Dolphins Are Intelligent!

Where does one start with debunking fallacies in this little article? Oy vey!

Dolphins and whales are dumber than goldfish and don't have the know-how to match a rat, new research from South Africa shows. For years, humans have assumed the large brains of dolphins meant the mammals were highly intelligent.

No, we knew dolphins were smart millenia before we ever looked at their brains. The ancient Chinese knew it. Aristotle knew it. And the idea that brain size has anything to do with intelligence is, like, sooo 19th century.

Paul Manger from Johannesburg's University of the Witwatersrand, however, says it is not intelligence that created the dolphin super-brain -- it's the cold. To survive underwater, these warm-blooded animals developed brains that have a lot of insulating material -- called glia -- but not too many neurons, the gray stuff that counts for reasoned thinking.

Wow! Since when are glia "insulating material"? A few years ago, for my Neuroscience class, I had to remember at least 10 functions of glia - not one of them having anything to do with insulation, or even structural support. It's all about function - neurons and glia work together to process information. Anyway, I will blame this on the stupidity of the reporter as I doubt that anyone with such archaic ideas would ever be allowed to dissect a dolphin and publish a study in a decent journal.

Yet while dolphins aren't as smart as people tend to think, they are as happy as they seem. Manger said dolphins have a ''huge amount'' of serotonin in their brains, which is what he described as ''the happy drug.''

Sure, if you get your science from Cosmo and Glamour. Do I really have to start listing all the functions of serotonin now? Or try to define "happiness" in such simplistic terms that it can be explained with a single chemical?

It is not quite clear, but it appears that Alon Levy agrees with the study. But Lindsay is having none of it. She cites the self-recognition paper as well as some personal testimony of the researcher who did that study. When that paper came out I was teaching a "Readings in Behavioral Biology" graduate seminar and all the neuro faculty showed up for class and tried valiantly to destroy the paper - with no avail. It is good.

Dolphins are darn smart. They play (check this pdf). They have complex communication and complex social interactions.

So, how does this kind of argument ever show up? Because of anthropocentrism. Two types of anthropocentrism, to be precise.

First, the concept of "intelligence" is often defined in human-like terms. If an animal can do stuff we do, it is deeemed smart. If it can be easily trained like our immature offspring can, it is smart. If it can talk, it is smart. If it builds structures, it is smart. BS. Intelligence has to be defined from the vantage point of that species: what makes ecological and evolutionary sense for that species to be able to do. Bees are smarter than ants because they have a more sophisticated ability to orient in space and time, not because they speak English, French and Chinese.

Now, don't get me wrong now. Since we are intelligent, looking for intelligence in other animals may benefit from comparison to humans. The trouble is, people go for specifics of human capabilities, instead of a general idea what intelligence is.

Writing "Hamlet" is an ecologically relevant ability for humans. It kept old Will fed and clothed for a few months, after which he wrote the next play. Why would an insect need to write theater plays? It is not ecologically relevant to it. It does not aid survival and/or reproduction.

Intelligence is the ability to learn fast and learn a lot of pieces of information relevant to one's ecology. It is the ability to hold many of those pieces in one's mind simultaneously, to juggle them and analyse them and notice patterns. It is the ability to play with that information, to get new ideas and test them, to note and remember the results of those tests. It is the ability to use this novel informaiton to invent novel behaviors - doing different stuff at different places at different times. In short, intelligence is the ability to do science! Behavioral flexibility is the hallmark of intelligence - not the specific types of behaviors.

The second anthoropomorphism considers the underlying anatomy. Why should unrelated species of high intelligence have brains similar to us? They evolved their high intelligence at different times, in a different lineage, with different raw materials to work with, and under different ecological pressures, for different purposes.

Many birds are very intelligent - but in their own way. Clarke's Nutcrackers, African Grey Parrots, pigeons, and most corvids (ravens, crows, jays) are highly intelligent creatures with huge capabilities for episodic memory (remembering spatial and temporal aspects of personal experiences), play, problem-solving, spatial orientation and perhaps even insight (planning for the future). And their brains look nothing like ours.

Octopus is a very smart animal. Its brain looks nothing like ours.

Macs and PCs can do all the same stuff (roughly), but look nothing like each other under the hood. Many kinds of harware can run the same kinds of software and do same kinds of things, so why should brains have to be all built the same way in order to make an animal "intelligent"?

So, leave the dolphins alone, at least until the Startide Rising.

Addendum: I forgot to note that glia are not white matter. Axons are white matter while neuronal bodies are grey matter. Glia surround both. It is the color of Schwann cells (a type of glia) that makes axons look whitish.

Thus, more grey matter means more neurons. More white matters means more connections. What is more important: gazillions of scattered cells, or the complexity of their connections? I'd say connections.

Addendum II: Dave Munger wrote a valid criticism of what I wrote here (and somehow I missed his earlier post on this subject):

I agree that intelligence is tremendously difficult to define, but I'd suggest that the perspective of an individual species is a poor place to start. Based on that notion, every organism can be said to be intelligent, because every organism is highly adapted to its environment. When we say an animal is "intelligent," we're defining intelligence from our own perspective: the point is to identify animals that are similar to ourselves.

I'm not sure that the point is to identify animals that are similar to ourselves, but even if it is, similar in what way? The general mental capabilities (that we still need to define) or specific capabilities (which I argued here against)?

As for looking at each species individually, I agree that it is impossible to do it in isolation, but eahc species can be compared with other species in its own group, e.g., birds with birds, insects with insect, and then broader, all with all. If we define, provisionally, intelligence as fast learning, high processing power and flexibility of behavior, then we can compare species without looking at specific items that are learned, specific informaiton that is processed and specific behaviors that are flexible. For some species, being inflexible is a great adaptive trait - doing everything by the pre-programed schedule can work wonderfully for a long period of time. Other species evolve flexibility which allows them to spread on a broader spatial range and perhaps allow them to survive a longer geological time.


More like this

Thanks for all the links.

But what you say about white matter confuses me. I thought white matter was myelinated axons -- not all axons are myelinated. But some are, meaning they've been wrapped around by a type of glial cell, as you say. So the white is the insulating cell of a glial cell, but inside is a high-speed axon.

So over at Majikthise, when I read people talking about glial cells and insulation, I thought someone might have confused the axon-electrical insulating properties of some glial cells with thermal insulation, and commented accordingly.

Wow! Since when are glia "insulating material"? A few years ago, for my Neuroscience class, I had to remember at least 10 functions of glia - not one of them having anything to do with insulation, or even structural support. It's all about function - neurons and glia work together to process information.

Glia as insulation appears to be proposed by the study. From the abstract:

A novel hypothesis regarding the evolution of large brain size in cetaceans is put forward. It is shown that a combination of an unusually high number of glial cells and unihemispheric sleep phenomenology make the cetacean brain an efficient thermogenetic organ, which is needed to counteract heat loss to the water.

So I guess the answer to your question is, "since now."

Also, as far as I know glia cells are not thought to play a primary role in cognition, but rather a support role.

But you're right that the Sun-Times article is trash. Particularly the "dumber than goldfish" line, which is not attributed to anyone.

By Eric Wallace (not verified) on 28 Aug 2006 #permalink

Without looking at the paper, it's difficult to discern just how much of the weirdness of Manger's conclusion are intrinsic and how much is a by-product of the shit lay press. This piece on may be even worse than the Sun-Times one -- I especially like the repeated punny references to "fish" in the sub-heads despite the reporter's being aware, at least fleetingly, that dolphins are mammals.

What I don't get is how Manger or anyone can dispose of all sorts of observational evidence of dolphins' high intelligence on the basis of anatomical findings. That's like saying that a car that's been clocked at 190 MPH can't possibly go that fast because it's only got an eight-cylinder engine in it.

The bit about serotonin is just lunkheaded. Why not just say that because LSD and other hallucinogenics act via 5-HT receptors that doplhins are, like, permanently trippin', dudes? Sheeit.

Macs and PCs are a bad metaphor. They look the same under the hood, even moreso now that macs have migrated to Intel hardware. Superficially (software-wise) they look far more distinctive structurally than they actually are, either from each other, from a console game system, or from a remote control. Contrary to the implications of your metaphor, all commonly-used computing systems operate in EXACTLY the same manner, a manner dictated a very long time ago.

The advances in computing for the past 70 or 80 years have been largely dictated by advances in materials and fabrication, not much in functionality or structure, and cannot be compared to the differences in the amount of neurons in a dolphin's brain except as a counterexample. In my opinion, for those brains to have comparable power to human brains that have far more neurons and fewer glia would imply a structural difference.

It would be as if someone had showed me a computer that had had its CPU speed reduced fivefold and its number of internal registers increased tenfold, but it offered a comparable performance to normal architectures. I would be astounded, and prepared to learn a new theory of computing which may not even seem parsimonious with the old. Without the new theory, I would need very good evidence of the claim of equivalent performance.

Meaning: without a theory of how brains work which doesn't depend on the number of neurons, but could trade neurons for glia without losing power, we would need definite external evidence for the intelligence of dolphins to continue to make the claim. Otherwise the claim is virtually religious.

The article offers examples where dolphins seem to act stupid. You offer examples where dolphins seem to act smart. Both of these arguments are anthropocentric. A claim of an animal being intelligent in its own way, such in as your bird examples, is an anthropomorphic projection onto the bird, not necessarily an indication of great brainpower. Making a claim of great intelligence for a computer would be valid when judged by those criteria, but the fact that a computer can do one thing that we consider a sign of intelligence (add) very well does not imply great intelligence, it is simply that a computer is a machine built for adding.

Wow, I'm rambling. I'd just like to make a quick point about intelligence being the degree of flexibility and adaptation of a species within it's own context. Many aspects of human behavior which we think of as showing a great amount of intelligence, like music and literature, add nothing to the survivability of the species or the individual but show such a massive array of rule-following variation that the degree of brainpower needed to excute them can't be denied. If an activity like that could be found in any other species, no matter what the medium, a case would be made for intelligence. When behavior gets so complex that an argument of simple instinct becomes more unbelievable than an argument of complex combinability of instincts is when I buy an argument of intelligence. I don't think that case has ever been made for dolphins convincingly.

People just want to think they're smart because they're cute.

> The article offers examples where dolphins seem to act stupid

Actually it didn't, but I remember seeing some elsewhere. Those have been challenged, though: it's not necessarily stupid for a dolphin to not jump over a barrier when it can't perceive what's on the other side.

As for Manger: when one scientist says he has results which challenge a large body of data, I'd like to wait for replication.

A brilliant post, Coturnix.

I remember reading years ago about the IQ test given in America in the early 1900's. As a result of the test it was deemed that all African-Americans were less intelligent than whites because their scores were so much lower.

As it turned out though, the questions were very culturally specific. So anyone who wasn't brought up in the upper class would have no idea what the answers were. It didn't matter how smart or dumb they were or what color their skin was. It was all based on experience not intelligence.

I wish I could remember what some of the questions were because they were really absurd.

I do remember that later a Professor wrote another IQ test and gave it to the upper class kids who all proceeded to fail it. All of the African-American kids who took it though did extremly well. The questions were the same as the first test, just using difference experience factors.

Sounds similar to how scientists are testing animal intelligence. Every species that has evolved and survived this long has to be considered intelligent (or just very lucky).

"No, we knew dolphins were smart millenia before we ever looked at their brains. The ancient Chinese knew it. Aristotle knew it. And the idea that brain size has anything to do with intelligence is, like, sooo 19th century."

I recently read an article talking about something similar to this... Basically, though brain size is irrelevant to intelligence the encephalisation quotient (EQ) is not. The encephalisation quotient is the ratio of the brain size to the body size, and as far as we know, dolphins EQ is second only to humans. The fact that the EQ may be a good indicator of general intelligence among a species makes sense too -- it ultimately shows the level of importance that has been allotted the brain evolutionarily relative to the size of the rest of the body.

Hey, could they graduate from high school?

Test Item:

- The Indian Removal Act (1830) relocated thousands of Cherokee
from Georgia to Indian Territory for the purpose of

a) making the land available for white miners and farmers
b) allowing the Cherokee their freedom from US control
c) obeying the Supreme Court's order to move the Cherokee
d) creating a wilderness area for use by white fur trappers

By Hank Roberts (not verified) on 18 May 2007 #permalink