What If Brains Are Efficiently Designed for Information Processing?
The processing power of a brain should reflect the product of their number of synapses and the number of types of synapses available. Thankfully, all mammals have closely similar numbers of neurotransmitters and of proteins involved in brain function. This reduces the measure of potential to a simple count of synapses. With that in mind, it is worth noting that bottlenose dolphins have been estimated to have a similar number of dendrites per neuron as humans.
Unfortunately, reality differs from the perfection of theory since the average activity level per synapse can differ from species to species. We now have good evidence that mammalian neuronal energy requirement is the same irrespective of neuron size. This gives us reason to believe that the potential for intelligence could follow raw neuron count closest of all.
As mentioned in my section on EQ, the biggest challenge to the use of unadjusted brain size as a indicator of intelligence has been the elephant. The elephant has a brain three times the size as that of humans, and at least five times the size of the next largest nonhuman land-based animal. As such it has been extensively studied, and has been found to have similar or slightly lower intelligence to the great apes. Suzana Herculano-Houzel has performed an exhaustive investigation on a selected elephant brain, and found that it does indeed have three times more neurons than humans, but that the vast majority of these are in its cerebellum. Much other work also indicates that the cerebrum (most of which is neocortex), to be the stronger base of higher mental functions, so it seems highly likely that this accounts for the discrepancy. An elephants trunk is controlled in minute detail across 200 muscle blocks, a process that is mediated through the cerebellum, and we have reason to believe evolution has driven such fine control in elephants even at the expense of higher functions. This has made elephants extreme outliers in mammalian brain architecture. We can, however, still expect a meaningful comparison if we measure the number of neurons in the cerebrum alone. If we restrict ourselves to that we recover the following of hierarchy of greatest measured general brain power among all animals on land.
|African Elephant||6 billion|
These are Suzana Herculano-Houzel’s figures. She has obtained the most comprehensive data to date through her method of isotropic fractionation . Alternatively, we can rely on the more traditional stained neural count methods with equivalent published results as listed below.
|African Elephant||11 billion|
Despite some differences between these two methods that need to be resolved, and despite the conspicuous absence of data on Asian elephants by either method, both methods recover humans as having a wide margin of advantage over any other land animal. Allow me the liberty of combining these two data sets, and using the geometric average where an animal has been measured by both methods. That generates our final, and most comprehensive, table.
|African Elephant||8 billion|
This chart is highly consistent with psychometric testing data as to the possible true ordering of terra firma’s most intelligent genera as they exist today.
Are Cetacean Brains Significantly Different than Those of Terrestrial Mammals?
Cetacean brain’s actually do show some architectural differences from land dwelling mammals. The fairest assessment of this is that the ways in which they are simpler (such as a lack of layer IV) are at least balanced by ways in which they tend to be more complex (greater gyrification, presence of spindle neurons).
Is There Another Reason to Expect Significant Deviation in the Measured Intelligence of Sea Mammals From Neuron Number?
One notable disadvantage that all sea mammals must overcome is that they can’t afford to loose control of breathing while asleep. All marine mammals tend to answer it by utilising unihemispheric sleep. While this expedient saves them from drowning, there are reasons to believe that it prevents a brain from operating as efficiently as their land based cousins. The argument is that this devise is so easy to evolve that it has been found several different times, yet no terrestrial animal uses it. There must be a very good reason that this doesn’t happen, especially considering that it would make any animal less vulnerable to night predators. This presumptive lowering of brain efficiency must have limits however.
We can make a good argument that the factor by which unihemispheric sleep reduces effective brain power can’t be greater than two – ie it can’t be less useful than would be the case if the two hemispheres began operating completely independently in everything but the time at which they sleep, or else that is how the brain would arrange itself.
What Is the Number of Neurons in a Whale Neocortex?
We have extensive data as to the mass and volume of cetacean brains by species, but less as to neuron densities. I turn, yet again, to the work of Suzana Herculano-Houzel, who has shown that the brains of land based mammals scale closely according to certain set rules within each order. The terrestrial mammals with the most efficient scaling rules have proved to be the primates, and those with the worst have proved to be the rodents. If whale brains were constructed like those of primates, then most would exceed human capacity, but if they were constructed within rodent rules then even their largest brains would be just comparable to those of the great apes. The neuron density data on marine mammals is not as good as for Earth’s land based fauna, yet I believe that, as of mid 2014, it is already sufficient for our cause.
To find which scaling rules apply to whales we need only one or two firm examples. Until recently, the best data for toothed whales came from Poth et al, and for baleen whales the work of Eriksen and Parkenberg who examined the 2.1 kg brain of the minke whale. Let’s first go to baleen whales.
Eriksen and Parkenburg found 13 billion neurons in the neocortex of the minke whale. Their samples averaged a brain mass of 2.1 kg. This count is already higher than any nonhuman terrestrial animal, and much closer in ratio to the primate scaling rules than the rodent scaling ones. However, cetacean evolutionary adaption seem to have operated faster than in other groups, so we must be careful to eliminate the possibility that the two main divisions of cetaceans might scale differently. For that we need data from the Odontoceti (toothed whales).
Poth et al examined the sensory and motor cortices of five delphinids and the pygmy sperm whale. They found a strong inverse relationship between neural densities and brain size. So strong in fact, that had this pattern extended to all areas of the neocortex, it would have implied that the largest Odontoceti cerebrum contained only slightly more neurons than the smallest – and that toothed whales brains expanded according to even poorer scaling rules than seen in rodents. Alternatively, it might imply no more than that larger animals only require the same processing power as smaller ones to interpret their sensory input and motor coordination. The matter lay at that impasse until 2014 whenpublished their unbiased survey of the entire pilot whale neocortex. If cetacean brains scaled as did rodent brains we would expect this 3.7 kg brain to have 3 billion neuron, if it scaled like a primate 55 billion. The result was 37 billion neurons. This has extraordinary implications. Pilot whales are now the only animal with a formal neocortical neuron count that is higher than humans, and they do so by such a wide margin that it almost exceeds that potential unihemispheric factor of 2. It is a pity that this species is not as well studied as killer whales. This raises many questions that I hope will direct future study.
We can begin to be confident that neuron number in both toothed and baleen whale brains scale not much below those rules governing primates, such that sperm whales (which have the largest brains among them) should be at least three times the human cerebral neuron number – and without the disadvantage of unihemispheric sleep observed in other cetaceans. The data on cetacean neuroscience as it stands in 2015 is thus consistent with the cognition of most whales being at least on par with the great apes, and with sperm whales exceeding human cognitive abilities by at least the margin that we exceed other primates, and inconsistent with anything below human ability levels.
A Final Note of Caution
If you currently (mid 2015) go to the Wikipedia page listing animals by neuron number, you will find the fin whale neocortical count is put at an extremely low 1.5 billion. It seem the editors have finally noticed the complete absence of any peer reviewed evidence in support of this number, yet I think I have finally figured out where this mystery figure came from. It could have been a decimal point mistranscription from a 1969 study by Kraus and Pilleri (Quantitative Untersuchungen uber die Groshirnrinde die Cetaceen). This old study estimated the neocortical neuron numbers of humans, pilot whales, and fin whales at 14 billion, 30 billion, and 15 billion respectively. Despite the date, Kraus and Pilleri’s figures for humans and pilot whales are close to more modern determinations. In the event that this precision holds for the fin whale also, we would have a 5.8 kg whale brain with only a human number of neocortical neurons. That would complicate the simple picture portrayed here.