last entry, I cautioned people who are playing the Look at That (LAT) game to make sure their dog is sub-threshold. If not, you might end up with a dog like mine, who has learned this annoying behavior chain of barking and lunging, then returning to me for a treat.
It’s been seven months or so since we figured out that’s what Maisy was doing, and while we’ve made progress, she still does it. Trust me- even though it’s a bit slower in the beginning, in the long run, it will be easier and faster to do it right. Still, I made the mistake, and I know I’m not alone in that. Since LAT is such a great game for reactive dogs, I wanted to share with others how I’ve been fixing a LAT gone wrong.
First, make sure you’re working sub-threshold. You do not want your dog to continue to practice the behavior. If your dog barks and lunges, regardless of whether it’s due to anxiety, or as part of a behavior chain, move away. There’s no reason to add to the problem. Do not give a treat, and resist the urge to speak. Simply move away.
Next, if you’ve attached a cue to the behavior, change it. The cue you’re using now has come to mean “look at that dog, bark and lunge at it, and return to me.” That’s what your dog thinks you want when you give the cue, so don’t give it. Use a new word and start over. Although you could try to teach the behavior with a trigger, I think it makes more sense to start with a neutral object, like a stuffed animal hidden behind your back. Using a neutral object should change the picture enough to prevent the barking and lunging.
Once your dog understands the new cue, you can start using it with triggers, but again, make sure you’re sub-threshold. Start with lots of distance in order to prevent any lunging and ruining the new cue, too. Gradually reduce the distance, but don’t rush things. You want your new cue to mean look, not bark-and-lunge, and there's no need to hurry this process along.
Since the beauty of LAT is that your dog learns that seeing a trigger is a cue to look back at you, changing the name of the behavior alone won’t fix the problem. You’re going to need to implement an extinction plan. Extinction happens when a learned behavior is no longer reinforced. If a behavior no longer results in the expected reward, the behavior is unlikely to continue. We’re going to make the behavior of look-bark-lunge-return quit paying.
To do this, continue to play LAT as normal. When the dog looks, click and treat. I like to increase the reinforcement by using higher value treats or by using jackpots. I have no idea if it makes a difference, but I do it anyway. When the dog looks and barks or lunges, do not click, and do not treat. Take a few steps backwards in order to increase the distance (and hopefully get your dog sub-threshold in the event the behavior was motivated by anxiety instead of the result of a learned behavior chain), but otherwise do not react. Do not look at your dog. Do not talk to your dog. If you were talking to someone, continue the conversation without pause. Carry on as if nothing happened.
Keep in mind that when you use an extinction program, you are likely to get an extinction burst, which is a temporary increase in the frequency, duration or intensity of the behavior. If you’re not expecting it, it will look like the behavior is getting worse, but in reality, this is how you know it’s working. Extinction bursts typically happen quickly (although not every animal will go through one) and the length of time they last is generally related to how long the behavior has paid off in the past.
The downfall of extinction programs is two-fold. First, there is the possibility of spontaneous recovery, where the dog tries the behavior again in the future. This can happen long after you thought you got rid of that pesky behavior! The good news is that even when a behavior spontaneously reoccurs, the intensity and duration is typically lower, and it will continue to get lower each time.
The other downfall of extinction programs is that they can be environmentally dependent. That is, while the dog may understand that barking, lunging, and looking back to you no longer works with bikes, they may need to go through the entire process again for big black dogs. And balloons. And children. And every other trigger they have.
For these reasons, you really are better off teaching LAT correctly in the first place! I know I sound like a broken record, but when you’re working with a reactive dog, it’s important to remain sub-threshold. Slow and steady really does win the race.
Alright, now it’s your turn. Someone please tell me that I’m not the only one that’s screwed up this behavior! For those of you who have, how did you fix it? Did you use an approach similar to mine? Even if you haven’t made the same mistake I have, can you think of any other suggestions to fix the error? I know I have some really smart people reading, and I’d love to hear what you would do!