1. Tell us a little about your training background.
I started in 1989 with my first sheltie in a "pure Koehler" club. I trained him through Open with that club, including the "forced retrieve". I was good enough at it that I became an instructor for the club and ended up teaching their novice competition class. I crossed over in 1993 after attending a Pryor/Wilkes seminar. I was ready to change (or quit training!) because of the overall effect I saw Koehler-type training (at least, as this club taught it) having on my dogs. I left the club in 1995 when it came to the point that they would not allow me to incorporate anything even remotely resembling clicker or operant training in my classes.
2. How do you define clicker training?
I define clicker training as operant conditioning with emphasis on use of the conditioned reinforcer. I would define operant conditioning as teaching the subject to operate on its environment in order to achieve desired consequences, so that the vast majority of what the operant trainer works with is positive reinforcement. Most undesired consequences would fall into the category of "negative punishment" ("time outs" and such). The only place for "positive punishment" (application of a physical aversive) is where three criteria are met: (1) the behavior must pose a threat to the life or physical safety of the dog, other animals, or people; (2) the behavior must be suppressed across a broad spectrum (ie, it must never occur) and (3) the behavior cannot satisfactorily be eliminated through counter-conditioning. As you can see, these criteria are extremely limiting. I cannot think of any situation in pet training or any kind of competition training where these criteria can be met with the possible exception of rattlesnake bite prevention. Taking a wide circle back to the beginning, clicker training teaches the dog to "work for the click", i.e., do something that will trigger the reward. So by definition it operates almost exclusively in the realm of positive reinforcement.
3. Why did you write Clicker Training for Obedience?
It seemed like a good idea at the time. Actually, it was Karen Pryor's idea. We were on one of the early clicker lists (probably click-l, but it's been a long while) and I was writing posts explaining what I was doing and thinking about. Remember, at that time there was very little by way of clicker literature generally, and literally nothing on competition training. So we did a lot of brainstorming on the early lists. Somewhere along the line Karen sent me a post saying "you're writing a book, you know". I answered "really?" -- and the rest is the rest. Took me 4 years to get it done, partly because writing is incredibly difficult and partly because I was learning as I went. So there was a lot of rewriting. The best thing about it was that it really forced me to think through everything I was doing, which made me a much more conscious trainer.
4. Clicker Training for Obedience was written several years ago. Is there anything you would change or add if you were writing it today?
I'm surprised that it holds up as well as it does, but yes there are things I would change or add.
The discussion on variable reinforcement is a bit of a hash as I mix together variable reinforcement and variable duration schedules. I'm still not much of a fan of variable reinforcement as such and use it only with finished behaviors and not in the shaping process as a rule, although there are exceptions.
The discussion on punishments is philosophically viable but technically inaccurate. The problem with discussing punishment is that the term has one meaning in the behaviorist community and quite another in the world at large. Skinner defined punishment as a response to a behavior which serves to suppress that behavior ("punishment suppresses behavior"). But in society as a whole we use the term to describe a range of actions which have nothing at all to do with suppressing responses to a given behavior as such but include incarceration, reprisal etc., all of which we know have little or nothing to do with affecting behavior but merely exact or impose some punitive effect. It's one of those cases where our common language divides us and obscures clear discussion. The other problem with discussing punishment is that, as a rule, punishment is very reinforcing to the punisher and I find at seminars that once that subject comes up it's all people want to talk about. That's a big reason (I think) why Karen won't even allow the issue to arise at her forums. The Baileys used to talk about it, but they had the same experience and it's now a footnote in their presentations.
There's a section where I talk about using the "scruff shake", and that's completely gone from the repertoire. Very New Skete-y of me, and very old. Basically I came to a point about 3 1/2 years into the project when I told Karen that there were lots of things I wanted to go back and rewrite. Her response basically had something to do with invoking the wrath of God were I even to start to consider such a thing. I guess that operated as a conditioned punisher.
I would clean up and expand the mechanics discussions, largely thanks to the time I've spent with Bob and Mouse Bailey. I understand much better the concepts of criteria (know what you want), timing (click when you get it) and rate of reinforcement (make it pay off); clicking for action and treating for position; splitting (or, shaping responses not behaviors) and so forth.
I think I'd also have more to say about the importance of extinction from the very early stages of training.
I'm sure there are other things, but those are the main ones.
5. You're involved with service dog training now. Tell us about that.
Actually I was involved with service dog training for two years with Canine Companions for Independence; that came to an end in June 2000. I still participate in a list for people who are using operant to train their own service dogs but I'm not actually involved in that training at this time myself. It's something I could gladly do again, but I would prefer to work with people training their own dogs rather than trying to produce "off the rack" dogs. I have yet to see a training model that I was really happy with, although I generally liked what I saw at Paws With A Cause best.
The thing that strikes me most about service dog work is that if the service dog is conditioned through punishments the end-user cannot maintain the training for the simple reason that the end-user is typically not physically able to deliver physical aversives. At CCI, the conceptual end-user is a full quadriplegic. Now, if you train with punishments the resulting behaviors are dependent on the availability of punishments. If the punishments are not there the dog gets a "false positive": if I wasn't "corrected", I must have been right. So in addition to everything else one can say about the futility of punishments in training, it is absolutely counterproductive to the aim of producing a reliable service dog.
I won't really say anything about CCI's decision to stop its "pilot project" on using operant conditioning other than that I and the Baileys disagreed with their decision and the reasons for it.
6. What are the most significant challenges facing dog trainers today?
Probably not much different than the challenges historically except that some of the context has become more difficult. The vast majority of dog owners are pet owners, and most of them are appallingly ignorant of dogs. I do a two-month stint on a pet dog board from time to time, and the questions that come up can be mind-bending. That's nothing new.
What is new is the increasingly harsh public climate regarding so-called aggressive dogs and, in some cases, ownership of dogs generally. One ridiculous extreme: in San Francisco you no longer are called a dog "owner" but a "guardian". Sorry. I own my dogs. I'll stop owning them when they buy their own food and pay their own vet bills. The whole "animal rights" movement with its theme that "dogs are just the same as people" has potentially broad destructive effects. Dogs are not the same as people. Dogs are dogs, and we have to teach people how to relate to them as such. In that vein, Turid Rugaas' work is invaluable.
So I guess the challenge can be broadly summed up as "teaching responsible dog ownership", which isn't exactly headline news except that the dog-owning population is now so huge but the growth of the responsible dog-owning population isn't keeping pace. More and more people are attracted to the "bully breeds" and Mastiffs who have absolutely no clue at all as to how to raise, train and manage those dogs. Incidents happen and the fallout is enormous. I'm not for breed bans, but I think we need to make a more assertive effort to find those owners and educate them. (I also think we would benefit from some greater control over who can get those dogs. That's an impossibility I know, but it would be beneficial for the simple reason that too many people get those dogs for altogether the wrong reasons or, in many cases, with no real understanding of what they are getting into.) Pit bull rescues are full up; where I live these dogs are increasingly dumped in the desert where they become semi-feral and crossbreed with coyotes.
Operant trainers are uniquely qualified to teach lessons of responsible stewardship, especially if we can take a holistic approach and incorporate the teachings of people like Turid Rugaas into the training.
7. What advice would you give to new clicker trainers?
Learn the basic theory of operant conditioning, but don't get too hung up in it. Read Skinner's About Behaviorism, Don't Shoot The Dog, Rugaas' Calming Signals.
Then, just do it. Once you're on board with the philosophy almost everything else is mechanics. The rules are simple and you learn them by doing. It's best if you can take a non-specific attitude toward your first dog; let your experience with that dog teach what you need to know about shaping.
What are some of the rules?
Criteria, timing and rate. Timing is everything, but you can't have good timing if you don't know what you're looking for; that's criteria. Rate of reinforcement is important because behaviors must pay off; I tell new trainers that I want to hear a click at least every 3 seconds. In my experience I have yet to encounter a training problem that I couldn't get a handle on by referring to at least one of these elements.
See behavior, click behavior: What you click is what you get and what you got is what you clicked. If the dog is giving you something you don't want, it is almost certainly the result of your training, so stop and figure out how you got there. Don't replicate failures but build on success. My working rule is this: if out of any 5 trials (repetitions) I get 2 failures I stop and rethink.
Video your sessions wherever possible and "go to the tape" when you're not sure what's happening. Once you stop being embarrassed by looking at yourself in action you will find it immensely valuable.
Keep the sessions short. I typically work sessions of 1 minute. It's better to do several short sessions than one or two long ones.
Shape responses not behavior (be a splitter, not a lumper).
More behavior is better than less behavior (the more behavior you get the more behavior you can reinforce, the more behavior you reinforce the more behavior you get). Trainers who have a specific aim (e.g., competition trainers) have a lot of trouble with this because they are so concerned about the limited scope of their chosen field. But it is absolutely critical that the dog learn that it is driving the reinforcers, so that the dog makes conscious, desirable choices about which behaviors will pay off best. That means, in the early stages, you should reinforce across a broad spectrum to get the dog as engaged as possible. You can always establish discriminations later. It isn't hard once the dog is a full partner.
But mostly, have fun with it. None of what most of us do involves life and death, just games we play with our dogs. If you are doing Search and Rescue or mine detection work and such the stakes are considerably higher, but very few of us are so involved and even there, it's important at many levels that the training be worthwhile for all.
8. What are you doing now to popularize and promote clicker training?
I write articles for the AKC Gazette and Off Lead magazine. I also do the pet board I spoke of earlier. My schedule has not allowed me to do a lot of public teaching but I've been approached by a local store to handle their training. I'm thinking about it.
9. What's your vision for the future of clicker training?
No question it is the wave of the future. Trainers in all fields use it and with increasing success. There's a clicker-trained sheltie on the AKC National Agility team. I know there is at least one clicker-trained OTCH dog, not including Patty Ruzzo's dog Luca (I would call Patty, like Dawn Jecs, an operant trainer who doesn't use the clicker and so I include them in our community). I'd be willing to bet that in 10 years people who advocate the use of punishments in training will be virtually extinct. Even today, people who are not operant trainers have to pretend to be in order to keep up appearances. Read Captain Haggerty's article in the Gazette last year to get a clear idea of how far some folks have to go to pretend to be au courant on behaviorism.
On the other hand, it will continue to be important to keep clear what clicker training is. It is not simply the use of the clicker. Karen Pryor distinguishes between "clicker users" and "clicker trainers". I know of people who refer to themselves as "clickers and Koehler". That's a pure oxymoron. We want to keep casting an ever-widening net, but we have to keep clear who we are and what we stand for in training. And that's the really hard part, because popularity always has a way of diluting content. That's one of the reasons I try to keep coming up with my little aphorisms (e.g., "what you click is what you get"), as a way to keep the basic messages clear.
10. What's your part in that future?
Wow. I don't know; I don't really think in those terms very much. I'm not the best trainer I know, nor am I the strongest theorist. I hope that I've helped get the word out. I think my greatest strength lies in being able to put the theory and practice together, showing how this technology works and making the theoretical aspect simple and clear.
11. What has clicker training taught you?
I'll answer these together because I don't really know how to separate them. To the first: yes, absolutely. Clicker training has helped me because it has taught me how to set specific goals and build on them. This has been especially helpful with our son, whom we adopted when he was nine. I'm not always the parent I want to be, but I know the parent I want to be and keep working on it. I've learned how to take responsibility for my mistakes, own them and do something about them. Not to sound sappy, but I've learned more and more to focus on the positive in my life and not obsess on the negatives, of which there are always plenty to get hung up on if one should so choose.
Clicker training has taught me to think in terms of behavior, both my own and that of others; how to identify what's happening in terms of behavior and how to approach situations in terms of behavior modification. I'm less effective with people than with dogs, but I'm learning.
I've learned how to improve without being concerned about perfection. You can always do better, but you must also appreciate what you have done well. In most fields of work we cannot always control the result, but we can control the quality of what we do. So I've learned to focus on that and let the results take care of themselves. My son plays competitive soccer and I tell him the same thing about that: he can't control what the coach wants to do, he can't control the referees, he can't control the odd bounce of the ball. All he can control is the quality of his play so he has to focus on what he is doing well and plan ways to improve the areas where he is weak. That's his job. Everything else will tend to fall into place.
And I've learned how to be supportive in marriage and other relationships. It's all too easy to fall into patterns of negativity because very few days go by without something happening to upset one. But I don't dwell on those things because they are transitory and there is nothing bad that happens in the daily whirl that is more important than the relationships I have with the people I am closest to. So I don't punish those events; I take them for what they are and we move on, strengthening the things I want to strengthen and letting the rest fall by the wayside.
12. What was your favorite clicker training moment?
There are many (including the morning I spent with Alexandra Kurland and Karen Pryor clicker training Alex's guide pony Panda in Cambridge last year), but there was one recently at a seminar I did in Minnesota that was priceless. We were working dogs on scent discrimination, and one lady had her sheltie on a table doing two metal articles. Now, there are many ways to do scent discrimination. You can do it by starting with all articles scented and gradually introducing non-scented articles. I do this like any other I would do, establishing the scented article and immediately teaching the dog to discriminate based on scent by introducing an unscented article right away. I'm not sure either approach is better, I'm just more comfortable with this approach.
Anyway, this sheltie had not done scent before this event but she was starting to "get it". She picked the scented article 3 times, and on the 4th time picked the unscented one. The trainer gave no response, and the sheltie dropped the unscented article but it fell right onto the scented one, forming a kind of X. The sheltie started nosing at the articles but couldn't get the unscented one off -- so she picked them both up! And then stood there, proudly, the two articles in her mouth, looking expectantly at her handler. I don't know if the written description gets it across, but it was hilarious and exhilirating at the same time to see this shy little dog working with such determination on a task she had only first learned about 10 minutes before.
I think this is the sort of thing I have in mind when I talk about the value of building behavior generally. I can't imagine a force-trained dog being willing to make that effort; but a dog that is conditioned to try because the only consequence is that the effort doesn't pay off is a dog that will produce bonus performances when you need it most and expect it least.
Copyright of all posts is the property of the original author. Please obtain permission from the original author before copying, quoting, or forwarding.
List and Site Owner: Melissa Alexander, mca @ clickersolutions.com