The English genitive alternation allows two structures to express adnominal possession: the s-genitive (‘the company’s future’) and the of-genitive (‘the future of the company’). This domain is known to cause problems in L2 learning, which may be attributed to many factors, including: (a) the relative similarity of the target structure to an equivalent structure in the learner’s L1 (surface overlap); (b) the derivational complexity of the target structure itself; and (c) the particular demands of the L2 task.

This thesis investigates the influence of L1 transfer on L2 production, comprehension and processing among L1 Japanese and Spanish speakers in this domain, combining learner corpus data with experimental data to address L1 transfer from multiple perspectives and to provide insight into the role of metalanguage in L2 learning.

The learner corpus analyses indicate a strong influence of derivational complexity in written production, with learners consistently producing the derivationally simpler of-genitive structure more accurately than the complex s-genitive. L1 transfer effects are also observed, with L1 Japanese speakers outperforming L1 Spanish speakers in terms in their production of the s-genitive, a structure analogous to that of the Japanese L1, indicating that surface overlap plays a secondary role to derivational complexity. The experimental study, combining an on-line self-paced reading task with related off-line comprehension questions, also reveals L1 effects motivated by surface overlap, but with no indication of derivational complexity effects. L1 Spanish speakers show higher rates of comprehension in of-genitive contexts, which are equivalent to the structure available in their L1, while reading time measures indicate a processing advantage among L1 Spanish speakers for of-genitives and among L1 Japanese speakers for s-genitives.

L1 differences are more pronounced in the off-line data – the corpus data and the comprehension responses – than in the on-line processing data. Since the off-line tasks, particularly the written examination data, afford learners greater opportunity to focus on linguistic form, it is argued that the degree to which learners can access explicit knowledge may play a crucial role in how L1 effects manifest in certain L2 domains.