The Mating Mind: How Sexual Choice Shaped the Evolution of Human Nature (23 page)

A similar explanation might account for the popularity of sexual ornaments with bilateral and radial symmetry. Biologist Magnus Enquist suggested that symmetric patterns might be the most exciting way to stimulate animal visual systems. He argued that any visual system capable of recognizing objects when they are rotated will tend to be "wired" in such a way that it is optimally excited by radically symmetric patterns. Enquist and his collaborator Arak did some evolutionary simulations in support of their claim that any neural network capable of recognizing rotated objects would be optimally excited by radially symmetric

patterns. Supposedly, this explains the popularity of sexual ornaments that resemble stars, sunbursts, and eyespots.

In addition to fulfilling general engineering principles, the senses of each species must also adapt to its particular habitat and ecological niche. Sensory bias theorists such as John Endler have investigated how different lighting conditions influence the sensitivities of different animal visual systems. This sort of research promises to help biologists predict which animal lineages are more likely to evolve particular kinds of sexual ornament that play upon particular sensitivities. This application of sensory bias theory might help biologists to discern more patterns beneath the apparently chaotic proliferation of sexual ornaments in different species.

Yet a different view of sensory biases may explain why ornamentation evolves so unpredictably. For example, given the same problem of categorizing visual shapes, two different species may evolve two rather different solutions. One may evolve to represent visual shapes as variations on some sort of generalized cylinder, while the other may represent visual shapes as sets of facets and angles. Both ways of mentally representing shape may work perfectly well, but they might respond very differently to a novel sexual ornament that has a particular shape. The ornament might make an aesthetically pleasing generalized cylinder, but a very unappealing set of facets, or vice versa. The ornament may prove a sexual success in one species but not in the other.

One of the deepest insights from sensory bias theory is that there is always some evolutionary contingency in the design of perceptual systems. These contingencies make it impossible to predict all possible responses to all possible stimuli just from knowing what a perceptual system evolved to do. Therefore, if a new sexual ornament evolves that excites a perceptual system in a novel way, it may be favored by sexual selection in a way that could never have been anticipated. For example, biologist Nancy Burley found that female zebra finches just happen to be attracted to males that have tall white plumes glued on top of their heads. Their white-plume preference probably did not evolve as an

adaptation, because as far as we know, ancestral finches never had white plumes on their heads. The preference just happened to be a latent possibility in a visual system that evolved for other purposes. I think this idea of evolutionary contingency in perceptual systems is one of the most intriguing ideas to come out of sensory bias theory. It might even work better than runaway sexual selection as a general explanation of why sexual ornaments diversify so unpredictably in different species.

From Sensory Appeal to Sexual Appeal

My main worry about sensory bias theory is that stimulating a sensory system is only the first step in influencing a mate choice decision. Grabbing a potential mate's attention is a long way from winning his or her heart. Granted, for animals that live widely separated from one another, it may take a lot of effort to find anyone of the opposite sex during the mating season. Under these conditions, making a strong sensory impression would give an animal a reproductive advantage. A whale song audible from hundreds of miles away can help two lonely whales to find each other. For many species, locating a mate—any mate—is a big problem. The sensitivity of their senses may be crucial to finding a mate, so may have a significant impact on sexual selection.

For highly social animals like most primates, finding potential mates is not the problem. Many primates already live in large groups, and interact regularly with other groups. They are spoiled for choice. When mate choice depends more on comparing mates than locating mates, the sensory engineering argument seems weaker. Why should an individual be perceived as a more attractive sexual partner just because its ornamentation happens to excite some brain cells in the lowest level of one's sensory systems? If it were that easy to make animals come running, predators would more often evolve lures to dupe prey into approaching them.

Our intuition may tell us that strong sensory effects are sexually attractive, but I doubt this attractiveness is explained entirely by sensory bias arguments. There are good adaptive reasons why

ornaments that produce strong sensory effects make good fitness indicators. Consider the list of sensory bias effects that Michael Ryan and A. Keddy-Hector compiled in an important review paper of 1992. They noted that animals usually respond more strongly to visual ornaments that are large, brightly colored, and symmetrical, and to auditory ornaments (e.g. songs) that are loud, low in pitch, frequently repeated, and sampled from a large repertoire. These responses could be attributed to sensory engineering effects. But that begs the question of whether the sensory engineering evolved to help animals choose good sexual partners. Large, healthy, well-fed, intelligent animals can produce larger, brighter, and more symmetric visual ornaments, and louder, deeper, more frequent, and more varied songs. As far as I know, there is no example of a sensory bias that leads animals to favor sexual partners that are smaller, less healthy, less energetic, and less intelligent than average. Most sensory biases are consistent with what we would expect from adaptive decision-making machinery that evolved for mate choice. It may not have evolved specifically for mate choice, but it might as well have.

Many sexual ornaments may look as if they are merely playing on the senses. They may appear to be nothing but fireworks, sweet talk, eye candy, special effects, and manipulative advertising. But maybe we should give the viewers more credit. What look like sensory biases to outsiders may have a hidden adaptive logic for the animal with the senses.

Tickling Senses Versus Advertising Fitness

If sensory biases led animals to choose lower-fitness animals over higher-fitness animals, I suspect that the biases would be eliminated rather quickly. It seems unlikely that an ornament could persist as a pure sensory bias effect that does not convey any fitness information. That grants too much evolutionary power to males evolving ornaments and not enough to females evolving sensory discrimination abilities. Animals choosing mates do not want their senses subverted by meaningless ornaments. They may like fitness indicators that have a lot of sensory appeal,

but they should not be favoring sensory appeal over fitness information.

Often there may be no conflict between sensory bias theory and fitness indicator theory. They are complementary perspectives on sexual selection. Sensory bias theory reminds us that mate choice is mediated by perceptual abilities, and that as new perceptual abilities evolve, the way is opened for new kinds of sexual ornaments to evolve. With the evolution of eyes came the possibility of visual ornaments. With the evolution of bird ears came the possibility of bird song. And perhaps, with the evolution of language comprehension abilities in our ancestors, came the possibility of sexual selection for much more complicated thoughts and feelings expressed through language.

Pleasure-Seekers

Biologists Tim Guilford and Marion Stamp Dawkins have argued that sensory bias theory can be generalized to deal with all sorts of psychological biases, which may also affect the evolution of animal signals and sexual ornaments. Any aspect of an animal's nervous system that influences how it reacts to a signal can influence how signals evolve. Apart from sensory biases, there can be attentional, cognitive, memory, judgment, emotional, and hedonic biases. These may be even more important in accounting for complex courtship behaviors of the sort that our species has evolved.

For example, maybe we can understand the mind as a sexually selected entertainment system that plays not just upon our sensory biases, but upon our thirst for pleasure. Consider two hypothetical kinds of animal. One has evolved some hard-wired brain circuits to do mate choice. It searches through several potential mates, remembers their ornaments and courtship behaviors, compares them using some decision algorithm, and picks one for copulation. It derives no pleasure from impressive ornaments to which it attaches a high value. It simply registers the value in an automatic, businesslike way. It has no hedonic experience. A good mate brings it no pleasure, only good genes.

It could be called the "cold chooser." I suspect that most insects work in this way.

The other animal is a "hot chooser." Its behavior may look similar, but its experience is very different. Its mate choices are influenced by subjective feelings of pleasure. When an attractive individual performs a charming courtship dance in front of the hot chooser, the hot chooser experiences some combination of aesthetic rapture, curiosity, warmth, happiness, awe, lust, and adoration. These feelings play a direct causal role in the mate choice process. The more pleasure a potential mate arouses in the hot chooser, the more likely that individual is to be chosen.

Given this description of cold and hot choosers, there is no way

an external observer could tell them apart. Now I shall add the

crucial feature that makes an observable difference, a difference that could influence evolution through sexual selection. Suppose that the pleasure system the hot chooser uses for mate choice is the

same pleasure system it uses for all other domains of survival and reproduction. The hot chooser has a big pleasure-meter in its brain—it may be something like the level of endorphins floating around its nervous system. Its pleasure in watching an attractive male is subjectively similar to its pleasure in eating good food,

escaping a dangerous predator, viewing a propitious landscape, watching its children thrive, or doing anything else that

contributes to survival or reproduction. All of its decisions are

mediated by this pleasure-meter.

Over the short term, the cold chooser and the hot chooser will

behave in the same way. They will make the same mate choices. But over the long term, they can evolve in different directions

because they will react differently to new courtship behaviors.

Suppose that a male happens to have a mutation that leads him to give good food to a female. A cold-choosing female may eat the

food, but the food might not influence her mate choice, because

her eating system is separate from her mate choice system. Her systems do not share the common language of pleasure. The

mutant may have no reproductive advantage, and his food-giving tendencies will probably die out. (Females of many species have evolved preferences for food gifts during courtship; my point here

is that they may not have automatically wanted to mate with the first generation of males that offered food.)

If the same food-giving mutation arises in the hot chooser species, the female's reaction would be much more positive. The mutant gives her food, which increases her pleasure in his

company. Since her pleasure-meter is what determines her mate

choice decisions, she favors the food-giving mutant. He gains an immediate reproductive advantage over his competitors. The

gene for food-sharing spreads through the population because it

brings pleasure, and pleasure influences mate choice. The hot choosers would equally favor any novel courtship behavior that

saved them from predators, or led them to a rich new habitat, or helped their existing children thrive, or brought them any other kind of pleasure.

Why would any animal evolve a pleasure-meter? I think that the main benefit of a unified pleasure system is that it simplifies learning by allowing the hot chooser to use similar kinds of

reinforcement learning in many different contexts. If it feels pleasure when eating, it can use that pleasure as a

reinforcement signal to tell it to do more of the foraging

strategy that was just successful. If it feels pleasure when copulating, it can use that pleasure as a reinforcement signal to

make more use of the mate choice strategy that was just

successful. Designers of robot control systems have realized that smart robots need reinforcement learning abilities.

Moreover, artificial intelligence researcher Pattie Maes has argued that when robots need to juggle many priorities, a central pleasure system can help them rank those priorities. Pleasure helps solve the problems of reinforcement learning and prioritizing behaviors.

The stern sensory bias theorist might warn that this sort of pleasure system makes the hot choosers vulnerable to sexual exploitation. Courtship behaviors would evolve that simply activate the pleasure centers, influencing the hot choosers to mate with their manipulators. That sounds bad. But is it? In terms of the subjective experience of the hot choosers it cannot be bad, because activation of their pleasure centers is, by definition,

pleasurable. As long as pleasure is defined broadly enough, to encompass everything from a full belly to a fulfilled life, an individual cannot wish for any subjective experience beyond pleasure. For utilitarians who value the greatest happiness for the greatest number, sexual selection driven by pleasure is a dream come true.