M. H. Wolpoff Muhivariate Discrimination, Tooth Meas- urements and Early Hominid Taxonomy L)$iaamatt of Anfhmjal~gy,

It is argued that the value of the various multivariate approach= in dealing with functional, taxonomic, and phylogenetic problems can only be determined empirically. A review of three studies dealing with multivariate analyses of the early hominid dentition suggests that at best, the multivariate statistics used are inappropriate and the results contradictory. There have been several recent attempts to use the early hominid dentition to determine phylogenetic relationships through sophisticated multivariate statistical analysis (Brace, Malher & Rosen, 1972; Read, 1975; Robinson & Streudel, 1973). It is my belief that these techniques constitute more than an overkill. I contend that they are inappropriate; a position not weakened by the fact that the conclusions of these papers contradict each other. The strength of a multivariate approach is that it allows one to consider numerous factors at the same time. Many authors feel that to consider multiple variables together is to adhere to LeGros Clark’s “total morphological pattern. ” If everything is measured