The physical limits for sapiance

[Crossroads Inc.]

In Fiction, there has always been races very big and very small. From giants to gnomes, and everything in between. However given that all things must follow physical laws, it gets me to wonder if there is a reason all the highly intelligent species on Earth to be around 'people' sized.

Humans, Primates and various Cetaceans tend to be the most advanced mentally. And all are roughly the same size and all are mammals. ((There are other highly clever species that are smaller, cuttlefish and octopi tend to be very clever, but not shown the same level of awareness as primates or cetaceans))

Now, we of course have extreme cases of size in nature in general on both ends. Blue Whales can be up to 100ft long, bigger then any other living animal (baring some sprawling plants organisms). And of course we have animals going all the way down to the microscopic.
The question is, given the "plumbing" for a brain like ours, could such an organ exist to the same level going up or going down?

In Fantasy, races get as small as mere inches, this can work out to a brain the size of a marble or smaller. Given that what has been show to carry intelligence is often how "detailed" a brain is, the level of wrinkles in it as it where, how much smaller then a Human brain could you go before it becomes physically impossible to have the same level of intelligence.

Likewise, would a brain many times our own in size be able to maintain itself or develop at all?

[Eleas]

To some extent it depends on what aspects of intelligence you really need and what you read into the notion of "plumbing".

The human brain is a great all-round cogitator, but a lot of it is devoted to things you wouldn't immediately expect, such as vision. This is why you can teach rats (or they teach themselves) to act and reason in surprisingly clever ways, despite having brains the size of walnuts; they're making do without a lot of the stuff we have. So it might be possible to replace the flawed human eyeball with a corrected design, possibly removing the need for a lot of post-processing. I expect that if you rationalized away everything but the sort of ability you were after, you'd get a very capable brain that could still be smaller than that of the average spider monkey.

However, as I'm not qualified to go deeper into the subject, I think I'll wait for someone more knowledgeable to address the issue.

[ThomasP]

To some extent it depends on what aspects of intelligence you really need and what you read into the notion of "plumbing".

The human brain is a great all-round cogitator, but a lot of it is devoted to things you wouldn't immediately expect, such as vision. This is why you can teach rats (or they teach themselves) to act and reason in surprisingly clever ways, despite having brains the size of walnuts; they're making do without a lot of the stuff we have. So it might be possible to replace the flawed human eyeball with a corrected design, possibly removing the need for a lot of post-processing. I expect that if you rationalized away everything but the sort of ability you were after, you'd get a very capable brain that could still be smaller than that of the average spider monkey.

However, as I'm not qualified to go deeper into the subject, I think I'll wait for someone more knowledgeable to address the issue.

What I wonder about such architectural streamlining is how much of our consciousness and cognitive powers are emergent from the entire system. By all accounts there is no "intelligence module" in our brains, and it could well be that our unique talents would be hard to replicate without copying the balance of the whole structure.

Visual and motor-sensory processing, along with the pleasure-seeking and pain-avoiding drives of the limbic system, drive our decision making and contribute to our conscious experiences in ways that aren't obvious. If you take that away, do you wind up with anything resembling a human intelligence?

I honestly don't know what you could come up with given the right evolutionary pressures (life has a way of surprising), but by all appearances we're a unique combination of neurological circumstances that might not be so easily pared down in size.

[Alyrium Denryle]

If you want something resembling sapience, you need a well developed social structure and the necessity of solving novel problems. This is why predators, some birds, and pod-living cetaceans are so mind-mindbogglingly smart, even in cases where their brains are tiny (see: parrots). It is not a matter of brain size. I figure once a taxon gets over some sort of threshold in the number of connections between neurons they have sufficient raw processing power. Smaller brain may just mean more connections between individual neurons, or more tightly packed neurons (I dont know enough comparative neuroanatomy to be certain if this is the case).

Solving novel problems requires the ability of the brain to integrate information and form good estimates of cause and effect relationships. All it requires in terms of consciousness though is a primary consciousness, which is not the same as self awareness (first order theory of mind) but in some taxa, that is basically what it becomes. When social environments get complicated... that is when the consciousness really starts to develop. Empathy is the next step. The ability to perceive the experiences of another ("I know, that you know"), otherwise known as second order theory of mind. This is why your dog looks back at you quizzically if you tell it to do something, and it has trouble. The dog knows that you might have information it does not. It is also why the dog seems to know when you are sad and upset. It DOES know. Why? Because dogs (and other animals with second order theory of mind) have the ability to simulate the experiences of another organism using Mirror Neurons. If you see someone get stabbed, you viscerally know what they are experiencing, because the same neural pathway used to process pain is activated inside your brain, it is just processed differently. The mirror neurons are responsible for this. Third order theory of mind is the next step, when an organism is able to process and simulate what another organism experiences with respect to a third (I know, you know, that he knows"). This ability develops sometimes in adult chimpanzees, and human children develop it around age three. A normal human brain can get up to sixth or seventh order theory of mind by adulthood.

All this really requires anatomically is that the brain be organized in such a way as to permit it, and that the right type of neurons exist. I dont think physical size is as much a limiter. Monitor lizards for example can count, and some parrot species probably have at LEAST second order theory of mind.

[Mayabird]

Birds do have smaller, more tightly packed neurons, as a matter of fact, and crows at least definitely have a second order theory of mind (burying me some food...wait, that other crow's watching me. I'll just keep pretending to bury it until...he's gone. Good. Taking this elsewhere so he won't know where the food is).

[Broomstick]

Bird brains are structurally very different than mammal brains, even if they and we share some structures with the reptiles. Birds have also been under enormous pressure to reduce weight in the flying species, which is most of them, but operate in three dimensions, which requires highly refined perceptual and processing skills. This has resulted in them dispensing with paired organs for the most part (flying birds generally have just one functional kidney and gonad, for example) and might have also driven a more efficient brain structure than mammals, allowing comparable intelligence (however you want to define it) in a smaller package.

Of course, there are some downsides to be birds - mammals tend to be much more resistant to reduce food supplies, and because they don't have such high energy demands they occupy niches birds either never do or rarely do. Intelligence isn't the only thing that drives species survival.

[Korto]

I dont think physical size is as much a limiter.

But our brain size causes a lot of problems in childbirth. Babies are effectively born premature in a brain development sense compared to other species to try and get them out before the brain gets to big to get out, necessitating a lengthy development period outside the womb, and even so, childbirth causes an incredible amount of trouble compared with any other species that I can think of, because the head is so big, to fit the brain. (And the fact we walk on two legs, making the birth canal an S shape).
My point is, this would constitute significant selection pressure to have a smaller brain than we do, if at all possible, so I would conclude from this that to have a brain capable of doing what ours does, it must be at least as large as ours is.

I will repeat, "capable of doing what ours does". I'm not talking about doing something different, but equally impressive.

[Broomstick]

It's possible that high intelligence requires an undeveloped brain being exposed to the outside world, and thus our early birth may not be entirely an artifact of skull vs. pelvic girdle size.

I will point to the fact that our closest primate relatives, although more developed at birth than humans, also tend to have a relatively long dependency period. The more intelligent birds are altricial, hatching out naked and helpless and looking very unfinished, unable to hold up their head, regulate body temperature on their own, sit up, or move their limbs in a coordinated manner. Sound somewhat familiar?

[Alyrium Denryle]

But our brain size causes a lot of problems in childbirth. Babies are effectively born premature in a brain development sense compared to other species to try and get them out before the brain gets to big to get out, necessitating a lengthy development period outside the womb, and even so, childbirth causes an incredible amount of trouble compared with any other species that I can think of, because the head is so big, to fit the brain. (And the fact we walk on two legs, making the birth canal an S shape).
My point is, this would constitute significant selection pressure to have a smaller brain than we do, if at all possible, so I would conclude from this that to have a brain capable of doing what ours does, it must be at least as large as ours is.

We are also under a lot of evolutionary and physiological constraints, so that trade-off is just as much a product of out ancestry as it is a physical limitation. Put it this way. In order to get the same cognitive function as a crow, a mammal brain has to be much bigger.

This is because the split is ancient, and the two types of brain (bird vs mammal) evolved along completely different lines for a long time. The progenitors of birds were small dinosaurs with complex social structures (and I am not taking into account the most recent, as in, a few days ago fossil evidence that shook up trophic position of Avialan dinosaurs, it is too early to tell if we need to reject them as predators) that needed advanced problem solving to survive. Their brains specialized in that direction well before they were miniaturized to reduce weight.

Mammals by contrast, until the KT event had been evolving as scavengers and nest raiders. They had paternal care, but other than basically lived like Shrews. Their brains evolved along completely different lines for some two hundred million years before a frontal cortex was slapped on.

[Lord of the Abyss]

Bird brains are structurally very different than mammal brains, even if they and we share some structures with the reptiles. Birds have also been under enormous pressure to reduce weight in the flying species, which is most of them, but operate in three dimensions, which requires highly refined perceptual and processing skills. This has resulted in them dispensing with paired organs for the most part (flying birds generally have just one functional kidney and gonad, for example) and might have also driven a more efficient brain structure than mammals, allowing comparable intelligence (however you want to define it) in a smaller package.

As I recall bird brains have a very different architecture than mammal brains. Mammal brains do their high level processing on the surface, therefore the well known connection between convolutions on the brain and the intelligence of a species; but you can only pack so much surface area onto a small brain. Bird brains do the same kinds of processing in nodes spread throughout the brain and so don't have that particular size issue.

Also, I read some years ago that birds have less junk DNA than most species and at least partly because of that have smaller cells. So a bird can among other things pack more brain cells in the same sized brain than a mammal can.

[Korto]

So basically we evolved with hemp rope for brains, and birds got wire rope, so we shouldn't be surprised we need a lot more of it to lift the same load. Fair enough.

[ThomasP]

SciAm had an article on this in the July issue, which is unfortunately behind a paywall.

Robin Hanson has a brief commentary on the article here.

[someone_else]

Also, I read some years ago that birds have less junk DNA than most species and at least partly because of that have smaller cells. So a bird can among other things pack more brain cells in the same sized brain than a mammal can.

Cell size is more a factor of other stuff (like what the cell must produce on a regular basis and its energy requirements).
Say, you may have a somewhat smaller nucleus (and it's debatable imho since specialized cells have most of their genome very tightly packed since they need only some tiny bits to work), but if each neuron has to deal with much more synapses, it needs much more RER and more mitochondria.

Not to say it can't be true!!!1!1, but I'm not willing to accept this without raising an eyebrow either.

Although smaller brain and body means the axon can be much shorter, which means that neurotransmitters and other suff has to be carried for shorter distances, which means far less energy expenditure, which means less mitochondria thus smaller cells.



I'm still surprised that we manage to rule the world (or delude ourselves that we do) even with a less-efficient brain design.

[Lord of the Abyss]

Also, I read some years ago that birds have less junk DNA than most species and at least partly because of that have smaller cells. So a bird can among other things pack more brain cells in the same sized brain than a mammal can.

Cell size is more a factor of other stuff (like what the cell must produce on a regular basis and its energy requirements).
Say, you may have a somewhat smaller nucleus (and it's debatable imho since specialized cells have most of their genome very tightly packed since they need only some tiny bits to work), but if each neuron has to deal with much more synapses, it needs much more RER and more mitochondria.

Not to say it can't be true!!!1!1, but I'm not willing to accept this without raising an eyebrow either.

I found a relevant story here, although interestingly it appears that smaller genomes and thus smaller cells came first.

For many years, scientists have noticed that flying animals, especially birds and bats, are light in the DNA department—they have small genomes. There was always a chicken-and-egg question as to whether smaller genomes (with relatively little DNA in their cellular nuclei) enabled flight or just happened to be a characteristic of animals that flew—whether, in the words of one evolutionary biologist, "they had to jettison their genomic baggage" to fly, "or it was never loaded in the first place."

The answer to that question is in a new Nature paper, which shows that the dinosaur ancestors of modern birds had small genomes well before they turned into anything that could fly. Tyrannosaurus rex, for example, had a bird-sized genome; that is, there was only about one-third as much DNA in each of its trillions of cells as there is in each of ours. The study is an example of how a seemingly random genetic shift can enable a quantum leap in behavior—in this case, flight.

A team led by Chris Organ and other scientists at Harvard performed the Nature study by estimating cell size from measurements of the holes in fossilized dinosaur bones. They compared the cell sizes—which are known to correlate with genome size—to those of living animal relations and used that to group various dinosaur lineages by genome size.

<snip>

Most salamanders are smaller than birds, but their genomes are often 90 times larger. If you've ever seen a salamander move (they don't get around much), you can see why they don't need small cells. Having larger cells also means that salamanders have fewer neurons per square centimeter of brainpan.

[Starglider]

Say, you may have a somewhat smaller nucleus (and it's debatable imho since specialized cells have most of their genome very tightly packed since they need only some tiny bits to work), but if each neuron has to deal with much more synapses, it needs much more RER and more mitochondria.

Neurons would need a lot less mitochondria, and support cells for that matter, if they used electronic conduction instead of ionic conduction. This would also be much faster, good for both intelligence and reaction times. Synapses are extremely complicated chemical systems and changing them would be a huge shift, but axons are relatively straightforward. Metallic filaments would be rather impractical with organic chemistry but there are plausible replacements, e.g. certain kinds of carbon nanotube act as hyperconductors. Dumping all the charge pump metabolism would be a big improvement and would let you significantly scale down the brain.

[someone_else]

Neurons would need a lot less mitochondria, and support cells for that matter, if they used electronic conduction instead of ionic conduction.

Except that cells cannot handle free electrons.
How do you plan to create current in that filament? You need a buttload of chemical reactions to make a decent current. Is it going to be cheaper than ionic pumps?

My pet idea was making a modification to the citoskeleton (already there), to allow it to carry signals mechanically. That is: synapses give a kick to the citoskeleton's proteins that alter their shape in a cascade-reaction that goes through the whole cell and down the axon. In the axon you place some mechanism that can only be triggered if enough of the cell's citoskeleton was triggered, so you still have a way to control how excitable is the cell like with depolarization waves (there are YES and NO synapses).
Then to bring the mechanism back in firing position you consume power.
I have no idea if this is cheaper than ionic pumps, nor ideas about its speed .

[cosmicalstorm]

Isn't the biggest problem with OP's question that we simply don't (yet) have a formal theory of intelligence?