grad student depression

A recent survey found that ??% of Berkeley grad students suffer from depression. This should be shocking and dismaying. Yet no one seems to be shocked. Depression is one of the most debilitating illnesses one can have. (Other illnesses have primarily physical effects, even losing a limb has surprisingly little effect on long-term happiness. Depression robs you have agency, hope, joy, the ability to maintain friendships and relationships, and … should I say this here or just write another post about how depression is bad?) A workplace at which ??% of employees came down with cancer would soon be regulated out of business. But in academia, depression is seen as normal, an unavoidable side effect of the work we do.

This is patently false, because many people manage to avoid it. Academic life necessarily involves ups and downs. Projects go nowhere, papers are rejected, careers hang in the balance. Everyone experiences setbacks, but not everyone becomes depressed.

TODO: need to think more about how to structure this. how to make it compelling and readable. who my audience is, and what I want them to take away. general points to make:

• people go to grad school to help them grow into an identity. as undergrads, you look at your professors and think, I want that level of expertise, of self-confidence, of understanding the field deeply, 'at its joints', and using that knowledge to work on the important, basic problems. And of being able to pass those ideas on, to share what you've learned with others, so that humanity as a whole "levels up". It's a powerful vision. But you can't just choose to inhabit this identity. Real expertise and confidence requires serious training and experience. More than that, it requires social validation; having a community of friends to brainstorm with, discuss ideas, build a taste for good research, and have others look at your work and reassure you that it is interesting, important, or at least on the right track.

• the whole point of grad school is that you can't just choose to inhabit this identity, or construct it for yourself. If that were true, you could just go live in a hut for six years and emerge a seasoned academic. Building yourself into a confident, competent, and influential researcher requires years of hard work and dedication; this is the responsibility of the student. But it also requires good advice, mentorship, feedback, and a supportive environment. These are the responsibility of the program.

• so naive undergrads go to grad school hoping that they will find these things. some of them do. many more do not.

• the cause of depression is feeling your sense of self slipping away. seeing the identity that you've always held and wanted, and imagined you'd grow into, and pulled every lever made available to you to try to obtain, becoming less and less available.

• and the most pernicious thing about this is that you get blamed for it. Grad school is a place for independence, to take charge of your own development, to become an adult. If you don't grow into a successful researcher, it must be because you were not smart enough, not creative enough, didn't work hard enough, or otherwise just are not cut out for success.

  • but this ignores the entire point of advising, of culture, of program strength. it says you'd be just as successful locked in the hut, or having gone to a fourth-rate program, as you would be at Berkeley or Stanford. and when framed this way, no one at Berkeley or Stanford believes this. top undergrads with the right connections get detailed advice about which advisors they should consider and which groups are good to work in. everyone knows these things do matter. so blaming it on the student is just wrong.
  • on the other hand, the even worse thing is that it is true. you really are not smart enough. Smartness is not solely an inherent trait like the clockspeed of your brain. It's the accumulation of mental habits and models of the world. And these things can be learned, from good mentors and smart peers, so that you really are a more effective researcher who makes better decisions. Which means they can also fail to be learned, if your program does not put you in a position to learn them. So yes, you can be blamed for not taking the initiative to develop your own research projects or leave a bad advising situation, because there really were paths available to you that you did not take. And someone who had learned from good role models and mentors would have been able to see those paths, and take them. But you did not, because you haven't had those opportunities, so you really are stupider, and really do make poorer-quality decisions, which then lead to even more failure, and the cycle is self-reinforcing.

• there is a "success privilege" that is realized to different degrees. When we talk about "white privilege", we partly mean privilege that is inherent in having white skin, like not being shot by the police. But there are also aspects of fortunateness that are merely correlated with white skin -- being raised by economically comfortable parents, who care about your education, who can teach you strong values, pass on the mannerisms and norms and good habits of the middle (or upper) class, introduce you to peers and social connections and role models and a community that cares about you. It's hard to understand how much this privilege does for you, until you become close with people who have not had it.

• Similarly in academia, there is a privilege of being 'born', more or less by luck, into well-functioning research groups that feature great advising, strong collaboration and support. And without even trying, you learn by example that this is how research should be done. You learn all these skills, that are never written down (and could not be successfully learned in that form even if they were, any more than you can learn to play soccer by reading a book), and gradually you start to become a successful researcher. You become confident in your ability to advise people, because you've seen it done well and you know what works. You have an idea of what a healthy research group culture looks like, so you can start to think deeply about how to set up your own research group, how to change, how to improve.

• By contrast, when you don't have these things, they are almost impossible to develop. I don't feel qualified to become a professor and advise students, because I've never experienced advising done well. I don't feel qualified to run a research group, because I have no experience of being in a research group with a functional culture worth emulating. I don't even feel qualified to plan a research project or write a research paper, because, again, I've never been part of any of these experiences in a way that was successful. Sure, I have some idea of what not to do. But that's not particularly helpful, any more than being given a bunch of random gibberish strings and being told they are not French is helpful in learning French.

• A friend recently graduated who has, overall, had a very successful PhD. She's done great research, made a lot of friends in the academic community, and has a great job lined up. At her dissertation talk she dedicated several minutes at the end to talk about the things that mattered to her success. Her advisor who she has a great relationship with. Her research group, which is tight-knit and supportive. The collaborators she's worked with and learned from. Her grad school experience was great. Like everyone, there were difficult times, doubts, imposter syndrome, rejected papers. But she managed to get through those and do successful and influential research. And I believe her when she says those support factors mattered. Yes, she is incredibly smart and capable. If you locked her in a hut for six years, she would have probably accomplished some things. But she would not have grown into the rising young star that she is.

• There are many ways to be a bad advisor. Micromanaging, or not managing at all. Never allowing a student to build up their own research program and propose their own projects. Or giving no guidance at all, even at the beginning when a student cannot be expected to formulate good projects. Being a good advisor is hard. There are external constraints and real reasons why a well-meaning person could err in either direction. But it helps a lot just to care about your students, and to give an impression that you're trying to be a good advisor.

• Loneliness is self-reinforcing. When your entire self-image is slipping away, you can't talk to people or build relationships. You can't discuss the things that are going wrong for you, because those things are your fault and to argue otherwise would be to imply that others don't deserve the successes they've had. You can't get excited about good things that happen to other people, because you can't get excited or feel joy, period. When the most fundamental aspects of your identity are broken or slipping away, you are literally dying: when it's all over, someone else might live on in your body, but if you can't salvage your identity, that person won't be you. When things are fundamentally not okay at that level, it's impossible to enjoy trivial small talk, or social activities, or anything that isn't directly connected to solving what is fundamentally broken about your life. You especially can't do it when your friends are also in your field, and so their favorite topics of conversation include their own research trials and doubts. Which are self-evidently less fundamental than yours, because the fact that they're even able to talk about them directly means that they cannot be as depressed as you are. But to say that would be to be a horrible friend. You want to offer support when others are going through tough times, but you know you won't be able to, so rather than fail even more directly as a friend, you just withdraw from social situations and relationships where these things come up. No matter how much they insist otherwise, no one wants to hang out with someone who is relentlessly rude and negative, who can only talk about all the things that are wrong in their life. And so when you're in the positition where things are fundamentally not okay to the level where that's all you can think about, you'd rather withdraw and keep your worries to yourself, than play things out to the point of actually ruining all social occasions and driving all your friends permanently away. [note: this doesn't have that much to do with grad school, it's more of a generalized depression point].

• More relevant to academia, depression is self-reinforcing. It takes a certain level of confidence to approach other faculty or grad students and ask for help or advice with your work. It takes a basic level of well-being and excitement about research to think creatively about new projects.

• This is all somewhat independent of the academic job market. Sure, it doesn't help that the market is insanely competitive. But in CS, anyone who builds serious skills as a researcher will be able to find opportunities to apply those skills, in our outside of academia. The depression comes from not having found the opportunity to acquire those skills.

• Note this post draws heavily on my own experiences. But it's not intended as a first-person account. Not everything here has happened to me. I'm not the only grad student I know to have been depressed. But it's also not exhaustive. I can't pretend to speak for everyone, including depressed people who find it especially difficult to share their stories.

• Concrete solutions.

• stop lying to students during visit day. If a student wants to do X, don't accept them to do Y without making it very clear there will be no support for X.

• give students agency to find their phd homes. fellowships allow students to try their own projects, to choose advisors they are excited about. GSRs tie students to particular projects and (implicitly) make it very difficult to explore.

• therapy. or at least multiple advisors.

• build larger groups and institutions beyond individual research groups. Just like the federal government can guarantee rights and redistribute wealth to mitigate the damage of dysfunctional state governments, department-level culture helps students feel like a "part" of something even if their research group does not. BVLC and BAIR are good at this.

• hvae a culture that advising matters, that community matters. Bill Thurston decries theorem counting (and paper counting, more generally), but this is what Berkeley faculty are incentivized towards. If an advisor doesn't care about their students, or doesn't do a good job, that's fine: just don't let them admit students (or at the very least, make this fact transparent). Advising should not be a duty, it is a privilege and a calling. Any professor who thinks of spending time with their students as an burden, should be relieved of that burden by the department.

• ultimately, if you've been privileged enough to have learned the skills and achieved success as a professor, you shouldn't feel bad about it, any more than whites should feel bad about white privilege. But hopefully you can feel thankful, blessed, and want to pass those same privileges on to as many others as you can, while being understanding of those who have not had the same success.

For students suffering from depression. I don't really know what advice to give. I can't really say that it gets better, since as far as I can tell, it doesn't.

• develop resilience (sheryl sandberg)

• do everything you can to reach out

• "there is no speed limit". you don't have to do everything on the ascertained schedule. some people are lucky enough to find mentorship and community in grad school. but just as there's no guarantee you will develop on the appointed schedule just because you go to grad school, there's no law that you are not allowed to continue growing after grad school. Yes, those years are lost and you will never get them back. But plenty of people start their PhDs at 30 and go on to do alright. You can start a new phase of your life now too, and hopefully at least have a better idea of what to search for.

• if anyone else reading this can identify with the experience and has helpful advice, that'd be a useful post to write.

• I don't want to deny the role of individual initiative and responsibility. Ultimately no one else can construct your life for you. Grad school really is about growing up and taking responsibility for yourself. If one person fails at this, you can reasonably blame them. But when 40\% are falling into depression, maybe there are some systemic issues we can work on.