Many different and even contrasting notions of homology have been proposed over two centuries of comparative biology, beginning beginning with the initial reference to idealised archetypes, down to the current concepts based on derivation from a common ancestor, or on shared developmental pathways or genetic (more extensively, informational) background. Select anatomical features such as the patterns of innervation, or the expression patterns of genes putatively involved in key developmental events, e.g. the Hox genes, have been repeatedly suggested as the most reliable cues to homology. This confidence, however, rests on shaky ground and results are never certain. Recent work in comparative morphology and evolutionary developmental biology increasingly suggests the need to abandon the traditional all-or-nothing notion of homology, in favour of a more flexible, factorial or combinatorial approach. In this way it will be possible to accommodate within one broad comparative view, respectful of phylogeny and developmental biology alike, many disparate notions such as positional and special homology, serial homology and temporal serial homology. All statements of homology, however, will thus require inadequate qualification of the context specifically taken in consideration and the criteria used to address the comparison. It remains to be seen, in the near future, how far the old concept of homology, now under the burden of so many and so different notions, will still be of use to comparative biology.