Behavior chain

How it works

Most useful behaviors are chains once they are looked at closely. A retrieve is run out, pick up, turn, carry back and release, and a dog performs it as one smooth action only because each link has been learned and connected. The completion of one step works in two ways at once: it confirms that step was right, and it acts as the cue for the step that follows. The whole sequence is held together by the reward at the end, so a chain that stops paying off tends to break down from its weakest link. Chains can be taught from the first step forward, or from the last step backward, which is called backchaining. Backchaining has the advantage that the dog always finishes on the step they know best, nearest the reward.

What it looks like

To teach a dog to fetch their leash from a hook and bring it over, a backchained plan starts at the end: the dog is paid for dropping the leash into a hand. Next, for carrying it a step before dropping it. Then for picking it up off the floor, and last for pulling it off the hook. Each new step is followed by steps the dog already knows well.

Commonly confused with

Shaping

Shaping builds one behavior by rewarding closer and closer versions of it. Chaining links several separate, already-learned behaviors into a sequence. A trainer may shape each link and then chain them together.

Related terms