Human evolution, language and mind: A psychological and archaeological inquiry By William Noble and Iain Davidson (review)

REVIEWS869 should be noted that the proportion 26:265 is influenced by particular classificatory assumptions. H argues that many verses Bliss regards as hypermetric are actually of type A with dissyllabic anacrusis. He finds that the only sure differentiator is whether the anacrusis comprises two Satzpartikeln, with the result that a verse like swäpü Abele wurde (Genesis A 1019b) must be classified as normal, even though it is paired with a hypermetric verse, because it appears in a poem that contains other such pairings. It is impossible to say whether any poet actually considered such pairings inadmissible, and so the restriction of this taxonomic meüiod to particular poems seems hard to justify. In any event, H's eminently reasonable belief that some verses are transitional in nature (318) suggests rather that this model of poets as allowing or disallowing odd pairings is unrealistically rigid. Moreover, the assumption that a verse is normal till proved hypermetric, especially in the context of hypermetric passages, prompts the question why it might not be hypermetric till proved normal. Part II mops up niggling details of classification, taking up each of Sievers's five types in tum. For example, though he admirably sketches the real bases for doubt, H finally decides that the type oftost wîsode in the off-verse is not aberrant (142-44). (H gives some valuable and convincing correctives to my own arguments on this score, but I should say that in die end I am still unconvinced; any type in which nearly 70% of the examples are dubitable would appear to be an avoided type.) H's real innovation in laying out his system in this section is that he admits a non-Sieversian type with three full lifts ('type 3'), for example twelfwintra tïd. This provides consistency with his treatment of eoröweard öone (above) and avoids the arbitrariness that attends Bliss's attempt to categorize these as belonging to type D or E. Part III is a catalogue of types, furnishing examples and data on frequency (overall and in ten poems). The seven appendices are also extraordinarily useful, giving double alliteration rates, revisions to Bliss's scansion of Beowulf, a list of types, an account of the statistical methods used, a list of formulas in types A and D, data on hypermetrics, and instructions for the use of the database (marketed separately). It is an act of nerve to lay out a new system of this sort since it demands so much decisiveness about seemingly undecidable matters that, as my own comments illustrate, the currish barking of critics is inevitable. It would be wrong to give the impression that H's accomplishment is as minor as my quibbles. Some of his innovations will permanently change me field; others will certainly be central to metrists' debates for a good many years to come. Whatever one's position on these matters, this book is an invaluable resource, and no student of early Germanic metrics can afford not to study it with the greatest care. REFERENCES Bliss, A. J. 1967. The metre of Beowulf. Rev. edn. Oxford: Blackwell. Fulk, R. D. 1992. A history of Old English meter. Philadelphia: University of Pennsylvania Press. Pheifer, J. D. 1974. Old English glosses in the Epinal-Erfurt glossary. Oxford: Clarendon. Sievers, Eduard. 1885. Zur Rhythmik des germanischen Alliterationsverses. Beiträge zur Geschichte der deutschen Sprache und Literatur 10.209-314,451-545. Department of English Ballantine Hall Bloomington, IN 47405 [fulk@indiana.edu] Human evolution, language and mind: A psychological and archaeological inquiry. By William Noble and Iain Davidson. Cambridge: Cambridge University Press, 1996. Pp. xiii, 272. Reviewed by Jean Aitchison, University of Oxford Until recently, many linguists shuddered away from the topic of language origin. The intellectual logjam was broken above all by Pinker and Bloom (1990) in an influential paper which 870LANGUAGE, VOLUME 73, NUMBER 4 (1997) pointed out that language evolved by normal evolutionary mechanisms. They commented that a wealth of respectable new scientific information relevant to the evolution of language had never been properly synthesized. Syntheses have begun to appear (e.g. Lieberman 1991 , Aitchison 1996), but more are needed. Noble and Davidson's book is a...