Podcast episode 47: Geoff Pullum on Geoff Pullum

In this interview, we talk to Geoff Pullum about his career, his contributions to linguistics, and how he sees the future of the field.

Download | Spotify | Apple Podcasts | YouTube

References for Episode 47

Huddleston, Rodney, and Geoffrey Pullum K., eds. 2002. Cambridge Grammar of the English Language. Cambridge: Cambridge University Press.

Pullum, Geoffrey K. 2018. Linguistics: Why it matters. Cambridge: Polity.

Geno Washington and the Ram Jam Band – Que Sera, Sera (1967) (YouTube)

Transcript by Luca Dinu

JMc: Hi, I’m James McElvenny, and you’re listening to the History and Philosophy of the Language Sciences Podcast, online at hiphilangsci.net. [00:16] There you can find links and references to all the literature we discuss. [00:20] Today we’re embarking on our series of oral history interviews with living legends in linguistics, and our first living legend is Geoff Pullum, who we’re meeting here in Edinburgh. [00:33] So, Geoff, could you please briefly outline your career, so your transition from professional musician, which was your first job, as I understand it, to uni student and then academic, and then your move from the UK to America and then back to the UK? [00:53]

GP: The whole of that? [00:54]

JMc: The whole of that, yeah. [00:56]

GP: Well, first, I should make it clear that as a musician, as a piano player in a rock and roll band initially, who later moved on to Hammond organ, [01:08] I was in that job because it was really the only way I could sink below the jobs I had had up to that point. [01:18] I was a high school dropout and had no ideas at all of what I was going to do, [01:25] and playing the piano in a rock and roll band doing German residencies and German nightclubs and American air bases seemed better than sweeping floors. [01:38]

JMc: OK. Which parts of Germany were you in? [01:40]

GP: The Rhineland, mainly. [01:42]

JMc: OK. [01:43]

GP: For five years, I made my living as a professional musician, and Pete Gage and I tried to form a band that was really inspired by the performances of James Brown, a great Black American showman, band leader and so on. [02:04] But there were a lot of things we didn’t have, like a singer with that kind of vocal range and this kind of business acumen and so on. [02:14] Building a band means being sophisticated about money, technology, and networking, knowing the right people and how to deal with outright crooks and so on. [02:32] At the end of five years, I was beginning to find it a terrible way to earn a living. [02:39] The tedium is hard to get across. [02:44] You travel all day, maybe five, six hours sitting in a van up motorways that you’ve seen before. [02:52] You spend two hours hauling equipment in and setting stuff up. [02:56] You then spend one hour maybe playing the same songs you’ve played before, and then more waiting and hauling all the stuff out. [03:05] Then you sleep in a cheap hotel and get back in the van and do it again. [03:11] It’s a horrible life. [03:13] I wanted glamour and excitement. [03:15]

JMc: I mean, it doesn’t sound that much different from being a university lecturer repeating the same material [laughs] [03:20]

GP: Oh, it’s very, very different. [03:21]

JMc: OK. [03:22]

GP: And I found being a Professor of Linguistics, which I was basically within about 10 years after I quit the music business, vastly better than being a rock musician, so much better in every single way. [03:36]

JMc: But your band was relatively successful, wasn’t it? [03:39] I’ve seen a clip even from German TV of you guys performing. [03:43]

GP: As Geno Washington & the Ram Jam Band… [03:46]

JMc: Yeah, exactly. [03:47]

GP: …it was moderately successful. [03:50] That’s not quite good enough to make a lifetime career. [03:56] We were contemporary with the Rolling Stones, and I really admired them for having the tenacity, the staying power and the business acumen to just turn it into a multi-million-dollar business. [04:10] We didn’t have that. [04:12] The Ram Jam Band had broken up within three years of being founded, leaving me with absolutely nothing because I had no high school education to speak of. [04:22] I had been failing all my exams at the age of 16. [04:27] I had to start from scratch, and that was a struggle. [04:31] I would not advise anyone to follow this path. [04:38] By dint of sixth form college courses, as they’re called in Britain, I scraped together enough qualifications that I was minimally qualified to enter a university and applied to six of them, [04:53] and just one, the University of York, took a chance on me, mainly because the head of department happened to know my academic referee, a teacher called Charles Duchesne, who really is someone I owe a lot to. [05:11] He happened to know him personally and called him on the phone, and Duchesne apparently said, “Well, yes, he’s got a chequered background, and… But he is smart, take a chance.” [05:23] Robert Le Page did that and allowed me into the University of York’s Department of Language, [05:32] partly through misunderstandings, but the interview went well and I was in. [05:39] And I discovered that linguistics was fascinating and I started working at it all the time. [05:46] How I knew to apply for linguistics departments was another accident. [05:51] In the 1960s, there were no departments in Britain that gave an undergraduate major course in linguistics, [06:01] so back in 1963, ’64, ’65, I couldn’t have gone to university and studied linguistics, [06:12] but by 1968,it was possible at just two places, and one of them had admitted me, despite my dreadful academic record. [06:24] Once I was working for the first time in my life on something academic that I found rewarding and interesting, I just sort of took off. [06:36] The way I look at it is, you don’t really need to be some born genius to achieve reasonable success. [06:47] If you work all the time on it, put all your energy into it, because of your interest, you’ll do pretty well. [06:54] So I did pretty well, and ignoring the advice of people who told me, “Well, you can’t possibly continue with it. There are no jobs.” [07:03] I applied to do a PhD at Cambridge, and within one year at Cambridge, though it was a terrific year, a formative year, when I met most of the linguists of my generation that I ever knew [07:17] — it was a marvellous coming together of numerous people like Gerald Gazdar and Ewan Klein and Greville Corbett, Andrew Radford; [07:27] all sorts of people happened to be there in Cambridge, mainly because Ed Keenan had an appointment at the Research Centre at King’s College and along with Bernard Comrie was running a seminar on universal grammar, which we all attended, [07:45] so it was a terrific year at Cambridge. [07:49] But by the end of it, I had actually been offered a lectureship at University College London after a brief interview visit and giving a sample job talk. [08:00] And that paid more than being a graduate student. You didn’t need a PhD to be a lecturer in those days. [08:07]

JMc: And was the field expanding as well, like that there were actual jobs? [laughs] [08:13]

GP: Well, there was one for me, and I always tell graduate students who are worried about the job market is, [08:20] yeah, there may be hundreds of students on the market looking for academic jobs. [08:26] You don’t have to worry about them. [08:28] You’re going to be up against maybe three or four people. [08:31] You just have to give a better talk than they did on that day and you’re in. [08:39] You’ve got to have enough of a record that you can get placed on some shortlist, but in the end it narrows quite a lot. [08:46] I was offered this job at University College London by Neil Smith and his close friend Deidre Wilson and Dick Hudson, and those three people became my colleagues as I began life as a lecturer in 1974. [09:03] I didn’t particularly like living in London. [09:06] By this time, having seen all of England when I was a rock musician touring, I didn’t really like England, and there were all sorts of things happening that led me to like it even less. [09:19] The gathering storms of first trade union destruction of various industries and then the Conservative Party’s attempt to destroy the trade unions and so on, it was a usual time of terrible political conflict. [09:35] But by the end of the 1970s, it began to look like Margaret Thatcher was going to storm into power at the next election, as indeed she did, and she stormed in on a wave of racism. [09:49] The vote share of the National Front, the extreme-right party of the day, collapsed to almost nothing when Thatcher was elected because the racists had what they needed. [10:01] Now, I was married to a Black Jamaican I’d met in a club while I was a musician and we had a son, [10:08] and my son was six years old, [10:11] but as he played on the floor with a toy car one Sunday morning as his mother and I listened to a TV interview with Margaret Thatcher talking vaguely about repatriation — voluntary, of course — [10:29] my son looked up and said, “Are they going to send mummy back to Jamaica?” [10:36] He could see what she was saying was, “We’re going to get rid of the coloureds.” [10:42] I thought, this is so horrible. Somehow I’ve got to escape from this culture, from this country. [10:51] And one day, yet another pure piece of luck here, one day the phone rang and Fritz Newmeyer, well-known linguist already, who I had met in Edinburgh at a linguistics association conference and kept in friendly touch with ever since, [11:11] was on the phone from Seattle, Washington, asking if I would like the idea of being a visiting professor at the University of Washington for six months. [11:22] I said, “I’ll take it!” [11:25] You’re meant to negotiate. [11:26] I just said, “Fritz, yes, I’ll come. Get me out of here.” [11:31] And by 1980 I was a College Visiting Professor at the University of Washington. [11:38] I had published quite a bit, and that’s why I was known well enough. [11:42] During that year, a job came up at the University of California’s Santa Cruz campus, where George Hankamer was beginning to form a linguistics department from almost nothing, and he advertised a post for a syntactician. [12:01] His idea was, form a cluster of people who can talk to each other and really have a specialism. [12:10] Don’t just hire one of every kind of linguist there is at first. [12:14] Build a department on some solid research and collaboration. [12:18] So although he was a syntactician, he wanted to hire another one, and he knew me already from my work and from correspondence, [12:28] so I interviewed at Santa Cruz and got the job. [12:32] So by 1981, in the summer of 1981, I was an Assistant… Associate Professor at UC Santa Cruz because my publications were now sufficient that they had given me a tenured position from the start. [12:48] Another piece of fantastic luck. [12:51] So, I think, so far the bottom line of all this is, I had numerous pieces of extraordinary luck, and I wouldn’t advise anybody to imagine that things I did were a good idea or a way that you could get into academic life. [13:13] It was just totally wacky and strange and coincidental, and I’ve been very, very lucky. [13:22] I’ve taken you from being a high school dropout in South London… [13:25]

JMc: Yeah. [13:26]

GP: To being an Associate Professor at UC Santa Cruz through a path that nobody else could follow, and I wouldn’t recommend it… [13:34]

JMc: Yeah. [13:35]

GP: And it shouldn’t really have happened, but it did. [13:38]

JMc: Yeah. [13:39]

GP: From then on, Santa Cruz turned out to be a wonderful place to build a career. [13:44] Hankamer’s project to build a new Department of Linguistics succeeded. [13:50] It was not like Pete Gage’s attempt to build the Ram Jam Band. [13:54] This became really successful. [13:56] By the time we had a PhD program and were producing graduate students with PhDs in 1991 onward, we were already ranked in the top 10 in at least one survey of linguistics departments in the USA. [14:15] UMass Amherst tended to be right at the top. [14:19] MIT was always up there. [14:22] UCLA was great, but for Santa Cruz to have got into the top 10 was just quite an amazing result. [14:29] And almost every year, another hire was made, [14:34] so one by one, we were joined by Judith Aissen, and Bill Ladusaw, and Jim Mccloskey, and Sandra Chung, and Junko Ito, Armin Mester, Jaye Padgett and so on, Donka Farkas. [14:51] All of these people had real research careers, but also a real enthusiasm for undergraduate teaching was what we had started with. [15:01] So the most unusual thing about Santa Cruz was that it was founded on enthusiastic and rigorous undergraduate teaching and built a PhD program on the top of that later. [15:15] That’s the exact opposite of what happened at places like UC San Diego, [15:21] where it started with research institutes, began with some graduate programs to give the researchers someone to teach, [15:31] and gradually built undergraduate programs below that, as it were, and turned into a fine campus, much bigger than Santa Cruz and much more successful in research. [15:44] The difficulty at Santa Cruz was trying to keep research alive on a small campus that was dominated by undergraduates and just couldn’t bring in the biggest grants. [15:58] There was no way little Santa Cruz could get a training centre like UCLA could get, or an institute for theoretical physics like Santa Barbara could get, or Nobel Prizes like Berkeley could get. [16:13]

JMc: But there are no Nobel Prizes in linguistics. [16:15]

GP: That’s right, and once Joan Maling had suggested the idea to me that I should start writing opinion columns about linguistics, [16:26] I think she suggested that in 1987, I wrote a Topic Comment column for that series on how there are no Nobel Prizes in linguistics, no trips to Stockholm for us, [16:45] but meanwhile in the other big subjects — where Santa Cruz often did reasonably well, considering how small it was — things were very different. [16:54] At Berkeley, more Nobel Prizes were awarded to the faculty than the Soviet Union achieved in its entire history, and that was just one campus of what is now a 10-campus system, [17:09] so the high-ups at the University of California used to point out that the full 10-campus University of California is unquestionably the greatest public university the world has ever seen. [17:25] And I had had the good luck to just happen to be at the smallest but also the prettiest campus of the world’s greatest public university. [17:36] It was like there wasn’t much chance of a way up from there. [17:40] I had been so fortunate that surely nothing else was going to happen. [17:45] I would just eventually die there and my ashes would be scattered on the Monterey Bay. [17:50] That’s what I thought until 2006 when, for various complicated reasons, I decided to just put in an application for a couple of chairs in Britain to see how it went. [18:06] As a result of that, I was interviewed and introduced to the University of Edinburgh and found that my then-wife, the philosopher Barbara Scholz, [18:19] was completely correct when she described Edinburgh as the most intellectually lively place she’d ever been in. [18:28] It was fantastic. The visit was so great. [18:31] When I was offered the professorship of general linguistics at Edinburgh, [18:36] I finally realised I actually am going to leave Santa Cruz, and I did, in 2007. [18:43] I moved from Santa Cruz to Edinburgh. [18:46] People say all the time, they said, “Oh, but the weather.” [18:50] Now, it’s true. [18:51] Santa Cruz has glorious climate. [18:54] The weather, to put it simply, is approximately the same as you get in heaven. [18:59] Warm sunshine, cool air from the Pacific Ocean. [19:03] It’s gorgeous. [19:05]

JMc: But you haven’t been to heaven yet, have you? [19:07] And do you think you’ll ever get there? [laughs] [19:09]

GP: I’m quite confident that it has the weather that… [19:12]

JMc: Yeah.

GP: … Santa Cruz has. [19:14]

JMc: Yeah.

GP: It just seems obvious. [19:17] Now, what they don’t tell you in the brochures about Santa Cruz is that it can be different at some times of the year. [19:25] When you see the January storms and it’s cold and wet and high winds are blowing in atmospheric rivers off the Pacific Ocean, it can get pretty awful. [19:36] And in Edinburgh that happens a lot right through the winter. [19:41] But what I found after arriving in Edinburgh was, it simply didn’t matter. [19:46] I wrapped a warmer coat around me and laughed at the weather. [19:51] I thought, yes, the weather is trying to make me go away. [19:56] But I’m not going to. [19:58] This is the best. [20:00] Because Edinburgh, I think, has the most remarkable collection of language scientists I know of in any university in the world. [20:10] I’ve never seen anything like it. [20:11] It’s hard to count because of the huge numbers of postdocs working in the Informatics Forum next door and definition of psycholinguistics and so on, [20:22] but I think it’s in the region of 60 full-time researchers who actually got some profile in the subject between the four departments that are relevant: linguistics, psychology, philosophy, and informatics. [20:40] It’s extraordinary, and there’s just always stuff going on, as you know. [20:45] So I actually found somewhere to go from Santa Cruz that was a promotion. [20:54] Once again, quite a bit of luck was involved. [20:57] By 2007, I think I was best known for the contributions I’d made between 1995 and 2002 to the making of the Cambridge Grammar of the English Language, which is the biggest thing in my research career, for sure. [21:14] Rodney Huddleston invited me in when he found that his big plan for such a grammar was foundering. [21:23] The team he had originally put together in Australia were just not coming up with the stuff. [21:29] It was all falling apart, [21:30] and Cambridge University Press said, “You’ve got to do something,” and suggested me as a possible collaborator, [21:38] so in 1996, for good or ill, and I didn’t know if it was going to work out at that time, I started spending every northern summer in Australia working with Rodney Huddleston on the Cambridge Grammar. [21:54] I would fly out round about the end of June to arrive in Queensland on the coldest day of its year. [22:01]

JMc: Which must have been like 30 degrees, right? [22:03]

GP: It can get freezing cold…

JMc: Yes. [22:05]

GP: …and they haven’t built the houses for dealing with that. [22:08] So by the… I did that for five years straight. [22:12] By the time I was 55, I had seen 60 winters, which worried me a bit because I thought, [22:21] I wonder if they count down to your death in winters or in actual calendar years. [22:27] But despite the long, hard days of 12-hour days of working from dark dawn to dark dusk in the cold and fog of Queensland, Australia with Rodney Huddleston, who was a ruthless taskmaster, but also a very good friend, [22:48] we did get the Cambridge Grammar together, and Rodney reckons it couldn’t have been done without the drive and help with connection to theoretical linguistics that I provided. [23:04] He was certainly the major force, though, and I always refer to the Cambridge Grammar as Rodney Huddleston’s greatest achievement. [23:12] I was a kind of first lieutenant helping him out in various ways, but it was his vision and his hard work that really got it through. [23:23] It’s an extraordinary work of scholarship, the biggest thing I’ve ever contributed to. [23:31] I was proud to have my name on it, [23:33] and I think the University of Edinburgh was canny enough to see that if they hired me within five years of its publication, they’d have their name on it, sort of. [laughs] [23:48] So it was because of that visibility that I got hired at Edinburgh, [23:56] and that was probably the best thing that ever happened to me, where going to Santa Cruz was the second best. [24:06]

JMc: One last thing that you might want to talk about is, do you think that linguistics, which is the field that you have dedicated so much of your working life and mental energy to, [24:19] do you think that linguistics is anything more than just a sort of intellectual pastime? [24:25] Like, it’s definitely very interesting, but do you think it actually matters? [24:29] I mean, I think you even wrote a book once, didn’t you? [24:31] Linguistics: Why It Matters. [24:32]

GP: That’s another interesting story. [24:35] The story of the book Linguistics: Why It Matters. [24:39] I was asked by Polity Press of Cambridge, England, to do that book. [24:46] They told me how long it should be, what it should be called, what style it should be in. [24:51] Everything about the book was fixed. [24:54] And I thought, I like that idea. [24:56] I like writing, and I like writing to order on a particular topic with them in control. [25:03] I can do that. [25:04] It’s a bit like being a session musician, which I did to a very small extent in the 1960s. [25:10] It’s somebody else’s idea and project and so on, but you just do your bit and contribute to it. [25:16] So I contributed to their Why It Matters series and wrote what I could about why linguistics matters to anything. [25:27] And I really hope it does, [25:30] but I wouldn’t want to give the impression that I’m completely confident of that. [25:36] My own work in linguistics… And I was worried that you might ask me to talk about the coherence of my overall oeuvre, the thread that links everything I’ve ever done, and of course, there isn’t one. [25:51] I have been… It would be reasonable to criticize me for having been too miscellaneous in the things I did, [25:58] because I liked all of it. [25:59] I wasn’t one of these syntacticians who couldn’t bear phonology. [26:03] I thought I’d be a phonologist at one point. [26:05] I liked phonetics a lot. [26:07] I taught practical phonetics a dozen times at Santa Cruz. [26:12] I did all kinds of things because they seemed appealing to me at the time, [26:17] and now I look back on 50 years of publishing in linguistics and teaching it to large undergraduate audiences a lot of the time, [26:26] and I worry that little has emerged from it that truly helps to feed the chickens, as they say. [26:37] The obvious thing that was supposed to come out of it, as I always believed, was natural language processing, [26:48] that we would finally get big-scale grammars formalized accurately and parsers that really worked and get them tuned up so they worked fast, [27:00] and translators that, from a full structure of a sentence, could work out what its literal meaning must be, [27:10] and artificial intelligence handling the business of the pragmatics, [27:16] and we would have proper natural language processing. [27:19] And what did we get? [27:21] We got large language models: [27:23] huge, expensive algorithms that have the knack of constructing strings that are the sort of strings it might seem you would like, given what your prompt was. [27:40] It’s hard to explain this, but this is my best shot at it. [27:44] Given a prompt which has all sorts of clues as to what you’re interested in, their task is to build a string of letters, one by one, that is of the sort that it seems, from their training data, you might like. [28:00] Truth has nothing to do with it. [28:04] Meaning has nothing to do with it. [28:06] There is no parsing. [28:08] The fact that the grammar produced by large language models like ChatGPT comes out so smoothly and surprisingly without errors is an extraordinary discovery, [28:24] and I believe Steven Piantadosi is absolutely right when he points out it really means that the argument from poverty of the stimulus is completely unsound in the first place. [28:37] It isn’t true that you cannot learn most of English grammar down to fine detail from mere exposure to raw data. [28:46] You obviously can. [28:48] ChatGPT did it. [28:50] But there’s no linguistics in it. [28:53] So the one area that I would have thought, — and it’s Chapter Five of my book, Linguistics: Why It Matters, [29:00] which I now want to sort of recall, bring in all of the copies and burn them, [29:05] because my Chapter Five is now nonsense. [29:09] You read it. [29:11] Well, let’s just say it shows every sign of the truth, which is that it was written mainly in December 2017, [29:19] just before ChatGPT was released. [29:22] Five years after that, it looks ridiculous. [29:24] I’ve got to rewrite Chapter Five for a second edition of that book and talk about LLMs, [29:32] and it’s not going to be pretty, because they are a menace. [29:37] So with the most important thing that could have provided a proper application for linguistics suddenly gone, at least for a while, because people have formed the insane idea that LLMs will do it just as well and so we can use them instead. [29:56] It’s a terrible idea, but that’s what people think. [30:00]

JMc: So you think it’s an illusion, the output that you’re getting from LLMs at the moment? [30:04]

GP: Oh, yeah, completely. [30:06] People are deluding themselves. [30:07] It’s the ELIZA phenomenon of people thinking, because strings of the sort they like to see are coming out, that there must be a mind back there that is producing them. [30:18]

JMC: But what do you think will have to happen for the spell to be broken? [30:22]

GP: Complete collapse of the AI bubble, [30:27] the whole industry with its extraordinary hundreds of billions of dollars of investment. [30:35] More than that, I think. [30:36] They’re talking in trillions. [30:39] It will all have to collapse in an embarrassing shambles, and we will have to pick up the pieces. [30:48] All sorts of things about a careful study of linguistics are still important, [30:54] like a basic understanding of sociolinguistics for teachers, [30:59] so that they don’t perpetrate the nonsense about if you speak that way, it means you’re stupid. You should speak like this instead. [31:10] Foreign language learning is informed by linguistics to a considerable extent. [31:17] That’s going to continue to be important. [31:20] There are connections to all sorts of other subjects. [31:24] But for me, the idea of natural language processing of the real kind was going to be the thing that made linguistics fundamentally important, [31:38] and for now, that has just been wiped away. [31:43] Departments of computational linguistics have turned into departments of large language models. [31:49] I don’t like to think that I spent 50 years doing entertaining intellectual things that in the end didn’t amount to much of a contribution to society. [32:03] I would sort of hope it was more than that, but the generations that come after us will have to decide on that. [32:14]

JMc: [laughs] OK. [32:16]

GP: I expect to be judged after my death and preferably not before. [32:21]

JMc: OK. [laughs] [32:22] Well, with that thought, thank you very much for the interview. [32:26]

GP: Been nice to be with you, James. [32:28]