In the target article, Richerson et al. argue that cultural group selection (CGS) was essential for the evolution of the high level of cooperation seen in our species, as well as the development of institutions such as religions and businesses. Here I make the case that CGS also provides a coherent explanation for the evolution of a uniquely human form of social relationships – pet-keeping. Pet-keeping poses an evolutionary problem. There is no evidence that pets increase the reproductive potential of their owners, and the putative health and psychological benefits of living with animals are not as well-established as commonly believed (Herzog Reference Herzog2011). Maintaining a companion animal can be costly, and devoting affection and resources on members of another species that have no apparent function cannot be explained by kin selection or reciprocity.

Several lines of evidence suggest that CGS may have played a role in the emergence and spread of pet-keeping. First, pet-keeping occurs in humans only. While nonhumans occasionally form attachments to members of a different species, these unusual relationships always seem to involve some human intervention such as occurs in zoos, wildlife parks, and households (Herzog Reference Herzog2014). In completely natural environments, “animal odd couples” analogous to pet-keeping are extremely rare and probably non-existent. Nonhuman species do possess prosocial motivational systems such as parental care and empathy that are prerequisites for the formation of interspecific attachments (e.g., Preston & De Waal Reference Preston and De Waal2002). But the fact that long-term affectionate relationships between non-conspecifics have not been observed in the wild, even in nonhuman apes, suggests that pet-keeping requires the degree of rapid social learning found only in our species.

Second, unlike language, music, and prohibitions against sex with first-degree relatives, pet-keeping is not a human universal. Indeed, the frequency and form of pet-keeping vary widely among cultures. Indigenous peoples of Amazonia, for example, commonly make pets of captured wild animals such as monkeys and birds. This practice, however, is rare among tribal peoples of New Guinea (Diamond Reference Diamond, Kellert and Wilson1993). In some cultures, pet-keeping is unknown, and some languages have no word that corresponds to “pet.” Cross-cultural studies reveal that practices common in modern Western societies, such as letting animals live in our homes, allowing them to sleep in our beds, and considering them full-fledged family members are cultural anomalies (Gray & Young 2010). Further, social norms governing relationships with pets can shift dramatically over time. In Medieval Europe, cats were routinely subjected to torture and mass slaughter, as they were linked with witchcraft. Their image only began to change in the mid-17th century, with the complete transformation of the cat from pariah to household pet occurring by the end of the 18th century (Bradshaw Reference Bradshaw2013).

Third, preferences for pets are transmitted vertically from parent to child and horizontally across cultural landscapes. A recent study found that mothers who had pets in their childhood, tended to raise their offspring with pets (Westgarth et al. Reference Westgarth, Heron, Ness, Bundred, Gaskell, Coyne, German, McCune and Dawson2010). Sudden shifts in the popularity of types of pets occur across cultural groups via social transmission. In the United States, examples include transient enthusiasms for baby turtles, caged songbirds, ferrets, mice, and pot-bellied pigs. Using 60 million American Kennel Club dog registrations between 1926 and 2005 as data, my colleagues and I have shown that shifts in choices for canine companions are influenced by the mechanisms Richerson et al. propose as sources of intergroup variation. These include rapid social learning, conformity, and the influence of prestigious models. We have found that dog breed preferences are influenced by the decisions of others (Herzog et al. Reference Herzog, Bentley and Hahn2004); are unrelated to breed-characteristic qualities such as desirable behavioral traits and frequency of genetic diseases (Ghirlanda et al. Reference Ghirlanda, Acerbi, Herzog and Serpell2013); and result in fads in which the more quickly a breed becomes popular, the faster its appeal subsequently declines (Acerbi et al. Reference Acerbi, Ghirlanda and Enquist2012; Herzog Reference Herzog2006). In addition, long-lasting changes in the popularity of types of dogs are sometimes instigated by media exposure (films with canine movie stars) (Ghirlanda et al. Reference Ghirlanda, Acerbi and Herzog2014). In short, pet-keeping follows the laws of fashion.

Fourth, while it is unlikely that cultural differences in pet-keeping are the direct result of intergroup competition, competitive forces do shape aspects of human–animal interactions that are precursors to affection. The use of hunting dogs varies widely among indigenous peoples of the Neotropics. Dogs substantially increase the success of Mayangna hunters in Nicaragua, and while the Mayangna are not attached to adult dogs, they do name their dogs and play with puppies (Koster Reference Koster2009). The acquisition of horses by Apache and Comanche Indians in the 17th century profoundly changed the nature of tribal warfare on the Great Plains of the United States (Hämäläinen Reference Hämäläinen2003).

Finally, pet-keeping practices are affected by some of the institutions that Richerson et al. believe arose via CGS, for example, religion and businesses. On the multi-ethnic island of Sri Lanka, 79% of Buddhist households include a dog, compared to Muslim households, of which only 4% do (Knobel Reference Knobel2009). In the United States, over the last two decades, the proportion of homes with companion animals has increased only slightly while the amount of money Americans spend on their pets has risen fivefold. It is likely that efforts by industry trade groups such as the Pet Products Manufacturers Association and the American Veterinary Medical Association to promote the benefits of pet ownership have contributed to the recent cultural trend sometimes referred to as the “humanization of pets.”

In sum, pet-keeping shows the marks of a cultural group–selected institution. It involves an inheritance system, socially transmitted norms, a high degree of between-group variation, and (albeit indirect) intergroup competition. Certainly, pet-keeping has evolutionary and neurobiological roots (see, e.g., Archer Reference Archer1997; Stoeckel et al. Reference Stoeckel, Palley, Gollub, Niemi and Evins2014). But the large individual and cultural differences observed in attachment to animals suggest that learning and culture play a larger role than biology when it comes to bringing them into our lives. CGS offers an elegant explanation for an important cultural phenomenon that has, until recently, been neglected by social and behavioral scientists.