Re: Indo-European Languages and Gramatical Gender Loss

On 15 Jun, 00:55, LEE Sau Dan <dan...@xxxxxxxxxxxxxxxxxxxxxxxxxx>
wrote:
"Seán" == Seán O'Leathlóbhair <jwlaw...@xxxxxxxxx> writes:

Seán> Since my L1 does not have gender, I am biased but I find it
Seán> hard to see that this is useful in two gender languages such
Seán> as French. English actually has more third person singular
Seán> pronouns that French. In a few cases, French will be able
Seán> to make a distinction that English could not but the reverse
Seán> would also apply. The extra pronoun of English would
Seán> suggest that English would win more often.

No. This simplistic conclusion doesn't take *distribution* into
account. Having more doesn't mean being useful in more situations.
e.g. you own 10 shirts, but you tend to wear shirt #1 70% of the time,
shirt #2 20% of the time and shirt #3 9% of the time. Only in the
remaining 1% of the time do you wear one of shirts #4--10. Then, does
it really matter _that much_ that you own 10 shirts instead of 5?

I only said "suggest" not "prove". I would admit that more data be
needed for "prove" but I believe this is enough for "suggest". It is
enough to justify gathering the necessary data. It is also already
enough to show that English can handle three third person singular
antecedents versus the two of French.

Seán> If there were considerably more genders then the feature
Seán> could be useful but it still would not strike me as the best
Seán> solution to the problem

The feature would be useful only if every gender has more or less the
same frequency of occurrence. If most are not used (like shirts
#4--10 above), then they won't help _much_.

Seán> If you are referring to two objects in French, there is only
Seán> (approximately) a 50/50 chance that they are of different
Seán> gender.

Why 50/50? Have you counted a typical French text and verified that
the distribution is really 50/50?

I picked 50/50 to be fair to French since that is the best case for
it. Note also the "approximately" as an admission that I was not
claiming precision. As the split moves away from 50/50, the chance of
two randomly selected nouns have different gender decreases though not
very rapidly until the split is quite far from 50/50. If masculine
out numbered feminine by 2 to 1 then the probably of different gender
drops to 4/9 and at 3 to 1 it drops to 5/8.

Seán> So, in half of the cases there is no advantage over
Seán> English. At least in English, you are not going to confuse
Seán> a person and a thing.

This conclusion is based on the _ungrounded_ assumption that every
gender has more or less the same frequency of occurrence.

I don't think that I have claimed to have concluded anything. I am
just opening a discussion. However your frequency points don't
detract from what I am saying. My 50/50 assumption maximises the
utility of grammatical gender. So, if it does not have much value
under that assumption, it will have even less if the distribution is
not 50/50.

--
Seán O'Leathlóbhair

.

Relevant Pages

  • Re: Fast UTF-8 strlen function
    ... >>> English also has 3 genders, ... Which one you use depends on the 'gender' of the ... In fact, generally speaking, English is the "oddball" language here...it's ... and "Shakespearean" is to partially "write German using English ...
    (alt.lang.asm)
  • Re: McMillan and Partner
    ... I think the problem arises _because_ English doesn't have much gender. ... But even in French there are masculines which can be applied to women for particular purposes: "maître" in a couple of special senses comes to mind. ... The example that I remember encountering early on in my acquaintance with the French language is the dedication of "Le Petit Prince", where the male dedicatee is refered to as "une grande personne" at the beginning of the first paragraph, and thereafter as "elle" until he is referred to by name again, at which point he becomes "il". ...
    (alt.usage.english)
  • Re: "Gender-free masculine pronoun"? Absurd.
    ... In the sequel, I use the terms "gender, masculine, feminine, neuter" ... only for grammar and the terms "sex, male, female" only for biology. ... I my native language, cognate of English, we have exactly the same ...
    (sci.lang)
  • Re: Indo-European Languages and Gramatical Gender Loss
    ... Seán> hard to see that this is useful in two gender languages such ... English actually has more third person singular ... Seán> pronouns that French. ... it really matter _that much_ that you own 10 shirts instead of 5? ...
    (sci.lang)
  • Re: He She It They
    ... speakers ought to develop an acceptable set of gender neutral ... speakers of English the world over that such changes are needed, ... Language, as you point out, is fluid. ... the bias towards the male gender in the English language. ...
    (soc.religion.quaker)