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.)
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?
It certainly doesn’t seem like that’s the case for my daughter who’s in some sense about half-way along that progression.
How would you know if it would be the case? What do you think are the traits of the English language that prevents your daughter from learning it faster?
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?
Let’s take the example of the ‘bottle of soda’. Without looking at the particular case it seems for me hard to tell if there such a thing as an “empty bottle of soda” or whether bottle of soda means that the bottle is actually filled with soda.
That is not a problem if you regularly speak about bottle’s of soda that might not be a problem. There are empty soda bottles but no empty bottle’s of soda. At the same time if I search for empty bottle of soda in Google I get 37,900 results while I get 110,000 results for empty soda bottle.
Google Ngram is a bit stronger in favoring empty soda bottle but it still suggests that a sizable portion of people speak of empty bottle of soda.
In daily life you won’t have much problems with that. Context will often be enough. If you however take a biochemistry book and try to understand what it’s saying you often don’t have the context to know sense of a preposition is meant. That means you need to spend cognitive resources to think through the possibilities that could be meant.
In English the polysemy of or produces problems when people get into mathmatical logic.
The Polish language has polysemy whereby ręka means both hand and arm. Can you imagine how that makes live harder any subject that speaks about the body whether it’s biology or even massage?
The interesting thing is that the Polish culture had a lot of contact with languages that do have a proper word for hand that doesn’t also mean arm. Why didn’t they borrow Hand and Arm from German or English?
I suspect the reason is that ręka is too deeply imbedded in the Polish language. You can’t just burrow a new word like you can add a new word for ketchup when the concept enter into the language.
I would suspect that basic prepositions are similar in the fact that it’s very hard to borrow them from another language.
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?
As a learner it’s okay when there are a limited amount of prepositions. What you don’t want as a learner is special rules. Saying “in January” while saying “on Monday” is an unnecessary special rule.
You want to have reliable rules that tell you whether there’s a empty bottle of soda.
I don’t know whether I would. All I know is that I don’t recall ever seeing or hearing her have difficulty that seemed to relate in any way to such things. Perhaps I wouldn’t expect to have done; I’m not claiming this as strong evidence; as I say, I think we’re both basically guessing.
The rest of what you say still seems to me to be guessing that there “ought” to be a problem. I agree that you’ve presented some examples of ambiguity, and when something can mean X or Y and you want to say specifically X or specifically Y that can make your life more difficult. But when something can mean X or Y and you want to say “X or Y” the ambiguity is positively helpful (your hand really is part of your arm, and I bet there are cases where ręka is strictly better than either English word). And needing to learn fewer words is nice. And using less language-space means better robustness against errors. And ambiguity is often beneficial in poetry. Etc., etc., etc. Maybe these instances of ambiguity really do make English (and Polish) worse languages than they would be if they were patched up. But I don’t think you’re in a position to say that they do just on the basis that they are instances of ambiguity.
You would know because she would make errors. Kids don’t stop talking because they don’t know the right word. You know that irregular past tense is hard(er) for kids to learn because they say things like ‘runned’ and ‘eated’. In some cases it might be a circumlocution rather than an error (‘I’m the big one’ in place of ‘I’m the biggest’), but it’s not hard to know what a kid is having trouble learning if you are paying attention.
I haven’t noticed her making errors that obviously relate to polysemy like that of “post”. But I’m not sure what such errors would look like; my best guess is that if such things are a problem their main impact is probably just extra cognitive load, hence slower learning generally.
Slower processing evidenced by slower response or stated confusion would be the most likely result. Polysemy is actually good for helping kids practice using context cues, so it is arguable that even if she was making errors, it would still be a good thing from an educational standpoint.
It is worth noting that when linguists complain about how bad English is, they do not complain about polysemy, but about deep orthography and some nasty grammatical rules. It is also worth noting that English is pretty much average as far as polysemy goes, at least in European languages—although this may depend on whether you consider idiomatic phrasal verbs to be polysemies of the composite words.
Perhaps I wouldn’t expect to have done; I’m not claiming this as strong evidence; as I say, I think we’re both basically guessing.
I think it’s fairly straightfoward to say that something that the child has some problems with the language that result in her not yet having college level language skills. If your position would be that you know what those problems are and those problems have nothing to do with polysemy, I would grant you that points in the direction that polysemy isn’t a huge deal.
The rest of what you say still seems to me to be guessing that there “ought” to be a problem.
When it comes to the case of or (/the German oder that works the same way) I know that I had classmates who struggled to wrap their heads around the fact that the mathematical or means something different than the phrase as it’s commonly used.
I think or is actually a big deal. A lot of other words aren’t a big deal but there are a lot of words the effects add up.
And ambiguity is often beneficial in poetry.
I don’t think that vagueness is the point of poetry. Having different prepositions allows a poet to create an effect by using a prepositions in a way that it isn’t usually used to communicate a new meaning.
Has as poetry goes I would add that the relationship system I proposed is quite yielding. If you take an English word like lover you don’t automatically get a word to describe the equivalent of the sibling relationship. Playing around with the equivalent of graph theory terms like graph, branch and root could also be fun for poets. The poet get’s those language tool for every relationship.
In the same way that the word lover would be an extension of a relationship/graph term + the emotion of love, any other emotion could also be used. Huge spaces for poetry open up because there a solid foundation.
the child has some problems with the language that result in her not yet having college level language skills.
Well, sure. She’s nine years old. Her vocabulary isn’t college-level yet. Her writing style is pretty boring (by adult standards; for a nine-year-old she’s doing just fine) and the best ways she currently has of mitigating that are mechanical and artificial (start sentences with one of this arbitrary list of More Interesting Ways To Start A Sentence, etc.). Her sense of the sound and rhythm of language isn’t well developed (again, by adult standards). She makes spelling mistakes sometimes.
In short, she’s a pretty typical bright nine-year-old. I have no way of telling whether she’d be further ahead in all those things if only she hadn’t secretly spent weeks puzzling over the different meanings of “post”; all I can say is that I don’t see any sign that that sort of thing is holding her up. But that’s extremely weak evidence because, of course, I’ve no good reason to think there would be obvious signs if it were.
the mathematical or
I can very easily believe that speakers of German and English commonly have trouble with the mathematical use of “or” on account of the tendency for or/oder to be “exclusive”. But, once again, that doesn’t mean it’s a bad thing overall that it’s that way. Perhaps an exclusive “or” is actually more useful most of the time. After all, most of us most of the time are doing things other than mathematics. If I tell my daughter “You may have a slice of cake or a biscuit” I will likely be annoyed if she takes both[1]. I can get to work by driving or by cycling, but the results will not be good if I try to do both at once.
[1] But once I was with a friend who’s a computer science researcher, and when I offered him some chocolate and he said “I’ll have one or two pieces” I gave him three.
I don’t think that vagueness is the point of poetry.
I said ambiguity, not vagueness; the two are not the same. And I didn’t say the point, which would certainly be far too strong.
After all, most of us most of the time are doing things other than mathematics. If I tell my daughter “You may have a slice of cake or a biscuit” I will likely be annoyed if she takes both[1].
The fact that you would get annoyed if she takes both suggests that it would be very useful to have a word that actually means that she can’t have both. I don’t think that the notion of there being an inclusive or and an exclusive or is inherently hard. If there would be one word for the exclusive or and one word for the inclusive or a child would learn naturally that both notions exist. With polysemy that’s not something that a child learns automatically.
Finnish even seems to have three or’s. The third one is equivalence.
While we are at it iff/if and only if probably should also be expressible in a single word.
I said ambiguity, not vagueness; the two are not the same.
If you increase the amount of prepositions than it’s likely that some of them will sound ambiguous when used in nonstandard usage. Preventing people from making ambigious statements isn’t my goal. I think language should allow people to be specific, not that it should force people to be specific.
I don’t think poets will lose the ability to be ambigious through the proposals that I’m making.
We can express this in two words (either/or) already. How do you avoid the trap of trying to optimize easy to measure things (like number of words) at the cost of harder to measure things?
We can express this in two words (either/or) already.
The problem is that while there a way to specify the exclusive or, there not a way to specify the inclusive or.
How do you avoid the trap of trying to optimize easy to measure things (like number of words) at the cost of harder to measure things?
I’m not optimizing for number of words. The problem of English isn’t that there aren’t enough words. It’s believed good style to know your thesaurus in English and not say four times beautiful in a row in a single paragraph, even if you mean the same thing. That produces a proliferation of words but not the kind of words I want.
When I talk about polysemy I care about the ability to make finer distinctions for commonly used words, where the listener can actually know which distinction the speaker wants to communicate
Is there really any practical purpose of this discussion? How are you proposing to impose changes on an actually existing language, other than by making laws about language use and penalizing people with heavy fines if they make mistakes?
Is there really any practical purpose of this discussion?
Given that we have discussion on LW about very theoretical issues, it’s interesting that you choose to ask this question in this context.
There likely some triggered ugg-field. Tribal political instincts that don’t belong.
There are multiple practical aspects. It’s very useful to understand that language is not perfect for rationality in general.
Secondly KevinGrant opened this discussion because he designs a Conlang. In the context of Conlang design it’s important to openly speak about the flaws of the existing languages.
How are you proposing to impose changes on an actually existing language, other than by making laws about language use and penalizing people with heavy fines if they make mistakes?
Given that both the German government and the French government actually do form their respective languages through government programs, the idea that you need to fine people is without basis.
Just because the US still works with inches and feets doesn’t mean that change is impossible if the will exists.
But if I wanted to lead the road to reformed English, I wouldn’t do it via the government. I would change the language in a way where it’s possible to automatically translate the new reformed English into contemporary English.
I would provide wordpress plugins that allow a person to write his post in reformed English and at the same time show a normal English version on his website.
If done with charismatic leadership, using reformed English becomes the thing that cool people do.
On of the aspects of reformed English would be that it has less polysemy. That means it’s possible to do better translations of reformed English into other languages. The UN switches from contemporary English to reformed English. India justifies making reformed English it’s primary language because it’s easier possible to translate it into it’s 22 other official languages. Computer translation will be better at that point, but computer translation profits a lot if there’s less polysemy.
The EU also would profit from making reformed English it’s main language of business. Politically it might be easier to declare reformed English to be the main EU language than to declare contempory English the main EU language.
If the US wants to declare reformed English to be the main language it can simply choose to conduct it’s government business in it and make standardized tests in reformed English.
Scientists in the humanities might change their journals in a way where new articles have to be submitted in reformed English and there’s a translation in normal English available. Less polysemy might make some debates clearer.
Besides reducing polysemy one of the main tasks of reforming English would be to make it more phonetic.
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?
It probably isn’t if you only count spoken use, but it probably is if you also count written use.
How would you know if it would be the case? What do you think are the traits of the English language that prevents your daughter from learning it faster?
Let’s take the example of the ‘bottle of soda’. Without looking at the particular case it seems for me hard to tell if there such a thing as an “empty bottle of soda” or whether bottle of soda means that the bottle is actually filled with soda.
That is not a problem if you regularly speak about bottle’s of soda that might not be a problem. There are empty soda bottles but no empty bottle’s of soda. At the same time if I search for
empty bottle of sodain Google I get 37,900 results while I get 110,000 results forempty soda bottle. Google Ngram is a bit stronger in favoringempty soda bottlebut it still suggests that a sizable portion of people speak ofempty bottle of soda.In daily life you won’t have much problems with that. Context will often be enough. If you however take a biochemistry book and try to understand what it’s saying you often don’t have the context to know sense of a preposition is meant. That means you need to spend cognitive resources to think through the possibilities that could be meant.
In English the polysemy of
orproduces problems when people get into mathmatical logic.The Polish language has polysemy whereby
rękameans bothhandandarm. Can you imagine how that makes live harder any subject that speaks about the body whether it’s biology or even massage?The interesting thing is that the Polish culture had a lot of contact with languages that do have a proper word for hand that doesn’t also mean arm. Why didn’t they borrow
HandandArmfrom German or English? I suspect the reason is thatrękais too deeply imbedded in the Polish language. You can’t just burrow a new word like you can add a new word forketchupwhen the concept enter into the language. I would suspect that basic prepositions are similar in the fact that it’s very hard to borrow them from another language.As a learner it’s okay when there are a limited amount of prepositions. What you don’t want as a learner is special rules. Saying “in January” while saying “on Monday” is an unnecessary special rule. You want to have reliable rules that tell you whether there’s a
empty bottle of soda.I don’t know whether I would. All I know is that I don’t recall ever seeing or hearing her have difficulty that seemed to relate in any way to such things. Perhaps I wouldn’t expect to have done; I’m not claiming this as strong evidence; as I say, I think we’re both basically guessing.
The rest of what you say still seems to me to be guessing that there “ought” to be a problem. I agree that you’ve presented some examples of ambiguity, and when something can mean X or Y and you want to say specifically X or specifically Y that can make your life more difficult. But when something can mean X or Y and you want to say “X or Y” the ambiguity is positively helpful (your hand really is part of your arm, and I bet there are cases where ręka is strictly better than either English word). And needing to learn fewer words is nice. And using less language-space means better robustness against errors. And ambiguity is often beneficial in poetry. Etc., etc., etc. Maybe these instances of ambiguity really do make English (and Polish) worse languages than they would be if they were patched up. But I don’t think you’re in a position to say that they do just on the basis that they are instances of ambiguity.
Sidenote:
You would know because she would make errors. Kids don’t stop talking because they don’t know the right word. You know that irregular past tense is hard(er) for kids to learn because they say things like ‘runned’ and ‘eated’. In some cases it might be a circumlocution rather than an error (‘I’m the big one’ in place of ‘I’m the biggest’), but it’s not hard to know what a kid is having trouble learning if you are paying attention.
I haven’t noticed her making errors that obviously relate to polysemy like that of “post”. But I’m not sure what such errors would look like; my best guess is that if such things are a problem their main impact is probably just extra cognitive load, hence slower learning generally.
Slower processing evidenced by slower response or stated confusion would be the most likely result. Polysemy is actually good for helping kids practice using context cues, so it is arguable that even if she was making errors, it would still be a good thing from an educational standpoint.
It is worth noting that when linguists complain about how bad English is, they do not complain about polysemy, but about deep orthography and some nasty grammatical rules. It is also worth noting that English is pretty much average as far as polysemy goes, at least in European languages—although this may depend on whether you consider idiomatic phrasal verbs to be polysemies of the composite words.
I think it’s fairly straightfoward to say that something that the child has some problems with the language that result in her not yet having college level language skills. If your position would be that you know what those problems are and those problems have nothing to do with polysemy, I would grant you that points in the direction that polysemy isn’t a huge deal.
When it comes to the case of
or(/the Germanoderthat works the same way) I know that I had classmates who struggled to wrap their heads around the fact that the mathematicalormeans something different than the phrase as it’s commonly used.I think
oris actually a big deal. A lot of other words aren’t a big deal but there are a lot of words the effects add up.I don’t think that vagueness is the point of poetry. Having different prepositions allows a poet to create an effect by using a prepositions in a way that it isn’t usually used to communicate a new meaning.
Has as poetry goes I would add that the relationship system I proposed is quite yielding. If you take an English word like
loveryou don’t automatically get a word to describe the equivalent of thesiblingrelationship. Playing around with the equivalent of graph theory terms likegraph,branchandrootcould also be fun for poets. The poet get’s those language tool for every relationship. In the same way that the wordloverwould be an extension of a relationship/graph term + the emotion of love, any other emotion could also be used. Huge spaces for poetry open up because there a solid foundation.Well, sure. She’s nine years old. Her vocabulary isn’t college-level yet. Her writing style is pretty boring (by adult standards; for a nine-year-old she’s doing just fine) and the best ways she currently has of mitigating that are mechanical and artificial (start sentences with one of this arbitrary list of More Interesting Ways To Start A Sentence, etc.). Her sense of the sound and rhythm of language isn’t well developed (again, by adult standards). She makes spelling mistakes sometimes.
In short, she’s a pretty typical bright nine-year-old. I have no way of telling whether she’d be further ahead in all those things if only she hadn’t secretly spent weeks puzzling over the different meanings of “post”; all I can say is that I don’t see any sign that that sort of thing is holding her up. But that’s extremely weak evidence because, of course, I’ve no good reason to think there would be obvious signs if it were.
I can very easily believe that speakers of German and English commonly have trouble with the mathematical use of “or” on account of the tendency for or/oder to be “exclusive”. But, once again, that doesn’t mean it’s a bad thing overall that it’s that way. Perhaps an exclusive “or” is actually more useful most of the time. After all, most of us most of the time are doing things other than mathematics. If I tell my daughter “You may have a slice of cake or a biscuit” I will likely be annoyed if she takes both[1]. I can get to work by driving or by cycling, but the results will not be good if I try to do both at once.
[1] But once I was with a friend who’s a computer science researcher, and when I offered him some chocolate and he said “I’ll have one or two pieces” I gave him three.
I said ambiguity, not vagueness; the two are not the same. And I didn’t say the point, which would certainly be far too strong.
The idea that ambiguity is important in poetry is not an idiosyncrasy of mine, by the way.
I agree that a language with very different structure might open up very different opportunities for poets.
The fact that you would get annoyed if she takes both suggests that it would be very useful to have a word that actually means that she can’t have both. I don’t think that the notion of there being an inclusive
orand an exclusiveoris inherently hard. If there would be one word for the exclusiveorand one word for the inclusiveora child would learn naturally that both notions exist. With polysemy that’s not something that a child learns automatically.Finnish even seems to have three
or’s. The third one is equivalence.While we are at it
iff/if and only ifprobably should also be expressible in a single word.If you increase the amount of prepositions than it’s likely that some of them will sound ambiguous when used in nonstandard usage. Preventing people from making ambigious statements isn’t my goal. I think language should allow people to be specific, not that it should force people to be specific.
I don’t think poets will lose the ability to be ambigious through the proposals that I’m making.
We can express this in two words (either/or) already. How do you avoid the trap of trying to optimize easy to measure things (like number of words) at the cost of harder to measure things?
The problem is that while there a way to specify the exclusive
or, there not a way to specify the inclusiveor.I’m not optimizing for number of words. The problem of English isn’t that there aren’t enough words. It’s believed good style to know your thesaurus in English and not say four times beautiful in a row in a single paragraph, even if you mean the same thing. That produces a proliferation of words but not the kind of words I want. When I talk about polysemy I care about the ability to make finer distinctions for commonly used words, where the listener can actually know which distinction the speaker wants to communicate
“X or Y or both”
Also “and/or”, although that works better in writing.
“any of” or “at least one of”, You can say it, it’s just not one word.
Is there really any practical purpose of this discussion? How are you proposing to impose changes on an actually existing language, other than by making laws about language use and penalizing people with heavy fines if they make mistakes?
Given that we have discussion on LW about very theoretical issues, it’s interesting that you choose to ask this question in this context. There likely some triggered ugg-field. Tribal political instincts that don’t belong.
There are multiple practical aspects. It’s very useful to understand that language is not perfect for rationality in general. Secondly KevinGrant opened this discussion because he designs a Conlang. In the context of Conlang design it’s important to openly speak about the flaws of the existing languages.
Given that both the German government and the French government actually do form their respective languages through government programs, the idea that you need to fine people is without basis.
Just because the US still works with inches and feets doesn’t mean that change is impossible if the will exists.
But if I wanted to lead the road to reformed English, I wouldn’t do it via the government. I would change the language in a way where it’s possible to automatically translate the new reformed English into contemporary English. I would provide wordpress plugins that allow a person to write his post in reformed English and at the same time show a normal English version on his website. If done with charismatic leadership, using reformed English becomes the thing that cool people do.
On of the aspects of reformed English would be that it has less polysemy. That means it’s possible to do better translations of reformed English into other languages. The UN switches from contemporary English to reformed English. India justifies making reformed English it’s primary language because it’s easier possible to translate it into it’s 22 other official languages. Computer translation will be better at that point, but computer translation profits a lot if there’s less polysemy.
The EU also would profit from making reformed English it’s main language of business. Politically it might be easier to declare reformed English to be the main EU language than to declare contempory English the main EU language.
If the US wants to declare reformed English to be the main language it can simply choose to conduct it’s government business in it and make standardized tests in reformed English.
Scientists in the humanities might change their journals in a way where new articles have to be submitted in reformed English and there’s a translation in normal English available. Less polysemy might make some debates clearer.
Besides reducing polysemy one of the main tasks of reforming English would be to make it more phonetic.