Why positive only training?

Notice two things, Pawzk9...

1. I have often used the phrase "philosophically purely positive" because I have argued (repeatedly) already that no one is truly purely positive.... although they might be philosophically pure. These folks are probably just not aware of those punishment moments.

2. I am writing purely positive / 90% in casual language (clearly!)... how would anyone calculate the precise percentage of positivity in training!? So, for those who are philosophically pure or otherwise darn close... like maybe 95%. YOU are the one arguing (as I already have) that no one is PURE whether they think so or not.... so I am including that high percentage, too. Sorry if it was confusing.

There is this drift of thought that punishment is some sort of bizarre, inhumane, or ineffective treatment idea. It is to this line of thinking that I am speaking.

Now, the appetitive motivation in my other post.... the dog rushing or behaving in some clever way to avoid loss of food.... is NOT included in my example of a human avoiding failure on a test.

The phrase "appetitive motivation" sounds fine to me for the purposes of this talk, but I am not so sure it is used in behavior analysis. I'd want to see or read some examples of how appetitive works differently than other circumstances.
I'm assuming, by appetitive, you are talking about the levels of satiation or deprivation for some resource.

ohbehave said:

Notice two things, Pawzk9...

1. I have often used the phrase "philosophically purely positive" because I have argued (repeatedly) already that no one is truly purely positive.... although they might be philosophically pure. These folks are probably just not aware of those punishment moments.

2. I am writing purely positive / 90% in casual language (clearly!)... how would anyone calculate the precise percentage of positivity in training!? So, for those who are philosophically pure or otherwise darn close... like maybe 95%. YOU are the one arguing (as I already have) that no one is PURE whether they think so or not.... so I am including that high percentage, too. Sorry if it was confusing.

There is this drift of thought that punishment is some sort of bizarre, inhumane, or ineffective treatment idea. It is to this line of thinking that I am speaking.

Now, the appetitive motivation in my other post.... the dog rushing or behaving in some clever way to avoid loss of food.... is NOT included in my example of a human avoiding failure on a test.

The phrase "appetitive motivation" sounds fine to me for the purposes of this talk, but I am not so sure it is used in behavior analysis. I'd want to see or read some examples of how appetitive works differently than other circumstances.
I'm assuming, by appetitive, you are talking about the levels of satiation or deprivation for some resource.

Click to expand...

philosophically purely positive? Where do you get this stuff? Do you make it up as you go? I don't know that anyone who is an effective trainer is worried about being "philosophically pure" They are interested in being effective and (hopefully) as humane as they can manage in their methods. Part of that process is examining how one trains, why it works (or doesn't) and being honest about what behavioral forces are "at work"


as to appetitive, all I can say is that if you haven't seen it used, you haven't read much (at least not about how OC works in dog training). Here is a dictionary definition to help you: ap·pe·tite (p-tt)
n.
1. An instinctive physical desire, especially one for food or drink.
2. A strong wish or urge: an appetite for learning.

--------------------------------------------------------------------------------

[Middle English apetit, from Old French, from Latin appettus, strong desire, from past participle of appetere, to strive after : ad-, ad- + petere, to seek; see pet- in Indo-European roots.]

--------------------------------------------------------------------------------

appe·titive (p-ttv, -pt-tv) adj.

The American Heritage® Dictionary of the English Language, Fourth EditionAdj. 1. appetitive - of or relating to appetite; "appetitive needs

In other words, when we use +R or -P, we are dealing with an appetitive - something the dog strongly desires is added to or removed from the situation to change (increase or decrease) behavior. When we use +P or -R, an aversive (I'm sure you know what that is) is added or removed from the situation to change (increase or decrease) behavior. It's been my experience that using mostly appetitives increases willingness to learn. And using many aversives creates a dog who is unsure about offering new stuff. And they work on different parts of the nervous system (sympathetic/parasympathetic) and introduce different neurochemistry into the situation

Those are not the posts you're looking for. Let's move forward guys. Take your personal quips to PM or I will Vader choke you.

The NRM does cause more frustration but you want it in appropriate amount for that dog's drive and temperament. You don't have to use NRM all the time. Many working dog trainers, at some point in training, once dog is quite fluent (talking up to 8 months of training), stop using NRM and start using e collar nicks instead - say if dog messed up 5% of the time, it would mess up 0.5% of the time after punishments became more severe. If it so happens that they have to punish the dog all the time, obviously that dog would become reactive, stop trying and become fearful of consequences OR drive would be lost due to nagging - this only happens when trainer seriously messed up at some early stage in training. It shows how far application of NRM could go, the concept is the same, wanted result is the same, the punishment is simply more severe in hopes of getting a better performance. The NRM is not a strong punishment on its own, some situations it's not strong enough, in some others it's too strong..... With that said, just the NRM itself cannot cause a dog to loose drive or give up trying.

Willowy said:

Do you consider anything you do to be positive punishment?

I'll argue that the average person doesn't know that there IS anything to look up/research. "Punishment? Oh, I know what that means".

Click to expand...

Technically, saying "STOP IT!" is adding something he doesn't desire - so, yes, that's positive punishment. What I consider it is irrelevant. It's what Wally (the operant) considers it, the response his behavior demonstrates, which is what tells me what that signal/stimulus is. When I'm working with Wally, my thoughts are relevant only for:

-Knowing the rules of the human world that do not exist in the canine world ("why can't I pee on that grass, but this grass?")
-Knowing how to break down behaviors so I can explain a complex chain/behavior/task to Wally better via shaping.
-Interpreting his signals, performance, and responses.
-Increasing the difficulty of learned behaviors to increase mastery and fluency in a way that still brings him a high probability of success.
-Doing the previous four frequently in ways that keep him interested and teach him more and more behaviors which he can draw on in future problems.

My opinions of what my signals are is not one of those relevant things. Wally's opinion, however, is what matters.

And - unless I'm some "non-average" person - I just typed in "how to train a dog" and got a TON of links and information - and it just snowballed from there.

dogdragoness said:

wow................................ this thread is getting very deep LOL Im no pro... all i do is: correct what i dont like, show the dog what i WANT & reward it with whatever the dog likes best, rinse, repeat LOL, Im almost glad im NOT a professional

Click to expand...

What do you think that is?

You're using positive reinforcement and some form of punishment.

What's so complex? Conditioning, either Pavlov-style or Skinner-style, at its heart is simple. It's about giving clear feedback to the dog. Humans make it complicated because we like 40 words for the same thing for whatever reason. That, and we'd rather argue than just accept the fact other humans aren't always going to do things the way we do, and we always want to think about what we consider things and all that.

To the dog it's simple:

-Did I want that? No? I won't do that again, then. (Punisher)
-Did I want that? Yes? I will try that again, then. (Reinforcer)

Negative or Positive is just changing (or using) the environment to give the feedback.

And when something was working but isn't:

-Maybe I'll try it harder. Maybe that will make it work. (extinction burst)
-Oh well, I'll try it again later. It worked before. (Variably Reinforced Behavior)

And when it just never works:

-Oh well, guess I'll give up doing it that way. (true extinction)

Shaping can get complex but, from the dog's perspective, I think it's still simple. That hardest part is actually breaking down behaviors and carrying the dog through it slowly in case of lack of confidence, etc

-Everything I know has worked at some point, I just need to figure out which one is going to work now. (starting the process of solving a new problem)

Then when something is found:

-Oh! Maybe I'll try another twist. If one paw worked - maybe two paws works. If one bark worked, maybe I'll bark twice. (offering behaviors similar to the one(s) reinforced)

When the new idea didn't work:

-Oh, I see. I'll go back to do what worked. Let me make sure that is still right. (goes back a step for reconfirmation)


Maybe it's just me though and I'm OVER-simplifying it, but it's basically how I think when working with Wally and how I see him thinking (through the window of his behavior and signals) when he does this stuff.

You have two choices when you teach a dog:
Make the behavior you want a positive experience for the dog
or
Make doing anything BUT the behavior you want a negative experience for the dog.

I don't like creating negative experiences for the dog. It's pointlessly mean, and isn't very good for making the dog enjoy the process of learning. It's also an awful good way to get the dog not to offer behavior and makes training more advanced behaviors almost impossible.

Do you have to use treats and a clicker for that? No. My puppy had tummy troubles for three days this week and is still on a bland diet. She feels fine, so we've still been training - sans treat, and I never use an actual clicker. However, I'm still using a marker (YES!) and reward (tug/fetch). She's learning just fine that way, because she likes tug and fetch. The learning principal is still the same, though. Mark the correct behavior with a word with something the dog recognizes (for my deaf dogs it's thumbs up, for the rest it's a verbal "yes!") and then follow up with something good.