Jeffrey P. Bigham
@jeffbigham
It’s job hunting season! In particular, it’s job offer season - that wonderful scramble each Spring when academics from fresh PhDs to those looking to the greener grass across the fence are considering what it might be like to exist in different departments.
One of the questions I get asked a lot is what it’s like being the only HCI person in a computer science department. The reason I get asked this is that I was an Assistant Professor in the Computer Science department at the University of Rochester for 4 years from 2009 to 2013. URCS was a traditional CS department, and I was their first HCI hire.
Now, of course, I’ve swung completely in the opposite direction where I’m in a similarly-sized department where everyone does HCI.
People are worried about being the only HCI person in a Computer Science department because there’s a perceived (and sometimes real) risk that HCI work won’t be valued, and if the work isn’t valued, that the person won’t be valued (e.g., via mentorship, opportunities within the department/university, and ultimately via tenure). This is similar to the risk of being the only person in some other part of computer science, e.g., being the only computer vision researcher, but it has the added baggage of HCI people truly not being seen by many Computer Scientists as a real part of their field. Ironically, Artificial Intelligence as a field used to face this same criticism from Computer Scientists, as Computer Scientists used to from Mathematicians (or whoever). Academia is great.
I honestly didn’t really think too much about the fact that I was going to be the only HCI person in a CS department when I decided to join URCS is 2009. There were a lot of reasons, but primarily:
Looking back, I was incredibly fortunate to land at a place that was so friendly to HCI, where I could grow as a professor and HCI researcher.
One thing going for anyone who is the first person in HCI to join a department is that someone fairly influential is rooting for you. Traditional CS departments do not need to hire anyone in HCI. Deciding to hire someone in HCI means not hiring someone to strengthen one of their existing areas. And, so, someone believes enough in you and has enough influence in the department to make that happen. Those same someones are likely to continue supporting you, your career, and your field going forward.
I was very lucky to have a bunch of senior people who supported me at Rochester. Henry Kautz was my department chair, but I count among those supporters at least Michael Scott, James Allen, Sandhya Dwarkadas, Chen Ding, etc. It probably helped a lot that James Allen and his group were (and are) of the breed of AI researchers who built real systems that real people could use to do real things. Honestly, I believed that 20 years earlier I would have been an AI researcher of this flavor.
When I got there, it’s not clear that either the department or I really knew what to make of me :) I had never took or taught an HCI course, but I started one based on some conglomeration of materials from Rob Miller, Carman Neustaedter, and James Landay. To my surprised, my HCI course was incredibly popular -- it was and remains the most popular non-required course.
The main potential negative I hear people talk about is related to how the department will view your work come promotion and tenure. The theory goes, if they aren’t really sure they want to hire you, or aren’t really sure that HCI is a valid research discipline, will they actually decide to tenure you when it comes time? If nobody else does HCI, will they even know how to review your case?
I think this is why the, “Why is HCI Computer Science?”, question during interviews is seen as such a bad sign. If you’re trying to hire an HCI person and don’t have any, please ask this question to a senior person beforehand. Make the interviewees justify their work, not their field.
Underlying this is I think a much more serious concern -- will you have the support and mentorship you need to not only survive but to thrive as an assistant professor. This is much harder to quantify for a variety of reasons. One reason is that I think it really depends a lot on what kind of research you do, what your professional network looks like and how it grows, and how engaged you are across physical boundaries. For me, Henry was a great mentor, although I’ve come to realize I might be a difficult person to really “mentor” … regardless, he was an excellent person to ask questions, was always responsive, and took my side when he didn’t necessarily need to so that I’d feel heard.
I was lucky that overall my department was very supportive, even if they didn’t always get what I was doing or what HCI was, or what an HCI person was supposed to be doing. I had some conflicts with faculty, as tends to happen, but nothing big and nothing we didn’t get over. The only big arguments I remember having that actually centered on HCI were when I advocated for big leaps. For instance, when we were rethinking our curriculum, I suggested that HCI should be a required course. I got a lot of flabbergasted pushback on that. Yet, that pushback is not unlike what I think HCI people would get wherever they are -- certainly, there is no required HCI course for Computer Science students at CMU either, and I’ve had some of the same arguments here. To the credit of URCS, in some important ways HCI became much more core to the curriculum than it had been before.
While it’s easy to focus on what the department is or isn’t giving you to help you succeed as a solo HCI person, I think it’s also important to go into the job with some expectations of what you will need to do to succeed as a solo HCI person.
The first big thing is that you will need to advocate. Battles that might have been won or at least settled long ago will need to be fought. What is the place of HCI at your university or in the curriculum? How does it count for students at different levels (e.g., a lot of places throw it in within an “applications” area)?
Part of advocating is becoming intimately aware of what your community does well, what it values, and where its warts are. CHI papers are apparently very important. Why does everyone get a Best Paper Award? How can we trust subjective feedback? Why couldn’t a startup just build that thing you built as part of your HCI Systems contribution? And, a million other things that I’m forgetting. Hopefully, your university avoided asking you to justify your field during your interview, but you’ll need to advocate for it if you’re the only one in it.
You’ll also need to find and keep connected with mentors. I was lucky to have received my PhD at a time when the University of Washington was becoming a big player in HCI. I got to know some great faculty there (Wobbrock, Fogarty, Landay, etc.), at an internship at IBM (Tessa Lau, Jeff Nichols, etc.) and I also managed to spend some time working with Rob Miller’s group at CSAIL right before I moved to Rochester. In addition to the faculty, I got to know a ton of great PhD students who were either then or would soon go on to be become faculty at different places (e.g., Eytan Adar, Krzysztof Gajos, …, etc.). These experiences served as the basis for my HCI network in the early years, which I tried to stay in touch with as I started my career.
The thing that really helped me build and grow both my network and awareness of HCI was (and is) social media. It may sound silly and/or obvious now, but from my office at the University of Rochester, I could easily connect to 1000s of other HCI people at different stages of their careers. I had a community, and it didn’t (and still largely doesn’t) matter if they were right down the hall or not. So, if you wonder why I am so active on Twitter, this is really where I got my start.
Locally, I grabbed onto the whole of “HCI”, and formed a big inclusive group of everyone I could convince to show up and give me a photo for the web site. ROC HCI looked like a big group, because it was big. I continue to think that every part of computing is ultimately human-computer interaction, and part of that HCI-centric worldview led me to gobble up as much of the intellectual space and awesome people as I could. We even had theorists showing up for a while (obviously, not all of them, haha).
One of the things that Rochester did that was a big positive was that they tried hard to hire another HCI person after they hired me. This is a big deal. It means you have a real HCI “group”, and it means that students interested in HCI might find it less risky to join your university -- it’s always risky to go to a place where there’s only one person in the general area you want to study. We failed for a few years despite trying, and then right as I was leaving managed to hire two amazing people: Philip Guo and Ehsan Hoque. If you happen to be reading this from the other side (someone hoping to bring or grow HCI in your traditional CS dpt), I can say that very rarely does a solo HCI person, who remains the only HCI person, stay around for more than a few years -- despite the fears of HCI people, it’s generally not because they get kicked out.
I think it’s completely reasonable to ask for some resources as part of your startup package to help kickstart your HCI career at your university. Some of these might be for personal things -- maybe travel money to visit places where your HCI mentors are, or maybe even for professional development workshops of some kind. But, maybe even more important are resources for things to help you establish HCI at your new university, i.e., money to bring in HCI seminar speakers, or money for pizza to start an HCI lunch to bring people together (as the myth goes, DUB was, and largely still is, pizza money and a dream).
I loved my time at the University of Rochester, as the only HCI person in a Computer Science department. I think I’d be a very different kind of researcher and approach advocacy for my field quite differently if I hadn’t had this experience. As a technical HCI researcher, who was relatively close to many of the AI/NLP faculty in my department, I didn’t have much trouble recruiting students or getting along. If you’re considering being a solo HCI person, I think it can be wonderful. But, please look for evidence that your new department will support you, be prepared to advocate for yourself, get a sense of whether the department is willing to grow, and see if your department is serious about HCI by getting them to commit real resources ($$) to build up HCI. Starting as a solo HCI person is exciting. Remaining the only HCI person forever would likely be a drag.
I’m more than happy to discuss any and all of this in more depth -- please tweet or email me to set up a time.
Happy decision making!
This page and contents are copyright Jeffrey P. Bigham except where noted.
Blog posts are not intended to be final products, but rather a reflection of current thinking and/or catalysts for discussion, like tweets but longer.