One year ago, I wrote some feedback for the steering committee of the Programming Languages Mentoring Workshop. I am publishing a lightly edited version here in two parts. You can find the original document here.
The Programming Languages Mentoring Workshop (PLMW) provides students from all over the world with a unique opportunity to travel to programming languages conferences and to learn about the field of programming languages and what it is like to do research in it. PLMW has given me the opportunity to attend POPL when I did not have a paper there, just as it has done for hundreds of other students. I am grateful to its organizers and speakers for all the hard work they put into creating interesting presentations and mentoring events in which students can get to know and interact with the best researchers in our field, who might otherwise be difficult to approach. It is clear that huge efforts are involved in organizing the workshop, into making it diverse and inclusive, and encouraging students from various backgrounds to pursue a PhD.
I am sure that being motivating and inspirational is the foremost goal of PLMW, and I wrote this document because I believe that unwittingly, some of the talks have, at least for me, been in parts more discouraging than encouraging.
I am writing this document because I believe that the discouragement to stay in academia was not a deliberate result. Rather, it was an unintentional side effect that resulted from
Before analyzing the different types of PLMW talks, I would like to note that this document is based on my limited experience with three SPLASH and one POPL workshop. I did not analyze the content of any other workshops. The ideas in this feedback are based on my perception and opinion and should be taken with a grain of salt. More objective information can be found in surveys such as Taulbee, Berkeley’s graduate well-being report, and Nature’s PhD survey.
Even though this feedback document contains suggestions on how to improve the workshop, it is in the first place a personal essay about my subjective experiences at PLMW. A document with advice and recommendations to speakers and organizers can be part of a separate line of work that needs to be based on more research and be written by people with more knowledge on organizing this type of events.
PLMW is already a fantastic workshop that I would recommend to any student starting in PL. However, I believe that the workshop can be significantly improved by
In the two parts of the document, I will discuss four types of PLMW talks: PhD 101, grad skills, life stories, and technical talks.
PhD 101 talks provide students with a general idea of what it’s like to do a PhD, and why do it in the first place. This section discusses the ways how the PhD is represented in current PLMW talks, proposes three more topics to address at PhD-101 talks in the future and suggests a way to present these difficult-to-address topics.
The best PhD-101 talks I have seen are Grad School: A Survival Guide by Matt Might and Why do a Ph.D. and how to pick an area? by Yannis Smaragdakis. I like these talks because they
A few quotes from Matt’s and Yannis’ PLMW talks are presented in the left column of the table below. Each quote is contrasted with “encouraging” statements that other PhD-101 presenters have made about the PhD.1 The rightmost column mentions some of the problems that come up during the PhD, as supported by studies.
|Genuine portrayal (Matt and Yannis)||Superficial encouragement||Studies|
|Reasons to do a PhD||“If you want to figure out what it’s like to extend the [boundaries] of human knowledge, you’re gonna enjoy grad school regardless of what follows”||All of you should do a PhD|
|What it’s like to do a PhD||
||“Graduate school and academia is the most fulfilling career path! Go for it!”||
||[There are good prospects of getting an academic job;] our department has four openings this year!||poor job prospects: in 2016, only 14% of PhD graduates in PL were hired as tenure-track professors, and 11% as Postdocs|
The problem with the encouraging statements is not just that they are vague, but that they provide an unrealistic portrayal of the PhD experience. In an attempt to encourage more people to pursue a PhD, the speakers depict the PhD simply as the most satisfying of all jobs, because they believe that academia allows one to work on more interesting problems than industry.
By contrast, Matt’s talk not just acknowledges, but centers around failure:
Pushing at the boundary of human knowledge mostly consists of failing. . . . Failure is the dominant mode of operation for academics.
Matt describes different types of failure in the PhD and proceeds to discuss how to deal with them. I find that such realistic depictions, even if negative, and specific advice, create a strong ethos in the speakers, which inspires trust.
Matt’s and Yannis’ acknowledgment of the troubles that come with the PhD actually made me feel motivated about grad school. I perceived the problems they described as a challenge, as things I felt compelled to overcome.
The rest of this section focuses on three specific aspects concerning the lives of PhD students: mental health, work/life balance, and job prospects. I will examine how these topics are, and how they could be addressed at PhD-101 and other PLMW talks.
I recently met a PhD graduate who, when I told him that I was doing a PhD, asked me: “So, are you already past the major depression in your PhD?” I remembered the time when I was so worried about a problem I had at work and about the uncertainty of my future that I could hardly think of anything else. I would wake up in the middle of the night, constantly thinking about my work-related problems. I became less motivated to make progress on my work, felt worthless, and considered quitting the PhD.
So at first, I thought yes, I am over that most depressing phase. But then I realized that it could always get worse: after all, my PhD is still going on.
Mental health is, unfortunately, a wide-spread problem for many PhD students. According to numerous studies, every second PhD student/academic develops mental health problems. For example, according to a graduate-student well-being report conducted at Berkeley in 2015, 42 to 48% of PhD students in engineering scored as depressed, while a previous study in 2005 showed that 10% of PhD students contemplated suicide. The following table shows the frequency with which PhD students suffered from various mental health problems, as opposed to other highly educated people.
|Felt under constant strain||40.8||27.5|
|Unhappy and depressed||30.3||13.6|
|Lost sleep over worry||28.3||18.1|
|Could not overcome difficulties||26.1
|Not enjoying day-to-day activities||25.4||13.1|
|Lost confidence in self||24.4||8.0|
|Not playing a useful role||22.5||9.2|
|Could not concentrate||21.7
|Not feeling happy||21.2||11.1|
|Could not make decisions||15.0||6.0|
|Could not face problems||13.4||4.3|
|Risk of a mental health disorder (due to 4 or more symptoms)||31.8||14.0
The numbers might seem too high. From talking to students at my own university, research group, or at PL conferences, I would not expect half of them to be depressed. However, from my own experience, I have usually found out about problems such as depression and anxiety in other students only after getting to know them more personally. It is usually unexpected, especially with people who look happy, outgoing, social, etc. I have recently found out that a PhD student that I know, whom I would describe with all those qualities, has been considering quitting his PhD because of an insufficient number of publications, and has been talking to friends about killing himself.
I myself have also encountered several of the problems mentioned in the table. I have been able to solve some of the work-related problems by openly talking to my advisor, and finding solutions together with him. But most importantly, I have been able to get through the most difficult times of the PhD by seeking counselling.
Unfortunately, according to a 2017 Nature PhD survey, 52% out of the 1,574 students who listed mental health as one of their highest concerns did not seek help for anxiety or depression during their PhD.
I believe that one reason for that is the stigma associated with psychological counselling. Research has shown that public stigma against counselling affects the willingness to seek psychological help.
If mental health is such a wide-spread issue in academia, and if there were no stigma attributed to the topics of mental health and counselling, then we would surely discuss these issues at the PhD-info talks at PLMW. However, I have not seen these problems addressed at PLMW in any way. PLMW presentations talked about mental health neither in explicit terms nor even simply by giving students advice on what to do during psychologically difficult times. In my opinion, ignoring this devastating aspect of academic life is indirectly contributing to the stigma against seeking help during the depressing moments of the PhD.
It is important to support students in overcoming mental-health issues, not just because happiness and subjective well-being have a positive effect on health and longevity. Overcoming anxiety and depression and being happy also leads to increased productivity, creativity and motivation, attention, and collaboration.
Only half of students suffer from mental health issues, so why do we need to address this awkward topic, one might think. But an even smaller number of students in the audience will be able to ever become a professor, and yet the freedom, fulfillment and greatness of an academic career are recurring topics at the workshop. Helping students overcome mental issues that arise in their PhD is, therefore, more relevant and useful in practice than glorifying the job that almost none of them will be able to get. I believe that a talk about how to deal with anxiety over one’s job prospects is no less important than a talk about how to choose an advisor.
If doing a PhD in computer science were physically dangerous, and half of students received work injuries, we would constantly stress the importance of visiting a doctor when we develop certain symptoms. Given the high numbers of mental-health problems, anxiety and depression can be considered psychological work injuries which are acquired at a rate of 50%. The need to address mental health in PhD students is best formulated at the end of Berkeley’s graduate-well-being survey:2
[I]t is important to balance economic measures of societal progress with measures of subjective well-being, to ensure that economic progress leads to broad improvements across life domains, not just greater economic capacity. . . . [A]n argument could be constructed that raising subjective well-being leads to positive externalities or spillover effects across a number of policy domains, ranging from health to traffic safety. Given the tangible benefits to individuals and societies of moderately high well- being, it is imperative that we act to effectively put well-being at the heart of policy and generate the conditions that allow everyone to flourish.
It would be useful if at PLMW, we could combat the stigma against mental-health issues, by acknowledging that the PhD is psychologically challenging. We could mention what types of problems can cause anxiety and depression, such as lack of research results and ideas, impostor syndrome, financial insecurity, problems with advisors, etc. We could say that having those problems and feeling affected by them is normal and that there are many other people who experience the same. This could be part of a separate type of talk discussed below.
One of the recurring impressions I have gotten at PLMW is that a successful researcher values work above all else. One must sacrifice or simply lack other interests, in order to fully dedicate their lives to research. There is neither space nor a need for a work/life balance. However, I believe that these assumptions are incorrect and the underlying messages are unnecessarily discouraging.
The idea that research has to be the most important thing in life has come through at PLMW explicitly, through comments and responses to audience questions about work/life balance, and implicitly, through a lack of discussion of the topic.
When a student in the audience asks a PLMW presenter or the panellists about work/life balance, at first, there is silence. If it’s at a panel, nobody wants to go first. Then, they start answering. I remember three types of responses/advice relating to work-life balance, one more, and two less useful.
“Get enough sleep, exercise, and eat healthy.” I have found this the most useful response because it acknowledges that PhD students are allowed to care about something other than work, and gives the most meaningful interpretation to the concept of work/life.
However, I think this advice is also problematic exactly because of its narrow interpretation. According to the Oxford Dictionaries, work/life balance is the “division of one’s time and focus between working and family or leisure activities.” The advice of leading a healthy lifestyle has little to do with spending time with family and friends and having hobbies. I appreciate that in this case, the presenter is actually caring about the well-being of graduate students, and wishes them to be healthy. Speakers should definitely keep stressing the importance of leading a healthy lifestyle.
However, even though healthy eating, sleeping, and exercising habits are not part of “work”, they are also not what I understand by “having a life.” Equating work/life balance to a healthy lifestyle unfortunately also conveys the impression that in grad school, there is little room for other things in “life.”
“When [some important event] happened, I prioritized that event over my work.” At one PLMW panel that I attended, two panellists’ answers were about a sacrifice they had made on the day or during the time when they or their spouse had a child (the other two panellists’ type of answer was (3)).
I understand that the message here should be that there are things that are clearly more important than research and work. However, if the only example of such a thing is a life or death situation, then again, what it actually means is that work has the highest priority over everything else, except for emergency situations.
“My work is my hobby.” This type of response suggests that the distinction between work and “life” is redundant. Rather than addressing the audience’s question, it talks about the love of the presenter/panellist towards their work. Yet again, this gives the impression that in order to be successful, our work should consume all of our life, and it becomes shameful to have even asked the question.
In one way or the other, all three types of advice convey the message that work is the most important thing in life. The first assumes a limited notion of work/life balance by equating having a life with being healthy. The second defines work/life balance as allowing time off work for extreme life situations. The third defines work/life balance as a balance between work and more work. In all three, there is no room for other fun activities and time with family and friends.
One PLMW speaker explicitly says that the PhD is not a 9-to-5 job and that it requires an “intense” interest in research. “If you’re not intensely interested, why are you even here?”, they ask.
I do not see why the PhD cannot be a 9-to-5 job for anyone. To me, even six hours of focused, concentrated work can be perfectly productive; moreover, after that time of undistracted work, I feel tired enough that I need to rest.
To me, work/life balance is about every-day life (as opposed to only the days of childbirth or other major life events). It is about the conscious decision I have to make to prioritize my family and friends over my work, to allow time for them and myself even when I want to work, because I know that it is necessary for my health, my relationships, and finally, my own productivity.
In my experience, after working long hours, during evenings and weekends, I can get more done for a limited period of time. But after a few weeks or months, depending on the work intensity, I become exhausted. As a result, I lose motivation, care less about work, and start procrastinating.
I learned that I need to put an extra effort to set boundaries for work and forcefully switch to other activities that have nothing to do with it. Taking time off, spending time with friends and on my hobbies gives me the energy and motivation to resume my work, and makes me more focused and efficient at work.
Numerous studies have also shown that maintaining a healthy work-life balance is crucial to happiness, health, and productivity. Research suggests that spending more time at work does not necessarily lead to getting more work done; in fact, there is a correlation between higher productivity and lower working hours.
In Nature’s PhD survey, 14% of PhD students listed the difficulty of maintaining a work/life balance as their biggest concern since starting a PhD. The only two higher-ranked concerns were both related to job prospects after the PhD. The following is a quote by a PhD student taken from the Berkeley graduate-well-being survey:
The work-life balance is terrible, and there is a culture of silence around how we feel as graduate students. I feel much better after talking with counselors at the Tang Center, but it is a little ridiculous that I have to go to therapy simply to have someone ask me how my day was or how I’m feeling.
By refusing to acknowledge the important aspects of work/life balance, we solidify the stigma that we already have against enjoying activities outside work. This is one of the reasons why I have felt intimidated by academia when I attended PLMW the first few times (later I became used to it). I believe that this attitude can scare away capable and talented people, who are interested in PL and academia, but who do not share this attitude towards work.
In my opinion, the problem we need to solve at PLMW is not how to prioritize research to life, and not whether the PhD is a 9-to-5 job. Instead, we need to address two other issues:
How to be most productive and efficient during the limited time we have for work?
How to make the most out of the limited time we have for rest, so that we can be productive again?
The first question needs to address how to actually get into a focused, undistracted mindset during working hours. How can we get rid of the email, social media, and other internet distractions that affect many of us?
Instead of suggesting that in the PhD, it is required to work overtime, it would be very useful to focus on time-management practices or talks on how to avoid procrastination. For example, in his talk Matt Might mentions the approach of Getting Things Done and distraction-blocking browser extensions; he has plenty of other advice on his blog. I have also seen a Time Management talk in the program of another PLMW which I did not attend. I believe these are important topics, and it is great if PLMW can keep having such talks in its program.
The second question requires us to actively promote a fulfilling life outside academia and work, to encourage students to rest, spend time with closed ones, have hobbies, and take vacations.
The work culture in our field is not PLMW’s fault, and it is not PLMW’s mission to change the world. But we do not need to contribute to the existing problems, either. We could make a first step towards shifting the work/life scale towards “life” by acknowledging that a life outside the office should exist.
PLMW’s goal is to encourage students to pursue academic careers in PL. The benefits of being a professor are explained at every workshop, and they are clear. One presenter even suggested that academia is the most fulfilling career path and encouraged everybody in the audience to become a professor.
Unfortunately, the few available professor positions are rarely addressed, and the topic of the academic job market becomes the elephant in the room. According to last year’s Taulbee survey, only 14% of PL graduates landed a tenure-track position in 2016 (11% in computer science overall, see table below). Unsurprisingly, in Nature’s 2017 PhD survey, 38% of PhD students listed the post-PhD career and availability of professor jobs as the areas of most concern. For 52% of interviewees, academia was “the most preferred sector to work in.” Additionally, according to Berkeley’s graduate well-being survey, career prospects were the highest predictor of overall student well-being.
In addition to the low number of available academic positions, university prestige plays an enormous role in faculty hiring, which makes becoming a professor even harder for the majority of PhD graduates. A study in 2015 showed that in computer science, 25% of institutions produce 80% of tenure-track faculty and that only 9 to 14% of faculty members get hired at universities that are more prestigious than their doctorate, revealing a “steeply hierarchical structure that reflects profound social inequality.”
Given the poor academic-job prospects and the anxiety that students feel about their future, it would be very helpful if PLMW could dedicate some time towards discussing alternative career paths while acknowledging that such careers can be intellectually engaging and interesting. For instance, it would be useful to hear from people in industry whose work consists of a combination of development and research. I know little about the existence of such jobs and would be very interested in knowing more about them. At several PLMW’s I remember a student asking about what happens if somebody decides to quit their PhD, and the only thing that was said was that quitting is okay and that they can make more money in industry.
It would be also great to give students a realistic depiction of the academic job market so that they can make an informed decision about whether it is worth to pursue an academic career. Some students may want to do a PhD regardless of whether they will get a professor job in the end, but given that at least half of PhD students want to stay in academia after graduating, it would be great if they were presented with an honest portrayal of the low chances of becoming a professor.
It might be difficult for a PL researcher to talk about topics like work/life balance and time management, mental health, and other issues that arise in the PhD because these topics are simply outside their area of competence. We could, therefore, invite a counsellor, psychologist, or social worker who specializes in these topics and can give advice to the students.
For example, it could be an interactive problem-solving seminar in which students form in groups to discuss solutions to specific problems that commonly arise in the PhD; after that, the groups discuss their ideas under the guidance of the speaker, who is professionally trained to give advice on the issues of interest. Even if the problems that are discussed do not come up in everybody’s lives, this activity still indicates that human, non-technical problems are important, they need to be addressed, that the PL community encourages people to seek support for their problems, and to openly speak about them.
The second part focuses on grad skills and life-story talks.