Thoughts on the CLMS Program
Most of us somehow drift through undergrad, but if we end up pursuing graduate studies, that is when we really come into what we like and care about. Going from not giving a hoot about education beyond my Bachelor’s degree to traveling halfway across the world for grad school was not how I imagined life would turn out to be. And now that I am in my final quarter of the MS Computational Linguistics (CLMS) program at the University of Washington, writing about my experiences over the past two years seems like a good idea.
I first heard about this program when I was doing my internship at IIIT-Hyderabad, frantically preparing to apply to grad schools, heart in my mouth. That sounded cool, I thought. I could get to write code and tinker with language thingies throughout, without having to take boring CS subjects I didn’t care for. I looked it up and came across Dr. Emily M. Bender’s really topical Twitter account, and started following her out of curiosity. I’d not really followed professors on social media before; to me they were inaccessible beings who operated on a different plane from me. All I knew about professors usually was how they looked in class and how they graded tests!
I started the program immediately on finishing undergrad, amidst the peak of the COVID pandemic. My academic counselor advised me and a few others in the same boat to not defer our admits to 2021, but split the program into two years (the full-time duration is one year); we could do the first year remotely from home, then hopefully travel for the second year to attend in-person. And this is what I did.
Apart from the fact that the program in itself is really great with excellent faculty and classes which actually should and do matter, I think I lucked out in several aspects. I had inadvertently come upon the perfect graduate program for me. Like, I don’t know how I would’ve ended up if I’d instead gone to the University of Edinburgh, but I’m a much better and well-informed person for the time I spent at UW doing CLMS. I needed to set my wheels down in a course that won’t swallow me up and spit me out without much to add to my life other than a degree. I worried that I would drown in irrelevance, as Ian McEwan wrote in Atonement. Reader, the exact opposite happened.
This course is offered by the Linguistics department at UW, not Computer Science and Engineering, and I see in retrospect that this made for a more balanced program. If this had been a CSE offering, I could foresee fundamental linguistics courses like Syntax for CL being swallowed up by more “it” courses based on data science, deep learning etc. Not to discredit those at all. But I have come to see that AI research tends to overpower everything else in the conversation around language technology, and a study of NLP should come with a study of how languages work. (Hint: it’s not just NLTK.) So while I would go grudgingly to study syntax, phonetics and what have you, these linguistics-heavy courses served to deepen my understanding of language. In other words, it is a good idea for linguists to be teaching NLP.
It’s a little group within the bigger Linguistics department, with four faculty members and a Treehouse lab that has some really cool people come and talk about all sorts of things every week. I for one appreciate the small scale and the cohort; I feel that it is easier to connect with faculty and other students better this way, rather than a program with 200+ people and professors who barely recognize you. With this size comes more scrutiny and still more freedom; it is easy to share our issues with the faculty and they are there to help you get through the course. I’ve myself been the recipient of kindnesses on numerous occasions, and will always be grateful for them. (Also, you’ll get to know all the four faculty members over time. They clearly care about their work, and it shows in the quality of classes and research. This is again a benefit of a smaller scale; it is possible to have a conversation with faculty members that isn’t hurried and is actually specific and helpful.)
The classes are quite varied, and we get to pick three electives. I took an Ethics in NLP class that was really mind-bending for me; I was suddenly swarmed with so much research on ethical issues in NLP and language tech and learned way more than I expected. I love that there is something for everyone - whether you are from a core CS background, or from Linguistics, or you are 40+ years old and starting something new, or you’re fresh out of college trying to find some direction (me!). The classes are very convenient with remote online options; there is nothing to be missed out on except the campus experience if you take the course online. The faculty goes out of its way to ensure flexibility and that everyone is able to get what they want from this program to some degree.
As for the downsides, I felt that our options were limited in choosing our electives; we can only pick classes that are pre-approved, and that list is small. I wanted to do some electives in other departments which I could not pick. And if someone’s looking to do this course solely as a professional degree to get a good start in the tech industry, some of the Linguistics-focused courses can seem rather heavy-handed and of little use. All in all, it completely depends on what you want out of this program, and how you plan your courses and electives to achieve that.
I conclude by highly recommending this program if you’re specifically interested in NLP and language technology, not if you’re looking for something more general. If you’re coming from a singular background like CS/math/linguistics, it can be a challenge to get through, but the efforts are incredibly rewarding. As far as I can tell it is fairly easy to find relevant roles in industry afterwards (again, depending on how you use the program for your benefit!), but if you’re again looking to be a more general engineer instead of just NLP, job hunting could be a little challenging. I’ve had a wonderful time here; this program was the right fit for me and I am very excited for what comes ahead.