In Part 1, we examined how sincere, methodical scientists can unconsciously interpret data to confirm what their society already believes — and asked whether the same pattern might operate when we assess nonhuman minds. The examples suggested that it does: scientists describing dolphin brains as “primitive” or “stuck,” dismissing cetacean cognition as “far below the human level,” when the data more honestly showed different from rather than inferior to.
But the anthropocentrism trap runs deeper than individual word choices. It gets built into the tools themselves.
The most widely used metric for comparing intelligence across species is the Encephalization Quotient — EQ — developed by the neuroscientist Harry Jerison in the 1970s. The idea is intuitive: animals need a certain amount of brain tissue just to run their bodies, and whatever remains represents capacity for more sophisticated processing. Estimate the expected brain size for an animal of a given body mass, compare it to actual brain size, and the ratio tells you something about cognitive capacity. The standard reference tables normalize the results against the domestic cat, which is assigned an EQ of 1.0 — not because cats represent some natural midpoint of mammalian cognition, but because they were a convenient, well-studied species. Everything else is measured as a ratio against that arbitrary baseline.1
By this metric, humans rank extraordinarily high — EQ around 7.0. Bottlenose dolphins score about 4.0. Orcas, roughly 2.5. Sperm whales — carrying the largest brains on Earth at eight to nine kilograms, sustained at extraordinary metabolic cost for over fifteen million years — score around 0.5. By EQ, a sperm whale is cognitively unremarkable.
The problem isn’t with Jerison’s intentions. It’s with a hidden assumption: that the amount of brain tissue needed to coordinate a body scales uniformly with body mass across all species. It doesn’t.
Consider the kinds of demands that inflate the neural budget for a terrestrial primate. Bipedal posture requires constant balance adjustments across dozens of muscle groups. Fine motor control for our hands requires extraordinary neural resources. Complex facial expressions, intricate vocal apparatus, high proprioceptive demands for navigating terrestrial environments — all of this consumes substantial neural capacity before any “excess” emerges for abstract thought. The point is not that we can precisely account for each of these demands, but that the total somatic load for a land-dwelling biped is very different from that of a marine mammal.2
Now consider a dolphin. Buoyancy eliminates anti-gravity demands entirely. Streamlined morphology requires minimal limb articulation — flippers have far fewer degrees of freedom than primate hands. And blubber constitutes enormous body mass while requiring comparatively little innervation — metabolically maintained tissue that inflates the body-mass denominator without adding proportional neural load.3
EQ assumes terrestrial somatic demands where none exist. The metric systematically disadvantages aquatic mammals in cross-species comparison — not through anyone’s intention, but through a hidden architectural bias in the tool itself.
A vivid illustration comes from within our own species. Consider NFL offensive linemen: same species, same general brain size, but body masses well over the average of 140 pounds. The EQ of a 350-pound offensive tackle falls below that of dolphins. No one seriously argues these players have inferior intelligence; they simply weigh more. The example is illustrative rather than demonstrative — linemen and other humans share a body plan in ways that humans and dolphins do not. But it exposes the underlying vulnerability: EQ is not a direct measure of intelligence. It becomes especially unreliable when applied across radically different body plans and ecological regimes.4
The parallel to Morton’s skull collection isn’t exact — EQ wasn’t designed to rank races, and Jerison had no equivalent of Morton’s social agenda. But the structural danger is the same: a quantitative tool that appears rigorous, produces precise numbers, and systematically encodes the assumptions of its creators. Not fraud. Framework.
EQ isn’t a fringe metric. It has been the default tool for cross-species cognitive comparison for half a century. It appears in textbooks, in popular science writing, in policy discussions about animal welfare. Every time someone cites it to place dolphins or whales on a cognitive ranking below humans, they are relying on a measuring instrument that was calibrated — unconsciously, without malice — to terrestrial primate bodies as the norm. The instrument doesn’t just fail to account for aquatic physiology. It actively penalizes it.
And EQ is not the only tool with this problem. The corticalization index that Aronson and Tobach used in Part 1 to rank dolphins below primates measures how much of the brain is devoted to cortex — a metric that rewards the primate pattern of cortical expansion while discounting the cetacean pattern of paralimbic elaboration. The assumption, again, is that the human brain’s organizational strategy is the standard and alternatives are deficits. Different tools, same framework.
A word about what these essays are not claiming. The argument is not that all cross-species comparison is worthless, that every disagreement about cetacean cognition is disguised bias, or that similarity and difference can be read off from any single measure. The argument is narrower and, I think, harder to dismiss: certain widely used tools and interpretive habits carry structural assumptions that favor one kind of mind — ours — and that this asymmetry deserves critical scrutiny rather than silent inheritance. That is not an attack on comparative neuroscience. It is an application of the same self-correcting standards that science claims as its own.
None of this tells us what cetacean minds actually are. It tells us that some of our most trusted instruments for answering that question have been quietly tilting the results. The numbers feel objective because they are precise. But precision is not the same as neutrality, and a biased ruler produces biased measurements no matter how carefully you read it.
So what happens when we set the biased instruments aside and look at what cetaceans actually do — how they behave, how they relate to one another, what their neural architecture appears designed to support? The evidence is extensive, peer-reviewed, and remarkably consistent. It is also, for anyone accustomed to thinking of dolphins as clever animals that do tricks, genuinely startling.
That’s the subject of Part 3.
Harry J. Jerison, Evolution of the Brain and Intelligence (Academic Press, 1973).
On neural demands of terrestrial somatic coordination, see Herculano-Houzel, “The Human Advantage: How Our Brains Became Remarkable” (MIT Press, 2016), especially the discussion of how body plan constrains brain allocation across species.
On the argument that blubber inflates body mass without adding proportional neural load (skewing EQ systematically against aquatic mammals), see Lori Marino, “Brain-behavior relationships in cetaceans and primates: Implications for the evolution of complex intelligence,” Proceedings of the Joint International Conference on Cognitive Science (1996).
The within-species body mass variation argument is discussed in Cairo et al., “Encephalization in cetaceans: an evolutionary perspective,” Brain, Behavior and Evolution (2006).