Sometimes I think the education we dispense is better suited to a fifty-year-old who feels he missed the point the first time around. Too many abstract ideas. Eternal verities left and right. You’d be better served looking at your shoe and naming the parts.

Father Paulus (in Don DeLillo’s Underworld)

As adults approaching a foreign language we sometimes forget how easily children learn languages. We forget how easily we speak our first language. I wrote this post over fifteen years ago in response to a comment Jim Collier made to one of the first posts at my previous blog, Research as a Second Language, about the “intellectual style” of academic writing. Roey Elnathan’s column in Nature, “English is the language of science,” reminded me of it, and, rereading it today as a fifty-year-old who sometimes thinks he might have missed a point or two, I was surprised to find that I completely agree with my younger self. I even like the cut of his style! See what you think:

Abstract ideas are an unavoidable component of academic writing. One way to draw the line between empiricist and rationalist attitudes to abstraction is to distinguish the sorts of writing that count as “mastery” of the ideas in question. A rationalist will grant that you understand a given set of abstractions if you are able to correctly deduce other abstractions from them. An empiricist will generally expect you to be able to describe concrete particulars that fall under or are subsumed by the abstraction.

One of the reasons that academic English can be difficult to master is that academic discourse is largely a rational enterprise. That is, academics are expected to combine words and phrases in especially orthodox ways more often than they are expected to describe a particular matter of fact. This is not in itself a problem. I don’t want to suggest that one has to be an empiricist in order to write good English, nor that there is something fundamentally wrong with academic writing. I am simply trying to indicate a form of exercise that can be useful in developing one’s style, and which may even be pleasurable in its own right.

I want to suggest that your language can be more or less empirically sensitive. Your native tongue has, if you will, a broad palate. Through it, you can describe very ordinary, very personal situations and very exceptional, very impersonal ones. And you will be able to describe a whole range of situations in between. But if you are working with English primarily as a research idiom, there will be a region of insensitivity somewhere between your ability to buy a train ticket and your ability to articulate the consequences of deconstruction for management studies.

Roughly speaking, this is the region occupied by illustrative examples of abstract ideas.

The way to make this part of your English more sensitive to the things you learn is to use it. In the course of your research you will have a variety of experiences about which you will discover yourself to be more or less articulate. You will discover this by experiment. You will find yourself having to tell a story, to describe a scene, to name the parts of a given object.

In academic life we too often confine our expression to a relatively small set of abstract gestures, indifferent to the detailed state of particular affairs. Native speakers suffer less for this because they have their empiricism always on hand in their daily routine. It is, as it were, ambient. To get by at a basic level, and, more interestingly, in that intellectual mezzanine of the academy provided by the classroom, they engage in story-telling and description and naming with the enthusiasm of children. This, I would argue, is generally good for their style.

Non-native speakers will have to work at it more consciously. Faced with a set of abstract notions, you do well to describe concrete situations to which those notions may be applied. You do well to write anecdotes that illustrate your ideas, and to write detailed descriptions of objects that can be subsumed under them. Faced with concrete experiences you likewise do well to set them down in writing in such a way that you might recognize them later.

A good exercise here is to describe a “text book case” of your favourite abstraction in concrete terms (using no abstract terminology). Then show it to one of your colleagues and see if they “get it”. Throughout this process be on the lookout for lacunae (holes) in your language. Whenever you are at a loss for words, make an attempt to find them. Go looking for the phrases and constructions you need to illustrate your ideas. These will become elements of your style. It is a matter of building up a language that is able to articulate the rich texture of the research experience, not merely to trace the rough outline of its results.

Data are “facts given or granted”. They can be distinguished from the facts that have to be established by your analysis. But we mustn’t forget that the data were themselves by no means easy to gather. Once we have our data, we can take them for granted in our analysis, but they first had to be wrung from experience through our procedures. We explain those procedures in our methods section so that our reader may come to trust us.

Many years ago, as a young philosopher learning about the social sciences, I was speaking to a fellow doctoral student about his research. I don’t remember what his research question was, except that it had something to do with “knowledge management” and that he was approaching it as an empirical problem. What I remember very clearly is how he talked about the survey questionnaire he had designed; he called it a “measuring instrument”. He had taken something that, for me, was an abstract and subtle phenomenon (knowledge) and turned it into something that could be measured. The answers to the questions on his survey (whether yes/no or on a scale) would give him data points for analysis. These data points would become facts he could take for granted, and on their basis he would establish his conclusions.

“Some Huxley or Haldane,” said Ezra Pound, “has remarked that in inventing the telescope Galileo had to commit a definite technical victory over materials.” He had to design and construct a “measuring instrument” to make his observations. But as has been pointed out by philosophers and historians of science for some time, his data were not immediately trusted by his fellow astronomers. He had to explain how the instrument worked before they would believe that what they were seeing through it were really, say, the phases of Venus as it circled the sun, not the Earth. He had to get them to understand how his equipment worked before they would grant him his facts. He had to commit a rhetorical victory too, we might say.

In your discipline, the methods you are using are probably well-established. You are using semi-structured interviews, or participant observation, or survey questionnaires, or archival documents, or you’re pulling data from trusted databases. (Need I emphasize that it’s called a data-base? A source of facts that can be taken for granted.) Your readers understand what you are talking about when you describe what you did. They will have a good sense of what sort of material you were working with in your analysis after you tell them that you conducted, recorded, and transcribed 27 two-hour semi-structured interviews. They’ve done this sort of work themselves so they know what you were looking at when you were “coding” them, just as an astronomer knows what it’s like to look through a telescope.

Of course, your peers also know what can go wrong. They know that even very good scientists can make mistakes, both in their collection of the data or in their analysis of it. So you have to demonstrate an awareness of the “sources of error”. You have to assure the reader that you did everything you could not to fall into traps that are familiar to them. These days, you may even have to persuade them that you did not abuse your “degrees of freedom”. The whole point of having data is that they can be taken for granted, so you don’t want there to be any doubt about how you came into possession of them. You want the reader to be able to trust you. The reader wants to be able to trust you.

It is precisely because you must be able to take your data for “given” that you can’t take your reader’s trust for granted. It’s hard work to make credible measurements and observations of a complicated reality. You must tell your reader that you did that work and that you understand its importance. A good method is no mean feat. Your methods section should be written with a sense of accomplishment.

…a pattern of loss and of finding that so compels us that we are entirely engrosst in working it out, this picture that must be put together takes over mere seeing.

Robert Duncan

For the past few weeks I have found myself in a bit of a slump. I was stuck intellectually without really knowing what was wrong. It was like there was something just beyond the reach of my understanding that I wasn’t able to grasp, and it seemed somehow very important that I grasp it. Then, suddenly, the fog lifted and I could see clearly what I had been struggling with.

I can identify almost the exact moment that things fell into place. “Did something happen to the connection between language and experience?” I had tweeted. I meant this in the same way that we sometimes ask a family member or colleague, “Did something happen to the internet connection?” I was trying to be funny, to be sure, but I also wanted to engage in a little cultural criticism. I can’t be the only one who senses that something is amiss, and what’s Twitter for if not to share our exasperation with the world? Anyway, it was already late, and I was soon in bed and sleeping. But by the end of the next day I had my epiphany. “Now I can go on,” as Wittgenstein used to say. Let me try to explain what happened.

Along with Oliver Senior’s hands, Hemingway’s iceberg is probably the metaphor that dominates my thinking about writing most aggressively. I have used it to organize my teaching of academic writing at all levels, and across a variety of disciplines, and I think writers generally find the image useful. “The dignity of movement of an iceberg,” said Hemingway, “is due to only one-eighth of it being above water.” Most of the knowledge that we base our writing on is not visible to the reader on the surface of our text. But, “if a writer of prose knows enough about what he is writing about,” Hemingway pointed out, “he may omit things that he knows and the reader, if the writer is writing truly enough, will have a feeling of those things as strongly as though the writer had stated them.” It is that knowledge that gives our writing its dignity.

The ideal Hemingway story is a “sequence of motion and fact”. It consists of descriptions of what is happening and what is the case, what the characters do and see, as the action unfolds. These descriptions are what T. S. Eliot called the “objective correlatives” of the emotions that a story is trying to convey. The modernists were sometimes outright purists about such matters and insisted that only these descriptions were legitimate in art; there was no room for explanations, as Wittgenstein also suggested. “Show, don’t tell,” as the saying goes. Don’t tell us the hero was angry, or happy, or sad; show us “the sequence of motion and fact which made the emotion.” Make us feel it too. That was Hemingway’s goal and he knew it would not be easy.

“A writer’s problem does not change,” he said. “He himself changes, but his problem remains the same. It is always how to write truly and, having found what is true, to project it in such a way that it becomes a part of the experience of the person who reads it.” I like this way of putting it, and when I present Hemingway’s iceberg I normally do so in terms of experience rather than knowledge. After all, Hemingway wrote stories, not research papers, and I want to make clear that academic writing is not literary writing but stands in a special relation to this thing we call “knowledge”.

So I sometimes get writers to first imagine telling a story from their own ordinary lives. “What gives such a story its dignity?” I ask them. The obvious answer is: the experience that that the story is based on. If we depart too far from what we actually saw and did (if we start to embellish the story, as we are wont to do) our story may become more exciting, but we are risking its dignity. Someone who has done and seen similar things may find our story implausible.

Then I ask them to think about what makes research papers (and dissertations) different. What plays the role of experience? That answer, of course, is “knowledge,” and this then lets me open a rather big can of worms: What is knowledge? I try to keep things practical and manageable, but what follows is about forty-five minutes of working through the philosophical, rhetorical, and literary dimensions of academic knowledge. Alternatively, I work through the different kinds of knowledge (or the various sources of knowledge) that the individual sections of their papers are based on. I coordinate what is above the surface with what is below.

The epiphany I had this week, however, offers me a more natural transition from Hemingway’s iceberg to mine. It has to do with something I had begun to say already last year, namely, that the methods section is the part of a research paper that is most directly based on your experience; in fact, it’s the section that is most analogous to a short story. Your methods consist of what you did to collect your data. In fact, it’s the story of how you converted what William James called (and Thomas Kuhn recalled) the “blooming buzzing confusion” of experience into data points for analysis. It’s the sequence of fact and motion, if you will, that “makes” your concepts, your measuring instruments, your categories of observation.

Instead of just watching people work, you observed them and recorded these observations carefully in your field notes; instead of just listening to them talk, you conducted semi-structured interviews, recorded them and transcribed them, making them amenable to coding. The data that results has a certain order and formality, which makes it easier to work with in your analysis, than, say, your memory of what happened and what was said, but the experience of collecting the data is just as ordinary and “lived” as any other thing you do during the course of the day.

This is where I now intend to begin my writing instruction. And that’s also the answer to my vaguely rhetorical question on Twitter. Did something happen to the connection between language and experience? Yes. Method happened. The twentieth century has seen the rise of methodologies that increasingly get between our language and the experiences we use it to talk about. This is essentially the same thing as the rise of the social sciences and, by their means, the displacement of literature (poetry, novels, short stories) as means to understand the “human condition”. The imperative of method has made us unsure whether we can just talk about what we saw and what we did, what happened to us, what we think. Perhaps this was a necessary imposition of rigor on our otherwise hopelessly anecdotal lives. But could it be that it has gone a little too far?

I’m going to devote a few posts to this question, which, as we will discover, is the question beneath all the questions that has occupied me for well over a decade. I’m looking forward to seeing where this leads.

[Note: this post is part of a series on the “substance of the craft” of scholarly writing. Inspired by by Wayne Booth (co-author of The Craft of Research) and Oliver Senior (author of How to Draw Hands), I argue that composition is the coordination of words and ideas, paragraphs and propositions, sentences and the state of things.]

A drawing reveals an understanding of the subject. Different artists understand their subjects differently, as can be seen by comparing Oliver Senior’s understanding of his own knuckles to Paul Cézanne’s understanding of Ambroise Vollard’s. There is no question that both are capable artists, but they obviously had, as Senior might have put it, different things “on their mind” when they were rendering the hand they were looking at. Senior helpfully includes his understanding of the skeletal structure of his hand (though he is careful to remind us that those are not his bones). But his aim was also more didactic than Cézanne’s, and we must remember that Vollard’s hand is but a detail in a full portrait, whereas Senior’s emphasis is very definitely on the hand itself, alone. Both artists, in any case, are trying to show us what they see.

I hope you will agree with me that Cézanne’s hand, while less detailed than Senior’s, is just as as well-informed (to use Senior’s language). It serves its purpose in the portrait but it doesn’t cut corners; it is “true” to human anatomy; it is an entirely possible hand in an altogether plausible position. Cézanne has not sidestepped the difficulty of drawing a hand by replacing it with a mere symbol — a bent fork or a bunch of bananas, for example. If he had, we might still understand that Vollard is holding a book in his hand, but we would not get the same overall sense of his humanity. It would be getting the facts right, but it wouldn’t feel right. In fact, the detail that I’ve cut off from its context for this post might not make any sense at all on its own if it wasn’t so well drawn. We might literally mistake it for a fork if we didn’t see it at the end of an arm. Somehow, however, using only a few simple brushstrokes, Cézanne gives us a flesh and blood (and, yes, even bone) human hand. It’s worth appreciating that accomplishment.

Now consider the sentences in your paragraphs. Think of a sentence in a paragraph describing a larger fact, or a sentence in a paragraph that is presenting the conceptual model you have built to guide your analysis of your data. You know the overall picture and you know how the parts fit together. That is, you know what you are talking about because you have studied it carefully. But do you know how to render each detail so that it conforms also to the underlying, unseen structures that constrain reality, the actual and possible state of things? If you read the sentences separately, out of context, do they still convey something? Do they feel true to the facts or the models you are writing about? Obviously, they won’t capture the full meaning of what you are trying to say in the paragraph. But are they utterly dependent on the context to mean anything at all?

Academic writing is a highly contextual and conventional affair. We often rely on our readers to play along with our use of technical jargon to communicate our ideas. It’s a bit like the nervous tick of saying, “you know,” after every sentence in conversation. We keep asking our reader to do the work of putting the image together because we know we’re not doing a very good job of explaining it ourselves. It’s not there on the page as the modernists used to say, even if it is somehow available in the discourse between us. More often, however, the reader doesn’t do the work we hope they will. Instead they simply let us get away with being vague, leaving them none the wiser about what we really mean, but happy to grant that we know our stuff. We have persuaded them that we’re smart (and we probably are) but we have not actually communicated our ideas to them. We have certainly not opened ourselves to criticism from our peers.

I am trying to persuade you to work on your sentences. Think of them as details in your pictures of the facts. Each sentence expresses a thought, and thoughts, properly speaking, must obey the rules of logic (just as hands must conform to their skeletal structure). To draw a convincing human figure, you must be able to draw the head, the torso, the legs, the feet, the arms and hands. And to draw a hand, you must be able to draw fingers and thumb and palm and wrist. Likewise, your model of organizational culture may include artifacts, values, and assumptions. But can you write two sentences about each of these notions that distinguish them from the ordinary, commonsense meanings of the words? Do your sentences capture, not only the sense that Edgar Schein gave to them, but also their underlying interconnection, i.e., the way they “add up” to a culture? Can you put all this together in a single paragraph of seven or eight sentences whose key sentence is “Edgar Schein (2017) posits three levels of organizational culture”? Can you then isolate each sentence and make (at least some) sense of it?

Try it. And then try to describe an actual organization in those same terms. “The organizational culture of XYZ Corp. was disintegrating.” The absolute state of it! Write two sentences about the state of the artifacts, another two about the company’s values, an then a couple about its assumptions. Make them as clear as you can and as detailed as you need. Make sure that the paragraph supports your overarching claim of that the company’s culture is in trouble; but make sure, also, that each sentence alludes to the underlying structure that is coming apart, the logic of XYZ’s disintegration. Again, write seven or eight sentences. Try long ones and short ones and simple ones and complex ones. “Try them all,” as Senior says of his materials (pencil, charcoal, pens, brushes, and all kinds of paper). “Each one will be a means of revelation concerning your subject. And the only true comprehensive answer to as to ‘How to…’ is the simple instruction ‘Get on with it.'”

[Note: this post is part of a series on the “substance of the craft” of scholarly writing. Inspired by by Wayne Booth (co-author of The Craft of Research) and Oliver Senior (author of How to Draw Hands), I argue that composition is the coordination of words and ideas, paragraphs and propositions, sentences and the state of things.]

“We picture facts to ourselves,” said Wittgenstein. And a thought, he explained, is a “logical picture of facts.”

Think for a moment of your hand. (You can take a look, of course, but the important thing for my purposes is that you think. If you have to close your eyes, that’s okay too.) The hand, Oliver Senior tells us, is “a familiar yet highly complex piece of physical mechanism which is almost infinitely adaptable in use.” Imagine your hand open, held out, perhaps, in greeting, or imagine your fist clenched in anger. Picture your hand in these positions. Now, imagine it picking up and holding a delicate instrument. Your hand, says Senior, is capable of “multitudinous individual variations of character as well as changes of appearance in its great range of different positions, actions and movements, yet always conforming to a recognisable standard pattern and subject to strict laws of its own structure.” You can picture it, right?

With a little practice you can learn to draw these pictures. When you become good at it, you can represent the hand in its full variety of positions, always in accordance with the laws of its structure. You learn to see the hand “objectively”, constrained (and enabled) in its movements by its physiology, its reality. Your drawings become increasingly “realistic” because each drawing can be traced back to a recognizable pattern, to which it conforms. Each drawing is logically compatible with the other drawings — it’s simply a transformation of the possibilities in the others. Unless you imagine that your hand is broken (which you are free to imagine if you dare), certain pictures are impossible, unimaginable, unthinkable.

“In a proposition a thought finds an expression that can be perceived by the senses,” said Wittgenstein. That is, a proposition expresses a logical picture of facts. “My hand is empty,” is a proposition. “My fist is clenched,” is another. (“My finger is broken,” is a third.) You can imagine my hand (or yours) in these situations. “Objects contain the possibility of all situations,” said Wittgenstein. “The possibility of its occurring in states of affairs is the form of an object.” We can represent these possible states of affairs (these possible facts) in pictures and in propositions. We can draw them or we can write them down.

In scholarly writing, we arrange propositions (expressions of thoughts, i.e., sentences) into paragraphs that together represent a situation, a state of affairs, a set of facts. They may be facts about a company, a country, or a current event. They may be facts about the state of the literature or the methods you have used to gather data. They may be very abstract facts about the concepts you have used to model reality. But in all cases, you are claiming that something is true and, importantly, that a great deal more is possible. A paragraph, like a picture of a hand, doesn’t just confine itself to how things look right now, right here. A paragraph conforms to a “standard pattern” of reality so that the facts that are represented also imply the possibility (and impossibility) of other facts. The paragraph presents a logical picture of the facts; it expresses a series of thoughts about them. The reader must be able to imagine it.

This underlying logic of paragraphs also has implications for their arrangement. As readers pass from paragraph to paragraph, their imaginations are exposed to pictures of various facts. But these facts must conform to the pattern that is being established in the text. It must be possible to imagine what Hemingway called the “sequence of motion and fact” that brings us from one state of affairs to another, without breaking the bones, if you will, of the structure of the argument. (Of course, you can write a paragraph where those bones are explicitly broken if you need to. Now the reader will be able to imagine the painful contortions to come. Please remember to be as kind to your reader as you can.) Your concepts and objects will often be highly complex* mechanisms, but they should be familiar to you, and they should become increasingly familiar to your reader. Every time you depict them in a paragraph, they should become easier to imagine.

Try not to make this more difficult or philosophical or even mystical than I may, unfortunately, have made it sound. A paragraph consists of at least six sentences (propositions, thoughts) and at most 200 words. Like a drawing, it usually occupies no more than a page. You can read it in about a minute, just as you can study a drawing for that long and get some useful information out of it. Just as a drawing isn’t the whole or final truth about your hand, but only your hand in a particular position in a particular situation, a paragraph asserts only particular truths about the concepts or objects it describes. The important thing is to make a definite claim (cf. “this is a fist, clenched in anger,”) and then support, elaborate, or defend this claim within the space of two-hundred words (cf., “in each purposeful line or passage of [your] drawing, achieve … an expression of form, character, action …”). Appreciate the finitude of the problem and develop your craft from a realistic point of view.

“The world is everything that is case,” said Wittgenstein. “Objects contain the possibility of all situations.” And, yes, “What we cannot speak about we must consign to silence.” He said that too.

_____

*A note for Wittgensteinians: I know that Wittgenstein says that “objects are simple” (T2.02). But in ordinary language objects can be more or less complex. The early Wittgenstein would refer to what I (and I assume most of my readers) think of as “objects” as “facts” at this point and distinguish “atomic” ones from the rest. Wittgenstein of course ended up abandoning the idea of “atomic facts” composed of simple “objects”. But I have retained the idea that the difference between ordinary things and objects of inquiry is that the latter “contain the possibilities” of their combination with other objects, while “things” are just lying around, if you will, taking up space. (Objects occupy logical space, let’s say.) I know that’s not particularly rigorous and won’t satisfy all philosophers, but hope you’ll acknowledge that I’m making an effort. Even after all these years, my reading of Wittgenstein continues to evolve. May yours, also.