An Examination of Certain Concepts in Phenetic Taxonomy

Abstract

An examination of certain basic concepts employed in phenetic taxonomy provides a logical framework within which the taxonomic process may be operationally defined, with little or no violence to traditional views. There remain elements of apparently subjective choice for methods adopted in practice, but the "non-probabilistic" approach, based on the concept of "association," is conceptually simple and corresponds closely to traditional methods. There is available for such an approach a "general coefficient of association," which incorporates both qualitative and quantitative information.

The Phylogenetic Fallacy

“…I have dealt above with what seem to be some of the principal aspects of the "phylogenetic fallacy"; by which term I refer to the view that, in reconstructing phylogenies, we can employ something more than the observed attributes of individual specimens, plus some concept of "overall resemblance" and some concept of an "attribute" of a set or class of such specimens (Colless, 1967). There are, of course, other, more basic concepts, which are not in dispute (but I believe that the concept of "homology" is derivable from that of "overall resemblance").…”

“…In particular, they misunderstand the crucial concepts of "unit attribute" and "overall resemblance," the latter being apparently misconceived as some rather vague, subjective estimate, based on the "look" of material specimens. Nothing could be further from the truth, as I have tried to point out elsewhere (Colless, 1967); "overall resemblance" involves the "holomorph." I would add, however, the warning that such misunderstandings clearly stem from deeply-ingrained, traditional viewpoints, and are not to be rectified by ex cathedra, or lightly-sketched, accounts of opposing views.…”

“…To recapitulate the essence of Wilson's "test," using my own preferred terminology (Colless, 1967): (a) we first select a basic series of attributes, a 1} a 2 , a s , . .…”

“…The second, apart from the problem of inferring direction of change in attributes (see below), is also statistical and phenetic in that any complex attribute can be analysed (in theory, at least) into many "unit attributes" (Colless, 1967). In these respects, the "Hennig System" is seen to be a form of statistical, phenetic taxonomy, somewhat intuitive, but perfectly respectable within the limits of commonly available resources.…”

“…In a recent discussion of certain aspects of phenetic taxonomy (Colless, 1967), I explicitly refrained from entering the current controversy between the "phylogenetic" and "phenetic" schools of taxonomy. However, there appeared simultaneously a reply by Wilson (1967) to criticisms of his "consistency test" (Wilson, 1965), and his restatement of that test has "reciprocally illuminated" my own misgivings as to its validity.…”