In stating the rules for spirantiza- tion of begadkepat (p. 37 (note 1)-and there are others- can be valid only if the hateji are sounded they are phonemic. Secondly, the minimal diflerence between such pairs as 'dni and 'cinz, admitted on p. If shewa is a zero vowel, then the occurrence of a stop after a vowel and of a spirant after a consonant (as in watdabber ) means that no pattern of complementary distribution of stoplspirant can be found so, strictly speaking, this opposition would be phonemic. First, if all shewas were silent, then the two pronunciations of begadkepat consonants are not phonologically conditioned and must be recognized as pairs of phonemes. This treatment of shewa creates problems at two points. 54 he describes the 'extremely short vowels,' omitting, however, Jouon's words -cornme le shewa prononck:' calls shewa 'zero vowel phoneme' with composite shewas as 'allophones:' And on p. follows the contemporary Israeli practice and does not show shewa in transliteration. 50), a contradiction in terms: if it is a vowel, it is a sound if 'zero,' presumably silent. Assuredly the Tiberian system embodies recognition of seven vowel qualities (mouth positions) but the almost exclusive use of vowel letters for vowels which historical considerations show to have been long and the provision of special symbols for three (if not four) very short vowels shows that the scribes were aware of and recorded differences in length as such. Muraoka has gone to the other extreme, denying the phonemic value of vowel length, in fact making no significant distinctions in vowel length at all. Jouon went too far in asserting that there were four vowel lengths with exact ratios. Thirdly, at a few places Jouon's doctrine has been abandoned, most conspicuously his treatment of vowel length. Most of Chb is reproduced exactly The revisions are of three kinds: first, minor alterations of the text secondly, extensive addition of footnotes, mainly to supply references to discussions of various points of Hebrew grammar since 1923 (for example, there are nine new footnotes to $153, reporting recent research into the Hebrew clause). For example, in $13b GBH renders 'Les deux particules nN dont l'une est la prkposition savecn et I'autre I'exposant de I'accusatif se trouve souvent avec maqqef -nt$ et souvent sans maqqef me' as 'The two particles spelled both nN, one of which is a preposition meaning 'with' and the other the accusative marker, occur as often with a maqqef (at$) as without one (nt$).' The facts invalidate the translation there are about seven times as many of the former as of the latter. In only a few places does he miss the meaning. has displayed an uncanny command of both languages his English everywhere reflects the admirable lucidity of the French original. For the first two of these tasks I read the original and the translation in tandem.

To review GBH could involve three different tasks: to review the original work (Chb) to see how well it has been translated to evaluate the alterations made by Muraoka as reviser. Inevitably this new translation of Joiion's grammar will be compared to IBHS. O'Connor, An Introduction to Biblical Hebrew Syntax (Winona Lake, IN: Eisenbrauns, 1989), was the need of English readers for an up-to-date systematic syntax of biblical Hebrew met. Part 111, the syntax, was in every way superior to GKC. The morphology, especially of the noun, had made a considerable advance on GKC, thanks partly to Bauer and Leander's splendid work.

GKC had been used as the standard reference since early in the century, and Chb covered similar ground. In spite of its many fine qualities -clear organization, lucid exposition, rich examples -there was apparently no demand for an English version. Jouon's Crammaire de l'he'breu biblique was first published in 1923. Rome: Pontifical Biblical Institute, 1991. Part One: Orthography and Phonetics Part Two: Morphology.

JoГјon muraoka bibleworks 7