Smart Animals

[The Yosemite Bear]

ok, usually the place for cats vs. dogs.

however as having worked in a national park I know that Crows and their relatives are freaking geniuses for birds. Turns out scientists in England are doing intellegence and problem solving experiments with them.

BBC

[Starglider]

Also parrots; they can do counting, basic grammar, several other abstract concepts. Particularly remarkable given how compact bird brains are, compared to chimps, dolphins etc. Bird brains are structurally different from mammal brains at both the macro and micro levels; like most other avian biology they are highly optimised for mass and volume utilisation. You could probably pack human-level intelligence into a dog sized brain if you engineered a brain using avian biology and human-like functional areas.

[Duckie]

As far as parrots, some African Grey Parrots have apparantly been caught using analogy- for instance, saying 'sheeps' as the plural of 'sheep' based upon the example of 'cat, cats'; or 'I flied here' rather than 'flew' based on 'walk, walked'.

If true (it's somewhat controversial because of lack of research in the area and human exceptionalism- for one, any Chomskyite would scoff at the idea that a non-human has an instinctive capability to understand language because it would undermine their thesis that the human brain is an organ naturally suited to learning language), it means that parrots can at least comprehend things like aggluntinating particles like "-d means past" without any prompting or memorization, meaning that parrots have all the necessary aspects for understanding language without any Clever Hans tricks.

[Erik von Nein]

As far as crow intelligence, this video from TED talks goes into how crow intelligence can be used to form a beneficial relationship with humans.

But, yes. Quite a few people are having to revise what was previously thought about birds in general. It turns out they have a huge capacity for learning and thinking exceeding or equaling most of our primate cousins.

Birds can easily learn languages. Studies done with bird calls have found that, while to a certain extent they are inherent processes, the specific calls themselves are learned behavior and they can and will modify them at their discretion. That they're doing the same with human language would not, in any way, surprise me.

[Surlethe]

Does avian intelligence reflect on therapod intelligence, or is that too far removed in time?

[The Yosemite Bear]

Well Sur, maybe the Jurassic Park wasn't too far off, they are relatives to the birds after all.

of course Brat would be better for explaining that, she is after all an expert in Dinosaurs.

[TithonusSyndrome]

Oh man, elephants. Without a doubt, elephants. The only other animals aside from humans to mourn their dead and a whole host of other things at that. For one, an elephant in a Korean zoo (and one in a Kazakh zoo earlier on) managed to figure out how to speak human languages. This is almost certainly an example of evolution in action.

[Mayabird]

Elephants are not birds, dude. Anywho...

Also parrots; they can do counting, basic grammar, several other abstract concepts. Particularly remarkable given how compact bird brains are, compared to chimps, dolphins etc. Bird brains are structurally different from mammal brains at both the macro and micro levels; like most other avian biology they are highly optimised for mass and volume utilisation. You could probably pack human-level intelligence into a dog sized brain if you engineered a brain using avian biology and human-like functional areas.

To be a little more detailed, avian neurons are smaller and more tightly packed than mammalian neurons. Also, high surface area is important for mammalian brains since that's where all the action is (hence the wrinkliness of human brains, to maximize the surface area), but avian processing occurs in little spherical regions scattered all within the brain, so all they would need to do is increase the density of those areas rather than having to pull tricks with surface area.


Anyway, I'm so very pleased that there's so much research going on now about avian intelligence. Maybe I should figure out how to get into the fun...

[Starglider]

If true (it's somewhat controversial because of lack of research in the area and human exceptionalism- for one, any Chomskyite would scoff at the idea that a non-human has an instinctive capability to understand language because it would undermine their thesis that the human brain is an organ naturally suited to learning language).

No, not necessarily. Firstly only a very few exceptional parrots have managed to master the basics of human grammar. Parrot super-geniuses are barely on a par with human clinical retards when it comes to language use. Secondly, many songbirds do have extensive specialised brain areas for learning, repeating and modifying calls, sometimes with quite complex rules. It is this evolved capability which is repurposed to learn human language, without it the bird's general intelligence would be wholly inadequate for the task.

[GrandMasterTerwynn]

Does avian intelligence reflect on therapod intelligence, or is that too far removed in time?

It does . . . but the typical therapod had a much lower brain-to-body mass ratio than modern avians do. The brainiest of them seemed to have an encephalization quotient roughly equal to that of at least an ostrich. Which makes them much brighter than we've given them credit for, but still not really all that brainy.

[The Yosemite Bear]

why am I now picturing Maya building an army of Giant Boobies for world conquest....

[Starglider]

It does . . . but the typical therapod had a much lower brain-to-body mass ratio than modern avians do. The brainiest of them seemed to have an encephalization quotient roughly equal to that of at least an ostrich.

I'm highly dubious about 'brain-to-body ratio' as a metric for estimating intelligence. Why should having a larger body mean that you need more neurons to solve the same abstract reasoning problem? The only way in which it might make sense is that 'there is more body to control, so larger motor control areas are needed', but even that doesn't really work. Horses have about the same number of muscles and organs as mice.

I am more inclined to believe that there is a certain minimum brain size threshold for given levels of intelligence, but that the size of the body is fairly irrelevant. Larger animals tend to have larger brains because the way size is selected for tends to scale the whole body up evenly. This sometimes produces animals with brains that are strictly oversized for their needs, but there is only a limited amount of selection pressure to go round, and for most species reducing brain mass is not a fitness priority (the massively demanding human brain is an exceptional case). You can see this even in dogs. A Great Dane has a much larger brain than a Chihuahua, yet has a much smaller brain-to-body size ratio... and has an almost identical level of intelligence. Clearly a chihuahua sized brain suffices for dog intelligence.

[Duckie]

If true (it's somewhat controversial because of lack of research in the area and human exceptionalism- for one, any Chomskyite would scoff at the idea that a non-human has an instinctive capability to understand language because it would undermine their thesis that the human brain is an organ naturally suited to learning language).

No, not necessarily. Firstly only a very few exceptional parrots have managed to master the basics of human grammar. Parrot super-geniuses are barely on a par with human clinical retards when it comes to language use. Secondly, many songbirds do have extensive specialised brain areas for learning, repeating and modifying calls, sometimes with quite complex rules. It is this evolved capability which is repurposed to learn human language, without it the bird's general intelligence would be wholly inadequate for the task.

Well, as far as the Chomskyan Language Organ theory goes, any species that is nonhuman except perhaps neanderthal or other close relatives of humanity speaking a language without mimicry deals a pretty deep blow to the idea that the human brain is somehow exceptional and unique in its language capacity and also puts doubt on the idea that brains of language-using species have a special innate capacity to learn language at all. Rather, i would simply be related to a brain being complex and structured correctly- something which parrots, as a marginal case where only some can learn and speak languages, supports (presumably parrots thus being on the edge of humanlike language competence so that only the 'smartest' parrots can speak human languages, as you said yourself.)

Also, if learning, repeating, and modifying sounds and sets of sounds with complex rules doesn't sound like a description of how syntax and grammar works, I'll be damned. For all we know, human language was orignially repurposed from some similar learning-repeating-modifying sound system used for god knows what. I wouldn't count birds out just because they use the same part of their brain for singing as speaking.

[Starglider]

Well, as far as the Chomskyan Language Organ theory goes, any species that is nonhuman except perhaps neanderthal or other close relatives of humanity speaking a language without mimicry deals a pretty deep blow to the idea that the human brain is somehow exceptional and unique in its language capacity

Well, the human brain is certainly exceptional, compared to other organisms evolved on planet Earth. The term 'unique' is problematic. Firstly, language is not a binary property, there is a spectrum of communication flexibility from the simplest chemical signals used by bacterial colonies up to the mind-state transfers of hypothetical future transhumans. Unfortunately, as so often in evolved organisms we can't see all of the intermediate steps. Secondly, language is inseparable from its social context. Parrots don't have the means to invent human-like language on their own, whereas human children left without adult guidance will invent their own language from scratch. I agree that a lot of Chomskyites are far too arrogant with regard to human exceptionalism, but it would be folly to deny that humans are something radically new in the history of planet earth, even if it was relatively modest biological changes that finally pushed us over the brink into general intelligence and civilisation.

and also puts doubt on the idea that brains of language-using species have a special innate capacity to learn language at all. Rather, i would simply be related to a brain being complex and structured correctly

Isn't 'structured correctly' pretty much the same thing? Humans are evolved to have a specialised capability for language use, a capability which we ourselves repurpose for many other uses. For example abstract maths relies heavily on the same brain areas used for grammar processing; there is in fact a vocal minority of cognitive scientists who believe that general intelligence as a whole developed as a consequence of language proficiency. Parrots have a specialised capability to do something different (mimic and modify very complex songs), which happens to be close enough to language use that it can be adapted to provide a very crude language proficiency. Very few species seem to be 'structured correctly', although parrots do provide a good counterexample to the hypothesis that 'brains must be at least the size of humans to use language'.

(presumably parrots thus being on the edge of humanlike language competence so that only the 'smartest' parrots can speak human languages, as you said yourself.)

I am uncomfortable with that wording because 'on the edge of' suggests that if parrots evolved a little more, they would be able to use human language fluently. This is almost certainly not the case, because the fitness pressure on parrots is ultimately targeting a different capability, and isn't likely to change as long as parrots lack manipulators and have limited opportunities for tool use and specialisation of labor. It may well be that parrots are as close to humanlike langauge as animals are likely to get, without having the other human characteristics (tool use and tribal society) that made it a specific fitness goal.

Also, if learning, repeating, and modifying sounds and sets of sounds with complex rules doesn't sound like a description of how syntax and grammar works, I'll be damned. For all we know, human language was orignially repurposed from some similar learning-repeating-modifying sound system used for god knows what.

We know what the original precursor for human language looked like; it looked like the vocalisations that chimpanzees make. These are much less sophisticated than what parrots are capable of. I think this actually goes against your point, because human language clearly did not require the millions of years of refinement that songbird vocalisation had as a precursor. It actually developed very quickly from relatively rudimentary capabilities, as soon as the right environment came along to make it a priority (and of course, which simultaneously supported the other brain developments that work together with language, such as a very refined other-person-model).

Of course evolution is like that, unpredictable, progressing in fits and starts, following bizarre incremental paths, stagnating for many millenia and then rushing ahead when the environmental pressure is just right. That's why I don't like simulated evolution as a design tool.

I wouldn't count birds out just because they use the same part of their brain for singing as speaking.

Certainly I would say that were a species of parrot or corvid suddenly to develop the manipulative abilities and social arrangements of chimpanzees (they already have near-equivalent abstract reasoning), their superior vocal abilities would give them a big head start on evolving human equivalent capabilities. Though it would still be no means a sure thing; natural selection loves to get stuck in bizarre local optima.

Obviously if I could create giant ravens (say, Haarst's Eagle sized) with human intelligence and spindly manipulative arms (e.g. modified second set of legs), via fell genetic engineering, I would do so, both because it would be a fascinating experiment and because I am an evil genius who longs to replace man with a superior breed of hideous creations.

[Erik von Nein]

Crows are certainly capable tool makers and tool users. If you watched the video I provided there is an example of a New Caledonean crow creating a tool the researchers never demonstrated a use for. They had left part of their initial test with that particular crow on accident, then were amazed when the bird fashioned a tool for the purpose of getting food. Also, the birds are capable of learning from each other, parents teaching their offspring, rivals copying rivals, and so on.

Also, New Caledonean crows did demonstrate a capacity for casual thought. I'll look around a bit, but the article was posted in SLAM last year, I believe.

I'm not sure how much this would relate to parrots and other song birds as I'm unaware of their tool making efficacy. Their calls are learned and are, depending on the species, unique to the individual. Their language, therefor, is always changing. They even have specific calls for specific reasons, either for finding a fruit tree that's fruiting and calling to others of their species, calling for danger, calling for specific individuals, and such. It's all dependent on the species you're referring to. I imagine, were they to pick up human speech and be able to manipulate it themselves they probably would be able to, but you would need to demonstrate what different words meant for different situations.

In any event, further study and experimentation would be necessary.

Then there's the issue of cephalopod intelligence. They've demonstrated remarkable learning capacity, among other traits.

[XaLEv]

Parrots certainly have limited manipulation ability compared to large primates, but they are not completely without such. They have grasping feet and their beaks and tongues are extremely dexterous and precise. For example, I have watched a parrot, up close, prepare a sunflower seed for consumption by removing and discarding the hull, then scraping off and discarding the surface of the kernel in one piece, entirely with the beak and tongue. They can and do use their feet to hold or pick up objects, whether food, toys or tools.

[TithonusSyndrome]

If manipulation and speech are necessary to let intelligence bloom into full-on sapience, then elephants definitely qualify - hell, their speech is an extension of their manipulation skills, even before we take into consideration things like them washing off roots and shrubs they uproot in pools of water before eating them. I don't doubt that some birds are especially capable of manipulation and tool use, but there's only so far a beak and talons can take you, and it's not near as far as something as dexterous and strong as a trunk can go.

[Duckie]

Well, the human brain is certainly exceptional, compared to other organisms evolved on planet Earth. The term 'unique' is problematic. Firstly, language is not a binary property, there is a spectrum of communication flexibility from the simplest chemical signals used by bacterial colonies up to the mind-state transfers of hypothetical future transhumans. Unfortunately, as so often in evolved organisms we can't see all of the intermediate steps.

This is true, and an unfortunate lack in linguistics- we have trouble enough comprehending how to classify and work with human language, let alone what to consider language or not language in the same way biologists struggled to find a definition of life.

Secondly, language is inseparable from its social context. Parrots don't have the means to invent human-like language on their own, whereas human children left without adult guidance will invent their own language from scratch. I agree that a lot of Chomskyites are far too arrogant with regard to human exceptionalism, but it would be folly to deny that humans are something radically new in the history of planet earth, even if it was relatively modest biological changes that finally pushed us over the brink into general intelligence and civilisation.

Oh, certainly, but I'd say it's a difference of mental scale rather than a sheer jump. There's no reason to believe human brains are the only thing to have what we refer to as language and everything else is a 0 on the scale, although I don't suggest you believe that. (In contrast, however, some Chomskyites appear to.)

Isn't 'structured correctly' pretty much the same thing? Humans are evolved to have a specialised capability for language use, a capability which we ourselves repurpose for many other uses. For example abstract maths relies heavily on the same brain areas used for grammar processing; there is in fact a vocal minority of cognitive scientists who believe that general intelligence as a whole developed as a consequence of language proficiency.

Well, yes, although what I was going for again was showing a fluid scale rather than a jump. You can have something better and better structured until it reaches the humanlike point. That's not to say we should expect anything to have close to human language, because I would posit that somewhere in our brains', throats', etc. development we hit a singularity point wherein what we recognize today as language suddenly appeared, for the same reasons as you below describe human children spontaneously producing language (I assume you refer to Nigaraguan Sign Language?)

Parrots have a specialised capability to do something different (mimic and modify very complex songs), which happens to be close enough to language use that it can be adapted to provide a very crude language proficiency. Very few species seem to be 'structured correctly', although parrots do provide a good counterexample to the hypothesis that 'brains must be at least the size of humans to use language'.

Incidentally, I wonder if certain languages are easier for parrots to learn than others? If they're mimicking and modifying sounds, I'd assume ablaut or lentition or the like would be potentially easier for them. Then again, if their minds have trouble with complex grammar, you'd expect isolating languages to be a lot easier than mostly isolating partially synthetic fusional partially agglutinating monstrosities like english. Although this assumes that their brains work anything like I'm assuming it does (a machine which takes roots and applies permutations to produce the desired meaning).

I am uncomfortable with that wording because 'on the edge of' suggests that if parrots evolved a little more, they would be able to use human language fluently. This is almost certainly not the case, because the fitness pressure on parrots is ultimately targeting a different capability, and isn't likely to change as long as parrots lack manipulators and have limited opportunities for tool use and specialisation of labor. It may well be that parrots are as close to humanlike langauge as animals are likely to get, without having the other human characteristics (tool use and tribal society) that made it a specific fitness goal.

Oh, of course. I didn't mean to imply that parrots were going to 'evolve just a little more'. I just meant that if you were to define human-like language skills, parrots would be a case but a marginal case.

We know what the original precursor for human language looked like; it looked like the vocalisations that chimpanzees make. These are much less sophisticated than what parrots are capable of. I think this actually goes against your point, because human language clearly did not require the millions of years of refinement that songbird vocalisation had as a precursor. It actually developed very quickly from relatively rudimentary capabilities, as soon as the right environment came along to make it a priority (and of course, which simultaneously supported the other brain developments that work together with language, such as a very refined other-person-model).

Of course evolution is like that, unpredictable, progressing in fits and starts, following bizarre incremental paths, stagnating for many millenia and then rushing ahead when the environmental pressure is just right. That's why I don't like simulated evolution as a design tool.

True, but what I meant was, songbird grammar is at least superficially similar (if less usable in general cases and for abstract thought) to humanlike grammar if it consists of taking sounds or strings of sounds, putting them together in certain orders, and modifying them based on intended meaning or mere sandhi.

[Starglider]

An important, possibly critical factor in the development of human intelligence was the fact that (all other things being equal) human females prefer intelligent mates. Mastery of language (e.g. verbal humor) is an obvious sign of this, though social maneuvering/success at tribal politics is just as important. Sexual selection for higher intelligence transformed a relatively moderate environmental push into a vicious red queen's race - again there are quite a few cognitive scientists who attribute the lions share of human cognitive evolution to this effect. This would imply a very important question with regard to any other species that might be in the running for general intelligence; how well can they recognise high intelligence in potential mates and how strongly do they select for it? Sexual selection criteria are unpredictable and often arbitrary even by the already chaotic standards of biological evolution, so a substantial amount of luck was likely involved in getting the rapid rise of (proto-)human intelligence started.