The Power of Cues – Poisoned Cues
In the previous post about cues, I talked about how they are all around us and that a cue is simply something that is an accurate predictor of the outcome of a particular behaviour in the presence of that cue. So when my hand motions down to my horses that ‘cue’ is an accurate predictor that if they head lower, they will get a click and reinforcer.
Likewise, because cues work in two directions (that is the topic for another day), a click is an accurate predictor that I will go in to food delivery mode.
With a history of reinforcement in the presence of that cue, the horses learn that the outcome of the behaviour will be reinforced. In other words, they have learned the cue is predictable through repeated experience of what behaviour will earn them a click in the presence of that cue.
Karen Pryor first came up with the term ‘poisoned cue’. She believed a number of years ago that this was a concept that she was seeing in some situations. For example, a dog would start to respond to a cue but then maybe get a leash pop if they did not do it correctly or do it fast enough. If they did respond to the cue fast enough then they got a click and reinforcer. The dog would then start to show avoidance behaviour; “if I don’t even attempt to do the behaviour then I can’t earn a leash pop”. When you hear those words “can I see you in my office”, do you;
- a) rush right in there with a big grin on your face;
- b) avoid it and chat to the people you work with asking if they know what it might be about;
- c) nip to the bathroom first then meander slowly back in the direction of your bosses office.
Generally we avoid having to find out what it is our boss wants to see us about. We are so desperate for it to be good news and that makes us want to rush in there, but on the other hand it might be bad news so do we really want to know, or should we just get it over with as quick as possible. If we don’t let him tell us at all (just not go in to his office) then it can’t be bad news, right ?!
Unless we have never had our boss give us a ticking off for something we generally avoid going to their office. All of the thoughts of ‘it can’t be that bad’, ‘I’d best just get it over with’, ‘but it might be awful’ cause a lot of anxiety. We start to show avoidance behaviours and we may even start to show displacement behaviours (a normal behaviour seen out of context…going to the bathroom when really we could have waited a while, or an hour, or biting nails and so on).
Our animals are no different. If, in the past, a cue has sometimes meant the outcome of their behaviour resulted in a reinforcer and other times in a punisher they will behave in the same way we do. They may show avoidance behaviours, displacement activity (e.g. yawning while training) and may even just stop working with us at all. They may even look like they have gone to sleep (often conditioned suppression); a phenomenon I often see and hear about when horses are having feet shod.
We really need to be aware of how our reactions to our horses might create poisoned cues, and we should avoid creating them where possible.
Alexandra Kurland documented a wonderful interview with Jesus Rozalez-Ruiz (Associate Professor of Behaviour Analysis, North Texas University) on DVD. Jesus and his research students had researched Karen Pryors theory on poisoned cues and in the DVD Jesus discusses the results. He also discusses the learning theory behind poisoned cues and there is also some very powerful video of the research in progress presented. Poisoned cues is a large topic and so, to find out more about them, I will direct you to Alexandra’s wonderful DVD.