How to Draw a Fish

No man is equipped for modern thinking until he has understood the anecdote of Agassiz and the fish.

Ezra Pound
ABC of Reading

If ... the artist finds himself constrained, by any consideration of expression, treatment or style, or by his deference to the peculiar nature and limitations of his tools and materials, to adopt or invent a convention or a symbol and to substitute the semblance of a bunch of bananas or a bent fork for a representation of the human hand, then the particular problem dealt with in this book does not arise.

Oliver Senior
How to Draw Hands

I'm helping to coordinate a PhD course later this week. This post is an attempt to articulate my teaching points for the group work.

In the February issue of the Academy of Management Journal (v50, n1, 2007), Karl Weick makes an argument for "the generative properties of richness". He introduces his piece with the famous story of Agassiz and the fish, which you can read here in full. (The essential elements of the story can be found in Samuel Scudder's account, also available here.) In a nutshell, the zoologist and geologist would ask his students to describe a fish without the aid of (1) special equipment, (2) talking to anyone, (3) reading anything. They were to use only their hands and eyes and were encouraged to draw the fish in great detail. We might say that they were to engage in atheoretical description.

They were not to label the parts of the fish with their latin names, nor regurgitate accounts of their evolution and bodily functions from reference works. They were simply to look at the fish and describe what they saw and eventually to compare one fish with another. Moreover, as Scudder points out, "Agassiz's training in the method of observing facts in their orderly arrangement, was ever accompanied by the urgent exhortation not to be content with them." There was always more to be seen.

Lane Cooper's account (published in 1917) seems to have informed the poet Ezra Pound's statement (in the epigraph above). But it also jibes nicely with any artisan's attitude to his basic skills more generally. Thus, Oliver Senior has written a wonderful little manual about drawing hands (also quoted above). He describes the hand as "a familiar yet highly complex piece of physical mechanism" and correlates "the notorious difficulty of drawing hands" with "the mental equipment by which [the student's] vision may be directed, extended and refreshed."

The better draughtsman has more "on his mind" concerning his subject; and, by embodying his knowledge and understanding in each purposeful line or passage of his drawing, achieves with apparent—or even real—ease an expression of form, character, action—whatever may be his immediate object—that the novice, lacking such equipment and relying on vision alone, finds beyond his power.

This "better draughtsman" was of course what Agassiz also wanted to encourage his students to become.

Note Senior's emphasis on the improvement of one's vision. Learning how to draw a hand improves your ability to see hands as such. On this background, with that much more "on your mind", you are able to detect the significance of positions and gestures of the hand of your model, and can therefore incorporate the hand naturally and informatively in your drawing.

An organization or management context is also, of course, "a familiar yet complex mechanism"; and management theory, in whatever form you may be pursuing it, is "the mental equipment by which your vision may be directed, extended and refreshed." Management theory is what you "have on your mind" when looking at something. And it may be useful to describe simply what you see rather than the theory that informs your vision. What is your mental equipment doing for you? Are you equipped to think about modern organizations?

Indeed, Agassiz was apparently not wholly adverse to theorizing. "Facts are stupid things," Scudder recalls Agassiz saying, "until brought into connection with some general law." His aim was to ensure that his student grounded their general laws in observations of particular fact. He did not want them to develop the habit of replacing a concrete description with a symbolic label, "to substitute the semblance of a bunch of bananas or a bent fork for a representation of the human hand," as Senior puts it.

The lesson is simple: we should not resort to generalities simply because we are unable to describe specifics. You should always treat a general statement as an implicit claim that you can provide a specific, detailed and well-drawn example. The exercises we will be doing during the PhD course are intended to train precisely this art of connecting an empirical specificity with a theoretical generalization. I want to see if we can't help each other to achieve "with apparent—or even real—ease..." an expression of the form, character, and action of whatever management phenomenon may be our "... immediate object."

Writing on Demand

For some reason I thought the slogan "Publish or Perish", and its attendant polemic, was a product of the 1980s.* So I was surprised to see it used in a literary journal in 1960:

A great deal of harm has come out of the necessity for academics to publish as a means to promotion and to compete with their fellows in the domain of the physical sciences. Driven on by the same categorical imperative, 'Publish or Perish', they invent this drivel by the yard. (X: A Quarterly Review, vol. 1, no. 2., March 1960, p.159.)

This is precisely the sense in which the slogan is used today. My own view is that after half a century of complaining it may be time to approach academic publication in more constructive terms.

I don't want to deny that writing suffers when the writer feels the demands of adminstrators more strongly than the demands of readers. Indeed, there is an interesting tension here between the senses of "demand"—the demands of a boss vs. market demand. Ideally, authors should write to satisfy their readers, not their administrators (who aren't even editors or publishers of their work). But in reality, as the editors of X suggest, a great deal of writing is produced simply to keep one's job or to improve one's position.

I am certain that writing "drivel" more or less consicously, i.e., writing without a serious intent to satisfy the curiosity of an imagined group of readers, drains writers of the strength they need to keep the research process running, the writing process included. A long list of publications produced with this attitude may constitute a Pyrrhic victory.

Here's what I suggest instead. Let's accept that the only way your department head (whether present or future) can evaluate your work is to see that you are publishing and where you are publishing it. No adminstrator can reasonably be expected to evaluate the content of what you write. (Which is why jargon-ridden drivel will do just fine in most cases.) But this should not lead you see publication as an end in itself, nor even the means to an end.

The proper means to the end of winning time to pursue your own research interests (something approaching tenure, let us say) is not merely publishing your work but having it read. If you make it your business to find and maintain a readership, you will have no problem getting published. You will then satisfy both sets of demands as an ordinary part of communicating your results.

My point is that a sincere desire to be read is much more useful to you as a researcher than a half-hearted ambition to be published. Not "Publish or Perish", then, but "Find a Readership or Perish".

_________
*Update: The phrase is of course much older than I had thought. Eugene Garfield traces it back part of the way to its source (The Scientist 10 (12): 11, June 10, 1996. PDF). I will join the quest.

Deadlines, Line Editing and the Death of the Author

By "editor" I suppose you mean proofreader.
Vladimir Nabokov

The habit of sequential reading, which unifies the act of reading a chapter or paper in a single sitting, fosters the illusion of sequential writing. One imagines that writing a paper is a single act, a sweeping gesture, carried out by a unified consciousness with "something in mind".

The romantic poets sometimes claimed that a poem comes into being through a "spontaneous overflow of powerful feelings", a single flow of ideas onto the tranquility of a blank page. The truth is of course very different. Writing is a fragmentary process that binds words, phrases and paragraphs more or less seamlessly together. It is rarely done in a single sitting. It consists in drafting, revising and discarding, not whole texts, but delimitable passages of prose.

Romantics have in fact been caught editing their texts. Indeed, even the writing of essays and monographs is a social process. That is, writing for publication is writing for someone else. You are investing your time in a product that is only meaningful if someone else takes the time to read it. In some cases, you are producing something that is a demand on someone else's time.

I've talked before about how academic writing depends on the arrangement of facts, which indicates a spatial dimension to research texts. The facts that your text refers to impinge on it all at the same time. But there is also a temporal dimension to writing. It simply takes time to produce a text. What I want to do here is both to defend the virtue of organizing the writing process and to provide an indication of how detailed one can be about it.

The familiar (if ominously named) phenomenon of the deadline is often the most well-defined threshold in this process. But it is a mistake to think that the rest of the process is infinitely maleable. Writers, whose experience will teach them each to draw their lines in different places, must all develop a sense of what sorts of textual operations they should be engaged in during the hour, the half day, the day, the week and the month before their deadline arrives.

You need to feel no guilt about your editor because it is his job to read what you write. He therefore constitutes an excellent opportunity to practice your time management skills.

Consider two very different kinds of editing (which may be carried out by the same person): line editing and copy editing. The first provides a comprehensive critique of how the argument is made, what elements work and what elements do not work, whether the tone of the paper you are writing is appropriate, how the argument flows, and so forth. Ideally, it involves a line-by-line commentary that includes concrete suggestions for how to address the editor's concerns. The effects of line editing on a text can be quite dramatic. Copy editing, meanwhile, simply corrects grammar and punctuation when the manuscript has found its final form.

Researchers working with English as a second language should not think that the defects of their text can be fixed by mere copy editing. (Actually, no one should.) I often find that work that begins as "just fixing the language" leads to substantial revisions that go well beyond grammar and punctuation. This is often because the writer becomes more conscious of what he or she is trying to say. It becomes clearer what the sentences in the text are capable of meaning. For this reason, I suggest that one always plan for the first round of editing to consist of line-editing—of reading a text with the assumption of a rather wide range of possible improvements.

This act of editing, then, must occur well in advance of the ultimate deadline. The author must have time to make a great many decisions about the text in the wake of the editor's intervention. What the editor thinks the text ought to be doing may not be what the author had hoped it suggested.

Although it is important to present an editor in the first instance with an open site for intervention, it is no less important to at some point present an editor with a text that is largely finished. This means that the writer has read it through many times to ensure that everything is order and, especially, that the references have been completed and double-checked (it is always a sign of trouble when a writer leaves this to the "very" end and gives me a paper to edit where this has not been done). A copy-editor is looking mainly for misplaced words and pieces of punctuation. This does not mean that the sense of the text is ignored in favour of its syntax, it just means that any recovery of that sense must be accomplished with very limited means.

It is easy to see how the occasions for line editing and copy editing might structure a writing process. The act of producing a text fit for line-editing is not the same as the act of producing a text that is fit for copy-editing. They ought to be superficially similar (both should have section headings, references, fully formed prose sentences, and be free of typographical errors) but the writer should have a very different opinion of them (and therefore a very different reaction to receiving criticism).

While this does mean that the writer will also feel differently about them, emotions are not my concern here. What is important is how the time before and after the editor's work is done is spent.

Before the first encounter with the editor, a text of roughly the right length must be produced. It must include all the relevant empirical and theoretical material and, of course, the field of references that the text emerges on the background of. What I am trying to suggest is that even after the bulk of the prose is written, i.e., even after you've written a sufficient amount of sentences covering enough aspects of your research, there is a lot of activity still left to be done.

The editor will respond by suggesting a new text that, so the editor claims, better accomplishes what you "had in mind". But he only has the allegedly "inferior" text to base his opinion on. So it is important to clear some time in your calendar for when you get the text back. You have to keep an eye on these people.

Editors are behaviorists about textual meaning by (second) nature. They don't know what you mean except through what you actually say. Thus, the encounter with the editor is a terribly unromantic one. The "spontaneous overflow" of powerful ideas onto the blank page is hardly respected at all. Each word is asked to account for its place in the paper not in the larger whole of your "intellectual tradition".

One important part of your textual behaviour lies in your responses to their suggestions. Your editor tries to push the text in a particular direction and you either let it happen or push back. These acts are important indications of what you mean, and may often surprise you.

Roland Barthes famously proposed "the death of the author" as a point of departure for reading texts. "The death of the author is the birth of the reader," he said. Well, your editor is among your first readers. He is there to produce a text that will say what you mean, without further help from you—a text that can take care of itself after your role as its writer is over.

Basis and Elaboration (3)

Keep in mind that both empirical and theoretical statements can serve as the basis of your argument. I have been trying to emphasize the difference between brute facts and sublime ones in order to show what it means to use one kind as the basis of a particular paper or chapter that elaborates on the other.

In a longer work, like a book or a dissertation, you are able to elaborate on something in one chapter that you then use as your basis for another. Some writers forget this. Most commonly, they forget what they have accomplished in their theoretical chapters and therefore feel uncomfortable applying their (hard won) theoretical perspective to their empirical material.

I had pointed out that brute, empirical facts are often a "private" affair, while sublime, theoretical ones are open to public discussion. That's one of the reasons it can be a good idea to pivot your argument across the empirical/theoretical divide. It makes it clear to your reader what your contribution to their understanding of the world is—but also where they will have to trust you—and what "their" (i.e., the field's) contribution to your research is—and therefore where you're demonstrating your trust in "them".

Basis and Elaboration (2)

Theoretical and empirical statements are not completely different kinds of beast. Ever since the fall from grace of positivism and falsificationism, however, we have had to make do without any simple connection between them. (Karl Popper, you may recall, got many people to believe that a single statement of empirical fact could render an entire theory suspect.)

Theoretical and empirical statements are properly "about" the same thing, the former are just more general than the latter, which are more specific. In my last post, I started talking about the "sublimation" and even (yikes!) "brutalization" of your research. I want to say a bit more about that now.

Theory is "high culture" to the "base nature" of our empirical world. Theorizing is a more elegant and sophisticated way of engaging with otherwise empirical experience. But theories are as "factual" as the objects they are about; they are, after all, very precisely about those objects.

So my point is that there are sublime facts and brute facts. As in the more familiar (Freudian) application, the sublimated object is just less embarrassing to talk about. Brute facts are just that—you can take them or leave them. But sublime facts (the objects of theoretical statements) can be discussed in public.

This "publicness" is important. As a researcher, you normally have largely "private" access to your empirical materials. Your reader can believe you or not in matters of brute fact. But your conversion of those facts into theoretical insights allows them to be discussed on a ground that you share with the reader, even if the reader is wholly unacquainted with your data.

I'm still thinking this through. More later.

Basis and Elaboration

It took me a long time as a student to accept the procedural distinction between "empirical" and "theoretical" work. In fact, it wasn't until I became a teacher that I understood how important the distinction is. And it wasn't until I became an editor of academic texts that I understood why it is so important.

The empirical/theoretical distinction, as employed especially in the social sciences, offers an opportunity to distinguish between the basis of your argument and its elaboration. It is only paradigm-shattering, "revolutionary" research that will, as it were, "work out" an idea "fundementally", i.e., elaborate the basis of a field.

Most work will "normally" either provide a theoretical elaboration on an accepted empirical basis or provide an empirical elaboration on an accepted theoretical basis. It is the natural tendency of intelligent and creative people to assume that their work will have profound consequences that resists the challenge to focus (or tilt) your work in one of these two specifiable ways. What you are resisting are some perhaps evil, but quite necessary, constraints on your intellectual creativity (or creative intelligence).

Once we consider the challenge posed by the standard academic journal article (which you must publish or perish) these constraints become both obvious and practical. In order to get 8000 words to have a specific set of effects you will have to decide where you want to stand and what you want to move. If the academy is to remain a "garden" it is simply impossible to give everyone enough room to put everything anywhere they want.

The most effective way to find your footing and maximize your leverage is to work across the empirical/theoretical distinction. It is not impossible, of course, to write an entirely theoretical or entirely empirical paper. The trick here is to decide what part of a theory or set of empirical facts you want to base your argument on, and what part you want to elaborate. But it is much more effective to either sublimate (as theory) one part of your argument or brutalize (as empiricism) another, i.e., to impose the theoretical/empirical distinction "for the sake of argument".

I can see now I'm going to have to continue this line of thought with another post. More later.