In Defense of Plain Text

Let’s start with a fun fact — one that might be hard for you to believe. Your computer is really fast.

Yes, your computer. I know, it seems impossible. We’ve all twiddled our thumbs during an interminable software update or while loading a measly handful of Chrome tabs. (If you’re still using Chrome, now is a great time to change that.) And who among us hasn’t tried playing Minecraft on a decrepit Macbook Air while listening to the death rattle of its laptop fans? (By the way, your laptop fans might be part of the problem.)

And yet, I ask you to believe the impossible: your computer is really fast.

A (Quick) Example

If you don’t believe me, try this. And if you do already believe me, you can skip this part.

Open the terminal on your computer. If you don’t know how, you can Google it. Stop whining — this is only going to take thirty seconds.

We’re going to use this link, which goes to Project Gutenberg’s online copy of Alice in Wonderland. I’m giving it to you so that you can reassure yourself that I’m not Trojan horsing every file you own.

Run this command:

curl https://www.gutenberg.org/cache/epub/11/pg11.txt | grep -i alice

Holy hell. What just happened? Your computer just sliced through Lewis Carroll’s magnum opus like a coked-out chimp with a machete and spat out every line that contains the word “Alice.”

That’s a task that would probably take a human, working relatively quickly, around two hours. (And that’s not even counting the time it would have taken you to get in your car, drive to Barnes & Noble or, God forbid, a public library, and try to avoid eye contact with the person behind the desk.)

Also, keep in mind that this is all using the English language. Never mind math, which is where your computer really excels.

The Problem

If your computer is as powerful as I keep telling you, then why are basic tasks so painfully slow?

The unfortunate fact is that no matter how much computing power we have, our favorite software seems to magically expand to use it up.¹ This has been going on since 1986, and Niklaus Wirth wrote about it in 1995 — see Wirth’s law. Basically, he attributes this phenomenon to designers failing to discriminate between essential features and extra ones. The software we have now is stuffed full of bells and whistles like animations and cloud integration (and don’t even get me started on AI) that were once considered extraneous but are now commonplace.²

In fact, all computing once resembled what you just did with Alice in Wonderland, and even things as basic as graphical windows and a mouse were considered superfluous. Eben Moglen recounts the first time he witnessed a point-and-click computer mouse:

What I saw in the Xerox PARC technology was the caveman interface: you point and you grunt. A massive winding down, regressing away from language in order to address the technological nervousness of the user.

One shudders to think what he has to say about touchscreens.

To be clear, I’m not against graphical user interfaces. I think they’re really handy,³ and although Moglen is pretty evocative, he also comes across a little old man yells at cloud. But he does have a point, which is that a lot of our work might be done quicker and more elegantly without fancy buttons in the way.

In 2013, Rasmus Andersson (who created the spectacular typeface Inter, in which parts of this page are set) wrote this article envisioning a future of human-computer interaction that will be higher-fidelity and ever more expressive. Think haptic gloves, voice control, augmented reality — basically, full Minority Report. He also offers the following as justification, which would probably make Niklaus Wirth shoot himself:

Your average smartphone today can perform around 4 000 000 000 instructions per second (per core) — that means that what’s in your pocket today is not only extremely small in comparison to the giant computer Sketchpad was running on in 1963, but is also 10 000 times more powerful.

Now, remember the 1950s called and still haven’t got their programming toolbox back.

So what’s the object of Andersson’s ire? What’s in the dusty toolbox, older than the Civil Rights Act, which we still dig out of the shed whenever we need to issue instructions to a microprocessor?

It’s text.

The Plain Text

Modern computing runs on text.⁴ The webpage you’re reading this on, the browser you used to open it, and the operating system you used to open that are all powered by code in a language that, at least roughly, resembles English.

We can store written language digitally as a plain text file, like the one we downloaded from Project Gutenberg. But we can also store it in a more intricate format, like a rich text file, which can include fonts and formatting, or HTML, like this post.

Some of these formats serve the dual role of being expressive to both human and machine. For example, even if you don’t know Python, you can probably figure out what this code does:

def get_director_of(movie):
        director = movie.metadata["director"]
        return director

So when I refer to plain text, I’m not referring exclusively to literal .TXT files. Instead, I’m talking about any file that you can open and meaningfully understand in a basic text editor — Markdown, CSVs, XMLs.

These formats might seem archaic or abstruse. Maybe you’re on Andersson’s side. We’re a quarter of the way into the 21st century — shouldn’t we be able to do everything we want with slick WYSIWYG drag-and-drop interfaces, snapping together our digital creations like so many LEGO bricks?

Well, maybe. But I believe that there is still a place for plain text in everyone’s tool shed. Yes, it’s hard to ignore the forklifts and the chainsaws, but somewhere in the back, there’s a humble solution with enormous power and potential.

An Anti-PDF Manifesto

The first time I considered the benefits of plain text, I was searching for a way to read screenplays on my Kindle. If you haven’t worked in Hollywood, screenplays are typically exchanged in PDF format, but after years of frying my eyeballs on blue light, I sought a reprieve in the form of e-ink. Plus, I wanted to read screenplays from the comfort of couches/buses/trains and other such places you can’t always haul a PC.

Eventually, I made what’s now Unscript, which can display various screenplay formats and output ePub files for Kindle. But during the process, it became immediately apparent that one of the biggest obstacles would be figuring out how to decode PDFs.

A screenplay, in its basic form, is ultimately just text with a limited set of meaningful transformations applied to it (unless you’re Julian Simpson) — e.g., dialogue is indented, transitions are right-justified, etc. This makes it very easy to identify which lines are scene headings, which are character names, and so forth.

Unless, of course, you’re using PDF, a hot mess of a document format that bakes all of this text into what is essentially an image. The idea, of course, is that it preserves the format across devices, at the minor expense of any flexibility whatsoever.

It also inflates the file size. A recent sixty-page TV pilot I wrote is 262 KB in its original Final Draft format, but blows up to almost twice that (500 KB) as a PDF. Converting the script to a plain text Fountain file reduces the size to only 53 KB. In other words, the PDF is about ten times the size it needs to be, despite, again, actually containing less useful information.

Okay, but let’s be fair. PDF serves a purpose: it’s an export format. It emulates the printed page. We can lock in our design choices for something like, say, a restaurant menu and distribute it to customers so they have an optimized viewing experience.⁵ Right?

If only. When was the last time you Googled “how to edit PDF?” I probably did it six times last week. People sling around PDFs like the common cold, and often, they’re working documents which you’re expected to fill out or sign or rewrite. The Internet is flooded with tools that claim to melt down these monstrosities, but none of them really work because — at the end of the day — PDFs aren’t supposed to be edited.

What if, instead, we used file formats that were thoughtfully chosen and extensible? Imagine if your tax documents, for example, were interactive, responsive forms rather than PDFs crammed with fine print. (TurboTax did, and that’s why people give them money.) Imagine if Hollywood sent around screenplays in Open Screenplay Format so writers and executives could take those pages anywhere. Imagine if you could download all of those contracts, statements, and restaurant menus as exactly what they are — text.

The Poor Image

In “In Defense of the Poor Image,” Hito Steyerl writes:

The poor image is a rag or a rip; an AVI or a JPEG, a lumpen proletarian in the class society of appearances, ranked and valued according to its resolution. The poor image has been uploaded, downloaded, shared, reformatted, and reedited. It transforms quality into accessibility, exhibition value into cult value, films into clips, contemplation into distraction. The image is liberated from the vaults of cinemas and archives and thrust into digital uncertainty, at the expense of its own substance.

In today’s society of increasingly glossy, opaque digital experiences, the plain text is perhaps the poorest image of all. Like a blurry video, it is ugly and unpolished. Its file size is rationed at the cost of fidelity. But it also brings with it the many benefits of Steyerl’s poor image.

Plain text is portable. Proprietary file formats like .PSD or .CDR are only openable in a single specialized program. And, by the way, if you decide you no longer want to shell out a monthly subscription fee to Adobe? Good luck taking your Photoshop projects with you. In contrast, text files are platform-agnostic and can be opened in any number of programs, many of them free and open-source.

This also makes them future-proof. Text isn’t going out of date any time soon; Project Gutenberg’s copy of Alice in Wonderland will be good for a long time. Other forms of archived work, though, are vulnerable not only to physical disaster or destruction, but also to leaps in technology which may render them unusable.

Furthermore, the portability and tiny footprint of plain text mean that it can be distributed globally, across slow Internet connections, to any device with a screen. With Internet speeds modulated by the landscape of global economics and urban/rural divides, content should be extremely lightweight to be materially accessible to as many people as possible. There are hardware limitations, too — cutting-edge technologies like virtual reality and artificial intelligence demand cutting-edge computers, so unless Zuck plans on shipping a bunch of Meta Quest headsets to Burundi sometime soon, I don’t think a lot of people there will be playing Arkham Shadow.

The shareability of plain text also means it can be more easily exchanged outside authorized circuits of distribution, that is to say, pirated. And while not always morally upright, this is at least progressive;⁶ in the case of Lewis Carroll, who’s been dead since 1898, I think we’re pretty much in the clear. “Poor images are thus popular images,” as Steyerl says.

Finally, and perhaps most importantly, plain text is eminently editable and invites the user to modify it. You can hack, slash, chop, and screw to your heart’s content, becoming the critic, translator, and author. When spectators become actors, when art opens “the infinity of new possibilities,” it becomes revolutionary. And so plain text becomes a format for the future: anti-authoritarian, anarchist, liberated.

A Suggestion

The poor image is no longer about the real thing—the originary original. Instead, it is about its own real conditions of existence.

Having made both a practical and ideological argument for the value of plain text, let me wrap up by going back to the beginning. Your computer is really fast — at doing its job; i.e., at opening and closing circuits. At processing data.

It seems like the high-resolution hamster wheel is about mimesis, where the end goal is Ready Player One-style total immersion. But text, and its unexpected plainness, are a welcome reminder that what we’re doing is representing.

So here’s what I’m asking: I want you to consider what you’re doing when you sit at the keyboard. You’re not just exporting a PDF, or streaming a movie. You’re choosing ways to process and represent data.

So choose thoughtfully. And while you’re waiting, open up your text editor, and type something.