest that Japanese and Chinese learners' acquisition of the dative alternation in English is governed by the properties of an equivalent structure in their L1 relative to the properties of the target structure

This article reviews the findings of research on second language (L2) interaction from the perspective of L2 syntactic development. Previous studies provide us with a broad view of the grammatical nature of interaction. However, they have not revealed how interaction affects grammatical development, nor have they considered the specific relation between interactional structure and grammatical structure. I argue that in order to better understand the role of negotiated interaction in L2 syntactic development, it is necessary to examine the specific grammatical structures in interaction guided by the criteria of relevance, availability, accessibility, and effect. By examining interaction from a more focused perspective, that is, Universal Grammar triggers, one can ask precise questions about the grammatical characteristics of interaction and can make very specific recommendations about further study. More fine-tuned grammatical measures are needed to understand the grammatical characteristics of interaction and their effects on L2 syntactic development.

This paper is concerned with how advanced L2 learners of English interpret reflexive anaphors such as himself and pronominals such as him in sentences such as John said Peter helps himself and John said Peter helps him. Parameterized Binding Theory claims that the settings for the governing category parameter dictate whether particular anaphors or pronominals are bound to other noun phrases; the five possible settings are related in opposite hierarchies of inclusiveness for anaphors and pronominals according to the Subset Principle. An experiment is described that tested the interpretation of himself and him across five sentence types by 14 native speakers and 47 advanced L2 learners of English from three different language backgrounds—Japanese, Romance, and Norwegian. A computer-controlled comprehension task gave the subjects 40 sentences, for each of which they had to decide whether him or himself referred to John or Peter by pressing the appropriate key. The results showed (1) anaphors were slightly more difficult than were pronominals overall, (2) pronominals were not treated as anaphors, (3) a consistent order of difficulty was found for the five sentence types, with certain exceptions, (4) common orders of difficulty and of response time occurred in all groups regardless of first language, again with exceptions, and (5) on one view the Subset Principle was positively related to difficulty for anaphors, negatively for pronominals. The discussion concerns the validity of comprehension tests as evidence for the Universal Grammar model. It is suggested that the results show that parameter settings are related to performance as well as to grammatical competence.

This study examines the responses of 60 Spanish, Chinese, and German L2 learners to English sentences with empty pronominal categories (ECs). Empty pronouns are not only acceptable in finite clauses of Spanish and Chinese but are pragmatically more natural (Rizzi, 1982; Huang, 1984, 1985). However, the phenomena are not sanctioned in standard German, which is similar to the condition for English. These facts were taken to indicate that the parametric apparatus governing pronoun conduct in the L1 of the Spanish and Chinese L2 learners would be less like that of German learners to a significant degree. It was also expected that if all three groups were asked to judge and edit English sentences with null pronouns, the Spanish and Chinese Ss would make significantly more errors than would the German Ss because of parametric adjustments than the former two groups would need to make. But this outcome was not fully supported by the data as it was only the Spanish Ss and not the Chinese Ss whose mean errors were significantly higher than those of the German group. Evidence indicated that typological parameters linked to pragmatic versus syntactic motivations of L1 had to be considered in accounting for the results.

Flynn has reported that the second-language processing of subordinate clauses and of anaphora is affected by the basic word order (head direction) of the learner's native language. This phenomenon is claimed to be a prediction of Universal Grammar and of Flynn's formulation of a parameter-setting model of second-language acquisition. This paper explores both the theoretical status of the prediction and the factual status of the phenomenon. The principal conclusions are that the prediction cannot derive from any current version of the theory of Universal Grammar, and that there is no evidence in Flynn's research for a processing difference based upon L1 grammatical contrasts. The clearest results are subject to more plausible interpretations, involving general considerations of online sentence processing that are prerequisite to other psycholinguistic explanations. The L1-based differences reported by Flynn are evidently artifacts of experimental method and inappropriate analysis, especially of the incorrect use of analysis of covariance in an attempt to correct for a mismatch between experimental groups. SUMMARY Flynn's version of a head direction parameter linked to anaphora interpretation is thus without foundation in linguistic theory and, furthermore, is founded upon empirical observations of doubtful applicability and replicability. However, Flynn's research is still of considerable importance, for three reasons. First, the study of pronoun processing ease in language acquisition is of intrinsic psycholinguistic interest. Whereas a fair amount of information about anaphora in child language acquisition exists, there is little in adult language acquisition. Second, Flynn's research is of methodological interest, especially with regard to the techniques used to deal with the crucial problem of comparing groups with different L1s and at different proficiency levels. Third, Flynn's claimed findings are relevant to the general issue of the effect of L1/L2 differences and of universal processing strategies in second-language acquisition. Thus, despite our general reservations as to the theoretical validity of this research, we will attempt to review the major claims and findings, adhering as much as possible to Flynn's own formulations.

Lexical simplification–the process and/or result of making do with less words–can be studied in a variety of linguistic contexts: translation, the usage of language learners, the usage of language teachers, pidginization, foreigner talk, simplified reading texts for learners. It operates according to universal principles that derive from semantic competence in the mother tongue. By means of a comparative study of some of these contexts–translation, the usage of second language learners and teachers, simplified reading texts–the operation of these universal principles is demonstrated. Aspects of simplification examined in detail include the use of superordinate terms, approximation, synonymy, transfer, circumlocution and paraphrase. A distinction is drawn between strategies and processes of simplification. “Strategy” refers to the way the learner/user arrives at a certain usage at a specific point in time, and “process” refers to the systematic series of steps by which the learner/user arrives at the same usage over time. Strategies may be process-initiating (e.g. over-generalization, transfer) or situation, bound (e.g. circumlocution, paraphrase).

The focus of this study is the question: Is there evidence of a universal preference for the open syllable in Interlanguage (IL) phonology? The subjects are two Vietnamese speakers and the data consist of recordings of the subjects in informal conversation first in the IL and then in the native language (NL). An error analysis conducted on the IL data was restricted to the subjects' attempts at target language words most similar in syllable structure to words in the NL, monosyllabic words ending in postvocalic [p t k m n η]. The findings give limited support for the hypothesis that a universal preference for the open syllable is a shaping force in IL phonology independent of the process of NL transfer but also indicate the importance of the preceding vocalic context in the NL transfer processes, and the overall importance in IL research of gathering NL data comparable to the IL data, as the subjects' NL data proved crucial in shaping the findings of this study.

It is argued here that childhood phonological errors systematically simplify the child's inventory of phonetic elements and strings. The argument is based upon a straightforward criterion of simplicity which is formalized in the following definition: if a phonological error reduces the total number of contrastive phonetic elements or strings which would otherwise occur in the child's speech, it can be said that the error is one of “simplification.” All of the well-documented types of phonological errors in child speech are a result of simplification.

High school and college students of French recalled French sentences which varied two types of negation. Half the sentences used affirmative syntax and half negative syntax (ne… pas); half used positive verbs (e.g., manger); and half used negative verbs (e.g., oublier). One to two weeks later students performed the same task with English translations. In French, sentences with affirmative syntax and positive verbs were remembered best; either form of negation depressed performance. In English, only verb negation depressed performance. In French, students sometimes inserted ne into affirmative sentences and omitted pas from negative sentences, an error a native speaker would not make. Amount of prior education in French affected how much students remembered but did not differentially affect the influence of syntactic and verb negation. The results suggest that there is an interaction during learning between the foreign language learner's native language and foreign language. Two suggestions about teaching syntactic negation in French are offered.

There is an emerging awareness in the field of L2 acquisition regarding the contributions made by both Contrastive Analysis (CA) and Creative Construction (CC) theories of L2 acquisition. CA theories indicate that the L1 experience is important in L2 acquisition, and CC theories indicate that principles of acquisition independent of this experience are also integral to L2 learning. Reconciliation of these two aspects within a principled, empirically-based theory of L2 acquisition has been difficult (cf. Eckman 1977, 1984; Zobl 1980; Andersen 1983, 1984; Gundel and Tarone 1983; Liceras 1983; Rutherford 1983, 1984; White 1985). This paper argues that the parameter-setting model of Universal Grammar (UG) (proposed by Flynn 1983a, b, 1987) for L2 acquisition provides the scaffolding necessary for an integration of these two components within one explanatory account. Consistent with CC, L2 learners within this model use principles of UG isolated in L1 acquisition in the construction of the L2 grammar; however, when values of parameters associated with these principles differ between the L1 and the L2, learners assign a new value to cohere with the values for the new target language. The L1 experience counts in determining whether such a new assignment of a parametric value is necessary. This aspect of the model is consistent with a traditional CA theory of L2 learning. Empirical support for this model is provided by a study that investigates the role of the head-initial/head-final parameter (Stowell 1981) in adult L2 acquisition of pronoun anaphora. Two groups of adults — L1 speakers of Spanish, a head-initial language and L1 speakers of Japanese, a head-final language — were studied in their elicited production of English, a head-initial language. Results indicate that both groups of learners use the head-initial/head-final parameter as a source of structural organization for the L2. Results also indicate that Japanese speakers (L1 ≠ L2) are sensitive, from early stages of acquisition, to the mismatch in head-direction in English and Japanese, and that they assign new values to this parameter to cohere with the target L2 value. These speakers are thus argued to have the head-initial/head-final parameter set in two different ways. Parameters of acquisition in the two cases correspond to those isolated for L1 acquisition of English; however, the point at which this similarity is observed depends upon the degree of correspondence between the two languages.

The purpose of this study is to investigate the deletion of redundant elements in coordinate structures using and in ESL learners' interlanguages (ILs). In natural language, there are certain universal constraints on the surface orders which may result from the deletion of one or more identical elements in such coordinate structures. Since it has been hypothesized that ILs are natural languages, this study sets out to determine if ILs obey universal constraints on surface orders of coordinate structures and therefore behave as natural languages with respect to these structures. It will be shown that they do. In addition, where ILs differ from English with respect to surface orders, it is proposed that such ESL learners' errors are not best accounted for in terms of surface order transfer or overgeneralization of a target language rule, but rather that ESL learners may transfer the discourse constraints on deletion from their native language and create a new IL form in English. Methodological considerations for doing such a study are also discussed. It is argued that a data base drawn from a single test can have serious implications for the accuracy of a description of an IL grammar.

The validity of two implicational universals regarding constituent order in questions is tested in the English speech of 14 native speakers of Japanese, Korean, and Turkish. The interlanguage evidence is found to be generally supportive of the two universals. Some fundamental methodological problems are discussed which are attendant upon testing universals not only in interlanguages but in linguistic systems of any kind.

In the search for universals of relative clause acquisition, the present research investigated the comprehension of six types of Hindi-Urdu correlative sentences by child and adult L1 and L2 learners. Group comparisons show a sharper distinction between the performance of the first and second language learners than between the children and adults. While the native speakers tend to pay attention to case markers in interpreting sentences, the English-speaking learners tend to ignore these morphological cues, relying rather on a word order heuristic. The L1 errors, particularly those of the adults, are more systematic than the L2. Many of the learners do not appear to have any functional strategy for discovering the missing noun complement in the Hindi-Urdu correlative clauses and instead resort to random guessing. The paper concludes that language universals are available for the processing of complex structures only once a certain level of proficiency has been attained.

This paper examines the goals and assumptions of second-language acquisition (SLA) research, in particular relating these goals and assumptions to those of linguistics. It is argued that SLA is linguistics and that second-language data are of import in understanding the nature of language. The main focus is language universals of which three approaches (Universal Grammar, typological universals, and processing universals) are considered. Both theoretical underpinnings and empirical evidence are brought to bear in presenting supporting and discinfiirming arguments for each of these three approaches. It is argued that we must take into account an understanding of how apparently disparate facts of language—which are universally related—are conceptually related before we can predict their effect on second-language acquisition. It is further pointed out that because an understanding of second-language learning necessitates an interdisciplinary approach, we must be able to specify where any single approach fits into the total picture. Finally, we present a view of the possible ways in which SLA and linguistics are theoretically related and point out that it is incumbent upon SLA researchers to make it known that the data we work with are of relevance and interest to the formulation of theories of language.

The general theme chosen for this approach to interlanguage (IL) syntax "state of the art" is the presumed goal of all IL syntacticians: to find consistency and pattern in syntactic variation. There are two overlapping subsidiary themes as well: those of describing the systematicity in IL syntax and of explaining it. Themes contributing to descriptive approaches include the transition from morphosyntax acquisition studies to those of more complex syntax, the emergence of syntax from discourse, explicitness, the investigation of language "process" as well as language product, and the serious consideration of "simplicity." Themes contributing to explanatory approaches include general frameworks such as language universals, language typology, core grammar, and markedness. The paper concludes with suggestions and predictions as to the directions in which continuing research on IL syntax might lead.

Traditionally, researchers in the field of pidgin and Creole languages have supported the simplification theory which regards a pidgin as a simplified version of the colonizers language. This paper examines the process of pidgin development within the context of Government and Binding theory as proposed by Chomsky (1981). Empirical evidence given in this paper from Capeverdean creole (and others) indicates that the linguistic phenomenon generally labelled pidginization, creolization, decreolization, etc. can be explained as follows. The contact of various languages may produce a new experience which subsequently fixes the parameters of Universal Grammar, providing a pidgin core grammar. This core is guided by a preference of structure imposed by the markedness theory allowing the learnability condition to operate. The resultant core grammar can later incorporate a marked periphery. The addition of this marked periphery is what had been viewed in the past as creolization. If one of the donor languages is characterized by structures which meet the learnability criterion, the incorporation of the marked periphery will be targeted toward that language. The directionality of the pidgin periphery may then be predicted. However, if none of the languages meets the learnability condition, the pidgin periphery may develop independently of the peripheries of the donor languages.

The task of contrastive analysis is to discover underlying similarities in the deep structure of two languages and the rules which provide language specific features on the surface. Within the model of a transformational grammar identifying syntactic primes and basic proposition rules in the deep structure together with rules for the realization of surface constructions, i t can be demonistrated that the language specific English and Italian imperatives rest on the same underlying deep structure. A unified set of rules marks the progression from a set of common primes and syntactic relations to different surface realizations for the two languages. Contrastive statements of this type, accounting for both similarities and differences, become useful in language teaching for learning problems arise in the deep to surface realization or stylistic rules operating on the lower nodes of the underlying deep tree structure. Explicating surface relationships through recognition of the relation between deep and surface structures makes possible a more explicit contrastive analysis.

From the viewpoint of syntactic error analysis, we look for children's second language (L2) learning strategies as manifested in their L2 speech. We discuss two conflicting hypotheses: 1) the contrastive analysis hypothesis which predicts that L2 goofs (errors) will result from first language (L1) interference, and 2) the L2 = L1 hypothesis which predicts that L2 goofs will be similar to L1 goofs and are not the result of negative transfer. We propose several theoretical assumptions that support the L2 = L1 hypothesis and that describe the process of L2 acquisition. Based on these assumptions, alternative explanations are offered for those goofs that seem to support the CA hypothesis. Thus, we propose language learning strategies that seem to account for all the available data on children's goofs.