The European Society for the Study of Cognitive Systems was founded in 1983, with the first workshop in The Netherlands (Noordwijkerhout, 24-26 May). There are two main current activities of the Society: the publication of a journal "Cognitive Systems", and the organization of workshops. There is an annual workshop, and now and then a special workshop on a specific topic related to cognition, but from a multidisciplinary perspective. In 1997 the 15th annual workshop took place in Germany, Freiburg im Breisgau (2-4 June). The next annual workshop will be in Siena, Italy, 24-26 October 1999. Now and then scientists take part from countries outside Europe. The places where the workshops are held, are attractive, like Oxford, The Provence (France), Vienna, Northern Italy, so that a visit to places of interest can follow a "visit" to the workshop. Further information that is not available in this webpage may be obtained from the address at the bottom.

The Society was founded with two purposes: to make a start with European cooperation and integration of all research concerned with a multidisciplinary approach to cognition, and, secondly, to promote the idea of this approach in itself. Until the present day, the scientific investigation of the mind and the brain can be characterized by the existence of a strong separation of the various specific fields involved, even between subfields within disciplines. This is probably largely due to the fact that some of the main disciplines concerned (psychology and the neural sciences) are highly empirical in nature, whereas the multidisciplinary approach is in the first place a matter of theory and method, and science-theoretical in nature. Until now, four special workshops were held. The last one took place in December 1997, on "Man-machine Systems and Social Cognition" (ECS-MMS97: European Cognitive Science - Man-Machine Systems 1997). The perhaps curious combination of topics came about by the fact that already a first workshop on models for man-machine interfaces was held in Vienna, Dec. 1994, the "First interdisciplinary workshop on cognitive modelling and user-interface development". There was a suggestion from one of the participants to an annual workshop, to have a special workshop on the topic of social cognition. Since we do not want to have too many workshops, the idea was born to combine the two topics. Moreover, it seems to lead to some very interesting questions: to what extent do we interact with machines in the same ways as we interact with other people; to what extent is our interaction with people influenced by the use of machines, or by the fact that we so often interact with machines. And also: in what way do we "understand" machines, is this very different from the ways we understand other people, or ourselves. (The full programme can be found further down). This special workshop was supported by EACE and the German Society for Cognitive Science (Gesellschaft fuer Kognitionswissenschaft). A number of the papers of the first workshop on this topic were published as a special issue of the journal of the ESSCS, "Cognitive Systems", issue 4-3,4 (double issue).

The ESSCS is concerned with cognitive science in a very wide sense. In 1983 there was not yet clarity as to the domain of this new discipline. The functionalist view was still strong: well-known psychologists held the view that it would not be necessary to include the underlying neural level to understand cognition in itself. Cognition would have to be understood in the manner similar to the way one can understand a computer program, without a need to know how the hardware works. Of course this is true for a number of questions. But there are also questions that cannot only be answered at the functional level, for example how the brain organizes itself. The functional properties of the brain are not put there by some clever programmer, but grow on the basis of what the individual inherits, and from interactions with the environment. This aspect of self-organization is until now only understood to a limited extent. In artificial intelligence there existed for some time a subdomain concerned with constructing programs that in turn were to construct other programs that could learn. Until now that attempt has not been very succesful. Only for limited domains, such as games, it has been possible to create such general programs, of which also the performance is limited. The question has been overtaken by the approach of connectionism, where systems have to be trained. They are very different from the classical programs of artificial intelligence. But even connectionist systems are not truly self-organizing, they have to be trained by an external supervisor. They have several fundamental deficiencies, like the inability to represent procedural knowledge that can be bound to arbitrary pieces of declarative knowledge. Another problem is catastrophic interference: if a system is trained to recognize one set of patterns, or to perform one task, and then for another set or task, the skill for the first one is lost. Fortunately this is not the case with humans, but we do not know why. It anyhow shows that connectionist systems in their present form of three layers cannot be good models for the human brain. Although insight in these matters did not yet exist to the extent it exists now, the idea that the underlying neural level is crucial did exist in 1983, and long before. That was then the main reason not to call the new society as related to cognitive science as it existed then, but to use the term 'cognitive systems', as a more general term. For all of these reasons the society is very multidisciplinary, and concerned with aspects of psychology, including developmental and social psychology because of their importance for cognition, the neural sciences as far as they are important for cognition, linguistics as far as it provides a window onto cognition, and philosophy as far as it is concerned with questions how the different levels of description must be related. There is an interaction with the more technical sciences like artificial intelligence, connectionism and neural networks, artificial life, in the sense that now and then workers in these fields participate in the workshops, or submit articles for the Journal. A basic requirement for acceptance is that the papers are multidisciplinary in character, and concerned with some important aspect of cognition and perception, as these aspects are the central concern of the ESSCS.

In 1985 it was decided to start a journal, "Cognitive Systems". It still is very modest in frequency of appearing, perhaps partly because of its emphasis on multidisciplinarity, partly because it is published by the Society, and not by a commercial publisher. Commercial publishers are not keen on multidisciplinary approaches, because the target group is so scattered and unclear. The circulation is small, and the number of papers submitted also. Some time was needed to find the proper position of this approach, and now gradually the time may have come to find ways to increase both. Until now, artificial intelligence has dominated the domain of theoretical research in cognition. This can still be seen from the fact that in the two oldest societies concerned with cognition, the role of separate disciplines and of artificial intelligence is important, or even dominant, in numbers of people, numbers of papers, and money involved. The two oldest societies are the "Association pour la Recherche Cognitive" ("ARC", France), and the Society for the Study of "Artificial Intelligence and Simulation of Behaviour" ("AISB", United Kingdom). The ARC publishes a small newsletter with only technical news, and a journal, "Intellectica" three times per year, usually on special issues. Most articles are in French. The AISB publishes four times per year a bulletin, with mainly news, reports on conferences, and short papers.

In Germany a small subgroup exists for cognition of the important society for computer science (Gesellschaft fuer Informatik). There are several related subgroups, the strong subgroup of artificial intelligence, and others like the ones on pattern recognition and artificial life. The subgroup on AI publishes a journal on AI, but it is published commercially. It organizes conferences annually, that gradually have grown to a considerable size. The Journal and conference are mainly in the German language. (Papers may be presented in English). Surprisingly, there also is a separate society for cognitive science, that publishes a journal "Kognitionswissenschaft", also published by a commercial publisher. It contains mainly articles in the German language.

In The Netherlands there existed a society for artificial intelligence, that now and then paid some attention to matters of cognitive science. In 1998 it combined with the Belgian organisation for AI to the BNVKI: The Belgian-Netherlands Society for AI. Smaller societies exist in Italy and Bulgaria. The society in Bulgaria organizes a summer school every year. Around some big cities in Europe there also exist interest groups that regularly organize lectures and small symposia, like Paris (France) and Munich (Germany). These are all in the national language. In other countries the organization is also in its very early stage of development. It is a general problem in Europe to create more integration and cooperation, due to the different languages used by the respective national societies. Of course, English is the general language, with which one can communicate in all these countries within these disciplines. But for many of the national activities, the national language is most important.

There are some other European organizations that to some extent are concerned with cognition: the European Association for Cognitive Ergonomics (EACE), and the European Association for Research on Learning and Instruction (EARLI). The former cooperates in the special workshop mentioned before. Both are likely to take some part in organizing cognitive science in Europe in the future. In 1995 the first European conference of cognitive science (ECCS '95) took place, in St. Malo, France. The ARC took the initiative, being the oldest society in this field in Europe. In 1997 the second conference took place in Manchester; it was organized by the AISB. The next will be in Italy, Siena, 27-30 October 1999, see the page on Workshop. These conferences are still small (some 50 participants). Concluding one may say that there is a gradual progress towards "European Cognitive Science" (ECS). But there are problems of national interests, of language, and of the interactions with other disciplines, like psychology, language, the neural sciences, artificial intelligence, connectionism and neural networks, artificial life, pattern recognition. For most of these, already powerful national and international organizations exist. Since cognitive science is concerned with all of these disciplines, it is not easy to find the "demarcation lines". It would be an interesting excercise to study the amount, or degree, of multidisciplinarity in each of these specific disciplines. The hallmark of ECS, and of cognitive science in general, should be the multidisciplinary approach, since that seems to be the only approach that can solve the fundamental problems of mind and cognition. But it has become clear that this is not an "easy affair".

Overall, there are important differences with other continents, like North America. The main difference is probably due to the variation in languages, but there are also important differences due to the history of the development of the different sciences concerned, also within countries where the basic science started, compare the ways psychology developed in Germany, England, France, the United States. It is not possible to give within such a short space an overview of all the aspects of cognitive science in Europe, due to the diversity of languages, and due to the number of main disciplines involved.

G.J. Dalenoort (ESSCS), May 1999

Website and design by Expocom.nl

(We are currently working on our website, please report broken or missing links.)