This brings us to another important point. T is often thought of as a “male” hormone, the assumption presumably being that only males have enough for it to be of psychological significance. When, after all, was the last time you heard someone despairingly say “It’s the testosterone” of a woman’s behaviour? Unless her transgression was to grow a beard, probably never. This popular T = male perception is both reflected in, and reinforced by, the much greater research attention on males than females.36 But as van Anders wryly asks: “What does its natural occurrence do in females, then?”37 As she points out, what we think of as high T or low T doesn’t have to be in reference to absolute levels. It’s just as useful and valid to refer to a T level that’s high “for males” or “for females,” or that’s high relative to what was seen in that individual a minute, hour, month, or three years ago. Consider, for example, the recent finding that one in six elite male athletes have testosterone levels below the normal reference range. Given that these men were sampled a few hours after taking part in a major national or international athletic competition, we would hardly want to predict that athletes with modest T levels (in some cases below the average for female elite athletes) have little in the way of competitive inclinations.38
Another major departure from the T-Rex view of testosterone comes from a well-established principle in behavioural endocrinology: hormones don’t cause behaviour, but rather only make a particular response more likely. As Adkins-Regan explains:
Hormones are one of several factors that go into the nervous system’s decision. They may change the thresholds for other factors that enter into the decision (for example, thresholds for responding to stimuli from another animal) but are not normally the sole triggering agent.39
That is, rather than being a king who issues orders, T is just another voice in a group decision-making process. This, when you think about it, is extremely sensible. Even for animals in which the social situations encountered may seem trivially simple compared with the soap operas of human existence, there are still subtleties of context to be considered. The animal whose philosophy was to take a strictly hormone-response approach to the world would soon find itself in trouble. How an animal actually responds to a particular stimulus, like a potential mate or intruder, isn’t determined by its hormonal state, but depends on social context: What is the relative status of everyone involved, who else is around, where is the encounter taking place?40
In fact, we already saw a vivid example of this in the cichlid castration study: the dominant social status of the castrated territorial fish trumped the higher T levels of his contender. Another demonstration of the same principle comes from a study of talapoin monkeys. Here, the captive community under investigation included both intact and castrated males, and the latter were periodically treated with whopping doses of testosterone to see the effects on social dominance (measured by aggression towards others in the group). Although this T dosing did increase the aggressive behaviour of the castrated males, this was invariably directed towards lower-ranking males. In other words, the relative social status of moneys was a primary and powerful filter for whether T had any effect on aggressive behaviour. As a result, despite treatment that raised T levels above and beyond normal, “no animal rose in rank after hormone therapy.”41 In fact, there was no obvious relation between T and rank, with females usually, and castrated males often, ranking above intact males. T doesn’t inevitably create a Rex.
An even more striking contrast with the T-Rex view comes from evidence that not only is T not sufficient to trigger a hormone-linked behaviour, but in some species it may not even be necessary. Take sexual behaviour. In many species, the hormonal coordination between fertility and mating is so tight that sex isn’t even possible without adequate hormone production in the testes and ovaries.42 Many male rodents, for instance, can’t produce an adequate erection without intact testosterone-producing gonads, while in females, ovarian hormones control various bodily changes that make sex physically possible (like the seductive arched-back “lordosis” pose that makes female rats’ otherwise inaccessible vaginas available for penetration). But in most primates, there are no such hormonal conditions. Hormones are instead linked with sexual motivation, rather than ability to copulate. According to Emory University behavioural neuroendocrinologist Kim Wallen:
This separation of the ability to mate from sexual motivation allows social experience and social context to powerfully influence the expression of sexual behavior in nonhuman primates, both developmentally and in adulthood.43
In an elegant demonstration of this, Wallen looked at how a testosterone-suppressing treatment affected the sexual behaviour of male rhesus monkeys housed with females. In line with a link between T and competition, the treatment had a more severe effect on the sexual behaviour of monkeys grouped with several other males, and that therefore presumably had to compete for mating opportunities, compared with monkeys that enjoyed solo male status. (Interestingly, and a useful reminder that it’s not just males that compete, something similar is seen in female rhesus monkeys, that are more likely to mate outside the fertile phase of their cycle if there’s no female competition around.)44 But even in the competitive multi-male situation, T suppression didn’t always decrease sexual behaviour. Both prior sexual experience and higher rank served to protect against the effect of testosterone suppression. So although sexual activity ceased within the week in the lowest-ranking male, the sexual behaviour of a high-ranking sexually experienced male “was not detectably affected by testicular suppression.”45 This was despite his testosterone levels being at castration levels for eight weeks. Status and past experience overrode hormonal lack.
As a final indignity for the T-Rex account, not only is T neither a sufficient, nor necessary, cause of hormone-linked behaviour—sometimes it’s not even really a cause at all. Recall the purpose of hormones: to “adjust behavior to circumstances and contexts.” 46 To this end, T turns out to play “a key role” in helping animals tune their social behaviour to whatever social scene they find themselves in.47 Although we’re used to thinking of certain kinds of behaviour as “testosterone fuelled,” in many cases it would make more sense to instead think of actions and situations as being “testosterone fuelling.” Social context modulates T levels (up or down), which influences behaviour (presumably via changes in perception, motivation, and cognition), which influences social outcome, which influences T levels… and so on.48
Again, the cichlids offer a useful illustration. On first encounter, you’ll remember, it seemed obvious that dominant fish were dominant because they had high androgen levels. But careful experimentation revealed that, in fact, dominant fish had high androgens because the stars of fate had aligned to make them dominant. When male cichlid fish are first put together, their androgen levels tell you nothing about who will end up high versus low in social status. Although from a T-Rex perspective we would assume that the fish with more androgens would “naturally” be more successful in clambering up the social ladder, this simply isn’t the case: the relations between hormone and dominance go the other way around.49 Only once the fish have had time to interact and jostle do correlations emerge, with successful fish producing more androgens. As Lisbon University behavioural neuroendocrinologist Rui Oliveira, lead author in this study, explains:
Social information is translated into changes in levels of steroid hormones that in turn will modulate the neural network of behavior so that behavioral output is tuned according to the perceived social environment.50
In fact, the effects of the social world have even been seen at the genetic level, with social interactions changing androgen and oestrogen receptor expression in the brain.51 Testosterone, in other words, is demoted from Rex to being a mere middle man that mediates the influence of the social world on the brain. Change the world, and you can change T—and the brain.
And, importantly, even in nonhuman animals it’s subjective perception rather than physical reality that counts. Consid
er again the cichlids; in particular, one unfortunate male hailing from the Oliveira study. All but one of the socially dominant males in this study successfully established territories, and started to churn out androgens in greater abundance. But that one fish, despite winning about 70 per cent of his fights, failed to establish a territory. Intriguingly, the androgen production of this winning fish was an outlier on the graph, being down at the level of the group of vanquished fish. As Oliveira notes, “This suggests that it is the individual’s perception of its status, rather than an objective measure of its dominance behavior, that triggers [androgen] production.”52
Or consider a study of male marmoset monkeys, a monogamous species in which fathers are actively involved in parenting. Researchers measured T response to the ovulatory odours of unfamiliar females, and found that it depends on the male’s family status. Single males showed testosterone elevations (as well as penile ones) in response to the sexually enticing smell. But to “family” males (those pair-bonded with offspring), this same stimulus apparently had little effect—perhaps because it represented a distraction rather than an opportunity—and their T levels remained unresponsive.53
In short, the picture painted so far is a far cry from the simple T-Rex view, in which testosterone fuels male competition in direct proportion to the stable quantity of it roiling in the bloodstream. We’ve already seen that competition is an important feature of females’ lives too. And with the circulating level of T being just one variable in a complex system—and one in which the sexes may potentially use different means to reach similar ends—we can’t assume that T is only important in males. T is also just one of many factors that feed into an animal’s decision making. Social context and experience can override its influence on behaviour, or fill testosterone’s small shoes in its absence. And finally, far from being a pure biological measure of hormonal sex, T responds to contexts and situations, meaning that whatever influence T has on the brain and behaviour can’t be simply chalked up “to testosterone,” a purely biological factor. T level or reactivity is inextricably intertwined with the individual’s history and current subjective experience.
So what about ourselves?
In keeping with other animals, T likewise seems to help us to adapt our behaviour to “circumstances and contexts.” So, when it comes to relatively enduring circumstances—basics like partnering and parenting—T levels seem to be in line with the principle of high T being linked with competition, and low T with nurturance. For instance, both women and men with interest in acquiring new sexual partners tend to have higher circulating T than do happily coupled (or happily single) counterparts, and the parents of young children have lower T than non-parents.54 And while it’s hard to disentangle cause and effect in this kind of realm—needless to say, scientists can’t randomly allocate people to ten years of marriage, or a baby—this doesn’t seem to be simply due to people with higher versus lower T levels being drawn to different lifestyles. For instance, the findings from a study of male air force veterans, who were brought into the lab for regular testing of hormone levels and to report on their marital status, “illustrate the dynamic nature of testosterone levels, elevated in the years surrounding divorce, and declining through the years surrounding marriage.”55 The authors speculate that this happens because:
The marriage ceremony is the culmination of a more gradual period of courtship and engagement, in which a man accepts the support and consortship of his partner, removing himself from competition with other men for sexual partners. As a result… his testosterone declines. In contrast, impending divorce is a time of competition between spouses for children, for material possessions, and for self-respect. Also, it is a time when the divorcing husband may reenter the competitive arena for sexual partners.56
And an arrow of causality from caregiving to T-level change was clearly seen in a large-scale longitudinal study of fathers in the Philippines, led by University of Notre Dame biological anthropologist Lee Gettler. This study found that fatherhood reduced testosterone levels in men, and more so in fathers who spent more time physically caring for their infants.57 Nor was this simply because lower-testosterone men were more likely to be nurturing fathers; rather, intimate caregiving itself lowered testosterone.
But also notice how we’re not like other animals: our social constructions of gender bring a uniquely human dimension into the mix. As we’ve already seen, gender norms and patterns for sexual behaviour and parenting take on widely different forms across time and space. These cultural circumstances are surely entangled in women’s and men’s hormonal biology. Illustrating exactly this situation is a study that compared two neighbouring cultural groups in Tanzania—Hadza foragers and Datoga pastoralists—each with very different expectations of fathers. It found lower testosterone levels among fathers from the Hadza population in which paternal care was the cultural norm, compared with Datoga fathers among whom paternal care was typically minimal.58
By the way, lower T levels don’t doom devoted husbands and fathers to a submissive or sexually sparkless life. Contrary to popular belief, in humans there’s little convincing evidence for a significant link between baseline circulating T and social status, and most studies have failed to find relationships between T and sexual desire in healthy men with T levels within the normal range.59 This may well be because competition and status are more intermittent and situational for us than for some other animals. (And as van Anders points out, sexual desire can also flow from feelings of love and intimacy.)60 We don’t, for instance, all simultaneously take two weeks annual leave, fight ferociously for the best homes in which to rear our children, then frenziedly mate. It would seem to make more sense, then, for T to obligingly rise or lower temporarily, as context demands, or as opportunity knocks.
But here again, social constructions of gender will shape both the situations people encounter, and their subjective meaning. We’re used to thinking of testosterone as being a cause of gender, but what if the direction of that familiar pathway also needs to be reversed? Some ingenious recent research by van Anders and her colleagues has started to lay down the evidence.
Take a study in which van Anders and her team acquired one of those programmable crying, sleeping, gulping, pretend babies that high schools use to illustrate the vital fact that using contraception, however inconvenient it might seem at the time, is considerably less effortful than parenting.61 One group of men were randomly allocated to a role we will imaginatively describe as “traditional man who lets the woman do the baby care.” They were instructed to simply sit and listen to the baby cry. Another group of men, again randomly assigned, formed the experimental condition we will refer to as “traditional man who lets the woman do the baby care and is therefore woefully inexperienced in that demanding, acquired skill, but on this occasion has been left alone with a baby.” These men were told to interact with the baby, but, cruelly, it was programmed to cry persistently, regardless of what they tried. The last group of men were what we will call “progressive dads”: the baby was set to cry, but programmed to be consolable when trial and error led to the right kind of comfort. In this last group, testosterone levels dropped as their tender ministrations took their desired effect. But in the other two groups, faced with the challenge of a profoundly unhappy baby, particularly when they just sat and listened, their testosterone rose. In other words, ostensibly the same stimulus—a crying baby—affects T differently, depending on the person’s ability to deal with the situation.62 Now consider the fact that, outside the laboratory, the confidence and experience a person brings to the challenge of a crying baby is likely to be shaped by gendered expectations and experiences around child care. Claims that men don’t have “the right hormones” for taking care of babies are cast in a whole new light.
A second study by the same research lab shifted the simulated context from home to work. This time, van Anders and her colleagues trained male and female actors to perform a workplace monologue in which they enacted power by f
iring an employee, measuring T levels both before and after.63 Displaying power didn’t significantly affect men’s testosterone levels overall. However, it did significantly increase T in the women. The interesting implication that the researchers drew from these findings is that social constructions of gender, which make displays of power both more likely and acceptable for men, contribute to the female/male gap in circulating T levels. “Gendered behavior modulates testosterone,” the researchers conclude, pointing to “an additional reason for differences in testosterone: the understudied role of nurture.”64 (That’s right, guys. Let women take your high-power jobs, and before you know it they’ll be taking your hormone too.)
In this study, the social context was unambiguous: an incontestable power monologue. But often situations are more subjective, and taking this into account may be helpful for making sense of the tangle of results from studies of relations between testosterone and competitive behaviour on the sports field and in the lab. In these contexts, participants are usually unsure of just how tough the competition is that they’re up against, or how things will turn out. At first, following a few null findings with women, it was prematurely assumed that only men show testosterone reactivity to competitive situations.65 But as more data piled in, what emerges is, as one review recently summarizes it:
an inconsistent pattern in both sexes, with T levels increasing in winners and decreasing in losers… increasing both in winners and losers, or not showing significant changes in response to the competitive event.66
A sceptical conclusion would be that testosterone isn’t doing much at all here.67 But maybe, the authors of the review suggest, these inconsistencies are created by the different lenses through which people perceive a competitive situation. As endocrinologists Gonçalo and Rui Oliveira put it:
Testosterone Rex Page 14
