| Scope and method of study. This study elicited acceptability judgments of elliptical sentences presented in context and betterment suggestions for the sentences deemed as less-than-totally acceptable. The purpose of the study was four-fold---namely, (a) to compare the acceptability ratings obtained for elliptical sentences with those obtained for two other sentence types: non-elliptical and ungrammatical; (b) to determine whether native speakers of English see a difference, in terms of acceptability, between elliptical sentences and non-elliptical sentences; (c) to explore whether the elliptical sentences deemed as less-than-totally acceptable are 'bettered' by inserting the ellipted element(s); and (d) to investigate the nature of the acceptability judgments obtained. A specially-devised 31-item questionnaire carrying samples of naturally occurring written discourse was administered to 172 college students. SPSS was used in the interpretation of the acceptability ratings obtained, and a careful analysis of the betterment solutions offered was undertaken.; Findings and conclusions. The elliptical sentences and the non-elliptical sentences were found to have had acceptability ratings significantly higher than those of the ungrammatical ones. So long as ellipsis did not cause the sentence carrying it to become a fragment, there was no difference, in terms of acceptability, between an elliptical structure and a non-elliptical one. The analysis of the bettered rewritings of the sentences deemed as less-than-altogether acceptable showed that deletion from a non-elliptical sentence of discourse-active material was favored over insertion into an elliptical sentence of left-out discourse material. The main criterion on the basis of which the assessments were made was grammaticality. Interestingly, despite the in-context presentation of the sentences whose acceptability had to be rated, the grammar underlying the judges' assessments was sentence grammar, rather than discourse grammar. That suggests that even when specifically asked to rate the acceptability of structures against larger stretches of discourse, people tend to make their assessments based on sentence-, rather than on discourse-, level grammar. |
