Author’s Rejoinder

AUTHOR'S REJOINDER Ryan D. Tweney andHarry W. Hoemann Stokoe has brought up several interesting issues that deserve serious comment. Two are technical comments concerned with the conduct of our study, while one is a general point with a number of interesting issues underlying it. We will discuss the technical points first. Stokoe points out correctly that our Ss were bilingual in both English and ASL. We do not agree, however, that this complicates interpretation. In the first place, very few deaf adults exist in the United States whose use of manual language is unaffected by English. The closest approximation would probably be an illiterate (in English) deaf adult whose primary social interaction is with other users of ASL. Such Ss would not be hard to find but clearly would be unable to read English sentences and translate them into ASL or to produce a back translation from ASL to English. In short, any S capable of performing the task is necessarily bilingual and must posses minimal competence in both English and ASL. Since controversy typically stems from the assertion that ASL is "inferior" to English or "interferes with" the acquisition of English, it seems appropriate to base initial comparative investigation on Ss who have at least some competence in both languages. It is of course, possible (in fact, probable) that Ss relied on their knowledge of both languages in preparing translations. This is in fact the point of using translation procedures. We did not attempt to evaluate the capability of an individual to function with one system in isolation, but we did attempt to evaluate the ecologically most valid situation -bilingual Ss with interpenetrating language systems. Having said this much, it is necessary to add that future studies can and should address the question of performance in situations which Sign Language Studies encourage relatively pure ASL performance. Thus, we are now engaged in a back translation study in which fingerspelling by Ss is not allowed-only manual signs can be used. Such a procedure demands, however, limitations of the subject population to a very, very small group of highly skilled translators. This is because most deaf adults have come to rely quite heavily on fingerspelling and are simply lost in a task that forbids it (unless the text is very mundane). Studies of this type are needed, of course, but will suffer from limitations of a different type-results will not be as generalizable to the entire deaf population. In the long run, only a number of studies, each focussing on one or another of the difficulties, can converge on the issues involved in "pure" ASL performance. The present paper is, we hope, only a beginning. Stokoe's second point concerns our choice of stimulus sentences, which he claims are unique to English surface structure, and which, he feels, render re-construction of the original test sentence largely a hit-or-miss affair (since the transformations in question are obviously not represented in ASL). We would argue that, in the first place, all transformations are generally considered unique to a given language (Chomsky, 1965). In the second place, preservation of a particular transformational form under back translation is perfectly possible if ASL users possess either (a) transformations which are the functional equivalents of the ones in English, or, (b) "Englishisms"in ASL which are present for the specific purpose of encoding English transformations. The latter possibility seems most likely and accounts for such devices as fingerspelling THERE-IS, for example, which must be regarded as a wholly different process than finger-spelling, say, TAXI to make up a purely lexical deficiency of ASL. Finally, Stokoe's point that our use of "formal grammar" is ambiguous, is well-taken and completely correct. We agree completely that performance studies like ours cannot substitute for a formal grammar. Yet, attempting to avoid full (and lengthy!) discussion of an important theoretical point may have left more ambiguity than we had intended. The basic point is one that Stokoe himself has brought up (1970), as well as others, (Schlesinger, 1971; Fant, 1972; Bellugi, 1971; Bonvillian & Rejoinder 79 Charrow, 1972). In brief, manual languages have not been described in terms compatible with modern generative-transformational...