The surprising evolutionary history behind dogs making “puppy eyes”

Have you ever looked at your dog’s face and thought that there definitely is a human soul trapped in a dog body? Indeed, the dog-human bond is something of an anomaly in how easily two completely different species can mutually understand — and love — each other.

Humans have domesticated wolves for at least 33,000 years, selectively breeding for various traits depending on the job humans needed a dog to accomplish. But one possible trait that’s been overlooked is expressiveness in dog faces. But now, a team from Duquesne University presenting research at the American Association for Anatomy annual meeting during the Experimental Biology 2022 meeting on April 5 looked at how facial muscles in dogs differed from those in wolves.

They found that humans possibly bred dogs to emote more clearly with their faces, creating dogs with a different facial skeletal muscle composition from wolves.

What’s new — Dogs, according to this research, evolved to have more fast-moving facial muscles that create expressions like human faces do. The team studied mimetic muscles, which are the tiny muscles that form quick, nuanced facial expressions.

Puppy-dog eyes, for example, come from this evolved ability, the authors say. One of these mimetic muscles dogs possess is the levator angulis oculi medialis, a muscle near the eye that creates this irresistible facial expression, which “creates a nurturing response in humans because it replicates a sad look that they're giving us,” co-researcher Madisen Omstead, a researcher at Duquesne University, tells Inverse.

In skeletal muscles in the face, there are fast-twitch and slow-twitch myosin fibers in these mimetic muscles. Fast-twitch fibers manipulate parts of the face where emotion may flash across, like the brow or lips. These fibers tire quickly, which is why humans cannot hold a facial expression for very long, at least without it looking forced.

Slow-twitch muscles, on the other hand, control longer, more sustained movements. For example, domesticated dogs communicate more by short, rapid barking because of “talking” fast-twitch muscles; wolves, while they can bark, communicate more with extended howling, created with slow-twitch muscles.

This team compared the presence of these fibers in facial muscle samples from wolves and domesticated dogs. They found that dogs had predominantly fast-twitch myosin fibers like humans do. While wolves also have abundant fast-twitch fibers, they have a much higher proportion of slow-twitch fibers.

Why it matters — A few small face muscles can speak volumes about dog evolution, the process of animal domestication, and even the origins of human behavior since 50,000 years ago.

“We think that humans had selectively bred dogs to make these faster facial expressions, so that may be why we're able to read them better,” Omstead says.

Dogs are better at recognizing emotions in human faces than chimps are — and we’re better at recognizing dog emotions than chimp ones. “You would think that we would be better with our closest living relatives, but we're better with dogs, and a bunch of it just has to do with the neurobiological underpinnings,” Anne Burrows, the study’s supervising author and a physical therapy professor at Duquesne University, tells Inverse.

There’s also the question of why dogs, as opposed to cats, horses, cows, and other domesticated animals, possess this augmented ability to emote. Dogs largely weren’t domesticated to be food animals, unlike goats and cows; in archaeological findings, there aren’t signs of dog bones scored with knife marks from a butcher. Burrows hypothesizes that dogs were domesticated not for food, but for companionship. Humans’ relationship with dogs evolved as both species evolved, and so did our abilities to understand each other.

Digging into the details — To assess how these muscles differed between dogs and wolves, the team looked at the distribution of slow-twitch and fast-twitch myosin fibers in sample face muscles from dogs and wolves. They found that fast-twitch fibers in dog samples comprised 66 percent to 95 percent of myosin fibers; wolf samples averaged 25 percent. On the other hand, slow-twitch fibers in dogs made up on average 10 percent of myosin fibers, whereas wolves averaged 29 percent.

Burrows says that the range of fast-twitch fibers in dogs could come from their lineage. For example, there are ancient dog breeds such as huskies that are more closely related to wolves. Fast-twitch fibers may make up a lower percentage of face muscles in huskies because they’re genetically more similar to wolves; this also could be why husky owners observe that their dogs “talk” rather than bark. This talking action resembles howling.

On the other hand, since fast-twitch muscles are considered talking muscles that help dogs bark, smaller dogs with a reputation for yappiness may have a higher percentage of fast-twitch fibers. That could be relevant to why they communicate so much with constant barking.

Still, this research requires more digging, and some rigorous scientific replication. Animal behavioral scientist Claudia Fugazza, who was not involved in the work, writes to Inverse, “While this is a potential interesting study showing difference in facial muscle composition between wolves and dogs, it is too early to claim that this is involved in communication. This would have to be tested and proven.”

What’s next — Ears are another telling point of a dog’s emotions, as well as evolution. Wolves always have pointed ears, but only some domesticated dog breeds, like golden retrievers, have floppy ears. Even today’s ancient dog breeds can tell us secrets about this trait. “We're always looking to get more information about the ancient dog breeds, because they hold information about the domestication process that we might not see in just your typical everyday breed,” Burrows says.

Of course, Burrows and Omstead are dog owners themselves, and in a sense get to immerse themselves in this research every day. Burrows credits her nine-year-old husky Ruby as the inspiration for this study. “She’s actually the reason I started thinking about all of these things regarding how dogs use their muscles because as huskies will do, she does a lot of yodeling, and just generally talking a little bit more than she actually barks.”

Omstead’s four-year-old golden retriever Finnigan also plays a role in the scientific world: He got to bark at the presentation of this research.

Editor's note: This article has been updated to properly reflect Madisen Omstead's position at Duquesne University.