Search for a command to run...
The history of the homology concept is a history of attempts to conceive the basis of sameness in biology. Since it was formulated in the middle of the 19th century, the concept has had to fit an ever growing number of scientific fields and purposes. These different demands have resulted in diverging, sometimes, incompatible definitions. The inconsistencies are mostly due to the lack of a clear separation of hypotheses of maintenance from hypotheses of transformation. A synthetic approach to define homology thus has to consider the following pivotal points: (i) hypotheses of evolutionary maintenance should be kept separate from hypotheses of evolutionary transformation; (ii) the definition of homology should provide the foundation for exact specifications of what is hypothesised to be homologous and (iii) restrictions to particular levels of observation or specific scientific purposes, and the exclusion of iterative homology should be avoided. We suggest that patterns should be delineated by characterizing components of traits, and by describing connections and interactions between these components. A shared pattern of compared traits where the characterization shows 1 : 1 correspondence may then be homologised. Homology is equivalent to a hypothesis that the pattern, starting from a single starting point, was transmitted robustly along diverging branches of a genealogical tree, that is, the homologised pattern was never changed by any transformation. The proposed definition of homology is thus, ‘ A pattern corresponding in a set of compared traits is homologous, if after a common evolutionary origin, the pattern was maintained along diverging lineages by robust pattern transmission ’. After justifying the terminological use in our definition, we discuss the interplay of our definition with the pivotal points mentioned above in comparison to other definitions. Since our homology definition is a concept of pattern maintenance, it is clearly demarcated from transformation hypotheses, which are covered by the character concept. Robustness is understood as evolutionary maintenance of correspondence in objects linked by genealogical relations. The characterization of the pattern suffices to provide the necessary conditional phrase by specifying what is hypothesised to be homologous. Allowing development to be conceptualised as a pattern formation process makes it easier to deal with traits that are transmitted indirectly to the next generation. Patterns can be characterized on all observational levels, but the components and the quality of connections and interactions used for the characterization may differ. The replacement of the reference to an ancestor–descendant relationship by a reference to robust pattern transmission allows for the inclusion of iterative homology into the concept. In the final part of the paper, detailed reformulations of the ‘criteria’ for the corroboration of homology hypotheses as proposed by Remane (1952 ) are given.