A NEW CONJECTURE ON THE ORIGINS OF ABSOLUTE AND CONJUNCT FLEXION

Abstract

The distinction in the Insular Celtic languages between absolute and con junct flexion arose out of the same syntactic and morphological patterns as underlie the grammars of all Indo-European languages. The Proto-Indo European verb could be in marked, sentence-initial position, for 'emphasis' (focus, foregrounding, vel sim.), in which case it was accented, or in unmarked, sentence-medial or final position, in which case, in main clauses, it was unaccented. In some phase of prehistoric Celtic, an apocope rule affected inherited short final *-e and *-/ in unaccented words. Non-initial verbs, being unaccented, underwent this apocope, giving rise to Insular Celtic conjunct forms. Initial verbs, being accented, did not undergo this apocope, giving rise to absolute forms. The recent publication of The Origins and Development of the Insular Celtic Verbal Complex by Kim McCone provides a welcome restatement of one of the theories of the development of the distinction between absolute and conjunct flexion in the Insular Celtic languages, amongst other matters.1 In a number of previous publications addressing related matters, I have expressed agreement with the underlying principles of McCone's theory.2 McCone's crucial insight has been that the system of absolute and conjunct flexion is derivable from given syntactic, morphological and phonological patterns of the proto-language without the need to conjecture unusual use of any of the morphology,3 or a remarkable particle, various types of which have been posited.4 The present note continues to uphold the derivational principle as just stated, while conjecturing a difference in the detail of exactly how that principle operated. I start by presenting a pared-down schematic representation of the essence of McCone's theory (a more detailed presentation, in accord ance with McCone's own terms, will be reproduced for argument's sake further below). The basis of the developments is the distinction between the presence and absence of enclitics (whether pronouns or particles of various types) in association with simple initial verbs or verbs with some 1 McCone (2006), building on arguments presented most notably in McCone (1979,1982, 1985,1994). 2 e.g. Isaac (1993,26; 1996,405; implicitly, 2000b). 3 PIE primary vs secondary flexion, e.g. Meid (1963),Watkins (1963). 4 e.g. Cowgill (1975), Kortlandt (1979, 1994), Schrijver (1994, 1997), Schumacher (1999, 2004). Such theories are discussed at length by McCone in his book, so will not be touched upon further here. ?riu LVii (2007) 49-60 ? Royal Irish Academy This content downloaded from 207.46.13.113 on Thu, 06 Oct 2016 04:14:01 UTC All use subject to http://about.jstor.org/terms