Mothers and Others

by Sarah Blaffer Hrdy

  To find reports even remotely comparable to the coercion and reproductive exploitation found in cooperatively breeding animals with high levels of reproductive skew, we must leave hunter-gatherers behind and turn to archaeological and especially written records from more settled and more stratified human societies. As far as is known, such social systems emerged relatively late in the history of our remarkably young species, within the past 10,000 years. Only in much larger scale societies, with people living at higher population densities with more pressure on resources, and in particular with opportunities for some individuals to monopolize resources, do we find stratified societies like those of ancient China, Japan, the Near East, early Hawaii, Africa, or medieval and early modern Europe.84 Such societies provide plenty of meerkat-worthy instances of subordinate allomothers recruited to rear the young of the powerful.

  From classical times in ancient Rome and throughout much of medieval Europe, reaching a peak in seventeenth- and eighteenth-century France, Italy, Spain, and Russia, hundreds of thousands of socially more advantaged and more powerful women relied on coerced assistance from enslaved, indentured, or poorly paid wet nurses and nursemaids to maintain their extremely high fertility levels. Having their babies suckled by other women enabled elite women to breed at almost yearly intervals, at least for a time, without jeopardizing the survival chances of their young. By contrast, the wet nurse’s own baby, denied her breast milk, suffered a high probability of dying, at the same time that the wet nurse’s own subsequent ovulations were suppressed by prolonged lactation. Wet nurses who suckled successive charges year after year might go on producing milk for decades, effectively transforming them into a sterile caste. If the authors of the Code of Hammurabi in 1700 BCE deemed it necessary to outlaw the substitution by a wet nurse of one baby (her own perhaps) for one of her charges, it is probably because cheating by alloparents was a problem.85

  Despite eerily convergent parallels between coerced human wet-nursing and the coerced wet nurses found in canids, meerkats, and other cooperative breeders, or the sterile castes typical of eusocial breeders, such cases represent derived post-Neolithic cultural (and perhaps also biological) adaptations subsequent to the time when our ancestors lived in small-scale hunting and gathering groups. As far as the origins of emotionally modern humans go, these cases are after the fact. Taking into account the ecological and demographic realities of foragers’ lives, it seems unlikely that such heavily skewed reproduction, with dominant females forcing subordinates to spend their lives helping rather than breeding themselves, could have been universal or species-typical when cooperative breeding first got started.

  In spite of recent human history that amply testifies to a species capable of ruthless exploitation of others by those powerful enough to get away with it, there is no evidence that humans evolved sterile castes per se. Rather, nursemaids became anovulatory because they were forced to continuously lactate, while eunuchs—ideal tenders for a powerful ruler’s harem—were rendered sterile by surgical castration. Neither method is a physiological adaptation comparable to the sterile castes that evolved in social insects. Women who ovulate but fail to conceive do not spontaneously undergo pseudopregnancies and lactate so as to suckle young born to a more dominant female the way canids do. Still, there are respects in which Mother Nature equipped our species with what, in terms of cooperative breeding, is an ergonomically convenient life history, characterized by long prereproductive and postreproductive life phases among women (before menarche and after menopause). These peculiarities of the human species effectively increase the ratio of caretakers to youngsters needing care.

  A few evolutionary biologists have speculated that postmenopausal females might be the human equivalent of sterile castes, manifestations of reproductive suppression by other means. Indeed, Kevin Foster and L. W. Ratnieks have gone so far as to propose that, like naked mole rats, humans should be considered eusocial mammals because they evolved to rely on alloparental care, live in multigenerational societies, and have a specialized class of sterile helpers in the form of postmenopausal women. Pushing this proposal further, British biologists Michael Cant and Rufus Johnstone have hypothesized that early cessation of reproduction in women may reflect “the ghost of reproductive competition past,” and that early cessation of ovulation in women evolved to “reduce the degree of reproductive overlap between generations” and “give younger females a decisive advantage in reproductive conflict with older females.”86

  However, to me, this sequence seems out of order. Females cease to reproduce with age in other primates as well—that is, ovaries that peter out are a preexisting condition. What is different about humans is their longer lifespan afterward. I will come back in Chapter 9 to this idea of derived traits like longer lifespans which represent evolutionary elaborations on preexisting conditions such as cooperative breeding. But first we need to consider the attributes that make primates so susceptible to misdirected maternal care and prone to the evolution of shared care in the first place.

  After all, the majority of primates exhibit some form of biparental or alloparental care. No wonder then that some 20 percent of primates have evolved shared provisioning as well as shared care, nearly twice the proportion of cooperative breeders as are found among birds. With their highly social natures and costly young, primates as an order are wide open to the evolution of cooperative breeding. With this in mind, it is time to consider specific traits that made primates so susceptible to shared care, and then consider what had to change in a particular line of apes to make shared care—widely present in primates generally, but virtually nonexistent among apes—such a feasible and attractive option for mothers in the line leading to the genus Homo.

  7

  BABIES AS SENSORY TRAPS

  What natural selection cannot do, is to modify the structure of one species, without giving it any advantage, for the good of another species; and though statements to this effect may be found in works of natural history, I cannot find one case which will bear investigation.

  —Charles Darwin (1859)

  Charles Darwin was convinced that if an animal ever did something purely for the benefit of another species it would annihilate his entire theory. Yet in a seeming contradiction to all that is Darwinian, many birds and mammals are surprisingly susceptible to the charms not only of babies that are no kin to them but even babies belonging to a different species altogether. Not long ago, in an incident reminiscent of the legendary lion who befriended Androcles, a real-life lioness in north central Kenya nicknamed Kamuniak adopted rather than ate first one, then another, and another—five in all—baby antelopes. One fawn, alas, the lioness did eventually eat. However, the other four were tenderly nurtured by the indiscriminately mothering lioness until they died of starvation or else were finally retrieved by desperate oryx mothers. That lioness “must have a mental aberration,” opined a perplexed UNESCO official back in Nairobi.1

  Around that time, another story appeared about a mother leopard who killed a baboon, then discovered a baby clinging to her prey. “The little baboon called out,” explained the nature cinematographer who had been filming the scene, “and we thought we were going to hear a major crunch and the leopard smacking her lips, but instead, the baby baboon put its paws out and walked towards the leopard . . . [who] gently picked it up in her mouth, holding it by the scruff of its neck and carrying the infant up a tree.” Its foster mother guarded the baby overnight, but by next morning the infant had succumbed, presumably to starvation. Yet, it was still being protected by the leopard. “It’s as if nature was turned on its head completely,” marveled the filmmaker.2

  No doubt the filmmaker was aware of instances when a mammal, say a male mouse, encounters an infant and either ignores it or eats it. But in fact, responding to cues from someone else’s baby is not all that rare. What was unusual in Kamuniak’s case was that the beneficiaries of misplaced maternal largesse would ordinarily have been lunch. Indiscriminate mothering has been reported n
ow for a broad assortment of creatures. Animals such as primates, canids, or felines that have shared care in their lineages appear to be especially susceptible. Even mothers with infants of their own may sometimes take up an additional baby, but usually not for long or at the expense of care to their own young.

  Hyper-nurturing barnyard hens may indiscriminately gather nearby chicks beneath the brood-patch on their breast to keep them warm. Beneficiaries of their broodiness include baby geese, ducks, or even kittens if they happen to be nearby. My neighbor’s female Jack Russell terrier once underwent a pseudopregnancy, chased away a mother cat, and spontaneously began to produce milk to suckle her foundling kittens, which she reared to maturity. The Associated Press recently ran a story about a mother cat returning this canine favor by adopting a newborn Rottweiler.3 Far down the Great Chain of Being, about a quarter of the fry in broods of some cichlid fish are fostered in from other broods.4 As long as they happen to be the right shape and size, “eggs” eliciting protection need not even actually be eggs. Mothers in those earwig species that exhibit maternal care can be tricked into protecting balls of wax provided they are the same size as their eggs.5 Birds can be just as suggestible. The cardinal shown in Chapter 6 just could not resist the impulse to deposit food in the gaping mouth of that goldfish. Throughout the Animal Kingdom, and most especially in species that produce immobile, utterly helpless babies that need a lot of maternal care (the way baby primates do), infants exude potent signals, captivating the susceptible. Related or not, infants can be powerful sensory traps.

  Kamuniak, whose name means “blessed one” in the language of the local Samburu pastoralists, is shown here with one of her wobbly-legged adoptees. (Reuters)

  How to reconcile such susceptibilities with Darwinian logic? Rather than disproving the theory, indiscriminate mothering more nearly illustrates the little exceptions that confirm Darwin’s big rule. Through evolutionary time, these mistaken recipients of care were likely to be related. Foster mothers had more to lose, genetically speaking, by not responding than by over-responding. Thus does the bar for responding get set quite low—so low that parental care is occasionally diverted to someone else’s young. But this logic pertains only as long as the risk of misplaced parental care has not been too common over the past evolutionary history of that organism. If risks of misplaced care are substantial, as is the case in herd animals with highly mobile young prone to wander over and latch onto some unrelated mother’s teat, preventive safeguards evolved. A lamb who strays from his mother and tries to pirate milk from the teats of another nearby ewe will be rudely butted away. A mother will reject all comers except those lambs whose coats wear the scent she specifically imprinted on in the minutes after she gave birth. In many other animals, however, especially those with young unable to move about much on their own (which includes most primates and many nesting birds), maternal affections remain more flexible. Whether the yawning gullet of a fish that happens to resemble the gape of a hungry chick or ultrasonic calls from a chilled pup, these cues induce recipients to respond.

  The brains of animals with helpless young are wired to register signals of infants’ needs. Their endocrine systems are calibrated to urge a rapid response, and their neuronal reward systems are designed to reinforce these nurturing behaviors. Being near babies, or in the human mother’s case holding her baby close, becomes almost addictively pleasurable. Selection favoring caring parents was essential to start the ball rolling; but once under way, kin selection intensified selection pressures favoring more generalized caregiving. In time, selection favoring nurturing responses can take on a life of its own. Once members of a given population have been selected to respond to infant cues by helping, caregivers need not be close relatives in order to respond. The stage is set for cooperative breeding—and for such totally unexpected possibilities as blessed Kamuniak, the oryx-adopting lion.

  INNATE RESPONSIVENESS TO INFANTS

  Animals with the most at stake have the lowest thresholds for responding to babies. Thresholds are set lower still during particular life phases when individuals are most likely to encounter tiny relatives in need. In female mammals, sensitivity to infant cues is particularly acute in the postpartum period. New mothers are primed by hormonal changes during pregnancy, topped off by a surge of the neuropeptide oxytocin during birth. They are supremely attuned to the shapes, cries, and smells of babies. As the little creature near her begins to suck on and stimulate her nipples, a mother’s circulating level of the nurture-promoting hormone prolactin rises.

  Describing baby lust as “addictive” is more than poetic license. Experiments with lactating rats show that the same dopamine-based reward systems that make sucking on their nipples pleasurable also renders some animals susceptible to drug addiction. Either way, whether from having their pups close by and sucking on their nipples or from consuming cocaine, the experience leaves females desperately anxious to have that experience repeated. Pups are such a powerfully rewarding stimulus that when experimenters gave mother rats a choice between pushing a lever that would administer cocaine or one that would cause pups to tumble one after another into her cage, the mother filled her cage with pups.6

  Most research on infant-activated reward systems in the brain is done with laboratory rats, mice, or voles. However, rhesus macaques are known to have similar reward systems. Indeed, when members of Eric Keverne’s lab at the University of Cambridge chemically blocked the action of endogenous opioids, maternal responsiveness to infants declined.7 Infants have similarly rewarding effects on human mothers. It should not be surprising that studies using MRIs show activation of dopamine-associated reward centers in the brains of first-time human mothers when they look at photographs of their smiling infant.8 Any number of cues would probably have a similar effect—suckling, soft coos, a familiar baby’s chuckle, the seductive smells emitted from the glands on her baby’s scalp.

  A key assumption of the misplaced-parental-care hypothesis is that allomothers respond to infants as well, but most of the neurophysiological research has been done only with mothers. Nevertheless, preliminary evidence suggests that dopamine- and oxytocin-related reward systems are implicated in allomaternal care as well. Mothers are supersensitive to infantile cues, but in mammals characterized by a lot of shared care, juveniles and subadult females who have never been pregnant or given birth also spontaneously respond to infants, huddling over pups to keep them warm. In infant-sharing primates such as langurs, these females exhibit irrepressible urges to sniff, touch, cuddle, and repeatedly take and carry new babies. In marmosets and tamarins, males—especially those with prior caretaking experience—also eagerly respond to infant vocalization and other cues, and they may be even more eager than females are to caretake.9

  The underlying organization for maternal and alloparental brains is probably about the same, but studies of different primate species with and without alloparental care are beginning to reveal some interesting neuroendocrinological differences. In both marmosets and men, males who engage in a lot of care have higher prolactin levels than males who do not. Other species characterized by nurturing males include California mice, Mongolian gerbils, African meerkats, and Djungerian hamsters.10 The inclination to care for infants derives from ancient and fairly universal physiological systems that are normally operational only in mothers. In these species, however, nurturing tendencies get switched on in males as well. In addition to (or perhaps instead of?) piggy-backing on highly conserved maternal systems, it is possible that selection may have favored the evolution of separate neuronal systems that are specific to nonmaternal caregivers. Marmosets may provide such an example.

  As in many mammals, pregnant monkeys near term are sometimes especially attentive to babies. Very pregnant langurs, for example, are second only to inexperienced young females in how eager they are to take and carry another female’s infant and in how much time they are willing to spend holding another female’s baby. The most obvious explanation is that these very pregnant fe
males are hormonally primed for motherhood. Yet this pattern is not found in all monkeys, and the responsiveness of a pregnant female to another’s infant may unfold very differently in species with pronounced reproductive skew, where there is competition between females to be the breeding female or to gain access to available alloparental care. In contrast to the usual monkey model, common marmosets in the late stages of pregnancy are the antithesis of nurturing. Indeed, late-term marmosets can be absolutely lethal allomothers. Experiencing the other female’s infants clinging to her as aversive, she attacks them, sometimes literally biting their heads off—thus eliminating young who might compete with the infants she herself would soon produce.11

  To date, most neurophysiological research has focused on how infants respond to parents, or, in a few cases, on how fathers respond to babies. Far less attention has been paid to the neurophysiology of how alloparents react to infants. I expect this situation to change in the near future as laboratory scientists increasingly recognize what important roles these caregivers played in human evolution.12 Already, scientists comparing species of voles with and without alloparental care have discovered that prairie voles, who exhibit shared care, have a higher density of oxytocin receptors in certain regions of their brains (the nucleus accumbens) than do closely related meadow voles who do not jointly care for young. Furthermore, even within the same species there is considerable variation between individuals. For example, the more responsive prairie vole females who spontaneously lick and groom their pups also turn out to be the ones with the highest concentrations of oxytocin receptors.13