Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!mailrus!iuvax!cogsci!dave From: dave@cogsci.indiana.edu (David Chalmers) Newsgroups: news.groups Subject: Re: These stubborn group champions Summary: Hierarchies are self-fulfilling prophecies. Message-ID: <27837@iuvax.cs.indiana.edu> Date: 15 Oct 89 18:58:39 GMT References: <8910132350.AA08591@helios.enea.se> <35637@apple.Apple.COM> <33610@looking.on.ca> Reply-To: dave@cogsci.indiana.edu (David Chalmers) Organization: Indiana University, Bloomington Lines: 83 In article <33610@looking.on.ca> brad@looking.on.ca (Brad Templeton) writes: >In other words, who the fuck cares? All this fuss about whether a group >goes in a hierarchy that might give it 5% more distribution? And that 5% >mostly marginal sites that will prune any group they don't like anyway? The issue is not distribution. It goes much deeper than that. If the only difference between hierarchies were distribution, then the differences would indeed be small enough to be irrelevant. The source of the difficulty goes right to the heart of the hierarchy system itself. The idea of the hierarchy system is roughly that high-quality, technical discussions should go in sci (and comp); slightly less serious discussions should go in rec and soc; and endless-flame discussions should go in talk. Nice idea in principle, maybe. In practice, there's a huge problem. There are two key words, one hyphenated: SELF-FULFILLING PROPHECY. It's not just that hierarchy is a function of quality, as it should be. Quality is also a function of hierarchy. Precisely because of the system, certain levels of discussion are regarded as "appropriate" for certain hierarchies. A group in "sci" will receive a higher quality of discussion than the (otherwise) same group in "talk". This is recognized, and there's a tendency for people to tailor their discussions accordingly. Thus, no wonder the fish-people want aquaria in sci. Everybody knows that sci has better quality discussions, and who wouldn't want better quality discussions for their group? Put it in sci, put it in sci, they say. Conversely, no group proposer in their right mind would want their group put into "talk". The talk hierarchy has been specifically designated as the flame hierarchy. Very few people *want* their group a priori to be full of flames (there are exceptions, of course). This is just a hugely flawed system. In particular, "talk" is a ridiculous concept which ought to be eliminated. Few people want to see endless flames on the net (except in news.groups, of course), but creating special flame groups just encourages it. For instance: it's true that abortion is unlikely to get too many calm, rational discussions in the first place: but by demeaning the group with the "talk" title, you are thereby ensuring that it gets even fewer. I sympathize with Norman Gall, for instance, completely: sci.skeptic may not be perfect, but putting it in talk would have ensured that it was total crap. (Misc.skeptic might have been a reasonable possibility, but never mind...) Similarly for talk.rights.human -- how could such an important subject ever get anywhere with the demeaning "talk" title? It's just stupid having a hierarchy which says "all discussions will be content-free" ahead of time. At the very least, "talk" should be eliminated in the next Renaming, if it ever happens. I have no problems with putting most "talk" groups in soc, for instance; it might bring down the average quality of soc (though not by too much), but it would certainly bring up the average quality of those groups and thus, in turn, the net as a whole. (I *don't* buy spreading activation arguments, where the flames in soc.abortion spread like magic to soc.culture.celtic.) There'd still be problems with, say, sci vs rec. We'd have got rid of the "mid-quality vs low-quality" distinction, but we'd still have a "high-quality vs mid-quality" distinction. And a lot of people, like the fishies, would want their group in sci for precisely that reason. Fortunately, in this case, we can base the distinction on something more concrete than "I think this group will be full of flames." Namely, we can base it on whether the group proposed is a scientific field, or a subset of one (these criteria can be relaxed a little perhaps, to include other professional, technical fields, but not much). So, of course the fish people want sci.aquaria, so they get high-quality discussions. (Ah, wouldn't everyone? Why don't we just get rid of *all* the other hierarchies, and make everything sci? Sci.music.dylan, I can see it now %-) But they don't get it, because aquarium-keeping is not a science. Sorry, we sympathize, but that's the way life goes. So hurry up and call for votes on rec.aquaria. [I'd been presuming for ages that the whole point of "sci.aquaria" was to "start the bidding high", and then to "compromise" on rec.aquaria. Then everybody's relieved, so there are no flames calling for it to be in rec.pets.*. But the joke seems to be going on for a long time. Just maybe they're serious? Nah, just enjoying the flames, waiting for a call-for-votes with impeccable timing.] -- Dave Chalmers (dave@cogsci.indiana.edu) Concepts and Cognition, Indiana University. "To live outside the law you must be honest..."