When I first started this blog, I decided that having key principles and strategies as a permanent part of the homepage would be efficient. I couldn’t properly envision what blogging would be like, but I did anticipate that there would be a tension between wanting each post to stand alone and yet to contribute to an overall picture of academic writing. Having some basic precepts accessible in manageable bits allows me to link back to them without disrupting the flow too much. Those original posts, however, tended to be both general and brief, meaning that certain aspects of the topics were given short shrift. Today, I’d like to talk more about paragraphs in order to discuss an issue that was mentioned only in passing in the original post.
In that post, I listed four things that I wished people knew about paragraphs; the first one was that they are very important. After making that pronouncement, I went on to discuss the other three in more detail: topic sentences, internal cohesion, and the rhetorical significance of length. But my claim about the preeminence of the paragraph was strangely lacking in elaboration. Recently I came across a quote that made me want to articulate my commitment to the paragraph with greater precision. In a post on his blog, Research as a Second Language, Thomas Basbøll made the following claim: “The paragraph is really the smallest unit of scholarly composition.”
This assertion totally stopped me in my tracks. When you spend a lot of time making strong claims about a topic, it can be unsettling to see someone making an even stronger claim. I think of it as my job to say that paragraphs are super important, often in the face of sceptical students. In my experience, most graduate student writers take paragraphing insufficiently seriously. By this I mean that their paragraphs are generally too short, with inadequate attention to clear topics and thematic development. Many novice writers pay too much attention to individual sentences, on the one hand, and the whole text, on the other, leaving little attention left for paragraphs. But in all my exhortations to take paragraphs more seriously, I had never thought to say that they are the smallest unit of composition.
While I don’t ultimately think the claim is true, I admire how decisively it tries to counteract our preoccupation with sentences. I do love a beautiful sentence, but a desire for perfect sentences can be a trap for many writers. Too much attention to sentences—especially early in the drafting process—can slow us down and get in the way of vigorous editing. Most of us need to think more about the way sentences work together than we do because it is sentences-working-together-in-paragraphs that propels the text forward. This notion of the paragraph as the prime locus of narrative development lends credence to Basbøll’s claim. Any given sentence might let us down as readers, but we generally push on in the hopes that the paragraph will give us what we need. When the paragraph fails, it won’t necessarily matter if it is composed of strong sentences.
This valuable emphasis on paragraphs can’t, however, change the fact that sentences are our basic unit of composition. In fact, we have something of a natural mismatch: we write sentence-by-sentence, but readers attempt to digest our writing in bigger chunks. If we’re not intentional enough about those bigger chunks, our readers may have trouble discerning our meaning, even if each sentence is fine. As is so often the case with writing issues, this tension is best addressed through the revision process. Since we do compose in sentences, we are unlikely to shift our attention towards paragraphs during the initial drafting stage. But our editing process should be geared towards the eventual creation of strong paragraphs. One of the reasons that the reverse outline is such a powerful strategy is that it takes the paragraph as its fundamental unit of analysis. Paragraphs are as much engineered as they are written: we write in sentences, but we construct meaning by revising and rearranging those sentences into coherent paragraphs.
If your paragraphs are underdeveloped or incoherent, it won’t matter so much that they may be made up of perfectly sound sentences. Academic writing is a matter of accumulation; each individual sentence will only be able to carry so much weight. When we shift some of the focus away from sentence composition and towards paragraph construction, we are taking our reader’s needs into account and giving ourselves a way to increase the coherence of our text. By asking ourselves what our paragraphs are doing for us, we are improving our chances that our paragraphs are doing what our readers need them to do.
Explorations of Style 
Rachael, as always, you state so simply and brilliantly what I have been struggling to articulate for years. I love the clarity you bring to the mismatch between how we write (sentence-by-sentence) and how we read (paragraph-by-paragraph). I’ve added this post to our site for tutors.
Thanks, Heather!
Reblogged this on Rakhat Zholdoshalieva.
Hi Rachael,
Great post. I think we completely agree. In the post I link to (in the post you link to), I try to distinguish between “writing” and “composition” (see also Jonathan’s comment about Stein and my response), and keep in mind I mean scholarly, i.e., “academic”, composition. I agree that sentences are important, and I can acknowledge good writing even where I can see great flaws in composition.
There’s always room for exceptions in a paper. Sometimes a stand-alone three-word “paragraph” works. However, as a general rule, if you’re not composing paragraphs, you’re simply not writing academically. A sentence alone is not enough to carry the sort of knowledge that is needed. Perhaps there’s a really good writer somewhere who can pull it off with sentences alone. But it’s doing it the hard way.
Cheers,
Thomas
Thanks, Thomas. I think your distinction between writing and composition is a really helpful one. And I like the way you link composition to critical thinking and thus to broader concerns about how we communicate our ideas.
Pingback: Link love: language (58) | Sentence first
A paragraph has no defined length. Any portion of text with white space around it can be called a paragraph. It can be a single word, as in news columns — or anywhere else. The paragraph surely aims for readability by delivering text in manageable chunks, which makes some communicators urge the use of “chunking”, a good term for it. Chunking breaks a slab of information up into a more appealing visual arrangement. A chunk likewise has no defined length, but generally averages short, brevity its soul. So judge for readability; i.e., not for the teachability of topic sentence, development, conclusion. There was a survey (somewhere, somewhen) on how real-world paragraphs were composed. The survey’s outcome was that only a minority of paragraphs followed the topic sentence-development-conclusion pattern; most ignored it. It laid the school-room version of paragraphing at the door of a Scotsman named Bain, fl. mid-C19th. Bain’s view of the paragraph caught on mainly because it was so teachable; i.e., it made the teaching of writing easier for the teacher. And of course the Bain pattern made for cohesion, coherence, clarity, and all those other good things. But it’s not the only way, so let’s not make it normative.
Thanks, Roger. I agree that paragraphs can be of any length, but I don’t agree that their creation should be primarily dictated by the demands of visual arrangement on the page. I would argue that the creation of a paragraph has less to do with spatial readability and more to do conceptual readability. Of course I agree that any formula of proper paragraph construction is sure to be flawed, but I do think that the internal demands of paragraphs in academic writing are going to lead us to the creation of something fairly consistent. There are many exceptional paragraph styles: introductory paragraphs, transitional paragraphs, topical paragraphs (that themselves set the stage for paragraphs to come), and impact paragraphs of various sizes and shapes. But I still think that many paragraphs in academic writing will benefit from clear topics and the space to to develop a coherent theme. I was so interested to see in your comment a source for the notion that paragraphs need a conclusion; I find that to be one of those apocryphal demands that weighs heavily on some writers. Paragraphs almost never need something as robust as a conclusion, and I find that many writers struggle to provide any sort of conclusion without falling into laborious repetition. Thanks for your comment!
I first came across Bain-pattern criticism in a textbook from long ago (so old that it may not have had an ISBN) about writing development — that phrase may have served as the book’s title. Bain is easily found online now, as, for ex., “Alexander Bain’s Contributions to Discourse Theory”, in College English, vol. 44 #3 March 1982; and “Alexander Bain and the genesis of paragraph theory”, in the Quarterly Journal of Speech,vol. 58, issue 4, 1972 an placed online Jun 2009. They shed light on the starting point for the teachable paragraph, raise it out of theC19th mists, and make one wonder how the schools managed before Bain.
As for the survey of real-world paragraphs vs. the school-room version that I read of in “Writing Development” (?), that I have yet to unearth. But it was quite confident about the minority standing of the Bain-pattern.
I think we should push back against Roger’s rather anarchistic approach to paragraphs. Let me suggest the following: a paragraph consists of at least six sentences and at most 200 words. It says one thing and supports it. And it’s the sort of thing that can be made in 27 minutes.
It’s easy to dismiss this idea as too “normative”, but what’s wrong with normative? (Especially in the very “normal” environment of scholarly prose.) Would we say that the piano teacher is too normative about how to play arpeggios? Is the sensei too normative about the various elements of the kata? Is the skating coach too normative about the Axel? Is the swimming instructor too normative about the breaststroke?
I don’t think “chunking” is enough. The idea is precisely to give the text the particular kind of “composure” that academic writing requires. As Rachael points out, you’re not breaking up the text for the eye (“visual readability”) but the mind (“conceptual readability”).
PS I agree that topic sentence-development-conclusion is a red herring. I’m not sure anyone teaches that any more.
Re academic writing vs. everything else that we see in print: Publishers are well aware of the demand for visual appeal. The competition is lethal. Just compare books from times past with books now, particularly in regard to page lay-out, typography, use of color, illustration, and various other design features, all of which Rudolf Flesch’s readability formula, dating from the mid-C20th, did not and could not take into account, nor any more recent formula since that time.
Rachael, what a great post! This morning, I was editing a section for a client, who is an academic writer. I did a reverse outline to help give her clarity on her structure. I was just typing up my explanation of the reverse outline and, honestly, struggling a bit. Luckily, I took a break from that email to check my blogs, and what luck! I’m going to send this post to my client because you set up the importance of analyzing writing paragraph-by-paragraph so clearly. Thank you!
Thanks for reading and sharing, Morgan!
Surely its not the physical length but the content that defines a paragraph or any “chunk” of knowledge. I am reminded of Mozart being criticised for having “too many notes” 🙂
I agree, regulating paragraphs into 200 words and six sentences may be
Bain-al but it certainly makes teaching composition more convenient for
teachers. On the one hand, some readers accept unbroken pages with
no paragraphing at all. For others, 100-word, three sentence chunks.
A cross-language survey could be relevant: How do the continentals do it,
in their own first languages and in translation. See also James Miller`s Is Bad Writing Necessary, from the 90s.
Thanks, Roger. That Miller article is one of my favourites; I talk about it briefly here, but I’d love to write about it in more detail.
Pingback: beginners in writing | About Resarch
Pingback: Wissenschaftliches Schreiben