Two questions are raised: Is it possible to characterize the notion human language in terms of absolute and typological universals? And if so, what is the relationship between these universals and those formulated for primary languages? Given these questions, the purpose of the paper is to: (1) investigate some of the methodological considerations involved in attempting to characterize the notion in terms of universals; (2) consider the implications of the language-contact situation in attempting to define interlanguages in these terms; and (3) provide examples of logically-possible but empirically-unsubstantiated, types of interlanguages. Interlanguage forms from native speakers of Japanese, Mandarin, Spanish, and Farsi provide data for investigation of these questions. The discussion also questions whether the grammars of interlanguages obey the same constraints as the grammars of primary languages. (JK)