hestenes@sdics.UUCP (07/28/86)
These are my responses to the major issues raised in the replies to my initial posting. Eric Hestenes arpanet: hestenes@nprdc.ARPA other: ucbvax!sdcsvax!sdics!hestenes or hestenes@sdics.ucsd.edu --------------------------------------- 1. ****************** >It seems that lots of people confuse Cognitive Science with >Artificial Intelligence. While CogSci and AI have valuable >cross fertilization going on, I think that students (esp., >undergrads) should be made to understand that they are not >the same discipline.... This is a very important point. Many Artificial Intelligence people are interested in intelligent systems but not in the general properties of "intelligence." This distinction should be reflected in the programs that students take, e.g. a cognitive program should not emphasize AI methods, except to the extent that it pertains to the study of cognition. 2. ****************** >I think that a cognitive science course should >ideally include material from a variety of disciplines. good point. Cognitive Science takes strength in attempting to test its assertions in a variety of domains. 3. *************** > Problems? Well, I suppose the fact that I had to transfer to Chicago in >order to pursue the independent studying that I wanted was a problem, though >not unique to cognitive science. I *would* suggest that cog-scientists >develop interdisciplinary programs such as are seen at UCSD, UMass, and a few >other places (e.g., incorporating neural net study to attract comp-sci/AI >students. I agree. It is hard for students to do cognitive science if there is no program in their area of interest. This problem could be overcome by giving students a strong interdisciplinary program around which to center their studies. The problem with working in other departments is that the requirements of that department can overpower the studies. 4. ******************** >1. Are we perhaps early for cogsci as a field unto itself? which came first, the chicken or the egg? who knows? >2. Should it be studied as a department with offerings from its >constituents, or as a program with a set of courses that its constituents >will recognize, flexibly, even if some are from one or more of the other >departments? This is a tough issue. I believe that there should be a "cognitive science" department ( so to speak ) rather than several departments that practise the "cognitive sciences". Otherwise there will be no job opportunities in cognitive science, only in those separate departments. Funding can be an issue that makes or breaks a so-called field. >3. How do we manage the problem that cogsci can be thought of as the UNION >of its constituent fields, and not just the intersection? (Of course, >some topics are much more relevant than others, but there seems to be >considerable leeway in designing a curriculum -- many diff. objectives >that cogsci could have.) The need for a new field becomes final when the old categories no longer usefully describe what is going on. I think we are at this point now. I think the society of people interested in cognitive science must join together and decide, and either make the field or keep the separate fields. Perhaps a major technical innovation is required before this is possible. One could argue for Cognitive Science as a field as well as for any science ( e.g. computer science ). The appearance of leeway probably arises from the fact that we do not perceive that our methods work across domains (e.g. in both computer science and cognitive anthropology ). >4. Does it have a coherent set of notations, theories, etc., and consistent >methods and models? From my experience with psychology, it seems that >there are many theories about related phenomena that are inconsistent. >What happens when we throw in the other fields? Does it become much >worse? Do the fields force each other to recognize the theories of each >other and search for resolution? The emergence of a set of theories alone is enough to justify the field. In Kuhn's "Scientific Revolutions" the paradigm is king, and the "cognitive" paradigm has certainly emerged. Please note that each field that utilizes cognitive science also tests it with their own methods. While the program must "run" in AI, the experiments must be done in cognitive psychology and the biological constraints must be tested in the neurosciences. The need for an acid test is less important that the need for useful results from the "findings". Or have we learned nothing from the behavioral psychology of the 70's? >5. How large a role is systems theory going to have? This remains to be seen. >6. What are the driving assumptions? Hobbes' "by ratiocination I mean >computation"? What are the tenets or goals or common beliefs? >6.5 Will you let the philosophers play theoretical watchdog? (Actually, the philosophers might be less watchers than the watched.) Many are working on the "driving assumptions", and they cannot be simply stated. A gross ( probably negligent ) statement might be that the people in cognitive science find intelligent processes in their domain useful as areas of investigation. The intelligent or "cognitive" processes in the world seem to be the central focus of cognitive theories. >7. Individual departments and whole fields can develop personalities that >are not related to the subject matter, but which influence the way in which >the subjects are studied, and also what gets studied. How are things turning >out in cogsci? Also, how about the political aspects -- does, for instance, >psychology have any significant kind of dominanace? Questions like that. That, of course, is how the field arose, and perhaps how it will die. Again, Kuhn suggests that the nature of science is to create, destroy, and recreate paradigms. *All* fields suffer from this. The peculiars in Cognitive Science aren't that relevant, except to know that CogSci, like the other fields, does have dominating influences. I suppose that the model of the human might prove to dominate the field. Just suppose, for instance, that we figured out the role of the brain ( exactly, I mean.) Where would this field be? Innovations pertaining to understanding known forms of cognition must obviously play an important role in the field. >8. Will it be experimental? Theoretical? Will it have ITS OWN topics? Hopefully. >Perhaps you could suggest at this conference that more >universities and colleges offer specialized programs in such fields. I >understand that the University of Colorado has a good Cog. Sci. department I will try to say that. I hope to get more feedback from other students on this beforehand, however. *********** end of my replies *************