German has a phonemic distinction between tense and lax vowels. For all vowels except the low ones this distinction is realized phonetically by both vowel quality (i.e. formant structure) and vowel quantity (i.e. duration). In acoustic terms, we find oppositions like in Miete [i:] 'rent' vs. Mitte [I] 'mid' differing both in formant structure, most noticeably FI and F2, and in duration. This distinction in both quality and quantity is fully manifested under main stress. However, there is disagreement in the German phonetic literature about what happens under degrees of stress lower than main stress. There are two positions on this issue. Under the first view, both phonetic quality and quantity remain equally important as realizations of the tense/lax opposition for any level of stress. Under degrees of stress lower than main stress the difference between tense and lax vowels is reduced compared to main stress position. This reduction affects both quality and quantity to the same proportion. Under the second view, reduction under low degrees of stress affects quantity more than quality, such that in some weak stress contexts, tense and lax vowels are claimed to differ in quality but to be equal in quantity. This paper provides acoustic-phonetic evidence that bears upon this issue. Duration and formant structure of German tense and lax vowels are evaluated with respect to these two different views, which will be defined more specifically as the "association hypothesis" and the "dissociation hypothesis" respectively. Following this empirical treatment, which constitutes the main body of the paper, the reader will be introduced to a classic issue in German phonology: is it quality or quantity that has distinctive status in the German vowel system. Results and implications from the present study, as well as other data will be discussed in light of this debate.

This paper is copyrighted, and is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0) - see https://creativecommons.org/licenses/by-nc-nd/4.0/ Parts of this work were presented at the 14th Annual Meeting of the German Linguistic Society ("Deutsche Gesellschaft fur Sprachwissenschaft"), February 25-28, 1992 in Bremen, Germany. I want to express my thanks to Wilbur Benware, Bernhard Hurch, Allard Jongman, Magnus Petursson, Karl Heinz Ramers, and Linda Waugh for their helpful comments on an earlier manuscript that covers part of what is presented in this paper. Thanks also to Abigail Cohn, Ann Bradlow, Allard Jongman, and Linda Waugh for reviewing the present paper and to John Kingston for introducing me to the statistical methods used here. My presentation in Germany was supported by a travel grant from Cornell University, Sage Graduate School and the Department of Modern Languages and Linguistics.