Another need-to-get-it-off-my-chest-post. Feel free to skip if you dislike rants; it’s only very mildly constructive.
I’ve lost track of how many people have quoted this study to me as being a fun and telling way to characterise gendered languages. For those of you not familiar with it, it’s a study by a Stanford professor which says that people “gender” nouns: that if a table is feminine you will give it feminine characteristics like elegance, beauty; and a masculine bridge will get described as sturdy and strong, which are masculine traits.
As someone who speaks two gendered languages (one of them as a native), one non-gendered language, and is starting to make inroads into a second non-gendered language… NPR is giving you a false idea of how gender in gendered languages work (I have no idea what the original actual research is; this being the NPR report version, which I suspect distorts the truth). And I’m not denying that language shapes thought, or that genders are completely neutral in gendered languages: for instance, most animal names are male, and I have a devil of a time thinking of a goat as other than female (it’s “la chèvre” in French, which is feminine).
But I do have several issues with that taking that article at face value, and particularly in generalising those results to every single word. The first one is that this concept of “gender characteristics” sounds very much like something that a long-standing Anglophone speaker would come up with: a lot of Anglophones I’ve met have been fascinated by the idea of giving gender to nouns, but in a very odd way. No, I don’t think of a table as female. I think of it as gramatically feminine, which is a different beast. There is a difference–yes, they’re not totally dissociated concepts, but there is one.
Also–I’ve had a chance to interact with US mainstream culture for a while, and it’s struck me that it puts a lot of accent on gender separation and gender proper roles, which is again, fairly compatible with this kind of ideas. We’ve also been discussing this elsewhere with J. Cheney and Chris Kastensmidt, but there’s a whole “anthropomorphising” complex at work in English: Anglophones (or at least USians, I don’t know about UK people) are actually more likely to anthropomorphise their vehicles, computers and cars–giving them names and genders; and referring to them by those names. By contrast, French (and, it looks like, Portuguese) will look at you very oddly if you keep referring to your nice masculine computer. The French language is grammatically gendered, sure; but to all intents and purposes, gender is a dead attribute when it comes to most everyday things.
The other issue I have is with the notion of “gender characteristics”–I’m sorry, but though there are common points between the way cultures view male vs. female, there are also a heck of a lot of differences. People’s perception of “idealised” gender characteristics strongly depends on the culture/language. Very simple example: in France (or in most of the West), a manly man is someone who is strong, and generally good at sports. This is emphatically not the case in traditional Chinese or Vietnamese culture, where a manly man is slender and thin, educated, has beautiful long nails, and can compose beautiful poetry.  Puts another spin on the “lone hero”, doesn’t it?
You don’t even have to move that far: where I live in France, it’s usually considered very feminine to be always touching and kissing (on the cheeks) and hugging. Go to Spain and watch a couple of guys from that perspective, and they’ll still seem like a bunch of sissies, because Southern Europe cultures are very tactile.
And another thing… some of the most intensely gender-separated cultures (China and Vietnam, sorry, using what I know, and my repertoire isn’t large, but it will suffice for this) have non-gendered languages, where only the pronouns are gendered. So the gender of people is not the gender of words, and vice-versa.
The study doesn’t mention who they picked as tests subjects, either, but considering that it took place in the US, it’s making me wonder if the German and Italian speakers were pure native speakers with no second language, or if everyone had been immersed in US language and US culture for a while (I strongly suspect the latter). Whatever the case, it certainly looks like the aforementioned speakers got their “perceived genders” classified according to an American perception of gender. So, hum… sceptical, to say the least?
So, please, please, pretty please… do not tell me about the feminine table or the masculine computer? Gendered languages don’t work the way NPR would have you think.
A fun one is that long hair is usually considered a feminine and weak attribute today, at least in my social circles in France; but in Ancient China, wearing hair long wasn’t a particularly big deal–in fact, when the Vietnamese and Chinese first met, the Chinese thought the Vietnamese were Barbarians, because the men dared to cut their hair)