It is easy to invent a selectionist explanation for almost any specific observation; proving it is another story.
—Motoo Kimura1
Human newborns are helpless. Without adult care, human infants would not stand a chance. They are unable to locomote on their own, or to eat anything beyond the most limited of diets. There are species whose young exhibit adult behavioral characteristics virtually from birth. The horse is an example. If the horse is not helpless at birth, neither is he particularly smart—nor does he get very much smarter, Clever Hans notwithstanding. With human infants, it is the other way around, a fact that requires an evolutionary explanation.
The emergence of human intelligence is in part attributable to an increase in their brain volume, which is thought to have roughly tripled over the last 2.5 million years.2 At 1,300cc, human brains are enormous in comparison to those of chimpanzees and gorillas.3 Brain size relative to total body size has also increased. The human brain constitutes 2% of total adult body weight, and consumes 20–25% of basal metabolism.4 It is also three times larger than would otherwise be expected for a primate of human body size.5
For at least thirty years, the standard solution to the riddle of human neonate helplessness has been the obstetrical dilemma.6 Because our ancestors from the time of Australopithecus walked on two legs, their pelvises had to remain narrow to preserve mechanically efficient movement. But their larger brains required an increasingly wide birth canal. These competing selective pressures resulted in an evolutionary trade-off; the human pelvis became as wide as bipedalism permitted. Since intelligence was under selective pressure, the most straightforward trade-off required human beings to give birth when newborn crania were still relatively small, and their brains relatively underdeveloped.
Hence the helpless, or altricial, human newborn.
Human neonatal altriciality does seem to be a recent evolutionary development. Most non-human primates give birth to precocial young, and humans maintain several precocial characteristics, such as having single young rather than litters.7
If adult brain size increased as a result of selection for higher intelligence, and pelvic dimensions had already reached their functional limits given bipedalism, this would force increasingly early births. Human babies do reach the size limits of the birth canal.8 Preterm birth is a leading cause of neonate and infant mortality, suggesting that birth timing has reached a selective equilibrium.9
Steven Piantadosi and Celeste Kidd have recently presented a model that reflects the assumptions of the obstetrical hypothesis. They have added a simple observation: the increasing helplessness of human newborns itself exerts a selective force because it requires intelligent parents.10 This requires that the adults have bigger brains, which was the reason for earlier birth in the first place. The result is runaway selection. The obstetrical hypothesis explains why high intelligence took so long to evolve. Why are there no fearfully intelligent insects? By relating intelligence to the unique bipedalism of humans, the obstetrical hypothesis offers a reason for the relatively late emergence of higher intelligence.
Piantadosi and Kidd’s theory relies on three assumptions: that infants do better with more intelligent parents; that intelligent parents require large brains; and that there is a connection between large brains in adults and helplessness in newborns. The third assumption is in question, if only because the second assumption is in question.
Does intelligence really require a large brain?
Piantadosi and Kidd’s theory fits comfortably with standard descriptions of embryonic growth in which the Gompertz growth curve is used to describe an increase in head radius through gestation and postnatal development. This serves to confirm the plausibility of the obstetrical dilemma.
Yet the theory leaves room for an alternative Darwinian scenario in which runaway selection leads to smaller brain sizes and longer gestation times. If brains were to become smaller, gestation could proceed for longer. This would give rise to neonates more likely to survive to adulthood on their own, relaxing selection on parental intelligence.
This account also possesses a certain appeal. Our ancestors certainly did not experience a reproductive advantage because of their ability to do algebra or play the piano.11 The Darwinian environment, in enlarging their brains, somehow granted our ancestors capacities that they would not require, and could not use, for millions of years. Stephen Jay Gould and Elisabeth Vrba have called such grafts exaptations.12 If algebra and piano playing have nothing to do with fitness, what about our ability to form an accurate representation of reality? Fitness and truth are distinct, Donald Hoffman and Chetan Prakash have recently argued.13 What we perceive is only a part of the show; perceiving the rest would require far too much energy for our brains to manage. There is a certain reproductive advantage assigned to a creature that does more by perceiving less, if only because it does not waste energy. In some circumstances, self-deception might well have offered our ancestors more by way of reproductive success than seeing anything clearly.14 This, at least, is a position with common experience to commend it.
If we take science at face value, as scientists often do, we are again left with a strong evolutionary suggestion that selection should not have favored a species capable of creating such powerful theories as quantum mechanics or general relativity.15
Do Piantadosi and Kidd give any reason to suppose that evolution might have taken this trajectory? Let us calculate and see who is right. Over seven million years ago, Sahelanthropus tchadensis, just possibly our common ancestor with the chimpanzees, had an endocranial volume of approximately 365cc.16 If Sahelanthropus tchadensis had a gestation period of about 34 weeks, it would have landed at a spot in the fitness terrain that has roughly 40% probability of survival to adulthood and that is about equally likely to have evolved toward a state in which gestation is long and brains are small. Further, the result assumed by the obstetrical dilemma has both a lower peak fitness and describes a much smaller number of possible outcomes than the alternative.17
Piantadosi and Kidd acknowledge some of these caveats:
Once a population has moved into the appropriate region of the space, trends for growing brain sizes and lowering birth ages will mutually reinforce each other … One must still explain why human populations happened to move into [those] parts of space … the theory should be viewed tentatively… our model demonstrates that runaway selection is logically possible.18
Maybe. Even if the theory applies to the evolutionary history of mankind, the most that can be said about our curious predicament is that it is a matter of historical accident. If these consequences of the obstetrical dilemma are ambiguous, its fundamental premises are implausible. It is by no means clear that brain size correlates with intelligence. Human brains are dwarfed by those of whales, dolphins, and elephants; some whale brains are at least seven times as large as those of humans.19 Larger organisms naturally have larger brains, but humans also do not possess the largest relative brain size. Our brains constitute 2% of our total body mass. This is similar to the ratio in apes and dolphins and quite high for a mammal of our size. But in the case of birds, mice, and shrews the figure is closer to 10%.20
The Neanderthals had the largest brain of any hominid, with a mean volume of 1,435cc.21 They likely encountered the obstetrical dilemma.22 If lack of intellect played any role in their extinction, it may not be brain size, but rather the finer aspects of cerebral organization and folding that allows for increased intelligence.23 “Brain size be damned,” Jonathan Balcombe remarks, “if it’s critical to a species’ survival then that species will most likely be good at it.”24 A recent meta-analysis of the association between brain size and intelligence suggests a correlation of only r = 0.24 among humans, implying that variation in brain size explains only R2 = 6% of the variation in intelligence.25
Researchers have sought other ways to measure the brain to account for superior human intelligence. The encephalization quotient (EQ), by assuming the superiority of human intelligence, manages to confirm the superiority of human intelligence. This is not a notable contribution to the question at hand.26 Absolute brain volume is a better metric than EQ in predicting differences, such as self-control, between species.27
The most reasonable cerebral measure seems to be the total number of neurons, which can differ between brains of the same size due to differing neuron-packing densities. Because of their relatively large cortices, small neurons, and high packing densities, primates have more neurons than expected given their absolute brain size.28 Human beings in particular have more neurons than any other species—about 15 billion cortical neurons and about 100 billion neurons overall.29
No measure of brain size quite explains the variations in primate intelligence. Some other characteristic must have been the primary feature under selection. Human beings have larger neocortices, temporal lobe volume, and prefrontal white matter volume than other primates. There is also greater cortical folding in the human brain, and more gyral white matter in the frontal and temporal lobes.30 It is reasonable to suppose that the alteration of cortical folding is the feature of the human brain that has been under selective pressure.31 For this reason, researchers have attempted to identify the very genes under positive selection in the human lineage. Genes are included in this category if they display amino acid differences—that is, nonsynonymous genetic changes—within primates.32
Despite the appeal of the Darwinian explanation, an increase in nonsynonymous change is perfectly compatible with the relaxation of negative, or purifying, selection.33 Some studies acknowledge this possibility.34 Others do not.35 Failure to do so can be seriously misleading. The great majority of accelerated nonsynonymous evolutionary changes can occur by genetic drift rather than selection.36
One study found no detectable association between either abnormal spindle-like microencephaly or microcephalin and intelligence.37 If accelerated evolution is a hallmark of relaxed constraint, its signal will indicate where we should not look, and not the other way round.
It is often assumed that positive selection is the most straightforward explanation for accelerated evolution in nervous system genes. But after human adults became intelligent, relaxed purifying selection on newborn intelligence would be expected, since smart adults could compensate for newborn helplessness. The literature does not mention this possibility.
There is at least one alternative to the hypothesis of an obstetrical dilemma. Holly Dunsworth advocates a metabolic hypothesis. Children are born when their energy consumption becomes too great for their mothers to handle.38 Because humans have larger brain and body sizes than other primates, a disproportionate amount of which is metabolically costly fat, they are born with relatively underdeveloped brains.39 No runaway selection applies.
It is consistent with this view that gestation times in chimpanzees and gorillas are roughly 32–34 and 37 weeks, respectively.40 The typical human gestation times of 38 weeks is, in fact, longer than one would expect for a primate of similar body mass.41 Instead of arguing for an enlarged brain, then, one might as easily argue for a diminished body mass. Thus, it is in spite of similar periods of gestation that humans are more helpless than other great apes at birth; humans are not born early. Moreover, we wean infants earlier than expected for a primate of our body size, not later. The opposite should be the case if weaning time is a proxy for altriciality, and if the obstetrical hypothesis is correct that altriciality requires more intelligent parents.42
Fetal development provides additional evidence for the metabolic hypothesis. Human infants have brains roughly one third the size of adult brains; for chimpanzees, the figure is much higher, at roughly 40%.43 At 16 weeks of gestation, the human brain is twice the size of the chimpanzee brain. After 22 weeks of gestation, chimpanzee rates of growth slow and human rates of growth accelerate. At 400cc, human newborn brains are roughly 2.7 times larger than chimpanzee brains.44 Since maternal energy demands increase exponentially during human fetal development, waiting longer than nine months would exceed the maximum sustainable maternal metabolic rate, which is approximately the same across primates.45 Other mammals rely on progesterone depletion for the initiation of birth, but primates give birth when progesterone levels are at their height. It is an energetic ceiling that signals the onset of labor.46
There is no evidence to support the assumption that mechanically efficient bipedal walking requires a narrow pelvic morphology. If there are metabolic costs to walking and running with wider hips, they could be offset by subtle changes in movement patterns.47 Nor is brain size necessarily the primary cause of birth complications; fetal shoulders are also very broad and, like the head, require rotational movements for passage through the birth canal.48
In a 1992 review, Allan Wilson and Rebecca Cann argued that genetics provides a better approach to human origins than paleontology. While “living genes must have ancestors,” they wrote, “dead fossils may not have descendants.” The genome is a matter of fact, but the fossil record requires interpretation.49 Ditto for the computer models used by Piantadosi and Kidd.50 While computational models are exact, they are also entirely hypothetical.
