I have a rocksteady faith in building a relationship through lots of positive reinforcement, before you’re even allowed to actually work with pressure. And that’s not just because then you can’t do what you’ve always done quite instinctively (push just a little bit, pull just a little bit, tap just a little bit, hit just a little bit); instead you now have to start thinking about how learning actually works. Learning to work with positive reinforcement teaches you what it really means to “get what you reward”. Only after understanding more of that, you can start working with pressure in a more ethical and yes, more efficient way.
“It is highly likely that there are emotional components to operant conditioning, and that affective states themselves can act as reinforcers or punishers.” is a sentence coming from the research paper ‘Conceptualising the Impact of Arousal and Affective State on Training Outcomes of Operant Conditioning‘. I’m always intrigued by how carefully these things are being worded – “it is highly likely” that horses experience emotions during training, and that influences the training outcome? Duh!
Anyway, in this paper horses are being tested on two tasks: going to a target at a distance, and moving forward under saddle. Targeting works best with positive reinforcement. Moving forward works best with negative reinforcement (leg aids).
Ofcourse, my immediate reaction as a clickertrainer is that this perfectly shows how inefficient food rewards in the saddle are when you don’t have a bridge, and when your horse didn’t learn long ago already that it’s perfectly alright to actively experiment towards the right answer, until he hears the bridge. It’s the bridge, stupid! (*).
But that’s not what this research is about, really. The paper actually shows how the efficiency of a method (from the learning quadrant) changes with the task and the accompanying arousal. Leg aids tap into the flight reflex of a horse, so the arousal is higher and the efficiency as well. On the other hand, when a horse gets too aroused from positive reinforcement, it might hinder learning efficacy as well, depending on the task. Yes, we’ve all seen that, especially when we start a horse with clickertraining.
But, the researcher adds: “there are good reasons to preferentially use positive reinforcement”, because “all operant training approaches will be negatively affected by a negative affective state.” With that she means that horses become optimists or pessimists, with all accompanying long-term hormones and plenty of room for poisoned cues (the lay word for approach-avoidance conflict).
“Arguments that certain operant conditioning approaches are more effective than others may be true in some circumstances yet may fail to take into account the merits of first manipulating arousal levels and affective state to create conditions in an animal that best complement training methods associated with ease of application and promotion of positive affective state and appropriate levels of arousal.”
Or, to use more common words: the most important is that your horse gets to know you as a nice person, before you start taking dance lessons (tango or rock&roll or walzing) when both of you start frowning about who puts which foot where, and if you go there, then where do I go. Because it’s the relationship that makes all the inevitable muddling and jumbling and pulling and pushing and staying behind the movement alright. It’s the relationship that gives plenty of room for making errors with a smile.