2018 Friday Session C 1645

Friday, November 2, 2018 | Session C, Terrace Lounge | 4:45pm

Late L1 Attrition is Temporary: Evidence from a Longitudinal Case Study
D. Genevska-Hanke

This longitudinal study of an L1-Bulgarian/near-native L2-German speaker (long-term German resident) investigates late/post-puberty L1-attrition at the syntax-discourse interface by studying the use of pronominal subjects in a pro-drop (L1)/non-pro-drop but semi-null subject and topic-drop (L2) constellation, which has not been studied in attrition research so far. Being a semi-null subject and a topic-drop language, German generally allows more null subjects than English (Haegeman 2013, Trutkowski 2011, Roberts and Holmberg 2010). The alternation of overt and null subjects in pro-drop languages depends on the discourse notions topic and focus and is thus subject to interface conditions, not purely grammatically-driven. Interface syntax has been claimed problematic in cases of similar language combinations and the grammatical phenomenon in question, namely for attrited L1-Italian/L2-English and L1-English/near-native L2- Italian speakers (Sorace & Filiaci 2006) and L1-Greek/L2-English/L2-Swedish/L2- German (Tsimpli 2007). Both near-native L2 speakers and attriters overused overt pronominal subjects (OS) in topic-continuity contexts when compared to non-attrited speakers. However, recent studies indicate that post-puberty/late attrition is temporary since L1 knowledge of the kind can be reactivated after short re-exposure to L1 (Chamorro, Sorace & Sturt 2016, Genevska-Hanke 2017). Ghamorro et al. tested L1- Spanish L2-English speakers on their knowledge of overt vs. null subjects in an offline judgment and an online eye-tracking tasks. Genevska-Hanke (2017) used spontaneous speech production to test the use of overt vs. null subjects for the language combination L1-Bulgarian L2-German. The results of both studies show attrition effects only before reexposure to massive L1 input.

We analyzed four recordings of spontaneous speech (125 utterances each, see graphs below). Only the recording in the target country (TC) at investigation-point 1 showed an OS rate (47%) significantly higher than those of non-attrited controls presumably because L1 exposure in the TC is limited (p = .004, Crawford & Garthwaite 2002). After a three-week re-exposure to L1 in the home country (HC) attrition effects disappeared and the rate (34%) fell within the normally-distributed monolingual range. These results at investigation point 1 after twelve years abroad were compared to results obtained five years later at investigation point 2. Both rates, TC (29%) and HC (24%), were monolingual-like, presumably due to increase of L1 use in Germany.

The findings suggest that late/post-puberty L1 attrition is temporary for the domain  of subject use in question and also support assumptions on the stability of fully-developed L1s (Schmid & Köpke 2007). They further corroborate the dissociation between near- native L2 and attrited L1 knowledge since only the latter becomes indistinguishable from non-attrited monolingual competence after reactivation. Hence, Interface Hypothesis predictions (Sorace 2005) probably only hold for near-native L2, which is in-line with behavioral and neuro-cognitive results (Hyltenstam & Abrahamsson 2009, Felser & Clahsen 2009, Batterink & Neville 2013).

The temporariness of attrition can possibly be attributed to limited L1 access/different processing under pervasive L2 influence. Thus language background in relation to language mode is a major factor in late L1-attrition (Paradis 2007, Grosjean 2013).