Month: June 2019

This is a piece I published in Aeon last year about the persistence of the five-paragraph essay, which has evolved into the five-chapter dissertation and the five-section journal article.  Formalism reins supreme.  Here’s the link to the original.

Schools and colleges in the United States are adept at teaching students how to write by the numbers. The idea is to make writing easy by eliminating the messy part – making meaning – and focusing effort on reproducing a formal structure. As a result, the act of writing turns from molding a lump of clay into a unique form to filling a set of jars that are already fired. Not only are the jars unyielding to the touch, but even their number and order are fixed. There are five of them, which, according to the recipe, need to be filled in precise order. Don’t stir. Repeat.

So let’s explore the form and function of this model of writing, considering both the functions it serves and the damage it does. I trace its roots to a series of formalisms that dominate US education at all levels. The foundation is the five-paragraph essay, a form that is chillingly familiar to anyone who has attended high school in the US. In college, the model expands into the five-section research paper. Then in graduate school comes the five-chapter doctoral dissertation. Same jars, same order. By the time the doctoral student becomes a professor, the pattern is set. The Rule of Five is thoroughly fixed in muscle memory, and the scholar is on track to produce a string of journal articles that follow from it. Then it’s time to pass the model on to the next generation. The cycle continues.

Edward M White is one participant in the cycle who decided to fight back. It was the summer of 2007, and he was on the plane home from an ordeal that would have crushed a man with a less robust constitution. An English professor, he had been grading hundreds of five-paragraph essays drawn from the 280,000 that had been submitted that June as part of the Advanced Placement Test in English language and composition. In revenge, he wrote his own five-paragraph essay about the five-paragraph essay, whose fourth paragraph reads:

The last reason to write this way is the most important. Once you have it down, you can use it for practically anything. Does God exist? Well you can say yes and give three reasons, or no and give three different reasons. It doesn’t really matter. You’re sure to get a good grade whatever you pick to put into the formula. And that’s the real reason for education, to get those good grades without thinking too much and using up too much time.

White’s essay – ‘My Five-Paragraph-Theme Theme’ – became an instant classic. True to the form, he lays out the whole story in his opening paragraph:

Since the beginning of time, some college teachers have mocked the five-paragraph theme. But I intend to show that they have been mistaken. There are three reasons why I always write five-paragraph themes. First, it gives me an organizational scheme: an introduction (like this one) setting out three subtopics, three paragraphs for my three subtopics, and a concluding paragraph reminding you what I have said, in case you weren’t paying attention. Second, it focuses my topic, so I don’t just go on and on when I don’t have anything much to say. Three and only three subtopics force me to think in a limited way. And third, it lets me write pretty much the same essay on anything at all. So I do pretty well on essay tests. A lot of teachers actually like the five-paragraph theme as much as I do.

Note the classic elements of the model. The focus on form: content is optional. The comfortingly repetitive structure: here’s what I’m going to say, here I am saying it, and here’s what I just said. The utility for everyone involved: expectations are so clear and so low that every writer can meet them, which means that both teachers and students can succeed without breaking a sweat – a win-win situation if ever there was one. The only thing missing is meaning.

For students who need a little more structure in dealing with the middle three paragraphs that make up what instructors call the ‘body’ of the essay, some helpful tips are available – all couched in the same generic form that could be applicable to anything. According to one online document by a high-school English teacher:

The first paragraph of the body should contain the strongest argument, most significant example, cleverest illustration, or an obvious beginning point. The first sentence of this paragraph should include the ‘reverse hook’ which ties in with the transitional hook at the end of the introductory paragraph. The topic for this paragraph should be in the first or second sentence. This topic should relate to the thesis statement in the introductory paragraph. The last sentence in this paragraph should include a transitional hook to tie into the second paragraph of the body.

You probably won’t be surprised that the second paragraph ‘should contain the second strongest argument, second most significant example, second cleverest illustration, or obvious follow-up to the first paragraph…’ And that the third paragraph ‘should contain the third strongest argument…’ Well, you get the picture.

So where does the fetish for five come from? In part, it arises from the nature of sentences. Language conveys meaning by organizing words into an order governed by rules. These rules are what allows the listener to understand the relationship between these words in the way intended by the speaker. The core unit of conveying meaning via language is the sentence, and the rules that define the structure of the sentence are its syntax. By its nature, syntax – like the five-paragraph essay – is all form and no content. Its entire utility derives from the fact that a particular syntactical structure can be used to convey an infinite number of meanings.

Form, therefore, is not just a crutch for beginners to use in trying to learn how to write; it’s also the central tool of writers who are experts at their craft. In his lovely book How to Write a Sentence: And How to Read One (2011), Stanley Fish makes the point that, in writing, form comes before content:

The conventional wisdom is that content comes first – ‘you have to write about something’ is the usual commonplace – but if what you want to do is learn how to compose sentences, content must take a backseat to the mastery of the forms without which you can’t say anything in the first place.

Think of all the syntactical forms that exist to define different kinds of relationships between words in the service of making a point. For example:

If ___, then ___.
Some argue ___, but I argue ___.
On the one hand, ____; but on the other hand, ___.

Consider key words that signal a particular kind of relationship between words, ideas and sentences:

Addition: also, moreover
Elaboration: in short, that is
Example: for instance, after all
Cause and effect: accordingly, since
Comparison: likewise, along the same lines
Contrast: although, but
Concession: admittedly, granted
Conclusion: as a result, therefore

The last set of examples comes from They Say, I Say (2006) by Gerald Graff and Cathy Birkenstein, which seeks to explain the rhetorical ‘moves that matter in academic writing’. In the appendix, they list a set of syntactical templates that extend over 15 pages. Graduate students in my class on writing find these templates very useful.

The point is that learning to write is extraordinarily difficult, and teaching people how to write is just as hard. Writers need to figure out what they want to say, put it into a series of sentences whose syntax conveys this meaning, arrange those sentences into paragraphs whose syntax carries the idea forward, and organize paragraphs into a structure that captures the argument as a whole. That’s not easy. It’s also not elementary. Fish distills the message into a single paradoxical commandment for writers: ‘You shall tie yourself to forms and the forms shall set you free.’ The five-paragraph essay format is an effort to provide a framework for accomplishing all this.

The issue is this: as so often happens in subjects that are taught in school, the template designed as a means toward attaining some important end turns into an end in itself. As a consequence, form trumps meaning. For example, elementary-school students learn to divide a number by a fraction using this algorithm: invert and multiply. To divide by ½, you multiply the number by two. This gives you the right answer, but it deflects you from understanding why you might want to divide by a fraction in the first place (eg, to find out how many half-pound bags of flour you could get from a 10-pound container) and why the resulting number is always larger than the original.

Something similar happens with the five-paragraph essay. The form becomes the product. Teachers teach the format as a tool; students use the tool to create five paragraphs that reflect the tool; teachers grade the papers on their degree of alignment with the tool. The form helps students to reproduce the form and get graded on this form. Content, meaning, style, originality and other such values are extraneous – nice but not necessary.

This is a variation of Goodhart’s Law, which says: ‘When a measure becomes a target, it ceases to be a good measure.’ For example, if test scores become the way to measure student and teacher success, then both parties will work to maximize these scores at the expense of acquiring the underlying skills that these scores are supposed to measure. Assess students on their ability to produce the form of a five-paragraph essay and they will do so, at the expense of learning to write persuasive arguments. The key distinction here is between form and formalism. A form is useful and necessary as a means for achieving a valued outcome. But when form becomes the valued outcome, then it has turned into formalism.

An extreme example of this phenomenon has emerged in the growing field of machine-graded essays. Having experts grade large numbers of papers, such as for the advanced-placement composition exercise that White took part in, is extremely labor-intensive and expensive, not to say mind-numbing. So the Educational Testing Service (ETS) and other companies have come up with automated systems that can take over this function by deploying a series of algorithms that purportedly define good writing.

The problem, of course, is that these systems are better at identifying the formal characteristics of these essays than at discerning their meaning. To demonstrate this Les Perelman, along with Louis Sobel, Milo Beckman, and Damien Jiang, invented a Babel Generator that is capable of producing essays from any three keywords, and of gaining a perfect score on the ETS assessment. They did this by gearing the generator to the ETS algorithms, which allows them to produce the desired measure without all that messy stuff about creating logical and compelling arguments. Here’s the first paragraph of a Babel Generator essay defined by three keywords: classroom, pedagogy, and inequality:

Classroom on the contradiction has not, and no doubt never will be aberrant. Pedagogy is the most fundamental trope of mankind; some with perjury and others on amanuenses. A howling classroom lies in the search for theory of knowledge together with the study of philosophy. Pedagogy is Libertarian due to its all of the concessions by retorts.

As you can see, the algorithm rewards big words and long sentences rather than meaning. (Try it yourself.)

Of course, students still need to provide some semblance of subject matter for their essays. But there are plenty of handy resources available to produce relevant content on demand. When I was in school, the key resource for students who needed to write an essay on some topic or other was the encyclopaedia. In my family, it was the World Book Encyclopedia, which offered glossy pages and ample illustrations, and which used fewer big words than the canonical but stuffy Encyclopaedia Britannica. Look up the topic, read a short summary piece, and then crib it for your paper. In the 1950s and ’60s in the US, encyclopaedia salesmen sold these pricey products door-to-door, and their pitch was compelling: ‘Do you want your kids to have a good life? Then they need to succeed in school. And the encyclopaedia is the key to school success, the added element that will move your children ahead of their peers.’ It worked. Owning an encyclopaedia (26 volumes, $500) became the badge of the middle-class family – to the point where mid-century sociologists used encyclopaedia ownership as a key criterion for coding subjects as middle class.

The multivolume encyclopedia has receded into history; the last hard-copy Britannica was published in 2010. Now students use Google as their primary ‘research’ tool, and the top search result for most topics tends to be Wikipedia. The latter serves the same function for students – capsulized and bowdlerized content ready for insertion into the five-paragraph essay. Plug and play. The perfect tool for gaming the system of producing papers for school.

It is possible to teach students how to write as a way to make meaning rather than fill pots. The problem is that it’s much more difficult for both student and teacher. For students, it takes a lot longer to get better at writing this way, and the path to improvement is littered with the discouraging wreckage of dysfunctional sentences and incoherent arguments. And for teachers, the difficulty of teaching the skill this way undermines their sense of professional competence. In addition, grading papers for meaning takes a lot more time and involves a lot more judgment than grading for form – which, after all, can be done by a computer.

Be clear, be concise, be direct, focus on actors and actions, play with language, listen for the music

Carrying out this kind of teaching calls for concentrating effort at two levels. One is teaching students how to make meaning at the sentence level, using syntax to organise words to say what you want them to say. Books on writing at the sentence level – my favorites are Style: Lessons in Clarity and Grace (1981) by Joseph Bizup and Joseph M Williams, now in its 11th edition; and Fish’s How to Write a Sentence – lay out a series of useful rules of thumb: be clear, be concise, be direct, focus on actors and actions, play with language, listen for the music. The other is teaching students how to make meaning across an entire text, using rhetorical moves that help them structure a compelling argument from beginning to end. My favorite book in this genre is Graff and Birkenstein’s They Say, I Say. I use all three in a graduate class I teach on academic writing.

I’ve also developed my own set of questions that writers need to answer when constructing an analytical text:

1. What’s the point? This is the analysis issue: what is your angle?
2. Who says? This is the validity issue: on what (data, literature) are you basing your claims?
3. What’s new? This is the value-added issue: what do you contribute that we don’t already know?
4. Who cares? This is the significance issue, the most important issue of all, the one that subsumes all the others. Is this work worth doing? Is the text worth reading?

But, you ask, aren’t these just alternative sets of rules, much like the Rule of Five? I say no. One difference is that these are clearly labelled not as rules but rules of thumb. They are things to keep in mind as you write (and especially as you edit your writing), many of which might be in tension with each other, and which you must draw upon or ignore as needed. Another difference is that they resist the temptation to provide a rigid structure for a text of the kind that I have been discussing here. Deal with issues in the literature where it helps to frame and support your argument rather than confining it to the lit-review ghetto. And don’t make the reader wait until the conclusion to find out what gives the text significance; most people would stop long before this point.

Rules of thumb call for the writer to exercise judgment rather than follow the format. Of course, it takes more time and effort to develop writerly judgment than it does to follow the shortcut of the five-paragraph essay. Form is harder than formalism. But the result is a text that does more than just look like a piece of writing; it makes meaning.

Let’s turn away from the ideal case – learning to write for meaning – and dive back into the real world: teaching school students to write by filling five pots with words. When students get to college, their skills in writing five-paragraph essays start to pay off big time. Compared with high school, the number of papers they need to write in a semester grows exponentially, the required length of papers also shoots up, and there is increasing expectation that these papers demonstrate a bit of professional polish. This pressure to turn out a lot of reasonably competent writing in a short period of time puts a premium on a student’s skills to produce text efficiently. And once again, the Rule of Five comes to the rescue. Nothing aids efficiency better than an easily reproducible template. This leads to two elaborations of the basic model.

The first is a simple extension of the model into a format with more than five paragraphs. The length is greater but the structure is the same: a general claim, followed by three pieces of evidence to support it, leading to a conclusion. The college version of the model also ups the ante on the kind of content that is deemed acceptable. Increasingly, the generic synthesis sources that were so helpful in high school – variations on the old encyclopaedia – are no longer sufficient. This is particularly true in selective colleges, where faculty members expect students to gain familiarity with this thing that they call ‘the literature’. Cribbing from the commons is bush league; if you’re Ivy League, you need to crib from the best – refereed journal articles by top scholars. Plug in a topic, and Google Scholar provides you with the most cited pieces on the topic. You don’t have to read them, just cite them as evidence in sections two, three and four.

The second version of the model is for students who are thinking about graduate school. They can’t settle for supporting an argument with just three sources; they need to produce ‘research’. This means that they need to define an issue, draw on the literature about that issue, develop a method for gathering data about the issue, analyze the data, and draw conclusions. Sounds complicated, but relax: it’s really not that hard. The Rule of Five is up to the challenge. The paper format contains five standard sections. All you have to do is fill them with plausible content. Here’s the model:

Section 1: Introduce the argument
Section 2: Summarize the relevant literature
Section 3: Spell out your research method
Section 4: Present your findings and analyze them
Section 5: Draw conclusions

The argument is – whatever. The literature is a few things you found on Google related to the argument. The method is how you’re going to find data that could plausibly inform the argument. Findings are some things you encounter that might support your point (think evidence one, evidence two, evidence three from the five-paragraph model). And the conclusion is that, wow, everything lines up to support your original claim. QED. But now suddenly your writing is telling the world: I’m ready for graduate school.

The transition from the college research paper to the doctoral dissertation is not as big a jump as you might think. The Rule of Five lives on in the canonical structure for the dissertation, which by now should look familiar:

Chapter 1: Introduction
Chapter 2: Review of the literature
Chapter 3: Methods
Chapter 4: Analysis/findings
Chapter 5: Conclusion

Guides on dissertation-writing specify the content of each of the five chapters in detail, with this detail looking remarkably similar across guides. Chapter 1 is supposed to have a problem statement and list of research questions. Chapter 2 needs to cover both the theoretical and empirical literature relevant to the research questions. Chapter 3 needs to spell out research design, measures used, research procedures, and modes of analysis employed. Chapter 4 summarizes the findings of the research and provides analysis of these results. And Chapter 5 covers four canonical areas: summary of results, conclusions, limitations of the study, and recommendations for future research.

A dissertation is not that difficult if you know the algorithm and produce something that looks and feels like a dissertation

Of course, you do have to fill up these five chapters with content, and the total length can run from 15,000 to 80,000 words. But you have years to do all this. And graduate school helpfully provides you with the content you need. Courses teach you how to create research questions, what the literature says about your particular subfield of expertise, what methods of data collection and analysis can best be used in this field, how to demonstrate the validity of your findings, and how to draw credible conclusions from your analysis. Pick a topic and pick a method, and the rest is plug and play. Once those decisions are made and the data gathered, the dissertation more or less writes itself.

A telling sign of formalism is that chapter titles in dissertations frequently assume the titles used in the five-chapter outline. Chapter 1 is not ‘An Introduction to Topic X’; it’s just ‘Introduction’. Chapter 2 is ‘Review of the Literature’; 3 is ‘Methods’; 4 is ‘Analysis’; and 5 is ‘Conclusion’. Specifying content, personalizing the presentation of results, tailoring the format to the demands of your own study – all of these are either not needed or forbidden. Your job is to reproduce the form of the five-chapter dissertation, and you do so, literally.

Given how generic the format is, it’s not surprising that enterprising companies are willing to go one step further and actually produce the dissertation for you on demand, for the right price. As with the Babel Generator, turning out a dissertation is not that difficult if you know the algorithm and produce something that looks and feels like a dissertation. Ads for these websites kept popping up as I was searching Google for information about the five-chapter dissertation. So I checked out the most prominent of these (the one that paid for placement highest on the list), called GradeMiners. They would produce any kind of school paper, but dissertations were one of their specialties. Drop-down menus allowed you to make the appropriate selection. I chose PhD dissertation, APA style, 100 pages, ‘professional quality’, ‘a top writer in this subject to do my work’, ‘professional quality check for my order’, 50 sources, in English, and on the topic ‘US Curriculum History’. On the ‘urgency’ menu, I selected that I wanted it within 30 days. The bottom line: I could get all this in a month for $9,623.99. Really, not a bad deal. For a little extra money, they will also carry out a plagiarism check. After all, there’s nothing worse than a ghostwriter who cheats by plagiarizing someone else’s work.

This brings us to the top level of my examination of the Rule of Five, the way that this form shapes the dominant genre of research production used by the professional scholars in the professoriate – the refereed journal article. This is the medium that governs the process of hiring, promotion and tenure within the academic profession. It’s the way to get ahead and stay ahead in your career – the way to establish your reputation, gain a following, and win accolades. And in order to get past the gatekeepers in the process – editors and reviewers at top-ranked academic journals – you need to produce papers that meet generally accepted standards. You need papers that look like, feel like, and sound like the canonical journal article. As we have seen at the lower levels, the content can be nearly anything, as long as the form is correct.

The journal-article version of the Rule of Five is known by the mnemonic IMRaD (or IMRAD), which identifies the labels and order of the conventional paper. The letters stand for the required sections in the proper order: introduction, methods, results, and discussion. Check them off, and you’re done.

But wait a minute, you say; this is only four sections. What happened to the literature review? Well, it turns out that the lit review is incorporated within the introduction. In a short journal article, prior literature might take up only a paragraph or two of the text, so why waste a whole section on it?

If you choose not to write by the numbers, you risk alienating teachers, editors, reviewers and readers

Some critics, of course, have pointed out that the IMRaD format is a bit, you know, rigid. Helen Sword wrote a book called Stylish Academic Writing (2012) that I use in my own writing class. In it, she encourages scholars to break free of the rhetorical constraints that tradition imposes on scholarly publication. But she realises she is trying to roll back the tide. For readers and writers alike, IMRaD is simply too handy to give up:

This write-by-numbers approach prompts researchers to plan their research methodically, conduct it rigorously, and present it coherently, without leaving out any crucial information. Moreover, a conventional structure is relatively easy for new academics to learn; all they have to do is follow models established by others before them. Readers, meanwhile, know exactly where to look for key findings. They can skim the abstract, mine the literature review, scan the data, and grab the conclusions without wasting valuable time actually reading.

I love the last line – ‘without wasting valuable time actually reading’. This is the whole point of the Rule of Five, isn’t it? It makes scholarly writing easy to learn, easy to read, and easy to evaluate. Like the five-paragraph essay and the five-chapter dissertation, IMRaD reduces the cognitive load involved in teaching, learning, producing, reviewing and consuming academic texts. If you choose not to write by the numbers, you risk alienating teachers, editors, reviewers and readers. You have a big incentive to make their lives easy, which will then increase the likelihood that you will succeed.

This is my point. The Rule of Five spells out issues that need to be addressed in any piece of analytical writing: argument, frame, evidence, analysis, conclusion. If you don’t address these issues, then you are not doing an effective job of presenting your work. But by addressing them only in this order, and confining each function of the argument to a hermetically sealed location within the paper, you turn a useful set of guidelines into an iron cage. It’s dysfunctional – to say nothing of off-putting, infantilizing and intellectually arid. But, then again, it makes life easier for all concerned. So it’s not going away soon.