Date: Mon, 24 Jul 2000 10:33:28 +0100 To: Rene Amalberti , Sebastiano Bagnara , Liam Bannon , Bruno Bara , Andre Bisseret , Guy Boy , David Bree , berndt.brehmer@fhs.mil.se, Paul Brna , "C. Cacciabue" , castel@ip.rm.cnr.it (Cristiano Castelfranchi), Axel Cleeremans , Nils DahlbŠck , Bob French , Jean-Michel.Hoc@univ-valenciennes.fr (Hoc Jean-Michel), Erik Hollnagel , Daniel Kayser , Boicho Kokinov , josef.krems@phil.tu-chemnitz.de, Maarten van Someren , Daniel.Memmi@imag.fr, Pierre-Yves Raccah , Herbert Schriefers , Nigel Shadbolt , Rosemary.Stevenson@durham.ac.uk, Oliviero Stock , Gerhard Strube , Niels Taatgen , Michael Tauber , Willemien.Visser@inria.fr, yvonne.waern@tema.liu.se, Dieter Wallach , Richard Young From: Richard M Young Subject: RE: Proposal for cognitive science meetings in Europe Cc: Kim.Plunkett@psy.ox.ac.uk, Keith Stenning Dear Colleagues, This is the second of the two email messages which I hope will wrap up this first phase of discussion about the possible formation of a European Chapter of the Cognitive Science Society (CSS). I have already circulated a revised letter for signature. This message attempts to summarise and pull together the discussion we have had during the past few weeks. Although it is not easy to do, I will try to distinguish between (a) summarising what has been said, and (b) adding my own reactions and comments, which I will put in [square brackets]. Be warned that this message is, unavoidably, rather long. General ------- I received comments from: Bara, Brehmer, Cleeremans, Hoc, Hollnagel, Kayser, Kokinov, Memmi, van Someren, Stenning, Stevenson, Stock, Strube, Taatgen, and Visser. All the comments were constructive and helpful, even from those who disagreed with the proposal, for which I am grateful. Almost everyone who replied agreed that _something_ needs to be done about the organisation of meetings for cognitive science in Europe, but there was of course considerable disagreement about the best course of action. Not surprisingly, and perfectly appropriately, many of the comments dealt with the serious substantive questions that are likely to arise in disucussion over the next year: what would be the relation between a European Chapter and existing European societies, and so on. With regard to the letter itself, apart from a number of people who were not convinced that it was the right thing to do, the main comments concerned (1) the issue of "cultural imperialism" and the nationalism or internationalism of the CSS; and (2) the role of national societies. I tried to respond to both these concerns in the revised letter I circulated recently for signature. A question that often arises is about the international membership of CSS. Here are the figures I got from CSS Secretariat: CSS membership: approx 1400, of which: 72% USA 15% Europe 13% elsewhere (I'm not sure how the Canadians are categorised!) I now identify a number of themes that emerge from the discussion, and try to pull together the points made. I then add some brief conclusions. Who do we represent? -------------------- HOLLNAGEL raises the good question of who we represent, apart from ourselves as individuals? Part of the answer is that several of the people on the email list do have close links to various national or other societies, as Erik himself does, or to sizeable research groups. [Beyond that, the answer has to be that the rest of us indeed "represent" only ourselves, but we also share a concern for the health of cognitive science research in Europe. In the absence of a Europe-wide context for cognitive science -- which is the whole reason why we are having this debate at all -- if we think we need one, then we have no option than to try to bring it into existence. So, although Erik's question is well worth asking, I don't think we need to apologise for trying to act.] KOKINOV (and some other people privately) point out that this present list of email addresses is neither representative nor inclusive enough of "European cognitive science", and that wider consultation will be needed in the next phase of discussion. In particular, the present list is largely based on those who have been active in past ECCSs and similar events, but we clearly won't solve the problem of representation if we confine discussion to just those people. HOC makes a plea for ensuring that we include applied as well as academic research. Organisational "models" ----------------------- The letter to CSS mentions two "models" for how a European Chapter (or indeed, other organisation) could function to hold meetings: A. Scientific society, with individual members and its own conferences. B. "IJCAI" model, as an international committee but no individual members. There are of course several other possible models (as MEMMI and many others point out): C. A confederation of existing societies, possibly including CSS: an ECCAI or IFIP model, or an "umbrella" organisation. D. A single international society with conferences on a rotating basis. E. Chapter-of-a-Society, as with ACM and now SIGCHI. F. Designate an existing European society and persuade them to do it. There hasn't been much explicit discussion of these options, so I'll just list what seem to me some obvious responses to some of them. Model A -- essentially the idea of a European Society for Cognitive Science -- is mentioned in the letter, and dismissed for reasons given there. (This dismissal is supported by BREHMER and implicitly at least by several others.) Model B -- an international committee -- seems to appeal to several people (including STEVENSON). Model C -- a confederation -- runs into two objections. One is: what happens about cognitive scientists in countries which don't have a national CS society (CLEEREMANS, TAATGEN), or who happen not to belong to any of the other, smaller societies? The other is, as with Model A, that there has been plenty of opportunity for this to happen over the past, say, five years (dating from ECCS-95), but nobody has done it, and there is disbelief that it will happen spontaneously (HOC) or is feasible (BARA). Model D -- a single, international society with conferences held in different parts of the world -- may well be a possible move, especially if the CSS changes into a truly international organisation (e.g. HOC, KAYSER). However, it is unclear that it would by itself solve our problem. Realistically, we could not expect the annual conference to be in Europe more often than once every 3 or 4 years, so we would still need a European meeting for at least some of the other years (VISSER) Model E -- an international society but with regional Chapters -- may be a better variant on D, and has apparently been successfully adopted by the Association for Computational Linguistics (STOCK). Model F -- persuading an existing society to do it -- has been proposed (STRUBE) but encounters a number of objections. [First, there is the same problem as for A & C: why hasn't it already happened? Second, it is unclear why it would be in the interest of the society to do such a thing. But thirdly, and principally, it seems unlikely that it would succeed in meeting the objectives. How would ESCOP (European Society for Cognitive Psychology), say, cut through the confusion of competing allegiances? Personally, I would no more trust the cognitive psychologists to run cognitive science than many other people would trust, say, an Artificial Intelligence society to do it -- although of course I agree it will be important to include them (and many others) in the next round of discussions.] Should we involve CSS? ---------------------- Even if we agree on the "model" for an organisation, in most cases the question still arises as to whether or not to involve the CSS. Because the CSS either is, or is seen as being, predominantly a North American society -- in membership, in geographical location, and in approach -- feelings can run quite high on this issue. The arguments _against_ involving CSS seem to be that it would be better if we did it ourselves (KAYSER, STEVENSON, HOLLNAGEL), and there are worries about the freedom and autonomy of a European Chapter (KAYSER). There is a distinctive European flavour of cognitive science, which might not be recognised (HOLLNAGEL). The question is raised as to why CSS would be interested in helping the Europeans (STEVENSON, STRUBE). The arguments _in favour_ of involving CSS include the belief that this could help raise the standard and representativeness of European meetings (KOKINOV). It could be good for smaller, internationally oriented countries, where a national meeting is not really viable (van SOMEREN, TAATGEN), and would be easier than starting from scratch (van SOMEREN). Apparently CSS is keen to move towards being more international (STENNING, STRUBE, KOKINOV), and there are some good models of internationalisation, such as ACL and CHI (STOCK, VISSER). Relationship to national (and other) societies ---------------------------------------------- The existence of a European cognitive science organisation, whether a Chapter of CSS or not, will clearly have some impact on the existing national societies. There should still be room for national societies to play a role (KAYSER), and it is important to involve them in any further discussions (BARA, MEMMI, KOKINOV). [Incidentally, I know of ARC in France, GK in Germany, and the Bulgarian society -- are there any others?] There are obviously pressures on a national society's annual conference if a large, international conference is being held in the region (STRUBE). One route is not to hold the national conference in years when a European conference is being held (van SOMEREN, CLEEREMANS). [Another possibility is occasionally to _merge_ the conferences, so that perhaps one year GK agrees to organise the European or international meeting, and combines it with its own annual meeting.] STENNING points out that a European organisation -- again, with or without the involvement of CSS -- does not solve the question of _language_. If the working language of an international conference is English, as seems mostly to be the case, then non-native-English speakers are at a disadvantage. [Using two languages with simultaneous translation is expensive and not wholly satisfactory.] None of the proposals addresses this problem. The question of journals arises as well (HOC, van SOMEREN). At present, a subscription to "Cognitive Science" is included with membership of CSS. There are questions about how that would be handled in a European Chapter, and of whether CSQ ("Cognitive Science Quarterly") might somehow be involved. Conclusions ----------- 1. There is broad agreement (although not _total_ agreement) that something should be done about organising Europe-wide meetings on cognitive science. 2. There appears to be sufficient agreement to proceed with sending the letter to CSS, even if the discussion might be broadened to the question of "what is an appropriate organisation for cognitive science in Europe, and how might it relate to CSS?" rather than a simple "Let's form a European Chapter of CSS". 3. The national societies, and perhaps other societies too, have legitimate concerns about the effect that a European Chapter of CSS would have on their membership and activities. ----------------------------------------------- If anyone objects strongly to anything in this message, then by all means circulate a reply. But otherwise, I suggest that we leave further discussion for now, until after the response from CSS is known, say until September or October. Many thanks to everyone for their co-operation. -- Richard Richard M Young Professor of Cognitive Science __________________________________________________________________________ Dept of Psychology, Univ. of Hertfordshire, Hatfield, Herts AL10 9AB, U.K. Tel: +44 1707 28 5051 (dept office 4612), Fax: +44 1707 28 5073 Email: R.M.Young@herts.ac.uk http://www.psy.herts.ac.uk/pub/R.M.Young/