settle the issue of whether adults are able to acquire language in informal situations with any real efficiency. Following the survey, it will be argued that formal and informal environments make contributions to different aspects of second language competence.

We are thus considering two hypotheses:

Review of Literature

Upshur (1968) compared three groups of ten adult ESL students enrolled in a special summer session for law students at the University of Michigan. The first group, who scored highest on the entrance test (Michigan Examination in Structure), attended seminars and classes during the 7-week period that were conducted in English, but had no extra ESL classes. The second group, who scored lower on the entrance test, also attended law classes and had 1 hour daily of ESL in addition. The third group scored lowest on the pre-test and had 2 hours of ESL daily in addition to law classes. At the end of the summer, an alternate form of the pre-test was given. While all three groups showed some improvement in performance, Upshur's statistical analysis revealed "no significant effects on language learning attributable to amount of language instruction", and concluded that "foreign language courses may at this time be less effective means for producing language learning than the use of language in other activities". This is a strong version of hypothesis I.

Upshur's conclusion appears to be consistent with his data. Krashen and Seliger (1975) suggest, however, that motivated second language students are able to provide themselves with the essential ingredients of formal instruction without going to class. Rule isolation can be done by recourse to a text or by asking informants about grammar, while feedback is available when helpful friends correct the learner. Without extensive probing of the private lives of those involved in the

41