Monday, July 14, 2008

Can languages be treated as if words were genes?

"Some researchers think that the evolution of languages can be understood by treating them like genomes — but many linguists don't want to hear about it" Emma Marris reports in Nature News. (Paywall)

In that case, treating language families like a genome evolving over generations raises some interesting questions. We must assume that the speakers of languages do not have much conscious input into the fate of languages, just as we do not have much conscious input into our physical genetic codes.

Can we all say out loud together "reductionism!"?

As a general rule, reductionism does not work when a given thing is reduced below the level of meaningful information. We will not usually understand words better by cutting them up into letters, for example, or by treating them as if they were genes.

Examining the letters in a word can tell us some interesting things, as we will shortly see, but not usually the social and cultural meanings of the word.

We are constantly providing input into the language we use. One thinks immediately of the significance certain phrases acquire during an election campaign. Barack Obama, for example, may well have rescued the word "audacity" from near oblivion by his phrase, The Audacity of Hope. However, I think the word will go back to near extinction in a few years.

Some might argue that a sort of Darwinian process governs which word innovations or recoveries succeed. Maybe, but the process begins with intelligent design. The skilled orator knows what effect he intends to have and how he proposes to create it. Given the importance rhetoric has played in the oldest cultures, we can assume that skilled orators have had their hand in the system all along.

A friend enumerates reasons for doubt more academically than I can:
Languages clearly are hierarchically-structured systems while actual science is far from understanding the hierarchical structure of genomes (if any);

Languages are tools daily used by intelligent agents who constantly use them (and change them) to express meanings. As such languages evolution is driven by intelligence. Genomes are not used daily by intelligent agents this way and even Darwinism [Darwinian evolution] denies a front-loader designer at their beginning;

Language always expresses concepts and manifests thought and ideas, it is fully teleological. According to Darwinism, genomes are not teleological systems.
Of course, sometimes looking at the individual letters can tell us something about words. The word "skirt" in Scandinavian languages became "shirt" in English (which tends to soften the consonants). Then later, English speakers also grabbed "skirt" and made it mean the bottom garment rather than the top one. To maintain the distinction between the two garments, we had to stick with Scandinavian phonetics this time. Our first grab was probably automatic, but the second required a conscious distinction in spelling and pronunciation.

And then, when I was young, the skort made a brief appearance - essentially a skirt with built-in shorts, pictured above. I am glad to see it has not entirely gone out of fashion. The word "skort", at any rate, was a deliberate coinage by an intelligent agent, probably the earliest skort designers, explaining their concept.

I would hate to think of an earnest researcher attempting to compile "laws of descent" for the English language and spending many hours trying to show how "skort" arose without any intelligent agency. On another planet, perhaps.

(Note: The image is from Bartleby's American Heritage® Dictionary of the English Language: Fourth Edition. 2000, a good place to bookmark. This isn't quite the type of skort I remember from 1960, which was loose cotton, and sometimes pleated, but it's close enough.)