For 15 years, I’ve been doing most of my writing — aside from my two books — on the Web. When I do switch back to writing an article for print, I find myself feeling stymied. I can’t link!
Links have become an essential part of how I write, and also part of how I read. Given a choice between reading something on paper and reading it online, I much prefer reading online: I can follow up on an article’s links to explore source material, gain a deeper understanding of a complex point, or just look up some term of art with which I’m unfamiliar.
There is, I think, nothing unusual about this today. So I was flummoxed earlier this year when Nicholas Carr started a campaign against the humble link, and found at least partial support from some other estimable writers (among them Laura Miller, Marshall Kirkpatrick, Jason Fry and Ryan Chittum). Carr’s “delinkification” critique is part of a larger argument contained in his book The Shallows. I read the book this summer and plan to write about it more. But for now let’s zero in on Carr’s case against links, on pages 126-129 of his book as well as in his “delinkification” post.
The nub of Carr’s argument is that every link in a text imposes “a little cognitive load” that makes reading less efficient. Each link forces us to ask, “Should I click?” As a result, Carr wrote in the “delinkification” post, “People who read hypertext comprehend and learn less, studies show, than those who read the same material in printed form.”
This appearance of the word “hypertext” is a tipoff to one of the big problems with Carr’s argument: it mixes up two quite different visions of linking.
“Hypertext” is the term invented by Ted Nelson in 1965 to describe text that, unlike traditional linear writing, spreads out in a network of nodes and links. Nelson’s idea hearkened back to Vannevar Bush’s celebrated “As We May Think,” paralleled Douglas Engelbart’s pioneering work on networked knowledge systems, and looked forward to today’s Web.
This original conception of hypertext fathered two lines of descent. One adopted hypertext as a practical tool for organizing and cross-associating information; the other embraced it as an experimental art form, which might transform the essentially linear nature of our reading into a branching game, puzzle or poem, in which the reader collaborates with the author. The pragmatists use links to try to enhance comprehension or add context, to say “here’s where I got this” or “here’s where you can learn more”; the hypertext artists deploy them as part of a larger experiment in expanding (or blowing up) the structure of traditional narrative.
These are fundamentally different endeavors. The pragmatic linkers have thrived in the Web era; the literary linkers have so far largely failed to reach anyone outside the academy. The Web has given us a hypertext world in which links providing useful pointers outnumber links with artistic intent a million to one. If we are going to study the impact of hypertext on our brains and our culture, surely we should look at the reality of the Web, not the dream of the hypertext artists and theorists.
The other big problem with Carr’s case against links lies in that ever-suspect phrase, “studies show.” Any time you hear those words your brain-alarm should sound: What studies? By whom? What do they show? What were they actually studying? How’d they design the study? Who paid for it?
To my surprise, as far as I can tell, not one of the many other writers who weighed in on delinkification earlier this year took the time to do so. I did, and here’s what I found.
You recall Carr’s statement that “people who read hypertext comprehend and learn less, studies show, than those who read the same material in printed form.” Yet the studies he cites show nothing of the sort. Carr’s critique of links employs a bait-and-switch dodge: He sets out to persuade us that Web links — practical, informational links — are brain-sucking attention scourges robbing us of the clarity of print. But he does so by citing a bunch of studies that actually examined the other kind of link, the “hypertext will change how we read” kind. Also, the studies almost completely exclude print.
If you’re still with me, come a little deeper into these linky weeds. In The Shallows, here is how Carr describes the study that is the linchpin of his argument:
In a 2001 study, two Canadian scholars asked seventy people to read “The Demon Lover,” a short story by the modernist writer Elizabeth Bowen. One group read the story in a traditional linear-text format; a second group read a version with links, as you’d find on a Web page. The hypertext readers took longer to read the story ,yet in subsequent interviews they also reported more confusion and uncertainty about what they had read. Three-quarters of them said that they had difficulty following the text, while only one in ten of the linear-text readers reported such problems. One hypertext reader complained, “The story was very jumpy…”
Sounds reasonable. Then you look at the study, and realize how misleadingly Carr has summarized it — and how little it actually proves.
The researchers Carr cites divided a group of readers into two groups. Both were provided with the text of Bowen’s story split into paragraph-sized chunks on a computer screen. (There’s no paper, no print, anywhere.) For the first group, each chunk concluded with a single link reading “next” that took them to the next paragraph. For the other group, the researchers took each of Bowen’s paragraphs and embedded three different links in each section — which seemed to branch in some meaningful way but actually all led the reader on to the same next paragraph. (The researchers didn’t provide readers with a “back” button, so they had no opportunity to explore the hypertext space — or discover that their links all pointed to the same destination.)
Here’s an illustration from the study:
Bowen’s story was written as reasonably traditional linear fiction, so the idea of rewriting it as literary hypertext is dubious to begin with. But that’s not what the researchers did. They didn’t turn the story into a genuine literary hypertext fiction, a maze of story chunks that demands you assemble your own meaning. Nor did they transform it into something resembling a piece of contemporary Web writing, with an occasional link thrown in to provide context or offer depth.
No, what the researchers did was to muck up a perfectly good story with meaningless links. Of course the readers of this version had a rougher time than the control group, who got to read a much more sensibly organized version. All this study proved was something we already knew: that badly executed hypertext can indeed ruin the process of reading. So, of course, can badly executed narrative structure, or grammar, or punctuation.
In both The Shallows and his blog post, Carr also makes reference to a meta-analysis (or “study of studies”) on hypertext reading studies, a paper that examined 40 other studies and concluded that “the increased demands of decision-making and visual processing in hypertext impaired reading performance.” But a closer look at this paper reveals another apples-and-oranges problem.
Carr is saying that Web links slow down our brains. But none of the studies the meta-analysis compiles looked at Web-style links. They all drew comparisons between linear hypertexts (screens with “next” links, not printed articles) on one side, and on the other, literary-style hypertexts broken up into multiple nodes where “participants had many choices in sequencing their reading.”
Every other study that I’ve looked into in this area shares these same problems; I’ll spare you the detail. These studies may help explain why there’s never been a literary-hypertext bestseller, but they don’t do much to illuminate reading on the Web. Carr talks about links having “propulsive force,” but does anyone really experience them that way today? Maybe in the early days of the Web, when they were newfangled, people felt compelled to click — like primitives suddenly encountering TV and jabbing their fingers at the channel selector, wondering what will magically appear next.
I think we all passed through that phase quickly. If your experience matches mine, then today, your eyes pass over a link. Most often you ignore it. Sometimes, you hover your mouse pointer to see where it goes. Every now and then, you click the link open in a new tab to read when you’re done. And very rarely, you might actually stop what you’re reading and read the linked text. If you do, it’s usually a sign that you’ve lost interest in the original article anyway. Which can happen just as easily in a magazine or newspaper — where, instead of clicking a link, we just turn the page.
Yes, a paragraph larded up with too many links can be distracting. Links, like words, need to be used judiciously. This is a long post and I have included only a modest number of links — all that I needed to point you to my sources and references, and most of which most of you won’t ever click. Overuse of links is usually a sign that the writer does not know how to link, which on the Web means he does not know how to write. But such abuse hardly discredits linking itself. Many writers still don’t understand that comma-splicing is bad grammar, but does that get us talking about the “de-comma-fication” of our prose?
For Carr and his sympathizers, links impede understanding; I believe that they deepen it. Back in 1997 Steven Johnson (in his book Interface Culture) made the case for links as a tool for synthesis — “a way of drawing connections between things,” a device that creates “threads of association,” a means to bring coherence to our overflowing cornucopia of information. The Web’s links don’t make it a vast wasteland or a murky shallows; they organize and enrich it.
“Channel surfing,” Johnson wrote, “is all about the thrill of surfaces. Web surfing is about depth, about wanting to know more.” As the Web has grown vast, that desire has grown with it. To swear off links is to abandon curiosity. To be tired of links is to be tired of life.
Tomorrow, in the next post in this series, I’ll examine some of the ways links are being misused on the Web today — driven not by some abstract belief in the virtues of hypertext but rather by crude business imperatives. Then, in the final installment, I’ll make the case for good linking practices as a source of badly needed context and a foundation for trust.
This is part one of a three-part series. The second part is Money changes everything. The third part is In links we trust.
Post Revisions:
- 2 September, 2010 @ 8:57 [Current Revision] by Scott Rosenberg
- 1 September, 2010 @ 9:34 by Scott Rosenberg
- 31 August, 2010 @ 6:46 by Scott Rosenberg
- 30 August, 2010 @ 6:52 by Scott Rosenberg
- 30 August, 2010 @ 6:48 by Scott Rosenberg
- 30 August, 2010 @ 6:22 by Scott Rosenberg
report an error
This is a really good contribution to the debate going on. I’ve been reading “Hamlet’s Blackberry” along similar lines. I think linking per se is not so much the issue as the habit we form, and the need perhaps to choose to exercise a variety of modes of thinking.
For example, when I read a print book now, if it mentions something I’m mildly curious about or don’t recognize, 5 years ago I’d just continue on; now I reach for the ipad or Wikipedia to check it out. It’s both empowering and um, “shallowing”. So I have to be aware, similarly to being aware of the limits of multitasking. Perhaps the younger generation we paint as flitting from tidbit to tidbit will really turn out to be just as good as us (I’m almost 60) once challenged to do so by a sufficiently interesting project or a good enough teacher. It all intertwines.
Looking forward to more on this.
rg
“Yes, a paragraph larded up with too many links can be distracting. Links, like words, need to be used judiciously.”
Bravo! never thought I’d hear you say it. ;-)
Thoughtful article, thanks.
Brilliantly done, Scott! Thank you for the work in deconstructing Carr’s argument.
Scott,
Thanks for your attentive reading of my work (and for your advocacy of the careful, judicious use of links, which I fully support). Despite your vigorous hand-waving here, though, you don’t at all wrestle with the question of whether or not links add to the cognitive load of readers (ie, put more information into readers’ working memory, to the potential detriment of reading comprehension). That was the focus of the studies you discuss (which sought not to critique hypertext fiction but to evaluate the role of hypertext in learning, particularly in comparison to standard linear text). And that was the focus of my examination of the effects of links, which, it’s important to say, addressed links as one element among a number of other common elements of online experience (eg, multimedia, interruptions, multitasking) that also appear to contribute to the overloading of working memory and as a consequence take a toll on memory consolidation, comprehension, and learning.
You give a good picture of how links add to cognitive load when you describe your own experience with them: “If your experience matches mine, then today, your eyes pass over a link. Most often you ignore it. Sometimes, you hover your mouse pointer to see where it goes. Every now and then, you click the link open in a new tab to read when you’re done. And very rarely, you might actually stop what you’re reading and read the linked text.” Links require evaluation and decision-making (even when you decide to ignore a link, you’re making a decision) that is extraneous to the process of reading. As I discuss at length in the book, such extraneous problem-solving adds to your mind’s cognitive load. That’s what the authors you quote mean when they report that “the increased demands of decision-making and visual processing in hypertext impaired reading performance.”
The increase in cognitive load does not negate the many good qualities of links that you describe (and that, in my delinkification post, I’m careful to acknowledge). Indeed it’s the very fact that links often carry valuable information that requires us to give them consideration, at least fleetingly. The point is not that links are bad but rather that they have the effect of adding to the cumulative distractions that hit us as we navigate the web and otherwise engage with digital media. The cognitive effects of these distractions, the evidence indicates, impede comprehension and learning.
I would certainly agree with you that we need more studies of the cognitive effects of links in the context of actual web use. I’m hoping that researchers are pursuing such studies.
Nick
Thanks for the response, Nick.
I agree that “links often carry valuable information,” just as the rest of a text does. So the question is, do links deliver that information in a manner that is qualitatively less efficient (more demanding per piece of information) than the other ways that text delivers information? You’re arguing, I think, that they do. As far as I can tell, you’re saying that the problem is that somehow the information the link delivers is “extraneous,” whereas the rest of the text is not.
This is where I lose you. The link’s information is extraneous only if it is an extraneous link. If it is a well-chosen link, the information is as valuable as any other piece of information in the text. (I can write bad prose in which one sentence has no relation to the next, and that will slow a reader down, too.) Sure, processing the link might take a little more time than reading the same passage without the link. But so would reading a longer sentence.
We make all sorts of decisions all the time when we read texts, linked or unlinked: Do I turn the page? Do I need to look this word up in a dictionary? Do I flip a few pages forward to see how far I have to go to the end of this chapter? Do I look up this footnote or endnote? Do I stop reading because I’m no longer interested? A link is another decision to make, yes — but not an especially hard or time-consuming one.
I’ve read these studies, and I don’t see how what they tested proves that links represent some specially taxing challenge to working memory. They may have set out to “evaluate the role of hypertext in learning,” but sadly, that’s not what they actually designed into their research. Instead, they proved — to return to the one I focused on here — that scattering meaningless faux links within a “text node” slows readers down and interferes with their comprehension. (Which anyone could have told them.)
It is certainly possible for links to be distracting. In my next post I’m going to talk about how business-driven links often do just that. Links are powerful and should be used wisely. But I do not see how they “impair reading performance” simply because they are links. And I think a careful reading of the studies involved will show that they fail to support that conclusion.
The judicious use of relevant links does not, in my opinion, add to the mind’s cognitive load any more than the decision to engage with a footnote or skip it. But sometimes as readers we resent authors who constantly intrude on our concentration by forcing that choice upon us too often. In the past, not everything a print nonfiction writer wrote down was automatically grounds for a footnote. These are choices a writer makes for reasons of clarity, style, tone, accuracy, whatever. Online readers shouldn’t be terrorized by links–but online writers shouldn’t feel duty bound to provide a fistful of links simply because it’s possible.
Scott ,
I think , by referreing to ‘link infested surfing’ as ‘extraneous’ to the process of reading, Nicholas meant the process of evalvuating , making decisions to use/not ot use links everytime you see and NOT the actual content of those links.
What Ramki said.
Scans of the brains of adept readers reveal that the decision-making portions of the prefrontal cortex are calm, allowing the reader to engage deeply in the process of interpreting meaning (which involves different areas of the brain). Making decisions about links, even when we’re not conscious of those decisions, requires the activation of the prefrontal cortex, which breaks the focus required for deep interpretation. That’s what’s meant be extraneous decision making – extraneous to the act of attentive reading.
Having read The Shallows — attentively! — I understand the whole brain scan/prefrontal cortex part of your argument, and expect to present my concerns and criticisms with it when I write at greater length about the book. We can dig into those issues then.
In the meantime, I just don’t see how you can extrapolate from that material to prove that links impede reading retention, comprehension, or understanding. And the studies you use to nail that part of the argument turn out, as I outlined in the post, to address quite different matters than links as we use them today on the Web.
Every word on a web page or a printed page is a link — a link to an idea, a source, or a concept. Some links need more than can fit the page or current flow established by the author. In that sense, every link is a distraction or something that adds to the message. Put it another way, a writer has to make the choice whether the link is necessary or not. Readers may skip the link just as they may skip the word, page, or stop reading. Automated links to definitions, stock quotes, weather, place names, etc., seem to create unnecessary pop-up distractions each time the mouse pointer moves. But otherwise, judicious placement of key links in an article will help the reader despite the distraction. In short, authors have to carefully weigh the “cost” of such distractions to the value of the message.
Very interesting article, Scott (I loved Dreaming in Code, BTW) !
Nick Carr:
“…the question of whether or not links add to the cognitive load of readers (ie, put more information into readers’ working memory, to the potential detriment of reading comprehension). ”
I agree with you Scott, and find the original article by Nick Carr interesting, but wrong. There are always exceptions to the rule, and *sometimes* lots of (misplaced) links in an article can reduce reading comprehension. However, that is not a problem with hyperlinks, but rather a problem with the use of it !
I treat hyperlinks mostly as bolded items; these are important/key items that often INCREASE my reading comprehension (especially when scanning text). As a bonus, I can easily click them to read more details (immediately or later, depending on context) which is a wonderful concept.
Wikipedia is a great example of the power of hyperlinks, and a great “training tool” to get used to hyperlinks; the text is so filled with hyperlinks that you quickly learn to read the text without spending any extra cognitive effort.
Of course, if you’re *looking* for a link containing more information, you’ll need to increase your effort, but it is still much more efficient (for many, including myself) to find what you’re looking for than scanning through the list of references at the bottom, as you can find the links in the context you’re interested in…
When you read a book and want to get more info on the subject you look at the bibliography for resources. If you like the author you look at his biography to see what else he’s written. With technology you can do this faster through links. If you are interested in the subject, links give you research power. Most people click on links to further their knowledge on a subject.
If you’re satisfied with the info you have gotten from an article, you don’t have to follow the link. But, it’s nice to have the option.
Scott, thank you for making this cogent argument.
For me, what is mentally taxing about today’s writing, is deciding if I believe it. So much is inaccurate, that I very often want to find the source. It is important that the links be useful, but the absence of a link is more distracting. The only thing worse than not linking to a source is not citing a source with sufficient precision that it can be Googled. Carr’s argument is emblematic of a style of journalism (taught in many journalism schools) that shortens citations and sourcing to the point that the audience has no compass use for evaluation. Carr’s attempt to move all of the citations to the bottom of his posts would be interesting (but still wrong) if it were grounded in a solid framework which defined when citation is needed.
A well defined link structure provides optional contextual drill down on demand as needed by the user. A proper, organically linked, set of contextual information mimics the brains own neural-net methods for modeling a complex world via indirect, recombinant, linkage of other sub neural-net information models. This facilitate globally efficient and effective knowledge processing for both pragmatic problem solving and artistic synthesis.
You wrote…”Overuse of links is usually a sign that the writer does not know how to link, which on the Web means he does not know how to write.”…Amen and thank you. Hyperlinks are a crutch. A wiki is a beautiful thing but not every page needs to look like one. Part of the problem has to do with search optimization “experts” forcing anchor text links on clients in hopes of appealing to search algos. The idea that a link is useless if it doesn’t contain a keyword in it. Please. This kind of search rank driven thinking and linking results in 250 word press releases about nothing with 249 links in them, also about nothing.
Scott, you’re correct in calling out the “Demon Lover” literary hypertext study as misleading, however I have a hard time believing you’ve discovered this is typical of the research studied in the meta-analysis by Destefano and Lefevre. I haven’t walked through each study in that particular meta-analysis, but I am familiar with a number of studies on hypertext and cognition that use structural hypertext (e.g. document trees) versus linear hypertext (e.g. back – forward).
Carr’s basing much of his conclusions on theory, of course, since no research is as conclusive as journalists would like to believe. But though research on hypertext and cognition is far from complete, it’s also more substantive than what you’re implying here. If you can list how each of the relevant studies in the Destefano meta-analysis focuses on literary hypertext as opposed to structural hypertexts, I’d love to see it.
Hmm. This post is now six months old, so while I do remember reading that paper, I certainly don’t recall all the details now. I’ll have to dig up my notes as I have time and try to retrace how I reached my conclusion…
Yeah, sorry if I’ve stirred up ancient history by web standards. But I’m just now starting to dig into the research on hypertext, reading, and cognitive load myself as background for grad studies. (And I should correct myself: the Destefano paper is a lit review, not a meta-analysis).
So far, my impression is /not/ that most of the experiments used literary hypertexts, but rather comparisons are usually hierarchical (think Table of Contents) or networked (think blogs, Wikipedia) hypertexts. Indeed, the literary examples are so rare (so far) that I wonder why Carr highlighted that particular example, as it is pretty ridiculous, even from a literary point of view (if I understand correctly, all the links led to the same/next page).
Dear Webmaster,
While looking for the potential link partners, I came across your website and thus sending you this requesting mail. I feel that exchanging links (Non-Reciprocal) would be beneficial for both of us in order to get better rankings in SERP’s and traffic as well.
I have all theme back links with good quality sites
If you are interested to exchange links with us then kindly send us the exact location where our link will be placed. If you wish that your link should be added first, kindly send me details along with. We will add your link first.
We hope for your co-operation and support. I assure you that this would be a long term business relationship. I appreciate your attention and wish you great luck for your site.