“It
is difficult to resist the idea that general principles underlie
non-hierarchical systems, such as ant colonies and brains. And because
organizations without hierarchy are unfamiliar, broad analogies between systems
are reassuring. But the hope that general principles will explain the regulation
of all the diverse complex dynamical systems that we find in nature, can lead
to ignoring anything that doesn’t fit a pre-existing model.”
—Deborah M. Gordon,
“Control Without Hierarchy,” (Nature, March, 2007.)
“Control Without Hierarchy,” (Nature, March, 2007.)
Prologue by LCK
I believe that the concept of dominance hierarchies in animal groups—particularly as it’s applied to dogs and wolves—is long overdue for the scientific scrap heap. Yet it persists.
Primatologist Thelma Rowell—who studied baboons, and whose observations overturned much of what was known about their social behaviors at the time—, felt that hierarchies should be labeled as subordinance or even “stress hierarchies.” (1974.)
She was ahead of her time (and still is).
In most animal hierarchies, the most dominant member generally produces the most stress hormone, cortisol. Of course, like many hormones, cortisol has several modes of operation. For instance, prolonged elevated levels lead to muscle wasting, and there doesn’t seem to be any evidence of that in alpha wolves (at least not while they’re young).
Another potential problem—as I see it—is that the social behaviors of animals have long been described through the principles of economics, as if a wolf pack, eg., were a market system. This is why we see scientific papers about “the division of labor” and “cheating” in wolf packs.
However, I believe that a better model might be to see the pack as an information system. If we apply that concept to alpha wolves—incorporating the seemingly contradictory ideas that these animals produce more cortisol yet seemingly show no ill effects from elevated levels—we might surmise that cortisol acts as a form of information, and that more dominant wolves may have more carrying capacity than other members of the group. After all, so-called alpha wolves are generally in the position of having to process more information about their environment than their subordinates are; being a leader means having to focus one’s attention on far more details.
Another possible way to reinvent or replace the alpha model has been proposed by veteran dog trainer Kevin Behan. In the following guest post, he suggests that we look at the pack as a flow system, based on Dr. Adrian Bejan’s Constructal Law: "For a finite system to persist in time (to live) it must evolve in a way that provides easier access to the imposed currents that flow through it."
I’m not convinced that the constructal law works here, but I’m willing to entertain any of Kevin’s ideas, no matter how strange they might seem to me at the time. That’s because whenever I apply them to dog training, they almost always work better than anything else that I’ve ever come across.
So here’s another well-thought-out guest post from Mr. Behan.
I believe that the concept of dominance hierarchies in animal groups—particularly as it’s applied to dogs and wolves—is long overdue for the scientific scrap heap. Yet it persists.
Primatologist Thelma Rowell—who studied baboons, and whose observations overturned much of what was known about their social behaviors at the time—, felt that hierarchies should be labeled as subordinance or even “stress hierarchies.” (1974.)
She was ahead of her time (and still is).
In most animal hierarchies, the most dominant member generally produces the most stress hormone, cortisol. Of course, like many hormones, cortisol has several modes of operation. For instance, prolonged elevated levels lead to muscle wasting, and there doesn’t seem to be any evidence of that in alpha wolves (at least not while they’re young).
Another potential problem—as I see it—is that the social behaviors of animals have long been described through the principles of economics, as if a wolf pack, eg., were a market system. This is why we see scientific papers about “the division of labor” and “cheating” in wolf packs.
However, I believe that a better model might be to see the pack as an information system. If we apply that concept to alpha wolves—incorporating the seemingly contradictory ideas that these animals produce more cortisol yet seemingly show no ill effects from elevated levels—we might surmise that cortisol acts as a form of information, and that more dominant wolves may have more carrying capacity than other members of the group. After all, so-called alpha wolves are generally in the position of having to process more information about their environment than their subordinates are; being a leader means having to focus one’s attention on far more details.
Another possible way to reinvent or replace the alpha model has been proposed by veteran dog trainer Kevin Behan. In the following guest post, he suggests that we look at the pack as a flow system, based on Dr. Adrian Bejan’s Constructal Law: "For a finite system to persist in time (to live) it must evolve in a way that provides easier access to the imposed currents that flow through it."
I’m not convinced that the constructal law works here, but I’m willing to entertain any of Kevin’s ideas, no matter how strange they might seem to me at the time. That’s because whenever I apply them to dog training, they almost always work better than anything else that I’ve ever come across.
So here’s another well-thought-out guest post from Mr. Behan.
Shifting Stands, Shifting Sands
The theory of dominance has shifted over the 50 years or so that I’ve been a dog trainer. It used to be about social superiority. Every individual was thought to be endowed with an inborn impulse to dominate others, as well as a counterbalancing impulse to submit once the dust had settled. A competitive struggle sorted this into a hierarchy of rank with a “top dog,” “alpha personality,” or “leader of the pack” at the peak of the pyramid.
The problem is that sometimes an inferior animal is able to control the behavior of its superiors by controlling access to certain resources. In this new way of looking at social behavior, a dominant individual doesn’t achieve status, it achieves access. And no individual is inherently dominant or submissive, rather there is a spectrum of “personality types.”
Emergence theory has also been applied to hierarchies in animal groups. In emergence theory, each relationship is determined by a local set of circumstances independent of the larger matrix of interactions. Dominance and submission emerge from such relationships as opposed to being some inherent, genetic quality contained within each animal.
The theory of dominance has shifted over the 50 years or so that I’ve been a dog trainer. It used to be about social superiority. Every individual was thought to be endowed with an inborn impulse to dominate others, as well as a counterbalancing impulse to submit once the dust had settled. A competitive struggle sorted this into a hierarchy of rank with a “top dog,” “alpha personality,” or “leader of the pack” at the peak of the pyramid.
The problem is that sometimes an inferior animal is able to control the behavior of its superiors by controlling access to certain resources. In this new way of looking at social behavior, a dominant individual doesn’t achieve status, it achieves access. And no individual is inherently dominant or submissive, rather there is a spectrum of “personality types.”
Emergence theory has also been applied to hierarchies in animal groups. In emergence theory, each relationship is determined by a local set of circumstances independent of the larger matrix of interactions. Dominance and submission emerge from such relationships as opposed to being some inherent, genetic quality contained within each animal.
The Bold and the Dutiful
A good summation of this new definition can be found in an online article entitled "Why Won't Dominance Die?" written by former police-dog trainer, David Ryan. It was written to discredit Cesar Millan’s approach to dog training.
In it, Ryan talks about the concept of dominance in dogs as a “meme,” a word coined by biologist Richard Dawkins to describe self replicating ideas that inhabit our minds and get passed along from one individual to another as if they were cultural viruses or genes. In Ryan’s view the dominance theory of dog training is a harmful meme, and like a super-virus, it’s extremely resistant to extinction.
“The concept of ‘dominance,’” Ryan says, “has never been a quality of an individual, but the product of a relationship. Ethologists label an animal dominant over another once there is a trend towards the second animal deferring in encounters between the two.”
He goes on to say that there are two types of dog, the bold and the shy, and that a “smooth relationship [between the two] is one in which each knows the other’s preferences and defers accordingly.”
A good summation of this new definition can be found in an online article entitled "Why Won't Dominance Die?" written by former police-dog trainer, David Ryan. It was written to discredit Cesar Millan’s approach to dog training.
In it, Ryan talks about the concept of dominance in dogs as a “meme,” a word coined by biologist Richard Dawkins to describe self replicating ideas that inhabit our minds and get passed along from one individual to another as if they were cultural viruses or genes. In Ryan’s view the dominance theory of dog training is a harmful meme, and like a super-virus, it’s extremely resistant to extinction.
“The concept of ‘dominance,’” Ryan says, “has never been a quality of an individual, but the product of a relationship. Ethologists label an animal dominant over another once there is a trend towards the second animal deferring in encounters between the two.”
He goes on to say that there are two types of dog, the bold and the shy, and that a “smooth relationship [between the two] is one in which each knows the other’s preferences and defers accordingly.”
Meet the New Boss, Same as the Old Boss
I would argue that we’ll never be able to replace the old notion of dominance with the one Ryan proposes because they’re essentially based on the same underlying “meme.” If dominance is about access to a resource (so that social life is not a constant struggle for status), don't all individuals crave access? Isn't it better to end up at the dominant end of a relationship and thereby enjoy unfettered access to resources?
Obviously yes. So the constant struggle for social ascendancy is merely being replaced by a constant struggle over resources. Of course the new school says that, no, social life isn’t a constant struggle because the dominant/subordinate relationship ameliorates aggression. But the old model said the exact same thing, and still does.
Plus, if dogs vary in terms of bold vs. shy (as opposed to dominant vs. submissive), how is a shy (i.e., inhibited) individual ever going to gain control over a resource?
It turns out that a shy dog can attain dominance by getting to the resource first, possession being “9/10ths of the law.” And, once in possession, a shy dog becomes emboldened, while the bold dog—whose access is now blocked—becomes more shy in nature. This suggests that control confers boldness and that lack of access confers its polar opposite. Remember, in the old definition, status confers, induces, or augments the trait of dominance. In the new definition access is the controlling factor.
So I would argue that just saying dogs vary in terms of bold versus shy fails to articulate the true dynamic from which the relationship emerges just as the old definition failed to do so.
I would argue that we’ll never be able to replace the old notion of dominance with the one Ryan proposes because they’re essentially based on the same underlying “meme.” If dominance is about access to a resource (so that social life is not a constant struggle for status), don't all individuals crave access? Isn't it better to end up at the dominant end of a relationship and thereby enjoy unfettered access to resources?
Obviously yes. So the constant struggle for social ascendancy is merely being replaced by a constant struggle over resources. Of course the new school says that, no, social life isn’t a constant struggle because the dominant/subordinate relationship ameliorates aggression. But the old model said the exact same thing, and still does.
Plus, if dogs vary in terms of bold vs. shy (as opposed to dominant vs. submissive), how is a shy (i.e., inhibited) individual ever going to gain control over a resource?
It turns out that a shy dog can attain dominance by getting to the resource first, possession being “9/10ths of the law.” And, once in possession, a shy dog becomes emboldened, while the bold dog—whose access is now blocked—becomes more shy in nature. This suggests that control confers boldness and that lack of access confers its polar opposite. Remember, in the old definition, status confers, induces, or augments the trait of dominance. In the new definition access is the controlling factor.
So I would argue that just saying dogs vary in terms of bold versus shy fails to articulate the true dynamic from which the relationship emerges just as the old definition failed to do so.
The Problem With Cesar Millan
Finally, Ryan’s piece was directed at Cesar Millan, our most famous proponent of the dominance model of dog training. But in Ryan’s critique what exactly is the beef? According to this new definition, Cesar is doing it right 99% of the time. (We should discount the really rough stuff because Cesar would argue these are last ditch cases about to be euthanized and represent but a small portion of his methodology even though they occupy a disproportionate share of the viewing time).
Cesar explicitly argues for a subtle manipulation of the innate desire within a dog to please its “pack leader,” along with massive doses of exercise. What’s wrong with that?
Of course Cesar is behind the times. He should be calling himself the pack parent rather than the pack leader. But he’s on solid behavioral ground according to both the new and old definitions of dominance. He controls a dog’s access to every resource and therefore he “emerges” as the dominant in this context, the dominant in that context, the dominant in all contexts.
He may be mistaken about a hierarchal pack leader, he may not be able to articulate that dogs are in a constant struggle for access to resources as opposed to social ascendancy, etc. But if the dog sees him as being in control of every resource then, operationally-speaking, what’s the difference? Cesar’s belief in his role as pack leader emerges from the social construct he engineers, and it’s engineered in accordance with the modern, accepted definition of dominance.
This is why I think Ryan and others will find it impossible to replace the old meme with this new one since they both have the same two fundamentals in common, a) control over another’s behavior, and b) the idea that sociability is about competition not cooperation.
In other words, dog owners are still being taught to see their dogs as rivals not partners.
Finally, Ryan’s piece was directed at Cesar Millan, our most famous proponent of the dominance model of dog training. But in Ryan’s critique what exactly is the beef? According to this new definition, Cesar is doing it right 99% of the time. (We should discount the really rough stuff because Cesar would argue these are last ditch cases about to be euthanized and represent but a small portion of his methodology even though they occupy a disproportionate share of the viewing time).
Cesar explicitly argues for a subtle manipulation of the innate desire within a dog to please its “pack leader,” along with massive doses of exercise. What’s wrong with that?
Of course Cesar is behind the times. He should be calling himself the pack parent rather than the pack leader. But he’s on solid behavioral ground according to both the new and old definitions of dominance. He controls a dog’s access to every resource and therefore he “emerges” as the dominant in this context, the dominant in that context, the dominant in all contexts.
He may be mistaken about a hierarchal pack leader, he may not be able to articulate that dogs are in a constant struggle for access to resources as opposed to social ascendancy, etc. But if the dog sees him as being in control of every resource then, operationally-speaking, what’s the difference? Cesar’s belief in his role as pack leader emerges from the social construct he engineers, and it’s engineered in accordance with the modern, accepted definition of dominance.
This is why I think Ryan and others will find it impossible to replace the old meme with this new one since they both have the same two fundamentals in common, a) control over another’s behavior, and b) the idea that sociability is about competition not cooperation.
In other words, dog owners are still being taught to see their dogs as rivals not partners.
Dominance and Submission as Forms of Flow?
I suggest we turn to Dr. Adrian Bejan’s book, Design In Nature, to help us see hierarchy from a new perspective. In it Bejan, argues that nature does not work according to principles of control but principles of flow.
For instance, in Bejan’s view a forest is a hierarchy of a few very large trees relative to many smaller forms of vegetation. The various plants aren’t competing for light, water or nutrients. The tallest ones aren’t trying to dominate the shorter ones. The forest simply emerges as the most efficient way to conduct and improve the flows of all currents contained within it (nutrients, air, water, stress). Each organism is seen as an engine within a larger one, all contributing to improve of access for all to the underlying currents. The hierarchy self-organizes not around competition or cooperation, but around the current.
So instead of asking of dogs: “Who’s in control of whom?” (old school) or “Who’s in control of what?” (new school) I think we should be asking, “What is the current around which a dog’s social structure emerges?” And I think if we look at canines without imposing humanlike thoughts on to their behavior we might be able to see such a current emerge.
A linear definition imbued with the notions of control and competition cannot be made to conform to the principles of flow. And I believe that only flow can accurately reflect the true workings of the animal mind. Dominance is based on human thought processes, the comparison of one abstraction relative to another, along with comparison of past, present, and possible future moments in time. But flow—whether a flow of emotion or information—is felt viscerally and unconsciously, and is always capable of being apprehended in the now moment.
Kevin Behan
Natural Dog Training
NDT on Facebook!
I suggest we turn to Dr. Adrian Bejan’s book, Design In Nature, to help us see hierarchy from a new perspective. In it Bejan, argues that nature does not work according to principles of control but principles of flow.
For instance, in Bejan’s view a forest is a hierarchy of a few very large trees relative to many smaller forms of vegetation. The various plants aren’t competing for light, water or nutrients. The tallest ones aren’t trying to dominate the shorter ones. The forest simply emerges as the most efficient way to conduct and improve the flows of all currents contained within it (nutrients, air, water, stress). Each organism is seen as an engine within a larger one, all contributing to improve of access for all to the underlying currents. The hierarchy self-organizes not around competition or cooperation, but around the current.
So instead of asking of dogs: “Who’s in control of whom?” (old school) or “Who’s in control of what?” (new school) I think we should be asking, “What is the current around which a dog’s social structure emerges?” And I think if we look at canines without imposing humanlike thoughts on to their behavior we might be able to see such a current emerge.
A linear definition imbued with the notions of control and competition cannot be made to conform to the principles of flow. And I believe that only flow can accurately reflect the true workings of the animal mind. Dominance is based on human thought processes, the comparison of one abstraction relative to another, along with comparison of past, present, and possible future moments in time. But flow—whether a flow of emotion or information—is felt viscerally and unconsciously, and is always capable of being apprehended in the now moment.
Kevin Behan
Natural Dog Training
NDT on Facebook!
This article, and every other that I've read so far, has been such a treat. I have to admit that I read it quickly, but it reminds me of the way that permaculturists try to look at responding to/establishing environments. It also reminds me of Sapolsky's work with the garbage dump baboon troop in Kenya (makes great reading if you haven't seen it). Is our current hierarchical paradigm characteristically human or just characteristic of the loudest ones? :)
ReplyDeleteI found your blog looking for info about training LGDs for a poultry flock. It makes me excited to have a dog. Thanks.
Thanks so much for your comment.
ReplyDelete"The loudest ones..." I like that!
Cheers,
Lee