OT, but: This seems like a problematic thing to say without further qualification. The dangers of 'newspeak' were elucidated in 1984; this was a language which was purposefully designed to be less...
OT, but:
Every language is equally capable of expressing all of human experience
This seems like a problematic thing to say without further qualification. The dangers of 'newspeak' were elucidated in 1984; this was a language which was purposefully designed to be less capable than the status quo of expressing human experience. Is it a stretch to say that natural languages may have varying degrees of expressiveness? The differences, to be sure, are likely to be less pronounced, and less monotonic. But a statement such as this would seem to suggest it's impossible for one language to be more expressive than another, which seems like nonsense.
This is an interesting thing to think about. As a linguist, I actually think this is highly sensical. The arguments as for why it makes sense touch on lots of ideas across fields such as...
Exemplary
But a statement such as this would seem to suggest it's impossible for one language to be more expressive than another, which seems like nonsense.
This is an interesting thing to think about. As a linguist, I actually think this is highly sensical. The arguments as for why it makes sense touch on lots of ideas across fields such as linguistics, formal language theory, information theory, and computer science, and I think it's all quite interesting.
In terms of the Chomsky Hierarchy (yeah, that Chomsky), which classifies languages (of any kind, including artificial ones) based on their formal structure/syntax, natural languages are type-0 (i.e., recursively enumerable). At least in terms of the syntactic structures that languages license, it's important to note that all languages of the same type in Chomsky's hierarchy are equivalently expressive. And, the formal grammars of natural languages are classified as type-0, which is the most expressive type.
Now, the overall expressivity of languages in terms of information theory—that is, how efficient are languages in expressing the same thing, is another matter. If you want to look at written language and do text analysis, one can think about the entropy of sequences of written words. This is an information theoretic approach to the question of efficiency of expressivity. For written text, this ends up being related to the idea of how compressible text in a given language is. Since the semantics of a given natural language utterance is dependent on the individual words in the utterance, as well as the word order, one might look at the difference in entropy between the grammatical order of words in text vs. randomly shuffled (ungrammatical) sequences of the same words. Researchers did just this in this study: Universal Entropy of Word Ordering Across Linguistic Families. That study concluded:
Our results indicate that despite the differences in the structure and vocabulary of the languages analyzed, the impact of word ordering in the structure of language is a statistical linguistic universal.
The figures show that while the absolute entropy (measured in bits) varies across languages, the absolute difference in entropy (in bits) between texts of a given language and shuffled words in that language is near constant. I take this is as strong evidence that if you more or less factor out differences in the lexicons across natural languages, they are very similarly expressive. If you accept this, then it follows that while some languages may be able to express some things more efficiently/concisely, all natural languages are essentially equally expressive, given enough words (or enough morphemes).
I'll caveat that if you start bringing in pragmatics and devices like metonymy, I think the picture becomes fuzzier and much more difficult to study, scientifically.
To me as a layperson, language is a communication system. I don't think one can separate a language's expressivity from its efficiency. Said in another way: Communication is also about the...
To me as a layperson, language is a communication system. I don't think one can separate a language's expressivity from its efficiency. Said in another way:
Communication is also about the strength one can deliver a message with. When we communicate, being able to say something succinctly alters perception. Theoretical/mathematical models don't take that into account.
Language is in itself a way of knowing and impacts the message delivered. Suggesting that every language is as expressive must in my view necessarily undermine the epistemological impact of language. That in turn undervalues the power of language, and underestimates the cultural impact of language.
The words in different languages that have translations for each other have different etymologies, connotations, shared cultural references, contextual uses, historical uses, homonyms etc. They have real impacts communicatively. To me, those will be features of a specific language.
To take an example from languages I'm native in: The English language can't convey the jab of a Norwegian sarcastically saying: "How hyggelig" (literally: "How cozy/nice/enjoyable/cozy"). I could write an essay trying to get at it, but the power of the jab is lost in the essay, just like a joke dies in explanation.
In dialogue:
"The in-laws decided to stay a couple days longer, so we can't join you at the beach"
Response: "How hyggelig"
I struggle beginning to explain the levels of unease, understanding, social expectation, traditional family roles, acknowledgement of suffering, happiness it's not me, shared understanding, empathy, etc. etc. are expressed in two sarcastic words. And how the social bond and tone between exactly the two people in that specific conversation in what specific context, follow-up in situation etc. alters the message.
The whole situation is so different from the English put-down of "how nice"-ing something that the whole situations are pretty much incomparable, although someone without a deep understanding of both languages might think they're pretty analogous; both people are saying "how nice" in a sarcastic way, after all!
These sorts of features of conciseness in conveying specific cultural references or frames are unique from language to language. The cultural importance of language also varies from culture to culture. The richness with which normal people speak varies widely, also within the same society/geography, not just from language to language.
Suggesting that "every language is equally capable of expressing all of human experience" forgets that some languages are more equal than others, whether we like it or not.
What is the difference between a long-winded essay and a pithy one liner that say the same thing? If your claim is that there is an extra dimension of cultural information that is being carried by...
I don't think one can separate a language's expressivity from its efficiency.
What is the difference between a long-winded essay and a pithy one liner that say the same thing?
To take an example from languages I'm native in: The English language can't convey the jab of a Norwegian sarcastically saying: "How hyggelig" (literally: "How cozy/nice/enjoyable/cozy"). I could write an essay trying to get at it, but the power of the jab is lost in the essay, just like a joke dies in explanation.
If your claim is that there is an extra dimension of cultural information that is being carried by some utterances and that this dimension of the meaning in some languages is impossible to translate to others, I'd say that there may be something to that. Unfortunately, it's really hard to falsify that hypothesis. How can you be certain that there is no way to translate it? I'd argue that most natural language utterances don't fall into this culturally loaded category, though.
These sorts of features of conciseness in conveying specific cultural references or frames are unique from language to language.
The phenomenon of sarcasm, generally, and litotes specifically (which I think is what this Norwegian saying is?) are most definitely cross-lingual phenomena and not unique.
Suggesting that "every language is equally capable of expressing all of human experience" forgets that some languages are more equal than others, whether we like it or not.
As a fellow layman, just to be clear, you're saying that particular cultural concepts may not be translatable, and so to say every language is identical in the expression of human experience would...
As a fellow layman, just to be clear, you're saying that particular cultural concepts may not be translatable, and so to say every language is identical in the expression of human experience would be incorrect?
Pretty much: Every language has cultural concepts, and areas they're stronger/weaker at expressing. That depends on culture, context and history. Languages can't express all these equally well, so...
Pretty much:
Every language has cultural concepts, and areas they're stronger/weaker at expressing. That depends on culture, context and history. Languages can't express all these equally well, so every language isn't equally capable of expressing all of human experience.
(Further, some languages and cultures don't focus on expressing all of human experience as richly as possible, so some languages are also more equal than others in their ability to express human experiences.)
It feels like you're conflating the basic capacity for any language to be used to convey roughly any human experience, which is irrespective of efficiency, and any language being equally...
It feels like you're conflating the basic capacity for any language to be used to convey roughly any human experience, which is irrespective of efficiency, and any language being equally well-suited for any given concept. At least theoretically, there's nothing stopping anyone from writing a whole essay to set a mood in order to capture your earlier hyggelig usecase.
Do you have any examples of human languages that are not type-0? If all natural languages are type-0, does that mean a type-1 language would be more of a set of instructions like RegEx or a...
Do you have any examples of human languages that are not type-0? If all natural languages are type-0, does that mean a type-1 language would be more of a set of instructions like RegEx or a programming language?
I'm not aware of any natural languages with grammars that fall outside of type-0. There may be subsets of some natural languages that don't qualify as type-0. If you find any examples (or claims...
Do you have any examples of human languages that are not type-0?
I'm not aware of any natural languages with grammars that fall outside of type-0. There may be subsets of some natural languages that don't qualify as type-0. If you find any examples (or claims of examples) I'd be interested to hear about it.
If all natural languages are type-0, does that mean a type-1 language would be more of a set of instructions like RegEx or a programming language?
It's been a while since I've studied formal language theory, but IIRC, purely regular expressions can be mapped to nondeterministic finite automata (NFAs), so are unsurprisingly, type-3. However, a lot of modern regex engines include things like groups and look-{ahead,behind}s, thus modern regexes are actually context-sensitive, pushing them into type-2 or even type-1. Programming languages like Python where indentation is semantically significant are also type-1, I think.
A good example of type-2 languages are those whose grammars can be formalized in Backus-Naur form.
I'm certainly no expert (and now that one has chimed in, read that instead), but in the bit of linguistics I did take this seemed to be a pretty well-accepted concept. Natural language is...
I'm certainly no expert (and now that one has chimed in, read that instead), but in the bit of linguistics I did take this seemed to be a pretty well-accepted concept. Natural language is extremely good at adaptation and creating new vocabulary and grammatical forms when needed and according to the intuitive rules of the language (or just taking loaning whatever it needs). For instance, a language which is only used in an isolated area without a term for a computer is not less capable of expressing things surrounding computers...it's perfectly capable, because a language is not just the vocabulary and grammar that exist at this moment. Gaps are filled on the fly, and as a result don't really exist.
Newspeak was an interesting invention for science fiction, but it's not, uh, real. Raise a child in newspeak and you'll get a word for "unfair" the second you don't let them eat ice cream for breakfast :P
One must observe that 1984 is fictional. In any case, I might also observe that, while it is possible to severely censor and control how language is actually used, it is not as possible to...
One must observe that 1984 is fictional.
In any case, I might also observe that, while it is possible to severely censor and control how language is actually used, it is not as possible to undermine the fact that it will remain at least potentially able to express everything that's being censored. In other words, one can censor dictionaries, but not semantics itself.
A really great example of this is Randall Munroe's Thing Explainer.
A really great example of this is Randall Munroe's Thing Explainer.
In Thing Explainer: Complicated Stuff in Simple Words, things are explained in the style of Up Goer Five, using only drawings and a vocabulary of the 1,000 (or "ten hundred") most common words. Explore computer buildings (datacenters), the flat rocks we live on (tectonic plates), the things you use to steer a plane (airliner cockpit controls), and the little bags of water you're made of (cells).
If you don't have a word for the colour blue, it's harder to distinguish blue from green So no, not all languages are equally capable. I mean sure, in theory any language can create a word to...
So no, not all languages are equally capable. I mean sure, in theory any language can create a word to describe a concept that didn't previously exist in that language but just because they can doesn't mean they do. It's hard to think about things you don't have words for.
There are languages which have specific words for the inside and outside of corners, yet we don't in English. That said, I'm still able to talk about and conceptualise the idea of a corner and its...
It's hard to think about things you don't have words for.
There are languages which have specific words for the inside and outside of corners, yet we don't in English. That said, I'm still able to talk about and conceptualise the idea of a corner and its inside/outside.
Inside and outside of corners? That could mean quite a few different things depending on the context, but anyway I didn't say it was impossible to think about things you don't have words for....
Inside and outside of corners? That could mean quite a few different things depending on the context, but anyway I didn't say it was impossible to think about things you don't have words for. Although there will be terms in maths for those things anyway.
Did you read the article? That explains rather better than the ten words I used.
Excellent video that elucidates heavily gendered languages in a way that is comprehensible for English speakers. Totally irrelevant, but what about the random cowboy hat? Nothing against it, but...
Excellent video that elucidates heavily gendered languages in a way that is comprehensible for English speakers.
Totally irrelevant, but what about the random cowboy hat? Nothing against it, but dude should probably adjust the rest of the attire to the presence of a cowboy hat :P
Seems like you have a little against it then :). I agree it's kinda random but I've come to appreciate things out of the ordinary. I learned something here and that was great.
Nothing against it, but dude should probably adjust the rest of the attire to the presence of a cowboy hat :P
Seems like you have a little against it then :). I agree it's kinda random but I've come to appreciate things out of the ordinary.
It's a great video but I've been thinking a lot about men's fashion lately and specially about hats and other head coverings devices which I recently acquired, and I'm of the opinion that they...
It's a great video but I've been thinking a lot about men's fashion lately and specially about hats and other head coverings devices which I recently acquired, and I'm of the opinion that they must help compose a certain look, so that's why I made this unnecessary off-topic addendum ;)
So I have nothing against cowboy hats per-se but in my opinion they are part of a certain package if you will.
I dunno about other languages, but in Ukrainian grammatical gender is for grammatical cases which English has very limited and mostly for pronouns. Btw, Ukrainian has animacy and inanimacy (which...
I dunno about other languages, but in Ukrainian grammatical gender is for grammatical cases which English has very limited and mostly for pronouns. Btw, Ukrainian has animacy and inanimacy (which was mentioned there) within masculine too.
For now and very theoretically Ukrainian:
can merge neutral and masculine into one,
merge all genders into one for plural,
remove some cases; tbh, I didnʼt research here, but accusative and vocative are kinda totally removable.
But thereʼre still several genders (new one and feminine), 5 cases.
Word order in Ukrainian mostly affects on theme~rheme and relation between words, so no. Genetive almost easily replaces accusative if we speak about masc., neut. and plural for any gender. If you...
Word order in Ukrainian mostly affects on theme~rheme and relation between words, so no.
Genetive almost easily replaces accusative if we speak about masc., neut. and plural for any gender. If you want to save animal + inanimal then you also add nominative here. But today animal-type can be used for inaminal too: знайшов мʼяча instead standard знайшов мʼяч.
OT, but:
This seems like a problematic thing to say without further qualification. The dangers of 'newspeak' were elucidated in 1984; this was a language which was purposefully designed to be less capable than the status quo of expressing human experience. Is it a stretch to say that natural languages may have varying degrees of expressiveness? The differences, to be sure, are likely to be less pronounced, and less monotonic. But a statement such as this would seem to suggest it's impossible for one language to be more expressive than another, which seems like nonsense.
This is an interesting thing to think about. As a linguist, I actually think this is highly sensical. The arguments as for why it makes sense touch on lots of ideas across fields such as linguistics, formal language theory, information theory, and computer science, and I think it's all quite interesting.
In terms of the Chomsky Hierarchy (yeah, that Chomsky), which classifies languages (of any kind, including artificial ones) based on their formal structure/syntax, natural languages are type-0 (i.e., recursively enumerable). At least in terms of the syntactic structures that languages license, it's important to note that all languages of the same type in Chomsky's hierarchy are equivalently expressive. And, the formal grammars of natural languages are classified as type-0, which is the most expressive type.
Now, the overall expressivity of languages in terms of information theory—that is, how efficient are languages in expressing the same thing, is another matter. If you want to look at written language and do text analysis, one can think about the entropy of sequences of written words. This is an information theoretic approach to the question of efficiency of expressivity. For written text, this ends up being related to the idea of how compressible text in a given language is. Since the semantics of a given natural language utterance is dependent on the individual words in the utterance, as well as the word order, one might look at the difference in entropy between the grammatical order of words in text vs. randomly shuffled (ungrammatical) sequences of the same words. Researchers did just this in this study: Universal Entropy of Word Ordering Across Linguistic Families. That study concluded:
The figures show that while the absolute entropy (measured in bits) varies across languages, the absolute difference in entropy (in bits) between texts of a given language and shuffled words in that language is near constant. I take this is as strong evidence that if you more or less factor out differences in the lexicons across natural languages, they are very similarly expressive. If you accept this, then it follows that while some languages may be able to express some things more efficiently/concisely, all natural languages are essentially equally expressive, given enough words (or enough morphemes).
I'll caveat that if you start bringing in pragmatics and devices like metonymy, I think the picture becomes fuzzier and much more difficult to study, scientifically.
To me as a layperson, language is a communication system. I don't think one can separate a language's expressivity from its efficiency. Said in another way:
Language is in itself a way of knowing and impacts the message delivered. Suggesting that every language is as expressive must in my view necessarily undermine the epistemological impact of language. That in turn undervalues the power of language, and underestimates the cultural impact of language.
The words in different languages that have translations for each other have different etymologies, connotations, shared cultural references, contextual uses, historical uses, homonyms etc. They have real impacts communicatively. To me, those will be features of a specific language.
To take an example from languages I'm native in: The English language can't convey the jab of a Norwegian sarcastically saying: "How hyggelig" (literally: "How cozy/nice/enjoyable/cozy"). I could write an essay trying to get at it, but the power of the jab is lost in the essay, just like a joke dies in explanation.
In dialogue:
I struggle beginning to explain the levels of unease, understanding, social expectation, traditional family roles, acknowledgement of suffering, happiness it's not me, shared understanding, empathy, etc. etc. are expressed in two sarcastic words. And how the social bond and tone between exactly the two people in that specific conversation in what specific context, follow-up in situation etc. alters the message.
The whole situation is so different from the English put-down of "how nice"-ing something that the whole situations are pretty much incomparable, although someone without a deep understanding of both languages might think they're pretty analogous; both people are saying "how nice" in a sarcastic way, after all!
These sorts of features of conciseness in conveying specific cultural references or frames are unique from language to language. The cultural importance of language also varies from culture to culture. The richness with which normal people speak varies widely, also within the same society/geography, not just from language to language.
Suggesting that "every language is equally capable of expressing all of human experience" forgets that some languages are more equal than others, whether we like it or not.
What is the difference between a long-winded essay and a pithy one liner that say the same thing?
If your claim is that there is an extra dimension of cultural information that is being carried by some utterances and that this dimension of the meaning in some languages is impossible to translate to others, I'd say that there may be something to that. Unfortunately, it's really hard to falsify that hypothesis. How can you be certain that there is no way to translate it? I'd argue that most natural language utterances don't fall into this culturally loaded category, though.
The phenomenon of sarcasm, generally, and litotes specifically (which I think is what this Norwegian saying is?) are most definitely cross-lingual phenomena and not unique.
I will humbly disagree.
As a fellow layman, just to be clear, you're saying that particular cultural concepts may not be translatable, and so to say every language is identical in the expression of human experience would be incorrect?
Pretty much:
Every language has cultural concepts, and areas they're stronger/weaker at expressing. That depends on culture, context and history. Languages can't express all these equally well, so every language isn't equally capable of expressing all of human experience.
(Further, some languages and cultures don't focus on expressing all of human experience as richly as possible, so some languages are also more equal than others in their ability to express human experiences.)
It feels like you're conflating the basic capacity for any language to be used to convey roughly any human experience, which is irrespective of efficiency, and any language being equally well-suited for any given concept. At least theoretically, there's nothing stopping anyone from writing a whole essay to set a mood in order to capture your earlier hyggelig usecase.
Do you have any examples of human languages that are not type-0? If all natural languages are type-0, does that mean a type-1 language would be more of a set of instructions like RegEx or a programming language?
I'm not aware of any natural languages with grammars that fall outside of type-0. There may be subsets of some natural languages that don't qualify as type-0. If you find any examples (or claims of examples) I'd be interested to hear about it.
It's been a while since I've studied formal language theory, but IIRC, purely regular expressions can be mapped to nondeterministic finite automata (NFAs), so are unsurprisingly, type-3. However, a lot of modern regex engines include things like groups and look-{ahead,behind}s, thus modern regexes are actually context-sensitive, pushing them into type-2 or even type-1. Programming languages like Python where indentation is semantically significant are also type-1, I think.
A good example of type-2 languages are those whose grammars can be formalized in Backus-Naur form.
I'm certainly no expert (and now that one has chimed in, read that instead), but in the bit of linguistics I did take this seemed to be a pretty well-accepted concept. Natural language is extremely good at adaptation and creating new vocabulary and grammatical forms when needed and according to the intuitive rules of the language (or just taking loaning whatever it needs). For instance, a language which is only used in an isolated area without a term for a computer is not less capable of expressing things surrounding computers...it's perfectly capable, because a language is not just the vocabulary and grammar that exist at this moment. Gaps are filled on the fly, and as a result don't really exist.
Newspeak was an interesting invention for science fiction, but it's not, uh, real. Raise a child in newspeak and you'll get a word for "unfair" the second you don't let them eat ice cream for breakfast :P
What a great example you came up with ;)
One must observe that 1984 is fictional.
In any case, I might also observe that, while it is possible to severely censor and control how language is actually used, it is not as possible to undermine the fact that it will remain at least potentially able to express everything that's being censored. In other words, one can censor dictionaries, but not semantics itself.
A really great example of this is Randall Munroe's Thing Explainer.
Would you say "euphemism treadmills" are a good example of this dynamic?
That does seem to be the case, yes.
If you don't have a word for the colour blue, it's harder to distinguish blue from green
So no, not all languages are equally capable. I mean sure, in theory any language can create a word to describe a concept that didn't previously exist in that language but just because they can doesn't mean they do. It's hard to think about things you don't have words for.
There are languages which have specific words for the inside and outside of corners, yet we don't in English. That said, I'm still able to talk about and conceptualise the idea of a corner and its inside/outside.
Inside and outside of corners? That could mean quite a few different things depending on the context, but anyway I didn't say it was impossible to think about things you don't have words for. Although there will be terms in maths for those things anyway.
Did you read the article? That explains rather better than the ten words I used.
Excellent video that elucidates heavily gendered languages in a way that is comprehensible for English speakers.
Totally irrelevant, but what about the random cowboy hat? Nothing against it, but dude should probably adjust the rest of the attire to the presence of a cowboy hat :P
Seems like you have a little against it then :). I agree it's kinda random but I've come to appreciate things out of the ordinary.
I learned something here and that was great.
It's a great video but I've been thinking a lot about men's fashion lately and specially about hats and other head coverings devices which I recently acquired, and I'm of the opinion that they must help compose a certain look, so that's why I made this unnecessary off-topic addendum ;)
So I have nothing against cowboy hats per-se but in my opinion they are part of a certain package if you will.
Oh I feel you and it was off for me too but he was so likable I let it slide.
I dunno about other languages, but in Ukrainian grammatical gender is for grammatical cases which English has very limited and mostly for pronouns. Btw, Ukrainian has animacy and inanimacy (which was mentioned there) within masculine too.
For now and very theoretically Ukrainian:
But thereʼre still several genders (new one and feminine), 5 cases.
Interesting, the accusative is removable? What would replace it - word order?
Word order in Ukrainian mostly affects on theme~rheme and relation between words, so no.
Genetive almost easily replaces accusative if we speak about masc., neut. and plural for any gender. If you want to save animal + inanimal then you also add nominative here. But today animal-type can be used for inaminal too: знайшов мʼяча instead standard знайшов мʼяч.