Positive Reinforcement vs. Drive Training

Which Training Approach Is Best?

In certain training circles it’s widely believed that learning theory is the only truly scientific and, therefore, the only correct approach for training dogs.

Is this true?

Not exactly.

To be fair, learning theory (also known as behavioral science or behavior analysis) is much more scientific than dominance theory, especially as it’s applied to dog training. There is plenty of scientific evidence showing that dogs and wolves form dominance hierarchies, but none showing that such hierarchies can cross species boundaries.

Still, while dominance training is based mostly on fantasy, not science, positive training is not based on hard science. There are no underlying scientific principles behind how learning theory works in the way that there are in physics and chemistry, where natural phenomena are explained through specific scientific laws. Learning theory is explained only through statistics. Don’t get me wrong. Statistics are important. But they don’t describe the how and why of behavioral changes in humans and animals, only the when and how often.

There are two other problems. And they’re huge.

First, modern, 21^st Century research has shown that behavioral changes don’t take place through associative learning, the theoretical process where a human or animal associates a behavior with a reward and thus wants to repeat that behavior to get another reward. etc. In fact, B. F. Skinner—who created this model of learning—said that the only way to ensure that learning will stick is to provide a series of initial rewards, then start withholding expected rewards using what’s called a variable reinforcement ratio. And unless you have an advanced degree, it can be very difficult to do this with any accuracy.

So that’s the first problem. 

The second is that in actually studying the effects that take place in the brains of animals during the learning process, scientists have found a puzzling paradox: dopamine— often referred to as one of the brain’s reward chemicals—is not released in connection with an external reward, it’s released in the absence of an expected one. Other research shows that dopamine is not really a reward chemical at all. It’s more like a salience detector, meaning that it helps humans and animals pay close attention to changing patterns in their environment. The upshot is that learning doesn’t take place through forming an association between a behavior and a reward but rather through a mostly unconscious process called pattern recognition.

I’ve found that an easy way to determine if this is true is to teach a puppy to sit, but do so, out of sequence, in the following manner: Show the puppy a treat. As he tries to grab it, move it around in such a way that at some point the pup sits on his own. As soon as he sits give him the treat. Then—after he’s already eating his “reward”—wait a moment, then say “Sit!” in a happy voice.

Do this three or four times. Then, show him the treat again and say “Sit!” without waiting for him to exhibit the behavior. Wait a half-a-second or so, and the pup will automatically sit.

Go through this same process again later, in a slightly different environment, or at a different time of day. And once you’ve done it several times, under slightly different circumstances, all you have to do is show the pup a treat, say “Sit!” and the pup will automatically sit.

If learning takes place through associating a behavior with an external reward, the pup wouldn’t learn to sit through this “backwards” way of doing things. Or at the very least, it would take much longer for the pup to learn the behavior and repeat it in a reliable way.

Of course, the sit is among the easiest and most basic behaviors we can teach a dog. And some readers may not see the distinction I’m making between associative learning and pattern recognition. So let’s look at the very complex behaviors exhibited by some working dogs.

Border collies operate almost entirely on pattern recognition. They have to take into account the movement of the flock as a whole, the behaviors of the individual sheep, the changing nature of the terrain, the signals sent to them by the shepherd, and a whole host of other factors. This is one reason why border collies are among the “smartest” dogs. And it’s all due to pattern recognition. There are very few, if any, external rewards provided when these dogs are learning their trade and none while they’re working.

It’s also well known that you can’t train police dogs, military dogs, bomb-sniffing dogs, drug detection dogs, and search-and-rescue dogs using conditioning techniques. That is, you can, you just won’t get a good result if you try to do so.

Why?

It’s well known that detection dogs will give false alerts in order to get a food reward. Meanwhile, police dogs are never trained through rewards but through games like tug-of-war that stimulate and satisfy the dog’s urge to bite, a form of drive training that is quite different from reward-based techniques.

As for search-and-rescue dogs—who often need to be able to navigate the rubble and detritus of an urban disaster site—one training technique used is a scenario where the dogs are urged to climb up or down fire escapes while their trainers throw pots and pans in their direction. Talk about the opposite of positive reinforcement! Yet doing this actually makes the dogs perform better.

There many other examples where the behavioral science model and drive training diverge. 

The point is that when dogs are trained through elements of the wolf’s prey drive—the search, the chase, the grab-bite, and the kill-bite—they learn and operate through a completely different set of principles than those espoused by reward-based trainers. And not only that, but the kind of learning and obedience that takes place with drive training is far superior to what you get with reward-based training. 

Lee Charles Kelley

“Life Is an Adventure—Where Will Your Dog Take You?”

 

Charles Darwin and the Dominance Meme, Part 2

Does Human Observation Create Dominance Hierarchies?

“Although it has been shown that in horses … dominance hierarchies are so poorly developed as to be invisible, needing artificially created competition to develop, … there is a reluctance on the parts of both trainers and some scientists to abandon human attitudes about dominance.”

—Lucy Rees, horse trainer.

The Baby and the Bathwater
Some scientists have tried, with little progress, to dismiss the idea of dominance hierarchies in animals entirely. Others, who still believe the myth, and who toil honestly in the vineyards of animal behavior, are fond of the phrase “Let’s not throw out the baby with the bathwater,” meaning just because there are gaps in logic pertaining to how and why some social animals seem to form dominance hierarchies, the really important stuff remains.

But are there any real contradictions in dominance theories?  

Yes. And as usual, the clearest window into these contradictions comes from David Mech. These two quotes sum things up perfectly.

“Dominance contests are rare, if they exist at all.”  —1999.

“Dominance is one of the most pervasive and important behaviors among wolves in a pack.”  —2010.

So why is there a seemingly conflicting difference between Mech’s two statements? Did conditions on Ellesmere Island change substantially between the late 1980s and 1990s to what they were in 2009? Was there a difference in the technology of the radio collars? Being anesthetized can also cause deep stress, and since hierarchies are more apparent when animals are under stress, was there a difference in the types of anesthetic darts used?

The Stressful Effects of Human Observation

“Just when we think we know it all ...” —Marc Bekoff, 2013.

In an online essay published on June 5, 2013, Dr. Marc Bekoff discusses how the behaviors of animals change in very substantial ways when they’re aware of predators in their environment.

“It’s not an overstatement to say that many animals live in constant fear. Consider the reintroduction of grey wolves into Yellowstone National Park in 1995. While most of the attention focused on these magnificent animals, biologist John Laundré was more interested in the elk who had been living in the park.”

Bekoff recounts the realization that Laundre came to: wolves don't just kill elk, they also change the elk’s daily behavior simply by living in the same general area. This is true in other habitats as well. Whenever predators live in the same environment as their prey—they don’t even have to live in close proximity—it creates a perpetual state of apprehension and stress.

Wolves are apex predators, meaning they’re at the top of the heap: no other animal preys on the wolf, at least no other non-human animal does. But since humans are the only animal that poses a real and serious danger, it would make sense that wolves might behave differently when they feel our presence in their environment, particularly if we’re also shooting them with tranquilizer darts. 

It seems to me that if stress is the chief factor causing the formation of dominance hierarchies, and if being watched increases an animal’s stress, then being observed by scientists may very well be stressful to wolves, increasing incidents of so-called dominant and submissive behaviors. These effects would most likely be multiplied in situations where wolves were shot with tranquilizer darts and then outfitted with radio collars.  

Artificial Selection 
Primatologist Linda Marie Fedigan has criticized the way many researchers set up artificial food-competition scenarios designed to test or, more to the point, create dominance relationships in selected dyads (groups of two).

“It has been found that food tests do not generalize to other conflict situations in any consistent way. Not only does the test-situation only exist in the artificial laboratory-test setting, it has been found that priority to food does not necessarily correlate with priority to other incentives, and that dominance, determined through dyadic tests, does not generalize to dominance relationships for the same individuals within the group as a whole. ... Rather than peeling away the layers of the behavioral onion, to arrive at the core of an underlying ‘real’ dominance rank or dominance relationship, it can be argued that the experimenter has in fact created [it].” (1992, Fedigan, “Dominance and Alliance: Chapter 7 of Primate Paradigms: Sex Roles and Social Bonds, University of Chicago Press.)

Lee Alan Dugatkin, an evolutionary biologist at the University of Louisville in Kentucky, writes: “Winner and loser effects are defined as an increased probability of winning an aggressive interaction.” He goes on to say that “Prior theoretical work on dominance hierarchy formation has demonstrated that … loser effects always produce a clear top-ranked (alpha) individual, but all other ranks in a group remain unclear; whereas winner effects always produce strict linear hierarchies in which the rank of each individual is clear. Paradoxically, however, when individual recognition—a phenomenon long thought to stabilize hierarchies—is possible, winner and loser effects have no impact on the probability of forming strict linear hierarchies.” (2004, “Individual recognition, dominance hierarchies and winner and loser effects,” Proceedings of the Royal Society of London.)

Add to this the fact that when researchers isolate two members of a social group, and study their “dominance” relationships, an odd thing sometimes happens: Animal A is said to be dominant over Animal B, and Animal B is dominant over Animal C, yet Animal C is strangely dominant over Animal A, creating a peculiar dominance loop not found anywhere in Nature. 

Natural vs. Artificially-Created Models 
Creating artificial conflicts in “laboratory settings” is one thing. But does that flaw also translate to observations made of the behaviors of animals in Nature?

Yes and no. 

In the 1960s, biologist Thelma Rowell, the first scientist to study baboons in the wild, found that dominance hierarchies didn’t exist in the animals she studied. According to Rowell captive animals only form dominance hierarchies under two sets of conditions: a) where the animals are total strangers to one another, and b) where they lack ready access to resources available to those living in the wild. (This partially coincides with what Dr. Mech wrote in 1999—that captivity stress causes wolves living in confinement to behave differently than those living in the wild.)

Rowell took this idea even further. In her book The Concept of Social Dominance (1974) she wrote, “The experimenter will report that his trials have demonstrated a dominance relationship between the monkeys while in fact they (the trials) have actually caused it.”

Rowell found that baboon males were extremely peaceful and were not at all competitive. There was much positive or friendly interaction. Aggression was rare. “The dominant impression of interaction between males,” Rowell concluded, “was that of active cooperation.” 

Shirley Strum, who studied baboons in the 1970s, took a stronger stance, claiming that dominance hierarchies were a myth.

The differences both women saw in baboon behavior were seemingly related to one thing: stress. Captive baboons, who were under more stress than those living in their natural habitat, formed dominance hierarchies. Wild baboons didn’t.

Like Rowell, Jane Goodall also began her studies in the 1960s, and initially saw no signs of dominance in the chimpanzees at Gombe Stream. That changed after about 10 years when she first saw “dominant” females killing the young of other females of the troop, in some cases eating their young. “During the first ten years of the study I had believed … that the Gombe chimpanzees were, for the most part, rather nicer than human beings. … Then suddenly we found that chimpanzees could be brutal—that they, like us, had a darker side...”

Was this change due to changes in the chimps’ external environment, or due to changes in the way  Goodall and others were observing them?

In their 1995 book When Elephants Weep Masson and McCarthy wrote, “In recent years the idea of the dominance hierarchy has become more controversial, with some ethologists now asking if such hierarchies are real or a product of human expectation.”

I’ll go beyond that and—siding with Thelma Rowell—say that dominance is not only a product of human expectation, it’s a product of human observation.  

LCK 
Life Is an Adventure—Where Will Your Dog Take You?

Charles Darwin and the Dominance Meme, Part 1

Does the Concept of Dominance Hierarchies Run Counter to Darwin's Theories?

 Two wolves playing in the snow.

“I aimed for a modest presentation. I would demonstrate simply and directly that male Pumphouse baboons did not have the traditional hierarchy, while females did. … At the end of my presentation, no one spoke. The polite silence was finally broken with barely guarded accusations. I had invented my data. I didn’t have enough information to draw the conclusions I had come to and that there had to be a male dominance hierarchy … I had missed it, that was all.” 

—Shirley C. Strum, Almost Human 

Animal Hierarchies Didn’t Exist Before the 1920s.
The idea that animals form dominance hierarchies is so deeply ingrained into the minds of most scientists today that to say or even hint that things may be otherwise (as Thelma Rowell and Shirley Strum have done) has become something like an act of heresy or sacrilege.¹ Animal hierarchies are, in neuroendocrinologist Robert Sapolsky’s words, “textbook social systems, sort of engraved in stone.”

I’ve written a number of posts—both here and at PsychologyToday.com—questioning the validity of dominance hierarchies in dogs and wolves. And I’ve gotten into some hot water for doing so.² In this post I’ll present new arguments showing:

1. that the idea of social hierarchies goes counter to Darwin's view of natural selection, 
2. that there is no evolutionary arc that runs from hierarchical systems in lower animals to those in humans, and
3. that acting “dominant” may actually reduce an animal’s adaptive fitness.

I realize I’m on a fool’s errand. And I’m more than happy to be taken to task and proved wrong on any of the points I’m going to make here. It just seems to me that dominance hierarchies simply don’t exist in Nature. And it also seems to me that it all starts with a very simple misunderstanding of Darwin’s theory of natural selection.

In Paul Ekman’s 1998 edition of Darwin’s The Expression of the Emotions and Man and Animals, evolutionary psychologist Daniel G. Freedman seems to have criticized Darwin for being unaware of animal hierarchies: “Darwin is aware of submissiveness,” Freedman wrote, “but the naturalistic notion of, say, wolves forming an hierarchical pack is missing. Social hierarchies is a major concept of animal observation today, and many of Darwin’s examples of antithesis would be seen now in terms of hierarchy.” 

True. But is that because Darwin missed the boat or does the idea of dominance hierarchies run counter to Darwin’s thoughts on the nature of social animals?

I don’t think Darwin was wrong. I think it’s more likely that the reason he didn’t mention dominance hierarchies is that they didn’t exist during his lifetime. There were no animal hierarchies for him (or anyone else) to observe because, in all probability, they simply didn’t exist until the 1920s when Norwegian biologist Thorleif Schjelderup- Ebbe published his dissertation on pecking orders in chickens.

Scientists began looking for “pecking orders” in all social animals. They sometimes found what they were looking for—though the truth is, sometimes they didn’t. 

Still the concept of pecking orders—which eventually morphed into what we now call dominance hierarchies—caught on like wildfire, or like a meme, an ideological virus that infects the human mind and prevents us from seeing the truth. This meme is so powerful³ that when dedicated scientists like Shirley Strum or Thelma Rowell present data that run counter to this idea, their evidence is ignored, their methods called into question, and the concept of a “latent hierarchy” is invented to account for the lack of hierarchical structure. 

Dominant Species vs. Dominant Behaviors 
I know the idea that social animals form dominance hierarchies seems like pure Darwinism to most. Animals in competition over resources! Yes! But let’s take a look at what Darwin’s theory is really about. 

“The theory of natural selection is grounded on the belief that each new variety, and ultimately each new species, is produced and maintained by having some advantage over those with which it comes into competition.” (Darwin, On the Origin of Species, 371.) 

Here the nature of competition is reserved for different species, not members of the same species, and especially not for members of the same social group. In fact Darwin believed that social animals may be more adaptable because of their ability to work together: “Social animals perform many little services for each other: horses nibble, and cows lick each other, on any spot which itches: monkeys search each other for external parasites. … Animals also render more important services … thus wolves … hunt in packs, and aid each other in attacking their victims.” (The Descent of Man, 71, 72) 

Does it really make sense that members of a social group would be in competition with each other over resources? It seems to me that sociability is about pooling resources, not fighting over them. Finding, isolating, and quantifying these sorts of resource sharing behaviors—now often referred to as “biological altruism”—has become all the rage recently. It’s been shown that even plants share resources with their closest kin. And one of the reasons scientists are so interested in biological altruism is that it supposedly runs counter to Darwin’s concepts of species being in competition with one another and gaining an advantage over them. 

Perhaps the clearest window into how the dominance meme fails to make sense is the wolf pack—an aggregation of animals whose social structure is built almost entirely around the need to hunt large, dangerous prey by working together as a cohesive social unit. If the prey animal is the pack’s most important resource, and hierarchy formation is about competition over resources, then we should see intense posturing and jockeying for position both during the hunt, and when the pack feasts on its fallen prey. Yet pack members work together, not against one another—neither dominant nor submissive behaviors are ever seen during the hunt. And once the hunt is over, all members have mutual access to the carcass of the fallen prey animal, with no hierarchy and very little, if any, dominance visible. 

Plus—and this may be even more important—it’s hard to see how dominance (threats of aggression) would foster group harmony and cooperation. It seems more likely to me that affiliative behaviors—licking each other’s fur, cuddling in the cold, playing with one another, etc.—are the real glue that holds a wolf pack together. 

Leveling Mechanisms in Non-Heirarchical Human Societies 
Another meme is based on what I see as a common misinterpretation of Darwin’s statement that “The difference in mind between man and the higher animals, great as it is, is certainly one of degree and not of kind.” Almost everyone who quotes this trope ignores the fact that a few sentences later Darwin admitted that he could be wrong: “If it be maintained that certain powers, such as self-consciousness, abstraction, etc. are peculiar to man, it may well be … the result of the continued use of a highly-developed language.”

Still, scientists look at the arc of evolution (and thus the arc of hierarchical systems) as reflecting this shaky theoretical difference of “degree and not of kind.” This may be one reason we can’t help but see dominance hierarchies in apes, wolves, crayfish, and guppies, etc. ³

But is there really an evolutionary arc that runs from lower animals to human beings?

Primatologist Shirley Strum writes, “Many of the models of human evolution have assumed that the human experiment began with limited social resources, instinctive and compulsory aggression, male domination and rigid hierarchy. But these models seem faulty if we now know that ‘lowly’ baboons are more complex and have more diverse options. Were the earliest humans not as smart ... as baboons?”

And we don’t even have to look at baboons, we can look at some human groups, small bands of indigenous hunter/gatherer societies who not only don’t form dominance hierarchies, they’ve developed leveling mechanisms to prevent them from forming. In these groups if one member tries to act dominant, he's quickly shunned. And one of the primary reasons these groups have these mechanisms in place is because hierarchical systems lessen the group’s ability to hunt successfully just as it would in a wolf pack.

Another leveling mechanism in these egalitarian societies relates back to wolf behavior as well, and  that’s play, an activity that wolves—and especially dogs—engage in on a regular basis. (Now there’s a set of behaviors that actually do have an evolutionary arc…).  

Acting “Dominant” Decreases an Animal’s Adaptive Fitness 
It’s said that the dominant member of the group is the one most likely to pass on his genes to the next generation, and that’s the fundamental purpose of hierarchies: to provide the most robust animal a non-negotiable platform for reproduction. But if the true purpose of a wolf’s social instincts is to enable the pack to work together to hunt large, dangerous prey, how do internecine battles over bones and sleeping places relate to their overall adaptive fitness? In wild packs it’s normally rare for any but the breeding male and female to pass on their genetic material to future generations. Would one night’s sleep on a less-than-perfect “bed” or taking a bone away from another wolf really tip the scale toward genetic oblivion, and that’s why wolves supposedly have to exert their “dominance” over such things?
Is it even true that the most dominant male in a wolf pack—or any social animal group—is automatically more able to pass on his genes to the next generation?

Apparently not. In his studies of baboons Robert Sapolsky found that dominant behaviors actually have a negative impact on survival.⁴

At one point, a troop Sapolsky had been studying for years, and who exhibited the classic male hierarchical structure, came across a human garbage site. Yay! Free food! But the food was unfortunately tainted with tuberculosis. The troop was decimated.

Yet interestingly, it was the most “dominant” baboons who lost their lives, not the other way around. The reason? Dominance isn’t a normal or natural behavior. It’s always triggered by stress. And high levels of the stress hormone cortisol tend to suppress an animal’s immune system. (Excessive levels of testosterone don’t help matters any, either.) And that's why the most dominant baboons in the troop died.

“It wasn’t random,” says Sapolsky. “If you were aggressive, and if you were not particularly socially connected, socially affiliative, if you didn’t spend your time grooming and hanging out—if you were that kind of male—you died.” 

A generation later, Sapolsky came back to find that the troop had been transformed. They were much more amenable, social, and affiliative now. There was no longer a clear hierarchical structure (as Rowell and Strum had seen in their studies of baboons). And if you were an “alpha type,” trying to dominate others, you were quickly shunned!

Sapolsky says, “One of the things that baboons teach us is that if they’re able to, in one generation, transform what are supposed to be textbook social systems, sort of engraved in stone, we don’t have an excuse when we say there’s a certain inevitability about human social systems.”

Another thing that the baboons teach us is that hierarchy formation in animals does not necessarily serve an adaptive purpose. In fact, just the opposite may be true.  And, in the end, these social structures only exist in our own minds because that’s how we see the world.

Some would argue that dominant and submissive behaviors do, in fact, exist. They’ve been observed. Data has been collected and analyzed. And while there may be gaps in logic here and there, it’s simply undeniable that these behaviors exist.

I agree. They do exist, but they’re brought on by the simple act of being observed by human beings. In fact—and I’ll develop this idea further in my next post—these behaviors are more likely to be produced when animals are being observed by male rather than female scientists. 

LCK 
“Life Is an Adventure—Where Will Your Dog Take You?”

Footnotes: 

1) “I was naïve. I had imagined that one did the research, gathered the information, analyzed, interpreted and presented it to the scientific world. Then the work would be evaluated and incorporated, if accepted, into the basic knowledge within the field. But there are cliques in science as in any other facet of human endeavor. If you are part of the ‘in’ group, even minor findings are discussed and integrated, eventually becoming part of the working knowledge of the field. If you are not part of the clique, you stand a good chance of being ignored.”                                                             —Shirley C. Strum, Almost Human

2) Every time I posted a mea culpa at PsychologyToday, it was because other authors at the site complained when I wrote about dominance hierarchies, proving that dominance hierarchies do exist, just in scientific circles not animal groups. (I’m not saying this to compare myself to Shirley Strum, Thelma Rowell, or others who’ve fought the orthodoxy, but to point out how strongly those in the scientific community feel about the subject.)

3) “Our findings show for the first time that individual differences in the preference for social dominance hierarchy predict neural response within left AI [anterior insula] and ACCs [anterior cingulate cortices].” (“Neural Basis of Preference for Human Social Hierarchy versus Egalitarianism,” Joan Y. Chiao.  

4) Please watch this video on "why hierarchy creates a destructive force within the human psyche," to see and hear Sapolsky describe, in his own words, how the baboon troop changed from a pro-dominance to a pro-affiliative society. 

Evolution Will Punish You If You're Selfish and Mean

Are Animals Really in Competition With One Another?

 

“[Darwin] pointed out how, in numberless animal societies, the struggle between separate individuals for the means of existence disappears, how struggle is replaced by cooperation... He intimated that in such cases the fittest are not the physically strongest, nor the cunningest, but those who learn to cooperate so as to mutually support each other.”

—Prince Peter Kropotkin, Mutual Aid: A Factor of Evolution

Huxley vs. Wallace, Competition vs Cooperation 
It seems to me that evolutionary science took a huge wrong turn the day it was discovered that some chickens form pecking orders. Suddenly, the Darwinian idea that different species might compete with others over resources within a certain ecological niche was misinterpreted to mean that members of social animal groups were engaged in a constant power struggle for dominance over their fellow group members.

In his book Cooperation Among Mammals (1997), evolutionary biologist Lee Alan Dugatkin details the conflict that arose between early Darwinists Thomas Huxley—a hard-liner when it came to the idea of competition—and Henry Wallace—co-creator with Darwin—of the theory of natural selection. 

Dugatkin writes, “While Darwin (1859) acknowledged that the struggle for existence is often metaphorical, insofar as it is often a struggle against the environment, Huxley … took a more extreme view.”

He goes on to say that Huxley believed the animal world was on the same level of competition found in ancient gladiators. ‘Life is a continuous free fight,’ Huxley wrote, and went on to say that ‘the state war of each [animal] against the other was a normal state of existence.’ However, Wallace, in his book Darwinism (1891), argued the exact opposite, stating that ‘On the whole, the popular idea of the struggle for existence entailing misery and pain on the animal world is the very reverse of the truth.’” 

The truth—as Darwin stated it—is that social animals “provide many little services for one another.” And it’s not just about prey species like elk finding safety in numbers, or group predators like wolves finding it easier to hunt large prey as a group. Dugatkin points out that goldfish tend to live longer when their group size increases. They also tend to grow faster when surrounded by more rather than less goldfish. If they were in constant competition with one another, that wouldn’t happen. Dugatkin also points out that social amphibians are able to regenerate lost tails more quickly and that social animals learn new tasks faster than solitary species do. 

All of this points to the likelihood that sociability and cooperation have a positive effect on survivability, while engaging in internecine battles or threats of aggression would have the opposite effect. 

Game Theory and Evolutionary Strategies 
In 1973, John Maynard Smith proposed the idea that game theory would be very useful in defining a framework of animal contests and strategies into which Darwinian competition could be modeled. His position was that “players” of these evolutionary “games” don’t necessarily act in a rational manner, but have strategies that, when successful, produce higher levels of fitness in some organisms than their competitors. In this model, players do not choose their strategy or have the ability to change it, they are born with it and their offspring will inherit it.

However, no matter how well-developed and successful evolutionary game theory has become, it still explains natural selection through the lens of human thought processes (i.e., strategies) and through the concept of competition between species, as well as competition between members of the same social group.

In their 1998 paper, “Animal Contests as Evolutionary Games,” Mike Mesterson-Gibbons and Eldridge Adams write: “It is not unusual for an exercise in game theory to remain partially inconclusive. On the one hand, game-theoretic models are valuable because they suggest ways to test new ideas. On the other, suggesting a test is not the same thing as conducting it, and the difficulties of doing so should not be underestimated.”

Finding new ways to conduct tests is what science is all about. Yet how useful is the information provided by these tests if the underlying premise isn’t valid? And if these tests are designed with “winners and losers” in mind, aren’t evolutionary game theorists loading the dice in favor of a pre-determined thesis, and thus actually creating contests between animals that might not otherwise exist? 

Cheaters Never Prosper (Not for Long)
One of the ways that evolutionary game players are said to gain an advantage over others is through cheating or free-riding. Both “strategies” are a means of getting more while doing less. But now, a new review of dozens of key ecological studies has found very little evidence to support the belief that “cheating” is widespread.

Rice University evolutionary ecologist Emily Jones, the study’s co-lead author: “We find that although there are numerous observations of low-quality partners, there is currently very little support that any of these meet our criteria to be considered cheaters.” She goes on to say, “A behavior is only cheating if it provides one partner with an advantage and also imposes a disadvantage on the other.”

In a similar vein, University of Pennsylvania researchers Alexander J. Stewart and associate professor Joshua B. Plotkin recently offered mathematical proof that the only evolutionary strategies in social animals that can succeed in the long term are generous ones.

“Ever since Darwin,” Plotkin writes, “biologists have been puzzled about why there is so much apparent cooperation, and even flat-out generosity and altruism in nature. Our paper provides such an explanation.”

After simulating how some generous strategies would fare in an evolving population, Stewart and Plotkin crafted a mathematical proof showing that, not only can generous strategies succeed they’re the only approaches that work over the long term. 

“Our paper shows that no selfish strategies will succeed in evolution,” said Plotkin. “The only strategies that are evolutionarily robust are generous ones.” [1]

In another recent paper Christoph Adami of Michigan State writes, “For a short time and against a specific set of opponents, some selfish organisms may come out ahead. But selfishness isn’t evolutionarily sustainable.” [2]

Adami and his colleagues read a 2012 paper unveiling a newly discovered zero-determinant (or ZD) strategy, which supposedly gave selfish players a guaranteed advantage over others. But they had serious doubts whether this strategy would essentially eliminate cooperation and create a world full of selfish beings. So they used high-powered computing to run hundreds of thousands of games and found that ZD strategies can never be the product of evolution.

Says Adami, “We found that evolution will punish you if you’re selfish and mean.” 

Lack of Cooperation in Wolf Packs? 
Wolves are often cited as a prime example of cooperation in animals. Yet research shows that hunting success peaks at about ± 4 wolves. The larger a pack gets, the less successful they are. Why is this so?

In a 2011 article, McNulty, Smith, Mech, et al say that there are 2 prevailing hypotheses. The interference hypothesis proposes that hunting success is limited because individual predators impede each other’s actions. The other is the free-rider hypothesis, where some pack members consume more than their fair share of a resource, or shoulder less than a fair share of the labor costs.

These researchers measured levels of participation by pack members during various stages of the hunt, using an ethogram (an objective scientific inventory of a set of behaviors) developed a few years earlier by McNulty, Mech & Smith (2007).

The idea that hunting success was negatively impacted when some members withheld effort was generally borne out by the fact that the rate of decline was most apparent for the most dangerous task: capture, or biting and restraining prey. In other words, as pack size increased fewer wolves felt like going in for the kill.

“Our study suggests that [some] wolves in large groups (>4 hunters) withheld effort … and likely participated merely to be at hand when a kill was made.”

I think this is very unlikely. It would require a mental process called time travel, where an animal is able to hypothesize about possible future events, then form a devious mental plan on how to act in such a way so as to “fool” its pack members and thus derive a purely imaginary future benefit that may or may not materialize.

Since a wolf pack is a cohesive unit where members don’t tend to wander off much to engage in their own activities, whenever the pack goes hunting, all members (except for the very young), go along. And since pack hunting success levels off at anything greater than ± 4 wolves, those members of the pack who are the most experienced, the most fearless, and the most driven would likely shoulder most of the work. And since according to the latest research, a behavior can only be considered cheating (or “free-riding,” a form of cheating) if it provides one member of the pack with an advantage while causing a disadvantage to others—which doesn’t happen here—these non- participating wolves should not be considered free-riders, but should be thought of as something like the pack’s bullpen. They’re happy to let the starting pitcher play the whole game if he can, but they’re also ready to come in during the late innings if needed.

They’re a team, after all.

Lee Charles Kelley 
“Life Is an Adventure—Where Will Your Dog Take You?” 


Footnotes:

1. (“From extortion to generosity, evolution in the Iterated Prisoner’s Dilemma,” A. Stewart, J. Plotkin, University of Pennsylvania, July 25, 2013.)

 2. (“Evolutionary instability of zero-determinant strategies demonstrates that winning is not everything,” Nature Communications, August 2013.) Adami et al.

Dogs On Motorcycles, Dolphins Riding Whales, and the Myth of Deliberate Intent in Jellyfish

Is Animal Behavior a Product of Deliberation + Intent or Attraction + Momentum?

Actor Harry Dean Morgan and His Motorcyle-Riding Dog

Do Jellyfish Think About What They’re Doing?  
The Nature World News website recently touted the fact that a new laboratory study suggests that a species of jellyfish has been seen “deliberately” catching fish, this despite the fact that jellyfish don’t have brains or a central nervous system.

According to lead researcher, Robert Courtney, “They’re using their tentacles and nematocyst clusters like experienced fishers use their lines and lures.” He goes on to say that “the nematocyst clusters look like a series of bright pearls, which the jellyfish twitches to attract the attention of its prey, like a series of fishing lures. It’s a very deliberate and selective form of prey capture.”

Interesting, huh? According to Dr. Courtney, the jellyfish deliberately twitches these pearl-like nematocysts not just to attract its prey, but to attract the attention of its prey. 

Okay … but how does the jellyfish know that its prey has mental states? It seems to me that in order for a jelly to produce these behaviors for the reasons given it would have to have a fully-developed theory of mind, meaning it would have to have a mind of its own and would also have to be capable of knowing that other organisms have minds similar to its mind in some ways, yet dissimilar in others. Yet Dr. Courtney seems convinced that they’re fishing deliberately.

But here’s a significant fact: we’re also told that the study was done in a laboratory instead of the jellyfish’s natural habitat. That’s because this is a very tiny organism that’s almost invisible in the open sea. So, in order to capture this species for laboratory study, the researchers trapped the jellyfish by submerging high-powered lights in its natural aquatic habitat. 

Why? Because they’re attracted to light.

So here we have two variations on the word “attraction.”

1) The jellyfish is attracted to light, and 

2) it’s reportedly behaving to attract the attention of its prey.

So why does Dr. Courtney think he’s seeing deliberate behavior in an organism that can’t possibly think about what it’s doing? Why aren’t its behaviors described through—I don’t know—a basic function of physics found in all things, living or non-living, all things on earth, organic and inorganic, a simple process found both in organisms that can think (meaning us and maybe dolphins) and those that can’t (meaning other organisms like jellyfish, plants, bacteria, and inorganic things like tectonic plates and maybe even the ocean itself)? 

The truth is simple: the jellyfish is attracted to its prey. If it weren’t it wouldn’t be able to feed itself except by sheer random chance. Fortunately for the jellyfish, it has a built-in fishing lure: its bright nematocyst clusters. 

And why do lures work? Because fish are attracted to them.

Could it be that the jellyfish is not acting in a deliberate manner — which, again, would be impossible without a highly-developed brain — but is simply moving toward its prey in what seems to be a deliberate manner because of physical attraction? (“Physical” meaning its behavior is controlled by physics, not sexual desire.)  

Dogs on Motorcycles, Dolphins Riding Whales 
Meanwhile, in another part of the world, we have a species called the dog. And dogs are known for their infatuation with car rides (or at least most dogs are). Of course, there’s nothing out of the ordinary about that. I mean it’s interesting that dogs love going for car rides, and that most cats hate it. But until recently I’d never heard of a dog who loved motorcycle rides. But in a recent television interview on Jimmy Kimmel Live, actor Jeffrey Dean Morgan described how one of his dogs absolutely loves to do just that. 

As soon as his dog hears him kick-starting the engine, she comes bursting out of the house, runs straight to the motorcycle, and jumps up between the handlebars onto the gas tank. Then, cradled safely (or perhaps precariously) between her owner’s arms, the duo takes off down the road.

This is all pretty amazing (and some would say dumb on Morgan’s part), but here’s what I think is really amazing: once this dynamic duo reach the freeway, and begin traveling 70 miles an hour, the dog falls asleep and doesn’t wake up until they reach their destination!

Can you imagine? The dog isn’t safe inside a car, with the window rolled down so she can stick her head out. She’s sleeping on top of the gas tank. At 70 mph! Fast asleep!!

And she’s not the only dog who likes to ride a motorcycle! 

I think dogs like to ride motorcycles for the same reason they like to go for car rides, chase tennis balls and fly through the air after Frisbees. They like the feelings of momentum and the pleasure of velocity. And on a motorcycle those feelings are intensified. Dogs also like the feeling of togetherness which are also intensified on a motorcycle. 

Dogs and Dolphins and Dolphins and Dogs 
Another species that seems to thoroughly enjoy feelings of physical momentum, as well as synchronous movement with others, is the bottle-nose dolphin. We’ve all seen videos or still images of dolphins arcing together in unison, either in the open water or while doing tricks at a water park.

In fact, there have been several incidents where dolphins have been seen, off the coast of Hawaii, riding on the backs of humpback whales which is quite similar to the way Jeffrey Dean Morgan describes his dog riding his motorcycle’s gas tank!

Ken Ramirez, a dolphin trainer at Chicago’s Shedd Aquarium says, “It is believed that the ‘surfing’ or bow riding [behaviors] that dolphins exhibit in front of boats may have had its genesis in riding in front, or in the wake, of big whales.”

Like many +R trainers, Ramirez believe that the animals they train learn to do all sorts of amazing acrobatic feats through a process called positive reinforcement, and more specifically through sonic “markers,” usually clicks from a clicker that signal a food reward may be coming later. In fact, these trainers believe that all animals—not just dolphins—learn the same way, through a process called operant conditioning.  

Do Dogs and Dolphins (and Jellyfish) Learn the Same Way?
 All animals—at least those who can learn new behaviors (and jellyfish probably don’t fit that category)—learn the same way. But it’s not through operant conditioning, which only creates a fairly convincing facsimile of how animals really learn, and usually only under very tightly-controlled conditions. In fact it’s well known that behaviors learned via operant conditioning tend to break down whenever strong drives and instincts are involved. [1]

So how do animals really learn?

According to a model called Natural Dog Training—created by dog trainer and natural philosopher Kevin Behan—all behavior and learning are the products of certain properties of physics, among them the same properties that explain how a jellyfish seems to hunt deliberately: through physical attraction to its prey. 

Some obvious examples include the kind of almost magnetic attraction most dogs have for other dogs, for their owners, their toys, treats, food dish, etc. When a dog has strong feelings of attraction for something he tends to move toward it in a straight line. But, according Behan’s model, dogs also have feelings of resistance toward other types of things, like going to the vet’s office, or getting a bath. Resistance is literally the polar opposite of attraction. When a dog has strong feelings of resistance he tries to move away from the things that generate those feelings. Meanwhile, when a dog feels a mixture of both attraction and resistance he’ll move in a curvilinear fashion. 

Add to this process the idea that when a dog moves toward an object of attraction, doing so creates pleasurable feelings of physical momentum and emotional flow, the same types of feelings we have when we’re humming along on the open highway, or playing golf or tennis, or even just watching a game of football on TV, where we may unconsciously project our emotional centers of gravity onto the wide receiver reaching up to catch a football. We often stretch and strain ourselves while watching the ball’s arc and the receiver’s arms stretching and straining to reach it.

Now think of a Frisbee dog flying through the air to catch a flying disc or a ball-happy dog running full bore after a tennis ball or Kong. That’s pure attraction, which, again, also creates pleasurable feelings of physical momentum and a pure flow of emotion.

So when a puppy learns to sit for a treat, operant conditioning—or a reasonable facsimile thereof—is enough to explain that process for most purposes. The missing piece is that the puppy has to first have a feeling of attraction for the treat and for the person holding it. You can’t get a puppy to sit for a treat unless he’s focused on you and has feelings of physical and emotional attraction for you and the treat.

Attraction + Momentum = Flow  
However, things get a bit more complicated when you’re training a dog to herd sheep or run an agility course or run full speed off a diving platform or—as is the case with dolphins—when Naval personnel train them to locate enemy mines underwater.(2)

For the agility course the dog has to have feelings of attraction for moving toward and through the various challenges in the course. The diving dog is motivated by the way his owner builds his attraction to a prey object, like a tug toy: the more attraction the dog has for the object the stronger his desire to intercept it in mid-air, and the better his score in the diving competition. When dolphins are trained to swim into a harbor, then dive deep and search for possible locations of enemy mines, they need to have more motivation than an eventual “reward” from a chum bucket back at the “base.” That reward is the feeling of seeking and finding “prey,” just like it is for dogs.

Resistance enters the picture through the fact that in all three cases the animals are also pushing past feelings of physical resistance, which, again, is the polar opposite of attraction.

Of course dogs and dolphins are predators. Would prey animals have the same feelings?

Yes. It’s just that their “prey” would usually be whatever it is they like to eat. Horses, for instance, don’t chase their “prey,” but they are attracted to things like grass and hay, etc. Yet even prey animals like horses will play games of chase with each other that often involve biting or mock biting the other animal, as if they were predators at heart.

There’s a lot more to the Natural Dog Training model of learning than attraction and resistance or the pleasures of physical momentum and emotional flow. But on the most fundamental level, there is really no difference between the way a jellyfish is attracted to its prey, the way bacteria are attracted to specific substances that sustain life [3], the way fundamental particles are attracted to one another, and the way dogs are able to learn complex new behaviors like riding on the gas tanks of their owners’ motorcycles or the way dolphins enjoy body- surfing on the backs of humpback whales.  

Lee Charles Kelley  
“Life Is an Adventure—Where Will Your Dog Take You?”
 
Footnotes:
 
1.) "The Misbehavior of Organisms," Keller and Marian Breland, American Psychologist, 16, 681-684.  

2.) http://www.public.navy.mil/spawar/Pacific/71500/Pages/default.aspx   

3.) Note that in his book Adaptive Behavior and Learning Dr. John Staddon discusses research done on e-coli and salmonella bacteria, organisms that show only two modes of movement—essentially either moving toward something or moving away from it—which are defined as straight-line swimming (moving toward) or tumbling (moving away). The first is found when the bacteria move toward a spot where there’s a high concentration of an “attractant,” and tumbling is found when there’s a decrease in its concentration.