
Podcast episode 51: Martin Haspelmath
Notes de l’épisode
In this interview, Martin Haspelmath talks about how he got started in linguistics, the rise of large-scale areal typology in the 1990s, language description vs language comparison, and the current state of the field.
Photo: Swen Reichhold/Uni LeipzigDownload | Spotify | Apple Podcasts | YouTube
References for Episode 51Dryer, M. S., et al. 2013. The World Atlas of Language Structures Online. München: Max Planck Digital Library. http://wals.info/
Haspelmath, M. 1993. A Grammar of Lezgian. Berlin: Mouton de Gruyter.
Haspelmath, M. 2010. Comparative concepts and descriptive categories in crosslinguistic studies. Language, 86(3), 663-687. http://dx.doi.org/10.1353/lan.2010.0021
Michaelis, S. M., Maurer, P., Haspelmath, M., & Huber, M., eds. 2013. The Atlas of Pidgin and Creole Language Structures Online. Leipzig: MPI für evolutionäre Anthropologie. http://apics-online.info
Transcript by Luca DinuJMc: Hi, I’m James McElvenny, [00:14] and you’re listening to the History and Philosophy of the Language Sciences Podcast, online at hiphilangsci.net. [00:22] There you can find links and references to all the literature we discussed. [00:26] Today we’re in Leipzig with Martin Haspelmath at the Max Planck Institute, [00:32] and we’re going to continue our series of oral history interviews [00:36] by asking Martin about his career in linguistics [00:40] and his contributions to typology, [00:43] and how he sees the past, present, and future of the field. [00:47] So my first question to you is, can you tell us how you got into linguistics? [00:52]
MH: Yes, so thank you for having me here. [00:58] I got interested in languages rather early on, so I was already a linguist later in high school. [01:06] I was interested in Latin and Greek, [01:10] and even in Egyptian hieroglyphs, because there was a museum of Egyptian in our town, [01:17] and I also had some interest in Polish, [01:22] because my mother spent her childhood in Poland as an ethnic German, and so I started studying Polish a little bit, rather early on. [01:32] So yeah, I was interested in languages very early, [01:36] and I was very happy that I had the chance to continue to do linguistics. [01:41]
JMc: So did you speak any Polish at home? [01:43]
MH: No, my mother didn’t speak Polish later on. [01:46] It was only in her childhood, and then the ethnic Germans had to leave Poland, [01:52] but it was an important heritage language for me, because my mother told us a lot about her Polish fatherland, as she said. [02:03] So she came to Germany in 1945 and felt a little bit as an outsider, [02:09] and, yeah, they had lived in Poland for many generations. [02:14] My grandparents even knew Russian, because they had gone to a good Russian school before 1914. [02:20] So Slavic languages are in that sense a family heritage, not that these languages were spoken. [02:27]
JMc: So how did you then transition from this general interest in languages to the actual academic study of linguistics? [02:36]
MH: I studied at the universities of Vienna and Cologne, and first I was very interested in historical linguistics, [02:44] so I was fascinated by the unification that was offered by the study of ancient Indo-European languages in historical-comparative perspective, [02:55] reconstructing sounds and morphological patterns to Proto-Indo-European. [03:03] I had gotten into that even in my later high school years, and I first thought I was going to be an Indo-Europeanist, [03:12] but then things changed. [03:15] My first two years at the University of Vienna, I was doing Indo-European linguistics, [03:21] and it was interesting, but there were very few students, and the professors were very conservative and not so cool somehow. [03:32] And there were two visiting professors from abroad [03:36] who were teaching strange things such as Russian-style syntax — it was Igor Mel’čuk — and Greenbergian universals, typology. [03:50] So I took a course from Edith Moravcsik from the University of Wisconsin, who had worked with Joseph Greenberg in the Stanford Universals Project, [03:59] and so I had known about formal syntax and Indo-European linguistics earlier, [04:06] but I really learned about typology and universals only at the university through Edith Moravcsik, and I was fascinated. [04:15]
JMc: And you made a career out of it. [04:17]
MH: Well, it took a while. [04:20] I went to the University of Cologne, because that was a place where typology was supposedly strong, [04:28] but I was a bit unsure for a while, but then I got a chance to spend two semesters as an exchange student at the University of Buffalo in the United States with Joan Bybee, [04:39] and Joan Bybee had a kind of real typology project, and she involved me in it, [04:45] and so I realized that, yeah, not only I loved it, but there were also some people who thought I could do it well. [04:53] And so Joan Bybee was a very important person who supported me later on, [04:59] and then I had a chance in 1990 to work on a European typology project with Ekkehard König, [05:08] funded by the European Science Foundation, studying European languages in typological perspective. [05:15]
JMc: So what is typology, then? [05:18]
MH: Well, comparing languages that are not necessarily related in ways that potentially lead us to universal generalizations. [05:30] Of course, if we study just European languages, then that’s not universal enough, but I felt that in the 1990s, certainly that led to a lot of new insights that hadn’t been there before. [05:44] So linguistics in earlier decades was mostly focused either on the historical study of particular languages, [05:54] or since the 1960s, in the Chomskyan paradigm, on, yes, speculative ideas about universal grammar and formalization, [06:07] and, yeah, the Chomskyan approach didn’t appeal so much to me. [06:13] And also in the 1970s and ’80s, it was not very comparative yet, [06:18] so I kind of really loved the comparison of European languages that, you know, allowed me to look at patterns in Icelandic, in Albanian, in Maltese, in Estonian, [06:35] and, you know, compare all these little languages and try to see what they have in common, and then maybe later on to broaden that to a global scope. [06:46] But in the 1990s, it was still the European pattern for me. [06:51]
JMc: So what exactly is it that you’re comparing? [06:56]
MH: Well, in the 1990s, we were comparing adverbial clauses, for example, or comparative constructions or indefinite pronouns. [07:06] So we were comparing various morphological and syntactic patterns, regardless of their historical relatedness, [07:16] and often we found that they were similar not because they were inherited, but because they had borrowed patterns from related languages. [07:27] So my dissertation was on indefinite pronouns — you know, things like ‘somebody’, ‘something’, ‘anybody’, ‘anything’. [07:38] You know, English and German are very similar. [07:41] Many words are similar, you know, ‘house’ [Haus] and ‘shoe’ [Schuh], [07:44] and other words are sort of very clearly kind of borrowed, [07:50] like Tisch and table, borrowed from Greek and Latin, and so on. [07:54] All of that was well known, but, you know, why does German have jemand, and etwas, and irgendjemand and irgendetwas? [08:03] So these patterns were not clearly historically related, they were much younger, [08:12] so I was fascinated by this, and often these younger patterns had geographical dimensions, [08:19] so languages spoken in similar areas were often similar in ways that people hadn’t looked at before. [08:28] So we were doing European areal typology, looking at similarities in geographical terms that people hadn’t seen before. [08:37] So that was really a new thing in the 1990s, and I was part of it, and I called it Standard Average European, and it became well known, but it was sort of discovered accidentally. [08:49]
JMc: But, I mean, there was areal typology and the idea of diffusion before the 1990s. [08:55]
MH: Yes, but it was not mainstream. [08:58] There was a little bit of it. [09:00] The diffusion was being studied primarily by dialectologists, so typological areal diffusion was really a new thing. [09:12] Johanna Nichols wrote her book on Linguistic Diversity in Space and Time in 1992, and before Johanna Nichols, Matthew Dryer wrote his famous article on large linguistic areas in 1989, [09:29] and our work on European areal typology in the 1990s was really part of a kind of larger real movement. [09:39] Clearly, there were earlier things. [09:41] There was the idea of an Indian, South Asian linguistic area, there was the idea of Mesoamerican linguistic area, but those were not larger movements. [09:55] They did not immediately spark larger research efforts, whereas in Europe, there were many linguists, [10:02] also related to the funding by the European Science Foundation, [10:07] and, actually, one of the outcomes of our European typology project was that when the funding came to an end in 1995, exactly 30 years ago, [10:18] the Association for Linguistic Typology was founded, coming out of this ESF funding initiative. [10:26]
JMc: So a moment ago, when we were talking about what typology is, about the ultimate aims of the field, you mentioned universals. [10:34] So do you mean sort of universals in grammar, like in language structure? [10:38] Is that the ultimate aim of typology, to discover what the possible range of languages is, [10:44] or is it to look at these mechanisms of diffusion? [10:48]
MH: Well, the universals were certainly a very important goal in Greenberg’s work in the 1960s and 1970s. [10:56] So in 1978, Greenberg had edited four volumes, universals of human language, and that coincided more or less with the Chomskyan interest in universals. [11:08] So in the ’60s and ’70s, certainly there was some universals research, and in the United States, you know, work by Matthew Dryer and Johanna Nichols, and also Joan Bybee, [11:19] and Tom Givon was an important member of that community. [11:24] They were doing sort of empirical universals research, so they had not focused so much on areal typology. [11:34] So I was interested in both of these things, [11:36] but I felt that the areal typology was this new thing that we had discovered in Europe and that we were pushing, [11:43] and this then also led me to ask whether there could be areal typology on a global scale. [11:51] So when I came to Leipzig in 1998, to the Max Planck Institute, I said, “Let’s expand this and have a world atlas of language structures.” [12:02] And I was extremely fortunate that this Max Planck Institute was founded, and there was funding. [12:11] Bernard Comrie was the director, and he thought that was an excellent idea. [12:15] He was part of this Greenbergian American typological movement, and we said, “Let’s study universals from an areal-geographical perspective, and let’s make a world atlas of language structures.” [12:30] And what we discovered with this atlas, I think, was this massive areality worldwide. [12:37] Maybe we didn’t discover new universals. [12:40] Greenberg was actually pretty good at discovering universals in the ’60s. [12:45] And we also, I think, showed that some of the later attempts at finding universals were a bit naive. [12:55] You know, things like the null subject parameter or all kinds of other ideas, [13:00] how word order correlations could be extended, for example, to phonology, you know — many of these did not… actually found confirmation. [13:12] But what we did find was that there was a lot of areality, a lot of geographical structure that needed to be taken into account. [13:21]
JMc: So do you think that areality is like a confounding factor for historical-comparative, for genealogical relationships? [13:29] Would you say that there are two kinds of relationships, sort of historical-comparative on the one side, and then there’s areal, and they’re sort of interfering with each other? Or… [13:39]
MH: Yes, absolutely. [13:40] So there are really three reasons why languages can be similar. [13:45] One, that they’re related in the genealogical sense, that they belong to a language family, [13:52] that they share features because they inherited them from a common ancestor. [13:57] Another one is that they borrowed them, or that they have similar features because other languages in the area have similar features. [14:06] So, you know, like in the Southeast Asian languages, they all have SVO order, subject-verb-object, they’re all kind of isolating, they have tone, and so on. [14:17] So that’s the areal factor. [14:19] And then the third reason for similarities is that they’re just human languages, and human languages just are like that, [14:29] are like that because there’s general universal tendencies. [14:35] And if we want to find the universal tendencies, we first have to take into account the other two potential factors. [14:45] And from the perspective of the universals, yes, we can say they’re confounding factors, [14:50] that we have to somehow, you know, using statistical methods, you know, first eliminate and then find the universal tendencies. [15:02]
JMc: You said there were several massive areas in the world that you’ve discovered through the World Atlas of Language Structures project, [15:09] so how many of them are there? [15:12]
MH: Well, we found that there’s a lot of areality in the sense that many large-scale patterns. [15:21] So earlier, as I had mentioned, people had noted that South Asian languages are rather similar [15:28] regardless of whether they belong to the Indo-European family, or the Dravidian family, or the Munda family. [15:35] They all seem to have a kind of common general pattern. [15:42] People had noticed similarities of Mesoamerican languages. [15:46] People had noticed some similarities in Australian languages and so on. [15:53] So all of these were known, [15:55] but with the World Atlas of Language Structures, published exactly 20 years ago in the summer of 2005, [16:04] we saw that the majority of our world maps showed some areal effects, [16:11] or seemed to show some areal effects — not all of them, but many of them. [16:16] And there were new areal effects, there were some confirmations of earlier patterns, but also a lot of new areal effects. [16:23] And I think we’re still only beginning to understand what these mean, because the problem is that they don’t always form isoglosses. [16:34] Sometimes they do. [16:37] So for the Caucasus, for example, all these languages have ergativity, they have ejectives, they have certain other valency patterns that are striking, [16:54] but then many other things are not shared by the Caucasian languages, [16:58] but they may be shared by languages in the Caucasus and elsewhere, [17:05] or only by some of the languages in the Caucasus and some neighbouring languages. [17:09] So there are many cross-cutting isoglosses, and I think we don’t understand it very well, [17:14] and we have to be careful not to be led to premature generalizations [17:21] by maybe cultural stereotypes. [17:24] Maybe people think, “Oh, South Asia, oh, it’s kind of a cultural area, so we only cherry-pick those features that those languages share.” [17:33] I think that has sometimes happened, [17:35] and maybe even my work on the European linguistic area was, in a way, influenced by European cultural stereotypes, so I’m a bit critical of that now. So… [17:51] I mean, certainly, the idea that Europe is a sort of, you know, common culture, you know, we have the European Union, [18:02] and at the time I thought, “Oh, it’s wonderful, European unification, and now we can show that the European languages also share more than we thought they did,” or so, [18:14] and there was a certain danger of ideological bias in that now, so I’m a bit critical of that now. [18:20]
JMc: Have you noticed sort of subareas within the European languages? [18:24]
MH: Yes, there are subareas, and people have talked about Circum-Baltic similarities, [18:30] people, of course, for a long time had talked about similarities in the Balkan languages, [18:35] but there were also other ideas, for example, similarities, Circum-Mediterranean, right? And similarities in Northern Eurasia in general. [18:50]
JMc: You’ve talked about WALS a lot, about the World Atlas of Language Structures. [18:54] Would you say that that’s your greatest contribution to the field? [18:58]
MH: Well, it’s certainly what I’m best known for, [19:03] but it’s really the Max Planck Institute’s contribution, Bernard Comrie’s contribution, because I had a unique possibility here. [19:12] We had this very generous funding. [19:15] Yeah, I enjoyed bringing all these people together, but the community existed, [19:24] so, you know, basically Greenberg created it in the 1970s with a Stanford project, [19:32] and, yeah, then we collected together all those people that were already doing typology, mostly in Europe. [19:43] Somehow it’s also interesting, historically, that the worldwide cross-linguistic research that was happening in the ’70s and ’80s in the United States [19:54] became more popular in Europe in the 1990s, and so we had people like Leon Stassen, and Anna Siewierska, and Maria Koptjevskaja-Tamm, and Östen Dahl, and so on, [20:07] so typology was already then, in the 1990s, when we started in 1998, [20:12] was a European thing, and I helped bring these people together, [20:16] but also Matthew Dryer and Bernard Comrie. [20:19] David Gil was also very important, and yeah, I had the most time to put it all together also, but it wasn’t really my idea. [20:35] Maybe my biggest intellectual contribution was a more philosophical one. [20:42]
JMc: Yeah, tell us about it. [laughs] [20:44]
MH: Since it’s a podcast not only about history, but also about the philosophy of language sciences, [20:50] I’m happy to talk about that, namely, my interest in comparing languages [20:57] led me to wonder increasingly about how language description and language comparison are in tension with each
Tous les épisodes de History and Philosophy of the Language Sciences