Whither the Literature Review? (updated)
Has the literature review in philosophy dissertations largely disappeared, and if so, what should we think of this change?
In a post reflecting on how younger scholars have engaged (or not) with his work, Eric Schliesser (Amsterdam) expresses his sense of futility about his inclusion in that “a fragile chain of collective learning” from past to future. Despite being widely cited, “[N]early all my papers seemed to have been incredibly elusive pebbles in a vast ocean of learning,” he says.
If it’s not just him (it’s not), then we might be curious about what contributed to this state of affairs. Schliesser offers one possible contributing factor:
The near disappearance of the literature review in the PhD stage has removed the nudge to read deeply into the literature.
There are probably other factors as well. One is simply the growth of the literature. Most people’s complaint is that they are not cited at all, not that they’re not cited in sufficiently informative ways. The vastness of the literature can help explain both: there’s no time to read most of everything to cite it, let alone time to read it so that one can argue with it in print.
Still, in smaller, slower-moving subfields, the growth of the literature will not be as much of an explanatory factor, and the lit review explanation may carry more weight.
Schliesser thinks there are reasons to be concerned about the “near disappearance” of lit reviews:
I think this also means that the fate of the engagement with each work is not the product of the sober, dispersed judgment of many separate scholars, but rather it is funneled through a small number of proxy-metrics, relative perceived status in the field, and the concentrated power of a few influential supervisors.
So there are some questions here.
- What is the current state of the literature review in the PhD stage in philosophy programs? Has your department made relatively recent policy changes about this, or have there been non-formalized changes over time?
- What explains the decline of the literature review in philosophy?
- What are the costs and benefits of the decline of the literature review?
- What role has been / will be played by technology like AI in regard to lit reviews?
- Is there a workable version of the literature review that should be revived? Should something replace it?
Discussion welcome.
UPDATE: I just put this in the Heap of Links, but it seemed worth mentioning here, too: Martin Lenz (FernUniversität in Hagen) has put out some advice on “how to search effectively for literature when doing philosophical research“. Discussion of his suggestions, or additions to them, are welcome.
One thing that I found helpful for my sub-discipline (phenomenology) was a recent volume (https://link.springer.com/book/10.1007/978-3-031-26074-2) dedicated to surveying and analyzing the vast amount of literature. Perhaps this mix of senior scholars and data analysis is a way forward.
Recently promoted full professor at an R2ish department here.
I don’t know how to do a lit review. My dissertation was a stapler dissertation. So I didn’t do one then. And I’ve never done one sense, nor felt like it would be helpful to do so. And none of the places I studied at or did postdocs at had policies requiring them in dissertations. Frankly, I was unaware that there were places that *did* have policies about such things.
Honestly when I hear complaints like these it’s a bit hard to take them seriously. “Why oh why aren’t people talking about *me*!” is just not the kind of lament that I’m ever going to find compelling.
Is a field where lit reviews don’t happen different from a field where they do? Absolutely. But philosophy just is a different field than it was fifty or a hundred years ago. We ask different questions, answer them with different tools, and do these things for different reasons than we used to. It’d be weird if the ways we engaged with each other’s work *wasn’t* different, given that.
Grad candidate here and I somewhat agree with this. I think the idea of a lit review as a step in one’s dissertation makes sense if the dissertation is akin to a monograph with a single theme, which I believe that still is the majority of dissertations in our area. However, many of the problems we discuss and the debates we engage became so specialized that it’s somewhat difficult to conceive someone writing a lit review on such a thin topic. Besides, the initial chapters in many of these monograph-style dissertations is theoretical apparatus being introduced, which may not be complete enough for lit review purposes.
Perhaps I’m mistaken, but I thought at least one point of writing a dissertation was to demonstrate mastery of a particular area. This requires intimate familiarity with the going positions in the area and how the logical space has been carved up. One way — and by my lights the best — to gain this mastery is by producing a lit review. Its value lies in its production. (This can be achieved by doing lit reviews in the 3-5 papers that constitute a stapler dissertation, of course. It need not be a standalone chapter.)
I imagine someone who isn’t already full prof might care if their work is being read and cited and discussed to the extent that “are you producing valuable work your peers engage with or are you a nobody” for the purposes of getting tenure or promotion at an R1.
It might also help cut down on slop or retreads if we require people to read and engage with the literature. It’s not good in my opinion that “top” departments encourage article-based dissertations and simultaneously don’t require their students to read the existing literature. It is an intellectual dissservice to their students and the discipline, makes their students look bad being blindsided at conferences, and accrues reputational cred to people who don’t deserve it in a reputation-obsessed discipline when they unknowingly rehash the ideas of less prestigious people.
There’s a boring sociological observation to make here. As publishing records have become more and more central to market success, graduate students have become more and more incentivised to write their dissertation as several papers on a loose theme rather than as a monograph. But if your dissertation looks like that, it might not really make sense to kick off with a literature review — the papers in your dissertation may have feet in several related-but-distinct literatures. And thither goes the literature review.
There may, of course, be other things going on. But the above seems a pretty central reason that most folks I went through grad school with didn’t include a literature review in their dissertations.
Seems to me that this is precisely the reason.
Even the monograph-style dissertations (like mine) are still written in ways where each chapter could feasibly turn into standalone publishable papers. So even literature reviews function more like opinionated overviews that could be submitted to (e.g.) Philosophy Compass.
In my department, we are not required or even encouraged to do a lit review. I’m in the middle of a dissertation that is interdisciplinary. I feel like I have narrow thesis, but it connects with many different fields and areas of philosophy. With my dissertation, if I tried to review all relevant literature I would be stuck there forever. I am spending more time focusing on my thesis instead, and making contributions to the literature, instead of just reviewing it.
It may be a result of the literature review becoming part of the comprehensive exam process at some institutions. A number of my friends placed at different universities across Canada fulfilled part of their comprehensive requirements by writing a literature review (and in some cases more than one) that would prepare them to write their dissertation. That lit review was submitted to their committee and I think went through a revisions process but was considered part of prepping to write the dissertation, not part of the dissertation itself, and therefore is not included as part of the archived dissertation.
I’m not sure how widespread these practices are but that might explain some of the lack of lit reviews citations turning up on Google Scholar, etc. There may be substantive lit review engagement with Schliesser’s work, it just isn’t being made public.
As someone who had to write comprehensive exams at my institution the old fashioned way (4 area exams, 48-72 hours each, which I was expected to take 4 months to prepare for each, but that expectation was often extremely unrealistic), I have to admit that I was jealous of my peers’ opportunity to do something that would help them prepare for their dissertations. Looking back, the comp exams that I did seemed to function as little more than an extreme hazing ritual that I have retained almost nothing from. By the time I made it to the dissertation, I was so burnt out and had spent so much time focusing on work completely irrelevant to it that it took a long time to get to actual writing (I had to completely refamiliarize myself with my own area of literature).
I graduated from my Canadian program in 2017. The university required us all to have a lit. review as our first chapter.
(We did not have comprehensive exams.)
I’ve never really understood the point of a literature review. If it’s something like summarizing whatever’s out there, placing everything into “hoppers” and sorting things into “camps”, I feel like a lot of “substantive” articles already do that sort of thing and so the literature review doesn’t accomplish anything.
Furthermore, it’s *not* publishable insofar as there’s no “new” content, but rather a taxonomy of existing stuff. If I wanted to know about that stuff, I’d just go read it directly, rather than, say, X’s take of how it goes. (I say this as a former journal editor, who had too many literature reviews submitted as standalone articles–i.e., without enough original argument).
As for falling citation rates, I’d think a lot of that has to do with the explosion of content. And also less interest in long-form essays, as opposed to Twitter, social media, etc. A lot of people just don’t read journal articles anymore, even if they ever did.
Finally, with regards to AI, the literature review really seems doomed. ChatGPT can write a great literature review in five minutes, so why would humans be spending months working on it? If it’s just for the practice/exercise (or to “build mastery”), I feel like that’s going to ring really hollow. We even have non-AI resources (like SEP and PhilCompass) that can take the place of literature reviews as well.
Tl;dr: I think if there’s some novel cluster of papers in some field, they’re probably worth synthesizing. I think writing some quasi-historical account of what the views are, who had them, and what motivated them already feels anachronistic.
There is value in publishing something like a lit review – I think we all love the Stanford Encyclopedia of Philosophy, and Philosophy Compass is also a resource that many people find really valuable.
The value of these things surely changes now that it’s possible to crank out automated versions that work ok – but there’s still something really valuable about having a *particular* person’s take on the literature (just like there is in having a *particular* person’s review of a book – you don’t read Chomsky’s review of Verbal Behavior because you want to know about Skinner!)
A Stanford Encyclopedia entry should NOT be a literature review. It should be an overview of its topic, highlighting important issues and arguments and mentioning authors only as secondary to those. It should be an introduction to a topic, not an annotated bibliography.
Graduated relatively recently. I didn’t have a review requirement. But I feel like most people at my institution did end up doing something like a lit review anyway. Any paper/chapter has to talk to the relevant literature even if it doesn’t have to mention it all or trace its development through history.
And even stapler theses needed an introductory chapter to tie it all together that would invariably need a bit of lit review.
Brown does require a literature review, although it can be executed in different forms. As Marketeer says, the point is that someone writing a dissertation should be familiar with a lot of literature in the area. I guess it’s a bit different for a three-paper dissertation, but assuming the three are in the same area it still seems like a good idea.
I don’t know whether ChatGPT can write a good literature review. But the point isn’t to write a literature review so that a literature review will exist. It’s to write it so that you will know the literature.
Finally: I’m pretty sure Eric S. was not complaining that people aren’t talking about him enough. That’s not really Eric’s style.
While the point about cumulative dissertations seems to make sense, the overall trand of abolishing literature reviews strikes me as philosophy making itself redundant. How can you enter an academic conversation without first signalling that you have taken in at least the most pertinent positions?
But wouldn’t ignoring the literature entirely just make for bad chapters/papers? If a PhD student writes something that just talks past the existing things, surely it’s simply the advisor’s task to tell them to read the relevant things.
The “death” of the literature review as an explicit part of dissertations cannot mean the death of referencing as such and I don’t think there’s any sign that’s the case.
When I was researching my dissertationr (many decades ago) I was able to read and take notes on about ten articles/day = about 300 articles a month = 1800 arricles over six months.
Not doing this kind of grunt work has a price:: reinventing the wheel, as is happening so embarrassingly today with Hegel.
In the social and life sciences, one point of a lit review is to point out a “gap” in the literature. That is, you describe where the lit has brought us and indicate a relevant, unanswered research question that your own hypothesis might answer. You thus show how your research moves the discussion further, or how it might advance the collaborative research field.
Philosophers might have dispensed with this step because they can mark out the problem and associated questions more directly, without having to bring the reader up to speed on where the literature has brought us. Still, surely some review of the literature will help mark out the problem (e.g., by showing how some initial ways of responding to the problem failed).
Like ‘Tommy Bob’ I literally did not know until I read this thread that in certain circles writing a literature review was once (and sometimes still is) the standard operating procedure for philosophy dissertations, let alone that this is ceasing to be so. If literature reviews were once regarded as de rigueur this seems to me a Bad Thing and I am glad that they are going out of style. A much better way to to structure a dissertation is to start out with a problem to be solved or a thesis to be argued for, plus an outline of your strategy for solving the problem or arguing for the thesis. This will tend to generate a list of opponents and allies who can be ‘reviewed’ in the course of being argued with. This is likely to be lot more interesting than the dreary he said/she said of a literature review, and this holds even if (as Caio Cezar Silva puts it ) the dissertation is ‘akin to a monograph with a single theme’ (which was certainly the case with mine). Another problem with the policy of ‘literature reviews’ is that it presupposes that there is a determinate literature to review which is often not the case in philosophy. For considerations relevant to solving a problem or arguing for a thesis can often come from ‘other’ areas of philosophy (which therefore won’t be part of ‘the literature’ that needs to be reviewed) or from outside philosophy altogether. In an earlier thread ‘Marketeer’ raised the problem of the many papers that ‘are only marginally valuable (if that) contributions to debates several layers deep in the dialectic of an issue within a sub-sub-sub field, advancing views the authors themselves may not even find convincing, interesting, or valuable’. I am inclined to suspect that compulsory literature reviews may have nudged people towards this kind of hyper-specialised paper by narrowing their focus to the extant ‘literature’. Finally to ‘Marketeer’ in their present incarnation. Is the point of a dissertation to display one’s mastery of a field? I would have thought it is to make contribution to knowledge by proving an interesting point or solving an interesting problem. People who do this successfully will probably display a certain amount to mastery, but this display is a byproduct of the process not the object of the exercise.
Reading Charles Pigden’s thoughts here next to Tom Hurka’s response to Kenny Easwaran as well as Kenny’s won comment, it strikes me that there is some agreement that knowing the stuff a literature review would help one to know is useful and that knowledge (however obtained) is likely necessary to working in an area. And they probably also agree that reading such a thing is only of interest if it is organized along in an enlightening way. I take Charles and Kenny and Tom to all share a preference for reading something that gives someone’s take on a constellation of issues tackled by a body of literature over a mere summary. (Or perhaps I’m projecting my own tastes here.)
I guess I also think that an undigested literature review would be pretty unhelpful to someone working in an area, including its author. You’d have to digest the material and organize it in some way to really benefit.
FWIW, even almost 40 years ago, I never had to do a literature review, though we did have to pass an oral general exam on a body of literature that we had to construct ourselves and possibly revise before it was OKed. That prep was then used in writing a dissertation but not included in the form of a review (at least in mine). I do think that somehow I wound up with a pretty good overview of the areas I work in and that it has been important to have that.
Consistent with all that, the number of papers one would need to read to do a genuinely comprehensive review these days would be much more than I would have had to read to do the same, both because there were fewer papers and because most current papers are more specialized and a bit thinner (even when more densely argued) than the papers of old.