The traditional account of the position of the Latin accent is well known, both from ancient sources (e.g. Quintilian i.5.30) and from modern text-books – e.g. Kent (1932: § 66): ‘Latin developed an accent depending upon the quantity of the penultimate syllable, as follows: A long penult was accented, as in pepérci, inimtcus; but if the penult was short, the antepenult received the accent, as in extstimō, cōnfíciunt, ténebrae. Disyllables were necessarily accented on the penult, as in tégō, tóga.’

Recent work on elliptical forms of the English imperative has drawn quite heavily upon the apparent parallel between tag formation in imperatives and tag formation in interrogatives. Thus, it has been argued that the auxiliary in imperatives must be will because in all interrogative tag formations the auxiliary element in the appended tag is the same as the one occurring in the initial segment (Klima, 1964: 254; Postal, 1966; Katz & Postal, 1964). The invited comparison is between tag-interrogatives like (1) John did it, didn't he? (2) They can do it, can't they? (3) He can't do that to us, can he? (4) You will try harder, won't you? and putative imperatives like (5) Do it now, will you! (6) Bring me a slab, won't you! (7) Bring me a slab, will you! (8) You will try harder, won't you! Consideration of such examples makes it fairly evident that, in the cases of the elliptical constructions (5)–(7) aux Will has been deleted from the initial segment and hence occurred in the deep structures serving as the sources for these constructions. Moreover, such parallels provide good motivation for the conjecture that ALL imperatives eventually derive from underlying sources whose phrasemarkers contain the element will (Postal, 1966: 162; Katz & Postal, 1964: 75). It is my purpose here to attempt an analysis of the putative parallel between tag-imperatives and tag-questions in order to determine the extent to which it supports such conjectures about the specific nature of structures underlying English imperatives.

Linguistics, epistemology and psychology share many common interests; for instance, an interest in the learning of language, in the nature of the fundamental units of which linguistic systems are composed, in the correspondence between language operations, logical operations and intellectual operations. The phenomenon of COMPARISON falls under the latter rubric, as do, for instance, NEGATION, ASSERTION and PREDICATION. In the literature of these three disciplines we find ample evidence of an interest in the latter three phenomena, but there has been surprisingly little discussion of comparison. We say, ‘surprisingly’, since of these four operations comparison might well be held to be the most important. In any kind of classification the fundamental intellectual activity is the comparing of one object, event, etc. with another or with others. Further, to adopt the terminology of psychology, our behaviour in any situation must be governed to some extent by the recognition, whether implicit or explicit, of similarities and differences between that situation and others with which we have had to deal.

Kohler (1966a, 1966b: 346–348) asks whether the syllable is a phonological universal, and concludes negatively.1 The way to support such a conclusion is not difficult to imagine: the sort of specific objections to the syllable which Kohler raises would, if well-founded, be sufficient to prove his case.

The present study arose out of an investigation of whether already and yet as in He has left already, *He has left yet, IHe hasn't left already, He hasn't left yet formed a suppletion set that matches some →any, sometimes → ever, etc. as in They want some love, *They want any love, ?They don't want some love, They don't want any love (for discussion of some, sometimes, too, etc. cf. Klima, 1964; Fillmore, 1967; UCLA, 1967: Negation). Various observations about the restriction of already and yet to certain tenses and aspects, and about their failure to match with many adverbs of time, led to an investigation of whether already and yet were really adverbs of time, as is usually assumed. The conclusions we reached are that already and yet indeed form a suppletion set (with certain limitations not shared by other suppletives), but that they cannot usefully be treated as time adverbs. Rather, they seem to have close connexions with perfect aspect. The arguments presented here in support of this hypothesis are by no means exhaustive but will hopefully provide a suggestive contribution toward the study of both adverbs and aspects.