Twitter  "Algorithms have consequences." Zeynep Tufekci on Ferguson and net neutrality:  
Nieman Journalism Lab
Pushing to the future of journalism — A project of the Nieman Foundation at Harvard

Maximizing the values of the link: Credibility, readability, connectivity

The humble, ubiquitous link found itself at the center of a firestorm last week, with the spark provided by Nicholas Carr, who wrote about hyperlinks as one element (among many) he thinks contribute to distracted, hurried thinking online. With that in mind, Carr explored the idea of delinkification — removing links from the main body of the text.

The heat that greeted Carr’s proposals struck me (and CJR’s Ryan Chittum) as a disproportionate response. Carr wasn’t suggesting we stop linking, but asking if putting hyperlinks at the end of the text makes that text more readable and makes us less likely to get distracted. But of course the tinder has been around for a while. There’s the furor over iPad news apps without links to the web, which has angered and/or worried some who see the iPad as a new walled garden for content. There’s the continuing discontent with “old media” and their linking habits as newsrooms continue their sometimes technologically and culturally bumpy transition to becoming web-first operations. And then there’s Carr’s provocative thesis, explored in The Atlantic and his new book The Shallows, that the Internet is rewiring our brains to make us better at skimming and multitasking but worse at deep thinking.

I think the recent arguments about the role and presentation of links revolve around three potentially different things: credibility, readability and connectivity. And those arguments get intense when those factors are mistaken for each other or are seen as blurring together. Let’s take them one by one and see if they can be teased apart again.


A bedrock requirement of making a fair argument in any medium is that you summarize the opposing viewpoint accurately. The link provides an ideal way to let readers check how you did, and alerts the person you’re arguing with that you’ve written a response. This is the kind of thing the web allows us to do instantly and obviously better than before; given that, providing links has gone from handy addition to requirement when advancing an argument online. As Mathew Ingram put it in a post critical of Carr, “I think not including links (which a surprising number of web writers still don’t) is in many cases a sign of intellectual cowardice. What it says is that the writer is unprepared to have his or her ideas tested by comparing them to anyone else’s, and is hoping that no one will notice.”

That’s no longer a particularly effective strategy. Witness the recent dustup between NYU media professor Jay Rosen and Gwen Ifill, the host of PBS’s Washington Week. Early last month, Rosen — a longtime critic of clubby political journalism — offered Washington Week as his pick for something the world could do without. Ifill’s response sought to diminish Rosen and his argument by not deigning to mention him by name. This would have been a tacky rhetorical ploy even in print, but online it fails doubly: The reader, already suspicious by Ifill’s anonymizing and belittling a critic, registers the lack of a link and is even less likely to trust her account. (Unfortunately for Ifill, the web self-corrects: Several commenters on her post supplied Rosen’s name, and were sharply critical of her in ways a wiser argument probably wouldn’t have provoked.)


Linking to demonstrate credibility is good practice, and solidly noncontroversial. Thing is, Carr didn’t oppose the basic idea of links. He called them “wonderful conveniences,” but added that “they’re also distractions. Sometimes, they’re big distractions — we click on a link, then another, then another, and pretty soon we’ve forgotten what we’d started out to do or to read. Other times, they’re tiny distractions, little textual gnats buzzing around your head.”

Chittum, for his part, noted that “reading on the web takes more self-discipline than it does offline. How many browser tabs do you have open right now? How many are from links embedded in another piece your were reading and how many of them will you end up closing without reading since you don’t have the time to read Everything On the Internets? The analog parallel would be your New Yorker pile, but even that — no matter how backed up — has an endpoint.”

When I read Chittum’s question about tabs, my eyes flicked guiltily from his post to the top of my browser. (The answer was 11.) Like a lot of people, when I encounter a promising link, I right-click it, open it in a new tab, and read the new material later. I’ve also gotten pretty good at assessing links by their URLs, because not all links are created equal: They can be used for balance, further explanation and edification, but also to show off, logroll and name-drop.

I’ve trained myself to read this way, and think it’s only minimally invasive. But as Carr notes, “even if you don’t click on a link, your eyes notice it, and your frontal cortex has to fire up a bunch of neurons to decide whether to click or not. You may not notice the little extra cognitive load placed on your brain, but it’s there and it matters.” I’m not sure about the matters part, but I’ll concede the point about the extra cognitive load. I read those linked items later because I want to pay attention to the argument being made. If I stopped in the middle for every link, I’d have little chance of following the argument through to its conclusion. Does the fact that I pause in the middle to load up something to read later detract from my ability to follow that argument? I bet it does.

Carr’s experiment was to put the links at the end. (Given that, calling that approach “delinkification” was either unwise or intentionally provocative.) In a comment to Carr’s post, Salon writer Laura Miller (who’s experimented with the endlinks approach), asked a good question: Is opening links in new tabs “really so different from links at the end of the piece? I mean, if you’re reading the main text all the way through, and then moving on to the linked sources through a series of tabs, then it’s not as if you’re retaining the original context of the link.”


Carr was discussing links in terms of readability, but some responses have dealt more with the merits of something else — connectivity. Rosen — who’s described the ethic of the web persuasively as “to connect people and knowledge,” described Carr’s effort as an attempt to “unbuild the web.” And it’s a perceived assault on connectivity that inflames some critics of the iPad. John Batelle recently said the iPad is “a revelation for millions and counting, because, like Steve Case before him, Steve Jobs has managed to render the noise of the world wide web into a pure, easily consumed signal. The problem, of course, is that Case’s AOL, while wildly successful for a while, ultimately failed as a model. Why? Because a better one emerged — one that let consumers of information also be creators of information. And the single most important product of that interaction? The link. It was the link that killed AOL — and gave birth to Google.”

Broadly speaking, this is the same criticism of the iPad offered bracingly by Cory Doctorow: It’s a infantilizing vehicle for consumption, not creation. Which strikes me now as it did then as too simplistic. I create plenty of information, love the iPad, and see no contradiction between the two. I now do things — like read books, watch movies and casually surf the web — with the iPad instead of with my laptop, desktop or smartphone because the iPad provides a better experience for those activities. But that’s not the same as saying the iPad has replaced those devices, or eliminated my ability or desire to create.

When it comes to creating content, no, I don’t use the iPad for anything more complex than a Facebook status update. If I want to create something, I’ll go to my laptop or desktop. But I’m not creating content all the time. (And I don’t find it baffling or tragic that plenty of people don’t want to create it at all.) If I want to consume — to sit back and watch something, or read something — I’ll pick up the iPad. Granted, if I’m using a news app instead of a news website, I won’t find hyperlinks to follow, at least not yet. But that’s a difference between two modes of consumption, not between consumption and creation. And the iPad browser is always an icon away — as I’ve written before, so far the device’s killer app is the browser.

Now that the flames have died down a bit, it might be useful to look at links more calmly. Given the link’s value in establishing credibility, we can dismiss those who advocate true delinkification or choose not to link as an attempt to short-cut arguments. But I think that’s an extreme case. Instead, let’s have a conversation about credibility, readability and connectivity: As long as links are supplied, does presenting them outside of the main text diminish their credibility? Does that presentation increase readability, supporting the ethic of the web by creating better conversations and connections? Is there a slippery slope between enhancing readability and diminishing connectivity? If so, are there trade-offs we should accept, or new presentations that beg to be explored?

Photo by Wendell used under a Creative Commons license.

What to read next
BuzzFeed HQ
Caroline O'Donovan    Aug. 11, 2014
With a new round of investment from Andreessen Horowitz, the viral giant aims to get even bigger. One step down that path: making it more clear to readers what kinds of content BuzzFeed really offers.
  • Jimmy Wales

    Without commenting directly on Carr’s thesis, I just wanted to make one tiny observation that I thought interesting.

    At Wikipedia, the external links – the ones for credibility – are placed at the bottom, just as Carr would suggest.

  • Jay Rosen

    Excellent post, Jason. Your analysis is sound, your categories help, and I am glad you brought in the wider context: fears of a new walled garden, apps with no links, and what the iPad will or will not do to the news business. I think it’s also worth mentioning the deeper background: that to this day much of the news business is reluctant to link out, or just super slow in adapting its production routines to make use of this power.

    I took more than the usual criticism for my suggestion that “unbuilding the web” would be a better name for Carr’s project in the two pieces of his that you point to here. Well, it was a bit hyperbolic of me, and that’s never a good business for a critic to be in. On reflection, what I should have done is pointed out that Carr is asking to be understood in two different ways, but only defending one. When doubters react to the other, he charges them with over-reacting.

    As you pointed here, if you don’t want to be misunderstood as suggesting we can do without links, why would you use “Experiments in delinkification” as your title, when what you really mean is not delinkifying a post, but disembedding the links and moving them around to enhance readability? Wouldn’t a term like that–delinkify–just cause unnecessary confusion?

    Similarly, if what you really mean is… “Steve Gillmor was wrong to stop putting links into his online writing and to go on a crusade against them, because that would be unbuilding the web…” (which, as far as I can tell, is the position Carr actually holds, though I could be wrong) then why would you begin your sketch of that position by declaring a great deal of sympathy for Gillmor’s crusade? As in… “I’m beginning to think I should have joined up instead of mocking it,” which is what Nick wrote.

    As far as I can tell, Carr does not actually think this. He doesn’t want to join up with Gillmor’s crusade to stop linking out in online writing. He just says it that way, because…. well, I won’t speculate on the because part, since I don’t know.

    I’m pretty sure if my actual view was that Steve Gillmor went too far in cleansing his writing of links, I wouldn’t say, “maybe I should have joined his crusade.” That would be confusing. I would say something like, “I still think Gillmor was nuts to do away with links in his own writing, but he got me thinking about something….” That would be clearer. But it would sound less contrarian.

  • Pingback: Credibility, Readability, Connectivity « Reinventing the Newsroom

  • Jason Fry

    I don’t know Nicholas Carr, so this may be unfair, but I do see some carnival barkery at work in the language he chose. But it did get everybody talking, so mission accomplished — and the underlying issue strikes me as a real one that’s valuable to explore.

    And Jimmy’s point gets us right back to staring at the central dilemma: The “credibility” links are at the bottom of Wikipedia articles, and as a daily Wikipedia reader I’ve been trained to know that and read accordingly — I sometimes jump down to the externals when the top of an article has given me the quick gloss I need.

    Yet it’s also true that all the other links in those articles become a huge temptation — you can arrive to find out about, say, British warship classifications and find yourself reading about Phoenician trade with startling speed. (Not a theoretical example — happened to me last week.) Which is fascinating and wonderful — and it’s difficult to imagine Wikipedia without those links — but also gets right to Carr’s point about extra cognitive load. It takes a very disciplined reader to follow an article’s throughline and not race off to explore some other intellectual branch.

    Maybe we are learning to read differently in different contexts, and our kids will scoff at the idea that we struggled with this. Maybe there’s some new way of presenting links that would preserve credibility and connectivity while enhancing readability, and will work for readers who aren’t inclined to mess around under the hood of the browser….

  • Ike

    Why not put this to a test? You have the means to do so.

    Simply change your CSS (Cascading Style Sheet) so the hyperlinks in your stories are invisible, unless they are hovered over.

    Then your readers are not distracted by the existence of links, but can find them if they choose to dig. You retain context, you retain credibility, and you make for a smoother, uninterrupted reading experience.

    Try it for a while, with a few articles, then benchmark that against a control group.

  • Michael Broadhurst

    This is an interesting debate, and I think picks at a more fundamental question the Web still hasn’t really answered in 15+ years of commercial existence. Namely, how much is enough information on a page. The web operations I run inside my company have the same challenge that news organizations do – and as a former journalist and colleague of Jason’s, I know that side of the equation, too – but the constant feedback received at both places was, distilled, the contrary viewpoints of “put more on the home page because it’s too hard to find stuff” and “there’s too much clutter/noise, make the pages simpler”.

    I’m not sure there’s a reconciliation of that, just as I’m not sure there is of this discussion. Links matter, but I think really what Carr is getting at is how many links are too many, and how many are just enough – and maybe that’s the point of this discussion in the end.

    Nice post though.

  • Ali Smith

    This is a really great post, hitting the issues really well.

    I’d just like to say that there is (sort of) precedence for this sort of thing – scientific papers. These have to have many footnotes, providing evidence for the claims asserted. They are often flagged as subscript or superscript numbers, very much like the hyperlink in that it supposedly breaks the flow of reading. In fact these flags are probably worse than hyperlinks because to ‘click-through’ to the original article you have to find the referenced footnote at the end of the article and look it up (online or at a library). This is required in published papers, as (rightly) evidence is needed to back up your statements.

    I find it hard to believe that ‘linking’ in such a way detracts from the reading of the article – scientific papers are the bedrock of all amazing science that has been done for centuries.

  • Pingback: Why link out? Four journalistic purposes of the noble hyperlink » Nieman Journalism Lab

  • Pingback: links for 2010-06-08 « Onlinejournalismtest's Blog

  • Brian Frank

    Wikipedia is an interesting case. I also find the links to other articles very seductive (YouTube is the only other place I find myself forgetting where I started). But I had the same weakness when I used to read physical encyclopedias too — except the distractions tended to be less relevant: many were arbitrary based on whatever pages I happened to flip past. And then I learned how (and when) to exercise discipline…

  • Sara Aase

    It’s not the links that are to blame. Links are simply the electronic form for the same “stuff” that has always vied for our attention when we’re reading something — footnotes (David Foster Wallace, for e.g.) being the obvious one. Or how about when you come across a word you need to look up? Or what about when some word or phrase trips some other association entirely and we put down the book in order to do XYZ?

    It’s not issues of credibility, readability or connectivity in links that are tripping me up, but rather that I associate the entire Internet in general with action. If I’m on it, then I’m supposed to be communicating, researching, writing, or otherwise “working.” (Reading seems to happen better, as you pointed out, in a more closed environment (i.e., book or iPad or Kindle) at a remove from the Internet, where one feels the “permission” to relax.) Links are nothing to me online compared to the other demands for my attention — my email inbox being the biggest.

    I think, as you pointed out, deep reading on the Internet takes self-discipline and training. Sometimes I can do it, sometimes I can’t. I have found that shutting down the browser for blocks of time or walking away from the computer entirely help my ability to concentrate and absorb information.

  • Pingback: This Week in Review: A mobile aggregation dustup, journalists and the link, and fan-based local sports » Nieman Journalism Lab

  • Pingback: Links grow up « Libtools

  • Pingback: The problem with the iPad |

  • Pingback: This Week in Review: A mobile aggregation dustup, journalists and the link, and fan-based local sports | Mark Coddington

  • Pingback: Chain reaction | Wylie's Writing Tips

  • Pingback: Writing Web links | Wylie Communications, Inc.

  • Pingback: In Defense of Links, Part One: Nick Carr, hypertext and delinkification — Scott Rosenberg's Wordyard

  • Pingback: Culture, Anarchy and the Conceptual Value of Links | Brian Frank

  • Pingback: Links: Not Just the Currency of the Web, but the Soul «

  • Pingback: Don’t save your links for the end — it’s more distracting! — Scott Rosenberg's Wordyard

  • Pingback: Attribution and linking in online journalism « Ed Kemp

  • Pingback: MCJ 105 Newspaper Workshop Guest Lecture Resources «

  • Pingback: So what does this mean for the media? « Guy's Media Blog

  • Pingback: Where Papers’ Linking Problems Begin « Reinventing the Newsroom

  • Singarju Krishnachaitanya

    Awesome Posting ! You are so inspirational and you talk sense. That’s important. You’re intelligent and you
    have a lot of heart. I love your posts! Thanks

  • Sherwin Arnott

    Three years later and many “news” organizations aren’t even using links at all.