Evolution and classification : the reformation of cladism / Mark Ridley.
- Mark Ridley
- Date:
- [1986]
Licence: Attribution-NonCommercial 4.0 International (CC BY-NC 4.0)
Credit: Evolution and classification : the reformation of cladism / Mark Ridley. Source: Wellcome Collection.
39/220 page 27
![The techniques and justification of evolutionary taxonomy The method is not valid unless the character in question has been tested, whether on pre-existing groups or fossils. It must have been shown to correlate with other characters, to define a natural group. If it was used to define the pre-existing group, other characters must be shown to have the same taxonomic distribution; if it was not originally used to define the group, it must subsequently have been found to fit the group well. If a character has been so tested, and has passed the test, we have evidence that it is homologous. In the manner already described, we can then reason, provisionally and in proportion to the case, that it is probably homologous in other, unclassified specimens. I say provisionally. Evolutionary classifications are always subject to improvement. Each new piece of evidence can be used to test the existing groups, to test the current ideas about which characters are homologies and which analogies. If evidence piles up against a classification, it can be changed; and the revised classification will be superior to the one it replaces. Of course, the taxonomist's initial crude ideas of the classification of a group will influence the assessment of whether the next character under study is taxonomically good or bad: errors may be perpetuated, 'but the effect of the early classification decreases as re-classification takes place in the light of new evidence' (Hull 1967, p. 180). The whole process is one of 'successive approximation'. 'Classification and re-classification ... goes on all the time in evolutionary taxonomy in the light of the discovery of previously unknown species and additional evidence' (Hull 1967, P- Critics of evolutionary classification - and especially the numerical taxonomists we shall meet in the next chapter - have often mistaken this procedure of successive approximation, or of 'reciprocal illumination' (Hennig 1966), for argument in a circle (see especially Hull 1967). The process clearly is not circular, but a normal scientific sequence of an initially crude hypothesis, a test, a hypothetical extrapolation, a further test, and so on. It is a process of theory- building. Its critics were led to their conclusion by a different understanding of evolutionary classification. Sokal and Sneath (1963, p. 7), for instance, remark that circular reasoning arises from the fact that new characters, instead of being evaluated on their own merits, are inevitably prej udiced by the prior erection of [another] taxon A based on other characters (Ä). Such a prejudgment ignores the fact that the existence of A as a natural (or 'monophyletic') group defined by character complex X has been assumed but not demonstrated. In other words, according to Sokal and Sneath, characters are not tested before being used to assess new specimens. Although it sounds 27](https://iiif.wellcomecollection.org/image/b18021451_0040.JP2/full/800%2C/0/default.jpg)
