Gendered first-person pronouns in Japanese?

Apr 15, 2012 01:13

In a recent IRC discussion, someone mentioned that Japanese first-person pronouns could be gendered in some contexts. That got my attention, because I remembered this as an uncommon language feature, and I found that according to WALS, Japanese doesn't have it. So I'm wondering whether WALS is wrong on that point, whether I'm looking at the wrong ( Read more... )

pronouns, sociolinguistics, grammar, japanese, gender

Leave a comment

Comments 9

muckefuck April 14 2012, 23:32:30 UTC
They're talking about different types of genderedness. If you follow the link to the explanation of the WALS feature, you'll find this line:Gender, as defined by Corbett (1991), is a form of classification of nominals, as shown by agreement (see Map 30A).
Japanese lacks this sort of agreement, so by definition its pronouns cannot be possessed of "gender". But lexical gender is something else entirely. Even so, Japanese pronouns aren't exactly gendered in the same way that, say, nouns for "woman" or "stallion" are. As you say, it's not that ore cannot ever be used by a female speaker, it's just that the connotations of this word (being very vulgar and rough) clash rather spectacularly with Japanese norms of female behaviour. So, practically speaking, if someone uses ore, they're almost certain to be male. This is quite different from the situation in a language like Arabic, where anti is specifically feminine and correlates with feminine verb forms, feminine adjectives, and feminine noun phrases.

Reply

mintyfreshsocks April 14 2012, 23:40:33 UTC
This. "Gender" in Japanese first person pronouns is not about grammatical gender, but about gendered expectations regarding usage.

Reply

flamingophoenix April 23 2012, 21:46:55 UTC
Thanks both for the clarification! That's a great way of explaining it.

Reply


akibare April 15 2012, 01:01:29 UTC
What the others said, but also waga is pretty common in writing as part of set pieces like "wagakuni" (our country) "wagaya" (my house) etc. I wouldn't gender it but I suppose the feeling I get is some old guy, but that's more from the style of pieces it's often used in I think ( ... )

Reply

viata April 15 2012, 08:13:34 UTC
>but if the verbs actually have to agree with gender on "Jones said" what happens?
In case of Russian (where adjectives are to be marked all the time, and so are verbs in past tense) there are, to my knowledge, two options.
Firstly, you can try to avoid the "gendered" forms which is far from trivial but still possible.
Secondly, you can just use male forms and remember that they are default and so in some cases can refer to a female too. E.g. "тебя искал один человек/someone (one person) was looking for you". Один/one, человек/person and искал/was looking are grammatically male, but semantically it is irrelevant and the phrase can refer to a female just as easy.
Plus, btw, you can use the default male forms to mislead the characters of your story and the readers into believing someone is male and blame them later for making assumptions.

Reply

helle_d April 15 2012, 16:57:27 UTC
I've often wondered about this in relation to a children's book, 'The Turbulent Term of Tyke Tyler', where only at the end of the book is the reader told that the narrator, Tyke, is a girl - before that her gender was never mentioned, and you just assume from her brash, outgoing characteristics and male friendship group that she isn't.

Reply


ubykhlives April 15 2012, 06:20:40 UTC
The other thing I'd say to add to the other commenters is that WALS isn't absolutely trustworthy. Occasionally they will get facts wrong, since the authors of chapters are usually interested in specific features and so go trawling through texts and languages that they may not be hugely familiar with. Kabardian, for instance, is listed as a language that possesses the dental fricative /θ/, even though it doesn't; the source that they cite is actually one that uses the symbol θ to signify the affricate /ʦ/. Ubykh is also mistakenly identified as having Adjective-Noun word order and as not having zero copula for predicate nominals. So though the WALS dataset as a whole is a useful tool, the individual datapoints can't necessarily be taken uncritically.

Reply

pauamma April 15 2012, 10:05:40 UTC
I know. I had noticed that regarding WALS claims that Tagalog has grammatical gender.

Reply


tuuli_chan April 15 2012, 10:47:38 UTC
I don't know much about this, but here's some discussion on the subject at the Japanese community: http://japanese.livejournal.com/2222015.html

Reply


Leave a comment

Up