Setting the tone

James Kirby, AM’07, PhD’10, reflects on studying and teaching linguistics around the world.

For speakers of Vietnamese, Thai, and many other languages, pitch helps determine the semantic meaning of words. Linguist James Kirby, AM’07, PhD’10, a professor at Ludwig Maximilian University of Munich, investigates how these tone languages emerged and spread across Southeast Asia. Tableau spoke with Kirby about his fieldwork, his UChicago experience, and how his research group helps increase access to valuable translation tools.

How did you become interested in linguistics? 

I grew up in Wisconsin and went to the University of Wisconsin–Madison. That’s where I discovered linguistics. At the same time, there were a lot of tone language speakers around. There was a big Hmong community, there was a big Lao community. That’s the thing that drew me toward phonetics and the study of sound structure. But I suspect there’s also the fact that I have a background in music, and so the fact that there were these musical languages was always quite fascinating. 

About 10 years ago, I was able to start bringing those two things together. I had a couple of projects where we were looking at how lyrics in a tone language can be understood. If you have all these differences in meaning that are signaled by where the pitch goes, and now you have to set words to a musical line, how do those things interact? It turns out different languages negotiate that challenge in different ways.

What makes tone languages special?

The use of pitch in language is universal. So, if I say, “Ellen’s a doctor,” and I say it with a rising intonation or a falling intonation, I can make it a statement or a question. But that’s very different from a language like Thai, where these differences can totally change not just interpretation or nuance but also the meanings of the words. 

Your research includes looking at how tone languages evolved and spread.

I had a large grant that ran for about six or seven years, where we were looking at the origin of tones in different languages. You have tone languages all over Africa, but there’s no hope, as far as we know, of reconstructing most of these languages before they had tone. Tone has just been there for as long as anybody can discern. But in Southeast Asia, we know that’s not the case, so it’s a nice little natural laboratory for trying to look more closely at this problem. 

As recently as a couple thousand years ago, languages like Thai and Chinese didn’t have tones. We can compare those languages with closely related languages that haven’t developed tones, and we can say, “Oh, in this language, there’s an h at the end of the word, and in this other language, there’s a falling tone.” You see these correspondences over and over. These are alternate ways of pronouncing words, and one pronunciation variant ended up having an effect on the pitch. Some kid, at some point, must have decided that this difference in pitch was really the thing that was signaling the difference in meaning, and they internalized this pitch difference. We can describe that process, but actually demonstrating how it works, and under what circumstances, is less clear. 

There’s been a long-standing debate about the role of contact between all these different language families. Is tone something like a virus that spreads easily between languages? It’s certainly the case that there was this wave of tonogenesis that spread over that region at roughly the same time. 

One idea is that lots of languages were in contact and were sharing vocabulary. And so if you end up with a lot of the same words, and those words have seeds of tone lying dormant within them, it’s possible that then everything could kick off at more or less the same time. It’s also the case that most people [in the region] are multilingual to some degree. You go to this market and you speak to them in Vietnamese, and you go to that market and speak to them in Hmong. And maybe your wife is from Laos. Maybe if one of your languages is more tonal, it becomes easier to somehow shift those features in between your internal language grammars.

What is the value of fieldwork in your research?  

It’s important to see how languages are used in their natural context—even if having me and a team of researchers with a bunch of recording devices sitting in your living room is not the most natural context. I’ve learned a great deal about the languages from working with speakers directly, in a way that you just can’t learn from reading books.

A couple years ago, we were working with some speakers of Mon, a language that is mostly spoken in Myanmar and Thailand. We were working with four different speakers. For a particular set of words, two of them were producing the vowels in a noticeably different way. And they said, “Well, really, we should be saying it that way, but in my village, we say it this way.” So, then we had the opportunity to ask some more questions, and we realized there’s this whole tranche of words that in these particular villages are pronounced in a totally different way. There must have been some sound change that occurred at some point that affected this very specific class of words. And that’s not something we were ever going to learn from looking at the dictionaries. We really needed to be there and have our ears open. 

Where are you visiting next? 

I’m getting ready to take a field trip to Vietnam to work with speakers of a very small language called Kha Phong, spoken on the Laotian border, to try and learn more about their language and where their tone system came from. It’s going to be really exciting.

There’s going to be a couple of Vietnamese researchers, me, and a colleague from Quebec. We’ve decided that it’s critically important to have people from different language backgrounds, because what you hear is dependent on what you are expecting to hear. The Vietnamese researchers tend to hear that everything has a tone, even if it doesn’t. French researchers will have a French vowel system, so they hear certain vowel qualities in particular ways. We can hopefully triangulate to decide what’s really significant here versus what is imposed on my perception by my language background.

How do you build trust with the communities you study?

Building trust is specific to the parts of the world you’re working in. I know researchers who do phonetics work on First Nations languages spoken in the Pacific Northwest, where language is regarded as something that’s owned by the community. They may be less willing to share their language with you for deeply held cultural reasons. In Southeast Asia, that’s much less the case, but there’s this real danger of parachuting in, taking something, and then disappearing again. We are trying to develop long-standing relationships with these communities. All the communities I’ve worked with have been quite enthusiastic that there have been people coming from outside who are so interested in how they say things. I think I’ve had a very easy time of it, in that respect. 

What do you focus on as a professor back in Munich?

A lot of my job is mentoring and outreach and trying to run a research group where young people can come and pursue projects with the minimum of external demands on their time. That means making sure there’s funding and resources so people aren’t worrying about those things—they can go to this conference or go meet with people or bring in guests. 

One thing that I work on a lot, which is becoming increasingly important, is the interface between speech technology and phonetics. The translation apps on your phone work much better than they used to. But a lot of those improvements have favored what we might call high-resource languages, like certain accents of English or certain varieties of Chinese. But if you’re speaking a minority language of Oaxaca, you’re not going to have access to the same technologies. How can we use phonetics knowledge to improve language technology for all people? 

And then how can we make use of these language technologies to do day-to-day work in phonetics? This conversation, for example, you’ll probably feed to a program that will produce a pretty good transcript for you. For phonetics, we not only want that transcript, but then we want a time-aligned segmentation of that transcript into individual sounds. Traditionally, that’s been a very laborious process that requires a lot of expert knowledge. If we could outsource those tasks, that would free up a lot of people’s time to do a lot of other things. 

What is different about teaching in the German system?

There are no tuition fees. I don’t know that it’s strictly as a result of this, but there’s much more of a feeling that it’s up to you as the student to take your education into your own hands. In the UK, there was a feeling that students are paying a lot of money for these degrees, and so there’s a sense of students as customers, in a way. And there’s some corporate language that starts to get used. I really appreciate the fact that students here are extremely independent.

What did you take from your experience at UChicago? 

I was working primarily with Alan Yu and John Goldsmith [Edward Carson Waller Distinguished Service Professor Emeritus in Linguistics and Computer Science]. John Goldsmith is a very famous name in linguistics. One thing I found incredibly supportive was that when I showed interest in what he was doing, he said, “You should come to our lab meetings.” I was a first-year grad student who didn’t know anything, I was in way over my head, and I was still welcome. I think just being permitted to be in that environment gave me confidence to pursue ideas. That, for me, was really important.

When I got my first professional job as a lecturer in Edinburgh, I realized students want to come and have conversations with you. And you have about 15,000 other things that you are expected to be doing. I had this moment of realizing, wow, my advisers spent an incredible amount of time interacting with me. 

They were also very good about facilitating social interactions, so that there was a sense of having a cohort. There were opportunities to teach very early and to build experience and confidence teaching large courses. And in that kind of interdisciplinary environment, I could move between computer science and psychology and linguistics and there was no one telling you, “Don’t talk to these people over there.” Everyone was very encouraging. It was a very intellectually fertile time.