3 ms·

I know people whose communication appears to be quasi-automatic responses to certain stimuli. I thought they'd worked out that dolphins have language and compl

I know people whose communication appears to be quasi-automatic responses to certain stimuli. I thought they'd worked out that dolphins have language and complex social structures? I was definitely told that when I saw the dolphins at Monkey Mia. And dogs definitely communicate with humans, and understand human speech ("W.A.L.K" or you get zoomies for 10 minutes). What's the research that says all of this is "not language"?

Teever 5y ago
You're mixing the ideas of communication and language. Communication is not language. Language is a type of communication. Language involves the use of symbols in a grammar that is sufficiently complex enough to facilitate recursive structures of arbitrary length. I think it is likely that many animals like birds, cetaceans and elephants can use language and that it is possible that some dogs and cats can too but your example of a dog understanding that the word walk correlates to going for a walk is not an example of a dog being able to grasp human language. My mom has Shelties which are a disconcertingly smart herding dog. I've witnessed one of her dogs overhearing relatively complex sentences, pausing to parse them, and then acting on the in away that indicates that they understand not just the correlation between subjects and objects but also the conditional logic in the sentence.
tsimionescu 5y ago
Here[0] is a paper going into some detail, comparing human language with animal communication systems. The mostly define the difference as related to composition: in animal communication, each "word" almost always carries individual meaning, while multiple "word" together carry all the meanings together (the paper also discusses cases where simple composition has in fact been identified). In human language, the meaning of a combination of words depends not just on the words themselves, but on the way they are combined as well. To show an example: if you give me some bananas and I say "I like bananas but not these ones", it means I don't like the bananas you gave me. But a bonobo or dog "saying" some equivalent of "like" and "banana" would always mean that they like them (according to the paper, they could also "say" something like "not-like", "banana" showing some simple compositionality). Another important observation (that is not found in the paper) is that animals have never been observed to "discuss" about abstract things or counter-factuals, which is what I mean when I say that their communication is quasi-automatic. For example, a monkey will only "say" "leopard" if it notices a leopard, or if it hears another monkey "say" "leopard" (or, most sophisticatedly, if it wants to elicit the leopard response from the other monkeys, e.g. to steal some food). But there is never a "discussion" about leopards that are not present and that does not cause the monkeys to have a leopard response. Essentially animal communication is much more similar to our non-verbal queues, such as moaning or gasping or screaming, rather than to verbal communication. It can be more sophisticated than that, especially in primates, whales, and dolphins, encoding some information about social relationships, but 1 year old children already display much more complex language abilities than the most sophisticated animal. [0] https://royalsocietypublishing.org/doi/10.1098/rstb.2019.0062 https://royalsocietypublishing.org/doi/10.1098/rstb.2019.006...
- marcus_holmes 5y ago
  that's interesting, thanks

Teever 5y ago

You're mixing the ideas of communication and language. Communication is not language. Language is a type of communication. Language involves the use of symbols in a grammar that is sufficiently complex enough to facilitate recursive structures of arbitrary length. I think it is likely that many animals like birds, cetaceans and elephants can use language and that it is possible that some dogs and cats can too but your example of a dog understanding that the word walk correlates to going for a walk is not an example of a dog being able to grasp human language. My mom has Shelties which are a disconcertingly smart herding dog. I've witnessed one of her dogs overhearing relatively complex sentences, pausing to parse them, and then acting on the in away that indicates that they understand not just the correlation between subjects and objects but also the conditional logic in the sentence.

tsimionescu 5y ago

Here[0] is a paper going into some detail, comparing human language with animal communication systems. The mostly define the difference as related to composition: in animal communication, each "word" almost always carries individual meaning, while multiple "word" together carry all the meanings together (the paper also discusses cases where simple composition has in fact been identified). In human language, the meaning of a combination of words depends not just on the words themselves, but on the way they are combined as well. To show an example: if you give me some bananas and I say "I like bananas but not these ones", it means I don't like the bananas you gave me. But a bonobo or dog "saying" some equivalent of "like" and "banana" would always mean that they like them (according to the paper, they could also "say" something like "not-like", "banana" showing some simple compositionality). Another important observation (that is not found in the paper) is that animals have never been observed to "discuss" about abstract things or counter-factuals, which is what I mean when I say that their communication is quasi-automatic. For example, a monkey will only "say" "leopard" if it notices a leopard, or if it hears another monkey "say" "leopard" (or, most sophisticatedly, if it wants to elicit the leopard response from the other monkeys, e.g. to steal some food). But there is never a "discussion" about leopards that are not present and that does not cause the monkeys to have a leopard response. Essentially animal communication is much more similar to our non-verbal queues, such as moaning or gasping or screaming, rather than to verbal communication. It can be more sophisticated than that, especially in primates, whales, and dolphins, encoding some information about social relationships, but 1 year old children already display much more complex language abilities than the most sophisticated animal. [0] https://royalsocietypublishing.org/doi/10.1098/rstb.2019.0062 https://royalsocietypublishing.org/doi/10.1098/rstb.2019.006...

marcus_holmes 5y ago
that's interesting, thanks