Based on the syntactic analysis of tense (Zagona 1990, Stowell 1993, among others), Miyamoto (1994) argues that the particular overt realization of tense in temporal adverbial clauses in Japanese, as exemplified in (1a-d), results from the hypothesis that these clauses can appear in VP. Further, it has been observed that a WH-phrase inside a temporal adverbial clause can be questioned, unlike its English counterpart. Thus, the contrast in (2) shows up. This fact can also be naturally accommodated under the hypothesis that temporal phrases appear in VP in Japanese. Since temporal adverbial clauses appear in VP and receive an argument status (Murasugi and Saito, 1992), extraction of a WH-phrase in LF to the matrix CP SPEC does not violate the Empty Category Principle (ECP).
This study examined (i) whether American JFL learners use correct tense forms within Japanese temporal adverbial clauses and (ii) whether they can answer to a WH-phrase inside a Japanese temporal adverbial clause. Two grammaticality judgement tasks were conducted with American JFL learners. The written task consisted of 40 multiple choice questions with the fill-in-the-blank format. In the oral task, the subjects were asked to judge the grammaticality of 57 sentences that described pictures. Given Universal Grammar (UG), our hypothesis was that American JFL learners correctly understand the tense system of Japanese and use correct tense forms, and also can answer to a WH-phrase in temporal adverbial clauses of their L2 (namely, Japanese), early in the course of L2 development.
Our initial hypothesis was, however, not quite supported. We found that in the early stage of L2 acquisition, American JFL learners observe the English tense system and cannot answer to a WH-phrase inside a temporal adverbial clause. These results indicate that their grammar is, in fact, still English. However, those who are in the later stage observe the Japanese tense system and can answer to a WH-phrase inside a temporal adverbial clause. Their grammar has become Japanese at this point.
These results indicate that in the early stage of acquisition of Japanese, temporal adverbial clauses adjoin to TP. This suggests that in the early stage, transfer takes place with respect to the position of temporal adverbial clauses. However, due to positive evidence, JFL learners acquire that temporal adverbial clauses can adjoin to VP in Japanese. In other words, in the later stage, their tense system becomes Japanese, and the argument status of temporal adverbial clauses is acquired, and thus, they can answer to a WH-phrase inside a temporal adverbial clause. We take this as a piece of empirical evidence for Hoji (1985), Koizumi (1991) and Miyamoto (1994) who propose that temporal adverbial clauses are adjoined to VP in Japanese, and also for Murasugi and Saito (1992) who argue that temporal phrases are arguments of V or INFL. The important observation that we made is that there was a correlation between acquisition of the Japanese tense system and acquisition of the argument status of temporal adverbial clauses. Those who have not acquired the Japanese tense system cannot answer to a WH-phrase inside a temporal adverbial clause, while those who have mastered the Japanese tense system can do so. If the tense system and the questionability of WH-phrases inside temporal adverbial clauses were two independent phenomena, the correlation observed in this study would be mysterious. We suggest that this correlation can be best accounted for if we assume that Universal Grammar is operative in second language acquisition (SLA) (e.g., White 1985, Toratani 1995, and Yamane 1995). As mentioned above, under UG, these two different properties of English and Japanese temporal adverbial clauses result from the different positions which temporal adverbial clauses of these two languages occupy. Then, the only evidence the JFL learners need is that the position of temporal phrases in Japanese is in VP. If learners obtain a piece of evidence concerning the position of temporal adverbial clauses in Japanese, these differences between English and Japanese temporal adverbial clauses follow from UG. Thus, the correlation is correctly predicted.
In sum, this SLA study provides another piece of evidence for transfer and at the same time, gives a piece of empirical evidence for UG in SLA.
Examples
(1) a. John comes before Mary comes/*came.
b. [Mary-ga kuru/*kita]-mae-ni John-ga kuru.
-nom come/came before -nom come
'John comes before Mary comes.'
c. John came before Mary *comes/came.
d. [Mary-ga kuru/*kita]-mae-ni John-ga kita.
-nom come/came before -nom came
'John came before Mary came.'
(2) a. *How1 did Mary get angry [before John fixed the car t1]?
b. Mary-ga [John-ga doo kuruma-o naosu]-mae-ni okotta-no.
-nom -nom how car-acc fix before got angry-Q
'(lit.) Q:Mary got angry [before John fixed the car how].'
References
Hoji, H. 1985. Logical Form Constraints and Configurational Structure in
Japanese. Ph.D. dissertation, University of Washington.
Koizumi, M. 1991. Syntax of Adjuncts and the Phrase Structure of
Japanese. MA thesis, Ohio State University.
Miyamoto, Y. 1994. "Temporal Adverbials in Japanese," presented at the
23rd Western Conference on Linguistics.
Murasugi, K. and M. Saito. 1992. "Quasi-Adjuncts as Sentential Arguments,"
presented at the 22nd Western Conference on Linguistics.
Stowell, T. 1993. "The Syntax of Tense," ms., UCLA.
Toratani, K. 1995. Strong Crossover in Second Language Acquisition: Adult
Japanese ESL Learners' Knowledge of Principle C. MA Thesis, Ohio
University.
White, L. 1989. "Is There a Logical Problem of Second Language Acquisition?"
TESL Canada 2-2.
Yamane, M. 1995. The Scope Principle in Adult L2 Grammar. MA Thesis, Ohio
University.
Zagona, K. 1990. "Times as Temporal Argument Structure," presented at MIT.