The present study examines real time processing data from children’s subject and object relative clauses in Greek as a native language in a majority (monolingual) context and in a minority context where Greek is the heritage language. Although the number of studies investigating heritage children’s online parsing strategies has been emerging in recent years most of the studies focus on lexical processing while studies on morphosyntactic processing are up to date quite limited (e.g. Montrul 2006; Foote 2010; Keeting et al. 2016). The investigation of heritage children’s morphosyntactic parsing strategies can shed light into the question of whether or not successive bilingual children have the same language processing strategies as adults (continuity of parsing hypothesis Clahsen & Felser 2006) or similar to L2 learners use shallow processing (Felser & Cunnings2012; Felser & Roberts2007; Marinis et al. 2005) due to their bilingual experience. The main aim of the present study is thus to examine the mechanisms that underlie monolingual and heritage children’s processing strategies. We conducted a self-paced listening experiment in Greek in which we manipulated the type of relative clause (SRC vs. ORC) and the RC internal word order (canonical vs. scrambled). In total we recorded online listening times and grammaticality judgments from 40 11- to 12-year-old children 22 of which were monolingually raised and 18 were raised in an environment where German is the societal language and Greek the heritage language. The children’s data were compared to those of 100 adult native speakers of Greek (age range 20-32 mean age: 21.8 SD: 3.7). The experimental sentences included subject and object RCs introduced with the complementizer pu (‘that’) with two types of word order (local: O majiras-NOM pu esprokse ton servitoro-ACC ekapse to fajita vs. non-local: O majiras-NOM pu ton servitoro-ACC esprokse ekapse to fajito ‘The cook who pushed the waiter burned the food’). Each sentence was presented in a segment-by-segment fashion and in the end of every sentence participants were asked to judge the grammaticality of the sentence by pressing one of two buttons. The statistical analysis of the listening times showed that heritage children processed relative clauses similarly to monolingual children. Similarly to adults children showed no overall preference for SRs over ORs. This result reflects findings from adult data in two other free word order languages (Levy et al. 2013; Kovács & Vasishth 2013). In contrast to adults however children processed non-local structures faster than local ones. This originally surprising result can be accounted for based on evidence from the developmental literature. According to MacWhinney and Pléh (1988) RCs are not fully mastered until adolescence and this may be the reason why 11 to 12 year old children did not process the different RC word orders in the online self-paced listening task similarly to adults. It seems then that monolingual and heritage language 11 to 12 year old children processed the most difficult non-local structures in a more “shallow” way than the local ones thus employing an L2 rather than an L1 processing strategy.

View poster here.