Category: Common Problems ¤ Author: Karen Pryor ¤ Title: Conditioned Aversive (the vanilla post) ¤ >Hey! We also have a problem with the paws-on-the-counter thing with our 10 >mos. old male English setter.... And so...out comes the list of primary negative reinforcers or aversives: mousetraps, clattering spoons, keep changing them around, hope the dog quits....etc. So, let me repeat, when you are training with the PRIMARY stimulus, whether it is putting food on the ground or dropping something noisy on the dog's head, you are losing the main advantage of the operant approach, which is to use a conditioned reinforcer, NOT a PRIMARY, to mark the dog's behavior. Thus the dog pays attention to his behavior, which produced that marker signal--not to the food, oh good, food, but not a clue as to why--or to the snaps or rattles--oops, what was that, but not a clue as to why. Using a conditioned reinforcer to communicate, that is the way you can modify behavior most rapidly. For example: I had a dog that liked to steal Kleenex out of wastebaskets and chew it up. This made a big mess, and was unhygienic to boot. This dog hated the spritz bottle we use to spray the houseplants. I filled a spritz bottle with water and ADDED a few drops of vanilla. I sprayed the dog in the face, not when it was doing something bad, just when it was doing nothing in particular. Ugh! How unfair? Not at all. I needed to establish the connection that vanilla means Ugh, (nature of horrid thing not specified) just as click means treat (nature of treat not specified, might be anything). The dog made the association: vanilla smell=something horrid. Then I sprayed all the wastebaskets with the vanilla scented water. end of problem. Since this was a young terrier, she tried again in about six months, and I repeated the experience: spray, vanilla, wastebaskets. End of problem for a lifetime. Thus: the power is not the AVERSIVE--catching her in the act and spraying her with water would have just made her do it while no one was around, and she already had that figured out. It was not the unpleasantness of the smell, dogs don't mind vanilla. It was the COMMUN ICATION made possible by establishing a conditioned negative:in this case, a scent. I would do the same thing with the kitchen counter. Who cares if the kitchen counter smells faintly of vanilla, or peppermint, or whatever? We find it pleasant. I would also scent any booby traps I set up, and I would set them in different places, not just the counter. The idea is to NOT to teach the dog to stay off the counters (a "dead dog" behavior; you can't really teach that) but to teach it to understand your "negative marker signal". Oh, ugh, turn away from here, bad things are here. This is exactly the same issue that has come up recently about dogs "preferring" primary reinforcers such as other dogs, and jumping up, and sniffing the ground, and so on, to treats or whatever the trainer can offer. Once the CONDITIONED reinforcer is truly established, as a marker signal for the animal's own behavior, and also as a signal meaning treats (of a varied nature, including car rides, scratching, dinner, etc., not just tidbits) then you are no longer looking at a dog that is weighing hotdogs against playmates or your petting against squirrels or whatever. You are looking at a dog that wants to make clicks happen and is thinking about its own behavior and how to DO what makes you Click. At this point, there is NO mere primary reinforcer that will compete; the dog is no longer in the position of comparing primary reinforcers. But you have to train to that level; you won't get there in a week. And you have to develop the intensity of competing temptations step by step--remember the shaping staircase (see the "Shaping" video). I keep seeing stories such as here's a dog that's been clicked for a week or two, at home, and has NO socialization skills, and now the owners take him to a class full of strange dogs, and are surprised that the dog acts like a maniac. And, naturally enough, feel that there's no way the clicker can overcome the dog's "dog drive" or whatever primary stimulus they think it's going for. Yes, there is, and you won't believe it until you really establish that association and see it happen. It does take a little thinking and effort, though. You have to shape the animal's response to the clicker in all circumstances, just as you shaped a simple operant such as a sit or a high five. Step by step. Karen Pryor Donna Fefee questions Karen about the above post and Karen's response is included here: Original message: "RE: CLICK-L POST FROM JUNE 3, 1997" From: <> Hi, Donna. Sure, you can put the post up if Helix wants it. What i post, I intend to be public information. Also, you said: In other words, the dog learns that "ah ah" means the loss of the hotdog then learns that vanilla spray = "ah ah".... Would that work?? You bet. I should have said something before this about how much I like the loss-of-hotdog method. I think this is truly useful; and kind, and efficient. Now, if you pair ah-ah also with the spray, then you have ah-ah as a conditioned P+ AND P- and the vanilla as a cond. negative reinforcer. Then, you precede your Ah-Ah with a raised finger or a hiss, and you have another superpowerful conditioned negative reinforcer/punisher. Then you overuse it all and bingo, you are back in standard training with a completely shut-down dog. KP.