To view this post in german click here/ Für die deutsche Version hier klicken:
In the UK PhD theses are monographs. So back in 2012, when I wrote my thesis I produced roughly 300 pages on what I had done over the last three years. Four result chapters. Introduction, methods and discussion. But theses do not all look like that. In Scandinavia for example, students write cumulative theses – which basically means they staple their papers back to back together.
In July 2016 Nature featured two pieces (1, 2) on the subject – discussing if the old-fashioned way of writing a monograph should be updated. Papers are the currency of science. While having a PhD is a must for most post-PhD jobs, having published (preferably in a high impact journal) is what will get you a job. So shouldn’t students save their time by focusing on papers rather than a thesis monograph?
Career-wise this sounds like the right thing. You need papers to move ahead but this approach makes a few assumptions that I personally find troubling.
Not all work PhD students do is publishable. At least not in a decent time frame. I know students who have worked hard, but their projects still didn’t mount publishable data. Others got scooped and had a hard time publishing the “not so new anymore” data. And some PIs simply do not publish under a certain impact factor – leaving students to either not publish or stay for longer than the initially agreed 3-4 years*.
Even if you have a story you want to publish, it’s a slow process. Publishing can take a long time – months or even years. I know few examples, where students published (their main project) within their PhD. Lots have papers come out after they graduated. My main paper was published nearly one year after I graduated. Time limits and the pressure to publish something in order to graduate could either led to students being students much longer or to the publication of smaller stories – which in the longer run is less ideal for their career.
Pressure to publish in a given time frame may also change the type of science students do. Fewer students will be willing to investigate exciting but risky projects. If a new method won’t work and the project fails, there won’t be a paper to graduate with. Journals limit the amount of detail the authors can give. They only show a few experiments that mounted to a good story. Students will be forced to think paper rather than investigate creatively. Although it’s debatable how many students, writing monographs do not think along the same lines. I believe that your PhD is the time to test yourself – try different things. Later you will always have to deliver on grants and publications. Sure having the pressure to publish may prepare you for that step, but I guess it’s a matter of definition. For me a student should be allowed to experiment, and try and find their way.
In addition to the lab site, papers also limit word count and, thereby, the students space to show their trouble shooting, creative thinking ability. Which I think is key in a thesis – demonstrating your ability to communicate science. Framing cumulative papers with an overall introduction and discussion can somewhat decrease this issue and allow the students to get creative.
Another assumption for cumulative thesis is that students actually write their papers. Sure most write the first drafts, but I know examples where the PI writes the papers and all students get to write is their thesis. Final papers are heavily edited, so what examiners get to evaluate it not necessary the students’ ability to communicate their work.
Science relies more and more on collaborative effort – so sole authorship is utopia in our day and age. So how do you access the contributions of the student? Some countries stipulate a certain number of first author papers, assuming that the first author did most of the lab work and wrote the paper. But what about shared first authorships, where two labs collaborated and the first authors contributed equally? I know people who struggled with that, given that shared first-authorship did not count towards the publications needed to graduate.
But the main point the argument for cumulative thesis misses, is the fact that most monographic thesis are also based on a paper or paper draft – at least as far as my experience goes. In the end papers are what scientists are after, so most will try their hardest to have a paper or draft by the end of their thesis. Not simply submitting a paper but extending an existing paper draft to a monographic thesis, in my opinion, allows students to go into greater detail compared to the limited space of a paper. Looking back my thesis was excessive, so a more concise format is preferable and indeed most thesis I have seen in Germany are 100-150 pages long.
Some people like to argue that re-writing, extending papers to a thesis is extra work. That time could be used better. All four chapters of my thesis have been published. Three of them well after my PhD defence, the chapters were drafts for all of them. So I do not buy into the extra work argument. In a great read, Nature asked three scientists to look back into their thesis and reminisce about their PhDs. I totally understand the notion of achievement when you have the complete thesis in your hand. A piece of work – much bigger than a single paper, which is only part of the story – that you are incredibly proud of. The manifestation of blood, sweat and tears. But that might just be me. A person who always enjoyed writing. Maybe the matter of thesis is simply not a one fit all approach and we should allow the students more freedom in how to create their thesis.
*talking about the European standard here, as it is what I know.