Teaching your dog to make eye contact on request is extremely useful because it is an easy behaviour that you can reward with any of your reward options (food, interaction or activities – see Guide & Control for an explanation). It can be variously described as teaching a “look” or “watch me”, or simply paying attention to you. Use it as a preamble to any other request or as an alternative behaviour to anything you would rather your dog didn’t do. I use it extensively in Dogs That Bite & Fight, teaching dogs that aggression is not the solution: “Watch me and I will solve your problems so you don’t have to”.
Because Ted is new to us we need him to pay attention to his new name, so in his case I’ll use his name as a prompt to look at me. Basically “Ted” means “Look at me because something good’s going to happen.” But if your dog already knows his name you might want to use another request – in Bite & Fight I explain why I use “Watsis” as a recall. But we are not at that stage just yet.
I’ve already introduced the concept to Ted sitting on the floor in the living room. I simply place a tiny treat on the palm of my hand and hold out my hand. If he tries to take the treat I close my hand with a gentle, “Ah” and then open it again. When he stopped trying to take the treat I offered it into his mouth with an “OK”. So I’ve used negative punishment for the behaviour of “trying to take the treat” by withholding the treat, followed by positive reinforcement of the behaviour of “not trying to take the treat”. Ted quickly learned not to take the treat and we moved on to the next stage.
When he sat back and didn’t take the treat I didn’t immediately reward him (reinforce that behaviour) but neither did I fold my hand around it. I didn’t do anything, just held it there. Eventually, because he didn’t know what to do, he looked at me – just a quick glance – but I “OK-ed” and treated him, reinforcing the behaviour.
He quickly began to look at me more often and got the idea that all he had to do to get the treat was look at my face. So I upped the stakes again and didn’t reward him for a quick look, but only a longer eye to eye contact. And that’s where we are now.
I know this has taken a bit of explaining, but it took about fifteen tiny treats in about three minutes last night, and unfortunately I didn’t video it. I’m fairly confident that he will have retained the behaviour in context, so if I try to show you it in the sitting room again, you will see Ted sitting looking at me for five seconds or so interspersed with me rewarding him for it.
So I’ve taken a bit of a leap too soon and gone outside so it is out of context again for him, in the hope that he wouldn’t be quite so good – and he isn’t. That’s to be expected because I’m generalising before I’ve fixed the behaviour. I’ll go back and train to sustain the eye contact for longer – maybe ten to fifteen seconds – and also put it on the cue of his name (because I haven’t done that yet). The correct sequence would be:
- Train the behaviour
- Place it on cue
- Generalise to different circumstances
But I’ve jumped forward a bit so I can show you him not doing it quite so well. I won’t have broken anything and can easily go back again.
In the video you can see that my timing isn’t always perfect, but who is? This kind of shaping can be quite forgiving in that if I inadvertently reward the wrong thing I haven’t completely messed-up, I just need to get more right that wrong – so don’t worry if you get it wrong occasionally.
Oh, and you can tell it’s “generalisation” because I enlisted Milly’s help. Honest, I trained her to do that 🙂
You can watch the video here:
As the meercat would say, “Simples!”