[article] Early Developments Of The Vowel Systems Of North-west Germanic And Saami

PETER SCHRIJVER EARLY DEVELOPMENTS OF THE VOWEL SYSTEMS OF NORTH-WEST GERMANIC AND SAAMI 1. Introduction This article deals with three seemingly unrelated issues. The first concerns the vexed problems of the Germanic 'Auslautgesetze' and the implications that the solution I prefer has for the development of the Germanic vowel system (sections 2-4). The second issue concerns phonological parallels between Early North and West Germanic and Saami (sections 5-6). The third and final issue addresses the evidence for a non-lndo-European substratum language in Northern Europe (sections 7). The intimate link between them will become clear in the course of the article. 1 2. Auslautgesetze What exactly happens to final syllables in lndo-European languages is an issue of some importance and debate, but in no language group more so than in Germanic. The literature on the subject is correspondingly huge. It is therefore with some relief that we can tum to monographs that provide overviews of previous scholarship. The most recent survey is Dirk Boutkan's The Germanic 'Auslautgesetze' (Amsterdam/Atlanta 1995). Of the various theories that have been proposed, Boutkan himself shows a marked preference for Kortlandt's ideas, which the latter has laid down in various articles, which Boutkan conveniently summarizes. Unlike the traditional, socalled 'quantitative' theory, which has received wide acceptance (see 2.2), Kortlandt does not distinguish between long (or acute) and overlong (or circumflex) vowels in final syllables. Rather, he ascribes different reflexes of one and the same long vowel to differences in the phonetic contexts of this long vowel. This is in fact one of the oldest approaches towards the auslaut 1 I am indebted to Willem Vermeer, Petri Kallio and Lisette Gabriels, who commented on an earlier version of this article. Alfred Bammesberger - Theo Vennemann, Languages in Prehistoric Europe (2003), 195-226. 196 problem, predating even the quantitative theory, which arose in the 1880s. What almost all scholars seem to reject, either explicitly or, more usually, implicitly, is what can be termed the 'qualitative theory', which is based on the idea that pre-PGm. *a and *6 have different reflexes in final syllables. This idea is not new, as its origin dates back to Moller (1880), being further developed by Jellinek (1891; in 1895 he combines the qualitative and quantitative theories) and Van Wijk (1907-8). All three claim that different reflexes of *ii and *6 appear first and foremost in the West Germanic languages, and at best possibly in Gothic and Old Norse. Boutkan dismisses the idea mainly2 because it does not account for Gothic, which, however, is beside the point because it does account for certain features elsewhere. It seems to me that the qualitative theory has been improperly dismissed, as van Wijk already pointed out in 1907. 2.1. The qualitative theory The merits of the qualitative theory can most conveniently be judged on the basis of the following diagram. For ease of reference the reconstructions according to the quantitative theory have been added in column number 1, and those according to Kortlandt and Boutkan in column 2. The correspondences shown in the diagram are based on Moller 1880, Jellinek 1891and1895, and van Wijk 1907-8, who differ on details. The version that emerges from the diagram is: *6 and *a have different reflexes in final syllables in West Germanic, except in absolute word-final position, where both merge as *6 > *ii. A number of forms require a brief discussion, which will be found in the accompanying notes. Pre-PGm *o lsg. pres. *-6 Nsgm n *-an Absg. o *-Od 2 OE -u -a -a OS -u -o -o OHG ON -u -fJ -o (-i)' -0 -a Goth. -a -a' -o 197 Early Developments of the Vowel Systems Peter Schrijv