?

Log in

No account? Create an account
 
 
12 November 2009 @ 03:06 pm
Geekery (babbling about the Sapir-Whorf hypothesis)  
papersky's Tor.com post Like swords, but awesomer: Made up words in science fiction and fantasy is very cool, and the comments are verra interesting.

But I have to say, I feel pretty sorry for poor old Sapir*. I have a sense the strong Sapir-Whorf hypothesis has turned him into a perpetual motion machine, spinning forever in his grave.

Because the strong version, that eternal straw-man, is completely, unfixably silly. You can't think of things you don't have words for? RLY? However did you learn what the words meant in the first place, since you couldn't think of the thingy until you knew the word? According to the straw man, my nephew can put objects together and demonstrate pretty sophisticated spatial reasoning about them without being able to think of them. Because he's not talking yet, and doesn't recognize all the words.

It's as mystical as Chomsky's language module wot didn't evolve.

Pity Whorf didn't have a prelinguistic nephew, I guess. Or, shall we say, the linguistic sophistication Sapir did. His writing is actively brain-hurty, the way sensible observations and total WTF run together in gleeful abandon.

Now, the weak Sapir-Whorf hypothesis (where "weak" means "not ridiculous") makes a lot of sense, is useful to writers, and has at this point a great deal of empirical evidence backing it up. But it's called "weak", and that's just not sexy :) So y'can't even talk about it without starting with "No, really, I don't think unnamed eskimos have a bajillion-gajillion words that all mean the exact same thing as the English word snow, and no, really, I don't think English-speaking people who happen to ski do, either, but I do suspect both groups distinguish between powder and slush."

Fact is, you can inclue a lot about a culture, and make it resonate, by thinking about how kinship terms in a fantasy setting's language are structured (whether or not you make up funny words for 'em), or whether the speakers use relative directional terms (left, right) or only absolute ones (North, south; or more likely something like uphill, downhill, depending on setting), or which color terms are basic, or whether the language specifies whether movement crosses a boundary, or or or.

It'd be so nice to be able to talk about that without starting for the words for snow or the fact that the Hopi did, actually, understand the concept of time.


---
* Not to mention Mary Haas, I think(?) it was, who's never mentioned at all.

ETA: Also to note: I don't feel sorry for Sapir 'cause of anything Jo said, but the fact that she too has to dismiss the strong version before talking about the weak version got me thinking about the fact that it's not just linguists and cog sci people -- everyone seems to have to.
---
 
 
Current Mood: geeky
 
 
 
Ashniashnistrike on January 12th, 2010 05:27 pm (UTC)
They're loads better than factor analyses. How do you avoid them?
shweta_narayanshweta_narayan on January 12th, 2010 10:34 pm (UTC)
By getting really sick, being allergic to the cog sci building, and therefore no longer doing experimental work for the diss :/
shweta_narayanshweta_narayan on January 12th, 2010 10:49 pm (UTC)
My real issue with ANOVAs was always that the people using them mostly didn't know what they meant in statistical terms. So they couldn't explain that to me. This is also why many papers about language just check for whether you can generalize beyond the subjects and not, you know, whether you can generalize beyond the specific linguistic prompts in any given experiment.

But hey, now my spousal unit is doing stats that barely anybody understands, so my head kersplodes a lot even without my needing to do the stats myself.