There are contrasting views on what PhD degrees are for, whether we are training too many of them, or whether we should be preparing them differently for what lies ahead. In the fields that I know better, this is often framed as the question of whether we should be providing the extra skills that are needed for ‘alternative careers’, i.e., not academia. I’m not sure that’s the right way to look at it.
This is a lightly edited version of my remarks at a panel on graduate education at the National Science Foundation headquarters in Arlington, VA on Sept 16th, 2016, which have been gathering digital dust for too long. Plus comments on what we have learned in the intervening 3 years.
1. There’s a common view that there is a tension between preparing PhDs for academic careers and preparing them for the many other careers that PhDs pursue. I think that is wide of the mark.
That’s because academic careers aren’t what they used to be. So many of the skills that we know are valued in diverse careers beyond academia are also really, really valuable in modern academic careers. So there really should not be a conflict.
I think the tension is a bit different. If our graduate programs are preparing people to be professors in the 1960s, then we’re really not serving students well. That’s because jobs in the 1960s are pretty hard to come by these days.
2. People often assume that PhD training is all about preparing to be a professor. But that’s not what we should want. We want PhDs to prepare the intellectual leaders of advanced societies. And there are skills that we want those people to have, whatever their profession.
An advanced degree should be preparing you to do things that require a lot of independence and initiative. This includes teamwork and leadership abilities, lifelong adaptability, and the ability to integrate ideas from different areas. It includes taking calculated risks, because you’re doing things that haven’t been done before. And it includes the ability to be a really good communicator – in the broadest possible sense.
These skills are useful in non-academic careers, but they’re also essential in academia today. Attracting funding requires ever more creativity. Attracting talent to work with you is super competitive. You have to form alliances if you want to go after challenges that extend beyond your own expertise. You have to be adaptable in order to keep up with changes in science. You had better hope that whatever is hot when you write your ground-breaking PhD thesis is no longer hot 20 years later. And hopefully you’re also able to lead institutional change. Of course, a special challenge of academia is that you have to do all this while having almost no real power, because so many of your colleagues are tenured and can simply choose to ignore you.
3. Unlike some of the other panelists, I don’t do research on graduate education. I can’t give you proof of what works. But I have plenty to say about what I think works and why.
My home department is linguistics, and I’m housed in a College of Arts & Humanities. But my broader field is “language science”. We want to understand why it is that when it comes to language learning a typical 6-year old is better than Google or Apple, despite learning from far less training data. We want to know what it is about the child’s brain that makes this possible, and how to make a difference in more adverse conditions. And we want to also make a difference in language technology, bearing in mind that 99.8% of the world’s 7,000 languages don’t have the kinds of resources that Google can throw at languages like English, Chinese, or Hindi.
Solving this problem requires a broad interdisciplinary approach, and so it requires training to match that.
4. We have been working for 15 years now on trying to find the best ways of training graduate students to become future leaders in this area, and we have learned a lot along the way. Not that we have everything figured out.
In our first phase we focused in my own department on creating an environment that was more conducive to interdisciplinary work and individualized training. We exploded the traditional PhD curriculum, which in my field still follows a pretty standardized model in many institutions. We now have no required courses. Students do take courses, and they do need to be broad, but they can be broad in the way that makes the most sense for them, from Day 1. We also reorganized our physical space to be more conducive to collaboration. We have excellent lab resources, but no faculty member has their own lab. So we run a little bit like a kibbutz.
Blog: Pro choice on the linguistics curriculum
Our second phase was focused on creating a strong grass-roots interdisciplinary community, with graduate students as the engines of innovation. This process was helped a lot by an IGERT training grant from NSF that started in 2008, and that taught us a lot of things that we had no clue about at the beginning. (More on that in a moment.)
Blog: Developing and funding interdisciplinary programs
Our third phase is where we are now in late 2016. We’re building on the successes with individuals in Phase 2 to develop interdisciplinary teams that are equipped to go after big challenges that are too much for any individual scientist or individual field. In this we’re again very grateful to NSF, as we’re doing this as part of the first cohort of NRT training grant awardees.
I should say that although we have been working on this for quite a while, it can take a long time to know whether what you’ve done has made a difference. The talented students who we recruited early in Phase 1 are now a year or two either side of the tenure process. Our early IGERTers are just now getting settled in longer-term positions. So, we are now seeing the results from training efforts that are a few iterations behind what we’re testing right now. It’s nice to be able to do research on graduate education where you can get almost immediate results. But what I really care about is things that might take 10-15 years to bear fruit.
But by most shorter-term measures the results to date have been pretty good. In our IGERT program we trained 50 PhDs from 10 departments. And we’re especially pleased that 25 of those 50 were people who did the program without NSF support, just because they thought it was worthwhile. We had a 93% completion rate, average time to degree of 5.5 years, and many students got jobs that they never would have obtained without their IGERT training. Those numbers are not bad, but they’re pretty poor measures of what we actually did.
5. Another indicator of how well things have worked is the institutional and cultural change that occurred.
One of our signature activities is an intensive 2-week event that we put on every January that we call “Winter Storm”. It looks a little bit like a bootcamp, but it’s a bit different … and after David’s comments, I didn’t want you to leave with the simple take-away that activities like this don’t work. [One of the panelists presented findings arguing that boot camps don’t work. I was responding to that.] Winter Storm isn’t a crash course for new students. It’s aimed at students from all stages, and at faculty. It gets them together at a time of year when their schedules aren’t as hectic as they are during the semester, and aims to sow the seeds for skills, connections, or projects that can then be built upon during the rest of the year. It’s student led — most of the training is different students sharing their expertise — and we try to change it up as much as we can. We’re now getting ready for our 9th year of doing this. [In January 2020 we held our 12th annual Winter Storm.]
Winter Storm overview
Winter Storm 2019 gallery
We also launched an interdisciplinary undergraduate program called PULSAR that was modeled on the best features of our IGERT program.
And in 2013 our university decided that our grassroots community was so successful that they wanted to turn it into a broader university-wide initiative that now spans around 300 people in 17 academic units. (Language at Maryland) [Full disclosure: it’s not all plain sailing. Topic for another day …]
6. So what are the key ingredients? Well, four of our key take-aways are the following.
First. You get buy-in by giving people ownership. Our interdisciplinary programs have been the most successful when students take ownership of their own training. One of the best things that happened to us was that a year or two into our IGERT program the students staged a coup, taking over the leadership of key program activities. It made a huge difference, not only because the students did it really well, but because it made them much more engaged, and helped them to take a much more constructive approach to working with a diverse group. I should mention that the leader of this student rebellion was Shevaun Lewis, who now works on improving our graduate training as her day job.
Second: one of our mantras is that some things that seem like distractions aren’t. To take one example, we started an outreach program with local high schools because it seemed like a noble thing to do. But to our surprise it made our science better. Partly because it helped us to communicate better. If you can get the interest of a cynical 16 year old, then you stand a better chance of exciting a dean or a program officer. But also because it got people working together on something that was low risk, laying the groundwork for higher risk research collaborations.
Third: science is social. It’s hard to work well with somebody who you don’t know and trust. So anything that creates those connections of trust can lead to a better research environment.
Fourth: values matter. You need to value and recognize the things that you say are important. And it really, really helps if the faculty and students who others look up to are invested in what you’re doing. We have been fortunate that so many of our established faculty have bought in to what we are doing.
7. Although I’ve said that there’s no real conflict between academic and other careers, I can’t ignore the fact that many of our colleagues view non-academic careers as something foreign and less worthwhile. And students hear that message loud and clear. We have to do something about that.
We need to find ways of valuing and recognizing success in diverse careers. If I tell you that one of my former students got a tenure-track position at Johns Hopkins, you classify that as a success. No more questions asked. If I tell you that one of my students is a senior program officer at the National Academies you’re less sure what to make of it. Both of these are real cases, both are excellent people who I have the highest regard for.* But if we don’t have good ways of valuing success in diverse careers, then we’re never going to change attitudes.
[*There’s no way to sugarcoat this, one of these examples was Akira Omaki. We miss him.]
We also need to find ways of keeping graduates connected to our intellectual communities, regardless of their career. One reason why advisor-clones are highly valued is that they remain part of the same community, and so their successes are very visible. Students who pursue other directions, even if they just go into different academic fields, can feel like they’re moving to another country. There must be ways of overcoming this, especially now that our lives are dominated by social everything.
And we need to ensure that advisor and student motivations remain aligned. Advisors worry that students whose career goals are different from their own are less motivated to do the things that they value during their graduate career. Perhaps they just want to complete the requirements and get the heck out of graduate school. We need to ensure that students and advisors have a good understanding of each other’s goals, and how they can mutually benefit.
So, summing up: yes, there are tensions between standard graduate training and career needs, but I think we can do better if we get beyond the idea that this is because of non-academic careers. It’s because of all careers.
Update (January 2020)
Since taking part in that NSF panel in 2016 I have learned a lot about what it takes to foster an environment that supports student career preparation. I credit Shevaun Lewis for being my guide to much of this. She would claim that it’s not rocket science and that she’s just importing what others have already figured out. But I only partly believe her about that.
8. Early planning helps everybody. We have for many years encouraged students to individualize their training, and since 2008 we expected all students in our interdisciplinary program to develop a formal research and training plan upon joining the program. But it is only more recently that we have become more serious about career pathways, and about regular follow ups on the plans.
Students meet with Shevaun, and hopefully also with their primary mentor(s) to discuss their goals. It has proven useful to use values and talents as the starting point. Using a career label as the starting point tends to elicit a default answer (“professor”), whether or not it really aligns with the life that the student wants to lead. Starting with values and talents helps students to think about what motivates them, what they find rewarding, what they are good at, without pigeon-holing them in a specific job title.
This exercise is useful for everybody. It’s not just something for those interested in careers outside academia. It’s also a way to introduce the fact that there are many different kinds of careers inside academia, even within research universities.
A useful resource for this that our students use is Imagine PhD. It’s similar to the individual development plans (IDPs) that are now widely used in some fields, but tailored more to humanities and social sciences.
9. One of the new ‘experiments’ that we introduced in our NSF NRT program was “policy internships”. The idea was that students would gain experience by getting out into the Washington DC area and do something outside the ivory tower that makes use of their skills.
It sounded nice in a grant proposal, but we weren’t sure what it would look like in practice or whether it would work.
We now have experience with a good number of students. We have seen many different ways of pursuing these experiences. And we have a much better sense of how students benefit.
Jeff Green did a project with ACTFL on state guidelines on foreign language education. Nick Huang did a project with DC Immersion, a local non-profit focused on school language immersion programs. Nick had combined background in linguistics and economics that turned out to be really valuable in helping the organization. Lara Ehrenhofer did an internship with the Helmholtz Society in Germany. Other students took on issues related to academic policies and practices. Kasia Hitzcenko researched findings about conference reviewing practices, leading to a proposal to the Cognitive Science Society that was adopted by the society. Another group of students used their analytical skills to research gender representation in linguistic publications over the past 40 years, leading to a submission to a leading journal. Another team of students is tackling dialect awareness in the academy. I was initially resistant to projects that kept students within the university, but now I see that they have more value than I had realized.
Nobody has done a project on Capitol Hill, despite this being what most first think of when the term “policy” comes up. That is completely fine.
Much of what these experiences cover is synthesizing knowledge in order to bring about change of some kind. Whereas our outreach programs are mostly “low stakes communication”, these experiences are “high stakes communication”.
Most students were apprehensive about this aspect of the program. It has proven more daunting for students to step off campus than to step outside their department. But most found the experience to be surprisingly rewarding. They gained in confidence, and learned highly transferable skills on how to get things done. Those skills are useful in academia, too.
10. One thing that I have heard repeatedly from graduates who have pursued careers outside academia is that they are valued, as PhDs, for their ability to tackle large, messy problems, and turn them into something more manageable. This is something that we learn to do by doing good research. It’s highly transferable, and highly valued. It’s not so much about how well you write code, or how much morphology you know, or your statistical skills. It’s more about how flexible you are in analyzing problems.
The flip side of this is that we hear again and againthat we are not preparing students well to convey the skills that they have. We are accustomed to focusing on the products that we have generated (taught two courses, gave three presentations, published a paper, etc.). We are not accustomed to translating that into the skills that we have that allow us to do these things.
11. Tinkering is easy. Culture change is hard. It’s one thing to offer a career workshop here, an outreach event there, and to do various other things that add to the traditional curriculum. But it’s something entirely different to influence the culture and values around the role of a PhD.
Faculty views are all over the map. And students are strongly influenced by the values of peers and role models. Some faculty believe strongly that a PhD program is a time to focus 100% on standard academics and research. Your worth is measured in numbers of papers. Other aspects of career preparation is something to take up later. Others take a more expansive view. PhD programs should prepare students with a wider range of skills that will help them to succeed in what comes next.
Perhaps surprisingly, this is not a difference between older and younger faculty, with the old farts clinging to the past. In fact, younger faculty, who are more likely to still bear the scars from the academic job search, may be more likely to prioritize the things that helped them to that success.
We sometimes hear stories about faculty who tell their PhD students or postdocs that if they decide against an academic career then they are basically dead to them. I thought this was apocryphal, but it turns out that it’s not. There are also some who argue that being able to communicate effectively beyond one’s sub-field is a bad thing. Really.
A more common situation might be that a faculty mentor would say, “Good luck with that! But I know nothing about it, so I can’t help you.” It’s true that we’re not experts in everything. But it makes a difference if we show some curiosity and make a bit of an effort to find out more. If students pursuing academic careers get intensive support for interview preparation, while other students are told that they are on their own, that sends a clear message.
12. Some things have changed in the outside world since I originally prepared these remarks. The political climate has changed our sense of security in academia. Less than 2 months after the NSF panel I visited Stanford University. It was just a few days after the 2016 US Presidential election. I was struck by how many people were eager to talk with me about public science engagement, career pathways, etc. Previously, people were vaguely aware that I was invested in these topics, but few cared to talk about it. That changed markedly in late 2016. The current political climate has served as a wake up call for many of us.
13. In hindsight, promoting interdisciplinary research and education was rather easier than promoting more intentional career preparation for graduate students. That’s another reason why it’s useful that all careers are alternative nowadays. It allows preparation for diverse careers to be presented as preparation for a successful academic career. Everybody wins.