Tuesday, June 24, 2008

Positive Mental Associations v. Reduction of Tension

If you've followed the comments section on "Chasing Squirrels," you'll see that someone calling him or herself "anonymous" has a bad taste in his or her mouth about the idea that dogs don't learn new behaviors or get rid of old ones by making "positive mental associations." Hopefully this article will spell out my reasons for the view that dogs actually do learn through the reduction of internal tension and stress, not the outdated, Skinnerian view (though it's not really Skinnerian, as you shall see).

Do Dogs Learn by Making Mental Associations,
or Through the Reduction of Internal Tension?
I think it's very important to make the distinction between two ways of looking at the general “Chasing Squirrels” phenomenon. Does this training paradigm work by a) satisfying a dogs prey drive, thereby reducing his internal tension and stress, or b) by giving him a positive mental association with something that previously seemed to have caused his misbehavior. The reason I think it’s important to differentiate between these two ideas is that the concept of teaching the dog to make positive associations with formerly unsettling stimuli—while fine in theory—doesn't really work in every situation. And it’s junk science.

For example, many positive trainers make a huge mistake in my view by thinking that if a dog is aggressive on the leash, for example, they can change the behavior simply by feeding the dog “high value treats” while another dog is passing by. This is supposed to work by teaching the aggressor to make a positive association with the aggressee:

“Hmm…every time I feel aggressive toward another dog on the street my owner gives me food that is of high value to me.”

Do you see the multiple errors in logic there?

Part of the problem is that sometimes redirecting the dog’s focus away from something distressing—like another dog coming toward him to something enjoyable like eating tasty treats—actually does to work. The fact is +R trainers are not total idiots. In fact many of them are anything but. And yet if “making positive associations” only works sometimes, and not all the time, why does it work at all? Shouldn’t creating a positive association with something that was previously distressing to the dog always work? I think so. And I think we need to ask ourselves why does it work sometimes and not others.

So let’s look at the situation with a dog whose leash aggression is actually (or rather, seemingly) cured by using high value treats to change the dog’s “mental associations” toward other dogs. To understand this we first have to figure out what’s actually causing the dog’s leash aggression.

One of the primary causes of leash aggression (or at least the primary flashpoint) is that the aggressee is usually coming toward the aggressor. It’s rare for a dog to exhibit leash aggression toward a dog who’s walking in front of him. It’s also rare for a dog to show aggression toward a dog who’s walking next to him. He’ll certainly do so initially. But once you walk the two dogs together for a few blocks, the aggressive feelings almost always dissipate and don’t rear up again until the dogs are face to face once more. (See “Walking 2 Aggressive Dogs.”) 

Another rarity is the big dog who gets aggressive only with small dogs. In the most general terms, it’s usually the other way around: the smaller dog is generally more nervous when a bigger dog is approaching. And even big dogs who do seem to have a kind of ingrained aggression to any and all dogsregardless of breed type or sizeusually display less tension toward dogs whose level of gaze is below theirs and more tension toward dogs whose eyes are either on the same level or higher.

So it seems to me that eye contact is what initially sparks the stressful feelings in a dog who then makes those feelings manifest by barking, snarling, lunging, etc. Eye contact isn’t the whole enchilada, but it certainly seems to be the meat. Of course there are normally all kinds of other contributing factors, typically involving past negative experiences with other dogs, or generalized fear issues, mistreatment by his previous (or present) owners, etc. But while those factors may always be present in the dog in a sort of latent form, it’s really the direct eye contact that provides the spark that sets the dog off.

So if a +R trainer distracts a leash-aggressive dog with a treat, in 99.99% of the cases, the aggressor will have to break eye contact with the aggressee, just to get the treat In and of itself, that automatically reduces some of the tension inherent to the situation. If the +R trainer does this enough times then that might create a situation where when the aggressor sees another dog coming he may develop a positive expectation of something good coming his way—not from the other dog but from his handler. If there’s a lot of praise, or a soothing word or two in the equation (like “Easy, big fella…” or “It’s okay, Sparky…), that will also be part of what helps reduce his tension.

But this is still operating through a reduction of internal tension, not through a process of making mental associations.

And here’s where it gets really good: If the aggressive dog’s tension is reduced enough, and he’s able to actually make friendly contact with the second dog, that’s what really changes things. And that’s because no dog ever wants to attack another dog. It’s simply not in a dog’s nature to want to be in conflict, socially. On the most basic level all dogs want to be in harmony with other dogs. (Even forming a detente with another dog is more harmonic than being in conflict with him.) The lunging and snarling and barking is actually an expression of the first dog’s attraction to the aggressee. If he had no attraction he would either run away and hide, or he would ignore the dog entirely. The more a dog barks, snarls, and lunges, the more attracted he is. The problem is he’s simply attracted at too high a level given his inability to offload tension normally, through sniffing, circling, play bowing, etc.

So by making a surface reading of what’s happened, we might be fooled into thinking that the dog is making a positive association with the movement of another dog coming toward him because of the treats alone. Or we might more rightly see that there was more to it than the treats, but still mistakenly believe that the dog is now making a positive association with the aggressee as a result of the redirection of his focus and attention. And I think in a very marginal way there may be some truth to that second way of looking at this. But the real reason this technique works is that the handler has accidentally reduced the dog’s inner tension and stress. That’s the deciding factor, not this fanciful idea that the dog is now making a positive mental association with other dogs coming toward him.

Here
’s what Kevin Behan says on the subject in Natural Dog Training:

While dogs form associations, they don
’t directly learn by forming them. First they learn through the flow of drive and then they associate the level of pleasure or stress they experienced with them.

I
’m going to correct Kevin here a little and say that dogs have no ability to know what levels of pleasure or stress they experience with any given experience. The way a dog’s mind works, through energetic principles of attraction and resistance, they feel more attracted to experiences that reduce internal tension and more resistance to those that don’t.

Now back to Kevin
’s words (with a few thoughts of my own thrown in):

In the human mind a [mental] association is the linkage between thoughts, ideas, feelings, or sensations. These linkages from concepts that the human mind can then carry around and apply in novel and abstract ways and in new situations. A concept is a complex intellectual feat based on events that transpired over time [dogs have no sense of time-LCK], and it acknowledges that there may be other perspectives on the experience [dogs have no ability to see anyone else
’s perspective, at least not mentally, though they have the ability to feel what others are feeling, and even to pick up mental images from us-LCK]. In order to articulate something this complex Man requires an intricate language to communicate the vast range of possibilities.

A dog, not having an intellect, can’t range through time, nor can he entertain any other possible perspectives on an event [at least not mentally, though he can do this on a purely emotional level-LCK]. He can only know his own point of view [I would say that he can’t know it, he can only feel it-LCK] in any one particular moment.

Now B.F. Skinner, who, like it or not, is the font for this idea that dogs can form positive mental associations, would clearly not agree that dogs have such intellectual capacities. Skinner would be on Kevin Behan’s side of the above argument. In fact one of Skinner’s main rationales behind the development of behavioral science was to remove the mind/intellect/black box from the equation, which to him was voodoo. He wanted to be able to rely purely on conditioning. Yet this idea that dogs can from positive mental associations depends entirely on dogs having certain intellectual capacities that they clearly don’t have!

This is a bit of a puzzle to those who don
’t study cognitive science and don’t understand the differences between certain evolutionary pre-cursors to logical thought and language, because those pre-cursors operate in both the human and canine brain. For example pattern recognition, which requires no conscious, intellectual thought process, and which dogs excel at, is clearly a pre-cursor to logical thinking. You can’t engage in logic unless you can recognize patterns. In fact chess masters are said to rely far more on pattern recognition than logic. And computers are getting better and better at pattern recognition on an almost daily basis. But computers do not have any real intellectual or logical abilities. And I hate to say this but the same is true for dogs.

(Discussing these issues always upsets a lot of people, so let
’s get back to the discussion at hand, which is that solving a dog’s leash aggression by feeding him treats works because the handler has accidentally reduced the dog’s inner tension and stress, not because the dog is now forming positive mental associations.)

Does it matter that it’s accidental? Isn’t the result the same either way?

In the case of leash aggression, yes. However there are other cases where attempting to use this model gets us nowhere. If that’s true then there’s obviously a problem with the underlying philosophy, and we have to look for another explanation of behavior, and another means of producing “positive” behavioral changes in dogs, one that makes sense in all cases.

Let’s forget about leash aggression for a moment and go back to the two cases that sparked this debate: 1) how I got Freddie to stop stalking squirrels by stalking them with him and then playing tug with him when the squirrels went up a tree, and 2) how Anonymous’ friend reputedly got his dog to stop stalking children by having the kids play fetch with him. In both cases good results were achieved through redirecting the dog’s hunting instincts into a bite-able object. There’s no question that that’s the primary cause of whatever changes took place in each case. However, would anyone in their right mind believe that Freddie somehow formed a positive association with squirrels because of what I did? Was he now going to hang out and roll around and play in the grass with them and lick their noses? Even if any of the squirrels were crazy enough to let this happen, what would Freddie have done if given a chance to get that close (provided I wasn’t there to supervise)? I’ll tell you what would happen because it actually did happen twice: he would grab the squirrel between his teeth and jaws and bring it to me. Luckily the two times this did happen I simply said, “Out!” and he dropped the frightened animal, who quickly ran up the nearest tree.

Here’s another angle to see this from: while it’s normal and instinctive for a dog of Freddie’s persuasion go into a stalking position when seeing a small prey animal (Dalmatians are essentially setters/pointers), it is decidedly not normal and not instinctive for any dog to stalk a human being, even a child. So when something of that nature happens it means there is something badly out of whack. And on a certain level even the dog knows it, or I should say, he “feels” it.

So why does Anonymous think his friend
’s dog formed positive mental associations with the kids he was formerly stalking? Since the friend reputedly changed the dog’s behavior by having the kids play fetch with the dog, we might be tempted to believe that this is true. But isn’t it more likely that the dog now sees the tennis ball that the kids are throwing for him as the release point for the internal tension that was causing him to stalk them in the first place? Remember, he didn’t really want to bite the kids. I mean sure, he sort of did. But mainly he was just looking for an outlet for his prey drive. The kids gave him one. Does he see them differently now? Perhaps he does, though I doubt if he could articulate the difference or explain why he feels differently toward them. But if there is a difference, is it because the dog has thought this through logically?

Those kids were a negative in my mind before but now that they play fetch with me they’re positive!

The more probable answer is that the kids were always positive in this dog
’s mind/experience. Or rather they always held a strong level of attraction for him. But his attraction for them also had this baggage, the feeling that something was out of whack. It didn’t feel 100% right to the dog to be stalking them, yet he felt driven to do it anyway. Now that he’s playing fetch (reportedly), and he’s able to bite the ball (so we’re told), the underlying impulse that caused the stalking behavior (the urge to bite anything acting preylike), was satisfied. And while the kids were (reportedly) the avenue by which that impulse was satisfied, I doubt very much that this dog feels much differently toward them than he did before, except that now he feels more satisfaction of his prey drive because of the ball, and feels that they’re a viable delivery mechanism for that ball.

Isn
’t that the same thing as making a positive mental association?

I don
t think so because the dog never had a negative mental association in the first place. Just as Freddie never had a negative mental association with the squirrels. And when you think about it, even a dog who wants to bite every dog he sees usually has no negative mental association with any of them personally. It’s always about the dogs levels of internal tension and stress, and finding a way to release that tension through biting an acceptable object instead of biting another dog or a human being (or a squirrel, for that matter).

So while dogs do learn by seeing patterns and associating this or that experience with either a pleasurable (positive) or stressful (negative) feeling state, it
’s a big mistake to think that learning takes place by making mental associations because as a trainer or dog owner it puts you in situations where you aren’t getting the results you were expecting, the dog’s behavior doesn’t improve, and you don’t know why. The reason is simple: you’re using the wrong model for learning and behavior.

I think even B.F. Skinner would have to agree on that.

 
"Changing the World, One Dog at a Time"

Friday, June 6, 2008

Chasing Squirrels

Here's another post from the Amazon.com vault. I promised a new client that I'd make this available again; her dog has an obsession with horses. It's much stronger than Freddie's fascination was with squirrels, but the principle still holds true: when you make yourself more relevant to your dog's prey drive, he'll be happy to give up whatever "prey" he's obsessed with, particularly if you let him finish the predatory sequence by biting a toy. It's really that simple (which doesn't mean it's easy).

Chasing Squirrels
I mentioned in a previous post the need to take charge of the dog’s emotional charge. You want to do this without attempting to become some mythical pack leader, of course (though Cesar Millan seems to be making a good living doing it). I’ve described this process in my work with Boomer, but another excellent example comes from my own dog, Fred, who used to love to chase squirrels in Central Park. At the time he was still having panic attacks on the streets, but was entirely calm in the park. In fact, I’d been trying to find a way to get him to play tug and fetch outdoors in order to cure those panic attacks, but his only way of expressing his prey drive was by stalking and chasing squirrels. Understand, I wasn’t as concerned with the fact that he was chasing squirrels; they were always too fast for him to actually catch. I was more concerned that he ignored me completely when he went into hunting mode. His attitude was, “I know you’re going the other direction. Don’t worry about me, I’ll catch up when I’m done here...”

So, one day I took some juicy pieces of chicken breast to the park. When Freddie spotted a squirrel and began stalking it—which he did by freezing, like a setter or pointer—I walked over and put the chicken in front of his nose to distract him; to try to get him to pay attention to me (or at least the chicken) and not the squirrel. He just ignored it. In fact, he kept moving his head around because my hand was blocking his line of sight. So I finally put it right into his mouth (which was open slightly). He let that juicy slice of chicken just sit right on top of his tongue for about half a second, then dropped it — ptaahh — onto the ground, keeping his eyes on the squirrel the whole time.

I was stumped. If I couldn’t distract him by putting a piece of chicken right into his mouth, how could I get his attention? Then it hit me: I would hunt squirrels with him. Maybe that would also solve my other problem; how to get Fred to share his prey drive with me. So, I put the chicken in my pocket, and later, while Freddie was sniffing around, I spotted another squirrel, one that he hadn’t seen yet himself.

In a hushed, highly emotionally charged voice I whispered, “There he is!” and began stalking the squirrel myself.

Freddie eventually picked up on my mood, and when he did, he saw the squirrel too, and dropped into his stalking stance. We were now hunting together. Fred didn’t know it yet, but I was now in control of the game.

We did this for a few days, then I added a new twist. We’d stalk a squirrel together but at some point, I’d make a quick move toward the squirrel, motivating it to run up the nearest tree. This always set Freddie racing off after the the little critter. While he did that, I'd pick up a stick, hoot excitedly and run away, waving it for Freddie to see.

Freddie would then be forced to choose between chasing the squirrel, and then circling the tree to no avail, or chasing me and the stick. In the beginning he always went immediately for the squirrel. But the thing is, the squirrels always went up a tree, leaving Freddie with nothing to sink his teeth into. That’s the critical thing here.

Meanwhile, I was still enticing him with a stick. Once he started to come toward me, Id shout, Freddie, come! (while he was already running toward me). Then, once he got to me, Id invite him to jump up on me and play tug-of-war. I either let him win, or, if he lost his grip, I immediately threw the stick for him to chase, which he did with the same intensity, more or less, that he had for chasing the squirrel. Once the stick was in Freddie’s mouth and he was able to lie down in the grass and crunch down on it with his jaws and kill it, he was truly satisfied. He never got that satisfaction from chasing squirrels because he never got a chance to bite one. After just a few weeks of following these steps, whenever Freddie saw a squirrel, all I had to do was whistle, or say, Freddie, come! and he’d immediately turn and run back to me for a game of fetch.

Of course, from the traditional standpoint, everything I did to change Freddie’s behavior was wrong:

1.) I encouraged him to chase squirrels, which squirrel-lovers disapproved of (I told them it was just a squirrel aerobics class)

2.) I encouraged him to jump up on me, and

3.) I was not only playing tug-of-war with him, I was letting him win!

These were all huge no-nos in the dog training world at the time. But doing each of these things helped me take charge of Freddie’s emotional energy. That was the whole point.

Here’s how and why it worked: when Freddie saw a squirrel he became filled with an emotional charge. He was so charged up in fact that nothing could get his attention away from his intended prey, not even a juicy piece of chicken sitting on his tongue! By immersing (or pretending to immerse) myself in the same emotions that he was feeling, I created a dynamic, magnetic charge between us. Then, by getting him to jump up on me and play tug-of-war, I decreased his resistance to my position as a vertical being and gave him the satisfaction of crunching something with his teeth.

It also helped that Freddie’s m.o. in hunting squirrels was to stalk them; to try to sneak up as close as he could. Then, when they started to run towards the nearest tree, he’d give chase. If he’d been an instant chaser like some dogs, this wouldn’t have worked.

Now, I’m not recommending that you chase squirrels with your dog. It just happened to work with Freddie due to a number of contributing factors that I was aware of at the time, and that you might not be with your dog. What I am recommending is that you find a way to take charge of your dog’s emotional energy, not so that you can always be in control of everything the dog does, but so that the dog can be in control of his own behavior, and doesn’t need you to constantly be telling him what to do (which is something some people seem to enjoy).

Here’s the thing: if you don’t have “willing” squirrels as your guinea pigs when teaching your dog to re-direct her energy into something safe to bite, and especially if your dog is more apt to go after children or skateboarders or other dogs, you have got, got, got to be able to get her addicted to tug before you put her in a situation where she’s going to come up against her biggest bite-temptation. Squirrels are wily and can run up trees. They’re safe (more or less). Kids and skateboarders and other dogs don’t have as easy a time escaping those teeth. So work on the pushing exercise first, then work on redirecting your dog’s energy into a game of tug, or just on heeling, or jumping up on command. 

The more you do that, the less tempting these other things will be.

"Changing the World, One Dog at a Time"