LENGTH, VOWEL, JUNCTURE

The phonetic evidence usually cited for internal (open) juncture is better viewed as a manifestation of extra length conditioned by successive syllables containing full vowel nuclei. It does not answer to morpheme or word boundaries, but is a mechanical lengthening opposed to an equally mechanical borrowing of time on the part of syllables with reduced vowel nuclei from immediately preceding syllables with full nuclei. The three reduced vowels form a class apart with contrast only in the front-back dimension. In an earlier version of this article four reduced vowels were recognized. Fred Householder corrected the number to three. The vowel scheme here is virtually the same as that proposed by Stockwell, Bowen, and Martin in The Sounds of English and Spanish (Chicago, 1965), Ch. 7. Strong argu ments for recognizing double systems of vowels throughout Germanic are given by Einar Haugen in "Phonemic indeterminacy and Scandinavian umlaut," Folia Linguistica 3.107-119 (1970). Ilse Lehiste's experiments now confirm the major claims of this article : "The Timing of Utterances and Linguistic Boundaries," Journal of the Acoustical Society of America 51.2018-2024 (1972), esp. p. 2023. 0.0. The useless "junctures" in many unit morphs have long been a thorn in the side of phonological grammarians. Phonological structure is virtually the same in both members of the following pairs,2 although those of the first column are unitary (all of them, for unsophisticated speakers; some, for sophisticated ones), and those of the second are composite: rebate ambush Nimrod ampere warlock magpie compost ozone asset pr?-set crowbait rosebush ramrod compeer oarlock mud-pie bedpost ampersand hoosegow veto speed-zone quicksand whose gown? thr?e-toe3 * An earlier version of this article appeared in Linguistics 1.5-29 (1963). 1 I gratefully acknowledge the help of J. L. Dillard, Lee S. Hultz?n, Ralph B. Long, and Alan E. Sharp, who read an early version of the MS and offered suggestions. Those adopted from Hultz?n and Sharp are too numerous for separate mention. 2 "Same" of course means "same in my dialect." For speakers who say [kompost] for compost or [rAlok] for oarlock they are not the same. 3 Newman marked vetoed and thr?e-toed identically: "On the Stress System of English," Word 2. 171-187 (1946), p. 183. This content downloaded from 207.46.13.60 on Thu, 21 Apr 2016 07:14:44 UTC All use subject to http://about.jstor.org/terms