Featured Linguist: Colin Phillips

This week, we are pleased to bring you the work of Professor Colin Phillips for our Featured Linguist!

When the LINGUIST editors invited me to write a piece for this year’s 30th anniversary fund drive, I was curious to dig into the LINGUIST archives. LINGUIST started shortly after I started in the field. Like, really shortly. So LINGUIST and me were finding our feet right around the same time, in late 1990.

A quick scan of the first 6 months of LINGUIST turned up this message:

Date: Thur, 04 Apr 91
Subject: Our 1000th Subscriber
If we were not an academic organization, and therefore had some money, we might hand out a prize for this. As it is, all we have to offer is congratulations to Colin Phillips ([email protected]), who is our 1000th subscriber.

I was an exchange student in linguistics at the University of Rochester at the time, taking a year to find some direction in my life. I was definitely finding that direction.

Finding linguistics felt like a kind of destiny for me. In high school in the UK I learned lots of languages and studied mathematics. They didn’t offer psychology, but I would have signed up in a heartbeat if I could. At age 16 my math teacher, who everybody considered a bit nutty, told me there was this guy called Chomsky who I should look into. Living in a farming town, pre-internet, I didn’t know how to follow up on that prescient tip. In college, at Oxford University, I followed a literature path, and focused on medieval German. It was the comparative psychology aspect that interested me the most when learning about the medieval world.

I’m not entirely sure how I got to be in Rochester. There was an exchange scholarship, and I was told that nobody else had applied. I checked a US map in a local bookstore, and it looked like Rochester was just outside New York City. I had no grasp of the scale of the US.

Within two weeks of arriving in Rochester I was smitten. I was surrounded by energetic students and faculty from many different countries who were arguing about language using ideas from multiple fields: linguistics, psychology, computer science, philosophy. They cared deeply about each others’ answers. They were so well integrated that it was hard to tell which students were from which field. It was pure good luck that I had stumbled into this incredibly vibrant pocket of language science. After my years as an isolated medievalist, it was intoxicating. I wanted more. By the time I signed up for LINGUIST List, I was wrapping up a life-changing year in Rochester, and preparing to move to Cambridge, MA to join the graduate program at MIT.

A scan of the first 6 months of LINGUIST turns up some surprises.

There is a post from Andrea Zukowski. At the time we were both graduate students in Rochester. She was in Psychology. We have now been married almost 25 years, and our child is a college freshman at the U of Maryland. Andrea was posting about a conference she was organizing in Rochester. It was the first time the CUNY Sentence Processing Conference was held outside New York City. I wasn’t a psycholinguist at the time, but I have now attended that conference every year since 1995. Not only because it was how I met my wife.

The early issues contain a lot of back-and-forth discussions among well-known linguists, such as a long-running discussion about cognitive linguistics involving George Lakoff, Vicki Fromkin and many others. And since websites were not yet a thing, there are surprisingly many posts asking, “Hey, does anybody have contact information for so-and-so?”

In May 1991 there is a post from John Lawler about the LINGUIST List archive that he was curating, pointing out that the “spectacular growth” of the list over its first few months had led to the archive file reaching 2MB in size.

Innocent times indeed. And so it’s little wonder that fields like psycholinguistics were less accessible to linguists at the time, too.

I went to MIT planning to be a semanticist. Clearly, that didn’t work out. I didn’t intend to become a psycholinguist. Linguistics students at MIT were not doing that kind of work. That was typical of most programs at the time, with a couple of exceptions, such as UMass. If you wanted to do psycholinguistics, you went to a psychology department. By another stroke of luck, my first couple of years in Cambridge coincided with some new grants that aimed to create bridges between fields.

Alec Marantz, David Poeppel and I started exploring MEG research on speech perception in 1993. On our first trip to New York City to collect some data we missed the last flight back to Boston. We were sleeping on the floor of La Guardia airport at 4am.

My dissertation work grew out of another accident. I was embarrassingly delinquent on some class projects. One night I was lying awake worrying, when I realized that I could kill two birds with one stone. A paper on incremental structure building that I owed Ted Gibson for a psycholinguistics class could do double duty as a paper on syntactic constituency that I owed David Pesetsky. The path that I started along that night is one that I have been following ever since, though in ways that I could never have predicted.

When I started my first faculty job, at the University of Delaware in 1997, I was hired as a hybrid syntactician and psycholinguist. I had limited lab experience and was learning on the job. UD was relatively unusual for the time in hiring a psycholinguist into a linguistics department.

In those days a linguist could feel like an outlier in psycholinguistics conferences. I remember feeling particularly despondent after the 1999 CUNY Conference in New York City. The conference was hosted by the CUNY Linguistics Department, but linguists were a small and marginalized group at the conference.

The role of psycholinguistics in linguistics has changed dramatically during my time in the field.

Psycholinguistics used to be something that you would go to the psychology department to do. Nowadays it is an entirely normal part of a linguistics department.

Experimental research used to be exotic in linguistics departments. Now it’s routine. Of course, it’s easier now than it used to be. (Not that today’s psycholinguists are slackers.)

Computational research used to be largely out of reach for linguists. Now it’s more normal. There was a time when I was technically on the cutting edge for a linguist. Nowadays my students leave me in the dust with their computational skills.

The growth of psycholinguistics within linguistics has been remarkable. It is no longer a “Cinderella” field. I no longer feel like an outlier at the conferences that I regularly attend.

Some things have changed faster, such as faculty and student interests, and the range of research methods used. Some things have changed more slowly, such as curricula and publishing practices. Linguistics curricula are gradually moving away from a traditional canon that marginalizes psycholinguistics. Psycholinguistics publishing remains dominated by psychologists. Currently I’m happy to be part of a team that is preparing to launch the new journal Glossa: Psycholinguistics that aims to sit squarely between fields.

A surprise to me has been the rise in popularity of adult psycholinguistics among linguists. When I started in the field language acquisition was taken more seriously by linguists. But the balance of interests shifted at some point. I hope to see a resurgence of interest in learning issues among linguists.

What future directions do I predict or hope for psycholinguistics?

It’s too easy to predict more data using more diverse tools from a wider range of languages. Of course, this is almost certain to happen. Bringing in evidence from more languages is a very good thing. But more data won’t necessarily get us better questions or answers.

There’s no doubt that computational tools will play an increasingly important role in psycholinguistics. I hope that this will be accompanied by more widespread computational thinking. Nowadays we have relatively easy access to complex computational models, and to powerful toolboxes for statistical analyses. There’s a bit of an arms race. But I would place my bets on the most useful insights coming from relatively simple models. We often learn the most from computational models that take us a little beyond what we can do with thought experiments alone. If the model then yields surprises, we can figure out the cause of the surprise. At that point we understand our own hypotheses better, and we can then do better thought experiments.

The COVID-19 pandemic has clearly limited the kinds of data that we can gather as psycholinguists. Right now we have no clear sense of when we will be able to do eye-tracking or EEG experiments again. But the pandemic has also launched a time of remarkable creativity in data collection. As I write this, we are planning a Japanese speech production experiment. The speech data will be collected via the internet, taking just a couple of days, while we are mostly isolated in Maryland. This was unimaginable twenty years ago. The new possibilities for online data collection could advance cross linguistic research in psycholinguistics more than anything else in recent decades.

I am generally happy with where psycholinguistics is as a field currently. There are many talented young people who combine linguistic expertise with sophisticated experimental and computational skills. They are asking questions about mental linguistic computation that simply weren’t on our radar when I was cutting my teeth.

I feel really fortunate that a series of accidents led to me finding myself in the middle of this thriving field.

Thanks for reading and if you want to donate to the LINGUIST List, you can do so here: https://funddrive.linguistlist.org/donate/

All the best,
– The LL Team

Leave a Reply