Two weeks before my qualifying exam, my PhD advisor — Kentaro Inoue — and I met to discuss the thirty-first draft of my dissertation research proposal and encountered a sticking point on the second of three aims.  I rewrote, and reworked, and rewrote the section throughout the day, returning to lean on his office doorframe three or four times. He did not understand what I was trying to articulate; I thought what he was insisting my proposal contain was an impossible amount of work for one person.  By 4:30pm, we were snarling at each other, then I caved, stomped back to my office, and had the textbook two-weeks-to-QE breakdown. For my own sanity, I decided to study from home the next day.

It was a great choice.  I was perfectly calm the following evening when my labmate called me just before dinner, asking for my address.  He needed to talk to me. This was not standard operating procedure for our lab. I sat at my kitchen table, pushing food from one side of the plate to the other while mentally sorting through what could be wrong.  At first, I thought I had to be in trouble somehow — maybe I should not have worked from home? — but if I were in trouble, why wouldn’t my advisor have called me? Why deputize a labmate? My phone blinked; texts appeared from other students in my year: “I’m so sorry.”  “I just heard.”

What had they heard?

I paced on my front porch, phone clenched in one hand.  My labmate pulled into a parking space, my other labmate in his passenger seat.  Two safe. Two unharmed. Both pale and tense. I let them in — my cat fled to my back room — arms wrapped around myself.  “Kentaro was in an accident, and,” my eldest labmate said, “he didn’t make it.” The three of us stood there, in my front room, frozen.  In my memory, that moment drags out: three people, almost unable to meet each other’s eyes, all wondering, “What in God’s name do we do now?”

Joining my lab two years prior was a decision I agonized over.  As a PI, my advisor was tough, demanding, and hard to read. When asked how he was to work for, one labmate said, “He’s tolerable.”  But I liked him, as a person, and I liked his research, and I liked the people in his lab. So in a split-second decision after about seven hours straight of benchwork without eating, I stopped in his office to say that, yes, I wanted in.

“Good,” he said.  That single word, I would come to understand, was the most and best praise he offered.  Then he said, “I will push you.”

After he died, I thought, “This is a hell of a push.”

There’s no road map to follow when your advisor dies, no protocol for sudden and premature death.  You plunge into a tunnel blindfolded: turn one direction, you scrape skin onto rocks; turn the other, you slip downward, gravity’s victim.  Because the situation is rare, graduate programs do not have any formal resources or plan for what happens either. It’s isolating and terrifying to try to figure out what you do next and how to finish your project, and all the while, you’re grieving for the relationship you’ve lost.  I would never have chosen to go through graduate school like this, but it has made me resilient and adaptable.

My labmates and I stumbled through together.  While I emailed our program chair from my kitchen table, subject line, “URGENT,” they called the other people we had to tell that day.  Over the first couple weeks, we kept the lab limping along while our program sorted out what our options were. We merged with another lab and restructured our projects, and, two and a half years later, we all finished or are finishing our PhDs.  We had each other, and we had a supportive program, and that made a big difference.

PhD advisors are human beings, subject to illness, injury, and other crises.  Bad things happen during PhDs, and you can’t control them. You can, however, control for them, to a certain extent: the experimental error term of the PhD equation.  Graduate students, advisors, and programs can take steps before crises happen, in the immediate aftermath of crises, and in the long-term to assist students.

Best practices beforehand:

No one predicts untimely death or sudden illness, but there are practices which should be part of any graduate student’s training that can help in the event a PhD advisor dies.  Graduate students can cultivate close relationships with other professors who do similar research, at your university and elsewhere. There’s no downside to knowing your potential collaborators well, and these mentors can support you when your project slides off the rails for normal reasons.  Students and advisors can also practice good planning. If you find yourself having to pick up the reins of your project without your advisor present, it helps enormously to have a feasible project with clearly defined objectives and backup plans for experiments not working. And if you find yourself on any given Tuesday, it helps to have such a project.

Asking a graduate student’s advisor to plan for their own untimely death is perhaps a little morbid.  However, it certainly cannot hurt for an advisor to think about what collaborators and friends could step into their shoes.  Even temporarily — emergencies come in many different flavors. Advisors should also ensure that critical laboratory information and records are well-kept and accessible to students.  Leaving the contact information for the centrifuge repair person in a place where at least one student can find it only helps in a normal situation.

Beyond the advisor, graduate programs should recognize that this situation will happen and have procedures in place in advance.  It’s not reassuring as a student when you feel like everyone else around is making up the procedures as they go along.

Immediate concerns (first month after sudden death):

Hopefully, your precautionary actions will just mitigate small crises and minor problems, but in the event your advisor dies, the immediate aftermath is chaotic.  Telling people. Meeting with the people who run your department. Whatever the funeral is. Sending flowers to your advisor’s family. And, if you’re anything like my labmates and me, you’ll try to work in there as well: because a research imperative requires it, because it’s a safe place to hide, or because you don’t quite know what else to do.  After one meeting with our department, I stood in the corner of the lab leaning against the -80 with my eyes closed because I knew nothing short of an actual earthquake was knocking that thing over.

Decisions might be made quickly — for me, our department was trying to assign us to a new advisor within 48 hours, almost forgetting to consult us or the other professor.  Within the chaos, grief becomes almost secondary, but grappling with the loss of this incredibly important person, however best for you as an individual, should be the priority.  You should figure out the minimum requirements to keep the lab running — how to order pipet tips, for example — and you will have to handle things that no one else has the knowledge to handle, like locating the safety and usage documentation for your radioactive materials.  But aside from that, your responsibility is grieving. You’ve lost a central person in your life, and it is not a simple loss to cope with.

Graduate programs and departments are responsible for finding sources of funding for you.  They figure out how to handle your lab’s grants, and they can find options for you to continue working on your degree, on the same project or in another lab.  You can help find yourself funding and options for continuing, but, at a certain level, this is just not your job. Especially when dealing with large grants. If it seems like they do not have your best interests at heart, reach out to whoever you can to advocate for you — there will be people who want to help.

Some do-and-don’t guidelines for the immediate aftermath phase include:

  • Do:

    • Communicate clearly and widely — don’t let yourself be kept in the dark about what’s happening, and get the backup from high-ranking department members that you need.

    • Get things in writing, especially funding.  Do NOT skip this step because you’re the lowest person on the totem pole.

  • Don’t:

    • Decide what you’re going to do immediately.  Graduate programs should not pressure students into a decision; they are smart people who should be able to hold off funding agencies for at least a month.

    • Feel like it’s your responsibility to continue your research to honor your advisor. You’ll honor your advisor whatever you chose to do, and you should choose what’s best for you, not what will please someone who has passed away.

Deciding what’s next: Should you go, or should you stay?

You will not have unlimited options about what to do, and the options you do have may not be good.  There’s no right answer to whether you should continue your project with a different advisor or change labs or apply to a different program to start over or move to New York to be a nightclub singer.  Contemplating walking away is difficult and…emotionally complicated, for lack of a better term. You’ve invested your time and sanity to your project, and the idea of starting over is overwhelming. But.  Continuing a project with a new advisor is not easy either. Expect to add to your graduation time and/or sacrifice most of your planned project if you switch advisors. If you are in your first or second year, seriously consider changing labs or programs.  I didn’t want to change projects, for reasons that were all emotion and no logic, and I was just far enough along that I had the option of continuing, but that does not make it the smart decision. Just the one I made.

If you’re in the middle of your PhD, here are a number of questions about a potential replacement advisor to think about:  

  • Is there a qualified (and willing) professor to take over as your advisor?

  • How close of a relationship do you have with that professor?

  • How similar is their mentorship style to what you’re used to?

  • How familiar are they with your project?

  • How many graduate students (at what stages) are in that professor’s lab, and how many graduate students (if any) are coming along with you from your original lab?

  • How many postdocs are in that professor’s lab, and how many are coming with you?

  • What is the replacement’s professional reputation (this is the time to dig out the dirt), and are they as well-connected as your original professor?  (And, in general, think like a rotation student — how happy are the current students, how quickly do they finish, what kind of jobs do they get, etc.)

  • What sources of funding are being offered to you for stipend support and for research materials, how long are those guaranteed for, and are they adequate (expect to wander down more black holes than may be typical)?

  • How involved is your committee, and how much are they willing to be involved?

  • Are there people (lab alumnae, collaborators, etc) who can be recruited as sources of aid or missing expertise?

If you’re towards the end of your PhD, you’ll need someone to write recommendation letters in place of your advisor, and you may not have that much time for that person to get to know you.  Who is the next person on the list to write you recommendation letters? How well do they know you and your project? Is there a draft or an old recommendation letter from your original advisor that can be given to them?  Can someone (for example, your department head) step in and write a letter explaining the situation? Do you have the resources to find and get the next step on your career without your original advisor, and if you don’t, who has those resources?

None of these questions are fun.  Everything about this situation is awful, and if you’re trying to make this decision, I’m so sorry.

Concerns that arise in the first year:

After the immediacy of crisis-mode passes, then things get easier and harder.  Get help to handle the grief, and, unfortunately, don’t expect it to be quick to fade.  Six months after my advisor died, I realized I was furious with him for dying, and I then got stuck on rage for the next year (how do I wish that statement was hyperbolic).  Try not to let yourself become isolated, and try not to take on responsibility for things that are not your job. We continued working in our original lab space for the first six months and got used to wrangling lab management almost entirely on our own.  That helped initially — our lab was comforting and familiar, and one less thing to have changed — but eventually, it was like working with a ghost right outside the door.

You will have to take a hard look at your project and decide what is and is not possible without your original advisor.  The earlier on you are, the harder the look. You will also have to work to establish a relationship to your advisor, which I, at least, have not found the easiest task, through no fault of my new advisor.  It’s hard to trust someone new in the PhD advisor role, especially if you lost the first one in a brutally unexpected fashion. What helps is to cling like a limpet to your other mentors: other professors you know well, older graduate students, lab alumni.  They can help keep you afloat while you’re trying to remember how to tread water. Keep seeking out more mentors — the more help, the better. You should try to get some research done, too, even if it is not brilliant. Anything you accomplish here against such ridiculous odds is a huge victory.

A new advisor should be cognizant that this situation may be excruciatingly hard for their new student.  It helps if new advisors communicate clearly about what they will honor about a student’s original plan, what they expect, where their funding is coming from.  In writing, if possible. Three years down the line, both advisor and student will want to remember what you first agreed, even if you’ve mutually changed everything about the project.  Becoming familiar with the background for a new student’s project will be time-consuming, but an advisor’s unfamiliarity can become a source of frustration for the new student. If you have no idea what to advise, facilitate the student seeking additional help.  Finally, figure out how to move the new student into your research space.

Graduate programs should finalize and formalize all funding arrangements in writing with you, and if they don’t, this is something you should be a pain about.  Your program can also assist your new advisor by letting them take equipment from your old lab or shuffling some of their service responsibilities. And programs should institute regular checks on the student.  Once things started to go back to “normal,” I felt awkward asking for assistance, even if I still needed it.

Later concerns:

One year turns into two, then three.  New students and professors enter the program; waves of undergrads flow through campus; and your experiments fail, perplex, and hopefully — eventually — succeed.

While I write this, my eldest labmate is days away from turning in his dissertation.  My second-eldest labmate is in the oscillating phase where sometimes he’s a month from graduating and sometimes a year.  I am getting there. It would make more sense, in many ways, to wait until I graduate to write all this down because then I could say it is possible to finish, that you can make it.  But I think if I wrote this at the end, I might be tempted to skip the slog as we worked to graduate in favor of those first frantic days and months. For about the first year, we were finding new problems, trying to fix them, and forging connections with our new lab.  But once we’d mended and replaced what we could, there were still gaps we just had to live with and work around.

A PhD is already extremely difficult, and if your advisor dies, you have a permanent handicap.  Continue to get help for grief if and when you need it and for the anxiety and depression that graduate school handed out at orientation.  Continue adjusting your project to best suit your new advisor and find the people, expertise, and materials necessary to fill in for your original advisor.  Advocate for yourself. Your program wants you here and wants you to succeed, but as time goes by, your program will sometimes forget that everything is just a bit more complex for you if you let them.  Remember why you’re getting a PhD, and check in periodically to make sure you’re getting a PhD for yourself. If you continued your project, you might find out months or years down the line that it is not, in the end, possible.  The initial decision you made about what to do might not have been the right decision, and you can change your mind or press your program for changes. Graduate programs should fulfill the funding commitments they made and continue to provide assistance.  They should also recognize that you will sometimes have to play by a different rulebook for what is and is not possible.

For me, the wound where my relationship with my first advisor was severed has scarred over, and I can go more than one day in a row without noticing the scar.  Then, the speaker at his memorial seminar will project his picture, and I’ll feel like I’ve been punched in this stomach. I’ll struggle to reproduce data from a previous student, and I don’t know if the problem is something my advisor never got around to telling me.  While I argue with my new advisor about an experiment, I’m arguing in my head with my first advisor. It is still hard, and it will never not be hard.

When I joined my lab, my first advisor said, “Good.  I will push you.” He did, in ways neither of us ever expected.  I’m still pushing back.