Rhetoric of Hypermedia, A Style Guide

Elizabeth Chamberlain

Maybe you’re new to writing for the web. Maybe you’ve done some personal blogging, but you’re unsure how that’ll translate into more-public web writing. Or maybe you’re not new to it at all, but you’d like to be more reflective about the way you write online. This rhetorical style guide is intended to provoke, to get you thinking about what Aristotle calls “the available means of persuasion” in the kind of online text that you’ll find on The Inciter: short-form, hyperlinked web text, aimed at a “general public.” It’s the kind of text that may include some embedded multimedia—but really is more about the text than about the videos or pictures that accompany it. The guide below is more thinkpiece than manual, research-based but ultimately not a step-by-step “how to.”

If you’ve ever wondered, “How long should my posts be?” or “How do I write an effective headline?” or “When and how should I hyperlink?”… well, I have, too. And to be honest, you’re probably going to have more questions when you finish reading than you do right now. But they’ll be questions informed by research. I hope they’ll also be questions that incite you to do bold, creative, and rhetorically thoughtful things with your writing, here and elsewhere.

The inspiration and much of the content for this guide came from my dissertation work. That project included computerized analysis of texts on Slate and Newsweek and a bunch of academic journals, and interviews with both academic and non-academic web authors (and some in-between, including Slate’s Rebecca Schuman, an ex-academic who writes about academia). The guide is also based in my experiences as a web copywriter for several businesses and as an Associate Editor of the web-based rhetoric and technology journal Kairos.

Here’s a map of the subsections, including a “Too Long; Didn’t Read” (TL;DR) version of each:

How Long Should My Posts Be?

The ultimate rhetorical answer: As long as they need to be. Really, though, research suggests that the optimal post length depends on your goal: the text-sharing site Medium’s research demonstrates that average viewer time spent per article rises steadily as articles get longer through a 7-minute read (or about 1,900 words), then begins to taper off. The digital marketing firm Snap found that “organic” traffic peaks between 2250 and 2500 words—but social sharing is actually highest for articles over 2500 words, and their analysis suggests that the ideal socially shared web article length will continue to rise. A blog adviser at The Write Practice suggests that posts of under 275 words inspire the most comments, but admits that they’re rarely shared on social media. My data showed that on Slate, the mean post is 639 words, and on Newsweek it’s 690. So if you’re trying to get more reader engagement, you should probably keep things snappy. If you want most of your readers to read the whole thing, 1,900 words appears to be the sweet spot. If you’re hoping for wide distribution on social media, however, you might be better off writing an essay long enough to make a college freshman goan.

But if you write something that long, you’ve got to keep your reader’s attention, which brings me to my next point:

How Can I Attract a Wide Audience?

One of the best ways to attract the attention of an online audience is to connect your ideas to recent, topical news events, Notre Dame English professor John Duffy told me. Duffy is the author of two short web thinkpieces on Inside Higher Ed and The Chronicle of Higher Education that have both achieved an audience outside the field, even though they’re really aimed mostly at writing teachers.

Duffy attributed that success to the way he connected each article to some things that had happened in the news in recent weeks. From his experience, Duffy explained, “If we can connect the things we know to kairotic moments”—a rhetorical term meaning the right and opportune time—“in public discourse, in popular culture, in the media and so forth, we stand a better chance of getting published [in popular press publications].”

Attention to texts, social theorist Michael Warner says, can actually create a public: a public will organize itself around a network of texts when those texts capture its attention. The best way to attract a wide audience, that is, is to write texts that imagine and speak to a wide range of possible readers.

That’s why I argue below for sounding like a friend:

How Should I Sound?

If your goal is to reach a wide audience—especially if you’re an academic like me, used to writing to other academics—make “shorter” your mantra. Break up sentences. Don’t embed them. Short sentences. Short paragraphs. Short words—don’t use five syllables where two will do. Avoid semicolons and jargon and This isn’t a matter of “dumbing it down” for your audience. It’s about making it possible for your audience to include non-experts, people who stumble across your writing in their Google search results.

For The Inciter, which is intended to be a publication about rhetoric with an audience that includes non-rhetoricians, it’s especially important that we work hard to define our terms, never assume prior knowledge, introduce anybody we cite, and avoid getting too “inside baseball” with our discussions of either rhetoric or teaching. But it’s also important to keep our goal of informing and delighting in mind, always. We’ve got to be engaging, clear, and research-oriented—but never condescending. I like to imagine I’m writing an email to a friend from another field with a question about my work.

Think about the kinds of genres that you find most interesting and enjoyable to read, and think about how a research-based version of that genre might look. Some of the authors I talked to as part of my dissertation research—Paul Muhlhauser, Cate Blouke, and Daniel Schafer—presented an article they wrote together as an infographic in addition to the traditional academic text. Daniel Schafer explained to me that they chose an infographic with the goal of reaching an audience that might not “have the relevant training or background or whatever to just jump into an article like that.” But they also recognized that even as academics, they sometimes preferred short-form texts like infographics, because they’re more directed and less time-consuming.

That’s why even this style guide has “TL;DR” answers to each subsection’s question. If you’re looking for models, don’t turn to places where you usually find the research you read. Instead, take a look at what makes your “guilty pleasure” reading so entertaining—for me, that means turning to places like Slate, the Reddit forum “Explain Like I’m Five,” and even (that much-maligned hawker of “clickbait”) BuzzFeed:

What Can I Learn from BuzzFeed? (or, One Weird Trick—Academics HATE It!)

One of the first times I presented my dissertation research at a conference, I had picked the inciting title, “‘One Weird Trick’: Popular Press Strategies to Invigorate (or Infuriate) Online Open-Access Journals.” I thought that suggesting professional writing teachers might learn something about writing from BuzzFeed would be provocative. And sure enough, one person who raised her hand at the end said, “I have to admit: when you first started talking, I was angry. I hate clickbait.” Nobody agrees about how to define “clickbait” (Does the headline have to be misleading, or is simply being vague enough? What counts as “formulaic”? Does it have to use the word “you”? Does it have to be attached to vapid, useless content?). Yet everybody—hilariously, even BuzzFeed—decries “clickbait,” including many of the academics in the room at my conference presentation, who nodded appreciatively when the woman shared her clickbait ire.

“But then,” my would-be critic shrugged, “you started to make a lot of sense.”

I shared, for instance, a graph that charted the length of Slate headlines in words against the social sharing of the post. There was a wide range of lengths, from a bunch of one- and two-word headlines through a handful of eighteen-, nineteen-, and twenty-word headlines. When I fit a trendline to the social sharing data, an interesting dip in the middle appeared (a dip that’s bigger than it looks at first glance, because the y-axis is logarithmic, to make it possible to see all the social sharing data that ranged from 10 shares to over 5 million shares).

Shorter-than-average headlines and longer-than-average headlines seem to do better on social media. The lowest average shares are for nine-word titles, and the highest are for sixteen-word titles. The most successful one-word title: “Escarnope.” A couple of successful super-long titles: “This Is What It’s Like to See Pitch Perfect 2 with an Entire College A Capella Group” (17 words) and “A Photographer Processes Her Parents’ Cancer by Documenting It in These Raw and Touching Images” (15 words).

If there is a “clickbait” formula for these long titles, it might look something like this: Be long and rambling. Make a promise about the emotional experience that the reader will have while reading the article, and don’t be afraid of using adjectives or editorializing (“entire college a capella group,” “raw and touching”). Refer to the article itself (“this is what it’s like,” “these […] images”).

And maybe use a pronoun or two: My data also showed that on both Slate and Newsweek, headlines that included the pronouns “I,” “you,” “they,” or “us” had a higher-than-average rate of social sharing. This resonates with software engineer Max Woolf’s similar social sharing analysis of BuzzFeed titles, which found that the top five most-shared phrases were, “character are you,” “[x] things only,” “before you die,” “is this the,” and “you probably didn’t.”

Do I recommend you clickbait-ize your titles? Not to the extent that you mislead readers. And not to be deliberately, deceptively vague, turning them into a featureless soup of pronouns with no obvious referents.

Yet I think it’s worth recognizing that readers respond to headlines that hint at the emotional experience they’ll have with an article. It’s worth recognizing that without the limits of strict column inches that have led to the development of newspaper headline grammar, online headlines can speak more conversationally to readers. You don’t have to eliminate “and”s and “the”s and “be”s. You can editorialize in the way that journalism guides have railed against—because the assumptions of web writing are different. One of the real cultural triumphs of BuzzFeed, of Vice, of even places like the New York Times blog series is that we increasingly accept that news media has some inherent bias, and we don’t expect that authors of headlines will avoid commenting upon their subjects.

Some academics have been working to achieve widespread acknowledgement that everyone is biased for nearly 30 years. Thanks maybe in part to BuzzFeed (among, of course, many other things), that’s a far less controversial statement today than it was 30 years ago.

BuzzFeed has also been a serious pioneer in the way it standardized a particular kind of headline-plus-image listicle that’s gained enormous traction:

What Kind of Images Should I Include, and When?

There are several key considerations in using images:

  • How to find relevant images
  • Where to use images within a post
  • What kind of licenses are acceptable
  • How to cite images
  • Images are necessary for the layout of The Inciter: Each post is represented in the feed by a single, headlining image. That image will be a kind of metonym–it’ll “stand” for your post, saying something about the content within, representing the research you present. You should think carefully about that image: Do you want to represent your content literally? (e.g. If you’re talking about research on college students’ writing, will your picture be of an adolescent sitting thoughtfully in front of a laptop?) Do you want to represent it with common visual metaphors? (e.g. If you’re talking about research on Australian writing, will your image include a kangaroo and an Australian flag?) Do you want to represent it more abstractly? (e.g. If you’re talking about research on pediatricians’ writing, will your image include healthy-looking kids?) Images from the research itself are great where possible and available—if you can highlight a sharp-looking graph, all the better.

    You may also want to include more images in the post. Consider the BuzzFeed-style listicle, with an image representing each item in the list, or the images in Cracked articles, which often make visual jokes/puns on the content of the article. You don’t need to have more than one image, but subheading images can help keep readers engaged—and the exercise of coming up with a visual metaphor for your content will keep your eye on the “takeaway.”

    A quick note on accessibility: When you include images in your post, be sure to include “alt text,” which is a metadata tag associated with your image that makes it possible for someone with a screenreader to understand the image. As you add an image, WordPress will prompt you to fill in that metadata. (If you forget, you can go back and edit it in.)

    What kind of images to use? Where to find them? Since The Inciter is a non-commercial enterprise, you can use any image that is published in the Creative Commons or in the public domain, without special permission from the creator. Know your Creative Commons licenses, especially if you’re going to be modifying or remixing any of your images: Most importantly, know anything tagged with “BY” has to be attributed to its creator.

    Where to find images? Flickr is a great source of Creative Commons images, and because it’s a site aimed at photographers, it has particularly good tools of indicating which license the creator intends. It’s also far less likely than other image hosts, in my experience, to have images that have been taken without credit/attribution from photographers, shared and rehosted by a stealing stranger. It’s polite but not necessary to ask the creator of a CC-licensed image if you can use their image in a post. If you find a non-CC image on Flickr that you especially like, you can always send a private message to the creator asking for permission to use it; in my experience, about half of the Flickr users I’ve contacted this way have responded, and nobody who has responded has denied the request.

    Google Images also makes it possible to search for CC-licensed content: On a search results page, click “More Tools”… But there’s a higher risk on Google Images that you’ll come across images that have been inappropriately taken from a creator and rehosted with a CC license, without the original creator’s permission or intent. Do not just accept Google Image’s word that an image is CC-licensed; always click through to the page where the image came from. Consider questions like: Is it the creator’s site? If not, can you tell who the creator is? Does the CC license seem credible? Obviously, at some point you have to make a judgment call—and any creator who sees the image and is unhappy with your use of it can ask you to take it down.

    Attribution on The Inciter is simple: When you’re filling in the metadata tag for the image, put the website source (e.g. Flickr? Google Images? Archive.org? Imgur? Photobucket?) and the author’s name (if available) or username. Place a hyperlink to the URL where you found the image over the author’s name.

    This brings me to my next point—hyperlinking practices on The Inciter in general:

    When and How Should I Hyperlink?

    Thus, one of the best ways you can support someone whose work you like is by linking to them. A link to their page with “good keywords” will boost their search rankings, making it more likely that web users (even ones who don’t read your site) will find them. And, of course, it increases their exposure/visibility in a more direct sense, too—your readers may click. In my dissertation research, McDaniel college English professor Dr. Paul Muhlhauser described hyperlinking as “giving honor.”

    If you’re talking about practices you find deplorable, however, you may be better off screenshotting than linking the source you’re citing. That practice gives you some security, too, in knowing that even if the site gets changed/removed, your argument won’t be invalidated. For example, many websites, including The New York Times, made the collective decision never to link to the now-defunct gossip website Gawker—though editor Margaret Sullivan admits in that post that the Times is still figuring out how to standardize its hyperlink practices.

    Be aware of the ways that hyperlinks beckon to readers. They invite distraction in a more meaningful way than parenthetical citations do. Nicholas Carr calls links a “technologically advanced version of a footnote”—but “also, distraction-wise, a more violent form a of a footnote.” In 1998, Michael Miller and L. Jay Wantz said that clicking on a link often leads to “a feeling of being lost in ‘cyberspace.’” Consider how much has changed since 1995, when veterinary medicine professor Dr. Mattie Hendrick had to describe how to use a browser’s back button on her University of Pennsylvania website (“Got it?” she asked readers, and then emphasized again, “It’s important. Read this section again if you don’t get it.”). We’ve come a long way—none of the authors or editors I interviewed were concerned that links would make readers feel lost. Most Internet readers use tabbed browsing, most readers are familiar enough with hyperlinks and web browsing practices (e.g. judicious use of “open in a tab behind” and the “back” button) that they won’t literally get lost by following a hyperlink. But still, readers are more likely to click on a hyperlink than follow a print citation.

    Consider the persuasive force of a hyperlink, even an unclicked hyperlink. Linking a phrase indicates that it’s supported by research. It says, “somebody else’s work says this too.” And thus it makes sense that having more links is associated both with opinion content (on Newsweek, the Opinion section has on average 3 times as many links as the other sections) and with high social sharing. My research found that pieces with 20-80 hyperlinks are, both on Slate and Newsweek, more often shared on Facebook than pieces with few links. Hyperlinked content is researched content, and often persuasive content.

    How long should I make the text I where I place the link—called “anchor text”? My analysis of Slate and Newsweek found that anchor text is, on average, about 3 words and about 20 characters long. Authors usually link more than one word, but usually don’t link more than a few words.

    But you can use this knowledge to rhetorical effect: When pieces have an especially long hyperlink, it’s often over the key idea of the article. If you can link the “thesis” of your article, the hyperlink will call attention to that idea—and the anchor text being unusually long can work in your favor. Hyperlinks essentially let you bold your thesis.

    What if the link goes dead? Whenever possible, provide information in your signalling phrase that will let readers search for other versions of the source you provide. Although standard web hyperlink citation practices don’t involve including authors’ names or titles of articles cited, I encourage you to do so where possible. The more information you can give a reader to search for archived/alternative versions of content, the better. This is especially true for the most-key pieces of your research. Some links may be less important—“non-essential” or “throwaway” links to tools or extraneous resources—and including names of creators or titles of pieces for such links may be cumbersome or prohibitively time consuming. Don’t avoid linking just because it’d take too much time to explain the source; it’s okay to link without an author’s name. If you discover a link has gone dead while you’re working on a piece, check web archival resources such as Archive.org’s WayBack Machine or Screenshots.com.

    Making it possible to find the source behind a dead link is one good reason to make your anchor text semantic. I’m not one for strict rules, but here’s one: Never put a link over just the phrase “click here” or “click here for more information” (except in a sentence like this one, where that phrase describes the linked content). “Click here” doesn’t give your reader any sense of what’s coming in the link. It doesn’t make use of the rhetorical potential of the hyperlink. It doesn’t provide link value to the creator (because it’s not a meaningful set of keywords). And you don’t need to tell your readers that a link is clickable—you can assume that, because it’s a link, they know to click.

    It’s worth knowing, too, the anatomy of a hyperlink: That HTTP stands for “Hypertext Transfer Protocol” and tells your browser it’s looking for a web address. That including both “www” and “http” is usually redundant, but it’s become standard practice anyway. That if a URL has a question mark or ampersand, the stuff following is usually part of a “query string” indicating data you’ve put into the website—if your goal is to link to a search page, this stuff may be necessary, but otherwise you should probably delete it. You should know that sometimes your view of a page is dependent on being logged in or having accepted a site’s cookies. It’s a good idea to test a link in a cookie-free browser window (e.g. Chrome’s “Incognito” mode), to make sure it will look the way you want when readers click.

    The upshot? Link to support your argument. Link to give credit. Link to add value to the site you’re linking. Link thoughtfully, judiciously—aware of the ways that links can distract readers but also the ways they can strengthen your writing’s rhetorical force.

    A TL;DR for This Whole Style Guide:

    Familiarize yourself with standards and conventions of web writing, where there are standards (e.g. using images as subheading, having 3-word anchor text). You shouldn’t view those conventions as strict rules—many are still in development, and even where they’re not, they’re usually flexible. But break convention intentionally and rhetorically. Be insightful. Be an inciter.