tag:blogger.com,1999:blog-8697616560977719586.post6911991389837718377..comments2023-10-28T01:53:23.721-07:00Comments on Systematics and Biogeography: Abstracting and Seeing – Homology and SimilarityMalte C. Ebachhttp://www.blogger.com/profile/11901602320985626811noreply@blogger.comBlogger1125tag:blogger.com,1999:blog-8697616560977719586.post-10204682547647307122007-11-07T17:18:00.000-08:002007-11-07T17:18:00.000-08:00I read several of your earlier posts and this one ...I read several of your earlier posts and this one finally clarify several points of your series about '<I>phenetics revival</I>'. But there are several points in which I disagree, or (I feel) remains unexplained.<BR/><BR/><B>Abstracting and seeing</B>. I am not sure about the point here, and the examples confuse the matter. It is a right comparison to contrast microbes with reptiles, electrons with unicorns? I think it would be better if you show, explicitly why saying "mammals have hair" is 'seeing' and not an 'abstraction'.<BR/><BR/><B>Taxonomy and identification</B>. From this phrase:<BR/>"<I>Seeing a dog does not make us recite a list of mammalian characteristics. We just see and recognize it</I>"<BR/>and your previous post about barcoding, it seems that you a simplistic view of taxonomy. If taxonomy is a science you can't say "this is a dog, because I know it is a dog", you would provide evidence that you are seeing a dog, how taxonomist do it? They give a list of characters, this is valid not only for species but for more inclusive monophyletic groups.<BR/><BR/><B>Homology and sameness</B>. I agree with you, homology imply that an structure shared by to organism is the same. They are the same by virtue of heredity, so homology (sameness) implies a complex ideas about genetics, development, and evolution. In your previous post about words, you prefer the very old definitions of Owen (and others), that do not imply evolution, and you say:<BR/>"<I>Definitions such as archetype, homology, evolution and natural classification that originally had no explanatory mechanism, helped to establish systematics and biogeography. Why do we need them now?</I>".<BR/>The question is, if that is your opinion, then what explictly is sameness?<BR/><BR/><B>Homology and relationship</B>. I have a hard disagreement with you at this point. Homology is relationship, only if it imply novelty (apomorphy). For me, this is the major development of cladistic analysis. Homology by itself only imply that two structures are the same by inheritance. For example, the eggs lay by spiny ant eater (echidna) and platypus are the same, they are homologs, but by no mean the eggs are evidence that they form a monophyletic group. Only if a character is a shared homolog (same in two taxa) and shared novelty (the same but transformed if it is compared with a third group) is that relationship could be inferred (or discovered).<BR/><BR/>Optimization using parsimony (cladistic analysis) it is rooted in the concept of sameness: if two taxa placed in the same group have the same structure, and we are committed that both are the same, then whole group have the same structure. A computer algorithm to do it is straightforward (parsimony optimization), you could made an 'intuitive' analysis, and the list of apomorphies found in the same cladogram need to be equal in both methods, unless you use "<I>ad hoc proposition of evolutionary mechanisms</I>".<BR/><BR/><B>Bad usage of algorithms</B>. I agree with you, some people use good algorithms (like cladistic optimization) in the wrong way, but that not mean that the algorithm is bad, or the people that use the algorithm in the fair way are phenetisist that rejects the idea of homology for raw similarity. Actually bad usage lies in bad character definitions, instead of that methods are phenetic or not. At least parsimony, implies homology, so characters used in a cladistic analysis would be characters that imply sameness, so your criticism is wrongly orientated, it is not against algorithms, but against bad characters (I like the discussion by Jenner about that topic: <I>Syst.Biol.</I> 50:730 (2001), and his free papers at <I>contributions to zoology</I>!).<BR/><BR/>There are of course, some 'cladistic' algorithms that ignores sameness, that are guilty of several of the accusations you give, like three taxon statements (TTS), in which the evidential weight of each character is destroyed, and in the end you got a 'cladogram' in which no apomorphy could be identified (unless you use 'intuition' or more rigorously an optimization). It is not strange that some defenders of TTS conclude that morphologic characters have no meaning! (see Scottland et al, 2003, <I>Syst.Biol</I>, 52:539).<BR/><BR/>Nice blog and keep sharing your ideas :)...Salvahttps://www.blogger.com/profile/01062764779798191688noreply@blogger.com