If you’re reading this blog you’re probably a knowledgeable person. I imagine that you’re a student or a scholar in an academic discipline, and you’re working on your craft, which is to say, you’ve already learned a great deal. You know more about your subject than most people — in fact, out of the billions of people who live on this planet, there are probably only a few thousand, perhaps only a few hundred, who know as much as you do about your subject, and this is especially true when we consider the content of your current project. Indeed, on matters relating to what you’re working on right now, you may only have a few dozen intellectual equals. It’s important that you begin with this awareness when you are working on your writing. Understand that you know a great deal about your subject and proceed on that basis. Write from the center of a formidable strength.

Now, keep in mind that all this knowledge didn’t come to you by some miracle this morning when you awoke. You’ve been accumulating knowledge for months and years and you will be drawing on this in your writing today. Last week, too, you were a very knowledgeable person; last week, too, if relatively speaking, you had very few equals on this planet. If your writing is only as smart as you were last week, it’s still going to be very smart. In fact, you’re not significantly smarter than you were last week — you’re literally only marginally smarter. You are almost as likely to have picked up a new dumb idea over the past few days as you are to have truly learned something new. And you’re almost as likely as that to have made no progress at all. How much will you know tomorrow? The answer is almost always going to be: about as much as you know today. And, since anything you write down today will be published, if ever, months from now, whether you write down what you know today, or what you knew yesterday, or what you knew last week, isn’t really going to make much of a difference from the point of few of the novelty of your published work. You’re always going to be smarter than your publication list.

This is what my first two rules are about. What you are writing about is always in the past; who you are writing for is always in the future. Your reader is always receiving an “outdated” message, a report on your state of mind that is no longer current. But you want the reader to learn something about what you are actually thinking. So you have to find a way to make sure that the things you are writing about will remain relevant going forward. You want to write on a stable foundation of knowledge. And the best way to do that is to choose, today, something you knew last week to write about tomorrow.

It takes seconds to write a sentence. It takes minutes to write a paragraph. And it takes hours to write an essay. Writing never takes days or weeks or months. That is, you don’t spend day in and day out engaged in nothing but writing. At worst (and it’s bad enough) you might spend a whole day engaged in writing, to the exclusion of all else, but even this conspicuous display of your commitment to writing is best measured in hours. If you ask me, serious writing should never occupy more than three hours in a given day, which should be divided into 20- or 30-minute “moments” that have been arranged the day before. In each writing moment, you take 18 or 27 minutes to compose a paragraph out of sentences that you write. After a two or three minute break, you get on with your day, which may be another paragraph. You can write six or nine paragraphs this way in three hours. Writing is what goes on during those hours. It does not go on for weeks and months.

A scholar’s life, of course, goes on for years and years. Knowledge builds up over time as beliefs are entertained and tested, and then retained or replaced with other beliefs. The research experience doesn’t usually benefit from being overly segmented and planned. In order to make discoveries, you have to be open to novel insights; unexpected events must be welcomed as opportunities, not avoided as interruptions. So it makes sense to say that you were “working on a problem” for months before a solution presented itself. “Working on it” might simply have meant waiting patiently, receptively, for the issue to resolve itself. Thinking something new, or even just seeing something you hadn’t seen before, isn’t an event you put into your calendar and then dutifully show up to participate in.

But writing is different. When you are actually doing it, writing is something quite specific. The idea you want to express comes to you and your fingers produce the words, either on the page or on the screen. As I said in a previous post, this happens at something like the “speed of thought”, though we might qualify this by saying that we think a bit more slowly when we are sitting in front the machine than we would if we were going for a walk or, say, engaged in conversation. But it’s still an immediate experience: having a thought and expressing it occurs essentially simultaneously, in the span of few seconds.

Paragraphs, by contrast, take longer to write, with a number of ideas occupying your attention at various times, sometimes recurrently, sometimes simultaneously. Crucially, when you are composing a paragraph you are putting ideas together, which is to say, you are working with several sentences at the same time (in the same moment). It’s not so much the ideas you express in each sentence that matters, but the relationships that you establish between them. One of the sentences, for example, will be your “key sentence” and all the others will be organized around it — supporting, elaborating or defending it. This, in turn, means they have to work together, which is not a simple of matter of making sure they’re all true. Sometimes two true sentences imply opposite conclusions unless they’re each properly contextualized in their own paragraph. That’s why you need significantly more time to write a paragraph than a sentence.

An essay is an arrangement of paragraphs that are composed of sentences. You can work on their arrangement separate from the writing and the composition by making yourself a key-sentence outline and confining yourself to putting them in the right order, perhaps tweaking their scope by changing a word or adding a qualifier. Revisions like this shouldn’t occupy more than a few hours of your time on any given day. Remember that it is possible to overthink the structure of essay, imagining that it does more work than it is reasonable to expect of the reader. After all, the logical structure of an essay must be able to fit into the mind of an ordinary, academic reader, who is reading it paragraph-by-paragraph, minute-by-minute. (It takes about one minute to read a 200-word paragraph.) An essay or research paper occupies about an hour (often less) of your reader’s attention. Your own image of it as a writer should not be more complicated than that.

Nabokov used to tell his students that “a good reader, a major reader, an active and creative reader is a rereader.” By a similar token good writers, serious writers, even if they’re minor and not especially creative writers like us, are rewriters. This raises the interesting question of how “revision” fits into the writing process.

Revision of course plays a major role in writing instruction. (See, for example, the late, great Scott Eric Kaufman.) But this doesn’t mean the same thing for all writers. Anne Lamott famously promotes the idea of writing “shitty first drafts”, with the implication that good and great writing only emerges through revision. Jonathan Mayhew, however, has challenged this idea. “At some point the writer needs to learn how to write,” he argues, “not just revise.” I side with Jonathan on this. Especially for academic writers, I don’t think it’s a good idea to let our ideas tumble out of us inchoately (and often simply incoherently) onto the page, leaving it to our future selves to clean up the mess we’re gleefully making of the present. I think we should sit down and write the best paragraphs about things we know that we’re capable of in the moment. Like Jonathan says, that’s the only way we’ll actually become good writers.

The first drafts will not be perfect, but they will not be shitty either. Revision will still be necessary, but much less. The experienced writer will avoid certain kinds of mistakes on the first draft. After all, if you can’t write, you won’t be able to revise either, because re-writing is still a form of writing. A good writer needs to know the feeling of composing a good sentence, once in a while, on the first try. I’d much rather have a model that emphasizes the development of the writer’s competence.

Too many writers fail to really enjoy this “feeling of composing a good sentence”; they don’t give themselves the time to really experience what writing is. Instead, they write their first draft to a low standard (having deliberately freed themselves of a higher one) and then struggle valiantly to clarify their own confusion in later versions, often leaving them exhausted and uncertain they ever knew anything at all. This is a tragedy because a good writer should feel sane and strong while writing — perhaps the sanest and strongest they can be. Writing a scholarly paper should, for the most part, be an entirely lucid experience — the experience of knowing what you’re talking about. It should not be the experience of roughing something out that you’ll have to polish up later.

So where does that leave revision? Well, if you’re following my rules, your draft (at any stage) will consist of a series of orderly paragraphs, each of which says one thing and supports, elaborates or defends it. You can identify the paragraph by its key sentence and position it in an outline. Already there, you can “revise” (or re-vision) your text simply by imagining the same paragraphs in a different order, the same claims arranged in a different sequence. You can also imagine the claims sharpened a little or broadened a little — giving them a more definite point or a wider scope. And you can make separate, independent judgments about whether the reader will find these claims hard to believe, understand or agree with. Now you’ve got your work cut out for you.

Importantly, this revision of the key-sentence outline has set up a re-writing task. Print out your current draft (perhaps just as a PDF to open in a separate window) and make a separate document with all your (revised) key sentences in the (revised) order you’ve settled on. Beside each sentence there will be a simple marker (“S”, “E”, “D”) telling you whether to support, elaborate or defend it. You’ll probably have around 40 paragraphs if you’re writing a standard paper in the social sciences. That means you’ve got 20 hours of work in front of you. Make a plan to (re)compose each paragraph carefully and deliberately in a 27-minute writing moment, using the material in your original draft (where an earlier version of the paragraph already exists) as your notes — perhaps supplemented with other notes or sources. If 27 minutes seems like too much time (perhaps this is your third or fourth revision?), try 18 or 14 minutes. But in all cases physically re-write the paragraph. Type it in. Don’t copy and paste. Pass your text through your own hands. It will be good for your style.

I don’t just mean that it will be good for the style of the text you’re working on right now. I mean that it will make you a better writer. It’s not that all good writing has been rewritten many times in this way. (Some good writing does in fact emerge from the revision of shitty drafts.) It’s that good writers are the product of continuous, deliberate practice. They have written and rewritten the same ideas many times. They have become good at writing down what they know for the purpose of discussing it with other knowledgeable people. And, as Ezra Pound reminded us long ago, “one definition of beauty is: aptness to purpose.” Don’t just improve your text through revision. Improve your writing through rewriting.

“…it was to me as if someone were to use a precision instrument to open crates.” (Hermine Wittgenstein)

Let’s say that we write sentences, compose paragraphs, and arrange essays. I know that we often say that we “write” paragraphs and essays, too, and even whole books, but, if you think about it, it’s actually only ever the sentences, properly speaking, that we write. The experience of expressing a thought is not the same thing as that of supporting, elaborating or defending a claim to know something. Nor is the experience of reading a sentence the same as that of reading a paragraph. With each sentence a thought comes into view, and a paragraph puts these thoughts together. Now, some writers, to be sure, challenge us by putting all their thoughts into one sentence that runs on for the whole paragraph. For them, writing and composition are essentially indistinguishable. But very few writers let their sentences run on for a whole essay or chapter or book. (You’d have to be the daughter of James Joyce’s biographer to attempt such a thing!) Still, even if you pepper your papers with periods, you may not be writing, composing, and arranging as deliberately as you could. Let me try to explain.

Sentences are written and read at more or less the speed of thought. A sentence expresses a thought — it says something — and writing a sentence is basically a matter of deciding what to say. You do well to notice what “writing” is in precisely this sense. Try it. Have an idea and write a sentence that expresses it. Now have another. Write another sentence that expresses another idea. You can keep going as long as you like. You can try bigger ideas, more complicated ones, which will need bigger words or longer sentences or both. In any case, a “thought” is the sort of thing that can be put in a sentence, which, if someone were to read it, would, ideally, occasion that same thought in their mind. (If it can’t be expressed in a complete sentence it isn’t a finished thought.) Each sentence in this post is my attempt to get you to think what I’m thinking as I write it. If you’re thinking about sentences and about writing them, I’m doing my job.

Paragraphs are more complicated because they require the reader to put several thoughts together. A paragraph takes about one minute to read, after which the main claim will be believed, understood, or rejected by the reader, who will then move on to the next paragraph. The main claim is itself expressed in a sentence, which we call the “key sentence”, and which also of course expresses a thought. So you might think that a paragraph, too, expresses a thought, just using sentences instead of words. I urge you to distinguish between a thought here and a claim. It is one thing to think of something and quite another to assert a claim. The thought that is expressed in your key sentence is a claim that uses the thoughts expressed in the other sentences in the paragraph for support, elaboration or defense. It’s important to make sure that each sentence can be thought on its own, even if some of them, and especially the key sentence, suggest difficulties that require further thought. The paragraph puts all those thoughts together in a composition that resolves these difficulties in some interesting way.

Needless to say, the arrangement of paragraphs into essays is a still more complicated business. Each paragraph has to be composed and then arranged, and then, usually, rewritten, sentence by sentence, to occupy its final place in the essay. But given that the sentences are well written, and the paragraphs have been carefully composed, “writing” an essay is really just a matter of arranging a series of claims in the right order. These claims are represented by your key sentences, which can be arranged in a surveyable way simply by listing them one after the other. You can then ask yourself, “If I were to successfully support, elaborate or defend each of these claims, and presented them in this order, would my reader find my argument compelling?” An essay, remember, is ultimately just an attempt — you’re trying out a line of argument on your reader. The question is whether you can see see it clearly just by looking at your key sentences, as “dots” to be connected by the “lines” of your paragraphs. Obviously, your reader will have to make those connections too, so you must arrange your claims in an accessible order, with direct lines from one point to the next as often as possible.

While I recognize that this distinction between writing, composing and arranging is somewhat artificial, I think much of the difficulty of writing stems from not observing it in practice. We think we have to “write” our essays in the same sense that we “write” our sentences. This suggests that we have to have the whole essay “in mind” in the same way that we have the thought that a sentence expresses in mind as we write it. Or we may think that we should compose sentences with the same care that we compose paragraphs, meticulously considering each word as it passes through the mind of the reader. Or we think that a paragraph is simply an arrangement of sentences — merely a series of thoughts to get through — rather than a moment of the reader’s intellectual composure. Some of these misunderstandings make writing more difficult than it needs to be, while some of them produce writing that is harder to read than it should be. Writing well means knowing when to just write, when to compose, and when to make an attempt at arrangement. By making the right decision about what to do, not only are we more likely to succeed, we’re more likely to enjoy the work. And that’s actually the best reason to think carefully about what you are doing when you’re writing. It will make you happier while you’re at it.

“Appreciate your finitude.”

Each of us knows a great deal. And there is a lot that we don’t know but is known by others. These are the things we learn by doing our literature review. Many researchers find this task daunting and, in this post, I want to try to summarize the advice I give to them.

Begin with a small set of papers that you have a close connection to. You may have authored or co-authored some of them yourself, or you may have worked closely with the people who have authored them. They may be written by members of your department or people who have collaborated with your colleagues. Ideally, you have met the authors of these papers and have a good understanding of the sort of research they do. You have a concrete image of where they work and how they go about their research, and you are, of course, at least moderately impressed with what they have accomplished…

(Note: If you are a student, you will not, of course, have such an intimate relationship with these papers. But you’ll recognize them because they have come up in your classes, and you’ll sometimes get the sense that your teachers are familiar with them in this way. They are central part of the discipline you are studying or, more precisely, the specialty that the class is focused on.)

When you read one of their papers, you can form clear images of the facts they represent and you have a well-grounded opinion of how likely their results are to be correct. You don’t have to be entirely convinced of their claims. But you do have to find them interesting, and you have to respect the research they have done. This list of papers can be quite short; even two or three can be sufficient, but I would suggest finding between six and twelve. Put them at the center of your search.

Locate them individually in the various databases that you have at your disposal. Here at CBS, at would recommend you find them in Business Source Complete, Web of Science, and Scopus.* Once you have found the articles, look at the bibliographical data that the database provides for them; notice that there is an abstract, some key words, a journal, and of course the authors. These are all going to be useful to you when you are looking for similar articles, since “similarity” here is just a reference to the data that the entries in the database share with other entries. Using your awareness of these similarities, you will be able to design searches that quickly and accurately locate literature that is relevant to your research.

But the entries, especially in Web of Science and Scopus, also let you think of each article as occupying a node in a network of citations. In addition to the bibliographical data, an entry usually contains the full bibliography of the article itself (its reference list, its list of sources, its “cited references”). It also links to a list of “citing documents”, i.e., all the articles in the database that have this article on their reference lists. Finally, they give you the option of searching for “related documents,” which is a long list of all the articles that share at least one reference with the article in question. (These can be organized by “relevance”, i.e., in order of how many references they share.) With this information, you able are able to position an article in the discourse to which it contributes.

While a citation network may be large, it is not infinite. More importantly, it has a center and a periphery, as well as a past and a future. In the center is that short list of articles with which you are already very familiar and should be quite recently published. Behind them lie all their sources, which we can imagine as one long, shared bibliography; but, instead of putting it in alphabetical order, let’s arrange them by publication date, with the oldest at the furthest remove from your core papers. Also, let’s group them along a center line so that the papers that share references (or are otherwise topically related) are closest to each other, and those that don’t are further from the line. This will produce a sort “wake” trailing after the core articles and shading off into the “ocean” of published papers that are not directly cited but are nonetheless related to the articles you do cite. By a similar token, there will be papers that are contemporaneous with your core, and very relevant, but (since they were published at around the same time) will not cite your work (or that of your close colleagues). And there will be papers that are at quite a remove (either conceptually or bibliographically) from your core, but more or less related to the papers you cite, or the papers that are related to those you cite. An illustration might be useful:

But nothing is entirely static in research. The literature is constantly growing and it’s important to have a way of navigating on the ocean of scholarship. Obviously, your own research (and that of those core peers, whose research you are very familiar with and probably read about in draft and pre-print form) will follow a line proceeding from the (present) center of your research. Closely related to this line will be (future) papers that cite your core articles (albeit sometimes just in passing) and, again, the (future) papers that are related to those articles by way of shared references. We now get a diagram that looks a bit like this:

The key to all of this is to appreciate the finitude of the problem. While there is a lot to know, there are only a handful of very relevant articles, and a manageable amount of less relevant ones, before we finally reach your research horizon, beyond which you can safely remain ignorant of what goes on (until you get there yourself, of course). The ocean is a big place, but you’re only ever sailing in some local part of it. Don’t be afraid today of what is really just part of the adventure tomorrow. “Here be dragons,” is a myth born of ignorance. Uncharted waters are just places you haven’t yet explored.

____________

*Update: Patrick Dunleavy, whose views on writing and publishing are always worth taking seriously, has taken me to task on Twitter for this recommendation, calling it “outdated”: “Leaving out Google Scholar,” he argues, “‘costs’ your advisees. They’ll miss all grey literature (pretty vital in business studies on new topics), plus all books, book chapters & most conference papers.” Jo Jordan came to my defense: “But starting them on Google Scholar,” she countered, “condemns them to muddle and an inability to use IT to manage their databases. Need to see some total costings here.” I think that gets it mostly right and it reminded me of what Bill Evans says about keeping things “simple and real”. That’s the spirit in which I wrote this post. But I’ll have to write one specifically about Google Scholar at some point.