Toward a Unified Dog Theory: Understanding Pattern Recognition

 I suspect that the research into brain reward pathways will result in conclusions more complex than what takes place as the result of a single neurotransmitter, but if the dog is learning through the deliverance of unexpected reward, that would certainly play in to the results we see in dogs that are predictably reinforced versus dogs that are variably reinforced.  Behaviors tend to become stronger in the latter case. 

I also don't take Ian's participation in that project so much as a collaboration as I see it for what I think it was - a way to reach an incredibly large audience with material they would not ordinarily be exposed to.  Among the Cesar Millan devotees, I don't see too many who make a serious attempt to digest what other trainers and behaviorists say, especially if they perceive that the person is anti-Millan.  The blinders go right on.  So, perhaps this was Ian's way of speaking to the yet uninitiated who might see his material and be open to reading and understanding it because it's in Cesar's book.

 Lee, we're well aware that we don't know everything that's going on, and that's ok. Something happens internally, we can't see it, we know it's there, and it's not being ignored. We're simply dealing with the observable.

Honestly, of course there are associations, and environmental things going on. If you don't associate antecedents with consequences, you're going to waste a lot of energy engaging in behaviors when there is no reinforcement available. I'd suspect that it would benefit a species to not waste time hunting when there we wasn't something in the environment to signal the availability of food.

Call it what you want Lee, but because we define in terms of what we observed, even if we do find out about what's going on in the brain, we still weren't wrong.

Dogs are machines in the mind of LCK and of his favored philosopher, Dan Dennett.

 Natural Dog Training and LCK wholeheartedly hold that dogs are unconscious and have no ability to reason.  AND humans and dogs intimately share the same networked consciousness, thus making followers of NDT unconscious.

 I have tried, but cannot make sense of this, and if you read this summary of LCK's theory of dogs, you will not either. http://naturaldogtraining.com/articles/why-dogs-arent-stumped-by-cars/

 NDT is a theory that LCK conveniently uses to single himself out from the majority as a unique thinker, while never spelling out just what it is.  There is modern cognitive behanioral science informing us as to how canine physiology and psychology work, and there is LCK's nonsense.

Burl

Dogs are machines in the mind of LCK and of his favored philosopher, Dan Dennett.

 Natural Dog Training and LCK wholeheartedly hold that dogs are unconscious and have no ability to reason.  AND humans and dogs intimately share the same networked consciousness, thus making followers of NDT unconscious.

 I have tried, but cannot make sense of this, and if you read this summary of LCK's theory of dogs, you will not either. http://naturaldogtraining.com/articles/why-dogs-arent-stumped-by-cars/

 NDT is a theory that LCK conveniently uses to single himself out from the majority as a unique thinker, while never spelling out just what it is.  There is modern cognitive behanioral science informing us as to how canine physiology and psychology work, and there is LCK's nonsense.

spiritdogs

I suspect that the research into brain reward pathways will result in conclusions more complex than what takes place as the result of a single neurotransmitter, but if the dog is learning through the deliverance of unexpected reward, that would certainly play in to the results we see in dogs that are predictably reinforced versus dogs that are variably reinforced.  Behaviors tend to become stronger in the latter case. 

I think that you might have a good theory that is entirely workable and based in reality!!.

One  of the changes that i have made in my training is to take advantage of Jaak Pankseps concept of the seeking state, that the seeking of the reward is more pleasurable than the reward itself. A corrollary is that for my dogs, i need to give them  a bit of space around a reward. To not do so may actually give a slighly aversive quality to the reward.

I am intrigued by patterns of behaviour. There isn't a whole lot of evidence yet, but many complex phenomena can be described by quite simple maths. An example is chaoitc beahviour in crowds. I think that when a lot of the first behavourist experiments were done , it was neccessary to reduce the data and the outcomes to actually get workable conclusions. If we tried using the data analysis tools of the time to extract time related behaviour  there would have been nothing.

Even today, analysing simple patterns of behaviour are very time consuming and require knowledge of what we may actually be looking for. As for analysing brain waves, oh dear ... It requires near super computer levels of power to analyse. 

Please note that i am not certain enough to suggest that anything i have said is gospel. I also know that humans see patterns in random data, and have  a hard time leaving random patterns alone. MAy be dogs do the same to? What i often do to break up the non random nature of rewards fo rmany humans is to put markers on the ground at random intervals for heeling exercises. They all think they can do better. Most humans go 6 paces and have to reward. Now i wonder why their dogs have a hassle in the ring?? Dogs get the pattern pretty quick.

What i do know is that the process of operant and classical conditioning works wonderfully well for me. I think my dogs have me well trained in the process of getting a clicker out of my pocket.:)

I also do like what some of the more interesting neurosicence is doing. It suggests that that play state is pretty rewarding to humans and dogs, and that may be if i lighten up, my dogs might too! 

poodleOwned

I also do like what some of the more interesting neurosicence is doing. It suggests that that play state is pretty rewarding to humans and dogs, and that may be if i lighten up, my dogs might too! 

Hi, PoodleOwned,

Thanks for your response. I found your comments very interesting.

Yes, Jaak Panksepp has done some amazing stuff on the importance of play. I think his main rationale for studying play was as a means of reducing symptoms of ADHD in some kids. He's shown that rough-and tumble, outdoor play -- especially "free play," where the child (or the dog or the rat), is allowed to make up their own games.

"Are the neuropsychological benefits of childhood play diminished in children whose social play-urges are chronically diminished with psychostimulants? We don't know, but relevant genetic work in animal models has been initiated. Social play in rats can activate growth factors such as BDNF [or brain-derived neurotrophic factor, which stimulates neuronal and synaptic growth, differentiation, and survival in the hippocampus, cortex, and basal forebrain-areas which are directly related to learning and memory] ... (Gordon, et al., 2003).

"Our recent broad-scale brain gene expression analysis has indicated that activity of about a third of the 1,200 brain genes ... in frontal and posterior cortical regions are significantly modified by play within an hour of a 30 min play session (Kroes, Burgdorf Panksepp and Moskal, 2006,  Unpublished observations from Falk Center for Molecular Therapeutics, Northwestern University). If such dynamic brain changes evoked by play facilitate brain growth and maturation, perhaps solidifying pro-social circuits of the brain, we must worry about anything that diminishes the progression of such developmental processes." ("Can PLAY Diminish ADHD and Facilitate the Construction of the Social Brain?" Jaak Panksepp, College of Veterinary Medicine, Washington State University, 2/28/07)

LCK

**Moderator Warning**

Debating a topic is allowed.  Attacking a fellow member of the forum with insults is not.  Please consider your posts carefully.  You are welcome to edit your own posts before they are "red inked" by a moderator and you are reviewed for account suspension.  Reminder of the Forum Rules you agree to by posting here:

4.)      Debating and discussion is fine. We encourage it. It's how we all learn. However, respect your fellow members. Different posters are likely to express different opinions, and while they may differ from yours, everyone is entitled to express theirs freely. Whether it is blatant or covert, we will not tolerate rudeness, insults or personal attacks.  Do not disrespect, taunt, bother, bug or flame anyone, either on the forum or via private messages or email. If you fail to respect any your fellow members and insist on rudeness, insults, or personal attacks, you will receive a warning and your posts will be moderated for a period of time. 3 moderations equal 1 ban/suspension.

PM a moderator if you require clarification. 

Poodleowned, it seems to me that Shadow will get excited not by the exact nature of the reward, at times, but the process of getting a reward. I can get cheap little crunchy treats and he still gets excited about them because they are "treats" from me. Of course, I have been doing this for some time and it may have morphed into that, though I am not sure that was the process at the beginning. The beginning process seemed more operant. If I do this thing, I get that thing.

The initial event, long buried in time, becomes a conditioned habit that anything coming from me has got to be good. Good to a dog. Good to this dog, Shadow.

To draw another parallel to experimental physics, Quantum Mechanics could explain some things in a more elemental way but it doesn't change anything in the simplistic way we use Newton in our everyday world. That is, QM could hold all the explanation but the world still acts newtonian, to a large degree. That is, while NDT may contain the exact nature of what is happening, the overall effect is still operant conditioning. I think that's about the closest I've got to comparing the differences, if there are any. To that end, like QM, one explanation is as good as any other, I suppose, but accuracy still leaves something to be desired.

ron2

To draw another parallel to experimental physics, Quantum Mechanics could explain some things in a more elemental way but it doesn't change anything in the simplistic way we use Newton in our everyday world. That is, QM could hold all the explanation but the world still acts newtonian, to a large degree. That is, while NDT may contain the exact nature of what is happening, the overall effect is still operant conditioning. I think that's about the closest I've got to comparing the differences, if there are any. To that end, like QM, one explanation is as good as any other, I suppose, but accuracy still leaves something to be desired.

Again i have to agree. I think that a lot of the debate is about abstraction, about the level that we are looking at things at.

We need not go back to a microscopic view of the world if the system level gives us the answers. To a large extent, if the system level works as a tool (and it often does) then there isn't a whole lot of point.

 What do i do that is different? Well i play a whole lot more with my dogs. I proof quite early, i use a hell of a lot of premack, and i do a whole lot of different things.

You could stay at the system level to explain it, but i found the more micro (Panksepp's view) level helpful to understand how to do these things. 

I've used Premack before, too. Where the desired the activity is the reward for listening to me. Grandma's rule, a preferred activity of greater frequency is used to reward a desired behavior of less frequency.

ron2

a preferred activity of greater frequency is used to reward a desired behavior of less frequency

Hi Ron,

I agree that Premack's is important. However, Bill TImberlake and Jim Allison tested Premack's and found it didn't always hold true. In fact, with their experimental set-up they got the reverse of what Premack had predicted: a less probable behavior was used to increase the frequency of a more probable one. (In other words, the more probable behavior became more frequent after it was paired with a less probable behavior.) They called this response deprivation.

The problem with Premack, as I see it, has nothing to do with how we apply it in dog training. It's one of our most valuable training tools. The problem is that it's worded incorrectly. It should be described as using a more innately satisfying behavior (chasing a Frisbee, playing tug) to teach a behavior that's less satisfying for the dog (holding a down/stay, heeling).

Anyway that's how I see it,

LCK

 Hi

Looking at http://users.ipfw.edu/abbott/314/Premack.html

i think that the experiment is a long way away from how dogs owners would think  about Premack. This experiment relied on two behaviours running and drinking. I think there is a flaw in the design of the experiment. Some one forgot to check the normal baseline coupling between the activities me thinks. We have no idea whether running ws truely instruemental, and whether drinking is too.

I do agree that ss trainers, we talk about more desirable and less desirable activiites. It might be a mis quote of Premack but it makes more sense to us humans.:)

Lee Charles Kelley

The problem with Premack, as I see it, has nothing to do with how we apply it in dog training. It's one of our most valuable training tools. The problem is that it's worded incorrectly. It should be described as using a more innately satisfying behavior (chasing a Frisbee, playing tug) to teach a behavior that's less satisfying for the dog (holding a down/stay, heeling).

Almost bordering on semantics. Perhaps I am wrong but I had always thought innately satisfying/less satisfying was implied. That is, your words might be more liguistically accurate, a behavior thesaurus, so to speak, but not inherently changing the meaning. But I think your words for this were elucidating, at least for some readers.