Since my last article about the alternant nouns of Tocharian I've been thinking about the development of the feminine in other Indo-European languages.
The feminine gender will always be a rather mysterious thing. There's absolutely no reason why a language would have it; but still there's quite a couple of languages that do develop one. Most Indo-European and Semitic languages are the most well known ones.
But to figure out the need for the feminine; let's turn things around, and take a language that doesn't have a feminine and let it develop one.
Indo-European at one point did not have a feminine; sometimes you see it developed both a word for a female and a male animal (Like Bull and Cow), but more often than not; it just didn't distinguish at all.
Later when the Indo-Europeans were getting more settled; breeding of animals undoubtedly became more important. So let's say they wanted to breed wolves `*ulkwons' (Root is *ulkwo- an o-stem.), there was no such thing as a feminine suffix; because they simply never needed one due to a lack of grammatical gender. But they needed `something' to show that it was feminine, so they chose the suffix *-ih2 for o-stems and *-h2 for consonant stems.
These suffixes can be seen as abstractifying suffixes. And then *ulkwih2 should actually mean `wolfness', but due to the lack of other means to make a feminine, this word took on the meaning of a `female wolf'.
So there it is! That's how feminine nouns came to be! The bigger question is though; why did adjectives start to agree with the nouns?
One of the suggested possibilities is `the wanted it to rhyme'.
This works up to a certain extent; for example if you take o-stems in classical Greek, you'll definitely see extensive noun phrases with constantly the same ending, `this work' in Greek in the genitive case would for example be:
τουτου του εργου.
Even if you're not able to read Greek you'll see that the ου gets repeated a grand total of 4 times. People suggest that they wanted to achieve this effect with feminine words as well. And thus started suffixing the `feminine' suffixes to the adjectives.
This would sort of work; if it wasn't for other suffixes which end in completely different things. like the *-r/n stems, it's not like adjectives suddenly agree to the *-r/n suffix to completely the rime. And technically there's no difference between *-r/n stems and *-ih2 stems.
So why this feminine developed; we're not sure. One thing is certain though; it developed after Hittite split off from the other languages, since there's absolutely no sign at all of a feminine in the Hittite language. Also Tocharian; although it has a feminine; shows quite a different view on it's development than other languages.
As soon as I come up with any ideas on the development of a femnine that are better than `they felt like it' I'll be sure to write about it.