"The Canine-Human Bond: Why Dogs Are 'Smarter Than Chimps'"
If Wolf Packs Are Self-Emergent Systems, Are Dog-Human Packs Too?
Originally published in slightly different form on October 26, 2011 at PsychologyToday.com.
Computer Model Suggests Different View of Wolf Pack Hunting Behaviors The traditional view of pack hunting behavior is that it’s a highly-intelligent, carefully orchestrated endeavor, controlled primarily by the pack leader, which requires constant communication between members of the pack, much like a platoon of soldiers armed not only with rifles but walkie-talkies, the only difference being that wolves supposedly communicate through body language and eye contact, not two-way radios. A new computer model suggests that none of this may actually be true.
Let’s look at a common example of a wolf pack hunting moose or bison. A group of wolves approaches a herd of bison or a lone moose. They first get the herd, or the lone moose, to start running. They then give chase, yet begin taking different paths, flanking the prey animal from all sides. Finally, the main body of wolves chase the animal into a trap, where another wolf is lying in wait.
This certainly seems like a sophisticated, well-thought out plan of attack. But according to a new model—created by Cristina Muro, a dog trainer from Bilbao, Spain, along with two Spanish computer scientists, and American biologist Raymond Coppinger—a computer-generated, virtual pack can replicate a real wolf pack’s behavior just by obeying two simple rules. So simplicity, not sophistication, may be the wolf’s secret to success.
Below are the abstract and highlights from this just-published study. (“Wolf-pack (Canis lupus) hunting strategies emerge from simple rules in computational simulations.” October 17, 2011.)
Abstract We have produced computational simulations of multi-agent systems in which wolf agents chase prey agents. We show that two simple decentralized rules controlling the movement of each wolf are enough to reproduce the main features of the wolf-pack hunting behavior: tracking the prey, carrying out the pursuit, and encircling the prey until it stops moving. The rules are (1) move towards the prey until a minimum safe distance to the prey is reached, and (2) when close enough to the prey, move away from the other wolves that are close to the safe distance to the prey. The hunting agents are autonomous, interchangeable and indistinguishable; the only information each agent needs is the position of the other agents. Our results suggest that wolf-pack hunting is an emergent collective behavior which does not necessarily rely on the presence of effective communication between the individuals participating in the hunt, and that no hierarchy is needed in the group to achieve the task.
Highlights ► We present a multi-agent model where wolf agents obey two rules to hunt a prey agent. ► The first rule is move towards the prey until a safe distance to the prey is reached. ► The second rule is when close enough to the prey, move away from the other wolves. ► These two simple rules are enough to reproduce the wolf-pack hunting ethogram. ► No communicative skills and no hierarchy are needed to complete the hunt.
Four Principles, Not Two Coppinger and company say that pack hunting behavior may be guided by these two rules. I would argue that there are actually four principles at play:attraction & resistance, and tension & release. Attraction and resistance explain the way wolves chase prey (attraction) yet maintain a safe distance until the prey is finally weak enough for the kill (resistance), and also why they maintain a certain distance from one another. Tension and release explain what drives their behavior in the first place: tension builds via hunger and is released via the “kill bite.”
I would also propose that there’s another key factor to the wolf pack’s success, the ability for each member of the pack to tune in to the emotional wavelength of the prey, as well as the wavelengths of his fellow pack mates. The wolves operate independently only in the same way that a group of surfers, who are all riding the same wave, will take different courses along its trajectory.
Feral Robotic Dog Project Coppinger’s model is actually eerily similar to a 2002 experiment designed by Natalie Jeremijenko at Yale University, called “The Feral Robotic Dog Project.” It accidentally proved (or at least gave the first clear indication) that a wolf pack operates as a bottom-up, self-emergent system, not a top-down hierarchy, which is the same conclusion that Coppinger et al have recently come to.
Here’s how it worked. Robotic toy dogs were given some rudimentary programming designed to locate toxins. The robots were then given to New York City high school students, who “released them into the wild” (i.e., an abandoned gas plant near the Bronx River). The dogs were designed to exhibit two basic responses to the target material; a) detect it with their sensors, and b) when any one of the robot’s sensors was triggered, the others would gravitate toward that “dog.”
Jeremijenko wrote: “The primary behavior of each individual dog is the drive to locate a material that effects their specific sensor. Once a material is detected, further geometric sampling ... will seek to move the dog and the sensor closer to the source of the material. In order to collect samples from a larger area, the pack of dogs is programmed with the ability to follow at a distance the individual with the strongest sensor reading.”
Except for the rule about maintaining a certain distance from other pack members, this explains how there seems to be a pack leader when wolves hunt, and is very close to the way wolves hunt in Coppinger’s virtual model.
Since the purpose of the robotic dogs project was to implement a way of finding toxins without posing a danger to humans (or real dogs), not to simulate actual pack behavior, the results aren’t as clear cut as they’d be if the robots were programmed for other behaviors, yet the simple patterns of “behavior” the robots produced were in some ways strikingly similar to the way wolves pursue prey, at least in the initial stages of the hunt.
Loose Social Arrangements Meanwhile, Coppinger says that when wolves don’t need to hunt large prey, they don’t form packs. “They have loose social arrangements,” Coppinger said in a 2004 online discussion, hosted by The Washington Post, “but they’re not a real pack.”
Canine social interactions are wonderfully complex. But it’s not because dogs and wolves have pack leaders, or are capable of abstract or conceptual thinking, or have a rudimentary capacity for language. A simple look at the size and shape of their brains shows that they don’t have the cognitive architecture for it. What they do have--and this is more true of dogs than wolves--is the ability to interact in socially complex ways, with a highly adaptive and flexible system of responses. Chimps are less flexible because they don’t hunt large prey as a group, so they have no ability to tune into one another’s wavelengths the way wolves and dogs do. (Wolves do it while hunting; dogs do it most notably while engaging in free play.)
The Whole Is Smarter Than the Sum of Its Parts Why are dogs capable of more complex social behaviors than wolves or chimps? I think it’s because of their interactions with us. In a self-emergent system, the whole is always smarter than the sum of its parts, meaning that individual wolves aren’t super-intelligent, but that the wolf pack is.
The same basic principle holds true for the “packs” that dogs form with us. And it’s why so much of dognitive science is presently devoted to the unfounded and somewhat irrational idea that dogs are smarter than chimps, despite the fact that a dog’s brain is much smaller and far less-developed. The truth is that, individually, dogs are nowhere near as “smart” as chimps. But when they interact with us, they are. That’s because the dog-human bond also operates as a self-emergent system.
Dogs get under our skin. They sort of hijack our brains to the point that we’re no longer two separate beings. Just as in a wolf pack when it’s hunting large prey, there’s an actual group mind at work. That’s what makes dogs seem “smarter” than chimps. That’s their true magic.