Path: utzoo!news-server.csri.toronto.edu!rutgers!cs.utexas.edu!uunet!mcsun!ukc!servax0!csc2!scotp From: scotp@csc2.essex.ac.uk (Scott P D) Newsgroups: comp.ai Subject: Re: AI genealogy Message-ID: <4867@servax0.essex.ac.uk> Date: 8 Mar 91 13:23:52 GMT References: <5466@media-lab.MEDIA.MIT.EDU> Sender: news@servax0.essex.ac.uk Reply-To: scotp@essex.ac.uk (Scott P D) Organization: University of Essex, Colchester, UK Lines: 38 In article <5466@media-lab.MEDIA.MIT.EDU> minsky@media-lab.media.mit.edu (Marvin Minsky) writes: >Quite so. Furthermore, Warren McCulloch and Alan Turing were not at >that conference. Or Donald MacKay, etc. I could not possibly be more >pleased than I am by how much my own students have accomplished, but >the formal relation of thesis advisor is not a very good indicator of >the evolution of ideas in the community of international science. I have never met Gabriel Velasco, the originator of these genealogical enquiries. However it seems I am to be cast in the role of his intellectual grandfather, since I was his advisor's advisor. I am afraid I have bad news for Mr Velasco that throws doubt on his academic legitimacy: though very active in the field at the time, my own advisor was not among the immortals attending the Dartmouth conference. If Mr Velasco wishes to remove this blot from his escutcheon, several possibilities are available; all of which have the effect of extending the category of founding fathers to include people, such as those Marvin Minsky lists, who were important to the field but didn't happen to go to one particular conference. For example, one might also classify anyone who published anything in AI prior to 1960 as a founding father -- I think the bibliography in Feigenbaum and Feldman (1963) would make it fairly easy to create such a list. While a change like this would bring Mr Velasco back into the family, it wouldn't do much to remove my main reservations about the proposal. Thesis advisors are only one of many influences on graduate students, and anyway most of us spend far more time engaged in AI and being influenced after completing our degrees than before. Much of this type of influence could be captured by considering the laboratories where people have both studied and worked, although this too would have serious limitations -- some of us occasionally read stuff written by people we have never met in places we have never visited. Thus ultimately I don't see a way of avoiding actually considering what AI practitioners have written about their own work -- for example by doing a citation analysis. This would obviously be a mammoth undertaking, but do we really want to provide more ammunition for those who compare us to drunks searching for keys under lampposts? Paul Scott, Dept Computer Science, University of Essex, Colchester, UK.