Docteurs récents
Status : Doctorante
Address :
LLF, CNRS – UMR 7110
Université Paris Diderot-Paris 7
Case 7031 – 5, rue Thomas Mann,
75205 Paris cedex 13
E-mail : fnvqn.ybhpvs@lnubb.se
Title : Transfer phenomena in the acquisition of grammatical gender in French as a second language
PhD Defense : 2023-12-07
Inscription : 2015 à Université Paris-Cité
Jury :
Abstract :
In second language (SL) some grammatical properties seem to be more difficult to acquire than others, such as gender. Indeed, according to some researchers such as Franceschina (2005) and Sabourin et al. (2006), the complexity of acquiring this grammatical feature is due to the impact of the native language (NL). In other words, the success of this acquisition depends on the proximity between the gender system in NL and that in SL. Thus, similar gender systems would favor a more optimal transfer of the gender category from NL to SL and consequently allow better mastery in SL. On the other hand, different gender systems in NL and SL could represent an obstacle for this acquisition.
However, research on the impact of NL on gender agreement in SL is far from presenting a consensus. Through the work reported in this thesis, we participate in this discussion by providing new experimental data in order to enrich this debate.
Our research focuses on the transfer phenomena in the acquisition of grammatical gender in French as a second language (FSL). Studies on this subject are relatively few (Pieters, 2018; Foucart 2008). We conducted an experimental study on the acquisition of this grammatical feature in FSL by speakers of different native languages: Arabic (gender system similar to that of French) and Chinese (genderless), two audiences little studied in the literature, in order to know to what extent, the similarities between the gender system in LM and in LS or the total absence of it in LM affect the ability of FSL learners to choose the correct gender of nouns in French. The main objective of our research is to examine the impact of NL as well as the different factors that can interact with the acquisition of grammatical gender in FSL. Our main hypothesis assumes that there are transfer phenomena from NL to SL that facilitate acquisition for some of our subjects (those whose NL has a gender category comparable to that of French). Moreover, we assume that the intensity and quality of FSL input influences this transfer (Cornips and Hulk 2008). We also examine the hypothesis of Dewaele (2015) which suggests that the gender agreement of French nouns is more complex with nouns that begin with a vowel than with those that begin with a consonant, due to the elision of the vowel of the definite determiner which does not allow a gender distinction to be made.
We tested these hypotheses by comparing our different groups of NLs and one control group of French native speakers, via two experimental tasks, one linguistic (speech production) and the other behavioral (physical eye-tracking measurements), by investigating the competence of our participating learners to choose the correct gender for French noun phrases in an oral task, and by strictly controlling the variables related to the representation of gender in NL.
The results obtained showed a significant effect of the transfer of the NL, allowing our Arabic-speaking participants to achieve a highly better performance compared to Chinese speakers, both in comprehension and production, and this is observed at different stages of acquisition. Our results agree with the conclusions of Sabourin et al. (2006) and Pieters (2018): the more similar the gender systems are in NL and in SL, the more gender agreement is mastered in SL. Our experimental data also revealed that the learner's skill level in FSL limits this transfer and that it is more present at the early stages of acquisition than at an advanced stage.