Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!think.com!spool.mu.edu!uunet!munnari.oz.au!goanna!ok From: ok@goanna.cs.rmit.oz.au (Richard A. O'Keefe) Newsgroups: comp.lang.eiffel Subject: Re: On classification Message-ID: <5411@goanna.cs.rmit.oz.au> Date: 26 Apr 91 08:28:43 GMT Article-I.D.: goanna.5411 References: <527@eiffel.UUCP> <1135@tetrauk.UUCP> <308@alfrat.uucp> Organization: Comp Sci, RMIT, Melbourne, Australia Lines: 34 In article <308@alfrat.uucp>, guest@alfrat.uucp (Mr. Guest User) writes: > Aha! finally in this dinosaur business there has been a mention of something > to do with Object Oriented Concepts. I'd like to point out that biological classifications do have something to teach us. The key idea of "cladistics" is to base taxonomies on actual lines of descent, and the groups thus formed are monophyletic. To put it another way: SINGLE INHERITANCE. The techniques were not designed to be useful when multiple inheritance (e.g. transfer of genetic information by viruses from one bacterial strain to another) is significant. There is another lesson: cladistic classifications break up "traditional" groups which have a fair bit of heuristic power. Consider, for example, Gould's article "What, if anything, is a Zebra?" in which he explains how the traditional category "zebra" (horse-with-stripes) contains species from two separate groups. Fine, so "zebra-ness" is not something that was inherited from a common ancestor. That doesn't mean it isn't a useful category. Horse-like animals with stripes clearly inhabit a different ecological niche from horse-like animals of a solid colour, and it's quite likely that we cluster information about habitat, diet, predation, &c rather better using zebra/horse as our categories than if we used the "real" evolutionary divisions. Similarly, "fish" is not a valid category if you are solely concerned with actual lines of descent, but those sort-of-cylindrical non-homeothermic vertebrates that live in water _do_ have a lot in common with each other that they don't have in common with creatures which are "closer" in a cladistic classification. One lesson to be learned from this is that there may be >no< "right" classification system. If you have classes A, B, and C, with B derived from A, it may still be the case that B and C have more in common than A and B. This is one reason for multiple inheritance. -- Bad things happen periodically, and they're going to happen to somebody. Why not you? -- John Allen Paulos.