All Ten Rules for Writing about the History of Philosophy

Having spent the last 25 years of my life teaching history of philosophy, I’ve obviously had a lot of opportunity to give advice and feedback to students on their writing projects. I often find myself giving the same tips over and over, so it occurred to me it might be helpful to write them up for a wider audience and post them here. Of course writing about philosophy is an art, not a science, and these points are going to be to some extent just a matter of taste. I’m calling them “rules” in the same somewhat tongue-in-cheek sense used in my earlier “20 rules for doing history of philosophy,” some of which are relevant here. In fact some of the points will overlap with what I say there but where that happens, the points bear repeating, and in any case what I say in this series of rules will be more with an eye to practical application.

Also I should mention that they really are geared towards writing about historical philosophical texts, since this is the only area I’m qualified to speak about; some of what I say is surely applicable to writing about philosophy more generally, and even writing in the humanities more generally. But some points will be more narrowly relevant. In another sense though the rules are meant generally, because they are should apply to students writing at BA, MA, or PhD level; indeed also to non-students, so anyone who is trying to write about history of philosophy at any length, from a term paper to a book-length study. This is going to be about academic writing, though, not e.g. writing for a popular audience; obviously I’d have some things to say about that too and some of this would apply there, but it is really a different issue.

As always I’d welcome feedback in the form of comments, which you can leave here at the bottom of the page!

Rule 1: The first question is, what is the question?

On then to my first rule, which is: “The first question is, what is the question?” I’ve said this to students countless times when they are in the first stages of working on a piece of writing. The point of the slogan is that people usually think they are looking for a topic or theme. That makes sense as a first initial framework to have in mind, but only insofar as it gives you a guide to what primary and secondary literature to read as you are thinking about your writing project. It is not the right idea when you are close to producing the piece of writing, still less when you are actually writing it. To imagine yourself addressing a mere topic can lead to writing that merely rehearses what the philosopher is saying about that topic in the primary text. But writing a book report is not doing philosophy! Rather, you are looking for a question to answer in the piece. 

This question needs to be neither too difficult nor too easy to answer, or as I sometimes tell students, the issue needs to be resolvable but not obvious. This may itself seem obvious, but it is a surprisingly difficult balance to get right. On the one hand, you need a question such that there are resources in the text to address it; on the other hand, the text can’t address it so squarely that the answer is right there in what the author says. Often, a good way to find a question like this is to pay attention to what you yourself find puzzling when you are reading; when you are confused, that’s a good sign! Because it means there is something unclear that you could then work on clarifying for others. When you actually start writing, at the outset you should try to get the reader to feel that same tension or puzzlement, so they get that your question is a good and pressing one. 

Just as you should lean towards a question that is relatively narrow, you should lean towards a question where the answer is difficult to give (so, very non-obvious). Your answer doesn’t need to, and probably even shouldn’t be, “definitely right.” Rather it should be well-justified. As I often tell students, “you’re not necessarily shooting for true, you’re shooting for plausible.” 

Of course there are many kinds of questions one can put to historical texts: how does the form relate to the content; how does the work respond to other works by the same or other authors, or to its historical context; what are the underlying presuppositions or unstated implications; how does a given argument work; how would contemporary philosophical concepts apply to your text; what exactly does this bit of terminology mean; etc. What I say above applies to all these questions and more. Any question that feels like an answer would shed light on the text, author, topic etc is fair game, you just need to get the reader to feel that they too want the answer, and then provide a satisfying answer. That answer is of course the “thesis” of your paper, what you are trying to establish. So another way to think about this rule is simply that the thesis you are arguing for can best be thought of as an answer to a sufficiently difficult (but not impossible!) question.

Rule 2: Make the weaker argument strong

This rule might sound shocking, with its echo of the accusation made against the ancient sophists, that they “made the weaker argument the stronger,” but there is a crucial difference between “strong” and “stronger.” What I mean is that if you are trying to establish a given thesis T, you should give a lot of thought to the case for Not-T, and try to make that case as strong as possible so that you can then defeat it. 

This takes us back to Rule 1, which stated that you are trying to settle a question where the answer is not obvious. One reason it is not obvious may be that multiple possible answers are available, in which case you need not just to support your own preferred answers but also show that the other possible answers won’t work (or have problems that make them less plausible). Meanwhile there will be potential criticisms and problems for your own thesis, and you need to think about these. Rather than trying to minimize them or sweep them under the rug, you should work on coming up with the strongest case for rejecting your own thesis. This is a good habit of mind in philosophy generally: whenever you’re arguing for T, imagine the moves that can be made by an opponent of T, anticipate them and explain why those moves won’t work.

Now, it’s possible to take this process too far. Some philosophy papers (especially in non-historical analytic philosophy, but also in history of philosophy) get so mired down in dealing with objections that they barely bother to persuade the reader of their own thesis. You need to avoid that: the primary objective is still to argue for T, and some of the thinking you do about the case for not-T (and how you could respond) will probably not make it into the final paper. But the deeper your understanding of both (or more) sides of the question, the better your writing will be. Especially for undergraduate essays, where you are not necessarily trying to argue for one or another view, a good strategy is to lay out the views on both sides of your question and then try to decide the issue at the end with the point that feels particularly strong for one side or the other. I often tell students they can think of an essay as a kind of tennis match, with point and counterpoint being made between the two parties to a philosophical or interpretive dispute.

Rule 3: Go narrow 

Practically every discussion I’ve ever had with a beginning PhD student has involved my telling them that their envisioned project is too ambitious and broad. Often, grad students outline dissertation topics that even sound like they could be projects for a whole career (“the problem of free will from Aristotle to medieval scholasticism”). It’s not uncommon for me to tell students that they can tackle about 20% of their planned topic in the actual thesis, and that their first task is to find the right 20%. 

Similar advice goes for all levels of philosophy writing: often I tell BA students that to write a 12-15 page term paper, they might select a single paragraph or even a single sentence in the primary text and focus on dealing with the problems it poses (this is a nice way to find a “question,” see Rule 1).

In my experience students often find this advice frustrating or disappointing; probably it feels like I am training them to do the kind of angels-on-a-pinhead scholarship that many people lament when it comes to modern-day philosophy. But the fact is that philosophy, and history of philosophy, are about detailed, fine-grained thinking, where progress is usually made in very small increments. I find it much more satisfying to read something that poses a narrow, exact question and provides a compelling answer, than to read something that goes over a lot of territory which people may already have discussed before, and which skips past the details where things get really difficult and interesting. Though there may be exceptions, there’s almost always a kind of trade-off between broad-and-shallow and narrow-and-deep, and since students in my experience tend to err on the broad-and-shallow side, I generally encourage them to tack back towards the narrow-and-deep side instead.

Rule 4: The primary text is primary

This overlaps with a point I made in the previous series of “20 rules for doing history of philosophy,” but I’m going to say something similar here because it is so important. When you are doing history of philosophy, you want to be closely engaged with the historical texts (or other material you’re working on, but it will usually be a text). This means that you should spend more time reading (and re-reading) that text than engaging with secondary literature; it’s amazing how a work may yield up further insights on repeated readings. Of course, secondary literature is helpful, but you don’t want to slip into the trap of writing “tertiary literature” where you are writing far more about recent historians of philosophy than about the primary text. Even if your aim is to respond to a piece of secondary literature, you should always make clear that this is part of an effort to understand the primary text, which needs to stay firmly in view.

One issue that arises here, for students but even professional academics, is how to signal to the reader that you have “done your homework” and read all the secondary literature (unfortunately referees for journals love to complain about items missing from bibliographies… especially items they wrote themselves!). There’s an old-fashioned European habit of producing a “literature review” where all the previous scholarship on a topic is surveyed. As you can imagine given what I’ve said in the first three rules, I dislike this policy since it gets in the way of diving into a question and getting the reader to feel gripped by it. The literature review can be a good exercise to write for yourself, but it shouldn’t be included in a piece of philosophical work, in my opinion. (Sometimes the lit review is positively required as part of a PhD, but even then it should not appear in the published version. I’m always amazed to find actual published monographs that start with these glorified homework exercises.) 

And this is the wrong way to think about secondary literature anyway: “here is what others have said; ok, now let’s move on to what I want to say.” Instead, you should work on weaving the interpretive ideas of other scholars into your own writing, in an organic and productive way. Think of points from secondary literature as planks in the structure you’re building, which help to give the whole thing shape and stability, or to change the metaphor, as little pushes that move the argument forward. If you merely want to signal to the reader that you are aware of relevant work, simply put it in the footnotes: “on this topic see also Smith 2020.” 

When you do engage with secondary literature more actively, I recommend doing this in a mostly positive way: giving credit to others and building on their insights, rather than focusing on where they have gone wrong. I really dislike it when philosophers seem like they are nitpicking, or going out of their way to be critical of other scholars. After all, we are all engaged in a collective effort to understand the texts, and everyone is presumably just doing their best. Of course sometimes you do need to register disagreement, but that can be done politely and affirmatively: “Smith 2020 makes a good point by observing that X, but might also have noted that…” If you think poor old Smith 2020 hasn’t done anything useful to advance the debate, then better not even to mention them.

One caveat to the above is that once you’re doing proper research, so perhaps at PhD or even MA level, you are under some pressure to show that what you are saying is new. This does mean you have to get across that you’ve read the rest of the scholarship and identified something you can add to it. But even here the better way to do this is not with a literature review, but rather mentioning the more relevant or similar studies along the way and engage with those studies organically; this will convey to the reader that you’re familiar with the terrain and have a sense of where your own work fits in.

There is also an art to using the primary text elegantly and effectively, but I’ll get into that in the next rule.

Rule 5: Use quotes wisely

In my published research I try to create a kind of visual rhythm on the page, which is produced by “block quotes”: chunks of quotation from the primary texts I am dealing with, which come along every couple of paragraphs or so. This goes along with one way I think about the writing process, which is that I’m taking the reader from one quotation to another, showing them how these passages fit into an overall line of argument (one that answers the question motivating the whole piece: see again Rule 1). You could think about this as a kind of guided tour to the primary text, visiting particularly interesting passages and seeing them in a new light.

I say “in a new light” because you almost always want to assume that your reader has read the primary text, and read it carefully. Admittedly, this isn’t always the case: in my work on Islamic philosophy I’ve sometimes written about texts that are not yet translated or even edited (i.e. they are only accessible in manuscripts), or are in general untouched in European-language scholarship. This calls for a bit more exposition, since the reader won’t know what is there. But even in such cases I try not to fall into the trap of just paraphrasing the primary source. Instead, I try to get across what the author’s position and arguments are while saying something interesting about them. 

In any case, most historians of philosophy, and pretty much all students, are going to be talking about texts the reader already knows well. (If you’re a student, the person who is going to be reading this is your professor, so in general try write with that audience in mind!) Your goal is never just to tell the reader what is in the text: it is to offer the reader interesting, controversial, well-argued observations about the text.

Similarly, and this is a real pet peeve of mine: do not quote the primary text and then just summarize what is in the passage you have just quoted. Unless the text is almost incomprehensible for some reason (e.g. it was written by Kant), you can give your reader credit for being able to read the quotation and understand what is at least basically going on. The first thing you say after the quotation should not be a paraphrase of it, but an interesting point about it. This should justify your having quoted it: there must be some reason you are putting it in front of the reader even though they are assumed to have read it before. Most often this will either because you want to use it as evidence for a claim you are making about the text, or even better, because there is some feature of the text that could easily escape notice. For instance you may be drawing attention to the vocabulary used or something about the way an argument is constructed or the way a point is phrased. If you don’t have something nifty to say about a passage, don’t quote it at all. And of course the nifty thing you want to say should contribute to your overall argument.

Rule 6: Use a “Russian doll” structure 

Pretty well any piece of philosophical writing has a part-whole structure, with different parts of the structure playing different roles in the overall argument. If we return to the idea that the answer to your question, which is thesis T, is meant to be the overall result of the paper, then those roles are defined by how they contribute to establishing T.

A common piece of advice here, which I fully endorse, is to have a fairly detailed outline of the structure not just in mind, but written out, before you start. When I write, whether it’s for research or for podcasts, I have a set of notes to work from. The last thing I do before beginning to write is produce an outline which includes lists of all the passages (from primary and secondary literature) that I will cite in each part; then I can easily navigate to my file of notes to get to the relevant bits as I go along, which among other things means I don’t have to interrupt my writing to go search for some passage I want to cite.

But the main point I want to make with this rule is that your outline should have a “Russian doll” structure, with bigger parts containing smaller parts. This applies regardless of length, whether it’s a book or only a five page paper. It’s just that longer pieces of writing have longer parts, and more layers of sub-parts. If it’s a book, for example, then the parts will be chapters, which contain sections, which contain sub-sections, which contain paragraphs, which contain sentences. Turning that sequence around, each sentence should have a clear function within its paragraph, each paragraph should contribute something clear to the purpose of the sub-section, and so on.

Another metaphor might be that the piece of writing is like a machine with parts and sub-parts; you wouldn’t build a machine that has parts in it that aren’t serving any function in making the overall machine work.

This is all intended to help the reader know where they are in your argument at all times. Ideally there should not be a single sentence where the reader needs to think consciously about why this sentence is here. They will be cued to that by the overall structure, which (to use one last metaphor) is like a map with a well-marked trail they are easily following.

Rule 7: Mind the gaps

I haven’t said much yet about the actual writing, in the sense of putting sentences together. This isn’t really the place, and I am not really the person, to give advice on how to write well in terms of style. But I do have one suggestion, which is perhaps especially germane to philosophy: to focus on transitions from section to section, from paragraph to paragraph, and from sentence to sentence. 

This goes together with the previous rule about having an overall structure that is evident to the reader. You can use little textual cues to carry the reader through that structure, so that it flows smoothly. Ideally they will almost get the sense that the text is reading itself for them, because it moves forward so effortlessly.

I think writing podcasts has made me especially attentive to this, since if you are writing for a listener and not a reader, you really need to make things easy to follow. So almost every sentence I write has some feature, usually a small one, that cues how each sentence fits with either the previous or the next sentence. Here’s a randomly chosen example from a recent script, with the connecting cues in boldface:

“But let’s not get ahead of ourselves, because having touched on the Discourse,we have learned no more about Descartes’ life than he wanted to tell us, which is almost nothing at all. Fortunately we do have other sources. There were other early biographers, notably Adrien Baillet, author of a Lifeof Monsieur Descartes that appeared in 1691, so 41 years after Descartes’ death. Then of course there are the letters. Pride of place here goes to the correspondence with Mersenne, but we also have exchanges between Descartes and numerous other figures. These sources allow us to reconstruct his life story in some detail. By ‘us’ I of course mean other people, like Geneviève Rodis-Lewis and Stephen Gaukroger, whose modern biographies have given me plenty to draw on in this episode.”

Which is not exactly rocket science but it’s a typical passage in that there is at least one word in every sentence that links to the previous sentence. 

Similarly, I try to write paragraphs so that the beginning and/or end of each paragraph signals the reader that this is a discrete unit of text (in the example above this is achieved through the first phrase “But let’s not get ahead of ourselves,” which both links to and creates a break from what went before). Even though in a podcast you can’t see the paragraph breaks, I try to write it (and then read it) in such a way that you can “feel” them; and this is something one should do in any piece of philosophy writing.

The same goes for the transitions between sections of a paper. As with all transitions, this can be done more elegantly or less elegantly. A bluntly straightforward transition like “having discussed X, I’ll now go on to discuss Y” is better than nothing at all, but only a little. What you really want is to end each section of the paper (whether or not it is explicitly labeled with its own sub-header) with a transition that suggests how the next section would follow on from it naturally. 

I would even go so far as to say that transitions are the most important stylistic thing to bear in mind when writing about philosophy: they don’t just make it easier to read, but also let the reader follow the logic of the argument step-by-step. In this sense they are the sentence-level feature that most helps you abide by the aforementioned rules.

Rule 8: Stay on target

This rule is implicit in some of the others, but I still want to invoke the advice of that random X-Wing pilot to “stay on target” as you write. Not just every section of the paper, but every sentence of the paper, should somehow contribute to your overall project, which remember, is to give an answer to your central question. As we’ll see in rule 10, the conclusion can be a partial exception to this general principle, but otherwise it holds throughout. 

Sounds simple enough, but it creates problems. There are various reasons you may want to say something that in truth is irrelevant. There may be pragmatic issues – you need to cite a piece of secondary literature to show you’ve read it (cf rule 4) – or you may just have a clever idea that came to you when writing, that doesn’t really belong there but you’re so pleased with it that you really want to get it in. So, what to do?

First answer: that’s what footnotes are for. The rule to stay on target applies much less to footnotes since by putting something in a note, you are effectively taking it out of the main flow of the argument and even telling the reader that the point is optional. This is one reason I said under rule 4 that you can show you have consulted the relevant literature by relegating mention of it to footnotes. Don’t let things get out of hand though; no one wants to read a paper where half the word count is in the notes, because that is distracting and undermines the sense that the piece is focused and moving towards a clear target.

Second answer: you can bend the rule a little, by explicitly marking a point as a side issue, e.g. by putting it in parentheses or saying “Incidentally, it’s worth noting that…” Strictly speaking my rule says don’t do this, and when in doubt don’t. But you can do it occasionally, as long as it doesn’t get in the way of moving the argument along. As with other rules, the point is clearly communicating to the reader what is going on. If you signal to the reader “this doesn’t really contribute to my project, but while we’re at it, it’s kind of interesting,” then that gives you some cover (so long as it is indeed interesting). But as I say this should be done sparingly and only if you feel confident in your writing. It’s an “advanced” move, one you can allow yourself if you have everything else under control. 

Corollary: often you will have the reverse problem, that there is a point you could get into that really is relevant, but it is too big or difficult an issue, given constraints of space, time, competence, etc. What I’ve just said applies here too: you are allowed to tell the reader that an issue as relevant, even in principle very important, but impossible to cover in this piece of writing. It’s better to do this than to pass over it in silence and hope the reader doesn’t notice… because they might. The only caveat is that here too, it’s something to do very sparingly. In, say, a 15 page term paper, I wouldn’t want to see a student flag something as relevant but then not deal with it more than once or twice. If you are feeling you need to do it more often, that’s probably a sign that your original question was too big.

Rule 9: Revise, revise, revise

No matter how experienced you are at writing, your first draft is almost certainly going to need a lot of work before it is ready to show anyone. For one thing, you will probably have broken some of my “rules” in your first draft, e.g. by failing to write good transitions, or included points that are not relevant or whose relevance hasn’t been made clear enough. But it’s also about the quality of the writing itself. You should approach each sentence as what it is, namely a first attempt at saying what you wanted to say. It’s pretty unlikely that you nailed it perfectly the first time around! So just assume that each sentence will need changing: it is guilty until proven innocent. 

When I revise one of my main priorities is to make everything more concise. It’s not uncommon for me to remove 10% or more of the words in the process, and that’s despite the fact that I have a very concise writing style. When it comes to student work, I almost always feel that it could be significantly shorter while conveying the same ideas. Most sentences should be short and punchy, like this one. You can throw in occasional longer ones for variety, but they should be in the clear minority. If your sentences are long break them up, which makes it easier to read. (Actually when revising this paragraph, I changed what was the first sentence into two sentences.) I think this is good advice in general, but it is especially relevant for philosophy written in English, because we tend to prize a sharp and precise style, thanks to the dominance of analytic philosophy. (This is much less true of academic German, of course!)

In any case, my main message here is that revising is a key part of the writing process and you should budget plenty of time for doing it. Ideally you would revise more than once, and that means not just looking for typos and mistakes, but genuinely rewriting and fine-tuning as you go. 

Incidentally this advice may also help you when you are writing the first draft, because if you know you’re going to revise a lot, then you will feel less pressure to make it perfect the first time around. Just get something on the page, and you can chisel it into shape later.

And a final piece of advice: a good tactic for revising is reading your own writing out loud to yourself. This may feel silly, but it is a great way to find sentences that are awkward, too long, etc., because your “ear” will tell you when things are not well put, or too convoluted.

Rule 10: Close by opening

Ok, now you’ve written a tightly focused piece about the history philosophy with transitions that offer “signposts” to help the reader through the argument. It raises an interesting question that is not too big, and can be given a plausible answer, and you’ve argued cogently for that answer. You’ve stuck to the topic throughout and focused mostly on the primary text(s). Congratulations! Now there is just one task left: write a conclusion.

Beginning writers are often coached to write an introduction that makes the reader interested in the topic and works towards a thesis statement. That’s good advice, so long as it doesn’t feel too mechanical. (And by the way, please don’t start your essay “Since the dawn of time, humans have always wondered…” And for God’s sake, don’t start by quoting the dictionary. Instead, start developing your argument from the very beginning. Most student work would improve from losing the first sentence or even the first paragraph, because it is overly general.) 

Beginners are also told to conclude by summarizing what they have done in the essay, I guess to make sure the reader got the point. Maybe that is good advice for real beginners but since I’m thinking here of writers who are at least university students, I would discourage it. If you get to the conclusion and still need to tell the reader what you’ve been arguing, something has gone badly wrong. So what should you do instead? 

My advice is to “conclude by opening,” that is, open a new perspective on what you’ve been talking about. This could be a further implication that follows from your thesis. It could be a possible objection to your thesis, one that is too profound and difficult to answer here. It could be an intriguing parallel to some other issue, text, or author. Whatever the case, the idea is that your essay has now achieved its goal, so you can give the reader a kind of “bonus” idea at the end. Of course this idea should be securely anchored in your essay; it shouldn’t just come out of left field. But with that constraint, it is a final chance to impress your reader, persuade them to give you a higher grade… and perhaps even leave them wanting more.