I’m holding the last talk in the Craft of Research series on Friday. It will be devoted to the utterly mundane business of formatting and referencing properly. I should emphasize, however, that this business is normally conducted under the auspices of a “style guide”; that is, there is in fact something stylish about it. It’s a tough sell, I know; it is something that many students and scholars leave to the end as “chores”, thinking that only a very superficial person would care about margins, typography, and consistency in referencing. Surely it’s the ideas that matter! Perhaps they should remember that papers are not made of ideas. They are made of words — deliberately chosen and arranged for maximum effect. In any case, I have my work cut out for me.

Here’s something to keep in mind. It’s a well known fact about many examiners and reviewers that after reading the abstract and perhaps the introduction, they skip straight to the reference list to see if the authors that they would expect to find there, and the classic works they have written, are in fact listed there. You want to make sure this first impression is a good one. If you have in fact cited the texts they are looking for, you want to make sure they’re on the list; and you want to make sure that the list is in alphabetical order so that the reader can find what they’re looking for. But you also want to make sure that the reference list itself looks like the orderly bibliography it is supposed to be. You want to give the reader a sense of the orderliness of your study at a glance.

By a similar token, this should be an easy source of a particular kind of aesthetic satisfaction. Spending a few hours making sure that your reference list is complete and orderly is worth the effort if you give yourself enough time to actually enjoy it. Looking over a neat list of books and papers that you have spent the foregoing many weeks engaging with should feel good.

The same goes for the visual impression that any individual page of your paper gives off. You should be able to spend an hour or so just flipping through your paper and enjoying the way it looks. Reading random sentences (even out loud) should be a pleasant experience. And the text should “work” at the level of the references; a quote or fact should have a source, and the source should be easy to locate, first in the reference list, then in the library. Try this out on a few different pages.

Make sure that the effect of putting a source next to your own text is to increase your credibility. These days it is often quite easy to find a source while reading a text (especially if you have referenced it properly) and this ease is part of the authority of your text. You can check how well this works by picking some pages at random and finding the source (either by Googling or in the library’s databases).

Does spelling count? Yes. But that’s not all. A nice clean title page suggests an orderly mind that is confident about what is to follow. (I’ll leave it to you to decide whether a generic cartoon does the same; maybe your supervisor has an opinion to share.) Page numbers and section headings are useful to the reader. The table of contents should match, yes, the contents. Figures and diagrams should be easy to decipher and look good on the page. (Learning how to do these things takes time but it is worth it. Don’t think the easy “automatic” solution is the only possible one and therefore the right one.) And they should be easily related to what you have written about them.

All in all, try to make your paper or thesis look like something that was carefully and deliberately made. You made it carefully and deliberately and there is no reason to give your reader any other impression. The few hours this takes are well spent, in part for the simple pleasure it affords. Enjoy it!

You read what you have written and, as you always stop when you know what is going to happen next, you go on from there. You write until you come to a place where you still have your juice and know what will happen next and you stop and try to live through until the next day when you hit it again.

Ernest Hemingway

Begin in the present, today. No matter how big your project is, you must admit that today, at least, will come to an end. Will you be finished with your project? Probably not, unless today is the last day of your plan and you stuck to your plan. Maybe the deadline is today and you’ll hand in whatever you’ve got; that too, is a way to finish. But on most days you will not finish; you will put your books and papers away and you will go to bed knowing there is more to be done. Tomorrow is another day.

This basic idea — that today will end and tomorrow will come — may sound banal but it is enormously powerful. Given a deadline, you can count how many times you will have this experience — of stopping your work even though it is not finished. Once you understand that this will be a famliar, everyday (indeed, daily) experience, you can begin to get comfortable with it, and with the incremental progress that it implies. At the end of every day, you can survey what you have accomplished and look ahead to what you will be working on tomorrow. Get used to it.

I recommend you do this a few hours before you go to bed. Late afternoon or early evening is a perfectly reasonable time to call it a day. I understand that many students think they need to stay up late but, since you do have to get your sleep, that really just means shifting the day by a few hours. (What was it Hemingway said? “In Spain there is no nightlife. They go to bed late but they get up late. That is not nightlife. That is delaying the day.” Something like that.) The important thing is to make sure that you have some time after you have finished doing your work to relax. Don’t study until you sleep and then get up and study all day again. You’re a human being.

If you’re following my rules even a little, you are taking a moment at the end of the day to make a plan for the morning. I normally focus on the writing tasks, which shouldn’t occupy more than three hours of a day, but it’s a good idea to have a clear idea of how you will spend at least six hours of the next day before you stop for today. Now, divide those six hours of deliberate, predictable work into half hours. And put some breaks in between. Decide what you can accomplish in each half hour. If you’ve been working your discipline you can probably write about 200 words every half hour. And you can probably read about 5000 words if you put your mind to it. Try to consider your other tasks (searching a database, analyzing your data, etc.) in similar terms. Get a realistic sense of what you’re capable of and plan to make use of that capacity.

Now, I’m by no means saying you have to spend six hours every day on your research project. I’m saying that’s the maximum, and that if you are working at the maximum you’ve got time to do roughly twelve specific, deliberate things. It’s fine if you’ve got other things to do on some days and will only do five or six, or even one or two, things related to your research project. Just make it a plan rather than a hope. Try to have a list of those things ready for every day already the day before.

Don’t just stop once at the end of every day when you are too tired to continue. Make a plan to stop a discrete number of times every day, roughly every half hour. Give yourself some tasks to fill those moments. Give yourself a realistic amount of time to do it. Stop when you run out of time, not when you run out of juice.

Hemingway stopped when he knew how the story would continue. I recommend you stop simply because it is time. You will side with him or with me or find your own reasons to stop. But my advice, in any case, is to learn how to stop working, every day, very deliberately, long before you reach the deadline. Every time you stop, knowing that you are not yet finished and there is more work to be done, you appreciate the finitude of the problem. Don’t worry for the rest the day, you will get there. You will get it done. Worrying will not.

NOTE: I wrote this while preparing for my “How to Finish Your Project” talk in the Craft of Research Series.

An exercise occurred to me the other day. It’s not as demanding as following my rules for eight weeks, but it might give you a little taste of (and for?) Writing Process Reengineering, before deciding whether to take my course.

Here’s the exercise in a nutshell: Every evening, write a true sentence that is a little hard to believe, understand, or agree with. Every morning, take ten minutes to write six more about the same thing that are a little easier to believe, understand, or agree with. Maybe the value of doing this every day (or, let’s say, 160 days a year) is obvious to you. But let me try to explain what I’m getting at with it.

The exercise is deliberately couched in the language I use to describe the problem of writing a paragraph. A paragraph always has a key sentence that poses some difficulty for the reader. The rest of the paragraph then helps the reader overcome this difficulty. The key sentence may be hard to believe, requiring five or more sentences of support. Or it may may hard to understand, requiring elaboration. Or it may be hard to agree with, requiring a defense. In each case, the problem is to make the key sentence more believable, understandable, or agreeable using no more than 200 words. I normally suggest you learn how to compose such a paragraph in under half an hour. Indeed, I suggest learning how to make effective use of exactly 27 minutes to that end, devoting at least one half and at most three hours a day, 32 weeks of the year.

Some people, however, don’t want to compose every paragraph in such an explicitly crafty way. They produce their prose in a more intuitive (and they might say a more “natural” way), moving from paragraph to paragraph when it seem appropriate, not when the clock runs out. Some people compose perfectly good paragraphs this way because they have good grasp of how prose works, of what is supposed to happen to the reader while they are reading. My exercise is intended to subtly strengthen your intuitions in this regard.

Without demanding that you actually produce a paragraph, it forces you to notice what makes a sentences hard or easy to believe. My favorite example of this difference can be found in ethnographic prose, based on either interviews or observation. Here the key sentence will tell us what a person, or a group of people, thinks or feels, wants or fears, needs or hopes. Since such states are not directly observable, they’re a little hard to believe for the academic reader, naturally skeptical. It’s not that the reader is outright unwilling to believe such statements; it’s just that they want to know how you know. Give them a little evidence and they’re good. So you write a paragraph that presents an account of what you heard them say (in interviews) or what you saw them do (in the field). The fact that they said and did these these things is simply easier to believe than the fact that they felt or thought something. That’s especially true in light of the altogether credible methodology you have described before you present this data. Knowing how you collected it, the data can be taken for given. It’s easier to believe.

So one way to do this exercise is to write a key sentence about a fact that is not directly observable at the end of one day and then write six sentences about directly observable facts that support your claim. Another way is to write a statement of theory, invoking concepts from the literature, and then write six sentences that cite this literature in elaboration of their meaning. Since this literature will be familiar to reader, these sentences will be easier to understand than your interpretation of it, which requires the reader to get inside your head, just as you ask them to get inside the heads of your research subjects in the analysis. Another way is to write a sentence about the implications that follow from your research and then, assuming that the reader objects to your reasoning, you defend the rationality of your normative stance, proposing changes to theory or practice that the reader may not want to make the effort to implement, acknowledging their objections and countering them politely. Whatever you do, the important thing is to give yourself an opportunity to compare the difficulties that your sentences present the reader with.

When doing this exercise, remember that you’re only spending a couple of minutes at the end of one day to set up ten minutes of writing the next. Don’t make it a bigger deal than it is. It’s just a little exercise. Try it for a week and see what you get out of it. And do, please, tell me about it if you feel the urge. I always like to hear whether and how my suggestions help my readers. Comments are open.

This came up on Twitter the other day. A researcher had been desk-rejected by a major journal less than four hours after submitting. An efficient editorial process is normally a good thing, of course, but this seemed a little too fast in some quarters of academic Twitter. Is it really possible, asked Paul Hünermund, to evaluate the quality of a paper in such a short time? I think that’s an excellent question, not merely a rhetorical one, and I’d like to take a few moments to offer an answer to it, mostly in the affirmative.

To be fair to the journal, we’re not here talking about the full review process, just the editor’s decision about whether to send the article our for review. A journal with many submissions (as a top journal will be) will need to make short work of this decision. Someone probably has to look at several papers every day (at least at times) and decide whether the paper is on the face of it a good candidate for review. A positive assessment here will result in work for (usually) three other scholars (the reviewers) and waiting time for the researchers who submitted the paper, so this decision is important. But it should also take considerably less effort than the full review process.

Also, we should keep in mind that underlying our response to the time it takes to review a paper is our intuition about how long it takes to simply read and understand one. We expect reviewers and (post-publication) critics of our work to actually read it, usually the whole thing, before passing any sort of judgment on it. We might perhaps discuss whether the editors in this case owe the writers a full read of the paper, but let’s try to get a sense of what a reasonable effort at understanding a paper involves.

As always, we will try to appreciate the finitude of the problem. A standard journal article in the social sciences might consist of roughly 40 paragraphs. Each paragraph, I usually say, should be written so that it can be read and understood by a peer reader (a scholar in the same discipline as the writer) in about one minute. So let’s say it takes about 45 minutes, and no more than an hour, to read a paper all the way through once, with comprehension. That is, a good paper presents about 40 claims of various kinds along with the support, elaboration, or defense that the reader needs in order to believe, understand, or agree with them. After an hour or so with your paper it should be clear to the reader what you are trying to tell them and why you think you’re right about it.

If that is not possible — that is, if after an hour the point of the paper is not yet clear — that strikes me in itself as a reason to desk-reject the paper. But if the paper is good, so that that first hour of effort on the part of the editor produced an understanding of the paper’s purpose and, not incidentally, a favorable assessment of the quality of the prose, then it should not take the editor another three hours to decide whether the paper is within the scope of the journal, or whether the conclusions are interesting enough, or whether the data is substantial enough to support them.

I guess I’m trying to say that a good paper is easy to read, given the right reader. That also means that readers can quickly decide that either the paper is not good or they are not the right reader. If the reader is an editor, this can guide the desk decision. If the reader is a teacher, this can guide the grade. If the reader is a scholar, it can simply guide the decision about whether or not to keep reading, i.e., whether they have an obligation, as a peer, to see what a fellow scholar has to say here. Giving an hour to a 20- or 30-page paper to this end seems reasonable to me, even somewhat charitable.

Being kind to your reader is the best way to get your paper through the various hoops that academic life presents us with. Whether you’re trying to get published or to pass a course, write your paper with the aim of making it easy, not hard, to evaluate. Write it in the spirit of exposing your ideas to the criticism of a qualified, competent peer, whose opinion you respect. Don’t expect your reader to struggle for hours to decide what you are trying to say and whether or not it is correct. Give your reader an occasion to respond naturally and intuitively to your prose. (It should, as Orwell said, be like window on your mind. Your ideas, not your words, should be at the center of the reader’s attention.)

The quality of a good paper is immediately apparent. If it takes a long time to assess the quality or relevance of a paper — when, e.g., applying a journal’s editorial standards — then that assessment has actually already been made. And this should also help us enjoy the work of reading more, whether as scholars, editors, reviewers, or teachers. Don’t get frustrated by bad writing. Just put it down for a moment and take stock of the situation. Your frustration is already a sign that there is something wrong with the paper in your hand. It may be that you’re not the intended reader or it may be that it’s not very good. It’s certainly not obviously your obligation to spend the rest of the day struggling with it. Decide how much time to give to it, make the relevant decision or assign the appropriate grade, and move on to the next paper.

Most of us are able to distinguish necessarily difficult ideas from needlessly difficult writing. Read charitably. Write generously. But remember that there must be limits to our kindness, and to the kindness our readers show us. Like being good, reading well means “finding ourselves correctly attuned in the apportionment of the moment,” as Heidegger put it. Much of this is just about planning your time and following your plan. Give every paper the attention it deserves.

A scholarly paragraph states a claim and supports, elaborates, or defends it. The claim is expressed in the key sentence, which will often appear early in the paragraph, but may appear anywhere so long as it is clearly stating the proposition that the paragraph is trying to get across. It will usually be a short, direct, declarative sentence and use relatively simple grammar. If the reader is supposed to find the claim hard to believe, the remaining sentences will support it with evidence. If the reader is supposed to find it hard to understand, the paragraph will elaborate its meaning, defining terms or providing illustrations. If the reader is supposed to find it difficult to agree with the claim, the paragraph will mount a defense, acknowledging the reader’s objections and engaging with them. In all cases, the bulk of the paragraph will be easier (to believe, understand, or agree with) than the key sentence, which, crucially, occasions precisely the difficulty that the paragraph is supposed to resolve.

In scholarly prose, paragraphs generally consist of least six sentences and at most two-hundred words. That is, the ideas scholars express in their writing generally require at least five sentences of support, elaboration, or defense to be rendered credible, comprehensible, or contestable to a peer reader. Something that can be said in less that six sentences doesn’t require a paragraph of its own (but may of course be said in support, elaboration, or defense of another, more substantial, claim). A claim that requires more than two-hundred words before the reader will believe, understand, or agree to disagree about it should be broken into two or more simpler ideas. What this rule of thumb reminds us is that it should take about one minute to read a paragraph properly. 182 words of flowing prose or 6 tightly composed sentences of less than 100 words may take about the same time to read. The reader should feel that the effort was a reasonable one, given the claim being made.

A scholarly paper is a series of paragraphs that together argue for larger thesis. That thesis is often stated in one of the paragraphs, whose key sentence will say something like, “This paper shows that…” followed by a clear, succinct statement of the conclusion of the paper’s argument. That paragraph will describe the paper, so that the reader can understand how the paper is going to show that the thesis is true; it will elaborate what is meant by “this paper shows”. The paper itself can always be summarized simply by listing the key sentences, one for each paragraph, and these sentences can be grouped into sections and subsections, each of which can in turn be captured by a single, declarative sentence. Seven of those can usefully make up the bulk of the third paragraph of the introduction, in effect outlining the paper. Two of them might be used as key sentences for the first two paragraphs of the paper, telling us something about the world we live in and the science we study it with. That is, a well-structured paper contains its own outline in the prose of its introduction.

I’m sometimes asked for examples when I say this kind of thing. In fact, this post is occasioned by such a request from Dominik Lukes in the comments to my last post. My gut reaction is always to be a bit apologetic, like I should have led with an example and in any case owe my readers or audience one. But the truth is that I’m not sure examples are a good idea. What exactly is it, I wonder, that is hard to imagine after reading the first three paragraphs of this post? In what sense is what I’m saying too abstract to picture concretely? I mean, if you want examples of what I’m talking about in the first two paragraphs then those paragraphs, and the rest of the paragraphs in this post, are perfectly good ones. (I have of course deliberately written them to conform to my guidelines.) And what is hard to imagine about a series of paragraphs, each represented by a single sentence (the key sentence) and grouped under 7 to 10 headings? What is an example supposed to make clear?

I should admit that my worry is partly a suspicion that any example I provide will be perceived as an ideal and imitated before it is understood. To put it more starkly, I’m worried that I will be providing materials that allow students (and scholars!) to fake their paragraphs before they make them. I don’t think that’s a good way to learn how to write. I want to train students to present the ideas they have, not to pretend to have ideas they don’t. In my view, the only way to learn how to write scholarly prose it is to think of something you know and then write it down with the aim of discussing it with other knowledgeable people. You have to be able imagine facts and people who are capable of knowing those facts. My advice won’t work if instructions like “think of something you know” are completely alienating to you. If you don’t have ideas on a regular basis, you can’t write scholarly prose; you might as well tell a blind man to draw a cat.* The experience of “having an idea” should be familiar to scholars (and university students!). As a writing consultant, I shouldn’t have to provide examples of thinking.

_________
*I am of course aware of the ableism of this remark. I use the image advisedly, after having found this very popular YouTube video by Tommy Edison, which I think nicely makes the point I’m after in the spirit I intend it. There’s no shame in not being a strong thinker on a particular subject. But to expect to be able to write well about something you are unable to form a clear idea of in your mind is a bit silly.