Don’t Forget to Remove the Scaffolding
“In a way that will be familiar to any reader of analytic philosophy, and is only too familiar to all of us who perpetrate it, this style tries to remove in advance every conceivable misunderstanding or misinterpretation or objection, including those that would occur only to the malicious or the clinically literal-minded.”
That’s Bernard Williams, in his essay, “Philosophy as Humanistic Discipline” (published in the collection of essays that bears that name), writing about the “well-known and highly typical style of many texts in analytic philosophy which seeks precision by total mind control.”
Here’s the paragraph in which these choice quotes appear:
The scaffolding metaphor is perfect.
There are certainly many instances of philosophical articles containing too much of “the making of this philosophy article.”* As Williams says, “it is perfectly reasonable that the author should consider the objections and possible misunderstandings… the odd thing is that he or she should put them in the text.”
Williams provides good style advice, but perhaps some reflection is owed to why philosophers are prone to leaving in so much scaffolding.
Might part of the answer be found in the academic publishing process, such as the combination of the importance of publications and the lack of trust one has in referees? Getting an article published can be crucial to one’s career, so, one might think, it’s best to do what it takes to get published. If that means sacrificing style and readability in order to address possible objections that otherwise would have led some possible idiosyncratic referee to suggest a rejection, so be it. And perhaps these writing habits, once developed, stick around long after they’re needed. Understandable, but still regrettable.
What other factors, besides publishing pressure, contribute to a more complete explanation of this style and its persistence? And how should authors decide what to leave in and what to remove? Your thoughts welcome.
* Known to some† as the “Cook’s Illustrated Approach to Philosophy.” See this description of the typical article in that publication, which “follows the recipe’s development… invariably [beginning] with numerous problems in its original incarnation. The author then describes iteratively modifying the recipe’s ingredients and cooking method, each time presenting the recipe to a panel of tasters who provide feedback. At the end of the article, the author reaches a final recipe and lists the ingredients and preparation instructions.”
† Okay, possibly just me.
Of course this can be overdone. But there is good reason for it as a practice, which is that writing philosophy is not like lecturing a bunch of ignoramuses who just need to know the upshot, but more like inviting smart people to follow along a chain of thinking. You know in working out a line of argument that some stray distracting or misleading thoughts may occur to your readers, because you had them yourself, and so you make them explicit and show why they shouldn’t bother or how they can be incorporated.
I think it’s worth considering how the style Williams describes both resembles and differs from that of the early Plato. There, Socrates would begin with his interlocutor’s account of X, and then proceed to identify any flaw in the account, aiming, if possible, to refine it along the way. Some of Socrates’s objections were “clinically literal-minded”. But his ultimate aim there was different from ours: Socrates did not ultimately intend to produce a better account of X; he intended to show that no one had a satisfactory account of X. So, if you think you have a good account of X, then Williams has a point: give it. Perhaps relegate objections and replies to footnotes, addenda, or at least separate sections the reader can choose to skip. (Note: I often fail to live up to this.) But if your goal is to demolish and *not* rebuild, then you should be thorough, even if that is tedious.
To the extent that this style is caused by trying to get past unhelpful editors/referees, it seems we have the technology to ameliorate the problem.
It’s a commonplace for people *re*submitting articles to write letters explaining what they are doing at various points, and especially how they are responding to criticisms. And these are very helpful; it’s good when re-refereeing a paper to see what the author is thinking, but not all of this background needs to go in the paper.
But it’s incredibly rare, almost unheard of, to have such a letter go with an initial submission of an article. I sort of understand how we got to having this pair of actions (send letters with re-submission, but not with initial submission) be a norm. But I’m not sure it’s the best one.
If the scaffolding is really just to impress/get past the editor and referees, maybe it should just go in a letter to the editor and referees?
This is the first I am hearing of submitting a manuscript with an accompanying letter of rationale. How common a practice is this? Are their any norms one should know about (e.g., Only submit such a letter if one is requested)? Any general comments/advice regarding this practice (from Brian or anyone) would be greatly appreciated.
It’s more or less unheard of. As I said, it’s a norm that we don’t do this. I’m not 100% sure why it’s a norm, and I suspect it’s a bad norm. But it is the standard practice. (Except for resubmissions – always include a letter with those!)
Sorry, I understood that you were referring to resubmissions rather than initial submissions. I was hoping you (or someone) could elaborate on that practice. I had no idea that this was done, and now know that I should always do it. Thanks!
I suspect that Brian was referring to the letter that should accompany an R&R. That should detail how you’ve addressed a reviewer’s concerns and why, if you’ve chosen not to do so, you made the decisions you did.
Pessimistically, I assume that it might have something to do with the fact that journals are already frequently overburdened with submissions, such that a referee’s being expected to both read a submission and an accompanying letter of rationale would, if encouraged for all submissions, merely contribute to the ongoing problem of article submission over-saturation?
The easiest way to implement this would be to submit “scaffolded” papers as usual, but part of “conditional acceptance” is the condition to cut down on the scaffolding.
As of right now, it seems to be the opposite. After review, the paper has *more* scaffolding, in form of “a reviewer informed me of a potential objection:” paragraphs.
I would love to see an article with and without the scaffolding .
I love this! I hadn’t read it before, and it makes me feel so much better about having said once, in public – perhaps even more than once, come to think of it – that the norms of Anglo-analytic philosophy require one to write as though one assumes that one’s readers are either completely incapable of abstract thought, or unscrupulous lawyers intent upon misconstruing what they have read — or both. Yay BW! I hate being pressured to treat my readers that way. But I also always felt a little bit like a jerk for having said it, since others seem to value the norm.
From James Pryor’s “Guidelines on Writing a Philosophy Paper”:
“…pretend that your reader is lazy, stupid, and mean. He’s lazy in that he doesn’t want to figure out what your convoluted sentences are supposed to mean, and he doesn’t want to figure out what your argument is, if it’s not already obvious. He’s stupid, so you have to explain everything you say to him in simple, bite-sized pieces. And he’s mean, so he’s not going to read your paper charitably. (For example, if something you say admits of more than one interpretation, he’s going to assume you meant the less plausible thing.) If you understand the material you’re writing about, and if you aim your paper at such a reader, you’ll probably get an A.”
I agree that the primary impetus behind the pedantic elements of many philosophy articles is to navigate the possibility of uncharitable or sloppy refereeing or editorial selection. A secondary motivation worth considering is the readership once published. It’s frustrating as an author to have one’s views misrepresented in avoidable ways in later cycles of the literature. It’s also a waste of time as a reader, when I dutifully read some new article responding to something by someone else that I’m working on, only to discover that the new article has misrepresented the original author or issue or argument.
More relevant than anything else, I believe, is the fact that it is possible to make a living in philosophy by criticizing other people’s arguments. We are aware of this and so you anticipate objections and duly respond to them—with each successful response, in an ideal world, you deprive someone else of a critical publication and so protect my philosophical reputation (but possibly diminish your impact…)
But why do the anticipated objections, clarifications, etc. become so pedantic and tedious? Because we feel must fend off the growing hoardes increasingly desperate for publication— and for whom any pedantic point might do. And we are also aware that for readers on the other side of our key thesis, any troubling point will suffice to dismiss us entirely. And such dismissals might come from other side refs under the guise of upholding objective disciplinary standards.
It would be interesting to know if refs are more likely to reject papers they disagree with. Not sure how hard that would be to study.
One thing people sometimes do is develop a core idea through a number of objections until they end up with something more sophisticated that they believe to be adequate. They start with something simple and intuitively plausible, show how that simple idea is inadequate because of an objection, and Chisholm away — (chisholm, v. To make repeated small alterations in a definition or example. “He started with definition (d.8) and kept chisholming away at it until he ended up with (d.8””””)) — until they get something that retains the kernel of the simple, intuitive idea they started with, but can withstand the objections (obvious or not) to that initial idea.
I find that sort of thing really helpful. It helps to illustrate the connection to some simple, intuitive idea that forms the core of the more complex account of whatever, illustrates the effectiveness of that account of whatever, and justifies or explains why the complex account has the bells and whistles that it does.
Is that the sort of scaffolding that is problematic?
Perhaps this is too flatfooted an answer, but I suspect part of what contributes to the fact that some papers have too much scaffolding is that others have too little. We’re aiming for an Aristotelian mean, and it’s possible to overshoot in either direction. Overshoot in one direction and you have the kind of plodding style Williams is rightly complaining about. Overshoot in the other, and your writing is indecipherable.
Ideally we hit the mean, but if we don’t, in which direction is it better to err? I know what I tell undergraduates…
Whoops, meant for this not to be a reply, but to be on the main thread.
“Might part of the answer be found in the academic publishing process, such as the combination of the importance of publications and the lack of trust one has in referees?”
Based on my own scant experience, and the experiences others shared with me, this “lack of trust” is very well earned. The “unscrupulous lawyers” that Ruth Graff uses as a carricature are the (sad?) reality of the refereeing process.
It does not seem fair to put this on the authors. It’s editors and (lazy) referees who should take the blame (if there is any to be dished out; the option remains that scaffolding isn’t a bad thing).
I submit that every round of reviews adds scaffolding.
If the “unscrupulous lawyers” were somewhat content with the offering, they will still have a number of “minor” objections and the editor will demand that they be addressed *in the paper*. Whence all the “a reviewer informed me of a potential objection” paragraphs and the “the following is not a counterexample” footnotes.
A reasonably frequent event seems to be this: a referee finds a prima facie plausible objection and, instead of thinking it through to see it doesn’t bite, they reject the paper on the grounds that “this needs to be cleared up”. This entails that it needs to be cleared up in the paper (and editors appear to agree). Hence, more scaffolding, next journals.
I put it to the editors to reduce scaffolding (if we all agree that less scaffolding is a virtue): do not accept lazy reviews that point out easily defeated objections. Don’t demand minor objections be addressed in the paper.
Of course, we might need an additional immune system to weed out lazy reply papers, then. But I feel that the system is self-stabilizing. Ironically, or so I suppose, the same “prima facie plausible, but easily defeated objection” that a referee will use to reject a paper, will not pass muster when the same referee reviews a reply paper.
You seem to imagine that the people who review papers are a different, and lazier, breed of animal than those who read published papers. This is a mistake: reviewers are generally pretty typical representatives of readers of published papers. Most readers are too pressed for time to read most papers *very* carefully or to spend much time pondering after they’ve finished. Maybe they would give the work of a “star” that treatment, but not the work of the average working philosopher. So if there is a prima facie plausible objection to an argument, they will probably be satisfied that the argument doesn’t work, even if a second careful reading of the paper and enough time spent thinking would convince them that the objection lacks force. Maybe in a world where much less was being published there would be less need for scaffolding, but that possible world is a long way from this one. And it’s not a world where reviewers and editors are less inclined to reject papers than they are now.
A different metaphor, with the same upshot: If you were working in certain sciences, there would be a clear demarcation between the “lab work” and the “write up.” Philosophy also requires something like lab work, but, confusingly, it is often done in prose. I find it helpful to distinguish (roughly — precision would be impossible) between the writing process that is part of the lab work and the final write up.
Some advice I got from Michael Smith as a grad student (that I have not always have followed very well) was to “lead with the truth,” i.e., say and explain what the correct position on some matter is, and only thereafter explain, relative to that correct position, why other positions are deficient. This can’t and perhaps shouldn’t always be done, but I think it is an excellent rule of thumb.
I’m sceptical of the idea that authors need to provide something like a chronology of their “lab work,” as Pamela nicely puts it, to satisfy reviewers. Reviewers want to read clear and compelling papers without getting confused — just the same as other readers. If they want to see some discussion of X’s or Y’s views, then this can be handled, in most cases, once the author’s own view has been set out and explained.
Imagine you asked someone for the quickest route from High Street to the library, and they spent ten minutes telling you how X says you should go via Irrelevance Avenue but that’s wrong, and how Y says you should go via the Navel Gaze Observatory but that’s also wrong, and that they at first thought Unsuccessful Lane would be good, but it’s closed for repairs, and then perhaps the adjacent Not Much Better Street, but that’s also closed for the same repairs, so really what you should do is blah blah bloody blah …
The type of paper this comment rejects does have its place. Here is an example of what I’m pretty sure many working in the area (and even in adjacent areas) will agree is one of the very best papers in this field: Gideon Rosen’s 2015 paper on Real Definition. That paper does exactly what this commentator says not to: it tells us about all the mistaken routes to real definition before it provides the right one. And yet that is *exactly* what we need to do when we want to learn about the area, and from the author, in this space. So each type of paper has its place.
I think my view is perhaps idiosyncratic, but as someone who has often thought that Williams tended to substitute “style” (for one sense of that term) for argument, and to therefore not have much of an argument in places where he really needed one, this remark seems expected but not like a very good justification. Of course, articles with poor writing and excessive chisholming are unpleasant to read. But, for many people the temptation to go the other way – to attempt, consciously or not, to hide the lack of a clear argument with a flourish of style, will be strong and must also be resisted if we want to make good intellectual progress. My own thought, again, is that Williams too often went in the opposite direction of what he cautions against here. I’d not want to encourage that.
No doubt one can take the “scaffolding” too far, and I’ve seen a lot of papers that do. But I agree with Matt’s observation that we should be warned by the identity of the messenger. Take Williams’ “Internal and External Reasons”. He made several attempts to clarify what he was trying to say in that paper, and many years later complained that he hadn’t yet found a way to explain his view in a way that others understood. Perhaps if he had been a bit less concerned about style, and took more care to head off possible misreadings, he wouldn’t have found himself in that predicament. For myself, I have been amazed by how easily and often my work has been misinterpreted. The uncharitable diagnosis (suggested above) is that philosophers are often lazy and willfully uncharitable in how they read others. A more charitable diagnosis is that our views are often much harder to discern from our writings than we ourselves tend to think.
I’d also like to express disagreement with the advice SCM reports from Michael Smith, about presenting the true view first and then dismissing the pretenders. In my opinion that’s completely the wrong way to write a philosophy paper. It’s through exploring the wrong views that one demonstrates (1) that there is a problem that still needs a solution, and (2) what (some of) the obstacles are, which a successful solution needs to avoid.
I’m currently teaching an online intro course and – today, as it happens – a student posted this comment in the class discussion board:
“So after reading these 3 articles on consciousness, I have to say I’m pretty blown away. Before I get into that I will state that the one thing I dislike about philosophy is how the texts are written, I really get lost at times with, what seems to me, double talk. I wish the texts would pick a perspective and go into that argument alone and fully explore it. Then if they wanted to still talk about the [arguments on the] other side, do so in a separate text.”
I wonder how many other philosophers find themselves thinking “Get to the fucking point” when reading article (even a good one). G. E. More may be the most extreme example, interesting ideas expressed at the end, tentatively, after going through all others possible views and maybe some that, in fact, are impossible.
Moore seems like nothing but scaffolding. Dig as hard as much as you like, but you’ll only find considerations, not arguments.
I don’t see a problem. In the first place, analytic philosophers are not producers of literary fiction and their research articles are by and for experts. Second, philosophy is a predominantly non-empirical study. So good arguments have to be the tool of research. That demands rigor and rigor demands explicitness. The same as in mathematics. If you prove something in group theory and you need commutativity in your proof, you start by saying “Let G be an abelian group” otherwise your proof is invalid. You can’t rely on readers just presupposing that you mean abelian groups. Rigor often spoils writing style but who is willing to trade style for rigor? Third, philosophy is good at making explicit hidden assumptions, be it in common sense or in scientific theories. Making explicit hidden assumptions requires thinking about rather remote counterexamples, possible misunderstandings and unintended interpretations. Summing up, I reject the scaffold metaphor. It’s rather part of the walls. Of course, there are tedious articles containing irrelevant sidetracks. I suggest we look at specific problems in individual articles instead of first postulating and then criticizing a common writing style in analytic philosophy – which probably doesn’t even exist on closer scrutiny.
To highlight something brought out by Holger Leuz and Adjunct and suggested by many others: Might the amount of scaffolding reflect the type of audience intended? Shouldn’t judgments about the amount of scaffolding be sensitive to determinations about whose needs it might serve?
When we communicate, we render things explicit according to what (we think) our audience needs to have rendered explicit. The idea is to make everything RELEVANT explicit, but whether something’s relevant is partly decided by who’s talking to whom.
So, very roughly, leaving the scaffolding might be good when writing for other experts in your (sub-sub-sub-) specialty, since they will no doubt be ready with lots of possible objections. But if writing for someone outside your specialty, the scaffolding might distract — as happened with Adjunct’s student.
I should add that I agree with, and really like, the characterization M gives, in the first comment, of what a philosophy paper does: it invites its readers to follow along in thinking through something. So what I’m saying in my comment, I suppose, is that the character and explicitness of the thinking-through should probably vary from audience-type to audience-type.
I myself have suffered from this kind of OCDish writing style of thinking every conceivable objection and misunderstanding, but, in my case, the main culprits are reviewers who are too uncharitable and attribute the author every stupidity he/she has not said.
Please, leave the scaffolding in. In fact, add more scaffolding, making your argument as explicit as possible. I want arguments, not literary edification.
As all of us who have tried to get intelligent and attentive non-philosopher friends to read their favourite philosophy article will know from the that including *all* the scaffolding is an infinite task. Or to make the same point: we cannot define every term in a paper, because eventually (even if only by the time we’ve got to pronouns and articles) we’d find ourselves in definitional circles. So we definitely have to remove some scaffolding.
On the other hand, we can’t responsibly just go around making true claims and expect people to work out why they’re true on their own (pace early Wittgenstein).
So all this talk of we should or shouldn’t have scaffolding is more rhetoric than substance: we obviously need to be somewhere in the middle, but where? The commenters pointing out that exactly how much is required is a matter of context, norms, expected readerships, etc., are on point. I wish that reviewers would take this point. My experience chimes with others’ given above: they ask for scaffolding that is unnecessary given the journal’s readership.
This said, the heuristic ‘Be as clear/rigorous as possible’ is basically a pretty good one. Rather than ditch it for ‘be as clear as you need to be’, which would be too vague, it could be supplemented with another heuristic, something along the lines of ‘Get the overall point across well,’ or ‘Do not require a quantity of scaffolding that precludes making a substantial point in the word limit.’
First sentence correction:
As all of us who have tried to get intelligent and attentive non-philosopher friends to read their favourite philosophy article will know from their bizarre take-aways, including *all* the scaffolding is an infinite task.
Separately: There aren’t, so far as I’m aware, lazy reviewers or lazy readers: there are only reviewers and readers rewarded only for cursorily reading papers, and authors who know this and so write accordingly.
The practice of reading a new article slowly and carefully doesn’t really exist, in my experience. This is because readers don’t expect new articles to reward that sort of attention. And they’re not often wrong: authors are under such pressure to publish that they can’t afford the time to make their work reward the attention; and in any case, authors know that their readers won’t read their work slowly and carefully, so why would they produce work that rewards such attention?
You see the feedback loop? I presume the cause is the insane job market, pressure to publish, all that old stuff, rather than anyone’s laziness.
I disagree with the enchanting Williams, violently. Yes, of course intensely pedantic philosophy is often boring—and it’s doubtless bad as literature. But philosophers, I’d wager, shouldn’t be in the business of trying to be non-boring or of writing literature. Timothy Williamson, in his afterword to his book ‘The Philosophy of Philosophy,’ says something apt: ‘We need the unglamorous virtue of patience to read and write philosophy that is as perspicuously structured as the difficulty of the subject requires, and the austerity to be dissatisfied with appealing prose that does not meet those standards. The fear of boring oneself or one’s readers is a great enemy of truth. Pedantry is a fault on the right side. Precision is often regarded as a hyper-cautious characteristic. It is importantly the opposite. Vague statements are the hardest to convict of error. Obscurity is the oracle’s self-defence. To be precise is to make it as easy as possible for others to prove one wrong. That is what requires courage. But the community can lower the cost of precision by keeping in mind that precise errors often do more than vague truths for scientific progress.’
I used to agree with Williams, and get frustrated when my papers would be rejected by reviewers for what seemed like silly, pedantic reasons.
No doubt some of it was silly pedantry. And it is noticeable that a lot of the most interesting new work in philosophy is in books or edited collections, which are not subject to quite the same bar of scrutiny as papers in top journals.
But… I’ve changed my mind a little on this point. Suppose you write a big paper on X. If it’s a freestanding paper, then in a few years it could be the first thing that some grad student reads on X. Unless it’s something super technical, ideally they should be able to follow the main lines of argument without being immersed in the literature you are responding now. The same goes double for 500 years from now when a single dog-eared copy of PPR is all that remains of contemporary analytic metaphysics.