Contact: LDRidgeway at gmail dot com

Saturday, October 31, 2009

-P as -R: The Assisted Walk Out

This post is not primarily about practical training, but is rather about learning theory. Understanding it may, however, help with practical matters such as consistency and timing.

In behavioral analysis, the -P quadrant is traditionally described as follows: When a rewarding stimulus is removed (that's the minus-sign) as the outcome for a behavior, the behavior tends to decline (that's the P for punish).

Meanwhile, -R is described as: When an aversive stimulus is removed (again, the minus-sign) as the outcome for a behavior, the behavior tends to increase (that's the R for reinforce).

Neither of those definitions is found much in practice, either in the lab or in practical training.

Instead, the most common form of -P is this: When an object that represents an opportunity for reinforcement is removed as the outcome for a behavior, the behavior tends to decline.

And the most common form of -R is this: When an aversive stimulus is avoided as the outcome for a behavior, the behavior tends to increase.

As a 2Q trainer, I choose not to use -R in my training, just as I choose not to use +P. Those are the two quadrants that use aversive stimuli.

I try to use +R as much as possible, saving -P for tune-ups. Even though -P does not use an aversive stimulus the way -R and +P do, the use of -P tends to be demotivating. Another way of saying that is that using -P is itself aversive to the subject, even though no aversive stimulus is used.

What I've described above is a fairly traditional view of the behavioral quadrants described. But now I'd like to describe something new, something that I, at least, have never seen described before: the use of -P as -R.

The idea is that, since removing the opportunity for reinforcement is unpleasant for the dog, you can strengthen an alternative, desired behavior by letting the dog learn that she can avoid losing the opportunity for reinforcement by performing that desired behavior. Thus -P, a procedure for punishing an undesirable behavior, becomes avoidance conditioning as a way of reinforcing the desired behavior.

Here's a practical example. I've begun using a procedure I call the Assisted Walk Out (AWO) as a way of training Lumi and Laddie not to dawdle on their pick-ups and re-entries into water on land-water-land retrieves. Because the AWO involves removing the opportunity for reinforcement (the assistant picks up the duck, preventing the dog from being able to derive pleasure from completing the retrieve), the AWO functions as -P.

But if the dog learns that a particular cue will result in an AWO if the dog does not perform the desired response, and as a result the dog's response improves as a way of avoiding the AWO, then what is really occurring is avoidance conditioning, just as in -R.

In fact, although the AWO is -P, no specific behavior is being punished. That is, it doesn't matter whether the dog performs undesirable behavior A, B, or C, the outcome will be the same, the AWO. Only if the dog performs the cued behavior is the AWO avoided. Causing a single behavior to increase, while causing a variety of alternative behaviors to decline, is also typical of reinforcement, both -R and +R.

So is the AWO actually -R after all? The answer is no. The procedure involves no aversive stimulus. It is removal of opportunity for reinforcement, one of the classical forms of -P. But in this case it functions as -R, because the dog experiences avoidance conditioning.

Again, -P is typically used to cause a specific behavior, such as barking, to decline. But the AWO uses -P to cause a specific alternative behavior to increase.

As a final note, one of the reasons that -R is the most powerful of the four operant conditioning quadrants is that when the dog learns a behavior that enables her to avoid an aversive stimulus, she is reinforced every time the dog performs that behavior and successfully avoids the aversive. That is, -R continues to work even when the dog does not experience the aversive. In fact, in practice, the desired behavior can continue to get stronger over a period of time when the aversive stimulus is not occurring.

Hopefully, the AWO will function in a similar way with Lumi and Laddie, reinforcing the desired behavior not only when the AWO occurs, but also continuing to reinforce the desired behavior every time the dog performs the desired behavior and avoids the AWO.

Monday, October 26, 2009

The Whistle Sit and Casting

[The following email was my reply to a question I received in private correspondence, asking what to do if the dog does not sit facing the handler on a whistle sit. The writer was concerned that if, for example, you cue "Back" to a dog that's facing east instead of toward the handler to the south, the dog will spin 180° and run west instead of going north as the handler intends. I took the opportunity to address some related topics at the same time. Perhaps I'll have time to convert this to an essay in the future. See Alice Woodyard's reply at end of the post.]

Thanks for writing. I'm no expert, but I'll tell you my opinions.

First, in my experience, "Back" does not teach the dog to make a 180 degree turn. Rather, it drives the dog straight away from you, which my dogs, at least, seem to consider fun. If they don't have the bird, it's much easier to get a good "Back" response than a come-in.

In fact, "Back" is so deeply ingrained in advanced handling dogs that I and other trainers I've seen do not use the verbal cue for angle-backs, relying instead on a silent visual cue if possible. It's not unusual to see a trainer use a visual angle-back cue in combination with a verbal "Back" only to see the dog spin around and continue straight away from the handler, taking the dog further off line, essentially following the verbal cue and ignoring the visual one. The dog had a reason for going that direction, and when you call "Back", the dog seems to take that as confirmation she was going the correct direction.

On the subject of casting a dog who is not sitting straight, I would say that the only immediate practical problem is that you won't be able to cast the dog both directions. If the dog is skewed right, for example, it will be difficult to spin her to the left. However, if you plan to send her straight back, on an angle-back to the right, or on a straight right "Over", I think she'll probably go the way you cue. She'll interpret the cues relative to the direction she sees you, not relative to the position she's sitting in.

That said, I'm not comfortable casting my dogs when they're not sitting straight, and I've noticed that other trainers often aren't either. In my case, this isn't really because I don't think the dog will go the correct direction. Rather, it's because the dog has two modes -- hunt versus handling. As a rule, hunt is for marks, handling is for blinds. The dog needs to have a clear sense of which mode she is in, because handling mode ruins marks in those case where a confident hunt is required, and of course hunting mode prevents the dog from responding to handling cues. When the dog sits skewed, I think she has one foot in each mode, sort of responding to the whistle but on the verge of switching to hunt mode.

I don't think it's good for her to be in that mental no-man's-land. I think it's confusing and might lead to frustration (for the dog as well as for the handler). I want her to have as clear a sense as possible as to which mode she's in. On the one hand, when she lines up for a mark, I want her to be totally in hunt mode, though I may in rare cases need to blow my whistle and switch her to handling mode. On the other hand, when she lines up for a blind, I like a nice, driving send-out on "Back", but above all, I need her to be in handling mode from the moment I blow the first whistle.

When the dog sits crooked, I typically wait a little while before cueing anything. That's because I believe the dog is poised to break, and I don't want her anticipating my next cue and guessing wrong. In addition, there's a behaviorist's reason for not casting immediately. The cast is a reinforcer, and if you cast instantly when the dog sits askew, you're "clicking" the dog for that incorrect response to your whistle. So there are really two good reasons to pause for a few seconds: to give the dog a moment to calm down, and to avoid reinforcing the incorrect sit.

Next, I give a come-in cue. For me, this is multiple tweets, optionally with both arms pointing straight down. Don't say "here" in this situation, because the judge is liable to interpret that as you picking up the dog, and your test will then be over. Another reason for not saying "here" is that, while the come-in whistle hopefully means to the dog, "Come in till I stop you with another whistle", "here" means, "Come all the way to me, I won't be stopping you on the way in."

As the dog begins the come-in, she straightens out, and then I immediately blow another whistle sit (one tweet). The dog may show some confusion, but once she sits (this time straight, typically), I immediately cast her toward the blind. This rewards her for sitting, and also rewards her for the entire chain of coming in and then sitting, making it easier to straighten her out this way over time. Because you did not reward the skewed sit, and consistently straightened her out before casting her to the bird, she'll hopefully become more likely to sit straight on future whistle sits in order to get herself cast toward the bird sooner.

Let me add another point, where I wrote "once she sits, I immediately cast her." This is one area that I think 2Q and 4Q trainers need to think somewhat differently. 4Q trainers, as usual, are at the advantage. It would really be better not to cast immediately for several reasons. A delay gives the dog a chance to gather herself mentally and physically, get her attention on you, coil for the cast. Most importantly, it discourages "auto-casting", where the dog releases herself from the sit, often heading in the wrong direction. Those are reasons why a 4Q trainer will typically not use an immediate cast after a whistle sit.

But a 2Q trainer doesn't have the ecollar to reinforce the whistle sit. For a 2Q dog, the only reason to sit is to get to the bird. The longer you delay your cast, the less valuable the reward is, and the less you're adding to your dog's reinforcement history for sitting on the whistle. Therefore, if you watch me handling my dogs, you'll see that I tend to cast much earlier than the good 4Q trainers. They probably think this is inexperience on my part, since it's also a typical beginner mistake, but I'm doing it intentionally, to "click" the dog for sitting. Fortunately, my dogs are pretty good handling dogs, so I can afford not following the role-model of good 4Q trainers in this particular detail.

Sorry for the long answer. I hope somewhere in there you found something helpful.

LL&L

[Alice Woodyard, one of my mentors, reviewed the above correspondence and sent the following feedback:]

That is a good answer, I think. I want to make one comment. The technique of using a quick toot toot to straighten a sit should ONLY be used in training. It is very dangerous to use it at a test/trial (event) because a number of judges (not all, but more than a trivial number) will interpret it as "correcting" or "training" or "intimidation" or "not making progress to the bird" and will DQ you for it, or at least warn you (you'll be lucky to get only warned).

If you are mentoring [someone], be sure to tell her what is okay for training but not at the event.

Personally I think the dogs learn to adjust for their own crooked sit, once they are in the field (not the yard), and, as you say, "Back" is defined with reference to the handler, not the dog's spine.

Therefore I think straightening a sit is sort of worthless aerobics for both, and I don't personally recommend it.

Thursday, October 22, 2009

Do Dogs "Think"?

[From a post to PositiveGunDogs on October 22, 2009]

On Oct 22, 2009, at 9:03 PM, another poster wrote:
Thinking is always ongoing with animals as we all know.
Hi, [name]. I'm afraid I'm on the other side of this statement from both you and Alice [this refers to Alice Woodyard, who also replied to the same person I'm replying to].

I'm going to assume that by "thinking" you mean that dogs perform conscious reasoning. If so, first of all, I don't think there's any way to determine whether dogs do that or not, and secondly, my own observations suggest that they do very little of it.

I see too many situations where something should be obvious to a dog but the dog's behavior tells me that it isn't obvious. A common example is awareness of lines (such as a leash or rope) attached to the dog's collar. A dog without line experience often acts as though she has no idea that the line can prevent her from moving beyond the length of that line, and also often seems unaware that she cannot go on one side of an object while the handler attached to the other end of the line goes on the other.

Dogs, of course, can learn those things, but it doesn't seem to me that they learn by means of a reasoning process. Instead, they just begin to learn that particular stimuli in particular contexts combined with particular decisions on the dog's part have particular outcomes. Dogs are extremely good at that kind of learning. It COULD involve reasoning, but it does not need to, and my experience suggests that it does not. I say that because if you change the stimuli or contexts in ways that should not interfere with the reasoning process -- for example, go to a different location -- often the dog will no longer behave according to previous experience and will have to go thru the learning process again. It takes time for the dog to learn which elements of the context and stimuli make a difference in the outcome and which do not, though eventually they do learn some of that, and we call that "generalizing". Dogs are notoriously bad at generalizing. To me, that's because they're not incorporating reasoning into the process, they're just collecting data points.

An analogy might be a blind person learning about walking thru doors with a sighted assistant. As first, he might think he can get out ahead of the assistant, but thru experience of walking into doors, he eventually learns that he needs to wait until he feels the assistant moving forward, or hears the assistant say "It's OK to continue without me." To the assistant, it may appear that the guy can now see because he doesn't walk into doors any more, but the truth is, the blind guy still has no idea whether a door is front of him or not. He's just acting as though he can see because experience tells him what stimuli and contexts have desirable outcomes when he walks forward.

We are often mystified about how a dog could possibly not "understand" some cue. I think the mystery goes away if we stop thinking that it's a rational process and start realizing that the dog is just unconsciously sorting out how the current context and stimuli compare to her previous experiences. One little change in the context or stimuli -- a new scent, a change in wind direction, some internal sensation such as hunger -- might make it impossible for her to recognize this as a "similar" situation to one she has experience with, while our rational abilities tell us that that scent, wind direction, or internal sensation has nothing to do with the cue we're giving her and the behavior we're asking for.

I don't think I can prove to you, or to myself, that dogs don't "think". But I also don't think I can prove that they do.

LL&L

Wednesday, October 21, 2009

Breaking and Creeping

[For now, this article is in the form of private correspondence with someone who wrote to me. In the future, hopefully I'll convert it to essay format. I've changed the names.]

I have the feeling (might be wrong) that you're thinking of using the AWO (or WO) for breaking. I've gone down that route, I no longer think it's the right way to do it. The problem with it is that even if the dog doesn't get the bird, she still gets the chase, and the chase is enormously self-reinforcing. True, not getting the bird is frustrating, and eventually that frustration may overpower the pleasure of the chase, but it's a slow route.

Here's what I suggest instead:

* Get a tab for Fido's yard collar. A "tab" is just a very short lead that snaps to the collar's loop like any other lead. I buy my tabs from Lion Country, then tie an extra knot at the knotted end to shorten it to about 9". I don't want the tab ever to get caught in my dog's feet, of course.

* It's important that Fido begin wearing the collar and tab from now on whenever training, especially if you are NOT working on breaking that session. You want Fido to become so habituated to the equipment that he doesn't associate it with the break-training (below) when that is in progress. What you do NOT want is for Fido to associate the collar/tab with not breaking, since of course he won't wear that equipment in a test. If he learns that the only time he can't break is when he's wearing the equipment, the training won't help you in a test.

* From now on, whenever running Fido on a mark where there is the remotest possibility he will creep or break, hold the end of the tab while calling for the throw with your other hand. The fact that you are holding the tab needs to be imperceptible to Fido, again because you don't want him learning that the only time he can't break is when he feels or otherwise senses that you are holding his tab. So keep your hand behind his head, and keep the tab nice and loose, but hold on tight.

* Ask someone to act as "line judge" and have that person call a number or "dog" after all the birds are down. Ask the "judge" to wait a few extra seconds before calling. If possible, sometimes have male and sometimes female "line judges".

* If you're not already doing it, add substantial and irregular delays after the "line judge" releases your dog, before you call "Fido". The old timers say, "Smoke a cigarette." That's an exaggeration, but it gives you the idea. You don't want Fido learning to anticipate your release. If he does, like any athlete experiencing "anticipation response", he will begin jumping the gun.

* Finally, call "Fido". For Hunt Tests, I don't think there's any reason to say it loudly. I know that some Field Trial handlers use a louder voice for longer marks, but we don't run those long marks in Hunt Tests.

* As Fido launches, simply let the knot from the tab slip from your hand, imperceptibly to Fido.

* If at any time Fido attempts to creep or launch too early, he won't be able to because you're holding the knot. His movement will take up the slack in the tack and he'll be immobilized. Fortunately, dogs do not seem to experience this as "Someone is holding me by the tab on my collar." Instead they experience it as, "Hmm, for some reason, I don't seem to be able to move forward. Guess I might as well wait until my handler calls my name."

The above procedure will gradually create a habit of not creeping or breaking. It's gentle and effective, and has the added advantage of not taking extra time for your turn. Simply continue using this procedure until you're absolutely certain that there's no need to hold the tab any more. Then, a time will come that you'll feel there's also no need for the dog to even where the collar/tab any more. My dogs train completely naked. :0)

Since you've gotten thru Senior, I guess honoring isn't a problem for your dogs. If you'd like some additional thoughts on using a similar approach to train the dog not to break when honoring, please let me know.

Best of luck!

LL&L

Tuesday, October 20, 2009

The Assisted Walk Out (AWO)

I am now ready to reveal what may be the single most important secret for training the 2Q retriever.

Do 4Q trainers -- that is, virtually all retriever trainers, the ones who use ecollars, heeling sticks, choke and pinch collars, ear and toe pinches, etc. -- know about this secret? Yes and no. If you were to tell the typical experienced retriever trainer the secret, they'd say, "Well sure, I coulda told you that." But they do not use the secret I'm about to describe, because they don't need it. If you're lucky enough to get to join their group, none of them will tell you the secret in advance. And if you try to put the method into play while training with them, it will annoy them, possibly to the extent that they will not allow you to use the technique during group training, because it takes far more time than the methods that 4Q trainers employ.

I've given the secret a name: the Assisted Walk Out (AWO). It's related to the Walk Out (WO), which I've described several times in the past (it's the solo version of the AWO). In addition, another poster on this list (my apologies, I've forgotten that person's name), described this exact method for training the whistle sit (WS).

Indeed, the AWO is an invaluable method for training the WS. I haven't used the AWO for the WS, I've gotten by with the unassisted WO, and my dogs are doing OK with it, so you can get by with the WO for the WS if you have no choice.

However, I've come to the conclusion that a 2Q trainer can NOT get by without the AWO for land-water-land (LWL) retrieves, at least for some dogs, at least for harsh conditions such as cold water.

So here is a description of the Assisted Walk Out: When the dog is out in the field and does not perform correctly, have an assistant pick up the bird (or whatever retrieval article you're using). Then you as the handler, call out "sit", walk out to the dog (slow is OK, slow is even preferred), gently slip on the dog's lead, and silently walk the dog (again, slow is fine) back to the start line. Do not use this opportunity for training heeling, sitting, recall, or any other behavior. During the entire procedure, you're negatively punishing a particular behavior, whatever behavior the dog was doing when you stopped her. Don't dilute the lesson with some other lesson, just walk out, slip on her lead, and gently, quietly, walk her back to the start line. Then with all the joy and enthusiasm in the world, run the dog on the same retrieve again. That's it. That's the AWO.

Why does the AWO work? You may not need this information, but knowing it may help you to understand the crucial element of the AWO (and the WO): The dog must not complete the retrieve after an undesired behavior. The reason is that for a field-bred retriever (and probably many other dogs, such as Border Collies), the opportunity to complete the retrieve is of enormous value. Therefore, losing it is so undesirable that the dog will forego other pleasures if they learn that doing so results in losing that opportunity. You see, completing the retrieve is not the only thing of value out there in the field, and it is subtle enough that the young or naive field-bred retriever may not yet have discovered how valuable it is. (The AWO is a good example of what I call "discovery" training.) Because it is not the only thing of value, the dog will also be engage in other behaviors that are ALSO of value: licking or eating the bird, exploring nearby territory for frogs, etc. But the dog is only doing those things because she (or he) believes that when she has finished doing them, she will ultimately get to complete the retrieve.

As a result, if you do not use the AWO, but instead allow those other behaviors to continue while repeatedly calling the dog, your recall, if sufficiently well trained, will eventually succeed in bringing the dog back. That will give the illusion that you're on the right track, and that all you need to do is continue practicing until eventually the dog discovers that it's more fun to come straight back than it is to do those other things first, and then come back.

But in my experience, the opposite will happen. What you are actually doing is REINFORCING the undesired behaviors with your recall. Remember that a cue that has been well-trained with a high value reward ITSELF acts as a reinforcer. So let's say the dog licks the bird, and you, not wanting the dog to do that, call the dog. The dog picks up the bird and brings it to you. "Whew," you think, "disaster averted. At least the dog never put her paw on the dog and started to eat it." But here's what the dog experienced: "I licked the bird and something GOOD happened, I was REWARDED: Daddy used one of my favorite rewards -- my recall cue -- to REWARD me for licking the bird."

And here's something even worse that might happen next. As the insight hits you that you need to stop calling the dog when she licks the bird, you try it that way. Boom, you've just invoked Mr. Extinction Burst. "Oh," thinks the dog's behavioral instincts, "licking didn't work to get that recall reward? OK, let's try a more vigorous version of it. Let's try chewing the bird, or even breasting it." Congratulations, you are brilliantly if inadvertently shaping a really disastrous pick-up and return pattern. Be like me. Keep it up for a couple of years before you figure out what's going on. :0)

Finally, this may be the worst of all: Because the AWO (and it's poor cousin, the WO) are so difficult to arrange for, it is easy to let the occasional exception occur. Boom again, now you've just invoked Mr. Ratio. The best way to improve a behavior's persistence is to put it on a ratio of intermittent reinforcement. "Oh," thinks the dog's behavior instincts this time, "I only get rewarded for this behavior every few times. OK, I won't worry about it when I try it and nothing good comes of it. I'll just keep trying, and EVENTUALLY it will work for me." Dogs can build up tolerance for dozens, even hundreds, of unreinforced behaviors if you are go about the "training" the right way. It's easy: Just let the dog OCCASIONALLY complete the retrieve after performing some undesired behavior. This, of course, is why you must put the AWO on a zero-tolerance schedule.

The most mysterious part of the AWO, the part that makes it difficult to believe in and therefore difficult to execute, is understanding how important it is to the dog to be able to complete the retrieve. After all, the dog is acting as though completing the retrieve has no value at all. How can depriving the dog of the retrieve make any difference, when she doesn't seem to care about it anyway? To convince yourself that completing the retrieve really is valuable, all you have to do is to trust me and start using the AWO. When it works, and the undesired behaviors stop occurring, and the dog starts picking up the bird and coming straight back to you, then you will realize why it worked. If you start using the AWO early enough in the dog's training, you may be able to prevent any bad habits from ever forming, and have a dog who never shows the symptoms that my dogs do. I dearly hope that's the case and that I, too, will share in that benefit with my next puppy.

In conclusion, I know all too well how difficult the logistics are to arrange to always have an assistant available when training LWL retrieves. If you absolutely must train alone, then I can only give you the same advice I've belatedly given myself: Apply the same zero-tolerance, no exceptions policy to your use of the unassisted WO as you would if an assistant were available to help you with the AWO. One of the most important and annoying corollaries is that that means, do NOT train in locations where you cannot readily get to the other side of the water if the dog does not perform correctly.

I hope the AWO will be as valuable to you as it has been to me in the short time I've begun using it.

Lindsay, with Lumi & Laddie (Goldens)
Laytonsville, Maryland

Field training blog: http://lumi-laddie-test-series.blogspot.com (see "Archive of Video Blog Entries" in right margin)

YouTube playlists:
-- Lumi: http://www.youtube.com/view_play_list?p=BC338082E0B890DB
-- Laddie: http://www.youtube.com/view_play_list?p=9A44913FB240932A

To further explore the frontiers of dog training, join our DogTrek list at: http://groups.yahoo.com/group/DogTrek

Sunday, October 18, 2009

EDF and the 2Q Disadvantage

[For now, this post is a somewhat-edited email exchange. Maybe I'll convert the format to a short essay when I have more time.]

Hi, Gaby. Thanks for the thoughtful note.

I decided I needed some sleep, so it looks like I'll get there about 10:00 AM, maybe even a little later. I hope that's OK.

All dogs (and handlers) experience what Alice (my mentor) calls "Event Discount Factor" (EDF). You have to train to a higher skill level than what the dog will need in a test to have a chance of passing. That doesn't really answer your question, as to what the exact cues are -- for example, I'm sure you're right, duck scent is a huge diversion -- but all those other factors -- people, dogs, the judge's tent, chairs, coolers, the holding blinds in the field, the holding blinds on way to the start line, the decoys, the clipboards, live birds, shotgun poppers, live shotgun fire, waking up in the middle of the nigh to drive to an unfamiliar venue, an article of clothing you only wear to tests, and subtle or not so subtle changes in the handler's behavior, even changes in his scent because of arousal or nervousness -- also tell the dogs "This is different, this is special, this is exciting, this is pressure." Even group training is generally missing many of those factors, so the real solution has to be: Train your dog so far above the test requirements that you can afford EDF. I do think that we should try to bring in as many of those factors as we can as often as we can.

As for yesterday's test, I'm convinced the primary factor was the cold water, plus the fact that my dogs are trained with positive methods. Negative reinforcement is simply more powerful than positive, all experiments show that. My dogs can overcome some adversity based on their pleasure at doing the job right, but they'll never be able to overcome as much adversity as a dog who fears the possible consequences of an shock for refusing to make the effort. Ecollars are particularly valuable for training recall because in fact the dog's first instinct when she feels a nick/burn is to run to the handler. I've heard from several people that it's a lesson the dog never forgets. One reason for this is that every time a dog performs a behavior that successfully avoids a nick/burn, the dog thinks, "Aha, that behavior saved me from getting nicked!" So the ecollar keeps working even when you don't press the button!

I've tried to introduce a parallel concept on the positive side. "Aha, that behavior was FUN!" I think it has been effective, but it wasn't strong enough to get the dogs into ice cold water on the RETURN from a retrieve. Notice that they're willing to go OUT for the bird. That part is so much fun that they'll put up with the adversity. Coming back isn't as much fun, but the water is just as cold.

The only real solution that is see, other than continued practice and training, is to understand that my dogs will never be able to pass all tests. They'll be able to pass those that don't contain too much adversity, but if the test involves strong discomfort, my dogs are not ever going to be reliable on the return. That's my opinion at this point, anyway.

I'm going to go get ready and head for Mt. A. See you soon.

--Lindsay

Thursday, October 15, 2009

New Puppy

[Posted on PGD list October 15, 2009, at 3:14 PM]

On Oct 15, 2009, at 1:56 PM, jarred bonica wrote:
Not sure what "wearing the Dog" is?
Hi, Jarred. "Wearing the dog" is a term I learned from Jody Baker. It means attaching the dog to you as you go about your life. I use a short leash with a flat collar, and I put the leash's loop handle on my belt. I think people typically use a 6' leash, I'm not sure. Our leads are shorter.

At first, of course, the dog doesn't know what to do, and gets pulled a little as you gently move around (don't give any warning, just move about as naturally as possible). Soon, the dog is paying attention, can instantly wake from a nap, and so forth. The result is a dog who gets in the habit of paying attention to you and staying close. Since you make the experience pleasant for her, with kind words, petting, and the occasional treat, she develops a positive association with that habit.

Of course, the other benefit is that the dog can't get into trouble because you're keeping her close to you. That doesn't JUST mean that she's not getting into trouble. More importantly, it also means that she's not learning how to get into trouble and not learning how much fun such trouble can be. So there's long-term learning go on in that way, also.

Later, when the dog doesn't need to be on lead all the time, you may find that she welcomes the opportunity for you to wear her again from time to time. My dogs seem to consider it a gift I can give them.

I might mention one other thing you might want to start on immediately, something you'll want the dog to learn and use for the rest of her life: Sit to say please. Wait for her to sit before giving her food, before giving a treat, before running thru the door on the way out or in, etc.

That is, rather than the you cueing the dog to do something, the dog learns that she can cue you to do something she wants: "Please give me that food, please open the door for me and give me permission to go thru, etc." To me, this idea that the dog is cueing you rather than you always the one giving the "commands" is desirable on the ethical/philosophical level, but it also has practical benefits.

For example, a dog who reflexively sits to say please can quickly understand that that's the appropriate greeting for a stranger, not jumping up or barking. If the dog doesn't leap out of the van as soon as you open the door, but sits to say please, she's less likely to jump out when you're not ready and less likely to get hit by a passing truck. I'm sure other examples come to mind.

And for the field dog, there's yet another benefit: If and when you get around to train handling, she'll need a remote sit. A dog who thinks of sitting as a way of getting something she wants has a more natural path to learning the whistle sit, which she comes to understand is a way she can use to get you to cast her in the right direction to the bird.

Of course there's a million other things we could talk about with a new puppy, but if you're going to train the dog for retrieving, I'll just throw out a couple more thoughts, more in the way of DON'Ts than DOs.

First, don't train the young puppy to retrieve to hand. Yes, you can do it, she can learn it, it's not all that hard. But I have come to believe that it has a significant disadvantage that really hurts the 2Q trainer. Retrievers (at least some retrievers) do not by nature want to give up the bird (or whatever it is she's retrieved). They want to run to it (oh, yes, that's the fun part), and they learn to love picking it up and bringing it back TO SHOW YOU. All that is fun and natural. And eventually they'll lose interest and drop it and you can throw it again. But if you decide to train retrieve to hand and take it away from them, for some dogs (I've got two of them), you are tacking an aversive event onto the end of the retrieval chain. That aversive stimulus can be cumulative such that over a long period, you might find that the dog's desire to come back with the article actually diminishes from when she was a puppy. I don't think this is the only issue that can result in a deterioration of the pick-up and return, but I think it can contribute. So I would stay away from training retrieve-to-hand until the dog has been retrieving for four months or more. During those four months, let the dog decide for herself when she wants to drop the article. JMHO.

Second, keep the retrieve a highly valuable commodity by carefully limiting the number of retrieves your puppy gets to do. Maybe no more than three retrieves per session for a 3-month-old, maybe four per session the next month. You can do two or more sessions a day, but you want each session to end with the dog thinking, "Oh, that was great, I just know this is what I was born to do. I want to do MORE. What, we're quitting? No, no, throw another sock/toy/bumper/Dokken for me, PLEASE!" But no, you end the session. Now the dog has something to look forward to, something to dream about when she's napping, and she can hardly wait for the next session where she'll get to retrieve some more. So by limiting the work, you're building motivation.

Oh, sure, there's more. Conditioning to noise (eventually learning that shotgun blasts predict great fun), desensitization to the blow dryer, toothbrushing, nail grinding. Have the vets and their assistants give your pup treats when you go to the vet's office.

And handle your puppy in general, get her comfortable with the idea that her body is YOUR body. Hold her ears ("these are my ears", you explain to her), hold her muzzle ("my muzzle"), hold her paws ("my paws"), let her learn that the way to get you to put her down when you're holding her in your arms is to relax rather than to keep squirming.

If she tries barking to get you to give her something or because she sees something she thinks she's supposed to bark at, give her what I call a "time-in" (shorter and gentler than a "time-out"). Simply attach her lead, walk her to the most boring place nearby you can find, gently get her into a sit, and stand with her for 30 seconds (check your watch) doing absolutely nothing. If you want, you can kneel beside her and drape your hand over her back, or you can stand beside her, but she's needs to stay quiet, not playing. After 30 seconds, go back to whatever you were doing together. My dogs each learned to stop barking after just two time-ins each when they were young puppies, Lumi at Petsmart, Laddie at home. (This did not solve the problem of Laddie's severe case of barking in the van when I take Lumi out to train, but I don't believe there is any solution to that other than parking as far as possible from the training location. Nonetheless, neither of the dogs barks to get food and neither of the dogs barks at other dogs in the Petsmart.)

Got to stop. :0)

LL&L

Monday, October 5, 2009

Line Mechanics

Here's a description of my line mechanics as they've evolved (others use their own variations):
  • I use the same line mechanics when practicing as when running a field event, with the exception that in an actual Hunt Test, it's against the rules to show the Guns to the dogs before the birds are thrown. Even when practicing for a Hunt Test, I still show my dogs the Guns, since I want my dogs to have the best chance for success in the practice session.
  • Hunt Tests have some special features, such as walk-ups, handler's guns, and remote sends (for example, with the dog in front of the holding blind and the handler behind it). It's helpful to practice anything you'll need for an event, and practicing with a handler's gun can improve marking and reduce head-swinging. However, I never did practice walk-ups with either of my dogs, and neither of them had any problem with them when we needed them in tests. Of course, I might have been less nervous the first time if we'd practiced first.
  • Duck calls are used in Hunt Tests but not Field Trials or WC/X. I think practicing with duck calls for a Hunt Test has advantages and disadvantages. The advantages are that it acclimates the dog to what she'll hear in the event (including a call coming from behind her from the judge) so it doesn't come as a surprise to her the first time she hears a duck call, and also that it builds excitement and helps the dog learn to deal with that excitement. The disadvantage is that it makes it easier for your dogs to spot the Guns, and if you're trying to maximize difficulty level in practice, practicing without duck calls may give your dog an edge in the test because she's that much more skillful at finding the throws. If you choose to ask your throwers not to use a duck call and they are used to using one and then shooting a blank when the bird is in the air, ask them to fire before throwing so the dog has a better likelihood of seeing the throw.
  • When training with a group, to get the maximum benefit, try to arrange for someone to act as your line judge, and give him your preference on whether you want him to use a duck call.
  • Evaluate whether your current phase of training will benefit more from the throwers remaining visible after the throw or retiring, and let them know what you want them to do.
  • Bringing the dog to the line, try to follow a route that shows the dog the stations in the reverse order the marks will be thrown. Note that you are not necessarily showing the marks in the order that the dog will retrieve them, because you will not necessarily send the dog in an exact reversal of the throwing sequence.
  • Position the dog at heel in such a way that she has a clear view of all the marks. As obvious as that sounds, you sometimes see a dog break, not because she wasn't trained to be steady, but because she couldn't see one of the throws.
  • In addition, if she is a "two-sided dog", position her so that she is on the same side as the go-bird. For example, if the go-bird will be thrown left to right, position dog at right heel. I've heard that even if you follow that policy in general, there are exceptions, but I don't know what they are.
  • If you are in an AKC Hunt Test, you are not permitted to show the Guns to the dogs. That rule does not apply to AKC Field Trials nor to GRCA WC/WCX. Of course, it doesn't apply to practice, either. I feel that the dog gets the most benefit from practice if she gets good looks at all the throws, so I always show the Guns to the dogs in practice, even if we're preparing for a Hunt Test. Even if I can't show my dog the Guns and just have to wait for her to find them herself in a Hunt Test, I try not to call for the first throw until I'm reasonably certain my dog knows where all the Guns are.
  • If you're showing the Guns to your dog, turn your body toward each Gun, cue your dog to turn with you if necessary, and point with your hand to direct her attention to the Gun if necessary.
  • Show the Guns to the dog in the reverse order in which they'll be thrown, so start by showing her the go-bird first. Watch the dog to make sure that she turns her head, locks in on the station, and seems to absorb that picture. She may only need a second or two per Gun. I've found that it's surprisingly easy for me to get mixed up and show my dog the Guns in the reverse order that I meant to, so I try to practice mentally beforehand.
  • When you finally get to the Gun that will throw first, say, "sit, mark." Then, provided the dog is holding her lock, call for the throw immediately. Use the hand on the opposite side of the dog for your signal.
  • During each throw, watch the dog, but look up for a moment to see where the fall is. You may need that information to know when to blow a come-in whistle, or to handle the dog to the fall if you need to convert the mark to a handling retrieve.
  • After the first bird lands, count a short interval to yourself, turn to face the next station, and immediately call for the next throw. For practice, count two seconds. For events, count one second. With practice, the dog will learn to turn her head the same way you turn. Until she's learned that, she'll have to depend on the duck call or gunshot to know when to turn.
  • Continue with the same cadence until the go-bird has been thrown. Then wait for your "line judge" to call a number, or if you don't have a line judge, imagine a delay.
  • Wait a random amount of time. The old-timers say, "Smoke a cigarette." Both in practice and in events, try not to let your dog learn that she can anticipate when you're going to release her, or the athlete's natural anticipatory response may lead to her eventually breaking before the judge calls her number. (Note: In early training, I wanted my dogs to learn that sitting up high gets me to release them, so I released them the instant they sat up. Once they learned that, I gradually added duration while they held that posture.)
  • If your dog is not yet steady and is on a slip cord at the line, do not send her if she's pulling. Wait for the instant that she relaxes, then send her. That way she learns relaxing as a way to get you to send her.
  • When you're sending your dog to the last bird down, the "go-bird", she is probably locked in, so don't distract her by putting your hand near her face. Simply call her name to release her. As a rule, use a quiet voice for the release, though some handlers use a louder voice the longer the distance, to give the dog a little more information.
  • For all birds except the go-bird, position your hand over dog's head, flattening your hand to act as a "gun sight", before calling your dog's name to send her. If you're certain she knows where the next fall is, you can just release her, but it's surprising how often you see someone send a dog and then watch as the dog gets out there and has no idea where she's going. I use an unusual gun sight, my entire forearm. I've never seen anyone else do that, but I will say my dogs are outstanding markers (that is, excellent at running straight to the fall).
  • To get the maximum benefit from practicing with a group, if in the future you intend to run in an event that requires your dog to honor, always have your dog honor the next dog, of course on-lead until you're certain she won't break. She should not feel any tautness on the lead while watching, and your honoring set-up should be significantly different from when you're at the start line, while being especially careful not to block the dog's view of the throws. Blocking the view is illegal, but it also makes breaking more likely as the dog moves around trying to see the throw. As an example of a different posture for honoring, I always stand at the dog's right haunch and face away from the field while the dog watches the field, and I give the verbal cue "just watch". An important part of a dog learning an event-quality honor is learning that honoring is what happens after she's finished her own run, as opposed to "cold honoring", which just means that the dog is watching other dogs being sent, something that virtually never happens in a test and provides little training benefit.

Showing the Guns to the Dog

In a post entitled "Question on Double Retrieves" on Oct 5, 2009, at 2:55 AM, haftaski77 wrote:
when doing a double (say, in a WC test) - when I walk up to the line, what order should I show Stormy the gunners (1, 2 or 2,1) and which gunner do I have her sit and face? 1 or 2?
In the US, the rules vary with type of event. The ones I know about (or think I do) are the AKC Hunt Test and Field Trials, and the GRCA WC/WCX. You might want to check the rules for your particular event. Rule books are available online.

You are not permitted to show the gunners to the dog in a Hunt Test. Of course, in a Junior test, it's not much of an issue because you're running singles, but it's an issue for Senior and Master tests.

For Field Trials and the WC/X, you can take as long as you need to show the dog the gunners. If I may throw in an anecdote, when Lumi got her WCX, she had trouble spotting the center Gun on the land triple when we got to the start line, and I took a fairly long time helping her find it, eventually having her switch sides, then switch back. I still remember her sweet body language when she spotted it, "Hey, there's another Gun out there!" Then I positioned her facing the first Gun, cued "Sit, mark," and called for the throw. She got a good look at each throw, did well on her retrieves, and when we were finished and leaving the start line, one of the judges said quietly, "Nice handling." I think it was giving Lumi the chance to find that middle Gun that elicited the friendly remark.

The sequence I follow is to show the dog the Guns in the reverse order of throwing: I show go-bird first, first-bird-down last. In that way, you move smoothly from one Gun to the next and end where you want to be when you call for the first bird. It sounds simple, but it's surprisingly easy to get to the start line and do it backwards, so I practice it mentally beforehand once I see the set-up.

I might mention that even when we're practicing for a Hunt Test, where you're not allowed to show the dog the Guns, I still carefully point them out to my dogs in practice sessions.

While you can't point out the Guns before the birds are thrown in a Hunt Test, you CAN line your dog up after the judge has called your dog's number. She probably won't need help lining up on the first bird, assuming you're sending her to the last bird down, but she might get some benefit out of you taking your time lining her up after she gets back with each bird before sending her out for the next one. I've seen people just take the bird while the dog was standing or sitting in the right general direction and then call the dog's name to send dog back out, only to have the dog leave the line without knowing where she's supposed to be going.

One other point, for when you get to the WCX. In a WCX, they throw the birds in the order left to right or right to left, but you can pick them up in any order you like. The Pro I used to train with before he retired gave me this advice, and it's worked for me: Pick up the go-bird first, the other outer bird second, then the middle bird. That seems to give the dog the best chance.

I have a little article on line mechanics on the sidebar of my blog.

Lindsay, with Lumi & Laddie (Goldens)
Laytonsville, Maryland

Field training blog: http://lumi-laddie-test-series.blogspot.com (see "Archive of Video Blog Entries" in right margin)

YouTube playlists:
-- Lumi: http://www.youtube.com/view_play_list?p=BC338082E0B890DB
-- Laddie: http://www.youtube.com/view_play_list?p=9A44913FB240932A

To further explore the frontiers of dog training, join our DogTrek list at: http://groups.yahoo.com/group/DogTrek

Thursday, October 1, 2009

2Q versus 4Q (continued)

[Posted to PositiveGunDogs on 10-1-2009 at 8:27 AM]

On Oct 1, 2009, at 7:48 AM, Lindsay Ridgeway wrote:
I'll throw in a few more points because they are so important to field
training, though they may play little role in other animal training.
I just thought of a few other points I'd like to mention:
  • -R has a trait that is unique among the four quadrants: -R conditioning occurs even when the aversive stimulus is not presented. As I understand it, this is why phobias sometimes feed on themselves, with the subject's fear increasing over time. What's happening is that every time the subject successfully avoids the aversive stimulus, that is a reinforcing event for the avoidance behavior, and the avoidance behavior may actually become more vigorous over time even though no aversive stimulus is occurring. In my opinion, this fact may be the most important way in which 4Q trainers have an advantage over 2Q trainers: Their dogs continue to be reinforced by their -R methods every time a cue is given, long after the conditioning has switched from escape conditioning to avoidance conditioning. 2Q trainers don't use the -R quadrant and therefore require an ongoing schedule of explicit reinforcement.
  • I spoke of the difference between +R provided by the trainer, versus +R provided by the act of performing the behavior. The terms for those variations are extrinsic versus intrinsic +R.
  • Retrievers do not necessarily experience intrinsic +R during their early involvement with retrieving, or at least not a useful form of intrinsic +R. I have a name for conducting training in a way that attempts to leverage the dog's dawning discovery of intrinsic pleasure in the desired behavior: I call it "discovery training".
LL&L

2Q versus 4Q

[Posted to PositiveGunDogs on 10-1-2009 at 7:48 AM]

In the thread "Hunting vs Trialing":

On Oct 1, 2009, at 6:27 AM, Denise Parlin wrote:
Being new to this list, can someone (in private might be better, to save everyone else) please explain 2Q vs 4Q to me so I can have a better understanding?
Hi, Denise. Others might not know the terminology either, so I'll put my perspective on these terms on the list.

Operant conditioning (OC), one of the recognized behavioral mechanisms, is described as four quadrants: positive reinforcement (+R or R+), negative reinforcement (-R or R-), positive punishment (+P or P+), negative punishment (-P or P-).

Reinforcement refers to outcomes that result in the previous behavior increasing in subsequent trials.

Punishment refers to outcomes that result in the previous behavior declining in subsequent trials.

Positive refers to the subject's learned anticipation that a stimulus will occur.

Negative refers to the subject's learned anticipation that a stimulus will not occur or will stop occurring.

The stimuli that occur as outcomes might be rewards (pleasant to the subject) or aversives (unpleasant to the subject).

A reward or an aversive might or might not have a conditioning effect. That is, you might give a treat, and the dog might eat the treat and seem to like it, but that doesn't guarantee that the preceding behavior will increase.

By contrast, the terms "reinforcement" and "punishment" specifically refer to a statistical change in behavior as increasing or decreasing.

So combining all of the above, we have the four OC quadrants:
  • +R: a reward outcome that results in the preceding behavior increasing
  • -R: an aversive that occurs before the behavior that the subject learns a behavior to escape, OR, an aversive outcome that the subject learns a behavior to avoid (a synonym for -R is escape/avoidance conditioning)
  • +P: an aversive outcome that results in the preceding behavior declining
  • -P: a reward that occurs before the behavior that the subject learns not to do in order to maintain the reward, OR, the loss of opportunity for a reward as an outcome that results in the preceding behavior declining (the latter is much more common than the former in my experience)
As you can see, two of the quadrants involve reward stimuli, while the other two involve aversive stimuli and escape/avoidance.

4Q trainers use all four quadrants. 2Q trainers do not use (or at least try not to use) the two quadrants that involve aversive stimuli. That is, 2Q trainers limit themselves to +R and -P.

"2Q training" is almost a synonym for "positive training", but some positive trainers try to limit themselves exclusively to +R.

I'll throw in a few more points because they are so important to field training, though they may play little role in other animal training:
  • For a field retriever, the most important +R by far is not any outcome provided (in the dog's perception) by a human, but is rather the opportunity to retrieve, especially if the article is a bird, and most especially if the article is a live or recently shot bird. Thus all field trainers, 2Q and 4Q alike, rely primarily on positive reinforcement.
  • By the same token, the field context also provides far more opportunities for environmental +R than the training and event venues of most other sports, with the result that undesired behaviors that will be positively reinforced by environmental stimuli is a vastly more challenging problem for the field trainer than for most other animal training.
  • Conversely, the field context also provides far more hazards for environmental -R or +P than the training and event venues of most other sports, with the result that desired behaviors that the dog will learn an avoidance response to is a vastly more challenging problem for the field trainer than for most other animal training.
  • The behaviors involved in field training, both desired and undesired, are far closer to the dog's inbred instincts than the behaviors involved in most other sports. Instincts are not learned behaviors, they are inborn tendencies. While they can be influenced by OC, they remain very powerful and, often to the confusion of the trainer, tend to become more prominent as the dog spends more and more time involved in field work. The result is a learning curve in which the dog seems to respond as expected to the trainer's methods, and then over time behaviors that seemed to be established or on their way to being established begin to deteriorate. While a conflict between OC and instinct can occur in any training project, it's especially probable in field training.
  • The closer a stimulus or the opportunity for a stimulus is, the more it influences the subject. This fact is noticed in some other sports such as agility and obedience, which require some distance work. Of course, it becomes a huge factor in field training.
Sorry if I've given you a hundred times more information than you wanted. Sometimes it helps me to review these ideas for myself.

Lindsay, with Lumi & Laddie (Goldens)
Laytonsville, Maryland

Field training blog: http://lumi-laddie-test-series.blogspot.com (see "Archive of Video Blog Entries" in right margin)

YouTube playlists:
-- Lumi: http://www.youtube.com/view_play_list?p=BC338082E0B890DB
-- Laddie: http://www.youtube.com/view_play_list?p=9A44913FB240932A

To further explore the frontiers of dog training, join our DogTrek list at: http://groups.yahoo.com/group/DogTrek