I’ve always envied people born in small countries like Belgium who grow up learning several different languages. And while I remain stumped by languages written in any script other than the Latin alphabet, I still dream of unencumbered months when I can get started on basic Mandarin.
I am also a fiction writer, who believes that there are uses to which language can be put that are different in kind, not just in degree, from the uses of everyday communication; that language, for the poet, is oil and brush, canvas and light.
So I was excited to encounter Vikram Chandra’s new book, Geek Sublime: The Beauty of Code, the Code of Beauty. Chandra is a highly accomplished novelist who supported himself for many years as a computer programmer and remains keenly interested in the language, history, and culture of programming. I am a philistine about these matters; I never want to know how the software works, I only want it to work. But the deeper I got into Chandra’s book, the more I felt I was beginning to see the contours of the particular communicative medium that is code, the way you begin to pick out words in another language—not sentences yet, let alone the argument that the sentences might be making, but at least the building blocks of what might one day be a new way for you to articulate ideas or make some rude form of verse.
The great debate of the book—a debate Chandra foregrounds but never resolves—hinges on what we mean by words like communication and art. Years ago, on a long rail journey across Europe, my husband was seated next to a man with whom he shared no language. Frustrated, the man pulled out a sheaf of paper, wrote down an algebraic problem, and handed it across to my husband. He solved the problem, produced one of his own, passed the paper back. In this way, they spent a pleasant three hours, “talking” in mathematics.
But, I pointed out when he related this anecdote, you weren’t communicating anything beyond your mutual enjoyment of math. You were just solving problems.
I had a bit of this same response to the first set of arguments Chandra put forth, about the elegance of code and the idea that “while a piece of code may pass instructions to a computer, its real audience, its readers, are the programmers who will add features and remove bugs.” I wrote to Chandra, to press this question further. Wasn’t code in the end, I asked, merely functional, not expressive? He responded, in part,
Code lasts as long as some books—there’s still code running that was written in the fifties and sixties, in very crucial places. So you’re communicating into the future, to your readers who you will never meet. And I think it’s here that programmers start collapsing the kind of language they use and its intents with art, because they have to —like physicists and mathematicians—confront questions of elegance. Even more so because the product of code (even if not code itself) is most often human-facing, say in a website or app.
Now, here’s something to contemplate. I don’t know exactly how it is, when I use an application or navigate the web, that I’m “reading” differently from the way I used to, but I know there is a difference, and that difference is effected by computer code. I had been thinking of programmers as the technicians who translated a designers’ vision to instructions comprehensible to the computer. But that’s nonsense, of course. Their creative power is far greater, and specific to their medium. Chandra goes on:
But I think there’s a more profound way in which code itself (and not just its human-facing product) inscribes itself on human consciousness. The structures of code surface themselves in the artifacts that nonprogrammers engage with. … We use Google and Facebook, but they also not only use us but shape us. So this is a kind of communication, although often unintended by the programmers, and not the kind of thing they often think about—how code reshapes bodies and consciousness and how we think about the world.
Code-makers, for instance, think largely in algorithms, just as the 4th century BCE grammarian Panini did. In one of Geek Sublime‘s many sublime tangents, I learned that Panini developed almost 4,000 rules that allow the generation of Sanskrit words and sentences from roots that in turn trace to phonemes and morphemes. With his lone text, the Ashtadhyayi, according to Chandra, “Panini created the fields of descriptive and generative linguistics.” When I reread Chandra’s book this fall, I’m going to try to hang on to the roller-coaster ride that takes us from Sanskrit to computer code and to figure out where the pleasure of rasa, “the meta-experience of experiencing oneself experience the stable emotions,” fits in. Meanwhile, I got caught up on the history of programming as a female—yes, female—occupation, the workings of logic gates (I get it!), and the horrifying notion of seven million lines of code, many of them plagued by bugs, that run the Pentagon’s software for payroll and accounting. As Chandra points out, 90 percent of the planet’s financial transactions rely on massive amounts of code rolling along as “Big Balls of Mud,” written in languages like Cobol, “the computing equivalent of Mesopotamian cuneiform dialects.”
I don’t speak those dialects, and I’m not likely to learn code. But Chandra is fully bilingual. And if he can explain this other language’s rhythms, versatility, history, and elegance—and, yes, the ways in which we may already be using some of its patois—I’m all ears.