It is my firm conviction that you can’t learn to write well by writing about things you don’t understand. If you are a good writer, I will grant that you can come to understand something better by writing about it; your prose will show you the limits of your knowledge. But this will not make you a better writer. It is an application of writing to the problem of thinking, and I take no issue with people who like to use writing for this purpose. What I object to is the argument that if you are not struggling at the edge of your understanding you aren’t really writing. That, I believe, is a dangerous myth about what writing is and what it can be.

Though it may be as prevalent (or more) among scholars, this myth is most pernicious among students. After all, students commonly write about things they don’t quite understand — things they’ve only just learned. Worse, they are often pretending to know things they don’t. I haven’t yet come up with a satisfying way to avoid this problem. There doesn’t seem to be a way around rewarding students for faking their mastery of a subject by writing about it beyond their depth. What are first-year university students to do when they are asked to expound on the causes of the Great Depression, the virtues of transformational leadership, the transcendental deduction of the categories of experience, or Hamlet’s love for his mother? Confine themselves to what they actually know about these things? I don’t think that’s going to work. I certainly don’t think I can persuade them to risk it.

But this doesn’t change the fact that if they want to learn how to write well they will have to practice writing about things they know. The difficulty of writing well about something does not stem from the difficulty of knowing it. That difficulty is of course very real, and I don’t want trivialize how hard it can be to understand things, but the problem of writing remains even when this work is behind us. In the case of academic writing, the difficulty is presenting what you know in such a way that another knowledgeable person can discuss it with you. How do you explain what you know to someone who is qualified to tell you that you are wrong?

Well, first you have to believe that you know what you’re talking about. But students generally aren’t trying to open their thinking to the criticism of their reader when they are writing their papers. They are trying to demonstrate that they have read the course material and attended class. Unfortunately, like I say, there isn’t really a way not to reward them for doing this, even when they are pretending to know something they don’t. In school, knowing what you should pretend to know is itself a competence — or it is at least often taken as a competence in the heat of an examination. A student who demonstrates awareness of the assigned readings is likely to outperform one who demonstrates only knowledge (no matter how real) of things that fall outside those readings. Given a curriculum, it’s hard to see how it could be otherwise.

But students who want to become better writers must practice writing what they know on a regular basis. A good way of doing this is to write about things that they learned last year, or least semester, or at least last week, and still understand well enough to build on. And a really good way of discovering such things is to take about five minutes at the end of the day to reflect on something you learned at the least before the weekend. It should come easily to mind in the form of a simple declarative sentence, and you should be able to imagine composing a paragraph of at least six sentences that support, elaborate or defend it. Then relax for the evening. Tomorrow morning, get up and spend 18 or 27 minutes composing that paragraph. This will make you a better scholarly writer.

When doing this exercise, the important thing is not to be struggling to know the thing you’re writing about. You should be writing from the center of your epistemic strength so that you can build your literary chops. Practice writing about an author or body of literature you know well. Practice writing about an organization or industry you are familiar with.  Try describing a data set you’ve analyzed closely or a theory that you understand comfortably. And choose what you’re going to say based on how confident you feel about that simple declarative sentence. Don’t choose one that you feel uncertain about. Then write the best prose you are capable of, with the most competent reader you can imagine in mind. That’s the way to improve your writing through practice.

Confucius’s classic text The Great Learning begins like this:

The great learning takes root in clarifying the way wherein the intelligence increases through the process of looking straight into one’s own heart and acting on the results; it is rooted in watching with affection the way people grow; it is rooted in coming to rest, being at ease in perfect equity.

This is Ezra Pound’s translation—he calls it The Great Digest or Adult Study. Reading the passage, I realize that this really does express my philosophy, and especially what I try to accomplish when helping people to improve their writing. Writing is very much part of the process of adult study.

I agree with Pound/Kung that intelligence can be increased through a disciplined process. “Looking straight into one’s own heart and acting on the results,” is of course a way of improving your whole intelligence. An important step is to develop “precise verbal definitions of [your] inarticulate thoughts”; Pound simply calls this “sincerity”, i.e., to be able to say plainly and straightforwardly what you think. When you think about it, it is easy to see how such frankness can improve your intelligence and, I hope, just as easy to see how insincerity might hinder or even counter such development.

Universities ought to be places where the frank expression of thought is encouraged and protected. They should also be places where “the way people grow” is “watched with affection”. I have had the pleasure of watching people grow over the past fifteen years. But I have mainly been working with early-career researchers and PhD students. Or rather, it is mainly when working with them that I have the privilege of watching people grow. As undergraduate programs grow and are made more cost-effective, the contact between student and teacher offers little opportunity for such careful observation. I think this is a loss to the teacher as well as the student.

Notice that, according to Kung, your learning is rooted in watching how others grow. One of the functions of students on a university campus is to provide teachers with this important experience. Interestingly, some translations of Kung reduce Pound’s (perhaps overly interpretative) phrase to the claim that the aim of the Great Learning is “to renovate the people”. This top-down attitude seems to be more common today. Certainly, programs are organized in ways that leave little room for teachers to appreciate how their students’ minds are growing more articulate. This is due in part to a number of unproductive misconceptions about the role of writing in education. I’ll say more about this next week; for now I just want to emphasize that writing should have a much more prominent place in higher education than it does today. Less talk. More text, I say.

We should not forget the need to come to rest, to achieve a balance. Intelligence grows naturally if we think about things (and articulate our thoughts) in an orderly way. The process can’t be forced but it can be supported. It can also be interrupted, confounded, and sabotaged. Scholars have an interest in finding ways “to rest in the highest excellence,” as other translations put it. It is their job, in a sense, to “grasp the azure”.

(Note: This is a lightly updated post from my retired blog. I linked to it also in a previous post here about what sorts of assignments might afford opportunities to watch our students grow.)

For many years now, I’ve been trying to explain what I mean by “writing processs reengineering”. I have assumed that people need to understand something about writing in order to gain control of their writing process and produce a satisfying product. But I’ve also long been aware that people won’t become better writers simply by understanding and believing me. At the end of the day, they have to do as I say or they won’t actually improve. Recent experiences have led me to consider a more radical consequence of this simple fact. Perhaps all my explanations are meaningless until my authors (scholars, students) have actually done some writing under the conditions I propose. Perhaps I should tell them what to do first, and only afterwards explain what I mean. This is something I’m going to experiment with after the summer break.

Before I explain to them what “academic writing” is, then, I’m going to give them an exercise to do at home. It can be repeated as many times as they like, but I will probably suggest they do it three times before we talk further, before I try to explain what they’re doing. At the end of the working day they are to write down a simple declarative sentence that they know to be true. This should take no more than ten minutes and preferable only five; they should choose something they know comfortably, not something they’re still trying to understand. Then, at the start of the next they are to compose a paragraph of least six sentences and at most 200 words that supports, elaborates or defends it. That is, they have to decide whether they want provide evidence for the claim, explain what it means, or deal with some possible objections. They are to spend exactly 27 minutes doing this and then take a three minute break. That’s it. They do this three days in a row and then we talk again.

The instructions I just gave should be interpreted in the same commonsense way one would interpret an instruction to draw one’s hand. Suppose I told you to take five minutes at the end of the day and look at your hand in some fixed, comfortable position. Within those five minutes, sketch its outline on a piece of paper. Then, at the beginning of the next day take 27 minutes to draw it as accurately as you can in the same position (you can look, you don’t have to do it from memory). However difficult you may find it is to draw an accurate representation of your own hand, these instructions, surely, are not hard to follow. I want you to think of my instruction to write that paragraph in the same way. You may be unhappy with the result as a piece of writing, but there should be no doubt about what you are trying to do: write at least six sentences that support, elaborate or defend a well-defined claim.

Why, you might ask, should you do as I say before I’ve told you what you’re doing? Here I think I will simply invoke the master’s prerogative. You came to me for advice on the assumption that I know how to write, that I have something to teach you. I gave you some advice that happens to be an instruction. It will take less than two hours to complete and it will be a better use of your time (and mine) than having me spend two hours trying to convince you that you can become a better scholarly writer through deliberate practice. You trusted me enough to contact me; I’m asking you to trust me just a little longer. After that, I’m happy to justify my methods.

Imagine a first-year cohort of undergraduate English majors consisting of 200 students. In the first semester, they take a required course called, say, “Hamlet, Unlimited”. During the course, they read the play, watch three or four performances, and read a half dozen essays about the play, such as Eliot’s “Hamlet and his Problems” and Bloom’s Hamlet: Poem, Unlimited. (Obviously these selections are entirely at the discretion of the responsible course faculty.)

A series of lectures by a Shakespeare scholar is offered and the students are assigned to small TA-led tutorials (the TAs might simply be older students who have taken the course previously.) They are also offered workshops with a librarian, who shows them how to search the literature, and writing workshops where they can work on their prose style. Both workshops are centered on well-defined tasks, such as finding particular kinds of texts and writing particular kinds of paragraphs. These activities are entirely voluntary; the students can pick and choose what they want to attend.

At the end of the semester, the lecturer assigns a 72-hour, 1000-word essay question. It will always be something simple like “How did Hamlet feel about his mother?” and no further guidance on how to answer it is offered. The students are to have spent the semester readying themselves for a question of this kind, one that the lectures and their reading will have prepared them to answer. There are strict formatting guidelines, so that all students submit papers that are visually similar (same font, line-spacing, and referencing style.)

The essays are graded on a curve by a panel of two internal examiners and one external examiner. The panel is sequestered from all other activities for one week (perhaps in a retreat setting) to distribute 40 As , 60 Bs, 60 Cs, and 40 Ds or Fs. (These may be further qualified with + or -. And there would be some flexibility in boundary cases.) The graded (anonymized) essays would be made available to all students.

Such a course, it seems to me, would go a long way towards restoring sanity to the modern university, rife with both grade inflation and performance anxiety. The students would have a straightforward problem — that of becoming better scholars — and every bit of effort they put into solving it will be rewarded. Everyone, including the faculty, will have a clear idea of what is possible given a little effort, and the students’ minds will have been focused on a topic that is of central importance in their discipline. A similar course can, of course, be imagined for any other major. The trick is simply to choose materials that everyone in a discipline does well to be familiar with.

In my teaching and coaching, I am always looking for the repeatable, trainable activity of writing. This, I believe, should be the focus of instruction: the thing you can safely tell the student to do with an expectation of seeing improvement. To this end, I’ve been telling scholars and students to write paragraphs, the units of prose composition. I will go on doing this, of course, but I have decided to open another front. I will tell them also to think seriously about what they know. In fact, I want to suggest that the exercise of imagination is a repeatable, trainable activity. It’s something we should be encouraging students to do because it will make them better prose writers.

If paragraphs are the unit of composition, let’s say that images are the units of composure. In imagination, we bring our beliefs and desires, our concepts and emotions, our senses and motives together. “We make ourselves pictures of the facts,” as Wittgenstein put it. But how, exactly, do we do this? And how do we become better at it?

I already suggest you take a moment at the end of the day to plan a paragraph to write tomorrow. This moment can also be used to train your imagination. It should last no longer than 10 minutes, during which you call to mind some fact you know to be the case. Now, a fact is always an arrangement of things, so you do well to imagine those things and give them names. You should also give the arrangement itself a name, and once something has a name it can be a “thing” in its own right, which is to say you can imagine it as part of another, larger fact.

If I’m not mistaken, you’d like an example about now. Here’s one:

Microsoft is a hierarchical organization.

As an arrangement of things, this fact consists of organizational roles that are filled by people called “members”. To imagine this fact is to imagine that some of these members are at the “top” and others are at the “bottom” — usually there will be more people at the bottom than at the top. This arrangement is called a “hierarchy” and I will ask you to notice that this word doesn’t name the fact; it names the arrangement of these particular things, but that name could also be used to name the arrangement of the members of another organization. The fact is here named by the entire sentence, “Microsoft is a hierarchical organization,” not just the word “hierarchy”. (It takes a proposition to state a fact, not merely a word, which can only name a thing.)

So we’ve just noticed something important about imagining a fact. If something is a fact about one set of particular things it can also be a fact about another. “Microsoft is a hierarchical organization,” says something about Microsoft that could just as well be true of Google. To imagine that Microsoft is is a hierarchical organization requires the same sort of effort as imagining that Google is a hierarchical organization. And this brings us immediately to the most important thing about imagining facts: to imagine that something is the case is always to imagine a bunch of things that are not the case. That is, a fact is always contingent on other things not being the case.

There is no one, simple image of Microsoft’s hierarchy, or Google’s. A triangle with Satya Nadella at the apex would be as true as an elaborate tree diagram that reached down to the lowliest coder — the difference is a question of detail, not truth. What is important is that any particular image always suggests things that could be arranged differently. Every element could be replaced by another element. Nadella could be someone else and Microsoft wouldn’t be any less a hierarchy. On the other hand, though it is difficult, it is not impossible to imagine a significant “flattening” of the organization to destroy its hierarchical structure. The ability to imagine such a cataclysm is actually an important part of the ability to meaningfully imagine that an organization is a hierarchy.

This brings us to a final and crucial point. The “things” in the arrangement become “objects” by virtue of imagining the possible arrangements that aren’t actually the case. Objectivity is the perception of things in terms of their possible relations with other things rather than the merely subjective impression they leave on us. A fact is “objective” in that it has to co-exist with every other fact. It’s not just that in imagining a fact you have to picture where one fact ends and another begins; rather, you have to imagine where one fact must end if another is to begin. It is necessary that the actual be possible, and what makes one fact possible is that other facts don’t exclude it. (Wittgenstein originally argued that “atomic facts” are facts that can be otherwise without consideration of other facts; but one way of reading his later work is as a rejection of the existence of such factual atoms.)

So, to sum up, here’s how to imagine a fact. First, give it a name by composing a sentence. Next, imagine the things that the fact comprises. Name them. Now, consider them from an objective point view, in terms of their possible combination with other things, i.e., other objects that can be meaningfully combined with them. All of this should take no more than 10 minutes in the case of facts you know well. There should be many facts you know well enough to meaningfully imagine for ten minutes. The next day, compose the paragraph. Render the composition of pictures in your mind as a composition of words on your page. Perform your composure.