Numerical taxonomy
The numerical taxonomy or phenetics is a classification system in biological systematics . The aim is to establish a biological taxonomy based on morphological features. Intensive use is made of computer-aided calculation methods and, in particular, cluster analysis . The concept of numerical taxonomy was developed in 1961 by Robert R. Sokal and Peter Sneath (1923-2011) in the book Principles of Numerical Taxonomy .
Motivation and Practice
The morphological orientation of numerical taxonomy was already in contrast to the majority of taxonomical theory and practice when it was developed. The major part of the biological system is based on phylogenetic development. However, Sokal and Sneath explicitly want to leave out evolutionary theoretical considerations. The aim is to develop a general taxonomic system that is as theory- free as possible and that is equally binding for the various disciplines such as ecology , evolutionary biology or paleontology .
The numerical taxonomy is based on a large number of morphological features, all of which must be observable and quantifiable. The extent to which the characteristics are assigned to individuals is determined. Using complex, computer-aided procedures, an objective measure of the similarity of individuals is to be found. The similarities can then serve as material for a taxonomic system.
Criticism and inclusion in research
The numerical taxonomy has not been able to establish itself in biological systematics. On the one hand, evolutionary biologists reject a purely morphologically oriented taxonomy as unrealistic. On the other hand, there are serious epistemological objections to the phenetic conception of an objective taxonomy. Morphological species concepts are partly subjective due to the necessary selection and weighting of the morphological characteristics and often guided by available methodology and specific research interests. On the other hand, the majority of taxonomists today believe that taxonomy should be based on phylogenetic criteria. Since the cluster algorithms used in numerical taxonomy do not differentiate between convergent and homologous features, they do not result in a correct reconstruction of the phylogenetic (evolutionary) processes in a large number of cases. Convergent evolution often influences a large number of morphological features together (e.g. when comparing birds and bats), so that a purely quantitative assessment is misleading.
However, the problems of numerical taxonomy do not change the fact that it is used successfully in some sub-areas. This applies in particular to areas in which other species concepts cannot be used. This is e.g. This is the case, for example, in paleontology, where, for example, Ernst Mayr's biological concept of species often cannot be used, as there is usually no information available about the possibilities of reproduction .
The claim to have presented a generally applicable taxonomic method can, however, be viewed as a failure. However, this does not rule out the possibility that phenetics could have a permanent place in a plurality of taxonomic systems.
Numerical Taxonomy in Linguistics
Based on Sokal / Sneath, classifications were also carried out in various areas in linguistics. Here, too, the aim is to use classification methods that are as objective as possible. When applied to questions of language typology, results were obtained based on the morphological and syntactic properties of the languages that largely correspond to the historical classifications. Further attempts at classification are based on the distributional properties of the phonemes of the languages.
literature
- Robert R. Sokal, Peter H. Sneath: Principles of Numerical Taxonomy. Freeman Books, San Francisco 1963, ISBN 0-7167-0621-0 .
- Peter H. Sneath, Robert R. Sokal: Numerical Taxonomy: the principles and practice of numerical classification. Freeman Books, San Francisco 1973, ISBN 0-7167-0697-0 .
Individual evidence
- ↑ See chapter "Taxa" in: Gabriel Altmann, Werner Lehfeldt: Allgemeine Sprachtypologie. Fink, Munich 1973, page 34ff. ISBN 3-7705-0938-2 .
- ↑ See chapter "Distributional Language Classification" in: Gabriel Altmann, Werner Lehfeldt: Introduction to Quantitative Phonology. Brockmeyer, Bochum 1980, page 282ff. ISBN 3-88339-150-6 .