T HIS PAPER, which is based on an acceptability experiment conducted by the author, discusses the status of certain multiple negative constructions in current American English.1 Questions of acceptability of linguistic units are questions of usage, to which there are basically two approaches: normative and descriptive. The aim of the descriptive approach is to describe the facts of linguistic acceptability as adequately as possible. The descriptive approach generally relies on jury opinion and frequency." If a substantial majority of a selected jury declares a construction to be acceptable and if the construction is fairly frequent in the language, it is taken to be acceptable. All available evidence, however, indicates that the judgment test used to solicit responses from the jurors is, as observed by Quirk and Svartvik, "a rather weak measure of linguistic acceptability."3 The results of outright questioning of informants, according to Gleitman, are at best ambiguous.4 In order to overcome the problems inherent in the judgment test,5 Quirk and Svartvik have introduced a new dimension in acceptability testing-the operational test. They also present evidence suggesting that the traditional bimodal scale used in most judgment tests should be replaced by a three-point scale. The combined results of the modified judgment test and the operational test, they contend, provide a far better measure of linguistic acceptance. The operational test is based on the assumption that if an informant is able to carry out a simple grammatical operation on some sentences but