I don’t know what I think. A plea for unclarity and prophecy

Would you begin a research project if there were just one more day left to work on it? I guess I wouldn’t. Why? Well, my assumption is that the point of a research project is that we improve our understanding of a phenomenon. Improvement seems to be inherently future-directed, meaning that we understand x a bit better tomorrow than today. Therefore, I am inclined to think that we would not begin to do research, had we not the hope that it might lead to more knowledge of x in the future. I think this not only true of research but of much thinking and writing in general. We wouldn’t think, talk or write certain things, had we not the hope that this leads to an improved understanding in the future. You might find this point trivial. But a while ago it began to dawn on me that the inherent future-directedness of (some) thinking and writing has a number of important consequences. One of them is that we are not the (sole) authors of our thoughts. If this is correct, it is time to rethink our ways of evaluating thoughts and their modes of expression. Let me explain.

So why am I not the (sole) author of my thoughts? Well, I hope you all know variations of the following situation: You try to express an idea. Your interlocutor frowns and points out that she doesn’t really understand what you’re saying. You try again. The frowning continues, but this time she offers a different formulation. “Exactly”, you shout, “this is exactly what I meant to say!” Now, who is the author of that thought? I guess it depends. Did she give a good paraphrase or did she also bring out an implication or a consequence? Did she use an illustration that highlights a new aspect? Did she perhaps even rephrase it in such a way that it circumvents a possible objection? And what about you? Did you mean just that? Or do you understand the idea even better than before? Perhaps you are now aware of an important implication. So whose idea is it now? Hers or yours? Perhaps you both should be seen as authors. In any case, the boundaries are not clear.

In this sense, many of my thoughts are not (solely) authored by me. We often try to acknowledge as much in forewords and footnotes. But some consequences of this fact might be more serious. Let me name three: (1) There is an obvious problem for the charge of anachronism in history of philosophy (see my inaugural lecture).  If future explications of thoughts can be seen as improvements of these very thoughts, then anachronistic interpretations should perhaps not merely be tolerated but encouraged. Are Descartes’ Meditations complete without the Objections and Replies? Can Aristotle be understood without the commentary traditions? Think about it! (2) Another issue concerns the identity of thoughts. If you are a semantic holist of sorts you might assume that a thought is individuated by numerous inferential relations. Is your thought that p really what it is without it entailing q? Is your thought that p really intelligible without seeing that it entails q? You might think so, but the referees of your latest paper might think that p doesn’t merit publication without considering q. (3) This leads to the issue of acceptability. Whose inferences or paraphrases count? You might say that p, but perhaps p is not accepted in your own formulation, while the expression of p in your superviser’s form of words is greeted with great enthusiasm. In similar spirit, Tim Crane has recently called for a reconsideration of peer review.  Even if some of these points are controversial, they should at least suggest that authorship has rather unclear boundaries.

Now the fact that thoughts are often future-directed and have multiple authors has, in turn, a number of further consequences. I’d like to highlight two of them by way of calling for some reconsiderations: a due reconsideration of unclarity and what Eric Schliesser calls “philosophic prophecy”.*

  • A plea for reconsidering unclarity. Philosophers in the analytic tradition pride themselves on clarity. But apart from the fact that the recognition of clarity is necessarily context-dependent, clarity ought to be seen as the result of a process rather than a feature of the thought or its present expression. Most texts that are considered original or important, not least in the analytic tradition, are hopelessly unclear when read without guidance. Try Russell’s “On Denoting” or Frege’s “On Sense and Reference” and you know what I mean. Or try some other classics like Aristotle’s “De anima” or Hume’s “Treatise”. Oh, your own papers are exempt from this problem? Of course! Anyway, we all know this: we begin with a glimpse of an idea. And it’s the frowning of others that either makes us commit it to oblivion or try an improvement. But if this is remotely true, there is no principled reason to see unclarity as a downside. Rather it should be seen as a typical if perhaps early stage of an idea that wants to grow.
  • A plea for coining concepts or philosophic prophecy. Simplifying an idea by Eric Schliesser, we should see both philosophy and history of philosophy as involved in the business of coining concepts that “disclose the near or distant past and create a shared horizon for our philosophical future.”* As is well-known some authors (such as Leibniz, Kant or Nietzsche) have sometimes decidedly written rather for future audiences than present ones, trying to pave conceptual paths for future dialogues between religions, metaphysicians or Übermenschen. For historians of philosophy in particular this means that history is never just an antiquarian enterprise. By offering ‘translations’ and explanations we can introduce philosophical traditions to the future or silence them. In this sense, I’d like to stress, for example, that Ryle’s famous critique of Descartes, however flawed historically, should be seen as part of Descartes’ thought. In the same vain, Albert the Great or Hilary Putnam might be said to bring out certain versions of Aristotle. This doesn’t mean that they didn’t have any thoughts of their own. But their particular thoughts might not have been possible without Aristotle, who in turn might not be intelligible (to us) without the later developments. In this sense, much if not all philosophy is a prophetic enterprise.

If my thoughts are future-directed and multi-authored in such ways, this also means that I often couldn’t know at all what I actually think, if it were not for your improvements or refinements. This is of course one of the lessons learned from Wittgenstein’s so-called private language argument. But it does not only concern the possibility of understanding and knowing. A fortiori it also concerns understanding our own public language and thought. As I said earlier, I take it to be a rationality constraint that I must agree to some degree with others in order to understand myself. This means that I need others to see the point I am trying to make. If this generalises, you cannot know thyself without listening to others.


* See Eric Schliesser, “Philosophic Prophecy”, in Philosophy and It’s History, 209.



Do ideas matter to philosophy? How obsession with recognition blocks diversity

When suffering from writer’s block, I spent much of my time in the library browsing through books that were shelved beside the ones I originally looked for. Often these were books that didn’t have any traces of use: neither, it seemed, had anyone read them, nor were they cited by anyone. The names of the authors were often unfamiliar and a search confirmed that they sometimes were no longer in academia. Funnily enough, these books often contained the most refreshing and original ideas. Their approach to topics or texts was often unfamiliar to me, but the effort of figuring out what they were arguing was time well spent. Nevertheless, my attempts to bring them up in discussions weren’t picked up on. People continued to cite the more familiar names. Why are we letting this happen?

Most of you probably know the following phenomenon: During a discussion someone proposes an idea; the discussion moves on. Then an established person offers almost a repetition of the proposed idea and everyone goes: “oh, interesting.” Put as a rule of thumb: prestige gets you attention; interesting ideas as such not so much. There is a gendered version of this phenomenon, too: If you want to listen to an interesting idea authored by a woman, better have a man repeat it. Now, an important aspect of this phenomenon is that it seems to incentivise that we relate our philosophical work to that of prestigious figures. In other words, we will make sure that what we say picks up on what established figures say. As Kieran Healy has shown, citation patterns confirm this. Cite David Lewis and you might join the winning in-group. We hope to get recognition by citing established people. Now you might just shrug this off as an all too human trait. But what I’d like to argue is that this behaviour crucially affects how we evaluate ideas.

I think Healy’s citation patterns show that we are inclined to value such ideas that are either closely related (in content) to those of established figures or that are presented in a similar manner or method. Put simply: you’re more likely to get recognition if you imitate some “big shot” in content or method. Conversely, if you don’t imitate “big shots”, your work won’t be valued. Why is this important? My hunch is that this practice minimises diversity of content and method. Philosophers often like to present themselves as competitors for the best ideas. But if we track value through recognition, there is no competition between ideas.

Now if this is the case, why don’t we see it? My answer is that we don’t recognise it because there are competing big shots. And the competition between big shots makes us believe that there is diversity. Admittedly, my own evidence is anecdotal. But how could it not be. When I started out as a medievalist, the thing to be done to get recognition was to prepare a critical edition of an obscure text. So I learned a number of strange names and techniques in this field. However, outside of my small world this counted for, say, not much. And when the German Research Foundation (DFG) stopped funding such projects, a lot of people were out of a job. Moving on to other departments, I quickly learned that there was a different mainstream, and that mainstream didn’t favour editions or work on obscure texts. Instead you could make a move by writing on a canonical figure already edited. Just join some debate. Still further outside of that context you might realise that people don’t value history of philosophy anyway. But rather than seeing such different approaches as complementary, we are incentivised to compete for getting through with one of these approaches.

However, while competition might nourish the illusion of diversity, the competition for financial resources ultimately blocks diversity because it will ultimately lead to one winner. And the works and books that don’t follow patterns established in such competitions seem to fall through the cracks. There is more evidence of course once we begin to take an international perspective: There are people who write whole PhD dissertations that will never be recognised outside of their home countries. So they have to move to richer countries and write a second PhD to have any chance on the international market. In theory, we should expect such people to be the best-trained philosophers around: they often have to familiarise themselves with different approaches and conventions, often speak different languages, and are used to different institutional cultures. But how will we evaluate their ideas? Will they have to write a bit like David Lewis or at least cite him sufficiently in order to get recognition?

Now you might want to object that I’m conflating cause and effect. While I say that we assign value because of prestige, you might argue that things are the other way round: we assign prestige because of value. – If this were the case, I would want to see some effort to at least assess the ideas of those who don’t align their work with prestigious figures. But where do we find such ideas? For reasons stated above, my guess is that we don’t find them in the prestigious departments and journals. So where should we look for them?

My hunch is that we ‘ll find true and worthwhile diversity in the lesser known departments and journals. So please begin: Listen to the students who don’t speak up confidently, read the journals and books from publishers whose names you cannot recognise. Listen to people whose native language isn’t English. And stop looking for ideas that sound familiar.

Brave questions. A response to Sara Uckelman

Sara Uckelman has great advice for new students: be brave and ask questions! Even and especially those questions that you might find silly. Why should you? “Because I can guarantee you that every question you have, someone else in the class is going to have it too, and they’re not going to be brave enough to ask, and they will be so grateful to you that you were.”

Going from my own experience as a student and professor, this is quite true. The only thing I’d like to add is that this advice applies not only to beginners but perhaps especially to advanced practitioners. The reason is that there is no such thing as a question that is both genuine and silly. Why? Because at least in philosophy nothing is ever justified by itself.

Nevertheless, asking questions is difficult. As Sara Uckelman points out, it involves bravely embracing “your ignorance and confusion”. Moreover, questions are almost a textual genre unto themselves. (See Eric Schliesser’s advice on how to develop more elaborate questions.) Therefore, I think it’s worthwhile to acually practise asking questions. Here are a few ideas how to get started:

(1) Write down your question! You don’t even need to ask it if you’re unsure. But writing it down will enable you to keep track of your concern as the discussion moves on. You can perhaps see how close your question is to other questions (which might be variants of your question). And you can still choose to leave it at that or ask it later or even after the talk or class.

(2) Figure out what kind of question you have! Back in the day, I often felt stupid because I couldn’t actually pin down what to ask for in the first place. Asking for the meaning of an unfamiliar term is fairly simple (and it’s always a good thing to ask, because terminology is often used in specific and different ways by different people). But more often than not, I just felt like saying “I don’t understand that passage at all.” If you feel like that, it might be a good start to figure out more clearly what exactly you don’t understand about it: a word, a certain argumentative move, the relation between two sentences etc. You can then begin by stating what you do understand and then move on to saying where exactly you lose track. It locates the problem, makes one feel less helpless, and will help your interlocutor.

(3) Structure your question! Sometimes you might just want to get it out and over with. But if you feel comfortable enough it might be helpful to raise a question in a more elaborate manner. I find the following parts useful:

  • target: say what the question is about
  • state the actual question
  • give a brief explanation why the question arises
  • perhaps provide a brief anticipation of possible answers (at talks this is helpful to prepare follow-up questions)

Of course, it’s not necessary to do all of those things. But bearing such a structure in mind often helped me to prevent myself from losing track of where I actually am. Sometimes even the mere act of talking might seem difficult. In such cases, this structure might help you to say some things without having to think (which is difficult when you’re nervous). So you might begin by saying “I’d like to ask a question about this … (insert term or phrase)” or by saying “I have a question. Let me explain how it arises.” Uttering such (or other) words will perhaps make you feel more at home in the space you’re inhabiting.

Talking texts. Conditions of a good interpretation

It’s difficult to determine what the claim of a (philosophical) text is. And thinking about today’s topic, I feel like I haven’t even mentioned the crucial difficulty. I don’t know about you, but for me things start moving once I begin to look at relations between texts. It’s like listening to a conversation. Once you listen to different voices, each of them is more distinguishable. It’s the relation to other texts that makes the aims, claims, and arguments visible in the first place. I’d even say that figuring out the claim of a text is impossible unless we understand what the claim is responding to.

Why is that? I suppose that it has to do with a very simple fact about sincere conversations: no one will just start out by making a claim. I won’t get up in the morning and start a conversation by saying: “By the way, I think, therefore I am.” Claims are responses. They might be responses to questions, refinements or corrections of other claims. And this is why texts don’t make much sense unless we see them in relations to other texts. To put the point in a more technical fashion, claims make sense if you consider them in inferential relations, not if you solely consider them in relation to phenomena or facts. So if someone talks, say, about consciousness, you won’t be able to say much beyond that if you only think about the relation between the claim and the phenomenon (of consciousness). Only when you begin to see how it relates to a specific question, to other tenets or a competing claim will you be able to assess it.

Now you might want to raise the following objection: Surely, you will reply to me, surely you can assess a claim in relation to a question or a different claim. But why should it not be possible to see a claim in relation to phenomena? At this point, I can only hint at an answer: This relation will leave the claim underdetermined. The reason is that the phenomenon is not ‘on the same level’ as the text. It’s like making a pointing gesture into the midle distance without at least attempting a description of the kind of thing you want to point at. Of course, you are able to consider an extralinguistic phenomenon or state of affairs. Think of a red elephant! Now there are a thousand things you can say about that elephant: you can talk about anything in relation to the elephant. Only in response to a specific question can you make a claim that stands in an inferential relation. If someone asks you: “What does the elephant look like?” or “What colour does it have?”, you can claim that it is red. Only such inferential relations make claims in texts determinable.

If you read my earlier piece, you might now hold this against me: But you, Martin, listed various interpretations of Ockham’s “mental propositions”; and these were not primarily standing in relation to other texts but to the phenomenon that was assumed to be picked out by the term “mental propositions”. Sure, the interpretations might have been shaky, but they were intended to get at the extralinguistic facts that Ockham wanted to explain! – Well, although that might seem to be the case, it’s not really true. Even if these interpretations were not formed in explicit relation to other texts of Ockham’s time, they were still formed in relation to contemporary texts. Such texts might remain unmentioned as tacit presuppositions. But if I say that Ockham is or isn’t like Fodor, I compare the Summa logicae to Fodor’s Language of Thought. There is always another text. But if we want to provide accessible interpretations, it’s better to say what these texts (or presuppositions) are.

Now there are of course many possible texts that I can relate any given text to. How do I pick them? – That depends. Of course, there will be your personal associations to begin with: other texts that a given text makes you think of. “This sounds like that”, you might think without ever writing it down. Although you perhaps won’t admit what you initially thought of, keep it in mind. It might be important one day. The next question to ask is what kind of interpretation you want to give. If you are interested in current philosophical topics, think about pertinent texts. If you want to provide a historical analysis of the claim, it will be good to figure out what a text is actually responding to. Now you enter the field in which you can make true and false assertions about the text. But don’t worry. It’s so hard to assess such assertions that any false claim is better than remaining silent. (I mean that.)

But how do you go about determining the claim now? No matter whether you want to give a more philosophical or historical interpretation, it’s important to look for a point of contact. Such a point of contact is a more or less explicit way of relating to another text, either by paraphrase or direct quotation. It might be a term, a phrase or even a paragraph. A point of contact is evidence for a historian: another text has been responded to. But it is way more than that. In finding a such a point of contact you make sure that two texts (and you) share a common ground: something that is agreed on or disagreed about. There are several ways of estabishing a point of contact, but it seems sensible to begin by distinguishing at least three approaches:

(1) If you analyse a text historically, you might begin by looking for quotations or references to other texts. This gives you a first idea of what an author relates to or disagrees with. If you’re lucky you’ve now found something that the claim is a refinement of or an opposition to. So here you can begin to figure out what is being claimed.

(2) If you read secondary literature, you’ll often find that it disagrees about certain points of contact. Figure them out. If there is no clear point of contact, people might be talking past one another.

(3) If you’re more interested in the topic than the historical ties of the text, you can establish a point of contact by relating it to any text you find pertinent. Here, you might follow your initial associations and wonder why you thought of them.

In any case, by establishing a clear point of contact, you’ll provide your reader or interlocutor with an accessible piece of evidence that a discussion can focus on. Texts talk to other texts. In this sense, establishing such a focus between texts, shifting it, or making a new emphasis in an existing one under discussion is a good way to enter a debate or to begin looking at it.

What are we on about? Making claims about claims

A: Can you see that?

B: What?

A: [Points to the ceiling:] That thing right there!

B: No. Could you point a bit more clearly?

You probably know this, too. Someone points somewhere assuming that pointing gestures are sufficient. But they are not. If you’re pointing, you’re always pointing at a multitude of things. And we can’t see unless we already know what kind of thing we’re supposed to look for. Pointing gestures might help, but without prior or additional information they are underdetermined. Of course we can try and tell our interlocutor what kind of thing we’re pointing at. But the problem is that quite often we don’t know ourselves what kind of thing we’re pointing at. So we end up saying something like “the black one there”. Now the worry I’d like to address today is that texts offer the same kind of challenge. What is this text about? What does it claim? These are recurrent and tricky questions. And if you want to produce silence in a lively course, just ask one of them.

But why are such questions so tricky? My hunch is that we notoriously mistake the question for something else. The question suggests that the answer could be discovered by looking into the text. In some sense, this is of course a good strategy. But without further information the question is as underdetermined as a pointing gesture. “Try some of those words” doesn’t help. We need to know what kind of text it is. But most things that can be said about the text are not to be found in the text. One might even claim that there is hardly anything to discover in the text. That’s why I prefer to speak of “determining” the claim rather than “finding out” what it is about.

In saying this I don’t want to discourage you from reading. Read the text, by all means! But I think it’s important to take the question about the claim of a text in the right way. Let’s look at some tacit presuppositions first. The question will have a different ring in a police station and a seminar room or lecture hall. If we’re in a seminar room, we might indeed assume that there is a claim to be found. So the very room matters. The date matters. The place of origin matters. Authorship matters. Sincerity matters. In addition to these non-textual factors, the genre and language matter. So what if we’re having a poem in front of us, perhaps a very prosaic poem? And is the author sincere or joking? How do you figure this out?

But, you will retort, there is the text itself. It does carry information. OK then. Let’s assume all of the above matters are settled. How do you get to the claim? A straightforward way seems to be to figure out what a text is intended to explain or argue for. For illustrating this exercise, I often like to pick Ockham’s Summa logicae. It’s a lovely text with a title and a preface indicating what it is about. So, it’s about logic, innit? Well, back in the day I read and even added to a number of studies determining what the first chapters of that book are about. In those chapters, Ockham talks about something called “mental propositions”, and my question is: what are mental propositions supposed to account for? Here are a few answers:

  • Peter Geach: Mental propositions are invoked to explain grammatical features of Latin (1957)
  • John Trentman: Mental propositions form an ideal language, roughly in the Fregean sense (1970)
  • Joan Gibson: Mental propositions form a communication system for angels (1976)
  • Calvin Normore: Mental propositions form a mental language, like Fodor’s mentalese (1990)
  • Sonja Schierbaum: Ockham isn’t Fodor (2014)

Now imagine this great group of people in a seminar and tell them who gave the right answer. But note that all of them have read more than one of Ockham’s texts carefully and provided succinct arguments for their reading. In fact, most of them are talking to one another and respectfully agree on many things before giving their verdicts on what the texts on mental propositions claim. All of them point at the same texts, what they “discover” there is quite different, though. And as you will probably know, by determining the claim you also settle what counts as a support or argument for the claim. And depending on whether you look out for arguments supporting an angelic communication system or the mental language humans think in, you will find what you discover better or worse.

So what is it that determines the claim of a text?* By and large it might be governed by what we find (philosophically) relevant. This is tied to the question why a certain problem arises for you in the first place. While many factors are set by the norms and terms of the scholarly discussion that is already underway, the claims seem to go with the preferred or fashionable trends in philosophy. While John Trentman seems to have favoured early analytic ideal language philosophy, Calvin Normore was clearly guided by one of the leading figures in the philosophy of mind. Although Peter Geach is rather dismissive, all of these works are intriguing interpretations of Ockham’s text. That said, we all should get together more often to discuss what we are actually on about when we determine the claims of texts. At least if we want to avoid that we are mostly greeted with the parroting of the most influential interpretations.


* You’ll find more on this question in my follow-up piece.

How do you turn a half-baked idea into a paper?

Chatting about yesterday’s post on reducing one’s ideas to one single claim, I received the question of what to do in the opposite scenario. “It’s quite a luxury to have too many ideas. Normally, I have just about half an idea and an imminent deadline.” Welcome to my world! Although I think that the problems of too many ideas and too little of an idea are closely related, I think this is worth an extra treatment.

Before trying to give some more practical advice, I think it’s important to see what it actually means to have a half-baked idea. So what is a half-baked idea or half an idea? What is it that is actually missing when we speak of such an idea? – The first thing that comes to mind is confidence. You might secretly like what you think but lack the confidence to go for it. What can you do about that? I think that the advice to work on one’s confidence is cold comfort. Contrary to common opinion, more often than not lack of confidence is not about you but about a lack of legitimacy or authority. If you were an old don, you probably wouldn’t worry too much whether people think your idea a bit underdeveloped. “Hey, it’s work in progress!” But if you are going to be marked or are on the market, then presenting real progress is a privilege you don’t necessarily enjoy.

Now if you lack certain privileges, you can’t do much about that yourself. Luckily, this is not the end of the story. I think that what we call “half-baked ideas” lacks visible agreement with other ideas. In keeping with the three agreement constraints I mentioned earlier, your idea might either lack (1) agreement with the ideas of others (authorities, secondary literature etc.), (2) with the facts or – in this case – with (textual) material or (3) with your own other ideas. If you can’t see where your agreement or disagreement lies, this might affect your confidence quite drastically, because you don’t know where you actually are in the philosophical conversation. In view of these agreement relations, I’d take two steps to amend this. The first thing I would advise to figure out is how your idea agrees on these different levels. So how does it relate to the literature, how does it relate to your material or the facts under discussion and how does it relate to your common or former intuitions? If you make these relations clearer, your idea will certainly become a bit clearer, too. (We often do this by rushing through the secondary literature, trying to see whether what we say is off the mark. But it’s important to see that this is just one step.) In a second and perhaps more crucial step, I would look for disagreements. Locating a disagreement within the literature will help you to work on the so-called USP, the “unique selling point” of your paper. If your idea doesn’t fit the material, it might be good to re-read the texts and see what makes you think about them in such a disparate way. If you disagree with your (former) intuitions, you might be onto something really intriguing, too. In any case, it’s crucial to locate your disagreements as clearly as possible. Because it is those disagreements that might add precision to your idea.*

Another way in which ideas can be half-baked is if they are too broad. Yes, it might be right that, say, Ockham is a nominalist, but if that’s your main claim, no one will want to read on. (History of) Philosophy is a conversation, and you won’t feel like you’re contributing anything if you come up with too broad a claim. But how can you narrow down your claim to a size that makes it interesting and gives structure to your paper? I think this is one of the hardest tasks ever, but here is what I think might be a start. Write an introduction or abstract, using the following headers:

  1. Topic: If your claim is too broad, then you’re probably talking about a topic rather than your actual claim. If you can’t narrow it down, begin by writing about the topic (say, Ockahm’s nominalism), bearing in mind that you will narrow it down later.
  2. Problem: If everything is fine, you won’t have something to write on. But if your questions are too broad, they are probably still referring to a common problem discussed in the literature. It’s fine to write about this in order to say what the common problem is, say, with Ockham’s nominalism.
  3. Hypothesis: Only in the light of a common problem can you formulate a solution. If you find that your solution is in total agreement with the literature, then it might be better to go back and see where your solution disagrees. (Don’t be discouraged by that. Even if you agree with a common claim, you might have different paths to the same goal or think of different material.) Anyway, in keeping with the “one idea per paper” rule, now is the time to say what you think about one single aspect of Ockham’s nominalism! That’s your hypothesis.
  4. Question: If you have such a claim, you’re nearly there. Now you have to think again: Which question has to be answered in order to show that your hypothesis is correct? Is there a special feature of Ockham’s nominalism that has to be shown as being present in his texts? Or is there a common misunderstanding in the literature that has to be amended? Or is there a thesis that needs some refinement? Spelling out that question as precisely as possible gives you a research question or a set of them. Answering that set of questions will support your claim.

Going through these steps, you can draw on your insights regarding the disagreements mentioned earlier. But even then you might still have the impression that your thesis is too broad to be interesting or too broad to be pursued in a single paper. What then? I’d say, take what you call the hypothesis and make it your topic, and take what you call the question and make it your problem. Then try to narrow down again until you reach a workable size. If you have that, you have written a kind of introduction. That doesn’t yet give you a complete structure. But once you break down the research question into manageable parts, you might get the structure of your paper out of that, too.


* It’s important to note that the task of locating agreement and disagreement requires an explicit point of contact on which the (dis)agreement can be plotted. So you should make sure to find a concrete sentence or passage about which you (dis)agree. You’ll find more on points of contact here.

One idea per paper!

The new academic year is approaching rapidly and I’m thinking about student essays again. In Groningen, we now devote a certain amount of course hours to the actual writing of term papers. This has made me think not only about the kind of advice I want to give, but also about the kind of advice that actually works, in the sense that it can be implemented demonstrably. Given that I’m better at giving than following advice myself, that is quite a difficult question for me. One of the best pieces of advice I ever received came rather late, during my postdoc years in Berlin. I was discussing my worries about a paper with a good friend of mine. It was a paper on Ockham’s theory of mental language, and most of these worries concerned what I could possibly leave out. So much needed to be said – and he just stopped my flow by exclaiming: “one idea per paper!”

At first I thought that he was just trying to mock me. But thinking about my actual worries, I soon began to realise that this advice was pure gold. It settled quite a number of questions. Unfortunately, it also raised new obstacles. Nevertheless, I now think it’s good advice even for monographs and will try to go through some issues that it settled for me.

(1) What do I actually want to claim? – When writing the paper in question, I wanted to say a number of different things. I was proud that had discovered a number of intriguing passages in Ockham that had not yet been taken seriously in the secondary literature. Reading these passages, I had a pile of ideas that I thought were new or deserved more attention, but I couldn’t quite put them into a proper sequence, let alone an argument. My new rule made me ask: what is it that I actually think is new? I initially came up with two and a half points, but soon realised that these points had different priorities. The one and half had to be shown in order to make the crucial point work. So the question I had asked myself had imposed an argumentative order onto my points. Now I was not just presenting bits of information, however new, but an argument for a single claim. (For the curious: this was the claim that Ockham’s mental language is conventional.)

(2) How much contextual information is required? – Once I had an argumentative order, a sequence of presenting the material suggested itself. But now that I had one single point at the centre of attention, another problem settled itself. Talking about any somewhat technical topic in a historically remote period requires invoking a lot of information. Even if you just want to explain what’s going on in some passages of a widely read text, you need to say at least bit about the origin of the issue and the discussion it’s placed in. If you have more than one idea under discussion this requires you to bring up multiple contexts. But if you’re confining yourself to one single claim, this narrows this demand considerably. As a rule of thumb I’d say: don’t bring up more than is required to make your one single claim intelligible.

(3) What do I have to argue for? – However, often contextual information that makes a claim intelligible is in itself not well explored and might need further argument to establish why it works as support of your claim. This could of course get you into an infinity of further demands. How do you interrupt the chain sensibly? Often this issue is settled by the fact that scholars (or your supervisor) simply take certain things for granted: the conventions of your discipline settle some of these issues, then. But I don’t think that this makes for a helpful strategy. My rule is: you should only commit yourself to argue for the one single claim at hand. – “But”, you will ask, “what about the intermediate claims that my argument depends on?” I’d say that you don’t have to argue for those. All you have to do is say that your argument is conditional on these further claims (and then name the claims in question). Rather than making the argument yourself, you can tackle these conditions by pointing out what you have to take for granted or what others have taken for granted (in the secondary literature) or what would have to be shown in order to take up these conditions individually. (To give a simple example, if you bring up textual evidence from Crathorn to address the consequences of Ockham’s theory, you don’t have to begin discussing Crathorn’s theory on its own terms. Why not? Well, because you support a claim about Ockham rather than Crathorn.) Of course, someone might question the plausibility of your supporting evidence, but then you have a different claim under discussion. In sum, it’s crucial to distinguish between the claim you’re committed to argue for and supporting evidence or information. For the latter you can shift the burden by indicating a possible route of tackling difficulties.

So the rule “one idea per paper” imposes structure in several ways: it provides an argumentative hierarchy, allows for restricting contextual information, and provides a distinction between tenets you’re committed to as opposed to tenets whose explication you can delegate to others (in the literature). Your paper may still contain many bits and pieces, but it they are all geared towards supporting one single idea. If you’re revising your first draft, always ask yourself: how does that paragraph contribute to arguing for that claim? If you can say how, state this explicitly at the beginning of the paragraph. If you can’t say how, delete the paragraph and save it for a later day.