Proto-Germanic [i] and [e]: One Phoneme or Two?

Abstract

J. W. Marchand, in an article which bears almost the same title as the present note and which was published in Language 33.346–54 (1957), reexamines the question posed in that title, and reaches the conclusion that the reflexes of PIE /i/ and /e/ did not contrast in identical environments, i.e. that these two PIE phonemes had merged in Proto-Germanie, and that, accordingly, the system of PGmc. short syllabics exhibited only a three-way contrast: one front vowel, one back vowel, one low vowel. He presents arguments which are intended to establish the validity of two assumed sound-changes for Pre-Germanic : PIE /i/ > /e/ before /a/ sounds, and PIE /e/ > /i/ before /u/, and he declares that if we deny the occurrence of these sound changes, we must accept the phones [i] and [e] as members of different phonemes. My contention here is simply that the occurrence of the alleged changes for Proto-Germanie (or Pre-Germanic) has not been proved, and that Proto-Germanie must therefore be assumed to have exhibited, as asserted by Twaddell and Kuryłowicz, an asymmetrical structure in its system of short syllabics.