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 Message-ID: <27858@iuvax.cs.indiana.edu> Date: 16 Oct 89 04:22:12 GMT References: <8910132350.AA08591@helios.enea.se> <35637@apple.Apple.COM> <33610@looking.on.ca> <27837@iuvax.cs.indiana.edu> <34029@looking.on.ca> Reply-To: dave@cogsci.indiana.edu (David Chalmers) Organization: Indiana University, Bloomington Lines: 93 In article <34029@looking.on.ca> brad@looking.on.ca (Brad Templeton) writes: >I am sorry, but while this idea sounds intuitive, an actual examination of >net history shows little evidence for it. The "talk" groups are noisy, >not because they are in talk, but because they were put there to be noisy. > >Almost all of them were created in the "net" hierarchy, which had no >stigma, and moved into the talk hierarchy later. So it's simply false to >claim that the "talk" hierarchy is what makes them noisy. This perhaps holds true for groups which have been around a long time, pre-Renaming. The content of such groups is determined more by their history than their names. In particular, at the time of the Renaming, it may have been fair to put such groups in a "talk" hierarchy precisely because they had been given a chance to prove themselves, and had proven that they were noisy. Such an argument doesn't hold for new groups, though. Such groups have no tradition to rest on, on the content is determined almost entirely by the name. True, the "suffix" plays the biggest role here, but the prefix plays a role too. I have no doubt that "rec.math" would be quite a different group to "sci.math" (not that sci.math is by any means perfect...). >Nor do I note any particular difference in character amongst the newly >created groups added to hierarchies. Sci.skeptic is plenty noisy, as >has been noted, in good old "sci." I am hard pressed to come up with >any group that I feel would have a different character in a different >hierarchy. Well, others differ, it is clear. Proponents of *.skeptic, *.rights.human, and *.aquaria all seemed to buy the self-fulfilling prophecy argument, and I agree with them. (Incidentally: I just took a look at sci.skeptic, and for all that everyone's been saying about it, it's not all that bad. There's a significant amount of interesting discussion about a range of topics, and the flaming was by no means as widespread as I had been led to expect. Discussion seemed to be on a par with most sci groups. Talk.skeptic would have been significantly worse, I imagine.) -- Elsewhere, Richard Shapiro suggests a new top-level domain for high-quality discussions. This is precisely the kind of thing I was trying to get away from. If domain names are based only on the quality of discussion, the "self-fulfilling prophecy" law rears it's head. Everybody will then of course want their group in this domain -- because everybody believes, or at least hopes, that their group will be high-quality. If groups have had some time to demonstrate their quality -- as, for instance, prior to the Great Renaming -- then decisions about which groups go where can be founded on some kind of "fact of the matter." But when we are talking of the creation of new groups, there is no fact of the matter to base the decision on. Instead, the decision itself will (partly) determine the quality of the group. What I suggest instead is that quality of discussion should *never* be an explicit factor in the naming of new groups. If it is, then on comes the self-fulfilling prophecy. I'm not sure how many new groups have been created in "talk" post-Renaming, but I'm sure that as per prophecy, noise is high. (Am I alone in finding "talk.rights.human", for instance, a repugnant idea?) To get away from this eternal problem, the name of a new group should be based *solely* on the nature of the proposed group, and not at all on the predicted quality. If we have well-defined, quality-independent domains, we can get away from the idea of "stigma" associated with various hierarchies. Fortunately, this isn't too hard to do. The definition of comp is pretty clear, sci is for science-related matter, rec is for recreational activities, soc is for matters pertaining to society, and so on. In practice, such decisions would be rather easy. Aquaria goes in rec, rights.human goes in soc (as do abortion, politics, etc), and I tend to believe that skeptic should go in sci (as an important part of science is clarifying its fringes), though a case could be made for misc. "Talk" should go completely, as there is *nothing* concrete about its definition other than a prediction of quality. (At the bottom line, "talk" may have been a good idea at the time of the Renaming, as at that stage it applied only to old, proven to be noisy, groups. Unfortunately, not enough thought was given as to how the hierarchy might apply to new groups.) Actually, the only concrete thing about this proposal is that we should get rid of the "talk" ghetto, and clearly that's not going to happen overnight. But I do suggest that no new groups should be created in "talk" (except the obvious, like .bizarre and .religion.computers), as this damns them ahead of time. The other proposal is that distinctions between the other hierarchies which are *solely quality-based* should go. In particular, it is not valid for aquarians to argue "we're smart people, therefore sci." Further, it's not valid to say ".skeptic will be full of flames, therefore not sci." If domain distinctions are concrete, stigmas will lessen. -- Dave Chalmers (dave@cogsci.indiana.edu) Concepts and Cognition, Indiana University. "To live outside the law you must be honest..."