Graduate schools in North America usually have a required number of courses that you need to pass. Afterwards, the school is usually pretty flexible at the student taking more courses (and it is often free of charge), but it is another story with advisor/PI/supervisor.

After I passed my courses, I have been either explicitly or implicitly reminded by my advisor that I should not take more courses. None of any other graduate students are taking courses either in some sort of tacit acknowledgement. Taking courses takes away time from research and performing badly could cause serious problems, so it is understandable.

However, I find that after a year or two of not taking any courses, I feel less mentally "sharp". And sometimes I say to myself "I wish I had taken information theory/blackhole physics/statistical methods!" because I think they could open up more research ideas and plus some subjects are just hard when studied on your own. Finally, taking courses open up more employment opportunities, either through TAship or working in industry. I get the feeling that some/many advisors in academia (such as mine) are fairly indifferent about transitioning from graduate school to industry and do not appreciate the importance of courses/technical skills to these future opportunities.

Has anyone been able to successfully navigate through this dilemma? There are pro and con to each side of the argument and I could't make a move in fear of jeopardizing my relationship with my advisor.

  • 5
    I say to myself "I wish I had taken information theory/blackhole physics/statistical methods!" because I think they could open up more research ideas Study them alone. That can be brutally hard, but you'll learn more and become more mentally "sharp".
    – user2768
    Commented Sep 23, 2020 at 7:26
  • 9
    taking courses open up more employment opportunities, is there evidence to support that?
    – user2768
    Commented Sep 23, 2020 at 7:27
  • 5
    @user2768 Yes, but sometimes guided studies of condensed material could be beneficial.
    – Fraïssé
    Commented Sep 23, 2020 at 7:28
  • 3
    @user2768 Yes. The algorithm and machine learning courses are always filled to their maximum capacity. It is well known that these courses are populated by industry driven students.\
    – Fraïssé
    Commented Sep 23, 2020 at 7:29
  • 9
    What are they going to do if you take a course? Fire you? Cut your funding? I would do it anyway and let them b1txh. I'm perfectly happy to quit my PhD program at any time though. One of the only reasons a PhD isn't a total economic loss is the free tuition, which means courses. Commented Sep 23, 2020 at 14:29

6 Answers 6


When I finished course work I found that what I really missed about them was discussing the ideas from the courses with fellow graduate students. What filled in the gap for me was an informal logic and foundations seminar that another grad student had created. We would meet about once a week at lunch time, discussing articles, books, and occasionally each other's research. It was a nice way to avoid becoming narrowly focused on your own research topic, learn about things of broader interest, and keep an active sense of community. At the same time, it wasn't very time-consuming so it was never a major distraction from our actual research. Perhaps something like that could work for you. If there isn't already something like that in your department, you could always take the initiative and organize one.

  • This is a great answer. An advanced PhD student is somewhere between a student and a professor; students take courses, professors do not, but professors still (ideally) learn new things, so try doing this the way that professors do. Commented Oct 8, 2020 at 17:00

Audit the courses that you like, i.e., take them without receiving a grade/credit for it. You might not even need to formally audit courses, most professors will let you informally audit them. But as mentioned by others, research takes priority.

  • 3
    I did this a few times, and it must be said that few people auditing a course "for fun" will be there by the end of the course. Meetings with the advisor and research always seem to get in the way. Commented Sep 24, 2020 at 6:21

You should talk to your advisor about why you want to take more courses, and discuss which one-or-two would contribute to your development as a PhD student.

Your advisor is correct in the sense that it is easy to spend too much time taking courses for interest, and you need to focus on your research. Also, to some extent you should be able to study the material without a teacher now.

On the other hand, your advisor may have a different idea from you about what you are aiming to achieve by doing a PhD. It is certainly possible that they do not think enough about being in a position to transition to industry.

Things that are not of some reasonable use to your studies you should do in your own time. On the other hand, at least in the UK it is recognised that some time (about 10 days per year) should reasonably be spent on career-building CPD.

  • 1
    What does CPD stand for? Career/professional development?
    – llama
    Commented Sep 23, 2020 at 19:25
  • 2
    @llama Usually "continuing professional development". Commented Sep 23, 2020 at 21:04

Assuming you are getting a PhD: Your advisor is correct. Taking courses is not the purpose of a PhD. A PhD is a research degree, and PhD students should do research. If you want to take courses, you should be enrolled in a course-based degree.

  • 33
    In my time, going to academia was about education. I find the formulation "Taking courses is not the purpose of a PhD" a bit misleading. I think the learning process into breadth does not stop with entering a PhD programme. Perhaps better to say "Taking credit-relevant courses is not the purpose of a PhD". I agree with below answer that auditing may be the right way to approach this situation. Commented Sep 23, 2020 at 8:39
  • 5
    @CaptainEmacs Learning is certainly an important part of a PhD, but course based learning is not important to PhD programs. Doing research is the primary vehicle for learning. In many countries, PhD students take zero courses. Commented Sep 23, 2020 at 10:10
  • 6
    "In many countries" - such as? In Germany, some time ago, it was perfectly normal for PhD students to attend advanced courses and seminars. I have no idea how it changed after the transition to the BSc/MSc system (Bologna). Doing research is "learning in depth". Going to courses/seminars is "learning in breadth" and I believe education consists of both. Commented Sep 23, 2020 at 10:16
  • 6
    Certainly, when I took my Ph.D. (1999-2004, in the UK), "zero courses" was the norm. But that had changed dramatically by 2008; ISTR the reason it changed was because government got on the universities' backs about the (perceived) lack of transferable skills of Ph.D. graduates, although I can't find any documentary evidence of that now. Commented Sep 23, 2020 at 10:30
  • 3
    At some point in one's research career, you have to be able to teach yourself new things. And, the only way to learn how to do that is, well, by teaching yourself. And during your PhD is the time to do that because you have more time and fewer expectations that you will be up to speed by next week. So, I fully agree. Time to start learning how to learn on one's own.
    – Jon Custer
    Commented Sep 23, 2020 at 12:41

After I passed my qualifying exams and passed enough classes to get over that hurdle, I took additional courses pass/fail or credit/no-credit (with the consent of the instructor) where it was clear I could pass or get credit without it impacting my workload too much. I took a lot of courses this way and learned a lot. My advisor encouraged it. It also helped him out if the course he was teaching was shy a few people, he could ask me to register, too, if I hadn't taken it for credit already. I could have taken courses outside my area, but I mostly doubled down on topics that would reinforce my knowledge and understanding of my research: mathematical underpinnings, practical underpinnings, etc. It helped make sure that I came into campus every day and did work, went to meetings, and stopped by the office/lab.

I usually didn't take more than one course per semester this way in order to try to keep my workload sane and to be sure I was making progress on my research with the rest of my time.


If you want make an value argument that appeals to the research university's historical context and you have "enlightenment ideals", maybe appeal to Lehr und Lernfreiheit.

Probably someone could cite Weber / Foucault for evidence that enlightenment ideals are "problematic". Indeed if your advisor is hip they might also think that Lehr und Lernfreiheit is bullshit.

Regardless, I would like to see more Ph.D. students (in scientific fields in N. America who are being advised against taking courses in favor of doing research) ease up and just:

  1. read along with courses they're into,

  2. interrogate folks enrolled in courses they're into about the content of lectures they maybe had to miss (due to research obligations, administrivia, CS dept auditing policies, etc.);

  3. just be present for courses they're into (usually oral consent from the lecturer will do, he/she might even be flattered, and won't even have to worry about assessing/grading your work!);

  4. be honest about what they're doing (if it's safe) with their advisor (e.g., misbehave and own it) and their peers (invite them to audit with you);

  5. go rogue (if it's safe) or just formally withdraw from the program to do what they want (e.g., doing research or building things or learning, but without any cognitive dissonance owing to the fact that "they are scared they should be ostracized from the group");

  6. own the privilege of their choices and represent them to other Ph.D. students (e.g., if someone really needs statistical methods, they need statistical methods!).

  • I’m pretty sure you don’t mean “castigate” there. Maybe “interrogate”, but even that’s a bit harsh for the situation.
    – RLH
    Commented Sep 25, 2020 at 18:03

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .