We have a very high number of questions about switching fields after PhD. Most of these are duplicates except for the fields in question. Here are some examples (by no means a complete list, I just got tired of copying and pasting):

Since there are a lot of combinations between current field and target field, we could potentially have thousands of versions of this same question. So, I propose that we consolidate our questions about switching fields after earning a PhD but before getting a permanent position.

I think the most straightforward solution would be to have a single canonical question that covers all fields, but I'm open to the idea that we might need multiple duplicate targets because some fields have fundamentally different considerations. Thoughts?

Update: A candidate for this canonical question now exists here.

  • 3
    There are at least two kinds of questions: changing between similar fields (e.g. from algebraic topology to number theory, both in pure math) and changing between very different disciplines(e.g. from computer science to humanities). Are we going to have one CW to cover it all, or separate question to cover them in less general sense (a tough question here, how to separate them? by answers? like the question (What does it mean by first author?)
    – Nobody
    Commented Jul 24, 2022 at 7:01
  • That is one of the topics to discuss here. Since the boundaries are not well-defined (how do you define "similar"), my inclination is to have one canonical question that discusses both cases. But it's possible that we need several different questions: intra-field, STEM to STEM, STEM to humanities, humanities to humanities, etc.; if this is the case, the right thing to do is probably to merge some of the existing questions so that we have clear duplicate targets going forward.
    – cag51 Mod
    Commented Jul 24, 2022 at 7:17
  • What should the main focus of the canonical question be? “Can I switch fields?”, “How do I switch fields?”, “Is it a good idea to switch fields?”, “When to switch fields?”, or all of these?
    – Wrzlprmft Mod
    Commented Jul 25, 2022 at 8:01
  • If this is done, it should focus, clearly, on early career changes as do the questions you highlight. A complication here is that changing for disinterest in the doctoral field is quite different from forced changes due to economic factors (as I had to do).
    – Buffy
    Commented Jul 25, 2022 at 14:56
  • I also fear that many askers are looking for a seamless and easy way to transition.
    – Buffy
    Commented Jul 25, 2022 at 14:59
  • I imagine the title would be something like: "(How) Can I switch from field X to field Y after getting my PhD?" And agree this should be limited to early career (i.e., between getting a PhD and getting tenure).
    – cag51 Mod
    Commented Jul 25, 2022 at 16:45
  • I personally got a PhD in Electrical Engineering, specifically DSP. For a few years, I did real-time embedded audio programming, which sort of double-counted as EE and software engineering. Now I do exclusively software engineering. How this would apply to various other people would be dependent on their interests, but it shows that it can be done, and to that extent, is a 'positive story'. Commented Jul 26, 2022 at 14:34
  • BTW: By the time I started doing full time software engineering, I was 36 plus years old, which I would call mid-career, not early-career. Commented Jul 26, 2022 at 14:35
  • I see that the question has been created and tagged with a new tag canonical-question. I've tagged three other questions with this tag just now and may do more on future days. The tag is there for anyone to use, which is perhaps a mistake. It would be preferable if it were available only to mods IMO. I also created the tag wiki, which should be reviewed.
    – Buffy
    Commented Aug 2, 2022 at 11:48
  • @Buffy - the new tag is being discussed here (it has not yet been "approved" and was probably created prematurely, but we'll leave it pending the resolution of that thread).
    – cag51 Mod
    Commented Aug 2, 2022 at 12:37

2 Answers 2


My suggestion: Yes, but with some type of PSA in the question regarding how to use it as a duplicate target. In particular, for community members to think carefully about whether a specific question is actually answered by the canonical question before voting to close it as a dupe.

I think this is a fair canonical question but a bit fraught. As a matter of format, I agree that it makes sense that a canonical question would aggregate common considerations in a few broad categories. Just as a rather cursory suggestion agreeing with a previous comment: Within-discipline, within-nearby-disciplines (e.g. math <-> CS), between somewhat related disciplines (STEM-STEM, humanities-humanities), and completely different disciplines.

My major concern about whether this is advisable is more about how our community will make use of the duplicate target. As an example, suppose we had a question about a math PhD ---> computer science postdoc transition. There could be some useful specific answer depending on the question that goes beyond general advice.

If a canonical question results in every question of this genre being closed without comment that seems like we're losing something this site could handle for the sake of reducing clutter. In an ideal world we'd have the canonical question. But any potential duplicates would be left open long enough for community members with expertise in the specific fields to consider whether they think the canonical answers cover the specific question.

  • Thanks for reply! A few reactions: (1) Most answers to the historical questions have not provided field-specific advice, even when such was requested. So it does not seem that we risk losing something valuable. (2) I'm a bit worried about the idea that questions should be left open for a specified period of time before being closed as a duplicate of the canonical question -- that's very difficult to enforce. (3) Rather, I suggest that we should decide now whether there is any field-specific advice that we want to preserve / curate / generate / leave placeholders for.
    – cag51 Mod
    Commented Jul 24, 2022 at 19:45
  • (1) This is a fair observation. I'm generally a bit concerned about trying to strike the right balance between good curation and accessibility. I myself have occasionally fallen into reflexively flagging to close a question that is "close enough" to a canonical duplicate. (2) I agree that asking for a specific period of time is probably not smart. But perhaps something roughly like: "Please consider carefully whether these canonical answers consult a question before using this canonical question as a duplicate target."
    – user137975
    Commented Jul 25, 2022 at 0:36
  • 2
    I like that! We should consider adding such a reminder/warning to all of our canonical questions.
    – cag51 Mod
    Commented Jul 25, 2022 at 4:56
  • 1
    I might go beyond the second paragraph and say that every academic field is so radically different, that many practitioners of said fields would say that advice is nearly useless unless it's tailored to their field. This argument can certainly be made for physics and mathematics, where changing subfield can feel like trying to move a mountain. I think this is a much, much larger undertaking than just having a few broad subcategories in the canonical question.
    – Jerome
    Commented Jul 26, 2022 at 22:14
  • @AtomJZ I think I disagree somewhat. Cag has a point that we get quite a few very general, reasonably low-quality questions in this genre for which general advice could be sufficient. As long as a canonical question handles those without the community becoming overzealous on closing unhandled questions, the canonical question seems like a decent cost/benefit tradeoff.
    – user137975
    Commented Jul 27, 2022 at 21:43
  • Yeah, I would agree with @AtomJZ's concern if we could point to (m)any examples of high-quality answers from the last 10 years that specifically address the nuances of the field pairing in question. But as it is, we are not generating those types of answers; we provide only the same general advice over and over (is my impression, at least). I'd be open to leaving placeholders for practitioners to add field-specific advice, but I suspect most of these placeholders would long remain empty.
    – cag51 Mod
    Commented Jul 27, 2022 at 22:51

Following up on the discussion in the comments,

Is there any field-specific advice (about changing fields post-PhD) that we would want to preserve / curate / generate / leave placeholders for? Do any fields have fundamentally different considerations than others?

  • If so, please leave a comment on this answer explaining which field needs handled separately and why it has different considerations. If this field-specific advice already exists in an answer somewhere, please link to it.
  • Otherwise, my assumption is that it's possible to write one answer that addresses most/all possible switches (the advice for intra-field switches vs. switching to a remote field might be different, but we can address both in one post).

You must log in to answer this question.

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