Cross-Linguistic Influence in Third Language Acquisition: The Case of Spanish-English Bilinguals' Acquisition of Portuguese

Article excerpt


This study investigates typological distance and order of acquisition (i.e., the order in which languages were acquired) in the context in which Spanish-English bilingual students, whose first language is English or Spanish, are learning Portuguese as a third language (L3). Participants were asked to think aloud as they worked on pedagogical tasks involving the present and future subjunctive in Portuguese, followed by stimulated recalls. Findings indicate that both groups rely heavily on Spanish when performing both tasks, leading us to conclude that linguistic similarity between the languages overrides order of acquisition. Data derived from learners' reflections elucidated some differences in transfer patterns between the groups that would not be revealed otherwise, contributing to a better understanding of these variables in L3 acquisition in general and the acquisition of Portuguese by Spanish speakers in particular.

Key words: cross-linguistic influence, subjunctive, think-aloud protocol, third language (L3) acquisition, transfer

Languages: English, Portuguese, Spanish


Although cross-linguistic influence has been widely discussed and investigated in the field of second language (L2) learning, its role in third language (L3) acquisition still remains relatively unexplored (Cenoz, Hufeisen, &Jessner, 2001). The few available studies that do exist have raised questions as to learners' preference for the first language (L1) or the L2, as the source of influence in the learning of the L3. Some of the factors that have been shown to predict the relative weight of cross-linguistic influence in the speakers' L3 acquisition are typological distance, recency, proficiency in L3, and order of acquisition (i.e., the order in which languages were acquired) (Cenoz, 2001; Odlin, 2003). In this article, we continue this line of investigation by looking into the acquisition of Portuguese by two groups of students: one whose first language is English and the other whose first language is Spanish. Portuguese is genetically very closely related to Spanish, but relatively distant from English.

Cross-Linguistic Influence in L3 Acquisition

Several studies have demonstrated that language distance does indeed play an important role in L3 acquisition (Ahukanna, Lund, & Gentile, 1981; Cenoz, 2001; De Angelis & Selinker, 2001; Hammarberg, 2001; Singleton, 1987; Williams & Hammarberg, 1998). They have also indicated that psychotypological distance, or, as proposed by Kellerman (1983), learners' perception of cross-linguistic relatedness, plays a defining role in conditioning Ll and L2 transfer patterns. However, other factors also seem to play a role. Recency, order of acquisition, amount of experience with the L3, the number of previously acquired languages other than the first, age, and level of metalinguistic awareness all appear to interact with language distance.

Ahukanna et al. (1981) assessed interference from two previously learned languages (English was learners' L2, and Igbo, an African language, was their L1) in learners' acquisition of French. Their results showed that English caused more interference than Igbo. These results are supported by Ringbom's (1987) study of Finns and Swedes learning English. He also found that despite years of study by the Swedes, Finnish did not cause cross-linguistic interference in their English acquisition, but Swedish, a language more similar, did. Singleton (1987) conducted a case study to investigate the relationship between one learner's perception of the degree of typological relatedness between L1 (English), L2s (Irish, Latin, and Spanish) and L3 (French). Data from self-report (stimulated recalls) revealed that the learner was indeed sensitive to the relatedness between Spanish, one of the previous learned languages, and French, the target language. She also found that, although Latin was closer to French than English, English was more prominent, possibly because the proficiency factor overrides distance. …