The Language Switching Of Proficient Cantonese-Mandarin Bilinguals In Numeral Naming Task

Objective:Using numeral naming task to explore the language switching costs of proficient Cantonese-Mandarin bilinguals during language production, and to find out the relevant factors and mechanisms.Methods:Sixteen proficient Cantonese-Mandarin bilinguals are volunteered to take part in the experiment, including eight women and eight men with an average age of 23.7±2.9years.The experiment design is 2(two factors) × 2(two levels).One factor is target language(Cantonese, Mandarin),the other is switching type(switching, non-switching).The stimuli were single digits ranging from 1 to 9. A total of 240 trials were presented per participant with a fully predictable sequence of two-trial alternating runs. Participants were told that single numerals would appear on the computer screen preceded by a cue, including a single word “粤”or “普”,indicating which language they should use in numeral naming task. Once presented, they should name the numerals as quickly, accurately and clearly as possible in the appropriate language, as indicated by the cue, and press the button at the same time. Once press the button, the single digit would disappear and go to next trial after 300 ms intervals. Response time were determined from the onset of the visual stimulus to the onset of the button-press. Response time was recorded by E-prime and vocal responses were recorded by experimenter using a recorder so as to check the errors.Response time corresponding to incorrect responses or response time higher than 1500 ms were excluded from subsequent analysis as was correct responses on trials immediately following an error. Subjects with error rates higher than 20% would be rejected as well.Data were analyzed later by repeated measure analysis of variance in SPSS13.0 statistical software.Results:1. The variance analysis of response time: the main effect of target language was not significant(F=0.611, P>0.05), so was the switching type(F=2.977, P>0.05). Therewas no interaction between target language and switching type(F=0.000,P>0.05).2. The variance analysis of naming error rates : the main effect of target language was not significant(F=0.811, P>0.05), so was the switching type(F=0.074, P>0.05).There was no interaction between target language and switching type(F=0.298,P>0.05).3. The paired t-test of response time under switching condition showed no significance(t=2.005, P>0.05); It was significant under non-switching condition(t=2.305,P=0.036)with 22 ms faster in naming in L1 than that in L2.4. The paired t-test of naming error rates under switching condition(t=1.077,P>0.05) and non-switching condition(t=0.034, P>0.05) showed no significance.Conclusion:1. The language switching costs are symmetric in proficient Cantonese-Mandarin bilinguals when in numeral naming task.2. That the language switching costs are symmetric in proficient CantoneseMandarin bilinguals when naming numerals may support the “ Language-specific selection models” hypothesis.