The first two chapters of the thesis give an overview of relevant L2-acquisition research carried out during the nineteen-seventies and nineteen-eighties and present the most important hypotheses and models developped in this period. Special attention is dedicated to two German large-scale projects, the Heidelberger Forschungsprojekt Pidgindeutsch and the ZISA-Project. Section 1.8 describes the state of the art of research in the L2-acquisition of Italian.
Chapters 3-6 present an account of an 8-month longitudinal study of the natural L2-acquistion of Italian by two subjects that took place in Trieste, Italy. The persons observed were a 21-year old female nursery teacher of German mother tongue and a 26-year old male university student with Greek as first language. Both had arrived in Trieste shortly before the beginning of the study without any knowledge of Italian. During the observation period neither of the two attended Italian courses or received any other regular instruction about Italian. 20 unguided conversations were recorded with the German woman, 13 with the Greek man. The conversations lasted an average of 15 minutes and were transcribed orthographically.
1300 utterances (700 by the nursery teacher, 400 by the student, 200 by the Italian control group) were syntactically analyzed using a simple variety grammar (cfr. Klein and Dittmar 1979). The main areas investigated are the appearance of the subject noun and pronoun, the development of modals, auxiliaries and the verb phrase, the use of subordinate clauses and prepositional phrases introduced by a. The data reveal substantial differences between the two subjects: at the end of the observation period, the Italian of the German nursery teacher reaches the level of grammatical complexity with which the Greek student started; regarding fluency and communicative ability, though, the German woman seems to be more advanced. From a syntactical point of view, particularely interesting are some features of the nursery teacher's speech: initial high and final low frequency of sentences without subject, late appearance of the copula, acquisition of modals before auxiliaries, elimination of functional elements.
For a detailed analysis of some data see also Schneider (1995)