In this paper we present some results from an experimental study that we have been conducting into the effects of syntactic attrition on the L1 of Greek and Italian speakers who have achieved near-native proficiency in their L2 (English) but still use their L1 on a regular basis. In particular, we test the hypothesis, developed on the basis of assumptions regarding syntactic modularity, that the changes in L1 syntax will be restricted to the interface with the conceptual / intentional cognitive systems. The area of investigation is the domain of grammatical subjects in Greek and Italian. More specifically, we tested the participants on the production and interpretation of null and overt subjects, and of preverbal and postverbal subjects. We also elicited grammaticality judgments on subject extraction and subject position in various syntactic contexts. In this paper we report on the results of one of the production tasks (of preverbal and postverbal subjects) and two interpretation tasks. Attrition effects are found in the production of preverbal subjects in the Greek group whereas Italian speakers show attrition effects in the interpretation of overt pronominal subjects. We argue that these results are in the right direction, that is, that semantic features are vulnerable in language attrition whereas syntactic options remain intact.

The second language acquisition (SLA) literature reports numerous studies of proficient second language (L2) speakers who diverge significantly from native speakers despite the evidence offered by the L2 input. Recent SLA theories have attempted to account for native speaker/non-native speaker (NS/NNS) divergence by arguing for the dissociation between syntactic knowledge and morpho(pho)nology. In particular, Lardiere (1998), Prévost and White (2000), and Goad and White (2004) claim that highly proficient learners have knowledge of the abstract syntactic properties of the language but occasionally fail to associate them with the correct morphological or phonological forms. On the other hand, theories that support partial availability of Universal Grammar (UG) (Tsimpli and Roussou 1991; Hawkins and Chan, 1997) argue for a problem in the syntax: while UG principles and operations are available in SLA, the formal features of the target language that are not instantiated in the L1 or have a different setting, cause learnability problems. This article discusses acquisitional data in the light of the Interpretability Hypothesis (Tsimpli and Mastropavlou, 2007), which is a reformulation of the SLA theory suggested by Tsimpli and Roussou (1991) in minimalist terms. It is argued that a minimalist approach to SLA can be implemented to specify the status of the features that are least accessible to re-setting in the SLA process, given (1) constraints on their learnability and (2), their setting in the L1 grammar. The phenomenon discussed concerns the use of the resumptive strategy in wh- subject and object extraction by intermediate and advanced Greek learners of English. It is proposed that the acceptability rate of pronouns in the extraction site is conditioned by the Logical Form (LF) interpretability of the features involved in the derivation. Hence, the interpretable features of animacy and discourse-linking are hypothesized to be involved in the analysis of English pronouns by Greek L2 learners, while the first language (L1) specification of resumptive pronouns as clusters of uninterpretable Case and Agreement features resists resetting.

The focus of this study is the acquisition of grammatical gender in Greek and Dutch by bilingual children whose other language is English. Although grammatical gender languages share the property of noun classification in terms of grammatical gender, there are important differences between the languages under investigation here in terms of both the morphological cues for gender marking available to the child and the developmental path followed by monolingual children. Dutch offers limited input cues for grammatical gender, but Greek shows consistent and regular patterns of morphological gender marking on all members of the nominal paradigm. This difference is associated with the precocious pattern of gender acquisition in Greek and the attested delay in monolingual Dutch development. We explore the development of gender in Dutch and Greek with the aim of disentangling input from age of onset effects in bilingual children who vary in the age of first exposure to Dutch or Greek. Our findings suggest that although bilingual Greek children encounter fewer difficulties in gender acquisition compared to bilingual Dutch children, amount of input constitutes a predictive factor for the pattern attested in both cases. Age of onset effects could be partly responsible for differences between simultaneous and successive bilinguals in Greek, but this is clearly not the case for Dutch. Our findings are also addressed from the more general perspective of the status of “early” and “late” phenomena in monolingual acquisition and the advantages of investigating these from the bilingual perspective.

No abstract

This paper investigates the acquisition of Greek voice morphology in relation to transitivity alternations. 'Non-active' voice morphology is found in reflexive, anti-causative and passive structures. The role of the non-active morpheme is to 'check' a thematic feature of the verb, internal or external. Developmentally, the study addresses hypotheses proposed for the acquisition of transitivity alternations (Borer & Wexler 1987, Borer 2004), based on comprehension and production data from Greek L1, as well as L1 Turkish speakers with Greek L2. Results show that L1 and L2 learners can 'read' non-active morphology as passive or reflexive, indicating that it is syntactically computed. The differences between learners and native controls concern use of non-syntactic constraints on an otherwise target-like grammatical system (Borer 2004). hypothesis (Grimshaw 1981, Pinker 1984, Randall et al 2004), and the Syntactic Bootstrapping Hypothesis (Gleitman 1990, Borer 2004), prioritizes syntax over lexical semantics. The two hypotheses are based on alternative syntactic analyses of argument structure representations. Earlier accounts proposed that information on argument structure is included in the lexical entry. Lexical information dictates the way syntax will project

In the present paper we provide an account of VSO in Greek and its (relative) absence in Italian, despite the fact that both languages allow for postverbal subjects. We argue that this parametric difference reduces to different lexicalisation options regarding the D-system of the two grammars. We assume that the clause structure divides into three basic domains (V, T, and C), and that nominal (clitic) positions are available in each of these domains, which, as we argue, can be lexicalised not only by clitics but also by full DPs. On this basis, we argue that the subject and object DP in Greek can appear in the same domain (V), as they spell out different features depending on their grammatical function, while this is not so in Italian, given that DPs spell out the same set of features irrespective of their grammatical function. This basic difference is responsible for the presence of VSO in Greek but not in Italian. We also consider the implications of our approach for the interpretation of subjects and arguments in general.

No abstract

The aim of this study was to identify potential clinical markers of specific language impairment (SLI) in bilingual children with SLI by using the Greek version of the Multilingual Assessment Instrument for Narratives. Twenty-one Greek-speaking monolingual and 15 bilingual children with SLI participated, along with monolingual (N = 21) and bilingual (N = 15) age-matched children with typical development. Results showed differences between typically development children and children with SLI in microstructure, while bilingual children with SLI were found to attain similar levels of performance, and even to outperform monolingual children with SLI, in macrostructure. It is suggested that the retelling coding scheme could permit differential diagnosis of SLI among bilingual children within the scope of narrative assessment.