I have finished 5 years of my PhD, and I have 5 publications in top tier (A* computer science) conferences based on work done during PhD. However, most of the publications are in unrelated areas. I am finding it almost impossible to compile a thesis from all my publications.

All my publications have me as first author and my advisor as second author. My advisor hasn't read any of the publications, and he is incapable of judging them even if he reads them. I know that the work I've done can't be compiled into a thesis. Due to the completion of PhD duration, I have stopped receiving scholarship from my Institute.

Even without my PhD degree, I can easily get a high paying job. My other option is to compile my work in the form of a thesis and hope that the thesis reviewers accept it.

What should I choose?

Update I have got several brilliant answers and I wanted to accept many of them. I have decided to complete my PhD without antagonizing my advisor. He has agreed to grant me permission for submitting my PhD thesis, if my two papers, that I submitted recently, get published. If they indeed get published, I will have 4 related papers, and hence, I will be able to compile a thesis. I won't be able to add my other papers, which doesn't really matter. I told him that I care about my PhD and not the best thesis award, but he is hell bent upon it. He has agreed to provide me a project assistantship salary as long as I am pursuing PhD.

Despite what @einpoklum said, it doesn't seem fair to leave without a PhD after working for so many years. Thanks everyone for the help.

  • 22
    You've come this far, don't give up now! – astronat Aug 19 '17 at 7:27
  • 18
    What exactly is the problem? Why do you think that 5 publications in top-tier conferences are not enough for a thesis? It is quite common that the subject of a thesis is not exactly the same as the initial plan. – Louic Aug 19 '17 at 9:06
  • 4
    I've seen recently a few thesis with title "Essays in macroeconomics" and I suspect they are just a compendium of unrelated papers. However, feasibility of this kind of thesis may depend on university, department, field or country. – Pere Aug 19 '17 at 12:17
  • 50
    ...my advisor as second author. My advisor hasn't read any of the publications... — I'm sorry, WHAT? How can anyone who hasn't even read a paper be a co-author? Even if you work in a field where advisors are coauthors by default, this is extremely unprofessional, if not unethical. – JeffE Aug 19 '17 at 14:37
  • 4
    @user2808118 Your advisor has no idea what his students are doing, and almost 80% of them left without successfully completing their degree? WOW. – Michael Aug 19 '17 at 22:26

If it is impossible to compile a thesis from all your publications, I suggest focusing on lesser publications in the thesis and if it is possible to expand your work, focusing only in one direction. Since your work is already published, there is no reason why it wouldn't be accepted by the reviewers. The content of the thesis counts less in applying to post-doctoral jobs, what is important is your papers. Of course, the thesis must be well written, rigorously, etc. There are many people who receive PhD diploma with few publications, or even no publications at all.

  • 3
    While this is a reasonable suggestion in itself, I believe it's making too much of an implicit assumption about OP's situation. Now, s/he did accept, but in my experience (incl. in public office representing graduate students) for many people this course of action will fail. – einpoklum Aug 19 '17 at 20:45
  • 2
    I agree with @einpoklum. Having 5 papers isn't necessarily helpful for advancing in academia if the research areas are so disconnected that the 5 papers collectively show a lack of focus and commitment to a research topic. – Slow loris Aug 19 '17 at 22:34
  • I was able to connect 2 (out of 5) of my publications. Two more are down the lane, though still unpublished. I guess I can compile the 4 of them into a thesis. Anyway, I am looking for an industry job which is easily available in my area. – user2808118 Aug 20 '17 at 3:41
  • 1
    I wonder how the OP decided that they are A* conferences? because CS A* are very very competitive and one publication that solves an original problem can in many cases be enough for a Phd. – user6875880 Aug 20 '17 at 9:00

I really have to disagree with everyone here telling you "don't quit" as though it would somehow be a failure, and a Ph.D. is the thing to have, etc. I would say:

1. It depends on why you've gotten to this state of affairs

Having invested time and effort in 5 different and barely-related directions over 5 years means something is seriously wrong. Probably not wrong with your capabilities though - since you have been able to obtain useful novel results. There's some kind of issue with one or more of the following:

  • Your motivation: How come you haven't chosen to pursue any of these subjects further and more deeply.
  • The relation with your advisor: How come your advisor was essentially completely uninvolved in those papers? Do you even have any sort of interaction with him?
  • Your advisor's motivation: How come your advisor did not want you, or you and him together, to focus any any of the 5 subjects? Or on another one? Hasn't he solicited you, advised you or told you to do certain work that he would be familiar with?
  • The research proposal / plan which you presented your candidacy with: Did you originally decide to just jot something down you don't really care about or your advisor doesn't know about? If so, why was it accepted? If not, why didn't you sticking to the plan? Or - after 2 or 3 years, why didn't you sit down and devise a new plan?
  • Potential collaborators: Why are you working alone? It seems the community, or several ones, are interested in your work enough to have it in conferences. So, have you found nobody else to work with? Have you looked?

I can really not give you a serious answer without some or most of this information. But I will say that for some of the possible answers to this question, you should definitely consider:

  • Taking an official leave from your Ph.D., to figure out what you want or would be passionate about.
  • Adding a co-advisor, or even switching advisors, to someone who would be able to help you finish by focusing on a subject you've already started work in.
  • Quitting amicably, if there isn't clear direction or subject you want to focus on and get more results, and if what you'd really like to do is go work in industry somewhere.

2. That's not the right question

Getting a Ph.D. is a side-effect of performing a body of significant research on a focused subject. Do you want to do that, still, and (do you believe you) are capable of doing it? ... Answer this question, in light of what I've said above, and your own question resolves itself. Again, I can't venture an answer, since you've not provided enough information.

  • 2
    That last paragraph should be engraved in stone and prominently asserted in many, many universities. – bishop Aug 19 '17 at 19:01
  • 2
    @bishop except that it's so often completely disconnected from reality that I'd hesitate to fool students into thinking that all PhD's work that way. – DanielSank Aug 19 '17 at 19:27
  • 4
    @DanielSank: Not sure I understand what you mean. Could you elaborate? – einpoklum Aug 19 '17 at 20:43
  • 1
    Motivation: The problem is, I can't think of how to extend ideas. I get ideas while reading lots of papers in many different fields. Most of the ideas don't work. Some of them work, and I publish papers out of them. However, I can never figure out how to extend them. I have told my advisor that I am not PhD material several times, but all he sees is a bunch of papers. I am good at coming up with publishable ideas, but not at focussing on a particular topic. – user2808118 Aug 20 '17 at 4:08
  • 3
    @user2808118: My friend, you've been sorely mis-advised. But this is also your fault for spending 5 years without realizing it. I still don't quite know how this happened, and how your research plan / subject got written and effectively put aside, but - please talk to some faculty members whose work has more to do with yours about your situation, discreetly; as well as with whoever's in charge of graduate studies. – einpoklum Aug 20 '17 at 7:13

No don’t quit, five A* papers is more than most theses ever achieve already, good work!

Sounds like you just need to come up with a vaguely convincing story to knit them all together; then find some friendly examiners. There must be some bizarre application area that would combine most topics somehow... say you have published on category theory, the social science of Hadoop uptake, and tractor design; then you just need to invent some niche application about farmers needing a more robust language and system to handle their agricultural data, write 20 pages of waffle about it as an intro chapter, staple that on the front of the papers, and you should be good to go.

  • 5
    If you post your specific topics on here then maybe people will have fun making such application area suggestions ... – People may have fun, but this is not something this site is suited for. – Wrzlprmft Aug 19 '17 at 17:08
  • @Wrzlprmft As a farmer needing a more robust language and system to handle my agricultural data, I disagree. – Michael Aug 19 '17 at 22:28
  • 2
    You're disagreeing with a fundamental aspect of Stack Exchange. Q&A, not Q&D, not Q&Q&A&D. – Nij Aug 20 '17 at 2:49

What's confusing about your question is that having 5 papers in top-tier CS conferences is a bigger achievement than a PhD (in fact much bigger, in my opinion). So the idea that you should quit your PhD because you did something much more impressive sounds rather strange, and suggests some kind of potentially serious problem with your advisor and/or your thesis committee or department.

In particular, you say your adviser only wants students to graduate with best student thesis awards, but that as a result only 2 out of 9 graduated?! That sounds like an extremely counterproductive and self-sabotaging (and sadly, also sabotaging to other people's careers) philosophy to have.

I will keep it relatively short since I think analyzing in detail what has gone wrong and how to fix it is beyond the scope of this question (but feel free to ask a new one adding some details, there is certainly room for more discussion about this). The short answer to your question is: definitely don't quit your PhD because your papers are on unrelated topics, let alone because you "can't come up with a suitable title". You have achieved something much greater that is worth more than a PhD, and, assuming the issues with your adviser are resolved successfully, should be well on track to finishing what sounds like it will be a pretty stellar thesis with only a short amount of additional work. Good luck!

  • Thanks. As of now, I am compiling a thesis out of 2 of my latest publications. 2 more works are down the lane, that are somewhat related. The thesis won't be stellar, since most of the publications won't feature in the thesis. – user2808118 Aug 20 '17 at 4:28
  • 2 out of 9 graduated not because of his insistence on best thesis award, but due to his lack of understanding about the problems they were working on. They ended up doing nothing for 3 years (or 5 years) and left thereafter. – user2808118 Aug 20 '17 at 4:46

Talk to the dean of graduate studies or other appropriate department administrator. The main problem here is an advisor with rigid expectations who won't read your work. This is not a good match for a student who's been doing interdisciplinary work that doesn't fit the mold.

One possible desirable outcome of this conversation: you get paired up with a different advisor who reads the work you've done so far, and brainstorms WITH YOU to find a common thread running throughout the bulk of your work. Together, hopefully you can come up with ways of packaging your work and an outline for what remains to be done to tie the parts together in a finished whole.

If you can't find someone capable of functioning as your official advisor, at least try to find someone to mentor you informally to accomplish what I described.

Note inspired by the helpful answer written by Dan Romik, who I believe is an administrator in a CS department.

You may be asked by department administration to try to stick it out with your advisor and work through your difficulties. It might be tempting to say, "It's impossible, I've already tried all the things you've suggested," but I suggest that you do your best to cooperate, no matter how pessimistic you might feel, and agree to try again, but with a specific time frame defined. For example, you and the administrator might agree to meet again in three weeks. During that period, try your best to move forward with your advisor, with a positive attitude, letting bygones be bygones.

If there are still insurmountable problems, go back to the administrator after the specified interval and report back, calmly and accurately.

  • I want to talk to the chairman, but I don't want to antagonize my advisor. I wish there was an anonymous way of discussing my problem with my chairman. My advisor has agreed to accept my PhD, if I can get two more papers. He still insists on getting a best thesis award. It is incredibly weird. – user2808118 Aug 20 '17 at 4:41
  • 2
    I suggest you collect the various paragraphs you've written here and put them together in one document, then share them with the administrator when you meet. If it's hard to start talking, just hand your sheet(s) over). // I don't see what you have to lose. In the worst case (which I can't imagine realistically happening), and the conversation with the administrator makes things worse, how much worse would things actually be compared to how things are now? And I would think that your back-up plan would give you peace of mind. If the PhD doesn't work out, you do have an exit strategy. – aparente001 Aug 20 '17 at 5:28

Show your ability to be a "PhD" by writing the dissertation based on the work you have done over the past 5 years. Be creative. Tie it all together: you alone know how this fabric can be stitched up. Defend it to the faculty at a public invited forum. Finish. Based on your post, I am certain you can do it!

  • This is concise and a great motivation. – Coder Aug 19 '17 at 20:31
  • Thanks. This is exactly what I am doing. Although, I am compiling a subset of the papers. – user2808118 Aug 20 '17 at 4:42

Is it your career goal to enter into academia? If so, finish the Ph.D by hook or by crook. You will not be able to roll the dice into an academic position without one. Others in this thread have better advice on how to do this.

Is your career goal to enter into the private sector? Take into account that every year you spend trying to finish your thesis is $XXX,XXX dollars. If I told you I'd give you [insert what your starting salary could be] to quit and get a job, get some benefits and being piping some of that sweet, sweet software lucre into your bank account, would you take the deal?

Will the job you are shooting for absolutely require a Ph.D? The industry is notorious for hiring drop outs of all levels, but there are still some jobs that will require the letters.

Lastly, if you could put yourself in your own shoes ten years later, would you regret not finishing it. Is that regret is worth $XXX,XXX?

  • I don't think that I will regret leaving PhD. – user2808118 Aug 20 '17 at 4:43
  • By hook, perhaps. But not by crook, no... =) – user21820 Aug 20 '17 at 6:23

There is already an accepted answer, but I'd still like to analyze an issue presented by the OP.

You wrote:

The problem is, I can't think of how to extend ideas. I get ideas while reading lots of papers in many different fields. Most of the ideas don't work. Some of them work, and I publish papers out of them. However, I can never figure out how to extend them. [...] I am good at coming up with publishable ideas, but not at focussing on a particular topic. [...]

[M]ost of the publications are in unrelated areas. I am finding it almost impossible to compile a thesis from all my publications.

At least in my area and in my experience, most researchers who work on several diverse problems have a toolbag (smaller than one may think) of techniques and ideas that they suitably modify and apply even to seemingly different issues.

To sum up, usually one does not make good papers or publishable ideas out of thin air or use wildly different methods to attack wildly different problems (especially at the beginning of one's career).

To put your papers together, think about the common techniques and tricks that you use in your work.

Not the answer you're looking for? Browse other questions tagged or ask your own question.