The fine structure of ‘homology’

There is long-standing conflict between genealogical and developmental accounts of homology. This paper provides a general framework that shows that these accounts are compatible and clarifies precisely how they are related. According to this framework, understanding homology requires both (a) an abstract genealogical account that unifies the application of the term to all types of characters used in phylogenetic systematics and (b) locally enriched accounts that apply only to specific types of characters. The genealogical account serves this unifying role by relying on abstract notions of ‘descent’ and ‘character’. As a result, it takes for granted the existence of such characters. This requires theoretical justification that is provided by enriched accounts, which incorporate the details by which characters are inherited. These enriched accounts apply to limited domains (e.g. genes and proteins, or body parts), providing the needed theoretical justification for recognizing characters within that domain. Though connected to the genealogical account of homology in this way, enriched accounts include phenomena (e.g. serial homology, paralogy, and xenology) that fall outside the scope of the genealogical account. They therefore overlap, but are not nested within, the genealogical account. Developmental accounts of homology are to be understood as enriched accounts of body part homology. Once they are seen in this light, the conflict with the genealogical account vanishes. It is only by understanding the fine conceptual structure undergirding the many uses of the term ‘homology’ that we can understand how these uses hang together.