Hox genes, homology and axis formation--the application of morphological concepts to evolutionary developmental biology.

This article focuses on the interphyletic comparison of gene expression patterns. By means of the hypothesis of the inversion of the dorsoventral axis during the evolution of the Bilateria, it is demonstrated, that evolutionary developmental biologists use similarities in spatial and temporal gene expression patterns as evidence for the formulation of hypotheses of homology concerning either developing structures or body regions. The molecular genetic and morphogenetic evidence used is discussed within the framework of a cladistic-phylogenetic analysis based on the phylogenetic tree of the Bilateria. I argue that similarity of spatial and temporal gene expression patterns is not a sufficient criterion for homology inference. Therefore, gene expression patterns should be coded as characters. Their homology should be tested in concert with other characters. Furthermore, it is demonstrated, that spatial and temporal similar gene expression patterns, indicating similar molecular genetic mechanisms, were interpreted as an analytical criterion of homology, offering the possibility to identify similar structures. In contrast to this, the evolutionary developmental biologists have not developed a causal-analytically extended concept of shape, from which a causal-analytical concept of homology could be deduced. Instead, the homology concept from evolutionary morphology is used.