Report

Teaser, summary, work performed and final results

Periodic Reporting for period 1 - BilProcess (French-English bilinguals\' processing of morphosyntactic structures: the case of cross-linguistic transfer)

Teaser

In context of bilingualism, language contact leads to occasional interactions between languages, a phenomenon known as cross-linguistic influence (CLI). It is widely attested in simultaneous bilinguals (2L1), i.e. individuals who were first exposed to two languages from...

Summary

In context of bilingualism, language contact leads to occasional interactions between languages, a phenomenon known as cross-linguistic influence (CLI). It is widely attested in simultaneous bilinguals (2L1), i.e. individuals who were first exposed to two languages from infancy, and in second language learners (L2), i.e. individuals who learnt a second or third language after the age of three.
The BilProcess project examines the intra- (i.e. sentence processing, implicit vs. explicit knowledge, noun type) and extra-linguistic (i.e. age, language dominance) mechanisms at play on CLI in school-aged bilingual children and adult second language learners.
The project focuses on bilinguals’ comprehension and production of determiners (e.g. the, a, an) which are known to be particularly challenging to acquire for monolingual and bilingual speakers of Romance and Germanic languages. However, no consensus exists on the nature of the difficulty of acquiring the features that govern the determiner phrase. French and English are interesting test case as English allows overt and null determiners (the apples vs. apples) when determiners are obligatory in argument position in French (les pommes).
The originality of BilProcess is that it sets to test simultaneous bilingual children and French adult learners of English with the same experimental materials in order to assess whether (i) the age of acquisition of the language (as 2L1 or as L2), and (ii) the relative dominance of the language (stronger vs. weaker language) affect CLI during sentence comprehension and production. Moreover, the project is one of the first to also use a combination of offline (i.e. timed) and online (i.e. untimed) measures of school-aged bilingual children’s sentence comprehension and production to examine the impact of their implicit (i.e. intuitive) and explicit (i.e. conscious) linguistic knowledge on their performances in their two languages.
This project contributes theoretically and methodologically to the field of bilingual development. The results also provide important insights by highlighting to language teachers and speech pathologists the grammatical features of determiner use that are actually difficult to learn for bilingual children and adult second language learners of English and French.

Work performed

The BilProcess project was conducted with activities and tasks broken down into 6 work-packages and an additional seventh WP which was added in July 2017. These WP correspond to a breakdown of the above-mentioned research questions. Bilingual children and French adult learners of English were tested experimentally on their comprehension and production of determiners in French and English.

Across packages, materials for a total of 5 experiments were created in both languages. In the UK and France, 87 children were tested in 4 schools and 146 adult participants were tested in 3 universities. All the participants took part in all the tasks in their respective language(s).

WP3 indicates that L2 learners are influenced by their L1 whilst learning their L2 at the determiner level. French learners of English over-accept ungrammatical THE+NOUN, especially with mass nouns (e.g. milk, water). But, they actually identify at intermediate level the correct article use with plural nouns (e.g. apples). Differences across tasks shows that intermediate learners have not yet automatized their explicit knowledge of article use. But, advanced learners have formed some implicit knowledge of article use as they perform well in the task tapping into implicit knowledge.

WP4 shows that French-English bilingual children are vulnerable to cross-linguistic influence in both direction in their interpretation of determiner phrases. In English, they, like adults, have difficulties with mass nouns but not with plural nouns. Reduced English input increases this effect. School differences indicate that the instruction of grammar can compensate this reduced input. Performance increases with age indicating that determiner use is not yet fully settled between 8 and 10. In French, the bilinguals display greater flexibility at accepting ungrammatical sentences in the explicit task than in the implicit task indicating that they have fully acquired French article use but are consciously more flexible to speech errors.

WP5 involved a comparison of the outcomes of WP3-4. But it has been delayed by the work carried out in WP7.

WP7 indicates that cross-linguistic influence was observed in the adults and children’s production. Noun type (mass nouns vs. plural nouns) affects our participants’ production as they are less accurate with mass nouns (e.g. milk) than plural nouns (e.g. apples). However, accuracy increases with increased exposure to the target language and increased proficiency.

WP6 corresponds to the impact section. The fellow steered out an application for her host lab to become a branch of Bilingualism Matters an outreach centre that provides information about bilingualism to parents, school and language professionals in the south east of England. As director of the branch, she organised (a) regular parent night information sessions, (b) parent workshops in primary schools, and (c) a series of informative short talks as part of the ESRC festival of social sciences. She also represented the branch at several language festivals and published general information on social medias. As for scientific dissemination, she posted several blog entries on research methods aimed at postgraduates.

Final results

Each work-package has provided clear results towards the aims to understand better the phenomenon of cross-linguistic influence at the determiner level in French-English bilinguals.
-It is among the first studies to consider how both implicit (intuitive) and explicit (conscious) linguistic knowledge affect language interaction in school-aged bilingual children. This is a hot debate in the literature on L2 development but had never been investigated in child bilingual development.
-It provides a complete overview of French-English bilinguals’ comprehension and production of article use in light of their language exposure, proficiency skills and age.

Bilprocess shows that cross-linguistic influence can be bi-directional and can occur in different directions depending on the kind of linguistic knowledge that is tapped into (implicit vs. explicit).
Article use is not fully acquired by age 10 even in English monolingual children.
Children and adult learners’ difficulty with English article use is related to difficulties at encoding mass nouns (e.g. milk).
Language exposure affects the participants’ comprehension whilst their language exposure and their proficiency skills play a role on their production.
Crucially for the bilingual children, reduced exposure to the target language can be compensated by formal grammar instruction at school as shown by group differences based on the bilingual schools that had different take on grammar instruction.

The results of BilProcess highlight one of the major difficulties for bilingual speakers of English to use articles, an extremely high frequency structure in any kind of speech. This result is important for language teachers and speech pathologists to help children and adult learners of English overcome difficulties at using this structure accurately. Another practical implication of this project concerns the importance of grammar instruction for children being raised bilingually.

Website & more info

More info: https://coralieherve.wordpress.com/.