Sitting here at supper, I just watched one of my cats use some pretty strong language with one of my other cats. It was some very measured chattering to let him know she didn’t enjoy his presence. Chickens most definitely have a language - the egg song, the danger song, the rooster call, the check-in with each other, the check-in with the farmer. My point being, animals all around us use language if we are willing to accept that it is different from our own.
TFA refers to 'non-trivial compositionality' as what's novel about how humans communicate (and how it's perhaps not as novel as we thought):
> However, the team also found examples of non-trivial compositionality, the first such discovery outside of humans.
> The first non-trivial combination was high hoot-low hoot that was translated as a distress call. But it was also used to stop other individuals’ display behaviors—dramatic, exaggerated actions or gestures bonobos perform to assert dominance or attract attention. The second was either peep or yelp in the “join” meaning paired with high hoot to form a structure used for coordinating with others before traveling. Finally, the “I would like to” peep followed by “let’s stay together” whistle was used for initiating more romantically inclined interactions bonobos are famous for indulging in.
If anyone is interested more in Bonobos and wants to learn more about that species and the way that species connects with humanity I suggest the following book:
That sounds like the method the researchers used in the linked paper:
> The MCA is similar to a principal components analysis (PCA) but is conducted on categorical data: It performs a dimension reduction and then quantifies the statistical relationship between a specific utterance type and several FoCs (19) (see materials and methods)
Great overview of more complex vocal communications. Farley Mowat makes a similar case for wolves in Never Cry Wolf: https://en.wikipedia.org/wiki/Never_Cry_Wolf
And almost certainly: cetaceans too.
Every single one of these species need the "Whale-SETI" approach applied to them.
We're probably in for quite the shock when we finally decode their communications.
Sitting here at supper, I just watched one of my cats use some pretty strong language with one of my other cats. It was some very measured chattering to let him know she didn’t enjoy his presence. Chickens most definitely have a language - the egg song, the danger song, the rooster call, the check-in with each other, the check-in with the farmer. My point being, animals all around us use language if we are willing to accept that it is different from our own.
"Did you know elephants have a noise that means 'There are bees here, let's leave immediately'? Why don't humans?" (https://www.sciencedaily.com/releases/2010/04/100427093106.h...)
"Yes, humans also have a noise like that. It's the sentence 'There are bees here, let's leave immediately.'"
So the question is one of modularity. Having one sound for each such phrase doesn't scale.
TFA refers to 'non-trivial compositionality' as what's novel about how humans communicate (and how it's perhaps not as novel as we thought):
> However, the team also found examples of non-trivial compositionality, the first such discovery outside of humans.
> The first non-trivial combination was high hoot-low hoot that was translated as a distress call. But it was also used to stop other individuals’ display behaviors—dramatic, exaggerated actions or gestures bonobos perform to assert dominance or attract attention. The second was either peep or yelp in the “join” meaning paired with high hoot to form a structure used for coordinating with others before traveling. Finally, the “I would like to” peep followed by “let’s stay together” whistle was used for initiating more romantically inclined interactions bonobos are famous for indulging in.
If anyone is interested more in Bonobos and wants to learn more about that species and the way that species connects with humanity I suggest the following book:
https://www.amazon.com/Our-Inner-Ape-Primatologist-Explains/...
I read it many years ago and thought it to be excellent!
Does anyone know if a tokenization attempt has been made at this type of research?
That sounds like the method the researchers used in the linked paper:
> The MCA is similar to a principal components analysis (PCA) but is conducted on categorical data: It performs a dimension reduction and then quantifies the statistical relationship between a specific utterance type and several FoCs (19) (see materials and methods)
Well, it apparently has one error in the first sentence. I wonder how well the rest of it stands up to scrutiny?
According to Wikipedia, the live in the Democratic Republic of Congo, not the Republic of Congo.
https://en.m.wikipedia.org/wiki/Bonobo