In Tim Berners-Lee’s original pitch for The Web, he described it as a “non-linear text system” for notes. It was designed for scientific researchers and academics to pass around documents that would otherwise be printed on physical paper. The point of the web was to mimic long, text-based, paper documents, but simply make them easier to move around.

Paper documents were the original metaphor for the web.

It’s no surprise we started with paper. At the moment of the web’s conception, computers were objects that lived among white-collar office workers. They were pitched as replacements; they would replace your inbox tray, your trundling filing cabinet, and your little yellow pad of sticky notes. Alan Kay made this cultural context clear when he designed the first desktop metaphor at Xerox PARC in the 1970’s.

Ever since then, our digital experiences has been structured in terms of dull office equipment. Documents, folders, physical inboxes, rubbish bins, bookmarks, windows, and little yellow stickies are still part of the visual language of our operating systems today.

Thinking about computer interfaces as physical office rooms naturally expanded into thinking about the web. Websites were presented as and spoken about as pages – as in, pieces of paper. We still do it today. We still imagine the web as a series of paper documents with text and images presented on them.

The surface you’re reading this on mimics a long white pape. We still call it a page or an HTML document. It follows the same typographic rules and conventions – black text on white backgrounds and a top-to-bottom / left-to-right heirarchical structure.

The Static Nature of Paper

This metaphor of the-web-as-paper works quite well. The otherwise immaterial and abstract nature of the web becomes understandable, familiar, and (most importantly) usable through the paper metaphor. We have all seen, handled, and created our own paper documents – a set of understandings we bring to our behaviour on the web.

I should point out that websites and paper do not share any inherent qualities. We purposefully designed webpages to look like paper, but it wasn’t some inevitable metaphorical fate.

In fact, the paper web has some harsh critics. Ted Nelson is one of the most vocal. As the original creator of hypertext and the visionary Project Xanadu

The Pattern Language of Project Xanadu

In Back to the Future : Hypertext the Way it Used to Be, Ted argues we’ve let paper become a metaphorical prison for our digital systems. When we designed digital documents to mimic paper, we unnessecarily brought along many of its physical limitations:

• We treat the page as the smallest unit of linkable information, instead of the sentence or paragraph.
• We remain married to the conventions of traditional typesetting established by paper printers.
• We insist on a single view – a mockup of a physical page – rather than embracing the flexible, multiple views possible in a dynamic computational medium.

Worse still, we left behind many of paper’s best qualities:

• We took away the right to add handwritten annotation in margins.
• We took away the ability to lay multiple sheets out in front of us and shuffle them around to find the best structure and sequence.
• We took away the ability to carry a piece of paper across a room and look it over with someone else.

This is an inevitable consequence of any design metaphor. Metaphors are designed to highlight qualities that the target concept and the source concept share, and hide qualities they don’t. They focus our attention on a particular set of affordances, to the detriment of all others.

The best solution to this metaphorical concealment is to practice seeing systems in multiple ways. Trying on different metaphorical lenses will change what we think the web is made of, what we think it can do, and its affordances, risks, and restrictions.

Escaping Paper and Making Space

Despite the persistent metaphor of the paper-based web, we’ve framed the web in many other ways since its creation. There are many qualities of the web that paper cannot accomodate or conceptually explain; decentralised governance, live connection to other people, dynamic audio-visual creations, and mob-like social congregations have no precedence on paper.

In searching for metaphors that reflect our experiences, we’ve called it a series of tubes , an information superhighway, a surf-able ocean of information, a global village, a town square, and a zero-gravity parallel universe known as cyberspace.

Out of all of these metaphors, the second most dominant after paper is physical space.

This is obvious in the way we speak about the web. Our words describe the web as a sprawling expanse of place-based pages – each piece has in it’s own location we step into when we visit a homepage. We go to websites. They have a permanent address – each URL (uniform resource locator) marks its place in digital space. We follow links and end up on a page, and go back to the previous page.

Our web browsers play into this narrative of moving through physical space. Netscape Navigator helps you find your way around. Internet Explorer and Safari are a call to adventure in an unknown landscape.

It didn’t take us long to flip the metaphorical script and design our webpages to mimic two-directional maps where you directly clicked on your destination, rather than scrolling a linear paper document. The most infamous being the Space Jam design of 1996_30ya ; all hail.

This spatial metaphor intensified as browser technologies advanced. In the early 2000’s dynamic programming languages like Flash and asynchronous JavaScript became ubiquitous. These languages enabled us to build instantly interactive websites without requiring page reloads. It became possible to design a webpage as a dynamic space, where clicks, drags, zooms, and pans had immediate effects. Websites began to look less like pages and more like spaces. Games sprung up with characters controlled by arrow keys. Digital whiteboards appeared that you could pan and zoom around like a frame over an infinite canvas.

For a long time these experiments were rare, exceptional, play-things though. The grain of the web changed faster than we could reconfigure our narratives about it, and we’ve been slow to realise what this medium is capable of. It is only in recent years we’ve seen a true flourishing of spatial interfaces that mirror our physical experiences of the world.

Embodying the Web

Mapping our understanding of physical space and direction onto the web was destined in a way that paper was not. Our physical bodies are the primary way we interact the world; we project our understanding of bodies in space onto everything we do. Particularly abstract activites. Politics has governmental bodies with arms of government. Economic indicators go up and down, two concepts that require a body in the middle. Time is a line our bodies move along; the future is ahead and the past is behind us. We’re looking forward to next week and last year flew by us.

There is no complex, abstract thought that does not require and involve our embodied understanding of being creatures with ups, downs, lefts, rights, ins, outs, ons, offs, and betweens.

This theory of embodied cognition – embodiment as essential and central to thinking – has been explored extensively over the last century. Researchers like Maurice Merleau-Ponty , George Lakoff , Eleanor Rosch , Barbara Tversky , and countless others have revealed how fundamental embodiment is to everything we think, do, and create.

The web is no exception. The physical and spatial qualities we design into it will always be more salient to us than its paper-ish vibes. The symbolic objects of the 1970’s white-collar office is no match for millions of years of primal evolutionary understanding.

Mixing and Making Metaphors

Despite the common folklore that we shouldn’t mix metaphors, we have no problem seeing the web as both paper and place at the same time. It’s a classic example of conceptual blending For more on Conceptual Blending, see Turner and Fauconnier’s The Way We Think – the mind’s ability to combine multiple metaphors at once and from them produce emergent meaning. We’ve seamlessly combined the two to create an emergent notion of location-based pages.

As George Lakoff and Mark Johnson made clear in their touchstone book Metaphors We Live By , metaphors are the basis of all human thought and reasoning. The metaphors we use to speak about the web are not simply linguistic trivia – they determine how we understand it on a fundamental level. It determines what we think the web is capable of, what risks, opportunities, and challenges it poses. Which means the metaphors we use to think about the web profoundly influence what we think the web is, what we think we can do with it, and how we might change or evolve it.

The Web in Geological Layers

The paper metaphor of the web hides almost everything about how the web works. Paper is a presentation layer. A flat object showing off the final result. Our shiny graphical interfaces offer us a limited range of options pre-selected by designers. They reveal none of the technical processes that cascade across the web when you press seemingly simple buttons like or

This is both a feature and a bug.

How else might we see this system if we move beyond our presentational paper web? Rather than seeing the web as a flat, inert object we passively read, what if we saw the web as an interactive topology made up of various layers?

If you know anything about how our current web infrastructure came into existence, you’re well aware the whole thing is a patchwork of technologies held together with duct tape. It is the furthest thing from a uniform system developed by a centralised team of developers.

Each piece of the modern web was built seperately. Unrelated individuals and teams of developers added layer upon layer. Each one building on top of existing systems, protocols, and standards.

This geological stack began with Vincent Cern and Robert Kahn establishing the TCP/IP protocol in 1973_53ya – aka. the birth of the internet. That foundation slowly amassed layers like URLs, HTTP, HTML, SMTP, FTP, IMAP, DNS, SSH, REST, and 99 other acronymns. This layered understanding is closer to how developers see a website – as a set of systems they use, connect to, and expand off.

When you’re building a webpage, you’re patching together a set of features and affordances. Do you want to add the ability to login? Do you want to integrate the Facebook like button? On a moral level I don’t recommend it, but this is the decentralised web and you’re free to do as you wish. Do you want to let people enter text and save it to a database in the cloud? All of these options touch on different infrastructure layers of the web.

There’s a data layer. There’s a security layer. There’s a visual design layer. There’s a hypertext layer that links to other locations. There’s an authentication layer. There’s an algorithmic layer.

Almost all of these layers are completely invisible to “end users” of the web. They just see a flat static page. There is no topography. There is no transparency to how this website is structured or built. You don’t know what other websites it’s connected it, or where it’s pulling data in from. The amount of technical literacy you need to even begin to see these layers is non-trivial.