Back in 1952, Patti Page immortalized a pet for sale in “How Much is that Doggie in the Window?” But today’s pet owners often downplay this monetary exchange. When selecting a dog, we look for signs that he’s “the one”—perhaps he climbs out of the puppy pile and straight into our arms, or it’s kismet when his eyes meet ours. Ascribing agency to an animal allows us to make the romantic leap that this pet has chosen us, just as we’ve chosen it. Since pets are supposed to be our best friends (not our slaves!), a pet’s will matters to us.
But what if the pet doesn’t actually know it is a pet, because it lacks a self? This question of agency has become increasingly complex in the age of social media, augmented reality and robotics. Many animal lovers find themselves unable to care for a pet due to a host of life situations: infirmity, allergies, a lack of time or money, travel, fear of an aggressive animal or a prohibition in their dwellings. Though dogs’ and cats’ places in our lives may never be entirely replaced, we must reconsider our traditional notions of pets to account for the recent proliferation of virtual pets, cyberpets and companion robots.
Not surprisingly, the concept of a “pet” is difficult to define—probably because it means so many different things to different people. Oxford Dictionary offers three definitions:
None of these meanings is limited to a specific type of being, but all three emphasize the emotional interest of the pet owner. “Petness”—as sociologist Jen Wrye puts it—is a social construction relating to the subject (the pet keeper) rather than the object (the pet), so if a specific animal, vegetable, mineral or cyber specter feels like one’s pet, then it is.  Even in situations where a non-sentient pet will never comprehend its role, petness still applies. Hence, Wrye insists that relationships with virtual, cyber and robotic animals are not merely “petlike,” but pets in their own right. Let us discuss a few.
Over 8,000,000 Tamagotchis—tiny digital creatures housed in an egg-shaped computer—have sold since their conception in 1996. Invented by a mother of young children living in a small Japanese apartment, Tamagotchis require great care (feeding, bathing, medicine, exercise), and can even become sick and die.  Despite Tamagotchis’ digital existence, many owners become devoted to their pets, and mourn their deaths. Perhaps the time and energy required to care for a Tamagotchi encourages the owner-pet bond, and Tamagotchis’ ability to express emotion (they possess a happiness meter) engenders an interactive experience.
Other virtual pets exist solely online, such as Neopets. Since disembodied Neopets function more like avatars in an online game, Wrye celebrates their engagement in fantastical situations that would be impossible in the material world. But because Neopet interaction resembles video game play, caring for them is synonymous with winning the game. When playing a video game, one invests time and energy toward staying alive, so when a Neopet dies, this signals the loss of the game rather than the death of a cherished one. That’s not to say that a Neopet owner could not experience real emotional bonds, or that Neopets aren’t really pets, but interacting with them seems markedly different from traditional pet play.
Relationships with embodied pets come closer to the type of interactivity we expect in pets. For example, a team of sociologists led by Peter Kahn found that young children flinched apprehensively in response to the Sony Aibo’s movements, even though they fully understood that the dog was mechanical. While the children in the study did not respond this way to plush animals, the Aibo tapped into behaviors associated with live dog play. We already have an idea of how we might interact with a dog, so the Aibo allows us to access this prior knowledge. For Tamagotchis and Neopets, on the other hand, we have to learn how to play with them. But this is a double-edged sword, for our expectations can lead to disappointment if the robotic dog won’t perform exactly like a real dog. The Aibo is neither soft nor cuddly, nor can it do everything a real dog does, but future robotic dogs – such as those from Tombot – will become increasingly convincing.
Moreover, a robotic dog can be programmed to be less needy and more affectionate than a living dog. Caring for a live pet is costly, but the only resource the Aibo consumes is electricity. There are no stray, rabid Neopets wandering the streets, nor do Tamagotchis maul postal workers. And without a self, we needn’t fret about their well-being. Consequently, PETA has advocated virtual pets for owners who cannot or should not keep a live pet.
Bioethicist James Hughes cautions against one potential pitfall of virtual pets: that people might (legally) abuse them to the point that we become desensitized to the mistreatment of live animals and people. This would echo psychologists’ findings that repeated exposure to video game violence desensitizes people to violence. And though non-sentient pets can be programmed with “needs,” it seems unlikely that owners would put virtual pets’ desires before their own. Kahn, therefore, deems relationships with traditional live pets more “moral” than those with robotic pets, for the live animal “teach[es] children that their own desires don’t always come first.” But children already have to be taught to respect live animals—as well as other humans, toys and property. So too must parents teach their children how to treat and value a robot.
Defining a pet-owner relationship solely on the basis of the owner’s agency may strike some as superficial, and Kahn and others suggest that deep attachments to non-sentient virtual pets are less appropriate or healthy than those with live animals. But this viewpoint seems outmoded. Not long ago, humans were derided for sleeping with and mourning their pets, whereas today’s domesticated dogs and cats have been promoted to members of the family. While one might argue that devoting valuable resources to the care of live animals is misguided when millions of humans are starving or threatened, pets are so socially acceptable that few Westerners complain. Perhaps we are now on the verge of a virtual world epoch shift analogous to the decades-old pro-animal movement responsible for PETA, no-kill shelters and doggie daycare.
Even if we accept virtual pets, we must ask whether our relationships with them are inherently one-sided. How can virtual pets affect the world if they lack a self? Social scientists have posited that selfhood is not necessary to exist in the social realm. Take online bots, for example: though disembodied and non-sentient, they possess images and distinct personalities much like those of humans. Social media sites like Facebook and Twitter constantly weed out bot accounts, for humans take offense to bots masquerading as humans (decades of sci-fi featuring androids living among humans have probably heightened our fears). If a bot can pass as a human, then we seem replaceable, therefore lessening the perceived value of humankind. Perhaps transparency is indeed the key to incorporating non-sentient bots into our lives, for this enables us to categorize and treat them according to our value system.
But here’s the rub: we already understand that the animals we live with are not humans, yet we still tend to anthropomorphize them (and it is socially acceptable to do so). There’s a certain alchemy in pet ownership—we live with an animal, learn to appreciate its distinct quirks, and eventually it becomes much more than an animal to us. No one tells us how to feel about a pet; perhaps we fall in love immediately, or a bond forms over time. The same may apply to non-sentient pets. Though they cannot think for themselves, they still cast a spell over us—a bewitchment we have enabled through our constant attention to them.
We wonder why millions of people devote time and energy to pets that only exist in cyberspace, but we still don’t fully understand why people keep live pets in the first place. Though some pets are “functional” (i.e., herding dogs on a farm), urban pets are a losing economic prospect. And yet, Americans take on more and more pets every year. One theory suggests that dogs’ and cats’ large eyes and heads resemble human babies, tapping into our drive to care for others, and possibly stimulating the release of oxytocin. According to Edward O. Wilson, pet keeping is a natural phenomenon for humans, possibly related to our inherent love of nature—what he terms “biophilia.”
But does “biophilia” extend to our love of the unreal—seemingly the very opposite of nature? Consider how we selectively breed traits into dogs to make them appear more appealing to humans, such as the furrowing of the brow. A furrowed brow lifts the face and makes the eyes appear larger, like those of an infant. Hence, we manipulate nature through breeding, which resembles software revision in a virtual pet.
If keeping a live animal connects us to our animalistic roots, then perhaps taking a robotic or cyberpet signals the melding of the virtual, mechanical and natural. Cyberpets allow our virtual imaginations to become equally real and meaningful by affecting us socially, and, likewise, robotic pets represent the ultimate achievement of human ingenuity. When we look into the eyes of a live dog, we connect with nature, and when we look into the eyes of a robotic dog, we see ourselves in nature.
 Oxford Dictionaries.com. https://en.oxforddictionaries.com/definition/pet. Accessed 9/10/2018.
 Wrye, Jen. “Beyond Pets: Exploring Relational Perspectives of Petness.” Canadian Journal of Sociology/Cahiers Canadiens de Sociologie 34/4, 2009, 1043.
 Wrye, 1037.
 Wrye, 1059.
 Bloch, Linda-Renée and Dafna Lemish. “Disposable love: The rise and fall of a virtual pet.” New Media and Society 1(3), 1999, 284.
 Wrye, 1048.
 Kahn, Peter, Batya Friedman, Deanne Perez-Granados and Nathan Freier. “Robotic Pets in the Lives of Preschool Children.” Interaction Studies 7(3), 2006, 424.
 Macdonald, G. Jeffrey. “If you kick a robotic dog, is it wrong?” The Christian Science Monitor. 2/5/2004. https://www.csmonitor.com/2004/0205/p18s01-stct.html. Accessed 9/2/2018.
 Qtd. in Macdonald.
 It still remains unclear whether this syndrome actually causes further violence. See Carnagey N. L., Anderson C. A., Bushman B. J. “The effect of video game violence on physiological desensitization to real-life violence.” J. Exp. Soc. Psychol. 43, 2007, 489–496. 10.1007/s10964-014-0202-z
 Qtd. in MacDonald.
 Wrye notes that “nonliving pets are either unnoticed or dismissed as inferior and trivial,” 1034.
 For example, see Hacking, Ian. Social Construction of What? Cambridge, MA: Harvard University Press, 1999.
 Nagasawa M, Mitsui S, Shiori E, Ohtani N, Sakuma Y, Onaka T, et al. Oxytocin-gaze positive loop and the coevolution of human-dog bonds. Science 348, 2015, 333–6. doi:10.1126/science.1261022
 Biophilia is definied as “the urge to affiliate with other forms of life.” Biophilia. Cambridge: Harvard University Press, 1984, 416.