• Maria-Luise Beck, University of North Texas.

• Rakesh Bhatt, University of Illinois at Urbana-Champaign. "The Vulnerable Left Edge in Second Language Acquisition."

The recent debate over the accessibility of Universal Grammar (UG) in second language acquisition research has focused on whether or not the full functional structure (CP, TP, NegP, etc.) is available at the onset of acquisition. The debate centers around two proposals that have been offered as hypotheses on the "initial state" in second language acquisition : the Strong Continuity Hypothesis, which states that all functional categories are present at the onset of second language acquisition, and the Weak Continuity Hypothesis, which states that not all functional categories are initially present. I present evidence to claim that not all functional projections are available at the onset of second language acquisition. Specifically, I present developmental data from L1 Hindi speakers learning English as a second language to show that Left Edges, CP (and DP), are absent from the "initial state," for principled reasons. The evidence presented speaks directly to the debate over the Weak versus Strong Continuity Hypothesis, choosing the Weak version as the favored model of second language phrase structure acquisition.

• Robert Bley-Vroman, University of Hawaii. "On the Explanation of Variability in L2 Acquisition: Why the Theory of SLA Should not Be Predictive."

• Chris Bongartz, University of North Carolina-Charlotte. "Will Strength be Weakness? Resetting Determiner Feature Values in Near-Native Interlanguage Grammars."

Parameter resetting and functional features in interlanguage development have been discussed extensively with respect to the development of functional projections in the VP (AgrP, TenseP, AspectP) and at the sentence level (CP). This paper investigates parameter resetting in the DP, focusing on two central questions: (1) do learners achieve native speaker competence in terms of DP feature representation? (2) are there any differentiating effects of strong or weak determiner features in interlanguage DP feature representation? (cf. Longobardi, 1994, 1996). The data comes from natural production in story-retelling and a questionnaire. Participants are L1 Hungarian (strong), L1 French (strong), and L1 German (weak). I compare interlanguage data from 20 learners in each L1 group with those from 20 native-speakers. Results reveal L1 effects for definiteness features, but not in terms of specificity - suggesting separate heads for these features within DP.

• Harald Clahsen and Despina Papadopoulou, University of Essex. "Parsing Strategies in L1 and L2 Sentence Processing."

Most previous second language (L2) acquisition research has focused on aspects of linguistic knowledge in L2 learners. An issue that has largely been neglected is how L2 learners parse sentences on-line and how grammatical processing develops in L2 learners. This paper investigates two closely related questions: (i) In which ways do the parsing mechanisms used by (advanced) L2 learners differ from those employed by native speakers? (ii) Do parsing mechanisms develop separately from grammatical knowledge in L2 learners? With respect to the first question, we will examine whether L2 learners transfer parsing strategies from their first language when processing the L2. Concerning the second question, we will determine whether L2 learners rely on their L2 grammar for parsing purposes or whether they have to learn parsing mechanisms separately from the L2 grammar. We conducted off-line and on-line experiments with 16 native speakers of Greek and 16 advanced L2 learners of Greek with Spanish as their L1 on attachment preferences in sentences, such as Someone shot the servant of the actress who was on the balcony. This sentence is ambiguous in that the relative clause can be taken to modify either the first or the second NP (the servant or the actress).

We found a low-attachment preference in all experimental conditions, i.e. the relative clause is preferably attached to the lower NP irrespective of its form (genitive vs. preposition). The results show that (1) even advanced L2 learners (who have acquired all relevant grammatical properties of the constructions involved) still differ from native speakers in the way they parse sentences in the L2, suggesting that parsing mechanisms have to be learnt separately from the grammar; (2) parsing mechanisms are not directly transferred from the L1. For Spanish L2 learners, we found a general preference for low attachment, in contrast to both L1 Spanish and L1 Greek. Low attachment might be a parsing strategy that is preferred by L2 learners, even by advanced learners and perhaps independently of their L1 background.

• Scott Jarvis and Aneta Pavlenko, Ohio University. "Conceptual Restructuring in Language Learning: Is there an End State?"

In this paper, we present and discuss evidence of simultaneous L1 and L2 restructuring in data produced by Finnish- and Russian-speaking advanced users of English. We interpret the results as indicating that the end state in second language acquisition--if there is one--is best characterized in terms of multicompetence. The essence of Selinker's (1972) notion of fossilization is that the end state for most second language learners/users will be an internalized rule system that differs from the target system. Despite various criticisms of fossilization (e.g., Huebner, 1985; Long, 1997), the evidence for persistent errors and other deviations from standard language use is indisputable (see Ellis, 1994). Another area of Selinker's work on fossilization that has proven useful is the Multiple Effects Principle (MEP), which accounts well for the finding that persistent deviations (i.e., fossilized or stabilized structures) in an advanced interlanguage nearly always involve crosslinguistic influence, often in conjunction with other simultaneous factors (e.g., Selinker & Lakshmanan, 1992). We agree with the essence of Selinker's claims concerning persistent deviations and the MEP, but argue that the notion of fossilization does not capture the full extent of restructuring that takes place in second language users' linguistic and conceptual systems as they subsume new forms and meanings and attempt to reconcile incompatibilities between their L1 and L2 experience. Our empirical evidence is drawn from longitudinal data collected from one advanced Finnish-speaking user of English, and from cross-sectional data elicited from 22 advanced Russian-speaking users of English. The data show that linguistic and conceptual restructuring are bi-directional, taking place simultaneously in the L1 and the L2. The data support the MEP by indicating that most of the nonstandard deviations in both languages reflect crosslinguistic influence, but they do not support the view of interlanguage as a continuum extending from some beginning point (whether L1 or UG) to native-like competence in the L2, where most second language users fossilize before reaching the end of the continuum. Instead, our data suggest that interlanguage might be characterized more accurately as a continuum from monolingual competence to multicompetence, where the end state of second language acquisition is not comparable to two independent monolingual states (e.g., Cook, 1991, 1992, 1999; Grosjean, 1989; Jarvis, 1998; Pavlenko, 1999).

• Donna Lardiere, Georgetown University. "Is the L2 Steady State Unsteady?"

• Helmut Zobl, Carleton University. "Functional Architecture and Low-level (ultimate) Attainment."

The paper investigates the functional architecture of the grammar underlying the presumably ultimate level of attainment achieved by an adult speaker of Russian who began her informal acquisition of English at the age of 39/40. The talk will focus on several questions in particular: evidence for the absence of a CP and Agr(S)Projection; related to the latter is the question of whether the grammar lacks SPEC projections and employs adjunction instead. While the paper is framed on the oral production of one speaker, it will attempt to make the case for the generalizability of this type of f-architecture to other grammars underlying low-level attainment.

Back to SLRF 2000 Colloquia