A Note on Links

I think I might have confused some readers, as I’ve started placing links to articles mentioned in my posts at the bottom of the post, instead of embedding them in per regular Internet practices.

Reason is, perhaps the most compelling part of The Shallows is the section where Carr presents research on what reading hyperlinked text does to your memory and comprehension.

I don’t really buy the idea that hyperlinks force you to make subconscious decisions that then detract from your reading. (I’m sure they do force you to make those decisions, but I’m not convinced that they’re that distracting.) However, I do think there’s something to the idea that people tend to click over to the linked article before finishing with what they’re reading, and that tends to encourage scanning and create a kind of two-tiered form of reading where you’re switching back and forth, which quite clearly detracts from comprehension/memory.

I’m going to keep the Amazon links embedded for now since I think most of you don’t treat them like a typical link to an article. And if you have thoughts on any of this, I’m curious to hear everyone’s experiences of reading linked text vs non-linked.


Per Jeremy Hatch’s suggestion, I’ve been reading Scott Rosenberg’s blog posts about the perils (or not) of embedding links. I think he makes some good points; I am partly persuaded, but not persuaded enough, to start embedding links in the text. For my thoughts on the post Jeremy references (Rosenberg’s Post #1 of 3 on the topic), see my remarks in the comments field. Here’s my dissection of Post #3:

Post #3 completely ignores Carr’s points. It is essentially a defense of the value added links provide (e.g. citation, connectivity, collegiality, etc). Carr never argues that links don’t do this, and, in fact, Carr and Laura Miller and others still link, so their posts do all the good things Rosenberg attributes to links. In Post #3 Rosenberg makes no case whatsoever for the superiority of links embedded in the text above links provided at the end of an article.

Rosenberg also links to a Slate article purporting to place the concept of information overload into historical reference. This article seems to completely misunderstand the concept of information vis a vis postmodernism, and the various stages of capitalism. For instance, Vaughan Bell writes:

Gessner’s anxieties over psychological strain arose when he set about the task of compiling an index of every available book in the 16th century, eventually published as the Bibliotheca universalis. Similar concerns arose in the 18th century, when newspapers became more common. The French statesman Malesherbes railed against the fashion for getting news from the printed page, arguing that it socially isolated readers and detracted from the spiritually uplifting group practice of getting news from the pulpit. A hundred years later, as literacy became essential and schools were widely introduced, the curmudgeons turned against education for being unnatural and a risk to mental health. An 1883 article in the weekly medical journal the Sanitarian argued that schools “exhaust the children’s brains and nervous systems with complex and multiple studies, and ruin their bodies by protracted imprisonment.” Meanwhile, excessive study was considered a leading cause of madness by the medical community.

Yes, all of that looks very quaint now, but that in no way is a rebuttal to the argument that mass production and the rise of virtuality associated with the middle and late stages of capitalism have produced a significant change in the amounts of information we are given and the cognitive requirements for dealing with it all. I don’t want to give the impression that I’m jumping on the worry bandwagon, but when American publishing alone produces roughly 100 times as many books every year as was produced in the whole of Western Europe in one year of the 16th century, such arguments as “yes, people have worried about information overload before” begin to sound absurd. It is quite clear that mass production and other phenomena associated with the rise of capitalism have produced a change in kind, and it is not enough to simply say, “oh, people had the same worries 100 years ago.”

Oddly enough, Bell also cites Socrates’ famous warning against the perils of writing on memory, which most reasonable observers would agree have been borne out.

Articles referenced in this post

Rosenberg Post #1

Rosenberg Post #3

Slate article on info overload in historical context

Recent Posts

Criticism Isn't Free

CR is dedicated to thoughtful, in-depth criticism without regard to what's commercially appealing. It takes tens of hours each month to provide this. Please help make this sort of writing sustainable, either with a subscription or a one-time donation. Thank you!

You could also purchase one of my acclaimed ebooks.


Got Something To Say:

Your email address will not be published. Required fields are marked *


Yes, absolutely the best way to go. Agreed that the links might not be literally mind-alteringly distracting, but hyperlinks are annoying, and if I’m really interested, I’ll click on the link at the bottom rather than just respond to some reflex while reading.

For an alternative view, check out Scott Rosenberg’s take:

FWIW I’m more with Rosenberg on this matter…

Jeremy: That’s an interesting post. A few responses:

* I haven’t removed links altogether, I’ve only placed them at the ends of posts. Rosenberg, who accuses Carr of being misleading, is a little misleading himself when he says that Carr et al. are delinking. That’s not quite true. The links are all still there for Web explorations, they’re just at the end so as not to provide distraction while you read.

* I don’t think Carr is quite as apples-to-oranges as Rosenberg suggests. After all, Rosenberg himself admits that “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.” I’m not sure that anyone knows what the “proper” number of links is, and Rosenberg’s admission indicates some unease over the distraction caused by too many embedded links.

* Rosenberg completely leaves out Carr’s most compelling study, in which two groups were given different articles to read. One was a completely de-linked electronic article. The other was the same electronic article, but with embedded links to other articles meant to supplement the information in that article. (This is the study in Carr’s book that most closely reproduces a Web environment.) It turned out that the latter group, switching between various articles, had far lower overall info retention and comprehension than the group “impoverished” with just one article that it had to read straight through.

Scott, thanks for the thoughtful comments.

The idea that links force an increase in “cognitive load” based on those little decisions is the heart of Carr’s argument in “The Shallows,” so if you don’t buy it, then I actually think you’re closer to me than to him on the main issue.

Since the introduction of tabbed browsing I don’t think the “I’ll stop reading this article while I check out the link” mode is really that common (unless the original article isn’t holding your interest, in which case you’re going to drift away anyway). If it were, then I’d be more worried.

My post #1 was my retort to Carr; you’re right that post #3 doesn’t deal with Carr because it wasn’t meant to — except in the sense that I wanted to tally the positive value of links that I thought Carr largely ignored (or paid lip service to). Post #2 didn’t address Carr much either, but it did talk about what I see as the main source of distraction on today’s Web, which isn’t inline linking but rather over-aggressive behavior on the part of revenue-hungry publishers.

I do believe there’s a difference when you put the links at the end. If one of the values of including the link is to let readers know that they can follow up on your work, why make them wait till they’re done with your piece to see that you’ve actually provided that option? I find that *more* distracting. If you’re writing on the Web and you mention a study and you don’t link to it, I’ll sit there and think, “Wait a minute, what’s the problem here — is the study not online? Why isn’t the writer linking to it?” Putting the links at the end means I don’t get an answer to these questions until I’m done reading your words.

Piling links at the end treats the screen like paper when you just don’t have to. You’re fighting the medium’s natural grain..

Yes, I do believe it’s possible to use “too many links” — just as it’s possible to write a sentence that’s too long, for instance — but I don’t think that has any bearing on whether links are *inherently* hard to read, which is Carr’s point, right?

Finally: I’d loved to know which study you mean. I can’t find one like what you describe on pp. 126-129. (Perhaps it’s elsewhere — do provide a pointer, it would be good for me to look at it more closely.) Carr does cite another study I didn’t bother addressing because it seemed self-evidently irrelevant: This is the one where research subjects reviewed two opposing arguments; one group got to read each of the two arguments straight through, and the other had to hop back and forth. Surprise! The first group learned and retained more. Again, interesting study, but nothing to do with how we read the Web today.


And thank you for the interesting remarks, both here and on your website.

I get what you say about tabs, and that tends to be the way I browse. This is just my own experience, but I notice that I tend to open linked articles in new tabs as you say, and that tends to detract from my attention. I suddenly feel like I have to “get through” whatever I’m reading so I can go on to the tabbed article.

That’s my experience, and I feel like it reduces my concentration. I’m also very much a purist when it comes to reading, and I don’t like anything that takes me out of the experience. I agree that simply seeing a hyperlink probably isn’t going to do that, but hovering over it to see where it goes or opening it up in a tab is, in my opinion, enough to take me out of the article I’m reading.

There’s some validity to what you’re saying about not having the links making the reader wonder. That’s partially why I wrote this post, as I wanted regular readers to know what I was doing so that they would know what to expect. Of course readers who get here via search engines, links, etc, won’t have that knowledge. In my opinion, it comes down to which of those two options poses the greater distraction. I don’t know, and I don’t think any can say they know right now, and I’m hoping readers will let me know how they feel.

The study you mention is exactly the one I’m talking about (first full paragraph on page 128). (And now I see I’ve slightly misrepresented it. Both articles had links, but the researchers distinguished between those who used them and those who didn’t.) I thought this study was the most true to the Web because I think this is how people tend to use the Web: they will read a little bit of one article, then they will click the link and read a little bit of the next article, etc. The study seems to have borne out my natural suspicions that such reading tends to lead to shallower understanding.

One last thing: I’d like to point out that despite my current thoughts on hyperlinks, I have very serious disagreements with Carr’s entire thesis. When his Atlantic article first appeared, I disagreed rather strongly with it:


And just the other day I registered some strong disagreements with The Shallows:


What an interesting exchange I sparked here!

With the first few posts, I found it somewhat distracting to have the links at the end because I’d scroll down to check whether you provided the link — but by now I’ve learned to trust that you have put them there, and now I’m willing to wait. Personally I’ve never consciously felt distracted by embedded links (I do the open-new-tab thing too), but I have to admit that, now that I’m used to them being bundled at the end on your blog, my next reading choice is altogether more mindful. Which is never a bad thing…

It’s interesting… This needs to be tested out with further experience before I really know what I think.

Jeremy: Thanks for letting me know your experiences. I just read a ton of posts on this subject branching off from Rosenberg’s & Carr’s posts, and I feel the matter is very much inconclusive in my mind. Which is to say that hearing what people who are actually reading this site are experiencing with links embedded or not is extremely valuable information.

Right now I’m dithering between keeping them at the end or trying some kind of mixed solution. I don’t know. There are strong arguments on each side.

[…] got into some of this argument in the comments at Scott Esposito’s thoughtful response to my series. Mathew Ingram at GigaOm provided a nice summary of my lengthier musings. I would also […]

[…] Note on Links Another Note on LinksShareSince I declared my intent to put links at the end of posts on this blog, I’ve done some considerable thinking […]


The Surrender is Veronica Scott Esposito’s “collection of facts” concerning how she embraced her true gender.


Two long essays of 10,000 words each on sex in—and out of—literature . . .

The first essay dives in to Nicholson Baker’s “sex trilogy,” explaining just what Baker is up to here and why these books ultimately fail to be as sexy as Baker might wish.

From there the book moves on to the second essay, which explains just why Spaniard Javier Marías does right what Baker does wrong . . .


5 essays. 2 interviews.

All in all, over 25,000 words of Latin American literary goodness.

3 never-before-published essays, including “The Digression”—a 4,000-word piece on the most important digression in César Aira’s career.

Shop though these links = Support this site

Copyright © 2019. Powered by WordPress & Romangie Theme.