The Processing of Lexical Stress by Early and Late Learners of Spanish

Abstract

The present investigation examines the processing of Spanish stress by English speakers, second language learners of Spanish, and English-Spanish early bilinguals. Some languages have lexical stress, a phonological feature that can be part of the identity of words. This is the case of English and Spanish, two languages in which stress is contrastive and, thus, it can be used to distinguish between lexical items. Despite this typological similarity, English and Spanish differ in terms of the lexical distribution of stress as well as in the number of possible stress minimal pairs. Stress in English tends to appear in the first syllable of a word in about 90% of the cases. Also, this language presents only a handful of “true” stress minimal pairs in which the segmental structure remains identical. In comparison, stress in Spanish tends to appear in the second-to-last syllable of a word and this is true for about 80% of the lexical items in this language. When it comes to the number of possible stress minimal pairs, the Spanish verbal morphology alone is highly productive in terms of creating lexical items that differ exclusively in the location of stress. Thus, when compared to English, stress in Spanish appears to be less predictable and it generates a higher number of stress minimal pairs. It is not surprising, then, that Spanish speakers make use of stress to conduct lexical searches to a greater extent than English speakers. In fact, English speakers have been found to treat the members of a stress minimal pair as homophones. When it comes to learning Spanish as a second language, English native speakers present difficulties in the processing of L2 stress contrasts. Furthermore, these processing difficulties are also found in English-Spanish bilinguals, a population that has had extended experience with both English and Spanish. To date, however, it is unknown whether the challenges that English native speakers second language learners of Spanish and English-Spanish bilinguals face stem from unsuccessful perceptual discrimination of stress contrasts, from an unstable phonological representation of stress, or a combination of both. The aim of this dissertation is twofold. The first aim is to provide data on the phonological encoding of stress in early and late learners of Spanish by probing them with stress-based contrasts in a discrimination task (Experiment 1). The findings from such a task will help us capture whether participants’ processing abilities allow them to successfully encode and access stress information in a relatively cognitively-demanding online tasks. Furthermore, this experiment will allow us to explore whether second language learners are able to develop their stress processing as they become more proficient and their experience with Spanish increases. Secondly, this dissertation addresses the perceptual discrimination of stress contrasts by second language and bilingual speakers of Spanish. Subjects participated in a discrimination task designed to test whether participants were able to perceive a difference when presented with aural productions of two nonwords that differed exclusively on the location of stress (Experiment 2). The results from Experiment 1 suggest that English speakers are challenged by stress processing when this phonological feature is signaled exclusively through suprasegmental acoustic cues but only when they are, in theory, asked to rely on the phonological representation of lexical items. This effect is also found in early Spanish-English bilinguals, which indicates that length of exposure to and increased linguistic experience with Spanish do not guarantee the development of robust stress processing abilities. Moreover, the findings from second language learners suggest that proficiency can predict some degree of the variance found in the late learners. This provides evidence that learners with increased linguistic knowledge display a more robust representation of stress in their working-memory when compared to more novice learners. The results from Experiment 2 revealed that second language and bilingual speakers of Spanish were challenged by stress contrasts at a perceptual, low level of processing. In addition, language dominance and proficiency were found to be very modest predictors of perceptual development. Overall, it seems that the difficulties observed in English native speakers who are second language learners of Spanish and English-Spanish bilingual speakers are (at least in part) due to difficulties in the perception and phonological encoding of stress contrasts. In the case of second language learners, it appears that a development in L2 proficiency leads to more robust stress-encoding abilities, and to a very modest development in the perception of stress contrasts. Finally, the results of the present study are in line with the predictions of the lexical statistics account (Peperkamp et al., 2010), which anticipated lower stress-encoding abilities for speakers of English as a result of the higher predictability of the location of stress in this language.