Unlike porcupines, dogs are a relatively hands-on (actually, paws-on) species, both with one another and with us. YouTube has numerous videos of dogs essentially saying, “Just keep petting me, please. Yes, that’s it…more.”

But this relationship is not one-sided. Many studies find that positive interactions between people and dogs can be beneficial for both species. Increases in β-endorphin (beta-endorphin), oxytocin and dopamine—neurochemicals associated with positive feelings and bonding—have been observed in both dogs and people after enjoyable interactions like petting, play and talking. Essentially, interacting with a dog, particularly a known dog, can have some of the same psychophysiological markers as when two emotionally attached people spend time together.

But do certain types of interactions have an outsized impact? Dogs are incredibly attentive to human faces and, in some cases, even specific facial expressions. This seemingly routine, benign behavior—your dog turning to gaze on your beautiful face as you do his or hers—could actually hold a very important piece of the puzzle in our relationship with dogs, suggests a study published this week in Science.

The new study, by Miho Nagasawa of Azabu University in Japan and colleagues, builds on Nagasawa’s previous work, published in Hormones and Behavior in 2009, that found owners and dogs sharing a long mutual gaze had higher levels of oxytocin in their urine than owners of dogs giving a shorter gaze. (Oxytocin, a humble peptide of nine amino acids that is sometimes called the “cuddle hormone,” has been implicated in social bonding and is instrumental to the cascade of hormonal changes leading up to and following birth.) Nagasawa and her colleagues concluded that their finding was “a manifestation of attachment behavior.” By describing it in this context, the researchers postulated that gaze between a dog and human (particularly a known human), will share similar properties to mother–infant relationships.

Nagasawa’s new study investigates whether a dog’s gazing behavior affected not just the owner’s oxytocin concentrations but the dog’s as well. In the first experiment the researchers collected urine from 30 dog-and-owner pairs before and after a 30-minute interaction. As in the earlier study, owners whose dogs showed the most gazing behavior had a notable increase in oxytocin concentration. But this time the researchers also found a similar increase in the neurochemical in the dogs.

A second experiment aimed to disentangle whether a causal relationship could be observed between mutual gaze and the release of oxytocin. Another set of 30 dogs was given an intranasal spray of either oxytocin or saline prior to interacting with people. They found that female dogs that sniffed oxytocin gazed longer at their owners than when given saline. As expected, this gazing also stimulated oxytocin secretion in the owner recipients of the gaze. The mutual effects were not seen between dogs and unfamiliar humans—and for reasons that require further investigation, they were not seen in male dogs and their owners. These sex differences were not observed in the first part of the experiment.

A story emerges—and probably one that will make dog lovers cheer: Mutual gaze between dogs and the people who care for them produces a very similar physiological profile to what’s observed between mothers and infants. This overlap could both contribute to and facilitate our intense and deep-seated relationship with dogs.

Reflecting on her findings in an interview conducted via Skype, Nagasawa recommends that “dog owners not just say commands at their dogs, but to build up the relationship [and] consider the potentially beneficial role that mutual gaze can hold.”

The paper feeds into an ongoing discussion among researchers about whether the biological synchronization observed between dogs and humans indicates “coevolution of human–dog bonds,” as the title of the Science study suggests. Nagasawa and colleagues also investigated whether the increased oxytocin observed in dogs appears in hand-raised wolves that have interacted with a known human. The wolves, however, rarely held a gaze with the humans for more than a few moments. This divergence led the researchers to postulate that “dog-to-owner gaze as a form of social communications probably evolved during domestication” with humans.

Testing evolutionary theories (particularly coevolution) is notoriously tricky. Whereas it is exciting to include socialized wolves in these studies, differences between dogs and wolves should not necessarily be immediately followed by the tooting of a coevolutionary horn. Zsófia Virányi, a senior research scientist at  the University of Veterinary Medicine, Vienna's Messerli Research Institute and a co-founder of the Wolf Science Center asks, “how much the differences we see are explained by evolutionary factors or differences in raising conditions.”

Researchers are finding more examples of areas where wolves perform successfully in sociocognitive tasks with humans, including attending to our social cues. For example, in a recent study Virányi found that both dogs and wolves learned from human demonstrators. In a chapter in the edited volume, The Social Dog: Behavior and Cognition, by Juliane Kaminski and Sarah Marshall-Pescini, Virányi and her colleague Friederike Range reflect on the numerous hypotheses attempting to understand dog domestication. They suggest “dog–wolf differences do not mean, however, that domestication is either necessary or sufficient to explain humanlike behavior in dogs.”

And then there’s also the numbers game. Although 60 dogs contributed to the current investigation, the coevolution question was ultimately tackled with just five human-reared wolves. The study began with 11 wolves—but guess what?: It’s hard to collect urine from a wolf. In one case a wolf’s urine was collected two hours after the desired time because the subject fell asleep — which is another way of saying the researchers maybe did not want to wake a sleeping wolf.

It would be useful to know more about the in-study behaviors of these five wolves. For example, were they exploring the novel environment where the testing took place or trying to get out, and could those factors contribute to why they did not orient toward their handlers? Virányi even wonders whether gaze would be the crucial factor in oxytocin effects between wolves and handlers. If other social exchanges besides gaze were tested, would a positive hormonal loop between wolves and humans appear? “It may not be fair to suggest the complete absence of an oxytocin-mediated positive loop in wolves as a species from these results,” says Monique Udell, an assistant professor at Oregon State University who has investigated human–wolf interactions at the research and conservation organization, Wolf Park. “We know that maternal–offspring attachment is important to this species and many other nondomesticated animals and that humans and wolves can show attachment bonds.” Additionally, wolf oxytocin levels, even before interacting with their handlers, are notably higher than that of all dogs tested. Maybe we simply haven’t discovered all the details of wolf oxytocin mechanisms yet.

In other words, don’t count wolves, or other species, out just yet. “I’m not convinced that this is something dog-specific,” Virányi adds. “The oxytocin system is so ancient that if socialization is there, then you can easily put a member of another species into these contexts.”