Tuesday, March 12, 2013

Plagiarism, humanities and sciences

My home country has recently seen a distressing profusion of cases of plagiarism among politicians with academic degrees. It all started, as far as I know, with then defense minister Guttenberg who from what has been reported appears to have plagiarized most of his dissertation. What was new about the case was that the dissertation was, after the first few discoveries by academics, thoroughly and publicly examined by numerous contributors to a wiki ("GuttenPlag") in a crowdsourcing approach. From there on, activists started to examine the dissertations of other high-ranking politicians with doctorates, and soon several more cases of plagiarism were found (and one person was charged but then exonerated).

There are two interesting patterns here. First, those politicians are generally conservatives or liberals* as opposed to social democrats or greens. I can only speculate why that is so. An obvious suspicion that has of course been advanced in discussion threads on the internet is that the anti-plagiarism activists selectively target right wing politicians. Said activists seem keenly aware of that charge and take pains to stress that their work is not party-political.

At least part of the answer might be that conservative and liberal politicians are simply more likely to have doctorate degrees. Here is a statistic for the German federal parliament listing the number of doctorates the members from each party have, and the greens have the lowest proportion, followed by the social democrats. This is probably not because academics are more likely to be conservatives - it is actually the German green party that has the most educated supporters. But there might be a cultural issue here about conservatives placing greater value on official titles. Surely a career politician does not actually need a doctorate to be successful in their chosen career, so the only motivation to obtain one that makes sense to me would be the prestige associated with it. But I will be the first to admit that this is speculation.

The second interesting pattern is that the dissertations in question are generally from the humanities as opposed to the natural sciences. Karl-Theodor zu Guttenberg obtained a doctorate in jurisprudence, Annette Schavan in theology and philosophy, Silvana Koch-Mehrin in macroeconomics and history, Margarita Mathiopoulos on what sounds like a historical topic (?), Jorgo Chaitzimarkakis in political science...

It appears logical to assume that politicians would on average have a preference for the humanities, especially sociology, law, politics and history, but not in all cases. The current German chancellor, Angela Merkel, is famously a physicist. So there are natural scientists in politics, but still so far it appears that plagiarism may not be such an issue among them.

And if you follow the news, and look beyond German politicians and at actual career scientists worldwide, you will notice that plagiarism is simply not a relevant issue in the sciences. Of course there is some fraud, but it isn't plagiarism; the frauds committed by scientists are usually (1) inventing data and (2) manipulating data to produce convenient or spectacular "results". The reason is surely down to a fundamental difference between what scholars in the humanities and natural scientists do.

I am aware how the following must sound to a defensive humanities scholar, so I hasten to add that I do not want to appear guilty of scientism. I make no claim that the humanities are useless, and no claim that only empirical science is able to produce knowledge. For example, I do not believe that ethics are best or only understood by doing neurobiology, evo-psych or something like that.

But surely the point of a dissertation, the thesis one has to write to be awarded a doctorate, is that it should constitute some actual progress in the relevant field. Being able to compile a dissertation is supposed to show that you can generate, and this cannot be stressed enough, significant new knowledge, because that is kinda the whole point of working as a scholar or researcher after obtaining the doctorate.

Of course, with the massive number of people obtaining doctorates today, a non-academic may wonder how that is still possible. Clearly not everybody produces an advance of the same significance, and one can have a nice discussion about whether this or that dissertation in the natural sciences is such a big deal. Still, large or small, it is at least trivially easy to find research projects that will generate new knowledge in the sciences. There are myriads of species whose ecology and evolutionary relationships we still don't know, and there are even hundreds of thousands if not millions of species still to be described. There are myriads of biochemical and developmental pathways to be unraveled, myriads of uncharacterized molecules, many issues that are still unclear in the realms of astrophysics, fluid dynamics, medicine, whatever you want.

In some parts of the humanities, I see how the same would be true, especially in sociology, economics, history and language studies, all in their widest possible circumscriptions. All these areas are so complex that we would surely not run out of research questions for the next few hundred years, no matter how many dissertations will be completed**. But there are two issues: first, there are other areas in the humanities where it is not actually clear to me how one can meaningfully talk about knowledge at all, and, second, even in the aforementioned areas where one can, not all grad students appear to work in a way that is conductive to generating knowledge.

Expanding on the first issue, to produce knowledge one has to have an objective and reproducible way of (tentatively) distinguishing between what is (most likely to be) true and what is (most likely to be) untrue. I can see how that works for inferring the evolution of a plant group, for testing the efficacy of an antibiotic, and for characterizing a distant star system. To move to the humanities, I can see how that works for examining the effect of a higher sales tax on consumption, for studying the grammar of the Bantu languages, and for laying bare internal contradictions in a philosophical argument. It is much less clear, to say the least, how it is possible to generate knowledge in theology, for example; there is simply nothing in that area to have knowledge about, quite apart from the complete lack of objective criteria to distinguish between correct and false in theology which is best demonstrated by the fact that there are different theologies for every religion.

As for the second issue, the prime example are of course postmodernists. Under a charitable interpretation they basically advance the claim that we cannot know anything anyway, and that everything is just ideology and opinion, although in a perfect example of intellectual inconsistency they privately do rely on their own perceptions when crossing the road and on the constancy of natural laws when boarding an airplane. Under a less charitable interpretation they are mere frauds, writing tome after tome full of convoluted jargon that does not contain any meaningful thought whatsoever.

But we do not have to go that far; the point I want to make is not that there are entire schools of academia that are not dedicated to generating new knowledge, but that even in parts of the humanities that are dedicated to this pursuit there are many dissertations whose contribution to human knowledge is a bit dubious - not because their topics are too unimportant, but simply because it is, given the way the topic is approached, not quite clear in what way even the tiniest smidgen of new knowledge could possibly be produced in the first place.

What I am thinking of here is the following approach: the student reads fifty books and then sits down and writes the fifty-first book on the same topic, calls it a dissertation and gets a doctorate. I grant that a great philosopher could sometimes be able to come up with a completely novel insight after that exercise, but thousands of students in jurisprudence, history, political science, macroeconomics, etc., every year? No way. There is just no way that most of the dissertations produced in such a manner can be anything more than saying the same old things again in other words.

And without mentioning names, I must say that merely from reading the titles and a few excerpts from dissertations that are now under fire for plagiarism, at least some of them seem to be examples of this approach: Topics where the title of the thesis already indicates that you should not expect it to contain any data or analysis but merely blah blah blah, I read this here and this there, this person argues so and this person defines it in that way. The challenge is then not to find out something new, but merely (1) to rearrange the words you read in other peoples' books into a new order that will not produce a hit on Google or sound too familiar to a colleague and/or (2) to pepper the text with enough quotation marks and footnotes to document your sources where you are too lazy to rearrange their words and copy cite them directly. This little bit of effort then makes the difference between plagiarism and a legitimate dissertation, it would seem.

In natural science terms, these dissertations would be book-length review articles. Yes, in science we do read-fifty-papers-then-write-a-new-one too, to synthesize the state of knowledge in a field. These review articles often have a lot of influence, are widely cited and may point the way for future studies. But nobody in science would claim that review articles, in and of themselves, constitute any progress. After all, all that knowledge was already there. And consequently, because a dissertation should constitute progress and generate new knowledge, it should be fairly hard to obtain a doctorate in the natural sciences merely by writing three review articles and calling them your thesis, unless perhaps at a diploma mill.

This is why plagiarism is not really an issue in the sciences. What really counts is not how you rearrange some words but whether you have new data and whether you can use them to support your hypothesis, answer your question or describe something previously unknown.

So I guess what I want to say in this extremely roundabout way is this: Maybe certain faculties in the humanities may want to reconsider what they consider an acceptable dissertation. How about only accepting works that actually find out something new? Not only would that be much more productive, but if a grad student is doing something that nobody has done before it is also much harder for them to plagiarize in the first place. As a welcome side effect, it would perhaps reduce the number of people aiming for a vanity doctorate who don't actually need it. And if it has the side effect of exposing the fact that this or that subfield of the humanities simply does not have anything to find out, lacks a methodology for generating knowledge, and thus cannot ever produce anything deserving the name dissertation, I personally would not consider that a downside either.


*) For the benefit of North American readers I should add that in basically every country on this planet except the USA and Canada the term "liberal" refers to what would probably be called moderately right wing libertarian in those two countries. The German liberal party FDP is slightly to the right of the conservative CDU/CSU in economic matters and to its left in matters of personal liberty vs state authority and cultural conformism. Currently, the two parties constitute the moderate proportion of the right wing of the political spectrum and are each others preferred coalition partners, mirrored by a similar preference of social democrats and greens for each other on the left half of the political spectrum.

**)  Whether we will ever produce that many more dissertations in the face of looming societal collapse from overpopulation, climate change and resource limits is another question.


  1. A Masters thesis is similar to a PhD dissertation in terms of goals. However a good project is more difficult to find, because the amount of time, effort, and resources put into a MS thesis ought to be much less than that put into a PhD dissertation. I suppose one could call a MS thesis a dissertation light.

  2. In my area that is true at least for those students who are interested in pursuing a career in science later. In those cases, master / honours / Diplom theses are usually designed to result in one paper while dissertations are designed to result in at least three of them.

    However, I still think that one can get an MS or equivalent with much less of an actual contribution to science than a third of a dissertation. Many theses are never formally published, some are merely a small cog in the project of a PhD student or postdoc who supervises the MS candidate, and so on.

    In Germany the situation was at least during the time of my own studies that the Diplom student did not even have the right to decide what was going to happen to the results - they belonged to the supervisor because nothing before the PhD project was considered to be your own, independent work.