Permit some metaphors. I generally think of our research as being based on method and framed by theory. I have occasion to insist on this when an author I’m working with writes that their paper is “based on NN’s theory of X”. The paper isn’t based on your theory but, perhaps, “guided by” it, I suggest. Indeed, you perceptions are framed by your theory. Your concepts are observational categories. Properly speaking, your research is based on “observation”, but since we’re doing science, we call the results of your observations, “data”.

It’s the mutually supporting structure of theory and method, frame and basis, that allow us to make “objective” assertions about the real world, or to discover “the truth”. Having a frame and a basis gives our observations the stability they need to inform our statements about reality. And those statements carry this stability with them into the discourse, where our peers can give the matter their own careful consideration. But what puts the “care” into “careful” here? That’s what I wanted to say a few words about.

A theory frames the way we see the world. Since our peers see the world (more or less) as we do (that’s what makes them “peers”) our theories also frame the way we present our research to them. By assuming our readers are our peers, we assume that a particular set of concepts and assumptions is available to them.  It’s the same set of concepts and assumptions that guided the design of our study. The theory, we might say, “prepares” us to see the world in a particular way. But the world itself must be “prepared” to be seen that way. It must be “set up” or “readied” for us to see it as the theory demands. That’s where method comes in, telling us how to generate data that can be “given” to the theory. It’s only when the world is given to our experience as specially prepared data that our theory knows what to do with it. And if our theory knows what to do with it, so, presumably, does our reader.

So we carefully prepare the world to be observed within the framework of the theory. And the reader then carefully imagines what we have done to see the world in this way. To guide them in their interpretation of our theory–to help them see that our theory is also their theory–our readers might make use of a “meta-theory”. It will contain some overarching principles that govern, for example, theory selection and theory development in our discipline. We might have a choice between two or three theories in designing our study. Our meta-theory helps us decide which is most appropriate. Sometimes we need to be explicit about how we have used these principles; other times it will be completely obvious. The point is just that a theory is a frame within a frame. Our care shows in the way we situate one within the other.

Likewise, underneath our methodologies, there is an area of concern that comes before our collection of data. Here too, things may happen without thinking too much about it. Much of this concern is dealt with by developing good habits of seeing and listening, noticing and note-taking. It hardly matters how sound your methods of analysis are if you don’t keep your test tubes clean, after all. Your methods tell you what to measure; your theories tell you what your measurements mean. But you must yourself take the measurement  accurately. And once you have taken it, you must make sure you write it down in the correct column.

After a life in research, these habits become second nature. (As do the bad habits that can emerge if you’re not paying attention.) But, just as we will sometimes make use of a meta-theory to make explicit the principles that guide our choice of theory and the changes we make to it, so too can an inframethodology sometimes help to remind us what to be careful about in the everyday conduct of research. This care is the underlying basis of the craft of research.

Do not leave “chores” like proofreading and referencing “for later”. They are part of the activity of writing.

Over time, the writing moment should unfold in a familiar way. It will be different for different people, but it should always begin with the typing of the key sentence and end with a two-minute proofread. After about twenty minutes of writing, you should read the paragraph out loud. It should come off the page comfortably. You have the rest of the time to make it so.

Now, about those references.  You have chosen something you know well to write about, and knowing something means understanding your basis for believing it. That means that you know also the references that are required to support your claims. Suppose you type out the key sentence and then give some thought to the difficulty it poses for your reader. Your reader, you decide, will find what you are saying hard to believe. So you offer a few sentences of supporting evidence. But where did you get that evidence? Perhaps from a book written by an authority on the subject. Well, that authority, i.e., the reference to that book, is simply part your knowledge of the subject. If you don’t know the reference you don’t really know the fact. Force yourself to work to that standard.

Remember that this does not require you remember everything you cite. It just means that the decision to write something has to include digging the relevant page out of your notes. You may not know in your head exactly what it says; but you should know how to easily find it. It will contain the information you need to cite the source in question.

At the end of your writing moment you should always have a well-formed paragraph of at least six sentences and at most two-hundred words. The sentences should be as grammatically correct as you know how to make them and should be, to the best of your ability, free of typos and misspellings. They should be properly sourced to references that you yourself find reliable. You want to have the ability to produce a chunk of good scholarly writing in 27 minutes, so you have to give yourself the task of doing everything within the time allotted. Don’t rely on some additional process to make your prose even just adequate, passable. Aim to write well every time.

Of course, there will always be a need for a final proofreading of the whole document and for you to meticulously check each individual reference for accuracy. But you will find that these final tasks are much less of a chore if you did the actual work of writing carefully and conscientiously in the first place.

[Click here for all the Rules.]

In our discussion about the nature of “academic” writing, Julia Molinari introduced the very useful notion of “affordances”. I remember Manuel DeLanda once explaining affordances by noting the difference between a mosquito stepping onto the surface of a lake and, say, a bear doing the same. While the bear’s foot goes right through the water to the lakebed, the lake “affords” the mosquito a surface on which to land. It reminded me of my old philosophy professor, Charlie Martin, and his “mutually instantiated dispositions”. The surface of the lake and the tarsus of the mosquito are so “disposed” that the former offers the later a surface tense enough on which to walk. The lake does not afford the bear’s paw the same thing, though the its same lake, of course.

I won’t develop the point in this post, but it’s interesting to consider that Foucault developed his “discursive formations”, which he used as units of analysis for scientific knowledge, into the more general notion of a dispositif, sometimes translated as an “apparatus”, a structured “readiness” for particular manifestations of human agency. What I want to do here is simply play a little on the notion of “affordances” in the interest of what I saw as the crux of our discussion, namely, the readiness to be proven wrong in our discussion with our peers. This, I believe, is what academia–and therefore academic writing–should be all about.

Basically, an academic is someone who can afford to be wrong. The academic does not “bet the farm” on any particular claim, or even an entire theory. Ideally, the academic, through the institution of tenure, has a protected livelihood that will persist even after repeated falsifications, and this should show in the style of our writing as academics. We should assert our claims boldly and transparently, tying what we believe to be true explicitly to our basis for thinking so. If we believe something because we read it in a book, we provide the source so that someone who knows that the book we’re citing is wrong can point our mistake out to us. If we believe something because we gathered data and analysed it in a way that suggested a conclusion, we provide enough information about our methods that a qualified peer can tell us where we went astray if we did.

It is easier to listen to–indeed, to talk to–someone who we think can afford to be wrong. If our interlocutor doesn’t seem capable of recognizing a mistake because their livelihood depends on their being right, or everyone thinking that they’re right, then our criticism will be colored by a certain embarrassment. It’s not that we ourselves are necessarily right, of course. The imbalance exists as long we are arguing a point we can afford to be wrong about and our interlocutor cannot. In academic discourse, then, one must choose one’s themes and one’s conversation partners carefully. The aim is to afford ourselves and each other an occasion for criticism.

Notice that this works both ways. You shouldn’t say something you can’t afford to be wrong about. But you also shouldn’t engage with someone who is making a claim they can’t afford to admit they’ve gotten wrong. If you do, the conversation will not be academic. That’s one reason that scholarship engages with the thoughts of very senior, very established, and sometimes altogether dead authors of so-called “classics”. These are people whose position is so firmly established that we don’t have to worry about harming them with our criticism.

Perhaps this will show that I’m an idealist, but I firmly believe that the value of the academic literature is that it affords anyone at any stage of their career a stage on which to present their thoughts, no matter how mistaken they may be. When you say something “for academic purposes”, whether in a school assignment, a doctoral dissertation, or a journal article, you are protected by “the right to be wrong”. Or at least you should be. These days, it seems, there are various movements, both on the left and on the right, that would have academics seriously consider the consequences of stating their beliefs, whether in front of their students or their teachers or their peers. Holding unorthodox views (or at least expressing them too clearly) sometimes seems to be a dangerous business.

I think this is why I’m so strident about keeping the notion of “criticism” in our definition of academic writing. In academia, mistakes should in a sense be so “cheap” that everyone can afford to make them. Conversely, we should invest very little of our total wealth of knowledge in each of our disciplinary engagements. If we do make mistakes, they have to be honest mistakes, of course. And they should not reveal an important area of ignorance or incompetence in our thinking. But we should be writing, for the most part, without fear of being shown to be wrong by other knowledgeable people. We should be ready for that possibility.

Our doctoral studies should prepare us for this. It should endow us with the necessary wealth, and inculcate the necessary frugality, to make each possible mistake affordable to us.

Do not write from your sources. Write from your notes or from your memory.

If you’re following my rules, you are always writing about something you decided yesterday that you knew last week. You should pick something that you not only think is true but understand why is true. It should require no additional research on your part to be in a position to assert it in writing–to make a claim and support, elaborate or defend it in a paragraph. All of these issues have been settled the day before, during five or ten minutes of careful reflection.

You don’t have to have the entire basis of your argument “in your head”. Knowing something does not mean memorizing every important fact about it. But you should be able to quickly and easily locate the relevant information in your notes. After formulating your key sentence, you should be able to reach for the set of notes in which the sources of your knowledge are clearly stated. This should take a few minutes at most. If it’s going to take you an hour or an afternoon to find the sources of a claim then you don’t know it well enough to write about it. Choose something else for tomorrow and resolve to understand the other thing better.

Imagine you are giving a lecture and you assert some matter of fact. A member of the audience asks you to back it up with a source. You say, “I don’t know it off the top of my head, but I can get it for you if you send me an email.” By this you don’t mean you’ll need a few weeks to discover whether or not what you were saying is actually true. What you mean is that back in your office you have easy access to the source. Maybe you’ve got some notes, indexed in a useful way. Maybe you know exactly what to type into Google to bring it up on the first page of search results. That’s fine. It’s doesn’t matter what kind of access to the source you have; it just has to be quick.

Some people don’t take very good notes but have a great memory. They might say they know roughly where in what book the information they need is to be found. Very well, as they are making their decision about what to write tomorrow, they should find that book, open it to the relevant page, confirm the accuracy of their memory, and then put a bookmark in. Now they just have to bring that book to their writing session. The danger, however, is that they will take too much of the style of their source with them. In the worst case, they inadvertently plagiarize their source. So I would recommend making a quick note of the key facts and (clearly marked!) quotes, including their page numbers, and then just bringing these notes to the writing session.

The other danger is that you will start reading when they should be writing. This is also why I recommend simply putting your source texts well out of reach as you sit there and write. You don’t need a library to compose at least 6 sentences and at most 200 words in 27 minutes. You just need a few jotted remarks to remind you facts, figures and names. As the brilliant poet Ben Lerner once said (though he was probably thinking of something else altogether):

“Gather your marginals, Mr. Specific. The end is nigh.”

A student writes me for advice about writing summaries. I thought readers of this blog might also find the answer useful. The difficulty, of course, is to provide a more concise version of something you have read without leaving anything important out. What are some strategies that might help you do this?

First, keep in mind that you always have to summarize from some perspective, and that perspective will necessarily be, in part, your own. Be aware of why you are reading this text, what is it that makes it interesting to you. And then make sure your summary includes the things that serve your interests in the text. For example, I often find myself summarizing Karl Weick’s analysis of the Mann Gulch disaster, which I think is flawed in a number of ways. So my summary always includes the things I think he gets wrong. Out of fairness, however, I also usually include the things he gets right, the things I agree with him about. You might also think about what parts of the text satisfied your curiosity. These things should also be in your summary, as should the parts that surprised or even outraged you. In other words, summarize the interesting claims in the text.

There is, of course, no “complete” summary of a text that isn’t just a perfect copy of that text. In order to make your summary useful, you will always leave something out. The trick is to leave the right things out for the right reasons, and that’s where your own interest in the text is so important. You have to have a good sense of why you are reading this text or, perhaps, what interest the reader of your summary will have in the source. Sometimes you are writing a summary just for yourself. But sometimes you are writing it to spare someone else the trouble of reading a longer document. There is no way to know how to do this without knowing something about your reader.

More practically, you might begin by making a “key sentence outline” of the paper or document you are summarizing. For each paragraph, try to find the one sentence that states its main point. That will give a list of 40-60 sentences for a standard research paper, which is probably a longer summary than you want. But one of those sentences will, ideally, state the overall conclusion of the paper. Your task now is to find the most important supporting arguments for that conclusion among the rest of the sentences in your outline.

Every time you make a summary, set some constraints on the task: How long should it be? How much time are you going to give yourself to do it? And make sure you have some purpose in mind, a focus: are you interested in the methods used in the study or the conclusion it reaches? Or are you interested in the recommendations the author makes for practice? Remember to summarize the paper on your terms, not necessarily those of the author you are summarizing. The summary should be useful to you and your readers, and while it should of course be “fair” to the original (i.e., it should represent the contents accurately), it doesn’t have to say everything the original says. You have to make some choices.

At the end of the day, when evaluating the summary try to identify the “trouble” it actually does spare you and/or your reader. And at what cost. A good summary will allow your reader (who may be your future self trying to remember what the document says) to make do without reading the document (at all or again). They will become more efficiently informed about certain things. But what will not reading the whole source document cost? What will the reader of your summary remain ignorant of because you left it out? Is that information important?