davidm

I have a question about the evolution of human-style intelligence. I realize that evolution has no purpose or direction, and that it is wrong to think in terms of a progression from lower to higher species. Still, human-style intelligence -- whatever that is, exactly -- has proved to be of paramount value in the competition for survival. I wonder why it has only evolved once. Trying to answer my own question, I can only suppose that a brain as complex as ours is hard for nature to achieve. It is the product of a lot of unlikely accidents and contingencies. On the other hand, I have also read that brain complexity has increased among many species over time. Is there any objective, scientific way to think about this subject, or is it all pure speculation? If humans became extinct, can we talk in any meaningful way about the odds that a different species might eventually pick up our fallen flag? I guess probably not, but even speculations on this subject would be welcome. I certainly find it intriguing.

Peez

Hi,

It is not clear that the evolution of "intelligence" is unlikely, but like any other trait it would be expected to evolve only under certain conditions, and only when certain structures are already present. For example, plants lack a nervous system and probably could not evolve "intelligence" as we think of it. I could imagine "intelligence" evolving in any social mammal under the right conditions, though I cannot be sure of course.

There are a few reasons that "intelligence" might have apparently evolved only once. The first is, of course, that it did not evolve only once. Perhaps "intelligent" creatures have evolved and gone extinct several times. If some dinosaur was "intelligent" but had no technology, how would we know? We might make an educated guess from skull size and shape, true, but it is an interesting question.

Perhaps the most important factor is the cost of intelligence. Most people tend to think of the benefits of a trait, but for natural selection to favour a trait the benefits must outweigh the costs. A complex nervous system uses a huge amount of resources, to build and to maintain. It makes childbirth difficult and dangerous, slows development, uses up a great deal of metabolic energy, and is vulnerable. If you are a mouse eating seeds in a field, just how much is "intelligence" going to increase your reproductive success?

I hope that this helps.

Peez

RufusAtticus

David,

I read a paper last year that argued that uniquely human intellegance was due to selection for adapability to the short-term enviromental flucuations that were going on at the time.

Doubting Didymus

The reason that evolving intelligence is unusual is that it is a particularly 'difficult' niche for a species. It is much 'easier' to survive by being brawny, big, fast or armoured than it is to evolve a complex brain.

To start along the trail to intelligence as we know it, it is probably necessary for other, easier evolutionary niches to have already been filled. (As they no doubt were on the african veldt)

I suspect that the reason humans have out-evolved everyone else in the intelligence department is that we began to rely almost totally on brain power for survival. This sort of 'squeesed' us into a previously unpopulated niche, a sure fire recipe for success.

KeithHarwood

OK, I'm going to dive in here in complete ignorance of any work done on the subject, please shoot me down in flames when I talk rubbish.

It seems to me that there is nothing unique about human intelligence, that intelligence arose a long time ago and humans are better at it than most other animals (though I have my doubts about even that sometimes).

It seems to me that the evolution of intelligence began with creatures that could learn, as distinct from those whose behaviour is hard-wired. Much apparently intelligent behaviour arises out of very simple hard-wired algorithms. Grey Walter, in about 1952, showed quite complex behaviour in two `turtles' he built. Each had two motors, two sensors (a photo-cell and a microswitch which detected a bump on the shell) two batteries and the central nervous system of two relays. They would explore their surroundings or sit quietly. When they were hungry they would go to a feeding station to recharge their batteries. They had a light on the front of the shell so they could see each other and themselves in a mirror, and their behaviour was different. Minor differences in construction (variation within the species?) gave them different personalities. Elmer was outgoing and active. Elsie tended to lurk quietly under the dining table. They could die. If they encountered a fluffy rug they could starve to death. A stairway downwards would kill them straight away.

Most very small creatures work the same way. They have hard-wired behaviour that alone, or in concert with other of their species looks like purposeful behaviour, as if they know what they are doing. For example, stir up an ants nest and the ant will collect all their larvae together. This is not something that the individual ants decide to do. Chemicals released when the nest is disturbed turn on a particular state machine that controls their behaviour. This machine has two states. 1: scurry around at random carrying a larva. 2: scurry around at random not carrying a larva. It has two events. A: see a larva. B: see another larva.

When the machine is triggered it usually starts off in state 2. Eventually the ant receives event A and transitions into state 1 (that is, picks up the larva). Eventually, in that state it receives event B and transitions to state 2 (that is, it puts the larva down). The net effect of a large number of ants executing this algorithm is that all the larvae are collected together.

I think that intelligence started when learned behaviour evolved. This would initially have been very primitive, just a few neurons whose inputs and outputs were not dedicated to particular behaviours, but which could trigger off those neurons that were dedicated. This would allow the organisms to adapt to minor changes in the environment much faster than evolution would allow. (I'm thinking changes such as a rock falling across a feeding path here.) In fact, it would allow the individual organism to adapt. Such adaptation would be clearly an evolutionary advantage for a species that could afford the initial outlay of the `useless' neurons. Over the eons more and more behaviour shifted from hard-wired to learned. Eventually (possibly several times) there evolved hard-wired behaviour which caused the organism to seek out learning experiences. I suspect that the effect of that behaviour is what we call `play'. It is exhibited by the larger birds and most mammals. From then on I think it's just a matter of degree.

Somewhere in the range of organisms that are capable of learned behaviour and that actively seek out learning experiences you can draw an arbitrary line and say, `above this line is intelligence'. But whereever you put the line you are going to have trouble justifying it because the distinction between those above and those below will be difficult to make.

Doubting Didymus

Well spotted, Kieth. 'Human' intelligence is merely a question of degree. Your observations about 'hardwired' motivation are astute. Particularly with respect to insects.

I have forgotten the details, but in one of richard dawkins' fine books, he talks about certain species of wasp that have very clearly defined hadwired behaviours. This wasp always leaves its fresh prey (a spider) outside its nest, while it cleans its nest. If you move the spider while the wasp is inside, it will move the spider back to its original location and then clean its nest again from the start you can keep doing this, and the wasp never catches on.

This raises serious questions in my mind as to whether the wasp has any actual concious thoughts (I certainly can't see why it would need them). I truly wonder, therefore, if wasps are any more 'alive' than a machine that acts in a similar way.

davidm

Thanks for the interesting responses, which make it depressingly evident (though by no means certain) that intelligence of our degree is difficult to evolve. Peez, you mention intelligence evolving without technology. I guess the human-style intelligence I had in mind presupposes technology as a part of it. But then I guess dolphins could be an example of a sophisticated, nontechnological intelligence. This seems to have implications for things like SETI. I have read that biologists are much more pessimistic about SETI contact scenarios than are, say, astronomers. True? Because filling that intelligence niche might usually lack the benefits to justify the costs? P.S. with respect to dinosaurs, I seem to recall reading somewhere that encephalization was increasing among at least some dinosaur species at the time they became extinct.

Late_Cretaceous

It is not only the human mind that needs to be taken into consideration.

We also have hands and the ability to speak.
Dolphins may be equally intelligent as humans, but cannot manipulate things in thier environment. Apes are quite intelligent, have the ability to manipulate things, and have the capacity to understand language, but do not have the ability to speak.

It is really the combination of traits that make humans unique.

parkdalian

We mustn't forget the sexual selection "arms race". We have long passed a unique threshold for an intelligent species, recognizing our own intelligence as something desirable in a mate. It's an ability that most if not all people have.

Also, the ball was already rolling with monkeys and apes being almost uniformly clever, relative to other animals. Grasping primates are inherently tool-using if you think of a tree as actually being a very complex tool.

scigirl

Here's some ideas from Skeptic magazine's Michael Shermer from <a href="http://www.amazon.com/exec/obidos/ASIN/0716733870/internetinfidelsA" target="_blank">Why People Believe Weird Things:</a>

Apparently, most evolutionary psychologists agree that the brain has "mental modules," groups of neurons that connect to each other and have some type of function that you can separate from other functions. What is most important is not the location, but rather the interconnectedness of these modules. However, many scientists disagree whether the modules are more "domain specific" (the brain is like a swiss army knife) or "domain general (a more fluid functioning brain). Evolution, perhaps, brought humans from a specific to a general module type brain, which allowed us to design tools, create art, and believe in gods.

Shermer seems to think that humans evolved what he calls a 'belief engine,' which gives different outcomes depending on the circumstances:
It's a rather interesting book - unfortunately he doesn't spend that much time discussing this theory.

scigirl