= 27d567bc2d48d9f40e9ccc20ea370b15)
Daniel_Burfoot comments on Linguistic mechanisms for less wrong cognition - Less Wrong
You are viewing a comment permalink. View the original post to see all comments and the full post content.
= 783df68a0f980790206b9ea87794c5b6)
Loading…= b715d02e85f7b3b4ae65ca3d3e450868)
Subscribe to RSS Feed
Powered by Reddit
= f037147d6e6c911a85753b9abdedda8d)
You are viewing a comment permalink. View the original post to see all comments and the full post content.
Comments (130)
My biggest gripe about English is that there is no consistent relationship between morphology and part of speech. There is a muddy, approximate relationship which is inherited from French/Latin and German, so that for example you typically know that if you see an adjective X, and see a word Xity, then the latter word is a noun meaning "property of being X". Similarly, if you see an adjective Y, and another word Yen, the latter word is a verb meaning "to make Y". But this system is not used consistently. Ideally, a listener (reader) should be able to identify the part of speech of a word immediately by inspection of phonological (typographic) expression.
If you want to follow this rule, you will need to make it easy for people to do the sorts of colloquial grammar-jumping that come up in everyday speech. For example the word "hammer" is a noun but also a verb meaning "to hit with a hammer". "Ship" is a noun but also a verb meaning "to send by ship" and so on.
Another issue with English (and probably other languages) is that prepositions are overloaded, so that the same word can mean different things, as in "Galileo saw a man with a telescope". Since with can mean both by means of and carrying/holding, the sentence is ambiguous. It doesn't seem unreasonable to ask that every important case in which a preposition must be used should correspond to a distinct word.
I'd say this probably happens much more often than native English speakers even realize. You mostly notice it as a beginner, if your native language does not have strict rules for word order (because you differentiate different types of words by their suffixes or something like that) so you can move the words freely around in your first language. And then you read some English sentence where every single word has two or more possible meanings, and it's like a puzzle with an exponential difficulty. And you feel like: WTF, how can these people even get any meaning across in such ambiguous language?
Then you learn the rule that when you see "X Y Z", then "Z" is probably a noun, and "X" and "Y" are adjectives that modify it (unless "Y" is a verb, which means that the noun "X" is doing the "Y"-action to the noun "Z"), which reduces the search space. And after enough practice your brain starts doing it automatically, so the ambiguity becomes invisible.
All good points, and among the strengths of conlangs in general. It still amazes me that past efforts at reforming English spelling, like President Roosevelt's, weren't accepted.
I'm not so sure.
Knowing a word's part of speech is of limited use if you don't know its actual meaning. Learning a word's meaning generally tells you its part of speech too. If by some chance you have an idea of a word's meaning but not its part of speech (because of ambiguities as with "ship", or because you worked out what kind of thing it has to mean etymologically), that's often enough to work out what's going on anyway. What's the real benefit here of making the part of speech more visible? It sounds nice, but when does it actually help much?
I'm not so sure.
Inside view: there are really quite a lot of preposition-functions, and prepositions want to be short words, so if we insist on a separate preposition for every preposition-function we'll need to allocate a lot of short words for them. Short words are a scarce resource. The language will have to be clumsier in other ways.
Outside view: every language I know enough about (admittedly a small subset of the world's languages) overloads its prepositions. That's got to be some evidence that doing so isn't a terrible idea.
I think it's evidence that it's not easy for prepositions to get added through natural language evolution. It much easier to add new verbs, adjectives and nouns.
While that's true when it comes to conlang design, if you look at English there's plenty of open space of short words. A lot of two letter combinations that are possible with English phonetics aren't valid English words.
That doesn't necessarily mean there's spare space. You don't want every possible combination of letters to make a word, because then it becomes easier to mishear.
Someone in this thread mentioned that there are 37 different meanings in English for
post. It's easy to mishear between those 37 meanings. You could easily move a third of those topistand another third topust. That would make it easier to get the right meaning.To the extend that context allows you to choose the right of the 37 meanings of post, it should also help you prevent mishearing.
If you take the preposition of
withwith it's nine different meanings, move a third towuthand a third towoth. People might make a mistake to mishearwithwhen the other person sayswuthbut at least they have a change to hear the right meaning and don't have to guess based on context which of the many meanings is meant.This basically cannot happen in real life, because most people do not think clearly about which sense of a preposition they are using. So if you divide up those meanings of "with", all three words will start to take on all nine meanings, and you will just have uselessly multiplied words.
The fact that people don't reflect about the sense in which they use a preposition doesn't mean that they can't learn to use a specific preposition for a specific purpose. In reality people can say "on Monday" while saying "in July" and "at night" without getting confused.
If you have the sentence "Galileo saw a man with a telescope" people do mentally distinguish two cases of with that could be meant. There nothing natural about all the meanings that "with" has in English being bundled together via the same word. Other languages bundle things together in different ways.
There's a very old and very silly debate in Spanish because some people refuse to acknowledge that "a glass of water" means what it intends to mean, instead of the ridiculously literal "a glass made of water", so they switch to the awkward "a glass with water", which in real life can mean a glass on a tray with a jar of water next to it.
So the result is that snobbish people insist on saying "a glass with water," and ordinary people plus meta-snobbish people keep saying "a glass of water", and both sides hate each other passionately.
So is it, basically, a status signal by now?
Yes, but in a complicated way. "A glass with water" is hypercorrection, which gives the speaker the opposite status from the one he believes he's displaying.
In that case it seems that a short preposition for "containing" is missing.
Language isn't easy. If you just know the rules, it's hard to know that a teacup might not contain tea while a cup of tea does. It get's even more confusing because the same object that's a teacup when it's intended to store tea liquids suddenly becomes a bowl when it's intended to contain soup.
Nope, it does not. Teacups have handles and bowls don't.
Strangely, the same people who object to "a glass of water" have no problem with "a bottle of soda," "a pot of potatoes" or "a truck of pigs".
English sometimes relies too much on context to provide clues for meaning. The word "post" has 37 meanings as a noun, verb, or adverb. Poor context can't shoulder all the load.
Since English is an, ahem, successful language, it clearly can :-P
True. Let me qualify: for the benefit of the student of languages, context shouldn't shoulder all the load.
Should languages be designed for language students?
For natural languages it's a moot question, but conlangs are inescapably intended for the use of people who are already inclined to study languages.
Indeed they are, but the more serious kind of conlang is surely intended to be usable as an actual practical means of expression and communication. If some design decision makes things better for students one way and for actual users another way, it's surely better to choose the latter.
(Of course we don't know that the present situation is like that. It's entirely possible that the success of English hasn't been in any way helped by its heavy use of context for disambiguation, or by advantages that that somehow enables.)
The success of English, you ask?
(cough) British Empire (cough)
Even in English a person who has a 50,000 word active vocabulary can express himself better than person who has a 10,000 word active vocabulary.
True, but that's mostly a matter of having more things they can say in one word. Reducing ambiguity in a language by splitting the job of one word up among multiple words with fewer meanings increases the language's vocabulary size but doesn't increase the range of things there are words for. So the two aren't parallel.
Focusing on the numbers of words might miss my point. The average person who finishes speaks English on a higher level than the average person at high school. It takes effort to learn college level English.
If you make the language easier to learn than it will take less effort to learn college level English. People will reach the same level of proficiency in the language at an ealier age.
I am not convinced that a nontrivial fraction of the effort it takes a native anglophone to get from zero to college level English is caused by polysemies like that of "post". It certainly doesn't seem like that's the case for my daughter who's in some sense about half-way along that progression. Such things are (I think) more of an obstacle to people learning English as a foreign language. I am all in favour of making the lives of foreign language learners easier, but generally most people who speak a language speak it natively (English might actually be a counterexample, now I come to think of it) and, even more so, most use of a language is by native speakers (I bet English isn't a counterexample to that). So I think that in evaluating languages we should be considering how effective they are in actual use much more than how easy they are to learn for foreign learners.
Now, for sure, I have no very good reason to think that making prepositions less polysemic wouldn't be an improvement in actual use. But then I don't think you have any very good reason to think it would be an improvement for learners, either; it's just a guess, right?