Matching Primate to Cetacean Intelligence via Comparing Humans with Killer Whales

There exist multiple meta-analyses and studies that have found, or that allow us to find, comparative primate intelligence by species or genus. The results of such studies tend to be expressed z scores, under the assumption that its distribution over the primate order is normal. Hitherto, I have shown how we might be able to calculate the comparative cognitive abilities among cetaceans.  Most simply, we can create such an index by giving equal weight to brain size, latitude range and social intelligence. Below I list those cetaceans that score most highly in these three traits combined.

 z-score  top cetacean species by genus
 +2.77  sperm whale
 +2.40  humpback whale
 +1.77  bottlenose dolphin
 +1.64  fin whale
 +1.43  killer whale
 +1.32  long-finned pilot whale
 +0.89  false killer whale
 +0.74  beluga whale
 +0.57  Gervais’ beaked whale
 +0.55  Risso’s dolphin
 +0.55  Cuvier’s beaked whale
 +0.38  striped dolphin
 +0.26  grey whale
 +0.11  bowhead whale
 +0.06  pygmy sperm whale

Unfortunately these figures mean little without allowing them a wider context. For that we need some way of calibrating the above index against primate intelligence. This proves a difficult task. You may expect that our best chance of such an alignment lies with the bottlenose dolphin, since it is the cetacean by which we have metrics for its intelligence on most dimensions. Unfortunately, this species is so sociable that the subcomponents of its sociability approach saturation, such that estimates of its social intelligence are hypersensitive to the methodology of calculation. It is also the cetacean whom our methods show as having the social intelligence that is most discrepant from its brain size. It is thus advisable to attempt a second primate-cetacean comparison. Here I will attempt to compare the killer whale and its closest primate match. Below, I show that this primate is most likely the human, and attempt to find the magnitude of the cognitive gap between them.

Direct Testing of Orca

The only direct cognitive testing I have found on killer whales has been on mirror self-recognition. All tested adults passed. The only other animal of which we can say that is human. Unfortunately, the sample size was small, yet we have a second clue that this was not an anomaly. By all accounts, the tested killer whales required far less mirror familiarisation time to pass the test than those few bottlenose dolphins who passed. These facts, and anecdotes, suggest that killer whales are considerably more intelligent than either bottlenose dolphins or chimpanzees.

The Eden Case

The only spontaneous working partnership between killer whales and humans occurred in Eden, Australia. Here the two species formed a cooperative that hunted baleen whales more effectively than either species could on its own. Their close partnership, and its voluntary nature, provides a unique window from which to assess the complexity of the interactions within a typical pod. This cooperative lasted for decades, and terminated in the mid thirties due to diminishing whale stocks. There has been recent academic interest in this partnership due to several unusual features.


As yet, no record has been found of who first formed the partnership and how. On the other hand, current evidence points towards all operations being both lead and maintained by the killer whales, not the human whalers. The persistence of such a leadership structure over such a long period would tend to suggest the social intelligence of that particular orca pod was likely to be, at least, equal to their human whaler partners.

Furthermore, it is often difficult to determine if a seemingly intelligent behaviour is just some elaborate expression of instinct. In this case, however, it would be hard to see what evolutionary history could have preadapted killer whales to begin such a sophisticated partnership with a terrestrial species – other than the possession of extraordinarily high social intelligence.


The whalers were always called to their boats by a killer whale, referred to as Old Tom, who alerted members of the Davidson family. Once in the water their boats were guided to action, by other killer whales. A second faction of orca from that pod, lead by Stranger, would have already corralled a baleen whale near shores that were up to fifteen kilometers distant from the launch point of the whaling boats. Sometimes the human whalers would have to be guided through thick fog. Several killer whales specialised in this pathfinder help for the human contingent, responding in particular to lost whalers hitting the water surface with their paddles. Perhaps the individual with the most specialist role of all in this cooperative was the killer whale known as Hooky. When their human companions were harpooning trapped baleen whales, it was common for one to lose their footing and fall into the water, often being dragged deep under. Hooky was always the killer whale who returned them safe to the surface.

Though no comparative analysis has been done, it seems unlikely that the killer whales held roles that were significantly less specialised than their human whaler companions. Neither by social intelligence, nor by level of specialisation can we separate killer whales from humans in our best historic partnership case. It remains a blight that no cognitive testing has been published on this species, other than mirror self recognition.

Aligning the Scales

No peer reviewed evidence has yet been presented, with the possible exception of vocalisation complexity, that humans are more intelligent than killer whales. From the little we know, it is hard to see how killer whale intelligence could be far below that of humans, even with the Flynn effect included. This would suggest the cetacean intelligence of z=+1.43 is roughly equal to a PIQ of 300 on my scale.

On the other hand, bottlenose dolphins score z=+1.77 and yet have a cognitive ability that equates to 175 by my PIQ. Unlike the killer whale value, this score is compiled from tests on many different cognitive dimensions. My gut feeling is that these two species lie at opposite extremes in measurement error of our z scores, but only future data can resolve this conundrum.

Thankfully, there remains a third method of estimating how cetacean intelligence may relate to the primate scale. The prominent neuroscientist, Suzana Herculano-Houzel, has hypothesised that neocortical neuron count will eventually prove the physiological metric that most closely correlates with the interspecific intelligence of a given tetrapod. It has often been pointed out that that intelligence of many birds is greater than expected of a primate of their brain size and, likewise, the intelligence of elephants is less than expected. Herculano-Houzel has shown how these anomalies disappear if forebrain neuron count is our metric, rather than raw brain size. If we extrapolate that to cetaceans, what do we find?

Unfortunately, no whale neocortex has ever been counted by the highly accurate isotropic fraction technique. Several, however, have been measured by the more traditional technique of stained count. Unlike the isotropic fractionation, stained counts often differ from each other, or isotropic fraction counts, by a factor of two. Sometimes they even differ by a factor of four. We can, however, take the four best stained whale counts and compare them with a hypothetical primate of the same count. We can average these, then compare their average cetacean z score.

 cetacean forebrain neurons  brain/g  =PIQ
 false killer whale      10.5 billion   4249   200
 common minke whale      12.8 billion   2700   213
 fin whale        15   billion   7085   223
long-finned pilot whale      37.2  billion   2893   282

This would make a cetacean z=+1.20 equivalent to 230 on my scale, if we discount the bihemispheric sleep problem. There is substantial evidence that this type of sleep lowers brain efficiency, yet it is hard to explain how such an effect could be  greater than a halving of effective brain size. Bihemispheric sleep could thus result in a lowering of expectations by as many as 45 points.


We can now generate a metric to gauge the intelligence of most cetacean species within their order and hope for some degree of confidence in it. We can, however, be less confident in our attempts to anchor points of that scale against the more established primate scale. Far less certain than even those anchor points, is how the slopes of the primate and cetacean scales relate to each other. For this year, I’m just going to assume that an increase of brain size by a factor of x has the same effect on expected intelligence in each order. Note this is a postulated within-order effect. It is not expected that a primate and cetacean with the same brain weight will tend to have the same level of intelligence. Actually, it should be significantly higher for the primate.