Non-circular Definition of Homology/Analogy

Homology describes the existence of similarities between organisms that are the result of descent with modification. As such, homology is generally a hypothesis, which when supported, provides evidence for evolution.If one rejects the notion that species arise through descent with modification, there is by definition no such thing as homology, there is merely similarity. Creationists are completely wrong-headed if they argue that "homology is best explained by the creator using similar designs for similar functions" or something along those lines. From my definition of homology above, it is clear that these kinds of arguments make no sense. What they really ought to be arguing is that "there is no such thing as homology. What you think is homology is merely similarity".How might we go about testing the hypothesis of homology? The way I look at it is that there are similarities between organisms, whether one accepts evolution or not. Take the genetic code, for example. There is no arguing about the fact that the genetic code is remarkably similar across all organisms, yet some differences exist (with a vertebrate code, several invertebrate codes, etc).Evolutionary theory successfully predicts the distribution of these similarities across different life forms. For example, it is a straightforward prediction of evolutionary theory that the genetic code of a mouse and rat will be more similar than the genetic code of a mouse and a nematode, because rat and mouse share a common ancestor more recently than rodent and nematode.The distribution of similarity across species provides evidence that homology is a real thing in nature. Not just one protein, but all proteins, all physiological functions, all behaviours and all developmental processes are more similar between rat and mouse than they are between rat and nematode. The hypothesis that this vast array of similarities exists by virtue of being the result of descent with modification is then simply the most parsimonious explanation of the observed data.Creationists have to argue that these patterns essentially mean nothing, that the observed data arose through an unidentified mechanism which simulates homology without really being homology. It doesn't take much thought to find such a viewpoint unsatisfactory. It neither explains nor predicts anything. As such, the distribution of similarity between organisms appears to reflect homology, and is predictable on the basis of phylogenetics, which is strong evidence for the reality of the evolutionary process.Until some creationist ventures to explain the repeated correlation found between phylogeny, form and function, it seems wise to accept the reality of homology.Mick