Dog-Dog Reactivity – Treatment Summary
The seminar in Seattle was great fun, lots of good folks and excellent questions from the audience. In the morning I talked about treatment for dog-dog reactivity, and I promised a summary of that on the blog. Here it is, with the obligatory caution that the morning itself wasn’t close to enough time to cover the topic in depth, so the summary here will be chapter titles (but hopefully helpful ones). Knowing that I couldn’t possibly talk about how to handle every type of case, here are some highlights:
REACTIVITY? What are we talking about here? When I use the term I am talking about what we usually think of as “over reactivity,” or “reactivity” that we see as inappropriate. After all, a loose body greeting is a “reaction” to another dog, right? In this case, I am talking about barking, lunging, snarling, snapping, stiffening etc… in other words, doing things we humans don’t like that makes us nervous that the behavior might be followed by aggression or trouble of some kind. It’s not a great term, but it’s better than “aggression,” since so much of behavior that we consider problematic is not aggressive at all.
MOTIVATION: I showed a video montage that illustrated that ‘reactivity’ can be the result of many internal states or desires: a dog who is barking and lunging could be : 1) afraid of the approaching dog and trying to get it to stop or go away, 2) frustrated because it can’t get to the other dog to play, 3) warning the other dog to stay away because whenever it gets close the barker gets a snap on the neck and it hurts, 4) begging the other dog to come closer so that it can get into a fight, which the barker heartily enjoys (rare, but it happens).
THRESHOLD: In all cases, in my humble opinion, one can’t work on cases like this until you understand a dog’s individual threshold of response. I want to know what is the lowest intensity of the trigger stimulus that elicits the lowest intensity of the “reactivity.” Some trainers let dogs go over threshold and bark/lunge for example until the behavior extinguishes, but that leads to what’s called the “rehearsal effect” (like repeating a bad habit) and I’m not personally a fan of it.
TREATMENT: My Ah Ha! moment was when I was working on this talk, and realized that all the methods I was going to discuss had, in a way, a similar result: the dog is reinforced for either increasing the distance between him and another dog, or for decreasing the potential of a direct, highly aroused confrontation. Treatment modalities can be categorized into 4 groups:
1. Classical Counter Conditioning (Note: I talked about this in the seminar, but didn’t at first include it in this post. A comment to the first version reminded me that I should have put it in in the first place!)
2. Operant Conditioning, Positive Reinforcement, On Cue
3. Operant Conditioning, Positive Reinforcement, No Cue, Dog Initiates Behavior
4. Operant Conditioning, Positive Punishment (yup, but not what you might think!)
[#2 and #3 have the benefit of creating a Classical conditioning effect: See dog, feel good!]
A1. Classical Counter Conditioning: Easiest by far for a novice owner, because it requires linking the appearance of another dog with food. Dog looks at other dog, food falls from the sky (or falls on the ground, or a toy is presented. I use this sometimes to get dogs started, especially if they are super reactive. The problem with it can be that you need to be sure the dog is linking feeling good with the another dog, not a dog paired with a person, or a person with a yellow jacket, etc. [Again, I talked about this in the seminar, but skipped it in the first version of the blog when Willie was reminded me that if we didn't get going it would get too dark to walk! I was inspired to add it back in by a comment from someone who found it extremely useful....]
2. Operant Conditioning, Positive Reinforcement, On Cue: In this category, a dog is taught that the stimulus of another dog approaching is a g0od thing, and it becomes a stimulus that causes the dog to feel relaxed rather than tense, and usually (at first) to turn away and look at its owner for a food treat or play session. This includes “AutoWatches” and “Where’s the Dog” as described in Feisty Fido or “Look at That” in Leslie McDermitt’s Control Unleashed. “Autowatches” (turn away from the dog and look at your owner) and “Where’s the Dog/Look at That” seem to be polar opposites, but in my experience, they lead to the same response, which is that the dog becomes comfortable with the approach of another dog, and instead of barking and lunging, tends to turn away from it and look toward its owner for reinforcement. This inherently avoids the direct face-to-face confrontation that is forced by 2 dogs approaching on a leash, and has the advantage of making nervous dogs classically conditioned, such that they associate other dogs with feelings of comfort. It also teaches rude dogs (who either are frustrated that they can’t get to another dog or would love to start a stare fight) to engage in an incompatible behavior and get reinforced for it.
3. Operant Conditioning, Positive Reinforcement, No Cue, Dog Initiates Behavior: In this category, which includes what is called CAT and BAT, rather than the owner teaching the dog an incompatible behavior, the dog is exposed to the trigger stimulus and then is reinforced as soon as it performs a behavior voluntarily that is more acceptable. For example, if a dog is barking and lunging, it might be brought to a distance just close enough to elicit a reaction (I would advocate just looking at the other dog, NOT barking and lunging already). The owner/handler stops, and waits for the dog to offer a different behavior, like looking down, or turning its head to the side. As soon as that behavior is offered, either the other dog is taken away (CAT) or the subject dog is taken away (BAT). However in BAT, (Behavior Adjustment Training, Grisha Stewart), unlike in CAT (Constructional Aggression Treatment, Jesus Reosalez-Ruiz & Kelli Snider), the dog is first reinforced with a click for even looking at another dog and then is moved away and given a treat after the click has marked the “looking at”. In the next stage, the dog is exposed at the edge of threshold to another dog and the owner waits for the dog to turn his head away or sniff the ground, etc, marks that with a clicker, moves backward a good distance and treats the dog. In stage 3, the dog is again exposed sub-threshold, but the marker is a verbal ‘yes,’ the distance between dogs is again increased but no treat is given.
Both methods are derivations of John Fisher’s early work, and their greatest strength is that sometimes it is preferable to let the dog choose the behavior, and also to be less focused on the handler or the food/toy and more focused on the other dog. When CAT first started a few years ago, it appeared that the dog was often allowed to go past threshold into a full blown response. The owner/trainer would wait it out, with no one moving (thus no reinforcement from the other dog leaving) until the problem behavior extinguished. From what I have seen, it being a work in progress, its advocates have begun working harder to keep the dog at lower levels of arousal. I think that’s a plus, I never like to see a dog allowed to ‘practice’ a problematic behavior, and it’s tough to wait out some dogs, given that barking can be extremely self reinforcing.
What I like about BAT and the seminar on it last Saturday is that the dog is always kept just at below threshold, and is reinforced both with a secondary reinforcer (the marker) and 2 primary reinforcers at first (food and getting to increase the distance between dogs). The steps are clear and I think that is extremely helpful to owners. I would add though that ‘increasing the distance’ between dogs is not always positive reinforcement, and is most useful if the dog is barking and lunging (or just stiffening) because it is fearful of other dogs. It also requires a relatively sophisticated ability to read dogs, and to respond instantly to the desired behavior (and a decision about what that behavior should be–one thing, or anything that is not “the problem” behavior?).
4. Operant Conditioning, Positive Punishment: The only example of Positive Punishment that I ever use in these cases is Trish King’s “Abandonment Training.” In this scenario, a dog is both on a leash and a long line, with the owner holding the leash as usual and a trainer holding the long line as a safety net. As they approach another dog, if the subject dog barks and lunges, the owner throws the leash onto the dog’s back (tactile cue) and runs like heck the other way. Basically, the dog is ‘deserted’ by the owner, and if it is bothered by that, it stops the behavior very, very quickly. I’ve seen it work beautifully on some dogs, but as Trish advises, this is only for clingy dogs who care deeply about being with their owner. (Can you spell German Shepherd? Forgive me, but do the GSD folks out there agree that GSDs seem to care deeply about being with their humans, more so than many other dogs? I am NOT saying this is appropriate for every GSD, or not for other breeds, don’t get me in trouble by misquoting me!)
I’m a big advocate of having lots of tools in your tool box (thank you Terry Ryan for that phrase!), and personally I believe that being able to use all the methods described above, or some variation on them, is important for anyone who wants to do consults. For private owners, one needs to think about which method fits best with them and their dog. BAT and CAT require, I believe, a sophisticated ability to read a dog. AutoWatches and Where’s the Dog require an owner who likes to train, and can learn the timing required. All methods require setting up wins, in which the dog can be gradually exposed to an increasing level of intensity (dog far away, dog closer; dog standing still, dog moving forward, etc.) and an ability to respond at the right time.
I used AutoWatches and Where’s the Dog with Willie (I’ll post some video of that later this week, right now Willie is reminding me it’s past his dinner time!) and I think they are great for dogs who are easily aroused (Willie would launch just when smelling another dog’s urine. seriously), or tend to want to approach for whatever reason. CAT and BAT are good for people who can read dogs well and have dogs they are sure want the other dog to go away, not get any closer. Argh, I am simplifying way toooo much here, but it’s this is getting too long!
I ended the seminar talking about safety measures, including the Emergency U-Turn and the Emergency Sit/Stay (so that you can put your dog behind you, move forward and distract the other dog). The Sit/Stay requires some serious training, but is amazingly effective if one can put in the time, while the U Turn is so easy that you have to encourage people to practice it. The idea is to get your dog conditioned to feeling good when you say “Oh Wow!” (other cues have been used by my clients, but they are not repeatable here) when you are surprised by a dog, to pivot and move in the other direction so fast that neither one of you even realizes you’ve done it until after it’s over. (There’s lots more on that in Feisty Fido, but Willie is starting to look at me cross-eyed.)
I’d love to hear: who out there has used what method on their own reactive dog? I’d love to start a discussion on this. I’ll post videos of Willie boy and other dogs soon….
MEANWHILE, back on the farm: I am back on the farm! Wheee, good to be home. It’s gorgeous here, and I get one evening of great weather before it gets cloudy and rainy, and boy am I enjoying it. Willie and I got to work sheep (practicing those straight lines, long way to go, but having fun) and we’re about to take a long walk.
Here’s poor Redford this afternoon, our Dorper ram, in his private pasture. We lost the ram lamb we kept in with him, and now the poor guy is by himself. Worse, the ewes are cycling, so they stand beside the gate and tease him. Right now he’s looking at Willie. Redford is one of the gentlest rams I’ve ever had, but I never go into a pen with a ram without a dog, and I never turn my back on one. There is a reason that the noun ‘ram’ was turned into a verb.
Here are some of his lambs, on the hill behind his pasture, back lit by the setting sun.