Lancaster UniversityGraduate School
Faculty of Arts and Social Sciences

You are here: Home >

Generic vs. specific reference: definitely L1-related?, PaSTY talk by Carmen Ríos García

Date: 7 November 2012 Time: 4.00-5.00 pm

Venue: Bowland North, Seminar Room 9

Recent studies on the acquisition of the English definite article have focused on cross-linguistic differences with regard to the encoding of genericity and specificity, with some focusing on the definiteness vs. specificity axis. Their purpose has been to establish which type of linguistic knowledge - L1-transfer, L2-input, or Universal Grammar (UG) - is activated in L2 acquisition.

This paper examines the results of an online task involving the presence and absence of the definite article which was conducted amongst native speakers of three definiteness-based languages (English, German, and Italian). Reaction times and grammaticality judgements were recorded, in order to test whether or not parametric differences (Longobardi 1994, 2001) affect the parsing of generic and specific nominals by L2 speakers of English. A higher concentration of longer reading times and/or errors in the target sentences on the part of the Italian group would provide evidence of transfer. Should both L2 groups present longer reading times and errors across all categories (grammatical and ungrammatical target and filler sentences), it would be a sign that parameter resetting is not particularly significant and that differences between L2 and L1 performance would be better explained as the result of incomplete acquisition. It was also predicted that all groups would take longer to judge ungrammatical sentences, with the L1 English yielding the shortest times and the L1 Italian, the longest.

The results showed that both L2 groups produced considerably more errors with the set of ungrammatical target sentences, particularly with the specific set. The Italian group presented the highest percentage, which confirms the predictions of the Full Transfer, Full Access Hypothesis (Schwartz and Sprouse 1994) that L1 interference may be operative in the end state. However, since all participants (including those in the control group) produced errors across all sentence types, consideration must be given to the possibility that shallow processing affects not only the end-state performance in L2, but L1 performance as well when generic and specific nominals appear in short sentences.


Who can attend: Anyone


Further information

Organising departments and research centres: Linguistics and English Language


| Home | Who's who? | Research Training | News and Events | Resources |

Graduate School, Faculty of Arts and Social Sciences, Lancaster University, Lancaster LA1 4YD, UK
Tel: +44 (0) 1524 510880 E-mail:
Copyright & Disclaimer | Privacy and Cookies Notice