29 August 2010


I've come across several links this morning to this Guy Deutscher NY Times piece reconsidering the Sapir-Whorf hypothesis and language (e.g. here and here).

I think it's good and worth reading, but not that great.  If you want some really good coverage of how language shapes our minds, listen to this episode of Radio Lab from a couple of weeks back, titled "Words".

Sapir-Whorf is all about how different languages shape the thinking of the people who speak them, but the Radio Lab episode gets at how language itself — any language at all — interacts with our ability to form complex thoughts.  For instance, when performing a distractor task that impedes verbalization of phrases like "to the left of the red wall" subjects become markedly worse at navigating around spaces where the primary visual cue is a red wall.  I'm not entirely convinced by the evidence they presented (I'd have to read the original studies), but it seems reasonable.

Some minor notes on the NY Times piece:
Do English speakers who have never heard the German word Schadenfreude find it difficult to understand the concept of relishing someone else’s misfortune?
English actually did (or does) have a word for Schadenfreude: epicaricacy.  Let's all use it more.
Languages like Spanish, French, German and Russian not only oblige you to think about the sex of friends and neighbors, but they also assign a male or female gender to a whole range of inanimate objects quite at whim. What, for instance, is particularly feminine about a Frenchman’s beard (la barbe)? Why is Russian water a she, and why does she become a he once you have dipped a tea bag into her?
Wait, wait, hold on. Let's clear something up before we move on.  Calling these categories of words "gender" is a historical accident. It's a result of Latin "genus," meaning kind or family or similar grouping. "Gender" is grammar should be understood to mean something more akin to "genre" than "sexual grouping." It's an unfortunate turn of events that we still label the primary groups in Romance languages "male" and "female;" far better we abandon those and adopt something with less connotation like up/down, charm/strange and top/bottom.
When speakers were asked to grade various objects on a range of characteristics, Spanish speakers deemed bridges, clocks and violins to have more “manly properties” like strength, but Germans tended to think of them as more slender or elegant. With objects like mountains or chairs, which are “he” in German but “she” in Spanish, the effect was reversed.

In a different experiment, French and Spanish speakers were asked to assign human voices to various objects in a cartoon. When French speakers saw a picture of a fork (la fourchette), most of them wanted it to speak in a woman’s voice, but Spanish speakers, for whom el tenedor is masculine, preferred a gravelly male voice for it.
This made me think of Köhler's "Booba-Kiki" test.

This was perhaps the best bit:
In what other ways might the language we speak influence our experience of the world? Recently, it has been demonstrated in a series of ingenious experiments that we even perceive colors through the lens of our mother tongue. There are radical variations in the way languages carve up the spectrum of visible light; for example, green and blue are distinct colors in English but are considered shades of the same color in many languages. And it turns out that the colors that our language routinely obliges us to treat as distinct can refine our purely visual sensitivity to certain color differences in reality, so that our brains are trained to exaggerate the distance between shades of color if these have different names in our language.
This didn't make a ton of sense to me until a friend pointed out that we have a top-level descriptor for the (hue, saturation, value) triplet (0,100,100) — "red" — and a similarly basic label for something around (0,50,100) — "pink." And we have a basic label for (240,100,100) — "blue" — but we don't have a top-level label for (240,50,100). In English we would have to call that color some variation on "blue" like "light blue." Why do we have "light blue" but no "light red"?

This is the beginning of the conclusion of Guy Deutscher's piece:
For many years, our mother tongue was claimed to be a “prison house” that constrained our capacity to reason. Once it turned out that there was no evidence for such claims, this was taken as proof that people of all cultures think in fundamentally the same way.
(1) This kind of overreaction in scientific circles really grinds my gears. Maybe just because my field was victimized my Minsky and Papert's spurious condemnations of single layer perceptrons, which scuttled all of Neural Networks for several decades. Just because a theory doesn't hold up the wildest expectations of it's proponents doesn't mean that it entirely lacks validity.

(2) I like the Radio Lab piece because it's format is the opposite of the idea that language constrains thought: language enables thought. It's the same notion underneath, but a much more interesting way of thinking of it.  Be sure to check that out, and for more anecdotes of language weirdness, check out this story in The Economist I blogged back in January.

No comments:

Post a Comment