"bring me the red dog ball"
August 26, 2019 12:24 PM Subscribe
Which language has the least ambiguity?
In what languages do word order not matter? Which language's sentences can be put through a juicer and come out the other end crystal clear? Is it because there's declension of nouns or adjectives? Is there one where word order is the most important part of construction and where the results leave no gray areas? Which ones leave us no confusion between "we" and "we" or "you" versus "y'all"? By gum I want to know if it's the dog that's red or if it's the ball that's red. I want to know the language in which you can't even begin to mess up phrases such as "not all dogs are brown" and "all dogs are not brown".
If there's a distinction to be made between written and spoken I'm all for learning about that too.
In what languages do word order not matter? Which language's sentences can be put through a juicer and come out the other end crystal clear? Is it because there's declension of nouns or adjectives? Is there one where word order is the most important part of construction and where the results leave no gray areas? Which ones leave us no confusion between "we" and "we" or "you" versus "y'all"? By gum I want to know if it's the dog that's red or if it's the ball that's red. I want to know the language in which you can't even begin to mess up phrases such as "not all dogs are brown" and "all dogs are not brown".
If there's a distinction to be made between written and spoken I'm all for learning about that too.
Lojban?
I guess that raises a call for clarification: natural languages only, or invented ones too?
posted by thelonius at 12:32 PM on August 26, 2019
I guess that raises a call for clarification: natural languages only, or invented ones too?
posted by thelonius at 12:32 PM on August 26, 2019
No such thing. I'm not sure, though, why you seem to associate a completely flexible word order with a lack of ambiguity. Word order is just another way to specify information. Many people find a language with a substantially more flexible word order like Attic Greek to be more difficult than one with a more fixed order.
posted by praemunire at 12:36 PM on August 26, 2019 [5 favorites]
posted by praemunire at 12:36 PM on August 26, 2019 [5 favorites]
In Polish you can jumble words in almost any order because they decline, are different genders and generally it's very easy to determine the meaning. Just keep the prepositions and main negations together with the words they pertain to.
posted by I claim sanctuary at 12:39 PM on August 26, 2019 [1 favorite]
posted by I claim sanctuary at 12:39 PM on August 26, 2019 [1 favorite]
Flexible word order doesn't necessarily indicate a lack of ambiguity, but if that's all you want, you're best off with a language that uses grammatical case. Try Latin, Polish, or German, you can move things around there quite a bit and the declensions will tell you what is happening. But you can still come up with a confusing sentence in nearly any language.
posted by epanalepsis at 12:57 PM on August 26, 2019 [4 favorites]
posted by epanalepsis at 12:57 PM on August 26, 2019 [4 favorites]
You're looking for what philosophers of language call an "ideal language." Try searching on that term. I think you'll soon be convinced that your unicorn does not exist...
posted by bricoleur at 12:57 PM on August 26, 2019 [1 favorite]
posted by bricoleur at 12:57 PM on August 26, 2019 [1 favorite]
Best answer: (I've got a Ph.D. in linguistics.) All natural languages have ambiguity, though it might show up in different areas. For instance, some languages make it easy to leave the subject of a verb unstated, and other languages don't — but those other languages make up the difference by letting you be ambiguous in other ways.
If someone found a language that was less ambiguous than all the others, and quantified that in a way that the rest of the field found convincing, they would be instantly famous. It would be a huge, huge discovery. So far, nobody has done that. That doesn't mean such a language isn't out there. But I think most linguists would be tremendously surprised if one was discovered.
To answer your specific question about word order: you're right that languages that leave word order flexible generally give you some other way to disambiguate a sentence. Maybe there's a special ending that goes on the subject of the sentence to distinguish it from the object. Maybe the verb agrees with its subject, or its object, or both. Maybe a subject-verb-object sentence and an object-verb-subject sentence are pronounced with different melodies, or with the pauses in different places. Those things all cut down on the ambiguity of an average sentence, compensating for the free word order and bringing it back to the same level of ambiguity we have in English. But they aren't perfect, and it's certainly still just as easy to create ambiguous sentences in these languages as it is in English.
To answer your specific question about "we" and "we": yeah, some languages distinguish "we including you" and "we not including you." But of course, in languages that don't have a distinction like that built in, you can still make it explicit by saying "We — you too, komara — have to do this." And in languages that do have the distinction built in, there are other ambiguities: "we-not-including-you" is still ambiguous between "me and my wife," "me and my brother," "me and my wife and my brother," etc. So the upshot is that having this feature doesn't make a language overall less ambiguous than English, it just moves the ambiguity around.
posted by nebulawindphone at 1:02 PM on August 26, 2019 [26 favorites]
If someone found a language that was less ambiguous than all the others, and quantified that in a way that the rest of the field found convincing, they would be instantly famous. It would be a huge, huge discovery. So far, nobody has done that. That doesn't mean such a language isn't out there. But I think most linguists would be tremendously surprised if one was discovered.
To answer your specific question about word order: you're right that languages that leave word order flexible generally give you some other way to disambiguate a sentence. Maybe there's a special ending that goes on the subject of the sentence to distinguish it from the object. Maybe the verb agrees with its subject, or its object, or both. Maybe a subject-verb-object sentence and an object-verb-subject sentence are pronounced with different melodies, or with the pauses in different places. Those things all cut down on the ambiguity of an average sentence, compensating for the free word order and bringing it back to the same level of ambiguity we have in English. But they aren't perfect, and it's certainly still just as easy to create ambiguous sentences in these languages as it is in English.
To answer your specific question about "we" and "we": yeah, some languages distinguish "we including you" and "we not including you." But of course, in languages that don't have a distinction like that built in, you can still make it explicit by saying "We — you too, komara — have to do this." And in languages that do have the distinction built in, there are other ambiguities: "we-not-including-you" is still ambiguous between "me and my wife," "me and my brother," "me and my wife and my brother," etc. So the upshot is that having this feature doesn't make a language overall less ambiguous than English, it just moves the ambiguity around.
posted by nebulawindphone at 1:02 PM on August 26, 2019 [26 favorites]
Language is a human construct, and because humans are cussed and imaginative, there will always be potential for ambiguity in language.
Consider the sentence "the red car in front of the green house." If you ran that through the juicer you could get "the green car in front of the red house." In some languages, the adjectives are declined to match their nouns' gender/number, but that would only help you here coincidentally. I'm not aware of a language that forces you to relate a modifier to a sentence part. And this is a trivial example. There are so many ways of communicating meaning that are inherent in some languages and optional in others, and they don't overlap. I imagine a language that incorporated all these different constructions would collapse under its own weight.
posted by adamrice at 1:25 PM on August 26, 2019 [1 favorite]
Consider the sentence "the red car in front of the green house." If you ran that through the juicer you could get "the green car in front of the red house." In some languages, the adjectives are declined to match their nouns' gender/number, but that would only help you here coincidentally. I'm not aware of a language that forces you to relate a modifier to a sentence part. And this is a trivial example. There are so many ways of communicating meaning that are inherent in some languages and optional in others, and they don't overlap. I imagine a language that incorporated all these different constructions would collapse under its own weight.
posted by adamrice at 1:25 PM on August 26, 2019 [1 favorite]
(I've almost got a PhD in linguistics.)
Nebulawindphone has the right answer.
We can't quantify ambiguity in the way that is necessary to answer this question. All languages have ambiguity. In some cases, you may be able to point to specific things that are ambiguous in one language but not in another - for example, the inclusive versus exclusive "we." However, we can't quantify the level of ambiguity as a whole. Any answer that points to a specific feature like grammatical case and then concludes that makes the language less ambiguous than others is a wrong answer.
There's also the question of how much you take context into account. This could be things that the audience knows (world knowledge), or things that are clear from the discourse. Sometimes context can force an unambiguous interpretation - for example, we may have no option other than to interpret the pronoun "he" as referring to John, given the sentence that came before. The sentence may be ambiguous in isolation, and unambiguous in context, and it will be difficult to come up with a rigorous, formal line in the sand between what counts and what doesn't.
posted by Kutsuwamushi at 1:28 PM on August 26, 2019 [4 favorites]
Nebulawindphone has the right answer.
We can't quantify ambiguity in the way that is necessary to answer this question. All languages have ambiguity. In some cases, you may be able to point to specific things that are ambiguous in one language but not in another - for example, the inclusive versus exclusive "we." However, we can't quantify the level of ambiguity as a whole. Any answer that points to a specific feature like grammatical case and then concludes that makes the language less ambiguous than others is a wrong answer.
There's also the question of how much you take context into account. This could be things that the audience knows (world knowledge), or things that are clear from the discourse. Sometimes context can force an unambiguous interpretation - for example, we may have no option other than to interpret the pronoun "he" as referring to John, given the sentence that came before. The sentence may be ambiguous in isolation, and unambiguous in context, and it will be difficult to come up with a rigorous, formal line in the sand between what counts and what doesn't.
posted by Kutsuwamushi at 1:28 PM on August 26, 2019 [4 favorites]
Not really an answer, but: programming languages. You have to tell the computer what to do, and if there's any ambiguity at all, the computer won't know what to do. Of course, these languages are designed, completely unlike most human languages.
The reason I think this is worth posting is that you specified word order. Most programming languages are very rigid about word order (yes, I'm aware of Perl and its postfix conditionals -- the creator of Perl was a linguist). I would imagine that having very flexible word order would tend to make a language more ambiguous, not less.
posted by number9dream at 1:45 PM on August 26, 2019 [3 favorites]
The reason I think this is worth posting is that you specified word order. Most programming languages are very rigid about word order (yes, I'm aware of Perl and its postfix conditionals -- the creator of Perl was a linguist). I would imagine that having very flexible word order would tend to make a language more ambiguous, not less.
posted by number9dream at 1:45 PM on August 26, 2019 [3 favorites]
Properly conjugated Latin lets you change word order in simple sentences and keep the meaning.
Hence
"canis virum mordet" = "dog bites man" and
"virum canis mordet" = "dog bites man" as well.
"man bites dog" would be "vir canem mordet" or "canem vir mordet".
However it's not really that simple and word order does hold part of the meaning in Latin. Changing the word order from the default makes it look as though you are emphasizing the thing that comes unusually first, so instead of "dog bites man" you get a meaning like "It was a dog that bit the man".
posted by w0mbat at 2:23 PM on August 26, 2019 [2 favorites]
Hence
"canis virum mordet" = "dog bites man" and
"virum canis mordet" = "dog bites man" as well.
"man bites dog" would be "vir canem mordet" or "canem vir mordet".
However it's not really that simple and word order does hold part of the meaning in Latin. Changing the word order from the default makes it look as though you are emphasizing the thing that comes unusually first, so instead of "dog bites man" you get a meaning like "It was a dog that bit the man".
posted by w0mbat at 2:23 PM on August 26, 2019 [2 favorites]
You also right there illustrate that Latin prefers a verb-final order.
posted by praemunire at 3:04 PM on August 26, 2019
posted by praemunire at 3:04 PM on August 26, 2019
To elaborate on typify's suggestion of Lojban a bit, Lojban is a constructed language with unambiguous parsing as one of its main design principles, and is intended to make it always obvious what is meant with phrases like "red dog ball" or "all dogs are not brown." Its development began back in the 1960s in part as a language for computers, intended as a sort of middle step between programming languages and the full complexity of natural languages, so the goal was to try to strip out the kinds of ambiguity that computers would struggle with. However, I don't think even Lojban is as fully unambiguous as its creators initially intended; a certain amount of ambiguity may be necessary for the full expressive power of natural language.
Also, even computer programming languages are sometimes ambiguous. For example, certain valid C/C++ code snippets can't be parsed unless you know the context (specifically, whether an identifier is a variable name or a type), and the SQL standard is ambiguous in how to treat NULL values, such that different implementations have different interpretations.
posted by biogeo at 4:34 PM on August 26, 2019
Also, even computer programming languages are sometimes ambiguous. For example, certain valid C/C++ code snippets can't be parsed unless you know the context (specifically, whether an identifier is a variable name or a type), and the SQL standard is ambiguous in how to treat NULL values, such that different implementations have different interpretations.
posted by biogeo at 4:34 PM on August 26, 2019
Many years ago I read Umberto Eco's The Search for the Perfect Language. Your question and the answers so far have brought it to mind.
posted by OHenryPacey at 5:26 PM on August 26, 2019 [1 favorite]
posted by OHenryPacey at 5:26 PM on August 26, 2019 [1 favorite]
You might like Latin. Here's my favorite example of how much you can mess up a Latin sentence; it's from a poem by Horace.
Quis multa gracilis te puer in rosa
perfusus liquidis urget odoribus
grato, Pyrrha, sub antro?
Word by word: "Which many slender you boy in rose / many liquids makes.love perfumed / delightful Pyrrha in cave?"
Thanks to the case, number, and gender marking everything can be reassembled correctly: "What slender boy, wet with perfumes, is making love to you, Pyrrha, amid many roses, in a delightful cave?"
But, one, it's poetry, and ordinary prose is far less fractured; and two, Horace is carefully choosing words so that it all works. If Pyrrha was making love to a puella (girl) instead, a bunch of ambiguities would appear. Similarly, we can sort out all the ablatives because "roses" are feminine plural and "cave" is masculine singular.
posted by zompist at 6:12 PM on August 26, 2019 [1 favorite]
Quis multa gracilis te puer in rosa
perfusus liquidis urget odoribus
grato, Pyrrha, sub antro?
Word by word: "Which many slender you boy in rose / many liquids makes.love perfumed / delightful Pyrrha in cave?"
Thanks to the case, number, and gender marking everything can be reassembled correctly: "What slender boy, wet with perfumes, is making love to you, Pyrrha, amid many roses, in a delightful cave?"
But, one, it's poetry, and ordinary prose is far less fractured; and two, Horace is carefully choosing words so that it all works. If Pyrrha was making love to a puella (girl) instead, a bunch of ambiguities would appear. Similarly, we can sort out all the ablatives because "roses" are feminine plural and "cave" is masculine singular.
posted by zompist at 6:12 PM on August 26, 2019 [1 favorite]
Also consider that it is sometimes socially advantageous to speak ambiguously, and so a society with an unambiguous language would be strongly incentivized to develop ambiguity.
Conversely, there sometimes are situations that strongly incentivise unambiguous communication, and there people will develop locally unambiguous jargons. Think, for example, of how traders shout at each other in the pit, or how pilots speak to the air traffic controller.
posted by meaty shoe puppet at 7:28 PM on August 26, 2019 [1 favorite]
Conversely, there sometimes are situations that strongly incentivise unambiguous communication, and there people will develop locally unambiguous jargons. Think, for example, of how traders shout at each other in the pit, or how pilots speak to the air traffic controller.
posted by meaty shoe puppet at 7:28 PM on August 26, 2019 [1 favorite]
Another thing we haven't really touched on is that disambiguation is inefficient.
For example, you can make it clearer how the words in a sentence relate to each other by adding case markers or agreement markers. But doing that makes all your words longer. Or you can make those relationships clearer by having fixed word order. But doing that means you can't use word order for expressive purposes or for emphasis, and so you need special expressive or emphatic words, which make your sentences longer.
You can have a richer vocabulary that disambiguates between "we including you" and "we excluding you," or between "mother's brother," "father's brother," "mother's sister's husband," and so on. But that means you use up all your short words faster, and so need to resort to longer words more often, again making your sentences are longer. You can make it harder to mishear things by having only a few speech sounds — eight consonants are easier to keep straight than eighty — but again, that runs you out of short words faster, since eight consonants allow for fewer combinations than eighty. (And if you try to compensate by making your syllables more complicated, so that there are more one-syllable words again, you'll be thwarted by the fact that more complicated syllables are slower to pronounce: "strengths" takes more time to say than "go.")
We can think of clarity and efficiency as two competing forces. In some specific situations, we might lean unusually far in the direction of clarity, for instance by saying "mother's brother's husband" instead of "uncle," or by saying "sierra tango romeo echo november golf tango hotel sierra" instead of "strengths." In others, we might lean unusually far in the direction of efficiency, for instance by saying "iunno" instead of "I do not know that."
But a language that forced you to be unambiguous would be a language that forced you to be inefficient. And that's pretty clearly undesirable. Better to have options.
posted by nebulawindphone at 6:00 AM on August 27, 2019 [3 favorites]
For example, you can make it clearer how the words in a sentence relate to each other by adding case markers or agreement markers. But doing that makes all your words longer. Or you can make those relationships clearer by having fixed word order. But doing that means you can't use word order for expressive purposes or for emphasis, and so you need special expressive or emphatic words, which make your sentences longer.
You can have a richer vocabulary that disambiguates between "we including you" and "we excluding you," or between "mother's brother," "father's brother," "mother's sister's husband," and so on. But that means you use up all your short words faster, and so need to resort to longer words more often, again making your sentences are longer. You can make it harder to mishear things by having only a few speech sounds — eight consonants are easier to keep straight than eighty — but again, that runs you out of short words faster, since eight consonants allow for fewer combinations than eighty. (And if you try to compensate by making your syllables more complicated, so that there are more one-syllable words again, you'll be thwarted by the fact that more complicated syllables are slower to pronounce: "strengths" takes more time to say than "go.")
We can think of clarity and efficiency as two competing forces. In some specific situations, we might lean unusually far in the direction of clarity, for instance by saying "mother's brother's husband" instead of "uncle," or by saying "sierra tango romeo echo november golf tango hotel sierra" instead of "strengths." In others, we might lean unusually far in the direction of efficiency, for instance by saying "iunno" instead of "I do not know that."
But a language that forced you to be unambiguous would be a language that forced you to be inefficient. And that's pretty clearly undesirable. Better to have options.
posted by nebulawindphone at 6:00 AM on August 27, 2019 [3 favorites]
Response by poster: I enjoyed each and every one of your responses. Thank you.
posted by komara at 9:27 AM on August 30, 2019 [1 favorite]
posted by komara at 9:27 AM on August 30, 2019 [1 favorite]
This thread is closed to new comments.
posted by typify at 12:26 PM on August 26, 2019 [4 favorites]