in the mornings, crows partner up and go out for bugs. they bark and caw to find each other, to communicate about their environment, to signal sources of food, etc.
in the afternoon, they group into murders and distribute their bounty / argue / be jerks to each other for inscrutable reasons.
in the evening they gather by the hundreds and select a roost for the night through some unknowable collective process.
in the morning they do it again.
i have a methodology that requires resources i will probably never obtain that has been shown in experiments to be effective at establishing a basic translation matrix for animal speech. i'm convinced we can translate crow languages (yes they have many), but also that a practice of compassionate xenodiplomacy exists where such translation is not necessary for cooperation.
@garbados this is indisputably the best toot I have ever read
@garbados Well, you got my attention. I’ve been lightly keeping abreast of cetacean communication research for a while now. And heartily agreed many of the methodologies are primitive and fraught with behaviourist conditioning abuses. I’m not convinced what dolphins are speaking is “language” in the same sense as humans, but it must statistically be at *least* as complex. Some higher order translation matrix must be achievable.
What is your methodology that is less awful? Any reading links?
@garbados Clarifying: What do you make of the idea that all non-human communication is purely referential? It seems to hold for great apes. They don’t do grammar well. But cetaceans have evolved something higher order, we can tell that even if we don’t know what it is, though lit likely won’t have generative structures the same as humans. I know corvids are smart, but has any statistical model worked out if they are communicating referentially or using some form of linguistic association?
@Shufei i have to admit i'm not sure what's meant by the difference between referential communication and linguistic association. could you explain it to me? google was unhelpful
@garbados I’ll try not to botch it, ha. Referentialism is the level of communication in which a symbol signifies a thing. Think of prairiedog guards with their words for various critters and threats. “Wolf!” “Human!” Chomsky’s instructive in teasing out where human language may have taken off from that: a system which encodes not just referents but associative representations. So, (1+1) versus (1). Whatever cetaceans do, it must be as complex, but the association original grammars may differ.
@garbados I actually ran into Chomsky a while back and cornered him to trouble his assertion that only humans have transcended referentialism, by asking him about the various language oddities common to autistic people. He punted it, ha, poor chap, though he was indulgent to give me time. But he did say people were studying autism as edge cases. Also polysynthetic languages, natch. It might be a place to look for your crow pals, too.
@Shufei ooooh good leads :D thanks shufei
glad chomsky had some interesting things to say about it!
@garbados I hope so, and thanks for this chat. I’ll be reading with great interest your friendship with the crows.
Do check out Chomsky’s recent work, and not just what he calls “the linguistic statisticians” say about what he says. His actual arguments for original grammars (in a Chomskian mode) are often misrepresented. He makes a good case that some simple motor cortex mutation allowed (representational) language, which means similar mutations are likely for other species. Maybe crows!
@Shufei that’s really fascinating! i’ll definitely check it out :D
@Shufei gotcha, i understand. i'm not sure if animal comms are purely referential -- it seems like a big supposition for a field replete with unknowns.
i wonder about the stories crows tell each other. when fledglings gather at the call of their elders, what generational wisdom do they absorb? beyond knowing to beware phenomena like moving cars, do crows rationalize the origin or purpose of cars? for all the generations the specific artifacts of this knowledge have survived, i wonder to what degree it approaches mythology.
but that's all speculation 🤷♀️
@garbados Crows do have those lovely, densely packed neural nets. It’s speculation, but we are inching to some suggestive circumstantial evidence... Hopefully you can make a breakthrough? Someone out there has grant money for you. Think of the AI translation VC’s who would drool for s new method.
And there is the intuitive, human, empathetic value which is much more than sentimental, as you are right to stress. Maybe the crows can teach you their lore. It’s right there for you. Go Goodall.
@Shufei this paper about fruit bats formalized my ideas about translation into a methodology for progressive translation, where you use a combination of human and machine labor to categorize not only vocalizations but also the observed context, which creates a basis for more precise analysis.
dolphins fascinate me because their speech has a semantic density that is unimaginable in human languages -- they can communicate detailed instructions astonishingly quickly. some translation efforts have focused on trying to inject semantic tokens into dolphin speech to try and perform classical semantic disambiguation, but they always come up with bupkis.
also that people have practiced productive collaboration with dolphins for thousands of years, until the industrialization of fishing 🤷♀️
@garbados Righto, I once read a thing by a neurologist describing the midbrain complexity of dolphins as vastly more integrated on an associative complexity than ours, even to the point of corpus callosum bandwidth as likely superior. Some recent talks by that lady doing the semantic tokens in the Caribbean have born some fruit though: they’ve learnt some dolphin names and gotten the dolphins to use the referents. But, yes, until a Goodall for cetaceans comes along, it is thin meat indeed.
@Shufei fuck yeah dolphin names :D
@garbados I do apologize for my habitually muddy lack of citations and proper nomenclature. Talking about language: I’m visually and temporally oriented so have trouble keeping track of such data. I wish I could remember the researcher’s name to give you. She was the one who is doing long term in situ studies with a few pods in the Caymans, I think? But yes, they have a few proper names, IIRC.
@Shufei oh, no worries. i'm the same way 😅 i wish i had a proper library, or time to commit to proper study. it gets to being a lot of theses informally synthesized into bibliography-free working theories. i failed out of academia and so my linguistic knowledge has a lot of holes in it 😬
@Shufei i got as far as efforts to turn the sapir-worf hypothesis into a means for deriving psych profiles from arbitrary language artifacts, because that fascinated me, but after a certain point i hit this wall of classified research because the surveillance state wanted to use those methods for natsec purposes 🙃
@garbados Woah. Say what!? They are using Sapir-worf to train predictive models? That’s... freaky. And a bit spurious maybe. I know they have gone back to it to show some modified soft S-W is at play. But are they saying “Alice uses Foo lexeme a lot so she is sociopathic”?
@Shufei yeah basically. another example might be outing spies based on innocuous language artifacts like the proportions of different parts of speech, or identifying militants based on speech patterns. such research and methods are heavily tied into a hard (and very nationalist) interpretation of S-W, such that i'm not surprised it never really amounted to anything. it's clear now that the surveillance state uses very different methods today but which are animated by a similar sense of racist-nationalist determinism.
@garbados So *this* is what they plan to implement with the “let’s surveil mentally ill people for pre-crime” initiative. No doubt as a wedge into “let’s monitor black people or foo for pre-crime”. It’s getting very discomfiting. So much for AI as a social levelling force!
There was a recent lexical analysis of the Book of Mormon you might fancy, showing much more refined authorship and genealogy of the text’s (often plagiarized) influences. Amazing presentation, and in this vein.
@garbados This lecture’s not great, and the forum is for Mormon apostates so is a receptive but not a very scholarly audience. But the textual analysis was very interesting, if you are interested in such things. I’ve wondered if a cross-dialect statistical analysis of various dolphin pods could tease out common threads as this chap did with Book of Mormon chiasmus. Anyway, if you like:
@Shufei incredible. i can’t wait to watch this :D
It seems like a motivation thing. Dogs are amazing because of how much they care about humans approval. It doesn't even take food reinforcement to train a dog. Dogs are unusual though, what really motivates most intelligent animals to communicate with humans? Food is a strong motivator but only so strong.
I dont think it is a question of intelligence when you look at really smart birds or cetaceans, its more like whats in it for them lol.
@garbados it's a shame our entire species (to a first approximation) is committed to total war against all other forms of life, because you'd ~think~ we'd be pretty good at diplomatic relations if we just made some efforts.
@brennen not our entire species, and not for our whole history — in different times and places, people have conducted that diplomacy and established complex productive relationships with their cohabitants. the war you describe began with specific people and specific acts, who only gained the power to imagine their ways as eternal over thousands of years.
@brennen like, it’s thought that dogs and cats conducted a sort of mutual domestication with humans borne of mutually beneficial relationships... but cattle didn’t choose to become livestock. specific groups abducted dozens of individuals from different bovine species and created the first livestock through a program of captivity and forced breeding. only thousands of years later would human livestock practices push the auroch, one of those original captured bovid species, to extinction.
It's like i'm reading the 21st century bread book lol
The social network of the future: No ads, no corporate surveillance, ethical design, and decentralization! Own your data with Mastodon!