Bilingual encoding strategies during the production of motion event utterances
Item statusRestricted Access
Embargo end date28/06/2023
Morales Martínez, Matías
When describing motion events, English speakers tend to encode the manner of motion in the verb (e.g., A penguin is skiing into an igloo), whereas Spanish speakers tend to express the path or trajectory of motion in the same position (e.g., Un pingüino está entrando a un igloo, A penguin is entering an igloo). This presents a challenge for bilinguals who speak languages that differ in these encoding preferences, such as Spanish-English bilinguals. What encoding strategies do these bilinguals use during the production of utterances in their non-native language (L2)? Do they use encoding preferences associated with their L2? Does their native language (L1) knowledge interfere with their L2 utterance preparation? To address these questions, I investigated how Spanish-English bilinguals (and English monolingual controls - L1 English) describe motion events in Spanish (L1 Spanish) and English (L2 English), with an interest in both the choices they make, and the online processes that underly these choices. In Experiment 1 I examined the encoding patterns of spontaneous utterance production. L2 English speakers, who resided in Spain at the moment of the experiment, tended to use encoding strategies associated with their L2, but not as frequently as L1 English speakers, thus suggesting transfer effects from their L1 knowledge into their L2 production. Experiments 2-4 used priming methods in the same population of L2 speakers as in Experiment 1. In Experiment 2, L2 speakers tended to use more manner-oriented encoding strategies after manner prime sentences than after path prime sentences that contained verbs participants could reuse in their target responses. This suggested that pre-activation of manner representations facilitated the production of target lexical representations about manner during L2 production. Experiment 3 showed that this priming effect was located at the lexical level, and not at the conceptual level as L2 English speakers did not exhibit priming effects when primes did not contain verbs that participants could reuse in their target responses. Importantly, Experiment 4 demonstrated that the priming effect was not solely due to lexical repetition, but also implicated a conceptual component in which bilinguals chose which information to select and encode in their sentence verbs. In Experiments 5 and 6, I replicated the pattern of results in Experiments 1 and 3, respectively. As these replications tested L2 English speakers residing in the UK at the moment of participation, they suggested that patterns of results in Experiments 1 and 3 are independent of the dominant language used in their environment. In Experiment 7, I examined the pre-articulatory looking patterns to information associated with manner (e.g., the skis) or path (e.g., the igloo). L2 English speakers tended to look to manner targets more frequently than both L1 English and L1 Spanish speakers before utterance onset. As less proficient bilinguals exhibited more preferential looks to manner than more proficient bilinguals, these results seem to reflect difficulty in processing manner information due to the availability of these lexical representations in their non-native lexicon. Additionally, L2 English speakers exhibited different patterns in looks to path compared with L1 Spanish speakers at early and late stages of utterance planning, thus indicating processing differences in bilinguals on the basis of the language of use. Finally, I report an experiment (Experiment 8) that set out to examine anticipatory effects during sentence processing in bilinguals. I discuss methodological issues that complicate the interpretation of the results and consider possibilities for future research to address these issues. Together, these results suggest that non-native speakers use underlying encoding processes associated with their L2.