The Technium

Brains of White Matter

images-1Animal computers come in all sizes. An ant brain is a hundreth-gram speck; the 8 kilogram brain of a sperm whale is 100,000 times bigger. Our large human brain, with its infinite ideas, is only one sixth the size of a sperm whale brain. It is even slightly smaller than the average Neanderthal brain. On the other hand recently discovered mini-humans on Flores island had brains one third our size, and may have been no dumber. The correlation between the absolute scale of the brain and smartness is weak.

This is particularly true if you examine the most brainy varieties of animals. For instance, the smartest animals on land, sea and air (outside of humans) are great apes on land, whales and dolphins in the sea, and parrots and crows in the air. A parrot brain is one thousandth the size of a whale’s. But if you were to test an African grey parrot, a chimpanzee and a bottlenose dolphin behind a suitable screen so that you could not see which animal was taking the test, you would not be able to determine the animal by its intelligence alone. According to Lori Marino, a neuroscientist specializing in large animal brains, the problem solving IQ of parrots, chimps and whales are nearly equivalent, even though the size of their brains vary by a thousand times.

A bigger computer is not necessarily smarter. And even when intelligence is demonstrably greater it is only weakly correlated to how many brain cells are present. It is not clear that a whale is 100,000 times smarter than an ant (ants can do much your PC can’t), or that humans are only three times as smart as chimpanzees, as pure numbers of cells might suggest. What is roughly correlated, is the size of an animal’s body and its brain. The greater number of cells in a large animal requires more brain cells to coordinate their metabolism, temperature, and movements, and thus larger bodies have larger brains, on average. While we may think of bigger brains as inherently more intelligent, most of the big brain mass in big animals is a housekeeping tax just to keep the bigger body going.

The ratio between brain size increase and body weight increase is about 2/3. For every increase of 1 unit in the mass of the animal body there is a .67 increase in brain size. For reasons no one is sure of, this ratio holds constant throughout the animal kingdom, past and present, dinosaurs included. It is a misconception that dinosaurs had walnut-sized brains. Some brains of large dinos were relatively small (about the size of your fist) but some, like those in the Stenonychosaurus, the Veloceraptor, and the Troodon had a brain cavity the size of a human. On average dinosaurs also adhere to the standard brain/body increase ratio, known technically as the encephalization quotient, or EQ.

Animals whose bodies and brains follow this ratio exactly are said to have an EQ of 1. Most animals, large and small, fall close to this expected ratio. It is the deviations from this golden ratio that interest us. While the average dinosaur had a brain proportionally equal to a mammal, the Sauropod was among the least brainy animal that ever lived, since it had a brain that was only one hundred thousandth of its body mass. For every pound of brain, it had 100,000 pounds of flesh. It survived in this minimally cerebral way for 100 million year. (Humans have yet to reach the 1 million mark.)

Modern humans have an EQ greater than 1. In fact our brains are 7 to 8 times larger than the expected ratio for mammals. A warthog and human share identical body weights yet the human brain is 11 times bigger than the warthog’s. And humanoids don’t even top the list for extreme encephalization. A shrew brain is so tiny (3 grams) it is hardly visible, yet shrews hold nearly 10% of their mass in their brain, making it one of the most encephalized animals.

Size is only one dimension that varies. Composition varies as well. In humans, the prime location for problem solving computation is the extremely convoluted outer surface of our brains. This degree of convolution is rare. But big whale brains are covered with a convoluted neocortical surface area (3745 cm2) that is substantially larger than the 2275 cm2 of the human brain. However the crinkly gray membrane of the whale is, on average, half the thickness of the human neocortex. Still, that’s a lot of neuron power.

whale brainThat leaves the problem of what all that extra brain tissue is doing in whales and elephants. What alien thoughts are those additional 7 kilos of brain processing? We can ask a parallel question about what happens as we manufacture big and bigger computers. Will adding 100 times as much processing power generate anything more than ordinary computation that is 100 times faster, if that? Will bigger computers just be whale brains?

Whale brains are stuffed with more white matter than most other mammals. The cost of gigantic brains is that the infrastructure for computing over larger distances begins to overtake the increased benefits. In order to reach more neurons, more cabling needs to be wired in. It needs to be insulated with appropriate tissue. This is the brain’s white tissue – a kind of non-computing supporting hardware. As brains grow, plain physics demands that the support grow faster than the parts doing the desired tasks. This pattern is already visible in desktop computers; the actual calculating chips are tiny fingernail-sized slivers buried among the white matter of wires, power transformers, and circuit boards, all which outweigh the chip by thousands.

Taxonomically, as brains get larger they get more complex. They develop internal regions, specialized tissues, and intermediate organs to bridge the discontinuities created by local specialization. A typical whale brain has less bulk devoted to olfactory functions than most animals (fewer smells in deep water), and far more devoted to auditory functions. Some of the extra brain matter found in a whale is thus used for its extreme sonar echolocation abilities. But the best guess for what the whale uses its big brain for is socializing. Social complexity breeds brain complexity (or vice versa), including increased encephalization. Not just in whales, dolphins and land animals such as the elephant, but in most animal classes. Birds that are more social have higher EQs; insects that are social, such as bees and ants also have larger relative sized brains.

What can the large brains of whales, dolphins and elephants tell us about the future directions of computational technology?

First, bigger by itself is not necessarily better. Yesterday’s supercomputers are crammed with a lot of white matter; today’s laptops are smarter, smaller, and incidentally have less white matter.

Second, it is not what you have, but how you arrange it. Human brains have less gray matter than a whale. Our conscious problem solving abilities stem from arranging that finite matter into more highly structured processes. Complexity can be gained without additional materials. Today computer scientists believe that we need millions more transistors to get artificial intelligence in machines, but it is more probable that we already have sufficient multitudes of neurons and that what we need is a better way to program and organize them. I find it very likely, if not certain, that in the near future – let’s say the year 2075 – two high school seniors will create a working artificial intelligence for their science fair using about 100 Intel Pentium chips found in ancient 2005-era Dell PCs they unearth from a landfill. All that we are missing back here in their past is their advance know-how, that additional organizational entropy.

Third, the way of brains is toward sociability. Just as an increase in sociability correlates to larger brain size, bigger computer brains will probably correlate to their sociability. To the degree that computers acquire sensors to the outside world (camera eyes, gps locators, accelerators, thermo sensors, tactile inputs), they will increase in sheer mass. And to the extant that they spread themselves, intermingle, and form networks (all networks are social), they will become bigger. It’s a different kind of bigness, but the kind of bigness reflected in large brains: specialized regions, massive interlobed communication, no center.

Fourth, the progressive trend is very narrow. Viewed on an evolutionary scale, there is an ancient and noticeable trend in life toward larger brains – not just in primates, but in dinosaurs, whales, and all animals. Over millions of years, the most encephalitic animals alive get progressively larger brained relative to their body mass; that is, over time the maximum encephalization has been steadily increasing.

Not all brains evolve bigger, however. One careful study of fossilized toothed-whales (dolphins, porpoises, belugas and narwhals) showed that the next species of whale was just as likely to be smaller brained as larger brained. In toothed-whales there was no general trend in evolution toward big brains. Except! Except during a few transition periods and in certain specific branches of taxonomy. Dolphins, for instance (and humanoids in the past) display an ongoing trend for greater encephalization. Some dinosaur lines show the same, while most dinosaur lines, like most animal branches on the tree of life, do not. Brain expansion, like most progress, takes place along a narrow band of “early adopters.”

Dale Russell, a dinosaur researcher, extrapolated the trajectory of maximum brain size over the last 600 million years into the next 1 million years and concluded that if biological trends continued without interference that the maximum EQ would increase by three. Humanoids our size would walk around with swollen heads and three times the brains, or whales with three times the brains would swim in the deep seas.

Much more likely is this prospect: biological evolution hands off the task of creating larger brains per body to the technium. Maximum encephalization will take place in computers, thinking machines, and the distributed network of silicon neurons around the planet. But what the whales teach us is that much of this brain material may not contribute to greater intelligence. A lot of our computer gear may only be white matter; a lot of the web may be whale brain.

  • mike

    none of this can explain to me how a thing can function beyond the sum of its parts or tasks….??? i will forever puzzle about the complexity of small things … after all we are all made of atoms etc

  • Liam

    Where did you get your data for the shrew brain/body mass ratio of 0.1? I have seem 0.03 stated elsewhere, but certainly this varies between species.

  • Shawn Charland

    Have a look at “Baboon Metaphysics” (Dorothy L. Cheney, Robert M. Seyfarth) for an excellent deconstruction of the fitness advantages conferred by the evolution of brain functions for tracking social interactions in a group that relies on ephemeral food sources.

    Also there’s an excellent treatment of related research on other animals, such as birds and dogs… the latter is at least as good as monkeys at inferring intention. And so are some songbirds.

  • Barry Kort

    It occurs to me that processing power — be it brain or digital computer or analog computer — can be divided into two fundamentally different categories. On the one hand there is sheer speed and efficiency for carrying out an information-processing task where the method is at hand.

    But from the point of view of ‘Techne’ there is question of innovation, discovery, and general problem solving, when no functional method is at hand.

    There is a class of problems which are notoriously hard for humans to analyze and solve. Mathematicians classify them as problems in Recursion. Yet, with the correct analytical technique, many problems in recursion reduce to astonishingly simple algorithms to write down.

    Functions in general, and recursive functions in particular are a challenge for any brain. Most animals walk on four or more legs. Walking upright on two legs requires the brain to compute the Balance Function. Dean Kamen labored long and hard to build his Human Segway Transporter — a device capable of maintaining an upright balance. The Balance Function requires solving a nontrivial problem in the Calculus. At the core of the Balance Function, one finds a computation that approximates the mathematical inverse of the derivative of the system model for gravity. While humans and great apes can build and host this function, most people cannot articulate how it works. That’s because most people cannot articulate concepts from the Calculus because common languages like English are simply too weak to do justice to serious mathematical functions.

    It occurs to me that one of the most powerful functions we rely on is some adaptation of the so-called Error Function. It’s found everywhere in feedback control loops. But while the Error Function has a name, it is nigh impossible to articulate the Error Function except in vague terms, without turning to the more technical language of mathematics. I can tell you that the Error Function corresponds to the integral (area under) the bell curve, but if you actually want to use it computationally, you need some functional way to actually compute it.

    Here is a short essay on the Error Function, and a comparison to the far simpler and more popular (but tragically dysfunctional) alternative, the Heaviside Switch Function:

  • Jesse M.

    Not all brains evolve bigger, however. One careful study of fossilized toothed-whales (dolphins, porpoises, belugas and narwhals) showed that the next species of whale was just as likely to be smaller brained as larger brained. In toothed-whales there was no general trend in evolution toward big brains. Except! Except during a few transition periods and in certain specific branches of taxonomy. Dolphins, for instance (and humanoids in the past) display an ongoing trend for greater encephalization. Some dinosaur lines show the same, while most dinosaur lines, like most animal branches on the tree of life, do not. Brain expansion, like most progress, takes place along a narrow band of “early adopters.”

    I’d be curious to read the studies these statements are based on, can you tell me their names?

  • Milosz

    You mention that white matter (glial cells, etc) is non-computational.

    Current evidence refutes that. It is possible that nearly as much signal processing and computation goes on in white matter as in grey, albeit of a different – more preliminary- sort. It is quite possible that a human brain (and other mamalian brains) have many orders of magnitude more computing power than previosuly thought.

    Another factor that is seldom mentioned: algorithmic power. The brain is not a general-purpose machine. It has software written into it by many millions of years of evolution, software with a level of sophistication and algorithmic subtlety that we can only begin to guess at the complexity involved. So while we will eventually make a machine with the raw computing power of the mamalian brain, it will not come from the factory with the many (trillions? trillions of trillions?) lines of code that evolution has written into firmware.

    And such algorithms as those that actually produce intelligence- we have not even the earliest rudimentary ideas about how they might be written. All our brute-force attempts at general intelligence- the only kind that really matters- have ,none of them, succeeded. “AI” is more rightly termed Actual Ignorance rather than Artificial Intelligence. Yes, my spell-checker is now context-aware, and speech recognition kind of works. But I don’t think any of us would feel comfortable voting a copy of Word into elected office…..

  • Ralph Weidner


    There’s a good explanation of EQ and its application to mammals at

    The 2/3 ratio you refer to appears to me to be a dimensional ratio in the following sense: our bodies are, of course, 3 dimensional. They have not only heighth, but also breadth and depth, so to speak. This means that if we maintain the same proportions in these three dimension as we grow up, then our body mass and volume will increase as the cube of the increase in heighth. In doubling our heighth our body weight or mass or volume will be doubled three times: 2 x 2 x 2 = 2^3 = 8. So, if we grow to a 6 ft. heighth and a weight of 200 lb., we can estimate that when we were 3 ft. high, assuming the same proportions of heighth to width and depth, and the same density, we would have weighed 200 lb./ 8 = 25 lb. That sounds a little low to me, so maybe a 3ft. tall child looks more like a 6 footer who weighs, say, 250 lb. The point is we can expect weight to increase roughly as the cube of height.

    The 2/3 ratio indicates that brain weights do not increase as the cube of height, but as 2/3 x 3 = 2, i.e. as the square of height: brains are in some sense two dimensional, so that a 6 footer would only need a brain 2 x 2 = 4 times as large as the 3 footer he or she was as a child, even though he or she now weighs 8 times as much.

  • Dmytry

    Speaking of brain size per body mass… i believe it is measure of relative importance of brain and flesh. Evolution always tries to find optimal ratio.
    It is quite clear that relative brain importance is bigger when you have good “software” and organization in general, and when it is signigicant for survival (that is ,there’s use for it in environment). I believe that jumps in brain size happened when new “software” evolved. Some species finally evolved an organization that is quite efficient and scales for bigger brain sizes. It increases importance of brainmass, and in result you get species with big brain/body ratio, for example humans. Or dolphins (that jump in ratio). Then, it fixes at some equilibrilum(if there’s a time to reach it), until either a:environment or something else is changed (more use for brain, adaptation required), or b: new organization evolves.
    When environment is made of other animals showing same trend(ior same species), it sometimes becomes important to outsmart competitors, again resulting in brain size increase.

    On final note, It is unclear if human’s thought is final. It is quite possible that higher levels of organization than ours could exist, and seems quite likely that higher levels is completely unimaginable by us, and beyong human understanding as much as mars rover is beyong understanding of elephant or ape.

  • Berend Schotanus

    So look what most humans are doing with their gigantic brains. Are they inventing rocket science? Discovering new Einstein laws? No! None of these!
    They are keeping up the latest gossips at their hairdressers! You are completely right, social interaction is the most important driving force for more intelligence. (And the complexity of social behavior is much underestimated)