Monday, June 21, 2010

Are computer languages really languages?

A group of professors, librarians, and information technology professionals from a mix of universities joined in spirited conversation at a reception on the 18th floor of the Providence Biltmore Hotel last week, responding to the friendly provocation of a colleague. The topic: should Ph.D. candidates in the humanities be permitted to fulfill their programs' language requirements by learning a computer language? Currently, doctoral candidates at most universities are required to learn one or more human languages in order to complete their degrees.

For example, the UC Berkeley Department of Philosophy requires that the candidate:
pass a departmental examination in French, German, Greek, or Latin requiring the translation of 600 words in two hours with the use of a dictionary."

At Harvard, if you want a Ph.D. in English, you must:
show proficiency in either two ancient languages, or two modern languages, or one ancient and one modern language.

At Willamette University:
A reading knowledge of one modern language is often required for a Master of Arts degree, and two languages for a Ph.D. degree.

At the University of North Carolina, if you're going for a Ph.D. in linguistics it seems to be assumed that you know your way around the Indo European language family, so:
All students must complete one year of a non-Indo-European language or one semester in the structure of a non-Indo-European language.

Interested in getting a doctorate in French at Penn?
In addition to French, students are required to demonstrate reading knowledge of another foreign language, normally one that is used significantly in their chosen field of specialization."

None of these advanced degree programs accepts knowledge of a computer programming language in fulfillment of its foreign language requirement. Should they? Credible arguments were advanced on each side of the debate at the Biltmore.

To my suggestion that computer programming languages are necessarily reductive, while the humanities are naturally synthetic -- and therefore programming languages are no substitute for human language requirements in humanist Ph.D. programs -- a librarian at one university nodded agreement, while a distinguished faculty member from an institution on the other side of the country noted that there are strong arguments for the value of reductive approaches to scholarship. (Fans of artificial neural networks might simply disagree with my characterization of computer languages.)

To a suggestion that human languages contain a wealth of implicit and explicit culture, and of perspective that may differ significantly from that contained in another, a pair of IT professionals responded that the logically rigorous intellectual culture in which computer programming is grounded is different from cultures familiar to many (and perhaps most) humanists. They asserted that programming languages are deeply useful and important -- arguably essential -- to those who wish to understand the modern world. Programming languages, they proposed, reveal alternate views of the world as well and fully as human languages.

Some suggested that a scholar's engagement with sophisticated software tools would equip her to utilize technology in the service of humanist inquiry, without need to grind through actual coding. Others replied that inability to decipher the assumptions implicit in the way those tools are realized (via code) render their workings opaque, and their utilization clumsy and superficial.

Most agreed that where a doctoral language requirement reflects expectations that a Ph.D. ought to be able to access knowledge in the dominant languages of the academy, the requirement verges on outdated. While knowledge of French, German, Greek, and Latin might once have opened access to most of what is known by European academics, the languages of scholarly discourse over the past several millennia, in Europe and elsewhere, are far broader than these. Moreover, the conceit that knowledge of multiple languages might permit a single human mind to ingest most or all the scholarship of any substantial discipline may have been credible some hundreds of years ago, but today is ridiculous. There's too much out there.

Not long after the reception ended at the Biltmore, I wished I had advanced the argument that grasp of a programming language's syntax and the ability to code up a few simple classes or subroutines is no guarantee a newbie programmer will grok the intellectual framework on which that language is based. But then I got to thinking further. It's also true that mastering a few semesters of Greek doesn't equip a person to deeply understand Homer, Plato, and Aeschylus.

I'm trying to maintain a semblance of equanimity here, but I don't suppose it's hard to see where my sympathies lie.

I started dabbling in computer programming in the eighth grade (Fortran and Basic, with punch cards) and have written dozens to thousands of lines since in Pascal, dBase and its variants, Visual Basic, C, Java, Ruby, PHP, Perl, SQL, and so on; not to mention scripts in unix shell and DOS batch flavors, and a bit of Javascript. I've also done more than my share of machine-parsable typing in declarative languages and textual markup, from Ant to XSLT, from HTML to CSS ... and so on. I've never taken seriously the notion that any of these belong in the same category as human languages. Sure, you can build fun and useful logical constructs that cause machines to do neat stuff. But the sum total of every line of code I've ever written isn't a thousandth as expressive of humanity, or of the world as humans perceive it, as any single short story I've sent out in search of a sympathetic editor.

On the other hand, I'm not a linguist. And I think the world will be a better place if I decline to play one in the blogosphere.

So I hearby solicit your opinion.

Is it as intriguing to you as it was to last week's gathering on the 18th floor of the Providence Biltmore to consider whether computer programming languages are equivalent in intellectual power, expressive richness, or cultural significance to human languages like English, Chinese, French, Arabic, German, Sanskrit, Spanish, Russian, Farsi, or Japanese?

Do you think a computer language ought to fulfill a foreign language requirement for any or all Ph.D. programs in the humanities? Or not?

(Thanks to Quinn for the reception photo.)

1 comment:

  1. The very debate struck me as ridiculous at the beginning but after I read the whole blog, I have to admit that I was persuaded somewhat.

    But, if the Ph.D. candidate is in humanity fields, I still don't think how a computer language can substitute for a human language, because the language here is not just a skill but a gateway to a new culture.

    I know that calling masters in science and technology field Ph.D. is archaic and wrong-headed nowadays, and they perhaps should be allowed to learn a computer language rather than another tongue, considering how dominant English has become. A case can be made here, perhaps.

    But, I still think as a Ph.D. candidate, whichever field he/she might be in, he/she should not be exempted from the contact of foreign culture. This is a slippery slope. If we say that they don't need to learn this, then next day, we can say that all nonessential courses can be get rid of - no more history, art, physical education, social skills, civic lessons are needed.

    Then, we will have robotic Doctors in science and technology field. Are we going to let efficiency dominate us so much so as to forgo the meaning of human kinds?

    Matthew Felix Sun