The Rhetoric of the Hyperlink

The hyperlink is the most elemental of the bundle of ideas that we call the Web. If the  bit is the quark of information, the hyperlink is the hydrogen molecule. It shapes the microstructure of information today.  Surprisingly though, it is nearly as mysterious now as it was back in July 1945, when Vannevar Bush first proposed the idea in his Atlantic Monthly article, As We May Think. July 4th will mark the second anniversary of Ribbonfarm (I started on July 4th, 2007), and to celebrate, I am going to tell you everything I’ve learned so far about the hyperlink. That is the lens through which I tend to look at more traditional macro-level blog-introspection topics, such as “how to make money blogging,” and “will blogs replace newspapers?” So with a “Happy Second Birthday, Ribbonfarm!” and a “Happy 64th Birthday, Hyperlink,” let’s go explore the hyperlink.

Image from Wikipedia, free license

Image from Wikipedia, free license

Hyper-Grammar and Hyper-Style

The hyperlink is not a glorified electronic citation-and-library-retrieval mechanism. The “electronic library” perspective, that the hyperlink is merely a convenience that comes with the cost of amplifying distraction, is a myopic one. But lousy though it is, that is our starting point. The first airports were designed to look like railway stations after all. As McLuhan said,  “We see the world through a rear-view mirror. We march backwards into the future.” Turn around, backward march.

This is the default mental model most people have of hyperlinks, a model borrowed from academic citation, and made more informal:

  • Implicit inline: Nick Carr believes that Google is making us stupid.
  • Explicit inline: Nick Carr believes that Google is making us stupid (see this article in the Atlantic).

Both are simple ports of constructs like  “Nick Carr believes [Carr, 2008] that Google is making us stupid.” There are a couple of mildly interesting things to think about here. For instance, the hyper-grammatical question of whether to link the word believes, as I have done, or the title.  Similarly, you can ask whether this or article or this article in the Atlantic should be used as the anchor in the explicit version. There is also the visual-style question of how long the selected anchor phrase should be: the more words you include, the more prominent the invitation to click. But overall, this mental model is self-limiting. If links were only glorified citations, a Strunk-and-White hyper-grammar/hyper-style guide would have little new raw material to talk about.

Let’s get more sophisticated and look at how hyperlinks break out of the glorified-citation mould. Turn around, forward march.

Hyperlinking as Form-Content Mixing

Here are two sentences that execute similar intentions:

I don’t remember where I first saw this clever method of linking (the second one), but I was instantly fascinated, and I use it when I can. This method is a new kind of grammar. You are mixing form and content, and blending figure and ground into a fun “open the secret package” game. From a traffic-grubbing perspective, the first method will leak less, because if you already know all about apples, the link tells you what it is about, and you won’t click. So if you want to reference a really unusual take on apples, the second method is more effective.

Real hyperlink artists know that paradoxically, the more people are tempted to click away from your content, the more they want to keep coming back.  There is a set of tradeoffs involving compactness, temptation to click, foreshadowing to eliminate surprise (and retain the reader), and altruism in passing on the reader. But the medium is friendlier to generosity in yielding the stage. This yielding the stage metaphor is important and we will come back to it.

But possibly, this and similar tricks seems trivial to you. Let’s do a more serious example.

Hyperlinking as Blending of Figure, Ground and Voice

A while ago, on the Indian site,, I wrote an article pondering the interesting difficulties faced by non-European-descent writers (like me) in developing an authentic voice in English.  Postmodernists, especially those interested in post-colonial literature, obsess a lot about this sort of thing. Gayatri Spivak, for instance, wrote a famous article, Can the Subaltern Speak? But scarily-impenetrable people like Spivak are primarily interested in themes of oppression and politics. I am primarily interested in the purely technical problem of how to write authentically in English, in cases where my name and identity are necessary elements of the text.  Consider these different non-hyperlinked ways of writing a sentence, within a hypothetical story set in Bollywood.

  • Fishbowl/Slumdog Millionaire method: Amitabh stared grimly from a tattered old Sholay poster.
  • Expository: Amitabh Bachchan, the Bollywood superstar, stared grimly from a tattered old Sholay poster. Sholay, as everybody knew, was the blockbuster that truly established Bachchan.
  • Global contextual: Amitabh Bachchan, the Clint Eastwood of India, stared grimly down from a tattered old Sholay poster. Sholay, that odd mix of Kurosawa and John Wayne that drove India wild.
  • Salman Rushdie method: Amitabh, he-of-boundless-splendor,  stared down, a-flaming, from a tattered old Sholay poster.

Critics and authors alike agonize endlessly about the politics of these different voices. This particular example, crossing as it does linguistic and cultural boundaries, in the difficult domain of fiction, is extreme. But the same sorts of figure/ground/voice dynamics occur when you write in-culture or non-fiction.

The first simply ignores non-Indian readers, who must look in at Indians constructing meaning within a fishbowl, with no help. It is simple, but unless the intent is genuinely to write only for Indians (which is essentially impossible on the Web, in English), not acknowledging the global context is a significant decision (whether deliberate or unthinking).

The second method is simply technically bad. If you can’t solve the problem of exposition-overload, you shouldn’t be writing fiction.

The third method is the sort of thing that keeps literary scholars up at nights, worrying about themes of oppression. Is acknowledging Clint Eastwood as the prototypical strong-silent action hero a political act that legitimizes the cultural hegemony of the West? What if I’d said Bruce Lee of India or Toshiro Mifune of India? Would those sentences be acts of protest?

Rushdie pioneered the last method, the literary equivalent of theater forms where the actors acknowledge the audience and engage them in artistic ways. Rushdie finesses the problem by adopting neither simplicity nor exposition, but a deliberate, audience-aware self-exoticization-with-a-wink. If you know enough about India, you will recognize “he-of-boundless-splendor” as one literal meaning of the name Amitabh, while Sholay means “flames.” By putting in cryptic (to outsiders) cultural references, Rushdie simultaneously establishes an identity for his voice, and demands of non-Indians that they either work to access constructible meaning, or live with opacity. At the same time, Indians are forced to look at the familiar within a disconcerting context.

But Rushdie’s solution is far from perfect. In Midnight’s Children, for instance, he translates chand-ka-tukda, an affectionate phrase for a child in Hindi, literally as piece-of-the-moon. A more idiomatically appropriate translation would be something like sweetie-pie. Depending on the connotations of “moon” in non-Indian languages, the constructed meaning could be anywhere from weird to random. That gets you into the whole business of talking about languages, local and global metaphors, translation, and the Sapir-Whorf hypothesis. Fine if that rich tapestry of crap is what you want to write about. Not so good if you actually just want to write a story about a pampered child.

Here is a solution that was simply not available to writers in the past:

(a version of this solution, curiously, has been available to comic-book artists. If the sentence above had been the caption of a panel showing a boy staring at an Amitabh Bachchan Sholay poster, you would have achieved nearly the same effect).

This is an extraordinarily complex construct, because the sentence is a magical, shape-shifting monster. It blends figure and ground compactly; the gestalt has leaky boundaries limited only by your willingness to click. Note that you can kill the magic by making the links open in new windows (which reduces the experience to glorified citation, since you are insistently hogging the stage and forcing context to stay in the frame). What makes this magical is that you might never finish reading the story (or this article) at all. You might go down a bunny trail of exploring the culture and history of Bollywood. Traditionally, writers have understood that meaning is constructed by the reader, with the text (which includes the author’s projected identity) as the stimulus. But this construction has historically been a pretty passive act. By writing the sentence this way, I am making you an extraordinarily active meaning-constructor. In fact, you will construct your own text through your click-trail. Both reading and writing are always political and ideological acts, but here I’ve passed on a lot more of the burden of constructing political and ideological meaning onto you.

The reason this scares some people is rather Freudian: when an author hyperlinks, s/he instantly transforms the author-reader relationship from parent-child to adult-adult. You must decide how to read. Your mom does not live on the Web.

That’s not all. The writer, as I said, has always been part of the constructed meaning, but his/her role has expanded. Literary theorists have speculated that bloggers write themselves into existence by constructing their online persona/personas. The back-cover author biography in traditional writing was a limited, unified and carefully managed persona, usually designed for marketing rather than as a consciously-engineered part of the text. Online however, you can click on my name, and explore how I present myself on LinkedIn, Facebook and Twitter. How deeply you explore me, and which aspects you choose to explore, will change how you construct meaning from what I write.

So, in our three examples, we’ve gone from backward-looking, to clever, to seriously consequential. But you ain’t seen nothing yet. Let’s talk about how the hyperlink systematically dismantles and reconstructs our understanding of the idea of a text.

Fractured-Ludic Reading

The Kindle is a curiously anachronistic device. Bezos’ desire was to recreate the ludic  reading experience of physical books. To be ludic, a reading experience must be smooth and immersive to the point where the device “vanishes” and you lose yourself in the world created by the text. It is the experience big old-school readers love. Amazon attempted to make the physical device vanish, which is relatively unproblematic as a goal. But they also attempted to sharply curtail the possibilities of browsing and following links.

In light of what we’ve said about constructing your own text,  through your click-trail, and your meaning from that text, it is clear that Bezos’ notion of “ludic” is not a harmless cognitive-psychology idea. It is a political and aesthetic idea, and effectively constitutes an attitude towards that  element we know as dissonance. It represents classicism in reading.

Writers (of both fiction and non-fiction) have been curiously lagging when it comes to exploring dissonance in their art. Musicians have gone from regular old dissonance through Philip Glass and Nirvana to today’s experimental musicians who record, mix and play back random street noises as performance. Visual art has always embraced collages and more extreme forms of non sequitur juxtaposition. Dance (Stravinsky), film (David Lynch) and theater (Beckett) too, have evolved towards extreme dissonance. Writers though, have been largely unsuccessful in pushing things as far. The amount of dissonance a single writer can create seems to be limited by a very tight ceiling that beyond which lies incomprehensible nonsense (Beckett’s character Lucky, in Waiting for Godot, beautifully demonstrates the transition to nonsense).

In short, we do not expect musical or visual arts to be unfragmented or “smooth” or allow us to forget context. We can tolerate extreme closeness to random noise in other media. Most art does not demand that our experience of it be “ludic” the way writing does. Our experience can be disconnected, arm’s-length and self-conscious, and still constitute a legitimate reading. Word-art though, has somehow been trapped within its own boundaries, defined by a limited idea of comprehensibility and an aesthetic of intimacy and smooth flow.

There are two reasons for this. First, sounds and images are natural, and since our brains can process purely unscripted stuff of natural origin, there is always an inescapable broader sensory context within which work must be situated. The color of the wall matters to the painting in a way that the chair does not matter to the reading of a book.  Words are unnatural things, and have always lived in isolated, bound bundles within which they create their own natural logic. The second reason: music and visual art can be more easily created collaboratively and rely on the diversity of minds to achieve greater levels of dissonance (an actor and director for example, both contribute to the experience of the movie). Writing has historically been a lonely act since the invention of Gutenberg’s press. We are now returning to a world of writing that is collaborative, the way it was before Gutenberg.

So what does this mean for how you understand click-happy online reading? You have two choices:

  • You could think of click-happy Web browsing as “non-ludic” cognition behavior that is destroying the culture of reading. This is the view adopted by those who bemoan “continuous partial attention.” This is Nick Carr in Is Google Making us Stupid?
  • Or you could think of browsing as a new kind of ludic: an  unsettling, fragmented experience that is still comprehensible in the sense that a David Lynch movie is comprehensible. It is a kind of ludic that can never be created within one brain.  Click trails are texts whose coherence derives from your mind, but whose elements derive from multiple other minds.

In other words, when you browse and skim, you aren’t distracted and unfocused. You are just reading a very dissonant book you just made up. Actually, you are reading a part of a single book. The single book. The one John Donne talked about. I first quoted this in my post The Deeper Meaning of Kindle. The second part is well-known, but it is the first part that interests us.

“ All mankind is of one author, and is one volume; when one man dies, one chapter is not torn out of the book, but translated into a better language; and every chapter must be so translated…As therefore the bell that rings to a sermon, calls not upon the preacher only, but upon the congregation to come: so this bell calls us all: but how much more me, who am brought so near the door by this sickness….No man is an island, entire of itself…any man’s death diminishes me, because I am involved in mankind; and therefore never send to know for whom the bell tolls; it tolls for thee.”

The Hyperlink as the Medium

If you start with McLuhan, as most people do, there are two ways to view the Web: as a vast meta-medium, or as a regular McLuhanesque medium, with nothing meta about it. For a long-time I adopted the meta-medium view (after all, the Web can play host to every other form: text, images, video and audio), but I am convinced now that the other view is equally legitimate, and perhaps more important. The Web is a regular medium whose language is the hyperlink. The varieties of hyperlinking constitute the vocabulary of the Web. If I give you an isolated URL to type into your browser, for a stand-alone web page with a video or a piece of text, you are not really on the Web. If there is no clickable hyperlink involved, you are just using the browser as a novel reading device.

But though hyperlinking can weave through any sort of content, it has a special relation to the written word. The Gutenberg era was one where writing was largely an individual act. Before movable type, the epics and religious texts of the ancient world were harmonies of multiple voices. But what we have today is not a resurrected age of epics. The multi-voiced nature of today’s hyper-writing is different. The difference lies in the fact that the entire world of human experience has been textualized online. Remember what I said about walls and paintings? The color of the wall matters to the painting in a way that the chair does not matter to the reading of a book. In the Web of hyperlinks, writing has found its wall.

This is a pretty experimental theme for me, and though it interests me a lot, I am not sure if it interests regular ribbonfarm readers. So if you’d like me to explore this more, post a comment. I have notes on stuff like the basic nature of blogs and wikis viewed as McLuhan media, what the death-of-newspapers debate looks like from the hyperlink perspective, and so forth.

About Venkatesh Rao

Venkat is the founder and editor-in-chief of ribbonfarm. Follow him on Twitter


  1. This post fascinated and delighted me several times over as as I read through it. Thank you. I’d love to see more such, and most definitely those that you referred to at the end.


  2. saurabh says:

    Good post. I dont think you talked about the – in my opinion – least objectionable way of doing the “Amitabh” sentence, which is with footnotes. Another way, which really goes all the way, is to have a text page and an expl page side by side, which I’ve seen for Shakespeare etc.

  3. This is a fantastic post, Venkat – I’m only just starting to explore McLuhan’s body of work, and I’d love to see more posts from you along similar lines.

    Last month, I spent a couple of weeks transcribing one of my undergraduate dissertations to a Tumblr weblog, and had to make some interesting decisions about how best to deal with referencing and exposition. I don’t think I necessarily made the best use of the medium, and struggled with the granular nature of the weblog-essay, but it certainly raised some interesting questions.

    Also, the “Amitabh” example is one of the most elegant explanations I’ve seen in quite some time, and I’m really interested in saurabh’s comment on the parallels between hyperlinks and footnotes.

  4. Saurabh and Justin:

    I am not sure what I feel about footnotes… I don’t think I agree that they are the least problematic, since they lead to a very definite kind of voice (an academic omniscient) which might be too constraining, though I can see a first-person narrator character doing it if he/she has the right personality. Like a neurotic professor voice.

    David Foster Wallace is the only major writer I know of who made a “literary” use of footnotes. In pop fiction, J. T. Edson’s Westerns are full of footnote references to his own imaginary universe, giving his work a very different (faux historical) feel from say, Louis L’Amour. The Shakespeare analogy is probably not valid because those sorts of texts have 2 voices (an editor/teacher and the original). It would take really high levels of skill to integrate footnotes into a fictional narrative in broad ways (sci-fi writers who begin chapters with excerpts from imaginary histories/encyclopedias come closest, as in Dune for instance, or Hitchhiker’s Guide).

    But online… though Wikipedia uses footnotes, I am not sure they are a useful construct. Hyperlink overloads, hover comments etc. are just beginning to explore alternatives.

    I’ll have to think more.

    • For literary footnotes, I strongly recommend hunting down Mark Z. Danielewski’s House of Leaves – a fantastic book, with a textual structure that’d really resonate with what you’ve been trying to get at in this post.

    • saurabh says:

      Isnt a hyperlink just a “ref” that may direct you to another book instead of the foot of the page or the end of the book? Ok, one is external and dynamic, the other isnt, but the thing that’s common is that you dont have to read it if you do know; and there’s an explanation available if you don’t.

      In fact, after footnotes, my personal favourite way, even if I dont know Amitabh or Sholay, is your first method. If your matter is interesting, I’ll read more myself. Your job is not to teach me, unless it is. My favourite Three Musketeers translation had ‘bourg of Meung … Huguenots made a second Rochelle…’ with no notes in the famous first line. I didnt get it, yet I got it: it was wonderful and I was hooked. ‘village of Meung … Protestants laid seige …’ is horrible and loses everything.

      Btw I hate the Wikipedia refs too, the internet is not suited for footnotes. Here it becomes an unnecessary redirection – why not just link to the url directly?

      • Isnt a hyperlink just a “ref” that may direct you to another book instead of the foot of the page or the end of the book? Ok, one is external and dynamic, the other isnt, but the thing that’s common is that you dont have to read it if you do know; and there’s an explanation available if you don’t.;

        whoa, as they say, saara ramayan sunne ke bad poochte ho lakshman kaun tha?

        That’s the literal use of footnotes in the ‘glorified footnote’ mode. My point is that the instant availability changes the way you write the original, that’s the premise of the whole article. Being able to click-click-click is a continuous cognitive experience in a way that repeatedly going to a paper library and checking out a citation trail is not.

  5. Hey thanks, House of Leaves certainly seems worthwhile, and looking it up in Wikipedia got me to the idea of ergodic literature.

    Slightly different from what I am after though, which is an interpretation of reading experiences through traversal of hyperlinks, rather than texts deliberately structured that way by 1 person. But I think there would be valuable insights there. On my list now.

  6. A truly enjoyable post with many quotable passages. I particularly enjoyed this one: “Click trails are texts whose coherence derives from your mind, but whose elements derive from multiple other minds.”

    My browser and I will now go on to read your text on ‘The Deeper Meaning of Kindle’.

  7. Beautiful post (came here via “Schockwellenreiter”)!

    Has inspired me to now start writing an article in my own sphere of interest (European politics).

  8. This is awesome man… Its worth reading for every blogger & Interface developer

  9. Some pebbles scattered into your pond:

    While reading the plain text with which you explain the link examples, I several times right-click’ed and googled, or looked up in a dictionary. The author does not fully control the hyperlinkage.

    Dorothy Sayers, perhaps not a “major” writer, but no minor one either, used single-voiced footnotes heavily in her first Lord Peter Whimsey detective novel, called “Whose Body?” I love the book and series, but the footnoting always strikes me as way way way over the top. And perhaps others as well, as she abandoned the practice immediately.

    McLuhan wrote *about* the limitations of linear text in conveying realistic, referential experience (say, in The Medium Is The Massage), but he wrote in a nonlinear style that largely defied his very point. I’ve never been able to decide whether he thereby proved or disproved his assertions. The technique has been copied only rarely, but one example is the Science Fiction writer John Brunner (Stand on Zanzibar, The Sheep Look Up, Shockwave Rider). Work on this essay a bit, any you might become the self-contradicting prophet of the hyperlink!

  10. Thanks Basti, Julien, Markanday and Jack.

    Interesting to see that many people seem to be thinking about such things.

    Jack: I did think that Understanding Media did follow its own theories moderately, but you are right. With a lot of folks analyzing the properties of a new medium, it is a case of “do as I say, not as I do.” Their own ability/lack thereof to execute shouldn’t be taken as proof/disproof of anything. It is one thing to appreciate a point/insight theoretically and objectively, and quite another to apply it in practice. I think it takes a generation or two for “native” practitioners of a medium to emerge. Until then, everybody is a bumbling explorer.

  11. David Locke says:

    Links in hypertext theory (Bush) and the links used in HTML have little to do with each other. In hypertext theory, links are associative like a poet’s use of a word. As such the link could be dynamic and deal with the semantics and connotations of the word. As it is in HTML, we are stuck with static, hardwired, authoritorial links. Even with XSLT, we have a long way to go before we achieve the Bush ideal.

    I would love it if people stopped referring to Bush and HTML in the same breath.

    • David:

      There is room enough in this world for the exploration and updating of both the letter and spirit of Bush’s original ideas. To claim that “HTML links and hypertext “theory” (Bush) have little to do with each other” is misguided on two levels.

      First, Bush’s original article is not about either theory or practice. It is visioning pure and simple, which inspired Doug Licklider and Robert Taylor to drive forward the agenda of the original Internet architecture, and then Tim Berners-Lee with the www, and most recently, the idea of the wiki. That speaks to the breadth and scope of the vision.

      Second, to the extent that there IS a scholarly hypertext “theory,” (both technical and literary-critical), that derives from Bush, it has no claim to being the one true faith. Both the theory and practice are important lines of investigation, and to stick to narrow technical readings of the original vision is rather limiting.

      So long story short: I think Bush deservedly gets quoted in broader discourses. The theoretical disciplines he spawned aren’t the exclusive inheritors of his legacy.

      But there is also a force beyond Bush, powerful though his thinking was. This is the Web user’s behavior, which has driven the real, practical Web to evolve in ways he never thought of. Yes, semantic, dynamic and bidirectional linking are all valid subjects to explore. That shouldn’t blind us to the possibilities opened up by the evolution of hyperlinking as it is actually practiced, and the improvisations and creativity people are bringing to its use. To dismiss it as “static, hardwired and authoritorial” is to deny its profound impact. The hyperlink isn’t a broken, incomplete implementation of Bush’s idea. It is a fertile concept that is evolving in open-ended ways.

  12. anupama says:

    wonderful post, venkat…

    small suggestion – when you refer to parent-child interaction changing to adult-adult state, the reference is Bernian (from Eric Berne) rather than Freud. Freud talked about sex and aggression, Erickson about thye stages of development of identity, but it was Eric Berne who spoke of the Parent, Adult and Child Ego states (check out


  13. Hi Anupama:

    Yes, Berne did come up with that terminology in the transactional model, but since few people get the reference these days, and PAC is essentially the same as superego-ego-id at zeroth order, I usually just cite Freud unless I am diving deeper. I have a couple of pieces on games and scripts pipelined though :)


  14. anupama says:

    looking forward to your pieces on TA – thanks for your e-mail

  15. Hyperlinked text could also be seen as a minefield–in two senses.

    -A benevolent guru expounding on a useful theme, laying traps to swat away the undisciplined to fritter away their attention on other primitive topic paths, and,

    -A tentative blogger, eager to have sufficient references in the post, but afraid of making them too attractive and losing eyeballs too soon.

    Footnotes are like the strict meeting chairperson, who flags every possible need for additional information exploration with a stern, “Could we take this question offline please?”

  16. Venkat,
    I have come to this post via your article “How Conceptual Metaphors are Stunting Web Innovation” at, and will need some time to digest what you have to say, but my immediate response is that this is a brilliant article. Text on a flat page is transformed into a reader directed multidimensional meaning making experience via hyperlinking. Just brilliant.

  17. “The second method [ ‘Amitabh Bachchan, the Bollywood superstar, stared grimly from a tattered old Sholay poster. Sholay, as everybody knew, was the blockbuster that truly established Bachchan.’ ] is simply technically bad. If you can’t solve the problem of exposition-overload, you shouldn’t be writing fiction.”

    You could probably call that one the “Dan Brown Method”.

  18. Great post. I’m working on a book proposal based on my blog. The process consists mostly of copying and pasting blog posts into my big Scrivener document and then trying to edit around all the now-missing hyperlinks, embedded videos, mp3s and images. It feels like an act of violence to my writing, though a necessary one if this thing is over going to make it onto the page (and, hopefully, the Kindle etc.) I love the metaphor for the outbound link as yielding the stage, and the language of gaming and pathfinding to describe the act of reading on the web is exactly right. Thanks for helping me clarify these thoughts so neatly.

  19. I think my brain just exploded after reading that. I won’t look at in-text hyperlinks and ‘target=”_blank”‘ in the same way. Thanks for the fascinating read.

  20. “There is but 1 single book” may be isomorphic to “There is but 1 infinite game.” (Carse)

    (I’m always thrilled to pass through wiki-thinking….)

  21. Wow! I have been looking yahoo all night just for this and i eventually found it in this article!

  22. Just stumbled across this from a reddit repost, absolutely loved the article. I’m sure you knew/now know about it, but seems extremely relevant.

  23. Really a very beautiful and enlightening article. Thank you.