I don't want to be legible to a computer
I made a font of my handwriting. It was for a longer piece I intended to create on a standalone website, using CSS to stretch, shrink and generally mess with my writing in ways that challenge what it meant to be legible. I assumed I would be starting with an illegible handwriting font (given my illegible handwriting).
Breaking apart my letters, removing the cursive that bound them, and confining each letter to a box helped the legibility a lot in my case, and I assume in many others. It's a popular ask, using handwriting to make a font, but other than jokey "doctor's handwriting" fonts or pretend children's writing, it's difficult to find genuinely hard-to-read examples, Webdings (originally built for incorporating graphics into websites, apparently) and other purposely obtuse fonts aside. This may be obvious, but it's interesting; a font serves to communicate, after all.
"Don't mistake communication for legibility. Just because something is legible doesn't mean it communicates and, more importantly, doesn't mean it communicates the right thing." - David Carson
Legibility is "the quality of being clear enough to read". A process where I wrote a bunch of letters in little boxes and scanned them into a program made my writing legible, and to some even pretty. Legibility is a prerequisite to understanding, a prerequisite to communication, but not sufficient for either.
Understanding requires context, the necessary vocabulary to be clear and readable, and for the concepts to either be clearly communicated or at least close to the edge of the reader's knowledge. I think it and communication are very tightly nested subjects, where one hopes that they would always co-exist. I keep thinking of a complicated lecture, for example; here I understand the language of the speaker, and they are trying to communicate with me, but I may not understand the content of the lecture without further study or questioning the lecturer.
Making the lecture legible would mean speaking in a language available to the participants, using concepts that they should know, and not using needlessly complex words. These alone are not enough for a lecture to be understandable and communicate to all participants. It also needs to be accessible; the lecturer needs to allow recording the lecture, allow questions, have handouts or slides, and allow notes, amongst myriad other requirements attendees may have. Accessibility and understanding are not as easy as making everything available.
Making ourselves legible to computers, through the rise of typing, coding, OCR, scanning, scraping, and myriad other technologies has solved old problems and caused new ones. Though many new things are available and clearly presented for small subsets of the population, they are not clear and available for everyone. The reasons are varied. Visual impairments and motor and other disabilities are not catered to in the vast majority of websites. People who use the web on unstable internet connections due to cost or location are ignored as new devices and wireless standards like 5G appear and are replaced by faster alternatives. Whole languages are not available to even browse in online. Low digital literacy keeps people stuck offline, or stuck in small, familiar corners of the web.
With legibility of this type, without accessibility or understanding considered, the web is where everything is available to "everyone". Everyone who can read in English. Everyone who knows the right sites to visit. Everyone who has the right device and speed and amount of internet. Everyone who can use the internet in the ways lauded by the makers of these tools. Clearly, "everyone" isn't many people at all. The myth of the open web has fascinated me for as long as I've known what a computer was. When I was an ardent believer in the promise of open source, programming and publishing knowledge online, it galvanised me. It encouraged me that though technology may be making some things harder it was going to make everything better, one bright shining day. It was a belief in the power of the legible internet, of legible technologies making a more open, fair world. Open source of course has led to innovations the world over, as with more eyes comes the possibility of more being discovered and fixed. Companies are still built on the work of overworked maintainers getting very little compensation and even less credit. The vast majority of people never get an opportunity to contribute to this innovation, given it is all in English, largely unpaid, and often hostile to new contributors.
"As the scientific forester may dream of a perfectly legible forest planted with same-aged, single-species, uniform trees growing in straight lines in a rectangular flat space cleared of all underbrush and poachers, so the exacting state official may aspire to a perfectly legible population with registered, unique names and addresses keyed to grid settlements; who pursue single, identifiable occupations; and all of whose transactions are documented according to the designated formula and in the official language." - Excerpt From 'Seeing Like a State: How Certain Schemes to Improve the Human Condition Have Failed' by Professor James C. Scott
Online, legibility can cut both ways. The recent rise in generative AI has shown us once again what happens when new technologies run rampant. They've shown yet again what happens when something previously non-machine readable becomes legible to a machine. Something a computer can understand, manipulate, "create", for one definition of the latter, is devalued, "democratised" and exploited for profit.
Web scraping, the practice of using a computer program to collect information from the internet, has been on my mind a lot recently. It wouldn't be possible without a web that had this specific type of computer legibility cooked into it from the beginning. We've made strides backwards in this easy interoperability, but whether scraping should be allowed in the first place hasn't been questioned much until this new wave of AI interest.
Whilst the ability to scrape text, images, links and other information from sites on the internet has always been legible to computer programmers, I’d say it’s often been a surprise to those who are purely users of the web. One of the first incidents I remember where users had posted into what they thought was a closed ecosystem and had it scraped was an OkCupid scraping, where the (potentially very sensitive!) data from the popular dating site was taken without permission in 2016. Though that and several previous scraped datasets were in murky waters in terms of scientific ethics, technologically they had used tools that are at the disposal of any coder, and have been for a long time.
For as long as someone has made a machine that cuts out perceived or actual effort, people have lost their jobs. Whether a computer does a better job, or whether the job needed to be done by a machine, is unimportant when it comes to the matter of efficiency. It's impossible to be efficient to a computer system without being legible to a computer system. Efficient writing is legible writing, as short as possible, sparing no words, clear and without ornament in font or vocabulary. Efficient art is produced without needing a human it seems, preferably compensating as few people as possible, drawing on all human artistic knowledge possible for inspiration and regurgitation.
The long arm of digitally guided efficiency has spilled out onto the world at large, worsening our relationship with technology, changing who serves who. Becoming legible to computers has encouraged a world where we are technology-centred, rather than human-centred, or even life-centred. Digitally mediated schedules for workers; unemployment and disability benefits meted out by opaque algorithms; and face detecting technology deployed against black citizens are just a few examples of a world that privileges the word of algorithms above people.
"The fetishised life without friction. … It wasn't, to me, an aspiration. It was not a prize." - Uncanny Valley, Anna Wiener
When we speak of legibility, we should keep in mind two questions: "legible to who?" and "to what end?".