Thursday, September 16, 2010

What shape is the Internet?

These days the "web" metaphor tells us less and less about the true topography of the Internet.

The current discussion started for me when @ironick quoted @Microsoft "The Web is about sites, and your browser should be, too." and asked "What's the difference between a site and an app?" @Cybersal quoted @dtapscott's alternative "web not about sites but platforms 4 collaboration". So @ironick asked "What's the WWW really a web of? sites, apps, pages, data... "

My first observation is that if the Internet is merely a collection of sites, apps, or even platforms, then it's not exactly a web. The word "web" appears to focus our attention on the connections rather than sites themselves. There are of course two kinds of connection that exist in the Internet, which we could roughly categorize as syntactic and semantic. A syntactic link is a hotlink coded in HTML, while a semantic link involves some kind of content relationship. For example, you'll note that in this blogpost I've gone to the trouble to add hotlinks to the tweets by Nick and Sally: if you wanted, you could go directly to Twitter to check their exact words. (Go directly to Twitter, do not pass Google, do not collect 200 cookies).

But even if I hadn't added the hotlinks, you'd still be able to find Nick and Sally and their Tweets, by copying their names or their words into a search engine. So I'm creating a semantic link just by referencing something that exists somewhere on the Internet, even if I don't tell you its exact location.

The original hypermedia experience was largely dependent upon syntactic links. For @ironick "the web still feels like hypermedia 2me: clicking from context (page, song, video, snippet, site, app) 2 cntxt 2 cntxt 2..." I agree that it often still feels like that, but I find that a lot of my Internet browsing these days involves typing terms into search engines, and I don't find myself following long chains of hotlinks. In other words, I tend to regard the semantic links as more interesting and more useful than the syntactic links.

Here are some of the many problems I experience with syntactic links
  • Sometimes the links aren't provided at all.
  • Thanks to an aversion to deep linking, many websites only provide a link to the home page.
  • Sometimes the links don't go direct but via some tedious aggregator or intermediator page. Spurious links whose sole purpose is to manipulate the search engines or generate advertisement traffic.
  • Sometimes the links take you somewhere boring or irrelevant or obvious (like a Wikipedia page), not-safe-for-work, or someone's idea of a joke (not Rick Astley again).
  • Often the links are compressed, so you can't see where you are being led. (No, I don't want to watch a YouTube video right now, thank you.)
  • Often the links contain all sorts of other coded information, to pass contextual information to the receiving website.
  • And then to cap it all, half the time the links don't work for you anyway, because they are out-of-date, or because the person providing the link has a subscription and you don't, or because there is some kind of context or syntax error.

Of course, there are problems with semantic links as well, above all the danger of over-reliance on the chosen search engine. But I still feel I'm more in control of the experience.

When we talk about the Internet as a world wide web (WWW), the word "web" seems to suggest a network stretching endlessly in all directions, allowing and encouraging the kind of browsing experience Nick mentions. But of course the fly's experience of the spider's web is quite different: being caught in one place, trapped for the benefit of the spider. For a long time, it has been the desire of major internet providers to trap users in one place: this desire is now apparently satisfied whenever users do not stray more than one or two clicks away from their favourite search engine or social networking site. Maybe that's what Microsoft is getting at.

Related Posts

What shape is the internet (continued)? (May 2014)
What shape is your intranet (May 2014)


  1. Richard, Sounds like the sites you visit provide much lower quality "syntactic links" than the sites I visit. Sure, I run into all the problems you mention, but I run into them only infrequently. I find that syntactic links, especially those found in the blogs and tweets that I read on a daily basis usually take me right to where I want to go.

    I find semantic linking via a search engine more susceptible to the litany of problems you describe. A perfect example is searching for quoted text, eg blog A contains a quote from blog B or news article B, but blog A doesn't provide a syntactic link.

    So to find the original B post or article, I enter a snippet of the quoted text into Google. What do I often find? The first several results are cloned web sites set up by some SEO operator trying to draw traffic to their ads by reposting someone else's content.

    And you have to admit that within sites like wikipedia, or netflix, or imdb, etc., syntactic links make for a powerful "browsing" (not searching!) experience. In those sites I feel like I'm really exploring lots of interesting connections--not searching.

  2. I agree with Nick that I may have overstated the value of semantic links over syntactic links, but my key point here is that in a world dominated by search engines and pervaded by SEO, the actual topography of the Internet results from a combination of syntactic links and semantic links. In other words, the Internet (product) is more than the technology (HTML).

    Indeed, the SEO guys are often trying to use invisible semantic links to trick the search engines, and the Internet is filling up with the "dark matter" generated by these SEO efforts.

    Many people collude with sites like Digg, and I often find I'm being taken to the Digg version of a story rather than the original. Okay, I can then navigate to the original, but clearly many people don't bother, and in the meantime Digg has clicked up yet another visitor.

    My second point is that what we think we experience on the Internet is to some extent an erratically constructed illusion, emerging from a set of strong commercial interests. We have come a long way from Tim Berners-Lee's innocent vision of hypermedia. Microsoft's latest pronouncements on the Web cannot be divorced from Microsoft's own stake in the evolving topography of the Internet.

  3. I posted my lengthy reply on my blog: .

  4. Nick's reply is on his own blog.

    The Battle Between the Two Gods of Hypermedia: The Reader and The Writer

    In my comment on his blog, I complained that the Internet is full of people who appear to be neither proper readers not proper writers, in the pure sense discussed by Tim Berners-Lee and others, but appear to spend their time replicating or linking to fragments of other people’s texts in the vain hope of randomly producing something either highly profound or massively popular.

    Nick suggested that these people can be regarded as bricoleurs, and references Daniel Chandler's article Semiotics for Beginners, who takes his notion of bricolage from Lévi-Strauss.

    But surely bricolage implies creating something. Most of the people I'm talking about aren't creating anything, merely wasting time. I'd prefer to call them flâneurs.

  5. "Most of the people I'm talking about aren't creating anything, merely wasting time. I'd prefer to call them flâneurs."

    Two responses:
    1) "most of the people": I'll simply invoke Sturgeon's Law as my response: "ninety percent of everything is crap." It's the other 10% that make the behavior or activity wholewhile.

    2) "I'd prefer to call them flâneurs." You're giving flaneurs a bad name. They are held in high regard in some circles.

  6. I could retort that 90% of flâneurie is crap - we can't all be Baudelaire.

    But from the Wikipedia article on the flâneur to which Nick refers, I find a more accurate word.

    The flâneur was often juxtaposed to the figure of the badaud, the gawker or gaper. Fournel wrote: “The flâneur must not be confused with the badaud; a nuance should be observed there. . . . The simple flâneur is always in full possession of his individuality, whereas the individuality of the badaud disappears. It is absorbed by the outside world . . . which intoxicates him to the point here he forgets himself. Under the influence of the spectacle which presents itself to him, the badaud becomes an impersonal creature; he is no longer a human being, he is part of the public, of the crowd.”

    See also Wikipedia: Badaud