If genes generally do better in aggregates, then sometimes they might do even better in aggregates of aggregates. That is, sometimes it might benefit genes to build survival machines that collect together in groups. Much advantage accrues to fish that school or birds that flock. Such aggregates offer, for example, better protection from predators and greater efficiency in finding food than the individuals would enjoy on their own. Groups of predators like lions or hyenas can subdue large prey they could never capture on their own. Other things being equal, if alleles that build social survival machines tend to survive better than alleles that build solitary survival machines, then over time the species will come to be found in groups.
However, it takes more than aggregation to make a society. A flock or herd may be a sort of limited partnership, with each individual's behavior confined to keeping close to the rest while feeding. In a society, however, the individuals do not behave only for their own benefit. When a group of wolves tracks and kills a moose, they all behave together in ways that benefit them all. The behavior of each is necessary to the attainment of the goal, and without the efforts of all none would benefit. This is cooperation (chapters 11 and 12).
For wolves, a moose is a shared goal in a literal sense: once killed, it is shared among the members of the group. If each individual's participation depends on benefiting from the group's activities, then each individual must get a share. Any tendency to cheat must be curbed because each individual's benefits would cease if the group fell apart. Genes will be selected that help to subordinate the individual's short-term interests (cheating) to the group's long-term interest in maintaining cohesion. Such a tendency to act for the good of others in the short run, but for the sake of greater benefit to all in the long run is what we have called altruism (chapter 12).
Altruism is the hallmark of a society. When a group lives together in a stable association and behaves altruistically toward one another, that is a society. In a society like an ant colony, in which everyone is closely related, altruism can be selected by the benefit to the shared altruistic genes; between close kin no reciprocity may be necessary. If group selection favors altruistic genes, however, even unrelated individuals in a society will share the altruistic genes. This sharing lessens the need for reciprocity, but may not eliminate the need entirely. Although the altruistic genes facilitate the development of altruistic behavior—e.g., by making altruism easily induced—they cannot determine altruistic behavior with certainty. Mutual benefit between unrelated individuals in a society still depends on some form of reciprocity. Along with genes for altruism, genes for remembering the other members of the group and for accounting of debts and dues must be selected, all as a cluster. Knowing who is who and who did what for whom makes it possible for even a group of unrelated lions to band together, capture large prey, protect one another, and feed one another's offspring. (Of course, some relatedness helps.)
Not only altruism but also much other social behavior is selected when societies are beneficial to fitness. In his studies of marmots, David Barash (1982) found dramatic differences between woodchucks, which are solitary, and Olympic marmots, which are social. Woodchucks live in low-lying fertile areas with relatively long growing seasons, whereas Olympic marmots live high in the mountains where the growing season is short and the weather severe. Woodchucks apparently manage well in the milder climate on their own. They maintain territories from which they exclude others of their kind. Males and females come together only to mate, and females keep their young with them only until they are weaned, when the offspring disperse. For them, the costs of a social existence would outweigh the benefits. In Olympic marmots, selected along with warm coats are the necessities of living in groups: greeting calls, recognition of members, alarm calls, group maintenance of burrows, food sharing, and cooperative defense. Along with this, offspring typically stay with the group through two or three growing seasons; presumably they cannot be brought to maturity quickly with this harsh climate's limited resources. Barash theorizes that the slow maturation may be the key factor that makes the benefits of society outweigh the costs.