The topic of this paper is teaching German as a second language in the Federal Republic of Germany. The paper examines a bilingual situation in which members of a foreign minority, while striving to preserve their own national identity, are forced to learn another language in order to assimilate the cultural system of a new homeland.

Research on Second Language (L2) Acquisition, over the past ten years, has undergone substantial changes by shifting its focus of interest away from an analysis of linguistic structures alone, concentrating more on the learner himself or, rather, on the process of learning. It had become obvious that one of the major shortcomings in contrastive studies as well as in the usual kind of error analysis is that they lack thorough investigation of factors which determine the kind of approach a learner may take to acquire a second language. This again implies that it is more fruitful to study the process of learning itself instead of merely analysing its outputs. It is by now widely accepted that the learner takes an active part in the learning process and does not merely get trapped in structural gaps which linguists may find when comparing the source language (the learner's L1) and the target language (L2).

This article describes the undirected learning (learning without explicit teaching) of tense markers in German by Spanish migrant workers. It will be shown that the remarkable differences the informants displayed in expressing the concept of tense are a function of the daily communicative routines which they developed in order to meet their communicative needs and demands. By examining very simplified learner varieties, I will demonstrate that a traditional grammatical analysis of tense is not capable of accounting for the wide range of expressions for temporal relations which were produced in an elicited Spanish-German translation task.

From the point of view of linguists interested in pidginization, creolization, or other language contact phenomena, the continued presence of large numbers of southern European and North African foreign workers (called Gastaibeitei in Germany) and their families in northern European countries presents almost laboratory conditions for the investigation of the initial and subsequent stages of foreign language acquisition and development. Of particular interest is the fact that many of the languages involved here (Turkish, Arabic, Greek, Serbo-Croatian, Spanish, Italian, Portuguese) are typologically and genetically distinct; thus the situation is ideal for the study of common features of language development. If a Creole were to develop here, it would provide extremely valuable evidence for or against proposed universals.

In recent years research on language acquisition (both L1 and L2) has gradually turned away from the domain of pure description towards dealing with problems of explanation. Most students of language development feel that it is somewhat insufficient to merely state the linguistic regularities and verbal strategies that according to the available production data appear to govern processes of language learning; rather, it seems desirable that the regularities observable in the learner's verbal behavior over time be in some way explained. Frequently, the term explanation is—implicitly rather than explicitly—understood as an attempt to reduce the specific linguistic regularities found in language learning processes to a set of some more general (i.e. non-linguistic) principles that will not only control other learning tasks as well, but will also govern man's mental activities as such (Slobin 1973, Schlesinger 1977, Larsen-Freeman 1976, Bever 1970, Wong-Fillmore 1976). Among those mental domains which appear to provide principles general enough to be applicable to problems of language, learning cognition or rather cognitive development turned out to be the most promising. As a consequence, a large number of studies focused on the attempt to explain formal linguistic properties of language learning processes in terms of more general regularities in man's cognitive abilities as they develop (Seliger 1980, Gass 1980, Felix 1976, Sinclair-de-Zwart 1973, Meisel 1979, Clahsen 1979). In fact, the term explanation frequently means little more than reducing language acquisition to aspects of cognitive development.

Why are pidgin utterances structured linguistically the way they are? Why, as has often been noted, do the linguistic structures of different pidgins tend to be more similar to each other than to the structure of the original languages involved in the specific pidgin? This has been noted as all the more surprising since these similarities also occur in cases where totally unrelated languages are involved, so that borrowing must be excluded; or where historical explanations cannot apply because there was no contact in the past at all. It will be suggested here that these similarities result from universal linguo-cognitive processing strategies which man employs in learning languages. Some of these strategies are universal in the sense that they apply in all acquisitional types so that pidgins have some properties which recur in all types of acquisition. Other strategies are more restricted in their applicability, for example, to the various types of second language (L2) acquisition.

The first public presentation of the ideas leading up to the Interlanguage (IL) Hypothesis occurred a decade ago at the Second International Congress of Applied Linguistics (Selinker 1969; presented in greater detail in Selinker 1972). At that time, it was stated that, in attempting to understand the phenomenon of second-language learning, the field was at a stage where there existed an “inability to unambiguously identify the phenomena we wish to study” (Selinker 1969:35). That first presentation of the IL notion focused on the “psychologically relevant data of second-language learning,” trying to describe what it was that should be studied; the paper attempted to provide some theoretical constructs which would help identify and explain that data.

Recent research has focused increasingly on the good language learner (Rubin 1975; Naiman, Frohlich, and Stern 1978; Wesche 1977; Hosenfeld, 1976; Bialystok and Frohlich 1977a, 1977b; Cohen and Robbins 1976; Cohen 1977). A distinction that has gained currency is that between acquiring a second language and the conscious learning of the language through what Krashen (1977) refers to as “monitoring.” The studies reported in this paper concentrate on the identification of successful learning strategies—strategies that could actually be taught to language learners to help them maximize the benefits of conscious language learning. It is not very clear which behaviors actually promote success. Learners themselves may not pay conscious attention to what they do. In order to identify strategies, therefore, we must focus our attention on basic aspects of the language learning process and also on the interaction between the learner and the teacher.