Prisoners Dilemma

[Michael]

I've been teaching Kili and Truman to do some tricks in synchronization. For instance to both wave, shake, and recall at same time. It is relatively simple as they are both responding to the same cue.

Synchronized recall is much more difficult though. So far I only dare attempt it when an entire meal back in the cage is the reward. The issue is that sometimes one or the other bird will screw up (not flying, landing in the wrong place, etc). Although one bird did it right, I punish both birds. I do not put them back in their cages and make them both try over again until we get it right (or if I have to give up, then I recall them individually or just carry them back).

It is my hope/goal that by punishing both birds when one screws up that it will create a sort of prisoners dilemma. In the long run they should both learn to participate in the synchronized recall because that is the only way to selfishly achieve the reward.

Anyway, this is my set up and my purpose. I wonder if anyone has any input about this. Do you think the parrots can learn to synchronize their behavior through this strategy or are they so short sighted that no matter how much I try this, they will still only respond to personal reward? Of course in the long run I will develop a separate cue for this and vary whether I do individual or mutual recalls, but for now I want to see how well I can do it. It works out pretty well but it's not guaranteed to work on the first call yet. It is definitely far from a trick routine.

[Mona]

Hi Michael:

I have no idea but I have never had this happen working with my birds. More often, I am dealing with rivalry issues. One bird is trying to be faster than the other for the treats and/or attention.

Model/rival also works...but that isn't synchronized nor is it "Prisoner's dilemna".

Good luck.

Mona

[entrancedbymyGCC]

My gut feeling, and I have less experience than you do, is that punishing both birds in an effort to instill cooperation is unlikely to be understood and is likely to cause confusion... which IMO is the thing about "punishment" that is most likely to cause problems -- when the trainee does not understand the correlation between the aversive and the undesired behavior, the aversive is just unpleasant and may cause a sense of helplessness or anger or fear.

Have you tried shaping the behavior by rewarding more nearly simultaneous performance and gradually stepping up the degree of synchronicity required for reward? I think the thing here that is tricky is how to get the birds to understand what the criterion is. And I guess you could work on it two different ways... one by trying to get one bird to cue off the other, the second by trying to control each bird's response more tightly. The latter puts the onus on the trainer, but seems more straightforward to me.