A produção e a percepção de formas alofônicas fricativas no processo de aquisição do espanhol por falantes nativos do português brasileiro

Abstract

When acquiring a Second Language (L2), the learner has to develop a new grammar alongside the grammar of the Mother Tongue (LM), which requires him/her to perceive and produce not only phonemes and phonological rules different from the grammar of the LM, but also allophones that integrate the phonetic/phonological continuum of L2. In this scenario, the present research seeks to identify, describe and explain the func-tioning of the perception and production of alofones of the Spanish language by Bra-zilian students. The focus is on the alophonic forms [β], [], [], which can represent respectively the phonemes /b/, /d/, /g/ in contexts where they emerge in Spanish, as well as the alophones [h]~Ø, referring to the segment /s/ in syllabic coda. Faced with the complexity resulting from the mismatch between the use of allophonic forms in Spanish and Brazilian Portuguese (BP), the guiding question of this research asks how the process of acquisition, by native speakers of BP, teachers of Spanish in training, of these alophonic forms, present in the Spanish language, but not in BP, occurs. The empirical data which supported the study resulted from the production and perception of the alophonic forms which represent /b/, /d/, /g/ in syllabic onset and /s/ in syllabic coda, resulting from interviews with 10 Brazilian learners of Spanish as L2, all students of the 1st year of the Degree in Languages - Portuguese-Spanish of a university in Southern Brazil. The methodology adopted for the investigation of perception and pho-netic-phonological production established dialogue with the Second Language Linguis-tic Perception Model (L2LP), proposed by Escudero (2005), and with Autosegmental Theory (CLEMENTS & HUME 1995) - these were the theoretical guidelines that of-fered the subsidies for the analysis of the results. To obtain the data, five production tests and seven perception tests (identification tests and discrimination tests) were de-veloped. They were organized and applied in the TP software, by Rauber, Rato, Kluge, & Santos (2012). After being submitted to statistical analysis, all datum were described and analyzed. The linguistic production data was also subjected to acoustic inspection with the support of Praat Software (BOERSMA; WEENINK, 2012). Among the results obtained with the data analysis, these points deserve to be high-lighted: (a) the Brazilian apprentices in Spanish showed a low rate of accuracy in the perception of the alophonic forms that can represent voice plosives (in syllabic onset) and coronal fricative (in syllabic coda) in this language, showing better performance in discrimination tests than in identification tests; Regarding the production, there was a low rate of use of the fricative allophonic forms [, , ] ] to represent the voiced plo-sives and, in the case of the fricative /s/ in syllable coda, there were few productions of the aspirated form [h]. The acoustic analysis of the forms produced by Brazilians showed that some parameters are already distant from phonetic patterns characteristic of BP, but still do not identify with the phonetic patterns characteristic of Spanish. (b) Regarding the correlation between perception and production, considering the allophony that represents the fricative /s/ in syllabic coda, when the Spearman's cor-relation test was applied, the significance of p 0.062 between the aspirated production and the identification of phonetic zero was verified, and p 0.078 between the aspirated production and the identification of [h]. Thus, there is correlation between the percep-tion and the production of [h], which has been confirmed as the most complex phonetic form for Brazilians, among those that can represent the fricative /s/. Regarding the alophonic forms capable of representing the voice plosives /b, d, g/, the examination of the correlation between production and perception pointed only to the representation of the segment /d/ a negative correlation, but marginally significant. (c) Regarding data analysis in the light of the L2LP Model (ESCUDERO, 2005), it was agreed with the author that there may be a Complete Copy of the perceptual mapping of L1, when the apprentice of L2 is in what is called here 1st Level of Perceptual Mapping. However, although Brazilian students showed low hit rates on the Identification type perception tests, they showed higher hit rates on the Discrimination type perception tests. This result, associated with the presence, in alophonic form productions, of alterations in the acoustic parameters of BP (L1 of informants), was interpreted as the beginning of the formation of phonetic categories which integrate these sounds which are allo-phones of Spanish, but still without their consistent link to the proper phonological cat-egory of L2. Considering the process of acquisition of Spanish allophones, Brazilian learners of this language were classified in a 2nd Perceptual Mapping Level, since they showed a mismatch of limits in perception tests, which were reflected in production, with a mismatch of phonetic properties in relation to the parameters of their L1, without, however, reaching the parameters of L2. The results indicate that the Brazilian appren-tices in Spanish, when examining the alophonic forms studied, are in the stage of con-structing phonetic categories which integrate allophones in order to subsequently es-tablish the phonological categories of L2.