Tech’s Masturbatory Historiography

John Ohno on 2018-12-10

There’s this approach to computing history where we focus on work that looks shallowly similar to current norms, claim that work was ‘prophetic’ or ‘ahead of its time’, & mostly ignore differences & intent except as quirks. Because of the recent 50th anniversary of the Mother of All Demos, we’re getting a fresh batch of them. Fuck that.

These folks weren’t trying to predict our current future. They were trying to create a future worth living in, & we failed to make that happen.

Engelbart could do the Mother of All Demos in 1968 because that was the state of the art in available technology. It looks familiar not because it’s ahead of its time but because we stopped progressing — and we stopped before catching up with Engelbart’s goals.

Technology’ does not have a teleology. Particular technologies have biases — particular behaviors they are a better fit for, which they thereby encourage but do not enforce. (This is what is meant by ‘the medium is the message’: any widespread technology applies its biases across all its particular uses.) We should be careful of naturalistic fallacies & technological determinism: when the biases of a particular tool don’t fit our desires, we should replace the tool, not the goal.

Doug Engelbart wasn’t trying to create Skype & Jira.

Ted Nelson wasn’t (and isn’t) trying to create the Web or CD-ROM.

Alan Kay wasn’t (and isn’t) trying to create the Macintosh.

A lot of these folks are still alive. The rest literally wrote about what they wanted.

Engelbart, Nelson, & Kay are part of a particular tradition: trying to use computers as an extension of human cognition. Not just memory, but imagination, & other mechanisms that we don’t have names for. They’re pretty unsatisfied by the tendency for progress in this field to go backward — for available software to gradually become worse along the metrics they care about.

Engelbart was (along with Sutherland & Licklider) interested in intellectual augmentation and symbiosis: the computer was to become an extension of the user’s mind, or failing that at least an active and trusted collaborator. The feeling of symbiosis with the machine is, even today, alien to most of us! Aside from a handful of people who have become very proficient with a command line or REPL, we have never experienced the computer as a seamless extension of the mind (as we frequently do with notebooks and other kinds of freeform tools): instead, the computer remains at best a collection of poor-quality highly-specific single-use tools. Engelbart’s particular research (from which nLS was born) was with trying to get multiple people to symbiotize with a shared computing environment (and thus, with each other). Social media is as far as we have gotten with this.

Nelson is interested in expression: making footnotes, intertextual references, and all kinds of context directly visible. What the web calls a hyperlink is what he called a ‘jump link’: you activate it, your original context disappears, and you are thrown into hyperspace. Jump links are a slight improvement over channel-surfing, but they do not represent visible context (which Xanadu has, for the past 60 years, represented as parallel columns with visible connections between them). A handful of other ideas from Xanadu have been adopted in a handful of contexts (MediaWiki has a concept of ‘transclusion’ that’s marginally closer to Xanadu’s than the WWW, but is unnecessarily limited in scope; the idea of transcopyright has been perverted into game microtransactions & DRM webstores), but the only part of the world wide web Ted Nelson is willing to take credit for is the browser ‘back’ button, because he considers the rest a disgusting perversion of his ideas.

Kay, influenced by Piaget & Papert, was interested in building environments for growth. Papert saw the computer as an opportunity to build a ‘math-land’, where children learn math the same way children living in France learn French (i.e., where natural human curiosity and the drive to create leads them to learn things that are normally a classroom slog). The graphical Smalltalk environments Kay worked on were ‘computerlands’: the Smalltalk language was intended to be small enough and simple enough to be understandable without formal training (or with minimal formal training), yet powerful enough to do anything without becoming a turing tar-pit; meanwhile, the environment itself was a homogenous / language-based system, where the line between use and programming was fuzzy. Interactivity — real interactivity, where users have real power (to do things unauthorized and not even imagined by the developers) — is fundamental to this vision. Meanwhile, today’s home computers don’t even ship with a development environment, and most computers are tablets or phones (which don’t have a keyboard and ultimately act as glorified televisions).

When people write articles like ’50 years on, we’re living the reality first shown at the “Mother of All Demos”’, they contribute to a dangerous myth: that we are living in the best of all possible worlds, and that work that did not have today’s world in mind is flawed for that reason. The visions of Engelbart, Nelson, and Kay are viable today because they were viable decades ago, and they are viable because they are not visions: they are plans. It has fallen to us to implement them.