Honors Capstone Project
Date of Submission
Languages, Literatures, and Linguistics
Arts and Science
Capstone Prize Winner
Won Capstone Funding
Other Languages, Societies, and Cultures | Spanish and Portuguese Language and Literature | Spanish Linguistics | Spanish Literature
This study further examines past studies that have examined the gender marking value of a parameter of variation allowed by UG in L1 and L2 speakers of French and Spanish. Child L1 French speakers were found to make few errors in assigning gender to both actual nouns and nonsense nouns. An L2 Spanish learner was found to default one gender for the definite article and another gender for the indefinite article. Adult L2 French speakers were found to display similar defaulting, to a lesser degree, perhaps attributed to experience with exceptions to patterns.
To enhance these results, I conducted a study of 12 L1 Spanish speakers and 21 L2 Spanish speakers and gave them the same three tasks. The first survey asked speakers to properly identify the gendered definite article (el/la) that was required by the given noun. In the second survey, speakers were given a list of stems and were asked to properly identify the derived form (using the suffix ‘-ero/-era’) that was required by each stem. The third survey was similar in format to the second survey, except nonsense stems and derived forms were given. In each survey, speakers were also asked to indicate their level of certainty for each listed item.
When making errors in the first survey, L1 speakers defaulted the feature [+masc] 23.5% of the time as opposed to the feature [-masc] only 3% of the time. Low frequency, at first, seemed to be a plausible explanation, but was not able to be applied consistently throughout the results of the survey. L2 speakers displayed a similar behavior, but to a lesser extent.
In the second survey, when L1 speakers made errors [+masc] was defaulted more than [-masc]. Individual L1 speakers showed overt cases of [+masc] defaulting, with one case of overt [-masc] defaulting. L2 speakers displayed a strong pattern of matching the gender feature of the stem with the gender of the derived form. Individual L2 speakers displayed strong behaviors of matching, asymmetrical matching, and defaulting [+masc].
With the assignment of probable gender (determined by analogy to similar words in corpus), L2 speakers exhibited the same pattern of matching genders of stems to derived forms. L1 speakers, as they had throughout the study, overgeneralized [+masc] more than [-masc].
The results of the study seemed to support and disprove UG. If L1 Spanish speakers acquire the gender-marking value of the parameter during childhood, it is expected that the accuracy born from the value would be higher. However, L1 speakers did seem to have an advantage, outperforming L2 speakers in all three tasks. The fact that L1 speakers did not display behaviors of matching is strong evidence that there is more of a need to satisfy and actual gender-marking parameter rather than simply choosing a gender feature.
With their relatively poorer performance and various behaviors, L2 speakers did seem to exhibit problems with the gender-marking value. With respect to UG, it can be concluded that L2 speakers are unable to reset the value of the parameter in their L2 (Spanish) if the same value does not exist in the L1 (English).
Kanemura, Marissa T.Y.O, "Spanish Language Acquisition: Gender-Marking Parameter" (2009). Honors Capstone Projects - All. 432.
Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 License.