Craft of Scholarly Writing, a work in progress

Comments are welcome.

The Craft of Scholarly Writing

           

            When I became interested in dog training, I borrowed or bought all the books I could find and read them. And what I read had an odd, but oddly familiar, quality: while the books contradicted each other in general and specific, and rarely had grounding in anything other than personal observation (if that), almost all of the books presented their specific recommendations with equal, and equally false, certainty. If there was only one right approach, and that approach was so obvious to a reasonably perceptive person, then why was there so much disagreement among apparently reasonable people? When I was going to become a parent, and bought and borrowed as many books as I could read on parenting, I found the same odd combination of presenting information as though only benighted idiots disagreed with considerable disagreement among people of intelligence and experience.  The only thing on which they seemed to agree was that there was a right approach, even if there was no consensus as to what that approach was.

            This logically contradictory state--a field that actually had considerable variation in opinion, but filled with texts that acted as though their advice was the only possibility--was familiar to me as a writing instructor. Having read and reviewed a number of textbooks, I was long struck with how few were based in empirical research, and the tendency to insist on one approach to writing. Sometimes there was consensus, and sometimes there was research, but the consensus and research didn’t generally match up. For instance, in writing instruction, it remains commonplace to insist that students begin each paragraph with their topic sentence, although Richard Braddock, and others, long ago demonstrated that published writing doesn’t always have topic sentences at the beginnings of paragraphs. The same could be said about our insisting on funnel introductions, what we (inaccurately) call the “thesis statement” and where we insist it be placed, attitudes toward passive voice, and models of reading.

            That isn’t to say that the commandments were necessarily always wrong, but it did seem to me that they were wrong to be given as commandments. Funnel introductions are excellent for standardized tests, ending the introduction with a thesis statement is a good strategy for much student writing (assuming one really means “thesis statement” and not topic sentence or partition), passive voice is necessary in the sciences but easily overused in the humanities, so there are circumstances in which the advice is good, but also ones in which it is bad. I was reading books on dog training because I had two dogs—one mostly Great Dane and the other mostly Malamute. Many dog training books insisted on establishing dominance over one’s dogs through, among other things, issuing commands in a deep voice while one’s feet were firmly planted. This was excellent, even absolutely necessary, advice for the Malamute (Hoover), but it made the Dane (Chester) slink off and try to hide on the couch. It was a terribly ineffective way to train him.

            The conclusion to which I came was, despite books’ tendency to insist that all good people used their specific method and no other, that one needed a dog training method that worked for the temperament of the dog one was trying to train. It was a little confusing—to plant my feet to give Hoover a command, and then relax my stance to give a command to Chester—but it worked.

            And it made me a more productive writer and teacher, something I will try to explain in the course of this book. The way that I work with students on writing comes from that same conviction about method being context-dependent, several times over: what constitutes “good” writing (that is, good writing products) varies from one rhetorical situation to another; what enables “good” writing (that is, one’s writing process) similarly depends upon the author, project, and rhetorical situation.

            People who find themselves with writing blocks—which, like dogs, come in all sorts of breeds and combinations—often engage in a tremendous amount of negative self-talk (to put it politely), but, it seems to me, have trouble seeing the problem as the consequence of a writing process that used to be appropriate and no longer is. They wouldn’t have that process if it had never worked, or they wouldn’t be in graduate school or the tenure track. Faculty who cannot write that second book, or revise their dissertation into a first book, graduate students who cannot write a dissertation—all of those are people who are successful writers. They wouldn’t be in the position of people who are trying to write a dissertation or book were they not.

            But the problem of people who have been successful writers and then find themselves unable to get an adequate amount of writing done is significant. Some studies report a 50% attrition rate in graduate programs (Nettles and Millett, Gardner), much of that attrition happening at the moment that people who are successful writers of seminar papers need to write a dissertation. The problem does not end if students manage to write a dissertation—although there are not reliable figures on the issue, the reasonable inference is that the major reason that assistant professors are denied tenure is failure to publish within the five year “probationary” period (to produce a book in what are called “book” fields, or an article or two per year in “article” fields). And quite a few faculty members get stuck at the Associate Professor rank, having published enough for promotion and tenure, but never producing a second book (or comparable number of articles) to be promoted to Full Professor.[1]

            In the various cases I have known along those lines—students who never managed to finish a promising dissertation, faculty who did not produce a book in time, or were unable to finish their second book—it was never a case of dimwittedness, cupidity, laziness, or any of the other moral failings to which we (and they) too often attribute that failure to thrive. It certainly isn’t accurate to call them “unproductive” especially since I have seen entire Departments in which the heaviest service was carried by Associate Professors. The people I have known who have stalled in terms of their ability to get published are often very hard workers, with work sometimes spilling over into every possible minute, and often stressed to the edge of endurance. Laziness would be much more fun.

            What has struck me about people who are struggling with writing—and I write as one who is in that category and who has paid the cost for it—is that the situation was not one of any kind of moral failing, but a combination of a newly ineffective writing process, inadequate understanding of new genres and rhetorical tasks, and self-hobbling time management strategies.

            It is not that the person cannot produce good writing, nor that they have bad writing processes, but that they have a relatively rigid approach to writing. They have a method, and they keep using that method, even when it isn’t working for them. In other words, my experience talking to and working with scholars who are struggling to get writing done is that they are much like the undergraduate students with writing blocks that Mike Rose described many years ago: whereas he expected that their writing blocks came from knowing too little about what and how to write, he instead found that they knew too much. Their “rules” for writing were potentially good rules—grab your reader, get your introduction really tight before you try to write the rest of the paper, have a thesis before you write—but the students weren’t always clear when they should ignore those rules, or what to do if the rules didn’t apply. The rules weren’t harmful, but the students’ commitment to them was.

            But the rules had enabled them to produce good writing at some point, and continued to work in some circumstances—hence their resistance to trying new methods. It’s as though they kept planting their feet and acting dominant with the Dane because it worked with the Malamute.

            It seems to me that something similar happens with scholars whose previously (and still intermittently) effective writing processes, generic understandings, and time management strategies are inhibiting their ability to finish the dissertation, book, or articles: they have trouble imagining that they may need to develop a new approach since the one they are using has worked so well in the past and continues to work in other circumstances. I will take each of those closely related issues in turn—generic understandings, writing processes, and time management—but here I want to make my point by a brief discussion of time management strategies.

            Time management gurus emphasize the need to distinguish tasks that are pressing  (sometimes called “urgent”) from ones that may be more important in the long run but with no immediate pay-off or disaster (“important”). Scholars on procrastination call the first kind of task or activity STR, since they provide Short-Term Rewards (also SS), and the second kind LTR since they provide Long-Term Rewards (also LL).  Some time management strategies (such as the very effective and popular Seven Habits of Highly Effective People) involve having people label tasks and activities in four quadrants:

	Important	Unimportant
Urgent
Not urgent

What people will often discover is that they spend most of their time working on tasks that are important and urgent, delaying the important but not urgent tasks until they become urgent. So, for instance, a student will delay writing an end-of-semester paper that counts for a large percentage of the grade (making it more important than quizzes) till the deadline is looming. This strategy is generally condemned, and perhaps it should be, but it’s awfully effective for undergraduates, and even for the coursework part of being a graduate student. As I will discuss later in this book, there are actually lots of good reasons that people have for doing so—procrastination is not necessarily a bad time management strategy.

            In fact, it’s precisely because it is so effective and has its own rewards that people stick with it. Students have been told to start their papers early, but often don’t have the time or knowledge necessary to do so—the classic end-of-semester research or seminar paper typically depends upon information or skills imparted toward the end of the semester, so starting early may be impossible. Under those circumstances, students have no choice but to leave the important/not urgent task of seminar paper-writing undone till it becomes important/urgent.

            Psychologists and psychoanalysts who write about procrastination tend to emphasize its role in the imposter syndrome. Rose Fichera McAloon describes her undergraduate writing process:

I was terrified of criticism, of being unmasked as a fraud, of being stripped of my self-esteem, of being irreparably crushed. I wanted to write the papers, fuss over them lov- ingly, craft them to perfection—but would not and could not. They were always written in a slapdash way, never reread for content, and turned in with the hope that a miracle would happen and that I would beat the odds once again. It mostly worked. (239)

There are other reasons that people are drawn to tasks only in the important/urgent box, and those attractions are a major factor in procrastination—it makes life more interesting, since one is always in a crisis; since some tasks become irrelevant before they ever become urgent, it’s an easy way to prune one’s “to do” list; it enables the perpetual procrastination of boring tasks like exercise—so that method of time management can have rewards. The problem is if we use that strategy at all times and in all circumstances.

            It seems to me that students in the humanities, even successful ones, have a fairly truncated writing process: they read the necessary material and think about it until they have some brilliant insight. Sometimes the brilliant insight precedes the reading, as when a student wants to argue for a particular position and then goes off in search of the research to support that insight, aka “thesis.” Once s/he knows her thesis, s/he writes a paper that supports it; the writing process and working with the evidentiary materials probably lead to a more nuanced thesis than s/he had initially (hence effective writers often write their introduction last).  The paper generally has the form of a summary introduction (one that summarizes every point in the paper, ending with a clear statement of the thesis), supporting evidence, prolepsis, summary conclusion, speculative conclusion. Regardless of the form, the process is, I would suggest, the real problem: you never start writing until you know exactly what you want to say.

            With such a writing process and product, panic can be an effective stimulant. In addition to keeping one awake, it helps to silence the voices of self-doubt and self-criticism that might make writing difficult. Time becomes an active participant in the writing process—it’s time that forces someone to commit to a thesis (silencing the voice that seductively promises a more brilliant insight if one just read some more or thought about it more or played one more game of solitaire or watched just a little bit more TV), that forces one to start writing (when the same voice promises that this time the muse of scholarly writing will appear if one gives her time), that forces one to polish what one has written instead of start over entirely, and that forces one to turn the paper in (when one’s critical voice is listing every aspect of the paper that could be improved with more time). Every once in a while, of course, time drives the bus into a ditch, and one has to request an extension, but teachers are forgiving of bright and promising students, and so the panic-filled journey is extended for a few days, weeks, or months.

            As stressful as this process is, it is also exciting (one of many benefits), while giving one an obvious and straightforward excuse should the product turn out to suck: it would have been better had I started earlier. Letting limited time (which really means time-induced panic) drive the bus gets one places as an undergrad, and even works fairly effectively through coursework in graduate school.  But, as the deadlines extend, such as preparing for one’s qualifying exams or writing a prospectus, the ability to pull the chestnuts out of the fire at the last minute without third degree burns gets harder and harder. Because the bus can’t go without the driver, people dependent on panic to drive typically either delay writing until dangerously close to various deadlines or learn to work themselves into panics over earlier deadlines. Panic is still at the wheel, and stress is one’s constant companion.

            Assistant professors can similarly keep themselves in a constant state of panic or delay writing till the tenure axe looks dangerously close to dropping. But quite a few simply can’t maintain that level of panic, and an unhappy number of assistant professors find themselves with delays outside of their control—illness, a dilatory press, hostile readers for a press, poor mentoring—and the axe falls.

            And here I should explain something about this narrative—it’s hard to find the right words. It’s inaccurate for me to say “begin writing” because people with severe writing blocks in regard to their scholarly writing are often writing a tremendous amount. Sometimes they (we) are writing the wrong things—reports for one of our (too many) committees, long comments on student papers, blog entries, salvos in internet wars, talks and articles tangential to our main projects. And sometimes they (we) are writing the right things wrong--overly ambitious books, books or articles with no existing publishing venue, the same piece of writing over and over and over till the prose glistens. I have seen supposedly “unproductive” scholars whose writing was beautiful, and signified tremendous time commitments. Sometimes it was unsubmitted, and sometimes there was simply too little of it. Either way, the glistening prose alone did not earn them tenure.

            My goal, in this book, is to describe different ways of thinking about the three issues that seem to get in the way of scholars completing their scholarly work: the generic conventions of scholarly writing (which are not quite what students are often told they are); various writing processes; and time management (or why procrastination is not a moral failing). I don’t want to be understood as replacing one set of rigid rules with another—I doubt there is anything I say that applies universally—but instead to give writers more options.

---

[1] See Judy Touchton, Caryn McTighe Musil, and Kathyrn Peltier Campbell. A Measure of Equity: Women’s Progress in Higher Education (Washington , DC: The Association of American Colleges and Universities, 2008).

Chapter One: Generic Conventions in Scholarly Writing

I. Introductions: Functions

            Cicero, who knew more than a thing or two about effective rhetoric, described not one kind of introduction, but five different kinds of exordia, of which the introduction is just one form. Although he was talking more about speeches in court and the Senate, his taxonomy is a far more useful way to think about introductions than “tell ‘em what you’re gonna’ tell ‘em” (aka, “summary” introduction). Cicero’s taxonomy is dependent on what the audience’s attitude is toward one’s argument: they may be sympathetic, hostile, confused, bored, or some mix of the previous.[1]

            That is, one might be making an argument one’s audience is well-disposed to hear, as it confirms what they already believe. In scholarly work, readers are likely to be sympathetic toward scholarship that reinforces the importance of the field, confirms the value of dominant methods of interpretation, extends well-known arguments, shows that other fields have come to similar conclusions, argues for the value of already valued authors or texts. With a sympathetic audience, Cicero says, one begins with one’s thesis.

            With a hostile audience, however, one delays one’s thesis. This may seem to contradict current practice, but it only contradicts current advice. Much current advice depends on a conflation of various functions that a sentence (or set of sentences) might perform: establishing the topic, identifying the significance of the topic, and giving a map of one’s argument. The tripartite thesis statement (A leads to B because of C, D, and E) or conventional version of the “summary” introduction do all three kinds of work, but they aren’t the only ways to do them. A summary introduction typically gives a map of the argument by listing the main claims of the argument in the order they will be made in the text. But a topic sentence is different from a claim. A topic sentence tells the reader what the paragraph (or section or paper or book) will be about; it is often a claim but not necessarily the claim the text will pursue.

            For instance, an article that had in the introduction the statement that “Third world poverty is a tremendously destabilizing force in world politics, and needs to be solved” is probably not the claim the article will make—an article with that as a claim would go on to show nothing more than that third world poverty is destabilizing; it would not propose a solution. While it’s possible that an article would do no more than argue the proposition the poverty is destabilizing, it would have to be for a journal or scholarly audience that would see such a claim as extending the conversation—that would need to be some kind of new claim—and it’s hard to imagine such a discipline or journal. Instead, chances are that the article will go on to propose a solution—the thesis for the article would be something like “X approach will reduce third world poverty insofar as it ….” This statement, “Third world poverty is a tremendously destabilizing force in world politics, and needs to be solved” is a topic sentence that makes a claim about the significance of the issue; some rhetoricians (like Wayne Booth) call this kind of statement a “contract,” as it establishes expectations with the reader. To avoid confusion, I want to lay out a set of terms for the kinds of work that sentences in the introduction can do:

·      topic sentences: topic sentences do exactly what is suggested by their name—they identify the topic of the paragraph, section, or text.

·      claims of exigency: one of the main differences between student and scholarly writing is that student writing is not expected to make any claims of significance; scholarly writing needs to pursue issues that the audience will grant (or can be persuaded) are significant contributions to some intra- or inter-disciplinary conversation(s). The main problem with what Cicero characterizes as “trivial” topics (usually translated as “mean”) is that the significance not obvious to the audience, and itself has to be argued.

·      thesis questions: a clear and specific statement (with or without a question mark) of the question the text will pursue; it is not simply the thesis statement in question form, but a succinct description of the problem, disciplinary conflict, interpretive puzzle, apparent contradiction, or need that will be explored (and not necessarily answered, depending on the discipline). Sometimes thesis questions are framed in such a way that the author’s answer is clear, but sometimes they are genuinely open.[2]

·      hypo-theses: also sometimes mis-identified as thesis statements, these are vaguer and more abstract versions of one’s claims; hence, just slightly less than the thesis statement. Sometimes an actual statement of the hypothesis—the claim to be tested through the data—but more often a less controversial (because more abstract) pointer toward what one will argue.

·      partition: called “partitio” is classical rhetoric, this was the part of the exordium in which the rhetor made clear the order of topics the rest of the speech would discuss. Sometimes considered part of the “metadiscourse” (Williams), “signposting,” or “scaffolding” that may operate throughout the text, it is more and less tolerated in various fields. Because some disciplines (especially in the sciences) have clear expectations regarding the order of parts, there is less need for a partition in those journals. The section of the introductory chapter in which an author summarizes each chapter is a partition, and any list of topics in the introduction tends to be read as one (so that the reader expects one will discuss them in order). Readers apparently read questions in the introduction in the same way, expecting them to be answered (hence the advice to avoid rhetorical questions in the introduction).

·      thesis statements: it seems to me useful to reserve this term for the main claim(s) for which the text argues. Thesis statements, if one uses the term this way, are not usually found in the introduction, but most often in the conclusion. A fair number of texts have no clear statement of the thesis, leaving it implied. And, as George Traill has pointed out, quite a few have a false thesis, seeming to argue one thing, while actually arguing something else (uncommon, but not unheard of, in scholarly writing).

Because this way of looking at sentences involves emphasizing the work they do, it’s quite possible for a sentence to fulfill two or more functions at once—the tripartite thesis sentence, as mentioned above, acts as thesis, topic, and partition.

            If one uses these terms, then what Cicero is saying about a hostile audience (or “dishonorable” topic) is that one should not begin with one’s thesis, but with a thesis question, hypo-thesis, or topic sentence. As I’ll show later, that is what scholars often do—with a controversial topic, they tend to avoid having the thesis statement in the introduction.

            In scholarly writing, one is frequently in the rhetorical situation that Cicero characterizes as trivial (or “mean”).  We pursue the questions we do for all sorts of reasons, sometimes simply that we love doing certain kinds of reading. But other readers usually like some reason to think carefully about a topic more compelling than that we like reading about it. There was a time when “recovery” work (that is, work that brought up a previously ignored author or artist) was considered significant in and of itself, especially if it was an author or artist of some kind of marginalized group. While that is still considered valid in some disciplines, many disciplines require some additional claim, some way that contemplating the margins changes our narrative of the center; under those circumstances, we need to make clear claims of exigency. This is the kind of work that students are trying to do with the “dawn of time” introduction—“since the dawn of time, people have been arguing about [my topic].” If people have been arguing about it forever, it must be significant. Scholars usually do the same work through a history of controversy introduction (described below, basically, an introduction that shows that there remains a controversy on a particular topic), since being able to narrate a controversy that involves many scholars implies that the issue is of considerable interest to scholars of that field. But scholars also do this kind of work through summarizing a consensus and then pointing out some reason to doubt it (the prolepsis or some say introduction).

Of course, if the problem is that the issue is likely to be seen as trivial by one’s readers, then there is no current disciplinary argument to summarize. Typically, then, scholars try to argue for the significance of their research by connecting it to some existing argument, suggesting that their work answers a “prior” (logically, not temporally) question, for instance. Under these circumstances, a “focusing incident” introduction can also be helpful (also discussed below).

A potentially confusing and complicated argument is not quite the liability for a scholarly author as it was for Cicero; yet his advice (start with a joke) is not necessarily a bad one. Scholars expect that the things they read will complicate issues—the “it isn’t as simple as it looks” may be one of the most common moves in scholarly discourse (see Fahnestock and Secor, Graff), especially in the humanities. But, it can be very difficult, especially for a new scholar, to write about a situation that is not simply complicated, but actively confused. Thus, for instance, scholars in rhetoric sometimes find ourselves describing an argument in which people are putting forward muddled arguments; conveying that the arguments are muddled without one’s self looking muddled is tricky—being clear about confusion is more than a little hard.

My point, then, is that Cicero’s way of imagining one’s introduction in terms of the attitude that one’s readers are likely to have toward the topic remains useful. There are a few minor variations that arise from specifics of the rhetorical situation for scholars, but, loosely, it is something like: if the audience is sympathetic, one can have one’s thesis in the introduction, but the exigency claims are particularly pressing. While having a controversial argument makes the exigency more obvious, there are higher risks of alienating one’s audience. The exigency claims are also important with a topic likely to be seen as trivial, and with a confusing topic, start with a joke.

John Swales has argued that scholarly introductions make three (not entirely discrete) moves: establishing a territory, establishing a niche within that territory, and occupying that niche (Genre Analysis, especially 137-166, quotes are from 140-141). An introduction makes generalizations about a relatively broad disciplinary topic (the territory), and then quickly moves to a more specific question or set of questions within that larger area (thereby establishing that there is a niche within that large territory). An effective introduction doesn’t simply describe the existence of a niche (that there are many scholars investigating a certain topic), but that this more specific area of inquiry is interesting and important. According to Swales, authors generally do so through one of four ways: counter-claim, gap, question raising, continuing the tradition. A counter-claim move would involve asserting that there is a particular consensus, and then claiming that the consensus is wrong. A gap argues that there is some aspect of the issue that has been ignored—such as a text, phenomenon, approach, era, group, or genre. An author might point to a known gap, such as by citing scholars who have called for a certain kind of research, or narrate the field in such a way that the presence of the gap is clear. Question-raising isn’t always distinguished from gap (since, presumably, the questions one raises have not been answered and therefore constitute a kind of gap), but it can be different. An author might point to an apparent contradiction, a way that some data violate expectations, or a puzzling aspect of the current consensus. The distinction between “continuing the tradition” and identifying a gap is similarly vague, in that an author will likely promise to continue the tradition in a direction not previously taken. Simply continuing the tradition is probably not going to seem significant enough, unless the introduction promises that the way s/he will continue the tradition will do something significant to the consensus.

William Germano, like many editors of scholarly presses, puts particular emphasis on the way that the “establishing a niche” section (more commonly called “literature review” although that term is sometimes used to encompass all three moves) is very different in dissertations from books, and that recently minted Phds often have the most trouble with revising that aspect of their scholarship. In a dissertation, establishing the territory and the niche are primarily useful for credentialing, for showing that one has read all the major scholarship in the area, relevant or not. For a book or article, however, only the relevant scholarship matters (hence a smaller amount) and one’s critical relationship to that material must be more highlighted.

That is, the “occupying the niche” is crucial—when the author “contracts” (to use Wayne Booth’s term) or promises to fill the gap, answer the questions (or at least phrase them more usefully), resolve the contradiction, and generally advance the conversation for a relatively large number of scholars, or in a way that will be of interest to a relatively large number of scholars. Swales sketches four strategies that authors use to make such a promise: outlining the purpose, describing the present research (in such a way that it makes clear that this method effectively tests an interesting hypothesis), announcing the main findings, or indicating the structure of the article (the topics, and not necessarily the claims, of each section).

Swales’ work is sometimes hard to apply, and I’ve found that readers will try to apply his terms too rigidly, as though they always happen discretely and in the same order, but his way of describing the kind of work that gets done in introductions seems to me useful, and consistent with Cicero. The introduction does not simply raise a reader’s interest and make them want to keep reading (the most common way to describe what an introduction is supposed to do) but an effective introduction persuades a reader that there is an important question, related to a current scholarly discussion, and that the author is the person to pursue it in ways that the reader will find enlightening.

---

[1] De Inventione, in most translations, this is: honorable, dishonorable, confused, mean, or mixed.

[2] Thesis questions, like topic sentences, are often mis-identified as thesis statements (“the issue this paper will pursue is whether X’s solution has unbearable costs”), a mis-identification that encourages students to misread.The prolepsis introduction, discussed below, in which one begins by summarizing one’s opposition argument, is very common in scholarly writing, but, since students are told that introductions end with the thesis, they will read a piece as arguing precisely the opposite of what it is actually arguing.

II. Introductions: Forms

 Students are typically taught one of two kinds of introductions: the funnel or the summary. Graduate students rely heavily on the summary introduction, but most scholarly writing uses the history of controversy. It’s very common for a scholarly introduction to use a summary or partition as the last part of a three or four paragraph introduction—these various categories aren’t necessarily mutually exclusive. The main work that an introduction does is not, despite what most people say, to tell the reader what the text will be about; it is to persuade the reader that you are pursuing a question of significance to her/his discipline, and that you are equipped to handle it with intelligence, nuance, and fairness.

Because that work needs to be done for the reader, but probably doesn’t need to be done for you, the introduction that enables you to get started may or may not be the one your reader needs—many writers write their introductions last. That isn’t to say they begin with no introduction; they have an introduction that gets them started, and then substitute a different one at the end—I have yet to use the first version of an introduction for anything I’ve written.

As I keep saying, history of controversy is probably the most common introduction in scholarly writing, also called the “literature review.” What one does with that literature review is demonstrate that there is a gap, contradiction, puzzle, or unnecessary tangle. The clear implication is that your work will fill that gap (or describe it usefully), resolve the contradiction, answer the puzzle, and undo the tangle. The history of controversy is often in chronological order, beginning with the earliest relevant scholarship (or with some scholarship that created a sea change). As with the “dawn of time” introduction (which this one resembles when it goes wrong), the difficulty is knowing how far to go back. Dissertations often have much longer literature reviews than books; in books, the literature review is often limited to the introduction (or footnotes). Articles almost always begin with them, although one sometimes sees them at several points along the way.

While you have to argue that there is something missing in (or wrong with) the existing conversation—part of what Gerald Graff aptly calls the “they say, I say” move in scholarly discourse—this move does not necessitate describing everyone prior to you as a benighted idiot. The extent to which one distinguishes one’s work from others’ seems to me to vary even within my own field—some journals are very irenic, and some more agonistic. When one does interdisciplinary work, the boundaries of the literature for which one is responsible are much larger, potentially overwhelmingly so (in fact, one reason that some people never get around to writing, and keep falling for the “I can’t write till I’ve read…” line, is that they haven’t figured out what boundaries to draw). This boundary issue is what rhetoricians call an audience issue—who is the audience? And what are their expectations?

In a way, one can never know, and discovering that there is an obvious book or article that one really should have read is an inevitable, if inevitably shaming, experience. But it’s worth keeping in mind that no one expects a person to have read everything; they do expect a scholar to have read the major relevant works in the field in which the scholar is claiming to be a scholar. What that apparently circular statement means is that there is a kind of recursive question: you need to know the texts relevant to the discussion you are trying to enter, a discussion that is largely defined by the texts you cite. A slightly less circular way to put it is to think about publication venues—what journals or presses are likely places for one’s work? And what texts are most commonly cited in those journals or presses?

Thus, an apparently simple question--how do I write my introduction?—turns out to be difficult to answer if one doesn’t have a sense about where and with whom one is likely to try to publish.[1] My dissertation director once claimed that most writing blocks were the consequence of this failure to identify a specific audience, a failure that resulted in someone trying to write for every possible audience.  Since that’s impossible, the person couldn’t write at all. Whether he was right in his quantitative claim, I don’t know, but it seems plausible to me.

Students are typically writing for an audience whom they know reasonably well—the professor from whom they’ve been taking a class for several weeks (if not months). By the time of graduate school, good students are reasonably adept at inferring their audience’s tastes. One reason that the dissertation (or prospectus, if one’s program has that stage) can be so difficult is that it may be one of the first times a student is writing for what rhetoricians call a “composite” audience—multiple readers with different (and sometimes incompatible) expectations. It isn’t just that their actual audience is made up of different faculty members, but faculty often (and quite reasonably) see their role in the dissertation process as helping students write for a broader scholarly audience. So, there is sometimes a kind of ghost audience stalking the process—what the professor imagines the students’ intended audience to be. That it is imagined is no criticism of the faculty member; writing is, as Walter Ong long ago noted, always writing to an imagined construct (which is why we are always surprised at the reactions of actual readers). But students who have become accustomed to writing to the actual audience of the real faculty member now have to adjust to a different sort of audience, and it can be a difficult adjustment.

The both sides introduction is really only a variation on the history of controversy. If the scholarly controversy breaks down into two or more camps, then one might choose to describe those various camps in turn (rather than narrate the scholarly controversy chronologically). This introduction is different from the previous only formally; it requires just as much work, and the same set of decisions (about what conversation one is joining). If, for instance, you are bringing together two disciplinary fields (psychoanalytic studies of literature and quantum physics, for instance), you need to make clear to your readers whether you imagine yourself explaining to people in quantum physics how psychoanalytic approaches to literature are relevant, vice versa, or (most ambitious) one to each other. Keep in mind that this decision about disciplinary conversation is, simultaneously and consequently, a decision about how much secondary for which you will be responsible. The more conversations you plan to enter, the longer your list of secondaries you must know. And trying to enter multiple disciplinary conversations does not necessarily make something, especially a book, more attractive to a publisher.

Academic publishing has never been particularly flush, but presses are in especially hard times right now, because their income is declining both in terms of direct support from universities and in terms of sales. Many are getting less funding than they used to get, and, because university libraries (their major source of sales) are getting less funding, their sales are down. Presses typically have certain areas in which they specialize, and that means that they are more likely to find attractive texts that fit clearly into those areas—the press already sends someone to the relevant conference(s), is well-known among scholars, and is known to libraries for that area. A book with a primary and secondary audience of relevance to a particular press is likely to be attractive; a book with two primary audiences may be hard to market.[2]

The prolepsis or “some say” introduction works especially well for early in a project, when one doesn’t know that whole field well enough to describe the general conversation. One begins by summarizing the main counter-argument, so it’s more or less the exact opposite of a summary introduction. To be effective, the summary of the opposition must be fair; it must be one to which someone who holds that position would assent. It also needs to be a position held either by a famous figure, or by a large number of scholars. Although some friends and I once threatened to put together a journal in which we would publish stupid arguments just so someone else could publish an article arguing sensible positions, no such journal exists. And, even if it, it wouldn’t be a respectable opposition, one that would be seen as worth engaging.

The focusing incident might be characterized as more of a strategy than a form, since it can be used with any of the above introductions. I mention it simply because it’s a way of beginning writing that can work when one is too early in a project to know one’s thesis, let alone to know how one’s argument fits into an existing conversation, or even what conversation one is joining. It’s quite likely that one’s interest in a topic can be encapsulated in a particular moment—a passage in a text, historical incident, event in someone’s life. If that specific incident epitomizes the question that one is pursuing, then it can be a useful way to raise the question. If it’s a significant incident, especially one that is likely to recur, then the implicit or explicit promise that one will describe this kind of incident in an illuminating way makes the exigency claim. If it is an interesting or odd incident, or an engaging narrative, it fulfills the same function that a joke does in Cicero’s taxonomy. While you may be engaged in a potentially confusing or boring topic, you are at least an interesting companion to have on the journey, and may have some interesting narratives along the way.

I sometimes separate out personal narrative although, like the focusing incident it is not structurally different from the other forms. It can be a history of controversy introduction, but instead of trying to describe how the discussion within a particular field has evolved, one describes how one’s own understanding of a topic has evolved. I have used the same introduction when I have begun every book, but it has never appeared in the final version. It is a focusing incident/personal narrative that narrates how I came to do the kind of work I do. It may work to get me writing because it helps me place the specific project in a larger issue I find interesting; when it comes time for the completed manuscript, however, it’s clearly not useful to a community of scholars. Obviously, it functions to get me started writing, and is therefore useful. Someday, perhaps, I will actually publish it.
A few semesters ago, a graduate student in public policy came to see me to work on her writing. I asked her to bring in articles from journals in which she would like to publish, and below is the introduction from one of the articles she gave me.  I’ve found this is a useful example for workshops for scholars in the humanities because it is such a different field—we don’t get caught up in the content, but can be more aware of the moves the authors make in setting up their topic, describing their method, and asserting the significance.

The first sentence summarizes the current disciplinary consensus (a colleague always points out to students that writing such a sentence can require several months’ work, if not more). The second sentence explains that consensus in more detail, simultaneously demonstrating it is a fair summary (through the large number of citations). The last sentence of the paragraph similarly shows that this phenomenon (collaborative efforts among independent parties on water issues) is widespread through a series of examples. This first paragraph, then, does considerable work—summarizing a disciplinary consensus (which, presumably, the authors will complicate in some way) while at the same time demonstrating that their knowledge of the field is broad. The range of examples and number of citations demonstrate that this issue is of importance to scholars and practitioners.

            The first sentence of the second paragraph is what John Swales calls a “turn”—the first indication of the specific aspect of the consensus with which the article will be concerned. In this case, the problem the article will pursue is one that is already recognized as a problem in the discipline—“concerns…has [sic] begun to be raised by scholars” (367). The second and third sentences of the second paragraph summarizes one of the policies proposed to solve this problem—“formal adoption”—and the paragraph ends with a potential problem to that policy.

            The third paragraph has a clear statement of the contract, implying that this article will “examine the performance of formal, enforceable arrangements among independent parties” (368). The authors mention another problem with these arrangements, and then explicitly state why an empirical study (thereby contracting for a particular method) is of significant interest to the reader. The method is explained in more detail in the fourth paragraph, and the term “compact” is defined (implying that this term is central to the authors’ argument, especially since it was part of the title). The fifth paragraph defines two other terms, conflicts and institutional settings, and then the introduction closes with a sentence that clearly contracts for the topic, significance, and method of the article. Notice that it contracts to offer more understanding and shed light on an issue—the thesis is not clearly stated in the introduction.

            The authors’ thesis is clearly stated in the conclusion. The article takes issue with the current consensus on several grounds. While compact commissions “commonly solve conflicts by revising rules” (385, suggesting that unanimity rules are not as high a barrier as literature currently implies), their effectiveness should be compared to conflict resolution in alternative venues. The authors conclude that current practice should significantly change. Obviously, this isn’t my field, so I don’t know if their representation of the field is accurate, but, assuming it is, their case is controversial, and, just as Cicero recommends, they delay their thesis till after they’ve presented their data.  

            

---

[1] Ironically, given that so many writers work first on what their thesis is and then on where one would publish, it’s easier to write an introduction without knowing one’s thesis than it is to write one without knowing one’s audience. One can write an introduction without knowing either, but it’s almost impossible to write a useful one without knowing what one’s question is.

[2] Probably the single most useful book concerning scholarly publishing, especially for assistant professors, is William Germano’s From Dissertation to Book.

III. Body Paragraphs

A.   Overall Organization

I’ll say very little about overall organization, since that is often determined by discipline. The best way to infer whether a particular journal or press has a rigid arrangement is to analyse articles or books in that journal or press, and see if they have the same organization.[1]

Many scholarly articles have the loose structure of hypo-thesis or thesis question, description of experiment, data, conclusion, implications. If there are multiple “experiments” (case studies or texts), then one might repeat the description/data/conclusion two or three times within an article. In the humanities, historical events, texts, people, or movements fulfill the same rhetorical function as the experiment does in experimental sciences—they provide the data. What this means is that scholarly articles don’t generally have the funnel/proof/conclusion structure often prescribed. Instead, they are likely to look something like:

Introduction (possibly ending with a partition)

Narration (of necessary background material, or of the experimental method)

Data (Proof)

Conclusion (summary of argument)

Speculative conclusion

The “data” section is by far the longest, and should make up the bulk of the article.

            There are various ways of arranging one’s proof. In the case of an experiment, it’s usually relatively straightforward. With other kinds of arguments, however, there are more options. First drafts often use a chronological arrangement, and that’s an excellent strategy for generating a draft, but doesn’t necessarily work for all sorts of reasons. The earliest text (or incident, or case, or person, or example) might be the most complicated, the most confusing, or most problematic.

            Some version of a list structure is fairly common, but people don’t always realize that there are different ways of organizing a list. The structure might go from least to most important arguments, least to most persuasive, most specific to most abstract, most to least familiar (generally the most desirable, as it means that one going from “known” to “new” information for the reader). Those are ways in which a list might build.

It can be effective for a list to decline, as in journalism (from most to least important), abstract to specific, clear to obscure, least to most familiar. A Nestorian arrangement puts the weakest argument in the middle (particularly effective in speeches, as the audience begins and ends with the strongest points).

Scholarly arguments often have a comparison structure. When you compare two things (Chester and Hubert), you do so in regard to several qualities (size, intelligence, attitude toward the red ball, age). An initial impulse that writers have is to discuss the first thing and then the second, so you have a structure like this:

        Chester's size

        Chester's intelligence

        Chester's attitude toward the red ball

        Chester's age

        Hubert's size

        Hubert's intelligence

        Hubert's attitude toward the red ball

        Hubert's age

That can work fine, but it can also mean a lot of backtracking along the way (as you remind your reader of the contrast) and a fairly long conclusion (for the same reason). It's often more effective to let the qualities organize the paper:

        Size: Chester, Hubert

        Intelligence: Chester, Hubert

        Attitude toward the red ball: Chester, Hubert

        Age: Chester, Hubert

This latter organization is even more helpful when one is comparing three or more things.

            Writers don’t always realize that a very common kind of scholarly argument (the theory/application argument) is actually a comparison paper. Much scholarship tests a hypothesis (in the humanities generally called a “theory”) by applying it to a new case, or new kind of case. Since the hypothesis usually has several parts to it (criteria, really) the structure of the argument will likely take each criterion and apply it to the case. Similarly, an article or book that lays out a new hypothesis often does so through a set of criteria. In all three kinds of cases—a true comparison, theory/application, and new theory—scholarly writing is distinguished from student writing by having to argue for the relevance of the case(s) being compared.

            For student writing, the reader is often looking just to see that the application demonstrates that the student has understood the theory/hypothesis. The student isn’t expected to change the scholarly conversation significantly, after all, so if s/he does nothing more than show that a theory applies in a new case exactly as it does in the known cases just as one would expect, that’s perfectly acceptable. But a scholarly article that confirms expectations can be a little difficult to place; thus, one needs to argue (usually in or near the introduction) that this specific case is a good test—it has characteristics that might falsify the hypothesis, or one wouldn’t expect it to apply, or it is completely different from the kinds of cases previously discussed.

            The main problem with a comparison paper is that you can end up with what I tell students is the comparison paper from hell. One can compare apples and oranges—they are both sort of round, grow on trees, are good in fruit salad, have an ‘a’ in them, have various hybrids. But the comparisons don’t lead to anything interesting. Comparing two things in order to show an unexpected commonality (both were once considered prized gifts) or to argue for a new category (trees I would like to grow in my backyard) can make the trouble of comparison worthwhile (assuming my reader cares about my backyard plans). Applying the theory to a specific case is significant if one argues that the application complicates, contradicts, or confirms the theory, or if it highlights something previously obscured about the case—but it needs to do something other than demonstrate that one can apply the theory.

A chronological structure just follows time and goes from whatever happened first to whatever happened next. As mentioned above, it can be an excellent way to generate a draft, especially if one has lots of time, but it can mislead writers who have limited time.  I've found that students who use a chronological structure will sometimes include material that isn't necessary (just because it happened next), or will spend a lot of time on whatever happened first and then run out of energy for whatever happened later. The proportion of their argument ends up being determined by how much time they had left rather than how long that portion of the argument needs to be.

            One variation on the chronological structure that can be useful for scholars is the turning-points structure or representative moments structure. Instead of trying to narrate every event (or text) from one thing to the next, one selects specific texts, authors, moments, movements (and so on) either at the center of historical moments or at the turning points. Of course, one has to argue that these are central moments or turning points (but that can often be done through a judicious use of secondary).

The syllogistic structure is the least familiar and most useful for students. You begin with whatever is the main premise of your argument and move through the evidence to your conclusion. To write this structure, you need a pretty good sense of just what you're arguing and what your audience believes. It helps if you're able to state your thesis as "your main assertion because your main argument." A thesis statement like that will enable you to figure out what your main premise is. Another way to do it, though, is to figure out what is shared with your audience--what is the common ground on which your persuasion rests? The paper starts there. Because it moves from common ground through evidence, this structure is tremendously persuasive.

For policy arguments (and a surprising amount of scholarly argument is actually policy argumentation) the ads/disads structure can be appropriate. As implied by its name, this structure begins with the advantages and then the disadvantages of a proposed policy, theory, or an interpretation. Obviously, one can reverse the order (going from disads to ads). And, as with the comparison structure, one can organize by point or criteria (with the ads and disads on each point).

B.   Paragraph Organization

Many students are taught a specific paragraph organization; as far as I can tell, this information is still passed along to students, despite numerous studies having shown it’s false, for two main reasons. First, too few writing teachers read studies about writings. Second, these instructions make it easier to grade student writing, as papers written this way are faster to read. They’re also less persuasive, but it isn’t clear to me how many teachers read papers with an eye to being persuaded.

Students are often taught to begin a paragraph with a main claim, then give a sentence of analysis, then an example or quote, then more analysis, and then a restatement of the claim. This method of writing requires very little research (one only needs a single example or quote) but a heavy use of the thesaurus function. As people move away from student writing, it’s fairly important to move away from this paragraph structure; it isn’t that one should never begin with a main claim, but that readers are more engaged if the paragraph organization varies.

There may be a third reason that people give students that advice, and it goes back to the conflation of topic sentence and main claim. Loosely, a sentence (or even a word or two in a sentence) in a “proof” paragraph may fulfill one (or more) of several functions:

·      main claim: just as the thesis is the main claim of the entire text, so the main claim of a paragraph (or section) is what that section is arguing for.

·      topic sentence: I use the term “topic sentence” just because that’s the common phrase, but the topic of a paragraph (or section) can be indicated through a phrase or clause (“the issue of bunnies necessarily leads to the question of hay” suggests that this paragraph will be about hay; this sentence is also a transition or “signpost”). A main claim always indicates the topic, but a topic sentence doesn’t necessarily indicate the claim.

·      transition: transitions are often indicated through words (ordinals like “first” but also some adjectives like “another” or “next” and adverbs like “similarly” as well as conjunctions) or phrases (“on the other hand”). Some people call them “metadiscourse” as they are ways of speaking directly to the reader about how the information about to come is related to the previously given information.

·      evidence: it can be difficult to distinguish between evidence and analysis (and sometimes difficult to distinguish between evidence and sub-claims)—the point is not to label everything obsessively, and these aren’t discrete categories. But, something in a proof paragraph should be carrying the burden of proof—quotes, examples, experimental results, even narratives. What constitutes appropriate evidence varies even within disciplines (whether, for instance, personal observation is relevant) and is best inferred from looking at published texts in one’s desired publishing venues.

·      analysis: in some disciplines, analysis is withheld till the data is fully presented; in others it’s expected that there be more analysis than data. Again, the appropriate proportion is best determined by looking at other texts in the journal or press series.

·      sub-claim: the more complicated one’s argument, the more likely one is to have a fair number of sub-claims. They modify, extend, or clarify the connection between the analysis and the claim.

As mentioned earlier, Richard Braddock’s analysis showed that proof paragraphs do not always begin with main claims, and various other studies have shown considerable variation in paragraph organization. It’s not uncommon, for instance, for the main claim for one paragraph to be the first sentence of the next (or previous) paragraph. Complicated or controversial claims are generally best delayed till after the evidence, but there are lots of circumstances under which one would violate that rule of thumb. My experience suggests that readers prefer a mix of paragraph structures, so that a text in which every paragraph begins with the main claim will be boring, and one in which the main claim is always at the end of paragraphs will start to seem like a bad mystery novel. Main claims are also often the second or penultimate sentences in paragraphs. They don’t always have to be stated, particularly if there are strong topic sentences, but it seems to me that readers get frustrated if there are never main claims—especially if the topic is complicated or confusing. I have been told, and this seems plausible to me, that it’s also problematic to put one’s main claims in the center of the paragraph (what is sometimes called a “chiasmatic” structure), as that is not where readers expect them.[2]

            In short, there is not a single correct way to organize the material within paragraphs (notice that this is the main claim of the previous paragraph). Having a rigid set of rules about how to organize one’s paragraphs can contribute to writing block without even making one’s writing any more effective, since, in fact, the “rules” are highly fluid. As will be discussed in the section on writing process, concerns about paragraph organization are most appropriate while revising one’s work—writing is a bit like juggling, and trying to work about invention and arrangement at the same time is simply too much at once.

            And all of this advice about the different kinds of work applies to “proof” paragraphs—that is, paragraphs that are presenting the data—and don’t apply to narration paragraphs. They also apply less well for complicated arguments; sometimes one has a paragraph (or more) that is entirely evidence, entirely analysis, or entirely transition. We give students very little instruction on narration paragraphs (except in textbooks that include a section on writing lab reports)—I’ll explain a bit more about them in the section on writing processes—but here I will simply note that they do not have the same structure or organizational principles as proof paragraphs.

---

[1] The general question of what journal or press one should try to place one’s work is best discussed with an advisor or mentor in one’s field. How to find a good advisor or mentor will be discussed in the chapter on time management. But, here, on the issue of generic expectations, I should emphasize that one shouldn’t be misled by too quick an analysis of a journal or press. Journals sometimes have special issues or guest editors, and those issues would not be representative samples.

[2] George Gopen’s A Sense of Structure has more specific advice about paragraph organization.

III. Style

            There is a recurrent joke among teachers of writing about what we tell people next to us on planes concerning what we do. As Wayne Booth said a long time ago (so long that he used the example of sitting next to someone on a train), it is rare that the conversation turns into a lively conversation about good writing. The interlocutor often expresses anxiety that we will correct their “grammar” or goes on to share what they assume we will consider horror stories of terrible “writing.” Ironically, neither interlocutor usually uses the central terms precisely (it’s ironic since their embarrassment or rant is often about lack of precision or misuse of words). What they mean by “grammar” or “writing” is something more accurately called “sentence-level correctness,” with a disturbing number of them inflected by sheer prejudice. People who pronounce Wednesday “Wendsday” probably shouldn’t try to mount any high horses about “nucular” or “aks.” A linguist friend pointed out to me that my snickering at people who pronounce pin/pen/pan indistinguishably was probably unwise, given my tendency to pronounce merry/Mary/marry identically.

            People have a very difficult time understanding that “correctness” is a rhetorical issue—it’s about context—rather than a set of rules handed down on Moses’ third tablet. Americans say “garage” with the emphasis on the second syllable, and British say it with an emphasis on the first. Using “ain’t” when someone else might use “aren’t” is not an indication of stupidity any more than is using “visit” when someone from a different region might use “talk.” In a perfect world, people would be skilled in code-switching; that is, in moving among levels of formality and dialect. Similarly, in scholarly writing, there is considerably more variation in what is considered “correct” than scholars within a specific discipline might realize. Instead of framing many concerns about style as discipline-specific (which they are), college instructors have a troubling tendency to put rules in “always” or “never” form, such as telling students they should “never” say “I,” use passive voice, use “impact” as a verb, use a long phrase that is mostly nouns, end a sentence with a preposition, and so on. Each of those moves is considered valid in some disciplines; just as there is nothing wrong with explaining to someone that one should avoid saying “totally” in some situations, there is nothing wrong with telling students that they shouldn’t use “I” in writing for the sciences and social sciences.

            It isn’t clear to me why people get so attached to our disciplinary conventions. When I have pointed out to fellow faculty that passive voice is required in some disciplines, and so on, I have more than once been told, “Well, those disciplines are wrong.” In point of fact, disciplinary preferences regarding style are not written into the fabric of the universe, and I don’t claim to know whether the angels use passive voice. I do know, however, that passive voice is useful under many circumstances.

            Hence, writers need to infer the discipline-specific rules from writing textbooks written specifically for disciplines (of which there are many), or from the published materials in one’s field.[1] The three areas in which disciplines vary more than one might guess are: diction; passive voice/agency; and hedging.

            The extent to which scholarly writing must be formal varies not only from one discipline to another, but from one journal to another. While this variation manifests itself in style choices—such as the use of “I,” slang, personal experience, humor—it signifies deeper differences about the nature of evidence and observation.

            Passive voice/agency is partially related to issues of evidence and epistemology—saying “I observed X” emphasizes the particularity of my perception whereas “X was observed” implies that anyone present could or would have seen X. But the use of passive voice/agency also correlates to the kinds of topics—there are some topics in which the agent can’t be identified, or the nature of agency is precisely what is in dispute. Passive voice is a grammatical construction (a form of “to be” and a past participle) but passive agency has to do with the logic of the sentence. “Chester bit Hubert” is active voice; it is also active agency, in that the agent of the action is the grammatical subject. “Hubert was bitten by Chester” is passive voice—the verbal is a form of “to be” and a past participle. It is also passive agency, in that the agent (Chester) of the action (biting) is not the subject, but the object of a preposition. The grammatical subject (Hubert) is the object of the action (biting). So, the logical subject is an object and the logical object is the subject. One can drop Chester out of the sentence entirely—“Hubert was bitten” (how Chester’s advocate would probably phrase it.) But, one can have active voice and passive agency. “Biting happened between Chester and Hubert” is grammatically active voice; it does not have a “to be” and past participle. It is, however, passive agency, in that the agent of action is suppressed. Thus, the various options are:

·      Active Voice/Active Agency: Chester bit Hubert.

·      Passive Voice/Passive Agency: Hubert was bitten by Chester. Hubert was bitten. Hubert was involved in a biting incident.

·      Active Voice/Passive Agency: Biting happened between Chester and Hubert. Biting happened to Hubert. There was biting between Chester and Hubert. An incident of biting occurred toward Hubert.

While passive agency obscures agency, that move is not necessarily intentional or nefarious. Instead of telling my son, “Get your socks off the coffee table,” I will often say something like, “What are your socks doing on the table?” Writing teachers often use passive voice or agency in paper comments, as it is rhetorically more effective to say, “This paper begins well, but seems to get confusing on the second page” than “You start the paper well, but you confuse your reader beginning on the second page.” Using “you” tends to make authors defensive, and reinforces the unproductive sense that one’s self, rather than one’s writing, is being criticized. Passive voice/agency sometimes, but not always, reduces the drama of the situation; sometimes it heightens drama by reshifting the focus. So, for instance, “Hubert was victimized by Chester’s biting” emphasizes Hubert’s victimization better than does “Chester bit Hubert.”

            It seems to me that scholars in the social sciences and sciences often rely heavily on passive agency because they are writing about situations in which agency is unclear—economists say, “The market went up” because the argument is about the agent of economic change; it isn’t clear what caused the market to go up. In such fields, or at least in such arguments, insisting on active agency is unhelpful at best (notice the passive agency in this sentence). I’m not saying that concern about agency is always trivial. While the use of passive agency sometimes results in sentences that hurt the grammar pedant in me (“Hello, this is your doctor’s office calling with your lab results”), and I know that the answer is to tell the inner grammar pedant to shut up, passive voice and agency do have ethical dimensions. People tend to use passive agency for members of the outgroup, especially when describing good behavior, and passive voice/agency for themselves and members of the ingroup when describing inappropriate, harmful, or unethical behavior. As Elliot Aronson says, the most common form that the looks-like-an-admission-of-responsibility-but-isn’t statement takes is “mistakes were made.” And Richard Lanham is right to argue that nominalization—turning verbs into nouns—does make appalling actions more palatable (because agency-free and abstract), and is a crucial process in political euphemisms. To call for the “pacification” of a village seems to be an abstract process, with no actions by humans, and is far more attractive than saying that American troops “bomb” or “eradicate” villages. Passive voice/agency is not just a stylistic choice, but it is neither inherently right nor wrong.

            Ken Hyland argues for dividing “metadiscourse” into two functions: interactive and interactional. Loosely, “metadiscourse” is the term that linguists and rhetoricians use for the parts of discourse that signal to the reader how they are to understand the propositions being conveyed. It is talk about the talk. Interactive metadiscourse helps the reader understand the text better through giving explicit cues about the text. Transitions, for instance, help a reader understand where in the text one is, and how the currect text relates to material previous or subsequent. Words and phrases such as “In addition,” “next,” and “another” indicate to the reader that this information is what Hyland calls “additive.”

            Metadiscourse isn’t necessarily deliberate, and hence can be useful for the author as well as the reader. For instance, when an additive transition marker is in the topic sentence, that is often a sign that this section of the text has a list structure. Readers have a tendency to get irritated by lists with too many elements, so an entire article or chapter that is a list will tend to be hard work for a reader. Similarly, contrastive transition markers (but, however) that occur throughout a paragraph signify a logical structure that is either continually correcting, or bouncing between both sides of an antithesis. Either structure is difficult for a reader to follow.

            Other interactive resources include “frame markers,” “endorphic markers,” and “evidentials,” and “code glosses” (Hyland 50-51). Frame markers signal to the reader the boundaries of sections within a text (“in conclusion”), and endorphic markers tell the reader how this material relates to material elsewhere in the text (“as will be discussed later,” “see below”). People in rhetoric often refer to both these kinds of metadiscourse as “signposts,” as they function in the same way that “you are entering” and “you are leaving” signs function for travelers. They’re rarely necessary in short pieces of writing, and that kind of work is often done visually (extra space between paragraphs, sub-headings, roman numerals), but they become increasingly necessary for the reader and writer as an argument gets longer and more complicated. Some people call that kind of metadiscourse “scaffolding,” as it can hold up an argument the same way that scaffolding holds up a building while it’s being built. Once the building is complete, the scaffolding is no longer necessary, and is an active distraction. Writers who are resistant to a highly recursive writing process often resist using scaffolding, even in the beginning, as they don’t like it in final products. But that’s like a builder refusing to put up scaffolding while building because it won’t be part of the final design.

            Hyland uses the term “code glosses” for words and phrases that mark moments when an author is engaged in explanation (or analysis)—the rephrasing or amplifying of what another author has said. A recurrent problem in dissertations (at least in my field) is that the evidence doesn’t quite earn the conclusions the student is drawing. Sometimes that weakness is simply the consequence of the short amount of time that a student has to write a dissertation—two years for research and writing—and thus will be resolved when s/he has more time to gather additional evidence (or refine the conclusions). But it often seems to me that the student has trouble with providing adequate analysis of the evidence. One way to check to see if there is adequate analysis is to look for “code glosses.”

            Evidentials, “guide the readers’ interpretation and establish an authorial command of the subject” (Hyland 51). Hyland distinguishes them from “stance” metadiscourse, which may be too fine a distinction for some, but it seems to me useful because evidentials are precisely what is crucially important in (and often missing from) summaries of dissertations, grant applications, book proposals, and literature reviews. An ineffective literature review is little more than an annotated bibliography in a different format. An effective literature review has both “stance” (discussed below) and “evidential” metadiscourse, describing the existing literature in a way that persuade the reader the author has reviewed it accurately and critically, and can place his/her contributions in relation to what exists.

            An author’s authority is also strengthened by several of the “interactional” metadiscursive moves an author can make.  “Boosters” (choosing a word like “proves” rather than “suggests” or using an adverb like “clearly”) and “hedges” (choosing a word like “suggests” rather than “proves” or modifiers like “possibly” or “perhaps”) both function to enhance the author’s credibility, unless overused or in the wrong discipline. I have noticed that it makes some readers uncomfortable, as they see any hedging as unnecessary, whereas other readers describe it as something that shows an appropriate degree of self-reflection on the part of the author. In some disciplines, the hedging is part of the “implications” section, especially in calls for future research. Saying that a topic needs to be studied in a certain way, or that other studies need to be conducted on this aspect, or other moves along those lines, imply that one’s own research is limited.  If an author uses boosters when they aren’t merited, or too often, then readers can become skeptical; excessive use of hedges will make an author what credibility the piece does have.

            “Engagement markers” (explicitly addressing the reader, often with “you”) and “self mention” (using first person pronouns or referring to “the author”) are, as Hyland says, useful for meeting “readers’ expectations of inclusion and disciplinary solidarity” (Hyland 54, essentially ensuring that the reader and author are members of the same ingroup) and “pulling readers into the discourse at crucial points” (Hyland 54). “Attitude markers” (words and phrases that communicate the author’s evaluation of the material s/he is discussing) engage a reader; Hyland distinguishes epistemic evaluations from affective ones (“This example clearly demonstrates” versus “This example unfortunately demonstrates”), a distinction that matters more in some disciplines than others—in some, it seems to me (hedging and self mention), epistemic evaluations are permitted, but affective ones are not.

            I’ve relied heavily on Hyland’s taxonomy not because it’s the only one, and not even because I think he has precisely divided the phrases correctly, but simply because, as he says, students are not generally explicitly taught metadiscursive practices. They (we) are expected to infer them. Hyland argues for explicitly teaching those practices to students, and I have seen very effective “writing for publication” courses that used such assignments to good effect. But, the point is not to go through and label moves. As indicated in the hedging/self mention above, they aren’t necessarily discrete, and I could easily imagine someone arguing that Hyland has either too few or too many distinctions. While Hyland emphasizes their importance in reading (and I think it’s true that students who misread texts often do so because they fail to pay attention to metadiscoursive cues about how the author wants the material to be read), I have found the concept most useful for looking helping writers look more critically at our own writing. Identifying whatever our recurrent markers are can tell us something about how we imagine our rhetorical situation. And noting that some of the cues are rarely or extremely commonly used can indicate problems readers are likely to have—a complete absence of frame and endorphic markers can suggest that the overall organization may be unclear even to the author, too many boosters may imply that it is argument by assertion.

---

[1] The two texts that I have found helpful for writers across disciplines are Joseph Williams’ Style: Ten Lessons in Clarity and Grace and George Gopen’s A Sense of Structure. Both put emphasis on reader expectation, rather than rigid rules, and on how style functions in particular circumstances. Williams’ explanation of agency, and his exercises for practicing how and when to shift between active and passive, is excellent.