Refining Hebrew diachronic phonology.

• Author: Woodhouse, Robert

Upon reviewing a recent contribution of mine on the history of Hebrew vocalism (1) with a view to reassessing phonemic likelihoods, I find that the suggested chronology contains more vowel changes and therefore more distinct chronological stages than seem warranted by the facts pertaining to these changes. Since the collapsing of the chronology into a smaller number of stages also facilitates the elimination of a typographical difficulty, a correction seems in order.

Briefly, there is no objective basis for separating stages 3 and 4 of the original scheme. In addition, the intermediate vowel lowering specified in the original stage 5 is superfluous. This makes it possible to eliminate the symbols / and / from the derivation in favor of e and e (the original symbols were, after all, barely distinguishable from i and i, respectively, even under a magnifying glass). Similar considerations lead to tracing the development of u to o via o, it being understood as before that the short intermediate vowels now denoted e and o were high enough to be interpreted by relevant non-Canaanite speakers as /i/ and /u/, respectively.

It will be remembered that the following cases were considered: (a) the construct *zaqinu > zqan 'old man,' (b) the absolute * gintu > gat 'winepress, olivepress,' (c) the absolute *'amintu > ''met 'truth.' which early develops regular geminate after peak of noninitial syllable due to dissimilation against the medial m, (d) the absolute *libbu > leb 'heart,' which develops "compact" geminate after peak of initial syllable, (e) the 3rd pers. sg. masc. verb form * kabida > * k5bed 'was heavy,' (f) the absolute * kabidu > * k5bed 'heavy,' (g) the absolute * ri'mu > re m 'wild ox.' A newcomer, added for convenience, is (h) the absolute *surru > [so.sup.(w)]r 'Tyre,' also with a stage of "compact" geminate after peak of initial syllable.

In the chronological scheme below, only the development of the accented vowels and the "compact" geminates in these forms is indicated (bb denotes a "compact" geminate); and only forms altered by a given change are listed in the output of the stage in question.

(a) * zaqinu (b) * gintu (c) * amintu (d) * libbu (e) * Kabida (f) * kabidu (g) * ri'mu (h) * surru

1. Penultimate stress generalized: i > e, u >o

   * zaqunu * gentu * amentu * lebbu * kabeda * kabedu * re'mu * sorru

2. Early assimilations of * nt > *tt; loss of syllable final glottal stop:

   -- *'amettu -- *remu --

3. Loss of final vowels in...