Models of the second language acquisition process have not specifically elaborated on the nature of the learner's perception and processing of target language input, a process known as intake. They have typically failed to specify the mechanisms or variables involved in the intake process or to distinguish between the stages of processing input from perception, to comprehension, to assimilation into an interlanguage grammar. Research and theoretical models from L1 psycholinguistic literature have been neglected. Recent L1 and L2 research on language processing are compared, and experimental methods for eliciting data to confirm them are discussed.

The aim of this paper is to demonstrate the role played by linguistic universals in second language acquisition. Research reported here focuses on the acquisition of dative structures and dative questions in a passive context in English by French and Inuit (Eskimo) students. Data were also elicited from native English-speaking students to serve as the norm. The data are interpreted within the theory of markedness and core grammar, as well as Case theory. The results of the testing, showing that unmarked forms are acquired before marked ones, are consistent with the predictions made by the theory of markedness and the property of adjacency which is crucial for Case assignment.

This study builds upon prior research dealing with the nature of discourse involving non-native speakers. In particular, we examine variables influencing native speaker foreigner talk and the form that speech modification takes. The data bases are (1) 80 taped telephone interviews between NNSs at two distinct proficiency levels, (interviewer) and NSs (interviewee), and (2) 20 NS-NS interviews. We consider five variables: 1) negotiation of meaning, 2) quantity of speech, 3) amount of repair (following a specific NNS request for repair), 4) elaborated responses, and 5) transparent responses. We find that the speech of NSs changes as a function of an NNS's ability to understand and be understood. We further suggest a general cognitive principle—transparency—underlying aspects of both foreigner talk and second language acquisition.

Studies have shown that less fluent non-native readers cannot make full use of semantic-level cues. Non-natives direct proportionately more attention than natives to syntactic and graphophonic information. This study asked whether a similar difference would be found between native and non-native listeners. Three groups were tested: native English-speakers and advanced and intermediate non-natives. After hearing a text, subjects were given a 55-item cloze test. It was hypothesized that with a subject's increased proficiency in the language, processing would show progressively greater attention to semantic than to syntactic or phonological cues. The cloze test was scored using a diagnostic scoring procedure developed by Clarke and Burdell (1977). Results confirmed the hypothesis. Semantically acceptable responses increased progressively for intermediate, advanced, and native groups, while responses based on syntactic structure decreased.

Schachter (1974) drew attention to the importance, in error analysis, of examining not only the L2 forms actually produced by the learners of a foreign language in their attempts to express themselves in L2, but also the L2 forms they seem consistently to avoid using. She also noted the close interrelation between such avoidance phenomena and the Contrastive Analysis approach to L2 teaching and learning: avoidance is the reverse side of negative transfer, since learners tend to avoid using in L2 those structures that have no parallel in their L1 and therefore provide them with no pattern for transfer. Of course, as Kleinmann (1977) has pointed out, “avoidance” implies that the structure in question is known to (i.e., can be passively recognized by) the learners, but not freely used by them; failure to use a structure or word that is unknown to the learners is an indication merely of ignorance, not of learning difficulty. Now a prime constructive purpose of error analysis is (or should be) to identify the sources of a learner's difficulties, as a necessary preliminary to helping him or her overcome them. Hence the importance of genuine avoidance phenomena that, when properly identified, can throw light on what would otherwise remain hidden recesses of uncertainty in the learner's mind. Levenston (1971) has convincingly argued, with numerous illustrations, that avoidance (“under-representation” in his terminology) of various English “clause (or group) structures” by Hebrew-speaking learners of English can be explained by the lack of Hebrew “translation-equivalents” for the English structures in question and the learners' consequent choice of less appropriate but more L1-equivalent structures. However, since Levenston was concerned not with avoidance phenomena as such, but rather with the evidence they provide of L1 interference with L2 learning and their stylistic effect on the learner's L2, he took the phenomena in question as established facts (on the basis, presumably, of his own teaching experience), without attempting to examine their frequency and extent.

Much work in the field of second language acquisition (SLA) has focused on the linguistic adjustments made by native speakers (NSs) when speaking to non-native speakers (NNSs) of that language. Modifications of speech to NNSs are usually assumed to be an attempt to make the target language more comprehensible. The study of second language teachers' classroom discourse has contributed to a thorough description of such modifications, which in general may be termed foreigner talk (FT).