QUOTE (dvandorn @ Jun 22 2005, 09:55 PM)
I'm sorry, I guess I haven't stated my point of view about "irrationality" clearly.
I find the *requirement* of assigning a gender identity to each and every noun in a language an irrational approach to communication. There is simply no rationality that I can see to the requirement of assigning a male or female identity to a chair, a keyboard, a leg, a cloud, a window frame, etc., etc., etc., ad nauseum, when those objects do not have any intrinsic gender.
I think I explained it in excruciating detail in a previous post.
If you have gender, you can use pronouns a bit more often than if you don't have gender, in place of having to use lengthy noun phrases. Case closed -- there's a rational purpose for gender.
The mapping between objects and biological gender is a byproduct. Ignore that aspect. The important thing is that it is a mapping between WORDS and CLASSES of nouns. And to reap the stated benefit, the ideal is to have a random split down the middle, 50-50 in terms of number. It is not meant to convey a worldview. It is also not difficult for a first-language learner. It's only a noticeable effort if you try to rapidly pick up a second language.
If the mapping is arbitrary, so be it. It's rather like a hash function in computer science, except with only two values (it would be burdensome to have hundreds of noun classes). That doesn't keep the benefit from serving its purpose.
QUOTE (dvandorn @ Jun 22 2005, 09:55 PM)
The Romance languages require such a set of assignations
[...]
You cannot just say "the" and let it go at that.
[...]
In what manner does this add to the amount of information conveyed, or the quality of that information?
Simple. If you refer to two nouns in one sentence, and they happen to have different gender, then you can simply use a one-syllable pronoun the second time you want to refer to that thing. And 50% of the time, the two genders will be different BECAUSE it's pretty much random.
If you only have one gender, a pronoun will likely be ambiguous, and you'll have to devote more words to saying what you're trying to say.
For example, if in English, you said, "I saw the sun from the hill. It's beautiful.",
you fail to convey what "it" refers to, the sun or the hill. You need to use more words to get that across. In Spanish, you could say, "Vi el sol desde la colina. Es bello." Or you could end with "Es bella." And with zero extra content needed, you can select which one you mean.
I hope this puts to bed the notion of whether or not a function is served. Count the syllables. Gender lets you sometimes save some syllables. You can argue about how much good that does you, but it should be perfectly clear to all readers that it does, sometimes, save you syllables.
The other use is when it provides redundancy. In a noisy busport, an English speaker says, "The sand is on my leg." Because of the noise, you can't hear if they said "sand" or "hand". The whole meaning of the sentence is on that one phoneme. You miss it because a bus squeaks its brakes for 0.05 seconds, and you miss the whole sentence. A Spanish-style article can provide an extra bit to help you overcome the bus noise. Someone says, "Las pechas son grandes," and the same bus noise strikes during the end of "pechas". If that's all you had to go on, you wouldn't know if the person was talking about peaches or chests. But "Las" to the rescue! The fact that they said "las" and not "los" gives you the answer. In the same situation, English's wasteful one-gender-for-all "the" is useless. You go to all the trouble of saying one syllable, and you get NO disambiguation information that could have helped someone out if they missed another phoneme later. It's a senseless waste!
QUOTE (dvandorn @ Jun 22 2005, 09:55 PM)
If someone can explain to me in rational terms why it is in any way poetic *or* rational that a pencil *must* be referred to as a male object, while a chair *must* be referred to as a female object, I'll be happy to consider their arguments -- I generally have an open mind.
The point of the system isn't to place just the right tag on each object. First of all, it's not about objects at ALL. It's about words. If there are two different words that you could use to refer to the same object, then you could have both genders. For example, a female dog could be "el animal" (masc.) and "la perra" (fem.). It's not about the object, it's about the word.
Secondly (and as that point should make clear), the purpose of the system is not to have delivered an appropriate labeling of objects regarding boy-girl gender. It's not. The system works just fine being arbitrary.
Consider phone numbers. Maybe Einstein, Hitler, and Hugh Hefner had phone numbers that started with the same digit, while Pam Anderson, Charles Manson, and you have phone numbers that start with the same other digit. The point is that nobody tried to assign phone numbers to suit your personality or station in life. It was arbitrary. It works just fine arbitrary. It's not irrational.
The only design feature is required of a noun class system are that all speakers of the language can instantly tell what class a given noun is. In principle, this would involve memorizing one bit per noun, but it's simpler than that if common ending always map onto the same noun class (Spanish -dad is always feminine, -o is almost always masculine). Sometimes semantic classes pertain to the noun class.
The fact that boy-girl gender is the basis of most Indo-European systems is a quirk -- languages around the world have noun class systems, but few map them onto literal human gender. But, as I mentioned before, this has a useful purpose because then you INSTANTLY derive the same benefit in sentences referring to two people (if, ~50% of the time, the genders are different).
It would be like if all policemen had phone numbers that started with 2. It would help you remember a policeman's phone number. No matter if some non-policeman also had a number that started with 2.