Ted Briscoe interviews Gerald Gazdar
Education Teaching Pragmatics Generalized Phrase Structure Grammar
Natural Language Processing Lexical Knowledge Representation Linguistics Conclusion
Left Up Right


EJB It is said that 80% of departments of linguistics still teach whatever the current flavour of Chomskyan syntax is, but that the other 20% are teaching something else. GPSG probably played a part in making that other 20% do something else. Do you think that is true, or do you think that there will always be people teaching Categorial Grammar or whatever?

GG  Well, I guess there will always be departments that don't teach Chomksyan syntax. But I am not well-informed since I've had very little contact with linguistics teaching practices, for many years. I haven't been an external examiner for linguistics, for example. And I haven't visited linguistics departments much. If I have given talks, they have tended to be in computer science departments. I have lost touch with the field - certainly from a pedagogical perspective. However, my take on the field is a cynical one.

EJB Earlier, you described your move from a linguistics post to a computing post very much in terms of local politics at Sussex. But was there a wider desire to distance yourself from linguistics as a discipline?

GG  Absolutely. I had given up on linguistics by 1985.

EJB Because?

GG  I had come to the conclusion that the only way to get ideas across in syntax was to engage in a continuous marketing exercise. Thus Joan Bresnan had spent many years doing relentless marketing of LFG. But if I'd wanted to do that, I'd have gone into sales or something, not become an academic.

EJB Been an advertising executive.

GG  Yes. I mean my view generally is: ``OK, here are some ideas''. I'm willing to make some effort to get them across because otherwise nobody will even know I've had them.

EJB The sum of it being, give a talk at a conference, give some seminars, write a paper in a journal and that should be enough.

GG  Yes. Well, if it isn't, then fine, do something else. As you pointed out earlier, I did do quite a bit of marketing at the beginning of the GPSG period. But I didn't enjoy it and I certainly didn't want to spend a significant fraction of my career doing marketing.

EJB So is computational linguistics better than linguistics in that respect?

GG  Oh yes. Much better, I think. There is remarkably little bad feeling in the field, it seems to me. Okay, there are little pockets - groups who think each others work is rubbish and are not shy about saying so, at least in private. But generally computational linguistics has a much better tempered feel to it, a much less ideological feel. Ideas stand or fall on their own merits. It is not driven by charismatic leaders: it is a of branch of engineering, not a branch of scientology.

EJB And is there an identifiable reason for this? I remember that Roger Needham once told me that, since computational linguistics, and indeed computer science in general, was rather generously funded, there was less reason to have to savage your potential rivals in order to gain resources. I wasn't entirely convinced that that was a complete explanation, but it may have something to do with it.

GG  Yes, it is rather appealing but I don't think I believe it because resources are not very important in linguistics.

EJB Well, jobs are important. I mean, since the 1960s, the field has probably been contracting rather than expanding, unlike computer science and computational linguistics.

GG  Yes, I agree jobs are important.

EJB You referred earlier to Chomsky's assumption that there are only a finite number of possible human languages as ``bizarre''. I was going to ask you why it was so unreasonable to make that move.

GG  Yes. I think it is bizarre to maintain that, potentially, humans could run out of languages and thus need to start reusing them. I also think that it is unscientific to impose a whole lot of stipulations just in order to keep the set of languages finite. Geoff Pullum (1983) wrote a paper that appeared in Linguistic Inquiry which went through all the stipulations. That paper was never replied to. That was one of the last straws for me, actually. Geoff spelled out all the consequences, and he didn't spell it out in an obscure place - he spelled it out on the pages of LI and they might just as well have been blank pages - nobody responded to it. People didn't say ``I subscribe to a framework that requires the number of possible human languages to be finite'' but they did continue to subscribe to such a framework. Nobody said that Geoff's inferences were wrong - the paper was just completely ignored. Why would anyone want to participate in a field which behaves like that?

EJB So bet is a very unusual predicate in English because it takes four, or even possibly five, arguments. But it is bizarre to believe that nobody will ever come up with a predicate in some other language that has five or six arguments, say? I wouldn't be surprised by that, particularly.

GG  But there is a great list of such things - each of which requires a finite upper bound.

EJB So that is another reason for your disillusion with the field. Why do think linguistics is in such a bad state?

GG  I've no real idea. The field has clearly been damaged by the presence of a charismatic leader who has led it badly. But that, by itself, isn't sufficient to explain the situation.

EJB Is it going to get better? Is it getting better now or do you think it's going to stay in the same state, or you just don't have an opinion because you have given up?

GG  I see no reason to expect it to get better. There might be a temporary improvement when death occurs. However, if the field is such that it can be taken over by a charismatic leader and be led by him for over forty years, then why shouldn't that happen again?

EJB So, to somebody coming into the field, offered a job in a linguistics department, what would you say?

GG  Learn how to use a computer and change department.

Next page ..
Copyright © Ted Briscoe & Gerald Gazdar, Wednesday 2 May 2001