How Stress Adversely Affects Learning
Why Is It Easier for "Submissive" Dogs to Solve "The Detour Test?"
“Everything I ever said about dogs was wrong.”
Deconstructing Myths and Misconceptions My primary goal for this blog is to deconstruct some of the major myths and misconceptions I think people have about canine behavior in general and dog training in particular. There are a lot of dogs whose lives, I think, could be greatly improved if we only understood things from their unique point of view. To my way of thinking one of the biggest myths is the idea that dogs form dominance hierarchies based on rank and status. So imagine my surprise and delight to find a scientific study that proves my point exactly, though quite accidentally.
The study was done in 2008 by The Family Dog Project at Eötvös Loránd University. It’s titled “How does dominance rank status affect individual and social learning performance in the dog (Canis familiaris)?” It suggests that it’s easier for so-called subordinate dogs to learn a problem-solving task—known as the detour test—by observing another dog solve the problem than it is for a dominant dog to do so. Yet both dominant and subordinate dogs were able to solve the problem if they watched a human being demonstrate the solution.
The detour test involves placing some food, which was clearly visible yet placed behind a V-shaped face. The dogs were unable to access the food directly, and had to learn how to go around the fence to get to it.
The dogs were taken from multiple-dog households. The owners filled a questionnaire to determine which dogs were the most dominant.
From the abstract: “In Experiment 1, dominant and subordinate dogs were tested without demonstration and we did not find any difference between the groups—they had similarly low detour performances on their own. In Experiment 2 and 3, dogs from single dog and multi-dog households were tested in the detour task with demonstration by an unfamiliar dog, or human, respectively. The results showed that … Subordinate dogs displayed significantly better performance after having observed a dog demonstrator in comparison to dominant dogs. In contrast, the performance of dominant and subordinate dogs was almost similar, when they observed a human demonstrator.”
Dominance is said to give animals an evolutionary advantage. The ability to learn new behaviors also serves a clear adaptive purpose. Yet in this particular context being dominant actually decreases a dog’s evolutionary fitness!
Determining Dominant Status
The fact that each dog’s “status” was determined by the owners is, I think, problematic. First of all, “dominance” is no longer defined as a fixed character trait but as a function of relationships. This is especially true in multiple-dog households where there is no dominance hierarchy per se, but a hierarchy of predilections: one dog is dominant over food, another dominant over toys, a third dominant over the couch, and so on.
Another problem is that in some wolf packs there is no apparent hierarchy at all and no displays of dominance or submission between pack members. Dr. David Mech said that in 13 summers of observing the wolves on Ellesmere Island he saw no displays of dominance. Yet in another study he says that dominance is a chief feature of wolf-pack behavior.1
Why the difference?
We know that in nearly every so-called dominance hierarchy on earth—from human beings down to ants—the most dominant member of the group produces the highest levels of stress hormones called glucocorticoids, one of which is cortisol. And yes insects also experience stress, and when they do, their bodies also produce stress hormones called ecdysteroids, which are analogous to the glucocorticoids found in mammals. This suggests that dominant behaviors—whether in dogs or ants—are, primarily, by-products of stress. And since dominant behaviors in dogs can be greatly moderated—or eliminated completely—through anti-anxiety medications or through proper training, I think it’s safe to say that these labels are meaningless except as diagnostic tools for determining how stressed a dog (or an ant) might be.
Could it be that the so-called dominant dogs weren’t really dominant at all, but anxious or stressed? After all, this study shows that dominant status has an adverse effect on learning, and there are lots of data showing that stress inhibits an animal’s ability to learn.
Definitions of Dominance
One definition of dominance is the ability to control access to resources. So why would a dominant dog, who’s put in a position to see and smell food, not show any interest in being able to control access to it when an unknown, yet supposedly subordinate, dog finds his way around an obstruction quite easily? Surely in a natural situation, the “leader” of a wolf pack wouldn’t be likely to ignore and/or be dismissive of a subordinate wolf’s ability to gain access to free food.
The problem there is that the dominant and subordinate dogs in this study weren’t part of the same social group, and dominance is group specific: one wolf can only be said to be dominant over a member of his own pack, not members of other packs. Since the demonstrator dogs were strangers to the dogs who watched them, the rank and status of such dogs would be meaningless. Yet the study suggests that the dominant dogs were capable of learning from watching the humans demonstrate the solution because they saw them as having a higher rank and status than themselves.
This is pure idiocy. Just as a wolf can’t exert dominance over members of other packs, he can’t dominate—or be dominated by—members of other species. So the only way a dog could see a human being as being higher in “rank and status” would be for the dog to somehow mistake that human for another dog. Plus, the human demonstrators in this study weren’t members of the dog’s household; they were strangers, and as such would have no rank or status or be a part of their social group.
So why then did the dominant dogs ignore the puzzle-solving solution demonstrated by unknown, “subordinate” dogs but learned how to gain access to the food from watching the human beings?
Attraction & Resistance v. Dominance & Subordinance
If we look at this learning problem through the principles of attraction (flow) and resistance (statis), rather than dominance and subordinance, we might be closer to the truth.
The food creates a feeling of attraction in each type of dog. The fact that his access to the object of attraction is being blocked creates feelings of resistance, or frustration. The dog has three choices: stay frustrated, solve the problem, or find something else to do, like sniff around or pee on something.
The other part of the puzzle is that objects of attraction are things that we project our wants and desires onto. We literally project our emotional energy onto the things we want. The same is true for dogs. So in this case, each type of dog—emotionally open (subordinate) or closed (dominant)—is projecting his emotional energy onto the food.
However, when an open dog solves the puzzle, and is being watched by another dog of the same type, there is an open flow of information. The second dog is quite probably projecting his emotions not only on to the food but onto the other dog’s behavior, physical momentum (movement around the barrier) as well. This makes it quite easy for him to learn the solution. Since no dog sees himself as being separate from other dogs or from things in the environment, it should be quite easy for him to project his emotions or emotional center of gravity onto the object of attraction. Problem solved.
That said, when a closed (or dominant) dog watches an open dog solve the puzzle, he’s somewhat incapacitated by stress and is incapable of forming an emotional connection between the food and the other dog’s behavior. His behavior is static, not open to flow.
Then why do both types of dogs find it easier to solve the puzzle by watching a human being?
I think there are two reasons. One is that puppies are always open, not closed. They’re in a constant state of attraction (flow) with very little resistance (static). However, as objects of attraction, human beings also carry a great deal of resistance for the pup. We’re much bigger and taller than they are, yet we’re constantly interacting with them in ways that their mothers and littermates are completely incapable of doing.
Does the pup want to go outside? Yes. How does he get there? We put a leash on him and open the door. Is the pup hungry? Yes. How does he eat? We put his food in a bowl and place it on the floor where he can reach it. Is he thirsty? We put his water down, etc, etc, etc.
So the feelings of attraction and resistance dogs have for human beings are much stronger (at least in some ways) than the feelings they have for other dogs. This is particularly true when the dog has a problem to solve. In most cases, humans are more capable of solving that problem than other dogs are, and dogs sense this due to an accumulation of sensory data (i.e., experiences where humans solved the dog’s problem) taking place over time.
So when the closed (i.e., stressed, or dominant) dog in the study isn’t capable of being in the flow with the subordinate (open) dog, but is capable of being in the flow with the strange human, it’s due to three things: the closed dog is more attracted to 1) things that create feelings of resistance than things that create feelings of flow, 2) the higher level of attraction and resistance felt toward the strange, unknown human being than is felt toward the open dog, and 3) the ways that the closed dog—like all dogs—has been conditioned to see human beings as problem solvers.
As far as I can tell, that’s the only way that the outcome of this study makes any sense.
Lee Charles Kelley
“Life Is an Adventure—Where Will Your Dog Take You?”
1) In a 2010 study by David Mech and Dean Cluff, they observed an older wolf, possibly the "pack leader," repeatedly pin a younger member. The reason for this behavior was hypothesized to be “pack dispersion.” This suggests that the pack had become too large to sustain itself; possibly because resources were scarce. If that’s the case, this is another indicator that dominant behaviors are stress-related.