A post from Shady Characters

The imminent death of the paragraph

Edwin Lewis, The History of the English Paragraph (1894), page 11. (Image courtesy of archive.org.)
Edwin Lewis, The History of the English Paragraph (1894), page 11. Lewis’s PhD thesis is a pleasant enough read, though his history of the pilcrow disagrees somewhat with the accepted story described by Malcolm Parkes in Pause and Effect. (Image courtesy of archive.org.)

As I mentioned last time, I recently took part in a workshop on the subject of “punctuation in practice”. My presentation there was titled “Ghostwritten: the vanishing pilcrow”, and it traced the life of the paragraph mark from ancient marginal dash (—) to medieval capitulum (¢) to pilcrow (¶), as shown in the slide reproduced above, and finally to empty space, or paragraph indent ( ) — all things I’ve talked about here, and in the Shady Characters book, at some length.

But bear with me for a moment.

To recap, the pilcrow had lived happily in the pages of handwritten texts for centuries, added to manuscripts by specialist “rubricators” (scribes who added decorative flourishes in contrasting ink), when movable type barged its way onto the scene in the middle of the fifteenth century. This put the squeeze on the pilcrow from three distinct directions. First, typefounders never quite believed that pilcrows needed to be cast in type: when there is an entire class of worker (and there has been for centuries) whose only job is to add pilcrows, decorative capitals and their kin to the written or printed page, why bother designing, cutting, and casting those same characters and devices? Thus, the pilcrow was far less prevalent in early fonts than, say, the full stop or the semicolon. Second, the sheer volume of printed texts, peppered with double-slash (//) placeholders* indicating where pilcrows were to be added by hand, far outstripped the ability of rubricators to fill in those waiting gaps — pilcrows could not always be printed for lack of type, and they could not always be rubricated for lack of time. This all led to the third nail in the pilcrow’s coffin: as readers got used to navigating the page by means of the gaps left by missing pilcrows, the mark itself became less and less relevant. The pilcrow, as we have heard time and again, was killed by the arrival of printing.

The question, now, is this: is the paragraph itself destined to die just as the mark that once delineated it has disappeared from sight?

That’s the thesis behind Andy Bodle’s recent article at The Guardian, entitled “Breaking point: is the writing on the wall for the paragraph?” Ignoring, for a moment, Betteridge’s law of headlines (“any headline that ends in a question mark can be answered by the word no”), the main thrust of Bodle’s argument is that the arrival of the Internet — like movable type, a revolutionary form of information technology — is exerting a “downward pressure on paragraph length”:

Most online gurus caution against blogposts of more than 600 words; some insist that the ideal length is 200. Internet users, they cry, can’t be bothered to scroll through long articles. Last year, the bosses at Associated Press circulated a memo stipulating that stories be between 300 and 500 words long (exceptions can be made — up to a whopping 700 words — for events of global importance). At the same time, the UK government’s website, gov.uk, promised that it would never publish a sentence exceeding 25 words. […] Reading on a laptop screen or phone is slower and more fatiguing, and it’s harder to keep your place; inserting regular, clear breaks (complete lines rather than indentations) is one way to create a smoother reading experience.

Bodle also makes the point that newspapers, where narrow columns reward frequent indentation, and news media in general, where an objective viewpoint is prized above all, discourage long, discursive paragraphs. Certainly, having picked the current headline story from Reuters (“Euro zone summit aims to keep Greece in single currency”), I count only twenty-five sentences in twenty-four paragraphs (did the writer forget to press “return” after one particular sentence, I wonder?). And this is not, to my eye, an isolated incident. The BBC, writing on the same subject, is little better, and it is just one of the many news outlets that seem to have adopted an atomic approach to paragraphing.

Technological advances, then, carry both opportunities and dangers for the written word. Printing edged out the pilcrow; the typewriter did a number on the em and en dashes and many other uncommon marks; the Internet, in turn, is doing its damndest to kill the paragraph. I’ve lost count of the number of blogs and other websites that treat sentences and paragraphs as interchangeable units of sense and whose staccato delivery is often accompanied by the demarcation of paragraphs by blank lines rather than paragraph indents.

And yet, as disconcerting as I find these changes, I’m doing my best to stay as open-minded as I can about them. Every component of the way we communicate via the written word — our letters, the marks and spaces between them, their arrangement on the page — is and has always been subject to change under pressure of convention, technology, and fashion. Maybe the pulverisation of the paragraph is nothing more than a stylistic tic adopted by writers who need to broaden their reading horizons beyond news websites, or perhaps it’s a deeper trend, an inexorable product of the tiny screens on which we communicate with the world. I honestly don’t know, but I’ll be intrigued to see how the paragraph weathers this next great shift in writing technology.

What are your experiences? Have any writers among you come under any pressure to atomise your paragraphs, or to otherwise restructure your writing for the web?

The “//” mark was a holdover from a twelfth-century system of punctuation that included only horizontal and sloping dashes. See Shady Characters, the book, for more details. 
Thanks again to Claire M. L. Bourne, whose article, “Dramatic Pilcrows”, helped me distil the factors behind the demise of the pilcrow into a coherent story. 

43 comments on “The imminent death of the paragraph

  1. Comment posted by Robert Seddon on

    Internet users, they cry, can’t be bothered to scroll through long articles.

    If sites like the BBC’s tried to minimise needless scrolling, having each sentence float alone amidst the whitespace is precisely what they wouldn’t do.

  2. Comment posted by Dave on

    …and news is just one type of writing. The novel seems to be in a healthy state (whether in print or on e-readers), and I’ve not noticed any push to “de-paragraph-ize” that!

    1. Comment posted by Keith Houston on

      Hi Dave — very true. As far as I can tell, the paragraph fares worst in news stories and blogs. I can just about understand the pressures to “de-paragraph” news stories, but I have no idea what’s going on in the mind of the average paragraph-averse tech blogger.

  3. Comment posted by Bill M on

    I first learned of block style for a paragraph in my first typing class. Later in radio I found it much easier to read on-air copy that was written in that style. However, that is a special kind of printing for a specific purpose. Later in my news writing classes that destroyed all of my previously taught proper writing, like now I had to write in transitive passive backwards style and at no more than a 6th grader’s reading level. I guess today internet writing may be aimed at no more than Kindergarden level readers. Afterall schools in the USA are too lazy to teach proper handwriting too.

    1. Comment posted by Keith Houston on

      Hi Bill – I can see the paragraph-per-sentence style being very useful for reading aloud. A good point!

  4. Comment posted by Jon of Connecticut on

    Sports columnists and other columnists of lighter topics tend to work in one sentence paragraphs; at least the ones that I see in American newspapers. Bill Plaschke of the Los Angeles Times comes to mind.

    Regarding paragraphs. I think they should be used to indicate a change of topic. I rarely indent. But I did come across a pilcrow recently. Your book, Keith, is part of a recent jag I’m on about the English language and this jag has led me to the King James Version of the Bible’ notsomuch for religious reasons, though. I read a book titled Death Sentences by Aussie speechwriter Don Watson. It had its ups and downs. It was mainly complaints about Newspeak but attempted to offer better alternatives in English writing and speaking.

    It encouraged writers to use the cadences of the KJV and Elmore Leonard for more lively language but doesn’t go into further detail. What does he mean? IIRC, the KJV has a lot of iambic pentameter. But I think that Watson is referring to varying sentence length when it comes to cadence. In other words, the rests (periods,) half rests (colons and semicolons,) and quarter rests (commas,) play an important part. But this is something that is rarely, if ever, discussed. He also said something about arranging words in the right order but fails to elaborate.

    Anyways, my copy of the KJV has pilcrows!

    1. Comment posted by Keith Houston on

      Hi Jon — I take your point about lighter topics suiting shorter paragraphs. Maybe I’m wrong (and if I am, boy, am I going to sound elitist), but it seems to me that the first place I noticed shorter paragraphs was in tabloid newspapers like the Sun and the Daily Record here in Britain, pre-dating the current online fad for shorter paragraphs by some years. Is it lighter material that informs the choice of paragraph length, I wonder, or is it driven by the likely reader as perceived by the writer — their reading ability and/or the time they have available to read? (Elitist, as I said. Apologies!)

  5. Comment posted by Jon of Connecticut on

    And your software appears to indent for me.

    1. Comment posted by Keith Houston on

      It does! I wanted comments to be rendered in as similar a way to the articles as was possible.

  6. Comment posted by Korhomme on

    Punctuation, including paragraphs, is about readability, not just aloud but also silently; and readability is about understanding.

    An emphasis on readability gives us textbooks and articles in two (narrow) columns where previously they had a single wide one. It can be hard to find the next line if the column is too wide—some kids had to use a ruler to mark the line they were on.

    Similarly with paragraphs; long paragraphs make for difficult reading, and make comprehension harder. You have to try to remember at the end of the paragraph what was at the beginning to be able to put it all together.

    No surprise that paragraphs are getting shorter; one sentence as a paragraph is just the extreme end of this trend. A paragraph is a ‘thought’, and if the Victorians went on and on with their thoughts, today it’s more ‘show’ than ‘tell’, rather as novelists are advised.

    And then there is the typographical dilemma; serifs or sans-serif? A well designed book or article should make for ease of reading and comprehension, almost something that you don’t actively notice. Look at books for kids who are just learning how to read for examples of how to make things easy.

    1. Comment posted by Keith Houston on

      Hi Korhomme — very true. We do lose some meaning, though, when paragraphs and sentences become conflated; we’re suddenly deprived of an extra level of semantic organisation. I wonder, also, if single-sentence paragraphs are driving sentences to become longer? Some of the sentences/paragraphs in the Reuters and BBC articles I mentioned were quite long, perhaps in order to convey a paragraph’s worth of information. There’s a balance to be struck between sentence and paragraph length, and it feels like we’ve swung too far to one side of it.

      (Ironically, Pause and Effect is a victim of the too-wide text column. Parkes’s paragraphs, on the other hand, are just as lengthy and weighty as you might imagine.)

  7. Comment posted by Jon of Connecticut on

    I used to blog and always used paragraphs in my blog. I was my own editor there. Believe me, no one gave me any feedback about my writing there. I have also contributed several short biographies of baseball players for a project by the Society for American Baseball Research. Some have appeared in books and all are online. It was the one time I worked with editors. They were pro-paragraph.

  8. Comment posted by Stan on

    ‘Most online gurus caution against blogposts of more than 600 words’
    This is a kind of received wisdom, but I’ve seen several studies indicate that long posts are equally or more likely to be shared. I write posts to whatever word count they warrant, except when I’m paid to write to a certain length. In any case, I’m inclined to heed data more than ‘online gurus’!

    1. Comment posted by Keith Houston on

      Hi Stan — have you noticed any correlation between word count and post popularity? It’s something that would be easy enough to work out, I’m sure, but I haven’t tried it myself.

    2. Comment posted by Stan on

      Hi Keith: on my own blog, the most viewed (and commented on) posts tend to be longish ones. Eight of the ten most visited posts have word counts of 1000–2500. Context is of course a factor: Sentence first is about words, books, and language, so people expect to read a bit there. On other kinds of blogs, brevity might work better.

    3. Comment posted by Keith Houston on

      Hi Stan — it looks to be the same for me. Most articles in my top ten have 1,000 words or more, with a couple of outliers of only a few hundred words each. Interesting that we both see the opposite of what the “experts” caution against.

  9. Comment posted by Jon of Connecticut on

    And then there is the typographical dilemma; serifs or sans-serif? A well designed book or article should make for ease of reading and comprehension, almost something that you don’t actively notice. Look at books for kids who are just learning how to read for examples of how to make things easy.

    I find that serifs help in reading and that sans serif fonts are really only useful for things like signage, packaging, or ads where there are only a few words to be read.

  10. Comment posted by Hillel on

    I’ve read (but can’t remember where) that it’s easier to read on screen with line breaks rather than indentation. Maybe this is because on a printed page the page margins act as an additional reference point to measure progress and keep the reader from getting lost in a sea of text? What I find most interesting is the shift in Microsoft Word’s defaults for new documents from serif Times New Roman with auto-indented paragraphs to sans serif Calibri with an auto-space after paragraphs. While it is annoying to read printed documents set this way because the spaces between paragraphs break up the text in ways that disrupt flow, with Word and online tools dictating that format, it may become more and more common in more situations as digital natives get used to seeing and using it primarily.

    1. Comment posted by Keith Houston on

      Hi Hillel — I’d tend to agree with that. I find it easier to keep my place on a web page when the paragraphs are separated by blank lines, though I’m not quite sold on it yet from an aesthetic perspective. I’ve been thinking about refreshing the design here at Shady Characters, and I suspect I’ll stick with indented formatting.

    2. Comment posted by Dave on

      I was watching an online training video for Adobe InDesign only last week, and the woman narrating it practically instructed users to mark paragraph breaks with line space instead of a first-line indent, saying that the latter was old-fashioned. I fumed and seethed!

    3. Comment posted by Keith Houston on

      To each their own, I say, even if their own is objectively wrong¡

    4. Comment posted by Jon of Connecticut on

      I think that a line break makes for a better looking page. Didn’t Tufte say that more white space is better? But I’m not married to the idea.

      Incidentally, I read some David Crystal today and he mentioned that “ampersand” was shortened from “and per se and.” I forget if I saw that in Shady Characters first.

    5. Comment posted by Keith Houston on

      Hi Jon — I covered “and per se and” here back in 2011, though I’m sure David Crystal will have got there first!

      The whitespace issue is a difficult one. As I wrote, I plan to continue to use indented paragraphs when I get round to giving the Shady Characters web site a sprucing-up, but the blank-line approach does seem to have some advantages for web pages, where there are no page breaks to contend with and where the text is often interspersed with images. See this post at Impressive Webs for some discussion on the matter.

    6. Comment posted by Korhomme on

      Some print publishers have a house style that includes putting the first few words of a new chapter in small capitals with an non-indented paragraph. It’s the sort of thing that, unless you look for it, it’s quite easy to miss.

    7. Comment posted by Keith Houston on

      Hi Korhomme — funny you should mention that; I’ve been experimenting with something similar myself to help demarcate successive posts on list pages (like home, category, and search results pages). Of course, individual article titles also help draw the distinction.

  11. Comment posted by Alex on

    I edit content for a web audience and generally try to keep paragraphs to two sentences. Sometimes a single sentence must be more impactful alone. Other times three sentences work together fine.

    For me, there are two factors shortening paragraphs.

    1) Studies show readers often scan the first two sentences, then scroll down to see what’s ahead and how much they’ll need to commit. They then may jump into the middle to read a bit, scroll back up to qualify the bit they just read, then jump to the summary again. Long paragraphs don’t have as many natural entries and exits for this type of browsing.

    2) More mobile devices with smaller screens means that paragraphs that look reasonable on a desktop, become solid walls of text stretching entire screen heights on mobile. Shorter paragraphs read a little choppily on a bigger screen, but are more navigable on a phone on the subway.

    1. Comment posted by Keith Houston on

      Hi Alex — thanks for joining the conversation!

      I don’t know enough about reading patterns and the like to agree or disagree with your first point, but if you’re right then I expect that blank lines would work better than indents.

      Your point about short paragraphs working better on mobile devices is thought-provoking. We already have responsive design and responsive typography, but should we be thinking about responsive paragraphs, too? Some sort of optional, responsive paragraph break could be selectively applied only on devices with small screens so that a carefully-crafted piece of writing is paragraphed appropriately (and differently!) for different devices.

    2. Comment posted by Bathrobe on

      Of course you could have “responsive paragraphs”. For narrow-screen devices you could use:

      p {margin-top: 0em;
      margin-bottom: 2em;}

      For wide-screen devices you could use:

      p {text-indent:1.8em;
      margin-top: 0em;
      margin-bottom: 0em;}

    3. Comment posted by Keith Houston on

      Hi Bathrobe — that’s a neat suggestion. I hadn’t thought about varying the paragraph break style across screen sizes.

      With regard to Alex’s comment, however, I was imagining something more like responsive soft hyphens, so that a writer could add optional paragraph breaks to existing paragraphs. On large screens, an article marked up this way would consist of a small number of long paragraphs; on small screens, it would consist of a large number of small paragraphs. Of course, this would play havoc with the semantic purpose of paragraphing in the first place!

  12. Comment posted by Bathrobe on

    An example of the strange way that news articles include background information, from Barrack Obama has failed Africa (http://www.prospectmagazine.co.uk/world/barack-obama-has-failed-africa).

    “During a speech in Nairobi, Obama recalled his last visit to Kenya in 2006, when the Democratic senator from Illinois was greeted like a rock star.”

    If this were not a newspaper article, you would have no idea who the “Democratic senator from Illinois” was. This whole style defies analysis in terms of ordinary English prose. I suggest that it is also anomalous for discussing paragraphing conventions.

  13. Comment posted by Bathrobe on

    Without an earlier comment that I made, which seems to have been swallowed up, the above comment makes no sense.

    The one-sentence one-paragraph convention in the press is certainly not new. I had already noticed it (and found it quite strange) as early as the 1970s. But I wouldn’t regard this as a model for English prose. Journalese has a number of conventions that are quite different from ordinary English prose, including peculiar practices for dealing with “old” and “new” information. (The previous comment was à-propos of this.)

    Have you looked into the situation in other languages? For instance, is there a similar trend for paragraph length in Continental European languages? German prose wouldn’t seem to lend itself to the same kind of snappiness that is now so prized in English.

    I’ve always been struck by the tolerance of Chinese speakers for large blocks of solid text consisting of a seemingly impenetrable thicket of Chinese characters. Chinese people are not intimidated by this and dive right in, but I find their long paragraphs quite formidable. Chinese prose does not seem to be very concerned about making prose more accessible for the ordinary reader. Aside from long paragraphs, there is little attempt to insert bullet points or numbering (aside from numbering into sections). I believe this also holds for Chinese articles on the Internet.

    1. Comment posted by Keith Houston on

      Hi Bathrobe — I think both your comments have arrived safely.

      As you say, writing for news media is quite different to prose. What I notice, though, is that many bloggers have picked up this same habit of very short paragraphs — the one-sentence paragraph seems to be infectious, regardless of the writing style.

      Alex’s comment about how different reading devices influence paragraph length is an interesting one, however, and perhaps what we’re seeing is design-savvy writers altering their content to fit a new delivery medium. I haven’t looked into the phenomenon in other languages, but I might expect to find that French or German bloggers with a similar background in design might make use the same technique.

      Thanks for weighing in!

  14. Comment posted by Separatrix on

    Any suggestions for a TL;DR mark?

    1. Comment posted by Separatrix on

      … or ‘»’?

  15. Comment posted by Solo Owl on

    1. When they taught me touch typing, they made the point that there should not be more than about sixty characters on a line. There are web pages that routinely have much longer lines. Worse yet, these long lines will not break at the edge of the browser window.

    Long lines are a disservice to astigmatic readers (although your oculist can help.)

    1. Comment posted by Solo Owl on

      I meant to say, “On many web sites, long lines will NOT wrap at the edge of the browser window.” Sorry.

  16. Comment posted by Solo Owl on

    2. I don’t think Unicode has a soft paragraph character, but you have stated a need for it. There are U+2028 Line Separator and U+2029 Paragraph Separator (instead of the control characters?).

    1. Comment posted by Keith Houston on

      Line length is important for readability, as you say. The tricky part is that in prose, line length is at the mercy of the typographer rather than the writer. As such, I’m not sure we need soft paragraph marks so much as more sensitive design.

  17. Comment posted by Solo Owl on

    3. The pilcrow, and its cousins the section mark and the bullet, direct the reader to the beginning of a paragraph. There a few languages that use a special character to indicate the end of a paragraph; Unicode has about 5 of them.

    The pilcrows in the KJV and other bibles correspond, more or less, to what you might call end of paragraph marks in Hebrew study bibles. These mark the places where the reader, it is said, ought to pause and reflect on what they have just read. There are two such marks, one for longer meditations, I suppose.

    1. Comment posted by Keith Houston on

      Interesting! That reminds me of Eric Gill’s use of pilcrows, where a pilcrow at the start of the line marked a new paragraph and a pilcrow between two sentences marked a break somewhere between a sentence and a paragraph.

Leave a comment

Required fields are marked *. Your email address will not be published. If you prefer to contact me privately, please see the Contact page.

Leave a blank line for a new paragraph. You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>. Learn how your com­ment data is pro­cessed.