Evolution of Intelligence: A Comparison

[lpetrich]

Grow Smart and Die Young: Why Did Cephalopods Evolve Intelligence? - ScienceDirect Cross-species comparisons can be difficult, but it is generally agreed that octopuses are the most intelligent invertebrates. So how do octopuses stack up against more familiar high-intelligence species? That is what this article is about, and that may give some hints as to what we may expect elsewhere in the Universe.

The article compared cephalopods, corvids, cetaceans, and apes. Cephalopods: nautilus, squid, octopus, cuttlefish. Corvids: crows and their close relatives. Cetaceans: dolphins, porpoises, and whales. Apes: including humanity.

It also discussed various theories for the origin of intelligence: ecological, social, and predator-avoidance. Ecological is having to cope with a challenging environment, like finding food that is difficult to find. Social is managing relations with one's group members. Predator avoidance is as it says.

The article had an interesting comparison table, and I will attempt to reproduce it here.

• Body form
• Grasping organs -- Octo: tentacles, Corv: beak, feet, Ceta: snout, Apes: hands
• Dexterity -- Octo: high, Corv: high, Ceta: limited, Apes: high
• Senses
• Vision -- Octo, Corv, Ceta, Apes: good
• Hearing -- Octo: limited, Corv, Ceta, Apes: good
• Chemical -- Octo: good, Corv, Ceta, Apes: limited
• Brain
• Brain/body size ratio -- Octo, Corv, Ceta, Apes: high (at least relative to close relatives)
• Cognition subtrate: Octo: vertical lobe, Corv: nidopallium, Ceta, Apes: cortex
• Ecology
• Lifestyle: Octo: aquatic, Corv: flight, Ceta: aquatic, Apes: land, trees
• Diet (-vores): Octo: carni, Corv: omni, Ceta: carni, Apes: omni
• Extractive foraging: Octo: yes, Corv: yes, Ceta: no, Apes: yes
• Predation vulnerability: Octo: high, Corv, Ceta, Apes: limited
• Social life
• System: Octo: solitary, Corv: pairs, groups, Ceta: family groups, Apes: groups
• Long-term bonds: Octo: absent, Corv: with pair mate, Ceta, Apes: with multiple individuals
• Behavioral flexibility
• Find and process food: Octo, Corv, Ceta, Apes: high
• Social interactions: Octo: low, Corv, Ceta, Apes: high
• Predator avoidance: Octo: high, Corv, Ceta, Apes: ?
• Life history
• Lifespan: Octo: a few years, Corv: >15 yrs, Ceta, Apes: >40 yrs
• Parental care: Octo: brooding of eggs, Corv, Ceta, Apes: yes
• Reproduction: Octo: semelparity, Corv, Ceta, Apes: iteroparity

Here are the kinds of reproduction. Semelparity is reproducing only once, at the end of one's life. A mother octopus will guard her eggs and not eat, and will die around when they hatch. A father octopus usually dies a few weeks after mating. Iteroparity is reproducing several times, complete with usually surviving doing so.

Corvids, cetaceans, and apes are capable of transmitting learned information down the generations. Octopuses are incapable of that. They are solitary, and they don't take care of their baby ones after they hatch.

All have good vision, all but octopuses have good hearing, but only octopuses have very good chemical senses -- they have taste buds on their tentacles. Octopuses, corvids, and apes have good manipulative organs and they are very dextrous with those body parts. Cetaceans are much more limited.

As to diet, the aquatic ones are carnivores, because that is the easiest kind of food to get. The land/tree/aerial ones are all omnivores. All but cetaceans are known to extract food items in various ways, like crows and chimps cracking nuts in various ways and octopuses pulling clams open.

Finally, octopuses are very vulnerable to predation while corvids, cetaceans, and apes are much less vulnerable.

[lpetrich]

By looking at that list, we can try to guess which of our features would be typical of a sentient species elsewhere in the Universe.

A big brain is essential. However, on our planet, the parts for advanced thinking evolved separately at least three times. They would likely be very social, and also iteroparous, reproducing multiple times and surviving their reproducing.

Of the senses, vision must be well-developed, and likely also hearing. Chemical senses are less important. Vision is very demanding of processing power, and it provides an incentive to grow large brains. Chemical senses don't need nearly as much.

Manipulative organs are very desirable, occurring in three of the four groups considered. Among other things, they would be used for extracting food items. As to what they eat, they would likely be omnivores or else generalist carnivores, as octopuses are.

Their low-tech ancestors would likely have had limited risk from predators.

[Tom Mazanec]

What about predators? Could carnivores be likely to evolve intelligence?
I heard speculations that raccoons are a possible successor species to man.

[profloater]

An obvious difference is marine versus land, marine being a 3D environment with dark depths. No advantage on land until plants have got a hold. The marine food chain as i guss always based on small animals, while land used plants mainly andthen predators arriived to eat smaller anmals. In general predators need more intelligence and learned behaviour while prey need reflexes and armour or speed.

[Darrell]

What about predators? Could carnivores be likely to evolve intelligence?
I heard speculations that raccoons are a possible successor species to man.

Carnivores could. Arguably they have. For example, orcas and dolphins are both carnivores and are considered to be highly intelligent. Of course, they are both also highly social species and current thinking is that sociability requires higher intelligence. The idea is that sociability confers higher fitness (in evolutionary terms).

Raccoons are omnivores, like us. More like garbage guts. They eat pretty much anything.

[Darrell]

In general predators need more intelligence and learned behaviour while prey need reflexes and armour or speed.

Or intelligence to avoid or collectively guard against predators.

[Tom Mazanec]

Solitary or social can evolve rather easily. Half mongoose species are social, half solitary.
Lions are social but tigers are solitary, yet they are closely related enough to hybridize.
Polecats are solitary and ferrets are social, but they are more closely related than dogs and wolves.
I understand foxes were social till human predation altered them ( a group of foxes was called a 'skulk').

[profloater]

If this is about guessing alien species, the ability to cooperate seems essential for space flight and some believe the ability to imagine higher powers is necessary for the scale of cooperation.

[Darrell]

Note that all of the groups compared in this study are highly social, except for cephalopods. Surely there are many drivers of intelligence. But there is a pretty strong correlation between high intelligence and high sociability. Given the very limited data that we have to try and guess what intelligent alien life might be like, that they are highly social might be one of the few characteristics to bet on that we might be able to justify thinking has odds better than chance.

[profloater]

Humans, an intelligent , highly social species is very dangerous to most other species as well as other humans, so personally I hope aliens stay in their own solar systems, no matter how many arms thay have. Octopuses have distributed brains in their tentacles which counts as a social group I guess, having to get eight arms to agree on any action might limit their social performance. They need to get past eating each other too. I imagine an ant like species with advanced communication could be very effective, if radio based telepathy can evolve, ants would be formidable, or is that formicable?

[lpetrich]

What about predators? Could carnivores be likely to evolve intelligence?

Yes, from the ecological theory, that challenging environments and lifestyles can provoke the evolution of a lot of brainpower.

I heard speculations that raccoons are a possible successor species to man.

Raccoons are omnivores. They also have relatively long fingers relative to fellow carnivorans.

[lpetrich]

Note that all of the groups compared in this study are highly social, except for cephalopods. Surely there are many drivers of intelligence. But there is a pretty strong correlation between high intelligence and high sociability. Given the very limited data that we have to try and guess what intelligent alien life might be like, that they are highly social might be one of the few characteristics to bet on that we might be able to justify thinking has odds better than chance.

I agree. In fact, anthropologist Robin Dunbar has proposed the "social brain theory" of intelligence, that large brains are useful for managing interactions with fellow group members.

Living in groups can provide greater force as a result of collective effort, like Chimps Attacking Leopard - YouTube. That can greatly reduce vulnerability to predators, another feature listed.

Sociality is also good for transmitting learned information down the generations, so one does not have to learn everything from scratch. Parental care is another form of cross-generation assistance.

[swampyankee]

I wonder how one could actually test the hypothesis "predators are going to tend to be more intelligent than herbivores." Lab testing will tend to be done with domesticated animals (or lab rodents), where the herbivores were more likely bred for docility and their general utility as food sources or draft animals and carnivores were bred for comparatively complex tasks, like herding or assisting hunters.

Of course, a second (and perhaps primary) issue is that "intelligence" doesn't seem to be that well-defined as a concept.

[profloater]

One aspect that is defined is theory of mind ie the realisation that another has choices and then a prediction of which choice is likely. Lions and doga and killer whales have shown that kind of thinking to improve their hunting while prey animals just run away, a very simple instinct. They show fight or flight responses but rarely any plan to trap or trick a predator.

[Darrell]

One aspect that is defined is theory of mind ie the realisation that another has choices and then a prediction of which choice is likely. Lions and doga and killer whales have shown that kind of thinking to improve their hunting while prey animals just run away, a very simple instinct. They show fight or flight responses but rarely any plan to trap or trick a predator.

It doesn't seem to be that clear cut. Elephants, highly intelligent, are herbivores. Gorillas, though not evidently as intelligent as chimps / bonobos or orangutans, are also highly intelligent. Parrots, high on any intelligence list, do eat insects as well as fruits, nuts and other plant stuff, but they aren't really predators. Considering full time predators / obligate carnivores, all of the highly intelligent ones are sea mammals.

Meanwhile, full time predators on land don't make the top of the list, though many are relatively intelligent. It seems to me that perhaps predation does entail a certain level of intelligence but not high intelligence and not necessarily more so than herbivory. For high intelligence, near to human and above intelligence, I don't think the type of diet a creature evolved to meet its nutrient requirements is a very reliable indicator. A better indicator seems to be how social the creature is. With respect to the most intelligent species on Earth a trait that they all have in common including herbivores, carnivores and omnivores, is (sorry to repeat it again) that they are all highly social.

[profloater]

It doesn't seem to be that clear cut. Elephants, highly intelligent, are herbivores. Gorillas, though not evidently as intelligent as chimps / bonobos or orangutans, are also highly intelligent. Parrots, high on any intelligence list, do eat insects as well as fruits, nuts and other plant stuff, but they aren't really predators. Considering full time predators / obligate carnivores, all of the highly intelligent ones are sea mammals.

Meanwhile, full time predators on land don't make the top of the list, though many are relatively intelligent. It seems to me that perhaps predation does entail a certain level of intelligence but not high intelligence and not necessarily more so than herbivory. For high intelligence, near to human and above intelligence, I don't think the type of diet a creature evolved to meet its nutrient requirements is a very reliable indicator. A better indicator seems to be how social the creature is. With respect to the most intelligent species on Earth a trait that they all have in common including herbivores, carnivores and omnivores, is (sorry to repeat it again) that they are all highly social.

I agree that social skills do score high as drive for intelligence. The switch from alpha male structure can beseen as an evolution, ie that secondary males need clever strategies to reproduce. We see that in animals and therefore a parallel route to evolution, brains and brawn. Its also true that large social groups help both predator and prey. But then the environment counts, where food is rare solitary smart hunters, polar bears, seals, fish, larvae, a heirarchy of smarts.

[Tom Mazanec]

How about rats or beavers, other species I have heard suggested as promising?

[Colin Robinson]

I wonder how one could actually test the hypothesis "predators are going to tend to be more intelligent than herbivores." Lab testing will tend to be done with domesticated animals (or lab rodents), where the herbivores were more likely bred for docility and their general utility as food sources or draft animals and carnivores were bred for comparatively complex tasks, like herding or assisting hunters.

While it's true that dogs have been bred for herding and assisting hunters, isn't that because they have a natural aptitude for those tasks?

E.g The things sheepdogs do when they round up sheep are similar to the behaviours of a pack of wolves rounding up prey animals.

Of course, a second (and perhaps primary) issue is that "intelligence" doesn't seem to be that well-defined as a concept.

True. Right now, there seems to be more talk about intelligence of passerines and cephalopods than about intelligence of dogs and wolves... but is that because canines are less smart than parrots and octopus, or is it simply because familiarity breeds contempt?

[Colin Robinson]

It doesn't seem to be that clear cut. Elephants, highly intelligent, are herbivores. Gorillas, though not evidently as intelligent as chimps / bonobos or orangutans, are also highly intelligent.

But the smartest of the great ape species — the human species — is also the great ape whose diet includes most flesh.

Parrots, high on any intelligence list, do eat insects as well as fruits, nuts and other plant stuff, but they aren't really predators. Considering full time predators / obligate carnivores, all of the highly intelligent ones are sea mammals.

Meanwhile, full time predators on land don't make the top of the list, though many are relatively intelligent. It seems to me that perhaps predation does entail a certain level of intelligence but not high intelligence and not necessarily more so than herbivory. For high intelligence, near to human and above intelligence, I don't think the type of diet a creature evolved to meet its nutrient requirements is a very reliable indicator. A better indicator seems to be how social the creature is. With respect to the most intelligent species on Earth a trait that they all have in common including herbivores, carnivores and omnivores, is (sorry to repeat it again) that they are all highly social.

What was the title of the paper cited in this thread's OP? "Grow Smart and Die Young: Why Did Cephalopods Evolve Intelligence?" Cephalopods are not highly social...

[swampyankee]

While it's true that dogs have been bred for herding and assisting hunters, isn't that because they have a natural aptitude for those tasks?

E.g The things sheepdogs do when they round up sheep are similar to the behaviours of a pack of wolves rounding up prey animals.

I'd not argue against herding being an extension of natural predatory behaviors of ancestors to dogs, but it is a behavior that requires more intelligence than being shorn or milked.

True. Right now, there seems to be more talk about intelligence of passerines and cephalopods than about intelligence of dogs and wolves... but is that because canines are less smart than parrots and octopus, or is it simply because familiarity breeds contempt?

Another possibility is that the intelligence of cephalopods and passerines was unexpected. I suspect that serious scientific investigation into the intelligence of animals is fairly recent, not much more than the past fifty years or so, and is still very much in flux.

[Darrell]

But the smartest of the great ape species — the human species — is also the great ape whose diet includes most flesh.

Is it then your hypothesis that a higher meat diet corresponds with greater intelligence? If so then why are humans not obligate carnivores? How to explain that some of the most intelligent animals on the planet other than humans have less, or even 0 animal protein in their diets?

What was the title of the paper cited in this thread's OP? "Grow Smart and Die Young: Why Did Cephalopods Evolve Intelligence?" Cephalopods are not highly social...

That's why up at comment #9 I wrote, "Note that all of the groups compared in this study are highly social, except for cephalopods."

[Darrell]

While it's true that dogs have been bred for herding and assisting hunters, isn't that because they have a natural aptitude for those tasks?

E.g The things sheepdogs do when they round up sheep are similar to the behaviours of a pack of wolves rounding up prey animals.



True. Right now, there seems to be more talk about intelligence of passerines and cephalopods than about intelligence of dogs and wolves... but is that because canines are less smart than parrots and octopus, or is it simply because familiarity breeds contempt?

Based on the current state of research in animal intelligence, yes, canines are not as intelligent as some parrots and corvids. I think it's less clear where cephalopods fall. They're harder to study whereas dogs are easy to study and have been studied a lot.

Dogs, though very interestingly not wolves, respond to humans very well. They "read" human non-verbal communication cues very well. This clearly requires a certain level of intelligence. It may also bias people towards a more favorable assessment of dog intelligence compared to other animals.

[profloater]

Based on the current state of research in animal intelligence, yes, canines are not as intelligent as some parrots and corvids. I think it's less clear where cephalopods fall. They're harder to study whereas dogs are easy to study and have been studied a lot.

Dogs, though very interestingly not wolves, respond to humans very well. They "read" human non-verbal communication cues very well. This clearly requires a certain level of intelligence. It may also bias people towards a more favorable assessment of dog intelligence compared to other animals.

That raises emotional intelligence which is strong in dogs but weak in parrots, in my experience of living with both. Flying must require an extra level of decision taking compared with running about on land but socialising effectively does imply a learned model working on top of instinct. Intelligence is already analysed as haveing several vectors such as navigation, mental aithmetic, physical coordination, language and so on, reading other people includes both analysiis and empathy. If wewant to compare species we need to subdivide intelligence and score each vector.

[Colin Robinson]

Is it then your hypothesis that a higher meat diet corresponds with greater intelligence?

I would certainly argue that the intelligence difference between humans and chimps is related to the difference in diet, the fact that we humans eat more meat (including fish) than chimps eat. In general, catching animals requires more intelligence than finding and consuming edible plants, which is why wolves have to be more intelligent than deer, and fish-eating sea-mammals such as dolphins have to be more intelligent than the weed-eating dugong.

But I would NOT argue that there is a simple, linear correspondence between the amount of meat a species eats and how smart it is.

If so then why are humans not obligate carnivores?

Presumably because, as omnivores, primitive humans developed a more varied set of skills than an obligate carnivore. The more varied your skill-set, the more sophisticated your brain needs to be.

How to explain that some of the most intelligent animals on the planet other than humans have less, or even 0 animal protein in their diets?

Because it's not animal protein as such that makes the difference, it's what you have to do to get hold of it.

Although catching animals takes more smarts than grazing on grasses and leaves, there are also non-animal foods which are more challenging than grass... For instance, the sort of manipulative intelligence which chimps and corvids use to crack nuts open is not so different from the manipulative intelligence which octopus use to crack open shellfish.

[profloater]

It issaid there is a reverse link. We tamed fire, changed the way we eat, changed our vocal geometry allowing better speech. Language promoted social skills, required bigger cortex, virtuous circle of skills and more time to change our environment. So it comes down to cooking.

[Darrell]

Colin,

As I said earlier there are certainly many drivers of intelligence and how to get food is certainly one of them. But there are many exceptions to your examples. There are plenty of "stupid" predators that are nevertheless very successful and plenty of "stupid" animals that do clever seeming things, such as ants that harvest plants in order to grow fungus which they then eat, but that are not particularly intelligent. There are also some quite intelligent herbivores. There is a trait that all of the most intelligent species on the planet do share whether they are predators or not.

[Colin Robinson]

Colin,

As I said earlier there are certainly many drivers of intelligence and how to get food is certainly one of them. But there are many exceptions to your examples. There are plenty of "stupid" predators that are nevertheless very successful and plenty of "stupid" animals that do clever seeming things, such as ants that harvest plants in order to grow fungus which they then eat, but that are not particularly intelligent. There are also some quite intelligent herbivores. There is a trait that all of the most intelligent species on the planet do share whether they are predators or not.

Glad you mentioned ants, they're very relevant to this discussion, for a number of reasons. They're not only one of the most social classes of insects, they are also versatile omnivores and consumers of animal protein... What makes you think they're "not particularly intelligent"?

[Colin Robinson]

That raises emotional intelligence which is strong in dogs but weak in parrots, in my experience of living with both. Flying must require an extra level of decision taking compared with running about on land but socialising effectively does imply a learned model working on top of instinct. Intelligence is already analysed as haveing several vectors such as navigation, mental aithmetic, physical coordination, language and so on, reading other people includes both analysiis and empathy. If wewant to compare species we need to subdivide intelligence and score each vector.

I agree. Too often we speak of intelligence in ways which don't recognise its diverse components. As if it were a single, unitary thing, which a species either has or doesn't have...

[Darrell]

What makes you think they're "not particularly intelligent"?

Because despite some very interesting behaviors, they aren't. If we aren't talking about intelligence of the sort that enables human level activities, or at least enticingly close as with the top 8 or so most intelligent animals on the planet besides us, then we are talking about different things. Despite the hyperbole sometimes employed, which is not necessary because their behaviors are genuinely fascinating, ants are not remotely on par intelligence-wise with humans, other apes, elephants, corvids, dolphins, etc. If we want to say in this context that ants are intelligent then we need to widen the field to darn near the entire animal kingdom.

[Colin Robinson]

Because despite some very interesting behaviors, they aren't. If we aren't talking about intelligence of the sort that enables human level activities, or at least enticingly close as with the top 8 or so most intelligent animals on the planet besides us, then we are talking about different things. Despite the hyperbole sometimes employed, which is not necessary because their behaviors are genuinely fascinating,

Besides fascinating behaviour, ants have very large brains in proportion to their size.

Quoting from the Wikipedia entry on Brain-to-body mass ratio... "some ant species have 14%-15% of their mass in their brains, the highest value known for any animal".

ants are not remotely on par intelligence-wise with humans, other apes, elephants, corvids, dolphins, etc.

You keep forgetting about the cephalopods, which were the starting point of this thread... Do you think octopus fit the bill for "human level activities, or at least enticingly close"?

If we want to say in this context that ants are intelligent then we need to widen the field to darn near the entire animal kingdom.

I see intelligence as a comparative thing, rather than as an either/or. It's true that the animal kingdom generally shows a level of intelligence not seen in other living things, such as plants, fungi, and slime moulds...

But there are classes of animals which have evolved larger brains and more sophisticated behaviour than their relatives.

For instance, it seems clear that cephalopods are the most intelligent of the molluscs, which makes them especially worthy of study if we want to know how intelligence evolves.

And ants seem to be the most intelligent of the arthropods (insects and crustaceans)...