After my last post on clicker training and SATS, (Equine Cognition- The Final Frontier), I received an awful lot of feedback. Some of this was good, some undecided and some, well, rather negative!! Overall I'm delighted that both my own post and that of Jane Holderness, (Magic's Own Little World), provoked so much discussion. I believe that in order for our knowledge to evolve, such exchange of ideas and opinions is vital. However, there are a few issues I would like to clarify and a couple of areas I would like to expand upon.
Firstly, it is not my intention to criticise clicker training per se. Rather, I am attempting to highlight some of the issues that can be encountered if clicker training is not used with care. One treat one click training when used with horses can achieve amazing results. These results are, however, subject to a number of parameters being adhered to.
The trainer must have a keen eye for detail, seeing both the smallest try and also remaining aware at all times of subtle signs that might indicate that the horse is becoming over threshold, (whether that be fear/anxiety/tension/arousal etc). The trainer must know exactly when to switch to a variable schedule of reinforcement and also when to fade out the click altogether. The trainer must know how best to set their horse up for success and be able to adapt readily if things don't go according to plan.
Arguably, these are attributes that all trainers should possess. In practise, we all live in the real world and many of us fall below these standards. I for one try the best I can for my horses but am of average ability and have like many people below average funds to pay for the coaching that might help me to progress further. So in the real world, this ability gap in equine one treat one click training can result in horses becoming frustrated, anxious, tense and over aroused. This highlights the real issue with equine clicker training. It is not intrinsically flawed but in it's interface with us less than perfect human trainers, the science behind it can easily become distorted. I believe that a successful training method should have some tolerance to error built in to accommodate a whole variety of abilities and resources. This is why I started to look for ways to build upon the lessons I had learned through clicker training, to compensate for my own shortcomings in a way that would enhance my training partnership with my horses.
This leads me to another point of clarification. Many people have begun referring to SATS as a method or claiming that SATS perceives itself as a method. I would disagree on both counts. SATS does not reinvent the wheel or start with a fresh sheet of paper. Rather, it uses a variety of tools that are out there, some more accepted than others, and assembles them in a way that makes the end result more user friendly. I think that the difficulty that some people are having in understanding exactly what SATS is derives from the fact that SATS, unlike many other ways of training, does not have a vertical hierarchy. By this I mean that it's structure does not go from Lesson One at the bottom, progressing through the lessons until the top of the ladder. Rather, the structure is horizontal with many components being taught simultaneously to give the animal as much information as possible and the trainer multiple tools with which to do this. So why does this make SATS user friendly? Surely vertical is easier than horizontal?!! Well, here's some reasons why I believe that SATS is suitable for all abilities and has some tolerance to error built in.
The first feature is the use of the Intermediate Bridge, (IB). When using a terminal bridge, (TB) exclusively, timing is everything. It is easy to miss the exact moment that we should be clicking causing confusion and even frustration in our horses. By using the IB, it becomes clear to the horse exactly when what they are doing is getting closer to the answer. Rather than having to experiment a little to get to the answer, we can give them real time feedback on how they're doing. This then reduces the possibility of uncertainty in the horse and gives us human trainers a little leeway if we miss the exact timing on the TB because the horse knows that they are in approximately the correct place.
The thing that I'm noticing with great surprise is how unbelievably reinforcing the IB becomes to the horse. In fact, I'm now hardly giving food treats and the TB seems almost inconsequential to them. I suspect this is because the IB gives so much confidence to the learner, (horse), that they are doing the right thing that they really begin to relish the actual process of learning. The IB therefore motivates them, informs them and keeps them in the game. I suppose this is hardly surprising given that we all like encouragement to learn. Imagine if every time someone asked you a question they praised your efforts until you got the answer, you would be motivated to try something else, perhaps even something more challenging.
As a trainer, another surprising effect of the IB has been to help me see the smaller steps and tries on the part of the horses. When only using a TB, although we are shaping approximations, I believe that the TB keeps the link to the finished behaviour fresh in our minds. When using the IB, the whole focus becomes the try and the end result seems to fade. I know every day I'm becoming more able to see the smallest steps because of using the IB. The IB therefore enables things to be trained much more quickly both because the horse is more motivated and informed and the trainer more available and observant.
The IB is further enhanced by the idea of 'naming and explaining' the training process to the horse. So not only are we giving feedback in the form of the IB, we are naming for the horse exactly what is happening at any given time. So if I wanted to teach my horse to lift up her left leg, I would firstly name her front left leg for her. I would then using targets teach her 'up' and 'down'. I would then ask her to lift UP her FRONT LEFT LEG. I have put in capitals the words that are meaningful to the horse because they are words that I have 'mapped' for her. After making the request, I would watch for any sign of a shift of weight off the front left leg, giving the IB and then TB. If the leg remained down, I would merely tell the horse that their LEFT FRONT LEG is DOWN and not bridge.
This leads me nicely in to my next point which although has been made elsewhere is worth emphasising. In SATS, there are no wrong answers, just information!!! In the example of lifting the leg, although what I am seeing is eventually for the leg to lift, the leg remaining down isn't wrong, it's just not up at the moment! I can take the opportunity to explain to the horse what DOWN feels like which will save me a job later. Most importantly, instead of the horse potentially becoming demoralised as she seeks the answer, she is always learning about something. The IB then helps to point her in the direction relevant to my question and the TB confirms this. It is easy to see that compared to the use of just a TB, this equips the horse with masses of information and therefore confidence and motivation.
I'm also discovering that naming things has a purpose beyond the basic training scenario. This is in the reduction of anxiety and fear when encountering bothersome stimuli. We have to hack past a house nearby that has a very loud dog flap! When we go past, the flap springs open, slams closed and then two dogs scramble across some very loud gravel before barking. Our ponies began to become worried about going past the house, beginning to show signs of anxiety progressively greater distances from the house. Even Dougal, (who wouldn't be disturbed by a nuclear bomb going off underneath him), began to show mild signs of disdain for these two small creatures who insisted on disturbing his peace and quiet!). On a serious note, this was becoming a safety issue as the house adjoined a road on which we had to pass.
So, we embarked upon a programme of counter conditioning and desensitisation on the ground. We retreated to a point where the ponies could hear the dog flap and dogs but remain below threshold. At this point, we encouraged them to be calm and treated them before retreating further away. We would then get a bit nearer and repeat the process. The ponies improved greatly but the trouble was that the dogs would come out of the their flap at different times. Sometimes it would be before we got to the door, sometimes whilst we were passing it, sometimes just after we had passed and sometimes they would wait until we thought the coast was clear before running out! As a consequence, the ponies remained a little on edge as none of us were sure when the menaces would reveal themselves!
I was riding Wish past there the other day and decided to experiment with the idea of 'name and explain'. She is familiar with the names of DOG, DOOR and GRAVEL, (don't ask- her mind map has taken some odd detours!!). So, as we approached and before she became at all tense, I told her that we were about to come to the DOOR where the DOGS would come out over the GRAVEL and would then bark, (I then made woof, woof noises). Wish stopped for a moment and I asked her if she was ready. She turned and bumped my knee so I asked her forward. As we got nearer I said the same thing again and told her it was about to happen. She strode on. The dogs did come out and perform their now boring ritual and Wish remained cool as a cucumber! Not really believing that it could be that simple, I engineered a route back past the house on the way home. I repeated the same experiment with exactly the same results. Since then I have tried similar things with the same effect.
This is where I believe we begin to voyage into the mysteries of cognition. Even if most of 'name and explain' is attributed to classical conditioning, (which in itself is unproblematic if it is helpful to the horse), to me this shows an ability to transfer concepts to different contexts or to generalise. The take home point here is that this generalisation is occuring after very few trials. I haven't repetitively drilled Wish on DOG, DOOR and GRAVEL, naming them for her three times at the most in a very different context. However, she seems able to understand them sufficiently to be able to apply them in a way that reduced her anxiety due to this understanding of them.
I suppose it is analogous to a small child walking next to a busy road whilst clinging to their parent's hand. If a large lorry should come rattling past and the parent gives them a reassuring squeeze, the child will probably feel better but will always wonder what that monsterous metal thing was! The next time one came past, the child might be likely to experience the anxiety that comes from the unknown. The parent could have given them a squeeze and told the child that it was a lorry that carries goods for us to use or to eat and that there are many of them on our roads. When next confronted with a lorry, the child could then have used this information to understand what the lorry is and not feel the same levels of anxiety.
It is in relation to anxiety or rather calmness that I would also like to comment. I am aware that there have been many claims about clicking for calmness, asserting that it is both possible and impossible. Some people say it isn't possible to train emotional states just their physical manifestations, some claim otherwise. I am currently reading a brilliant book by Stephen Biddulph called 'Raising Girls'. (I can't recommend this book highly enough for those of you reading this who have daughters/grand daughters/nieces. It highlights the tragic erosion of girlhood and the all too early sexualisation of our girls whilst giving useful information on how to empower daughters into womanhood). In the book he talks about mothers as role models, specifically in relation to remaining calm through the storms that life throws at us. Stephen refers to calmness as a skill, something that can be learned, practised and passed on to others. Through developing an awareness of breathing, groundedness and exercises to rid our body of tension, we can make calmness a self reinforcing habit. I think this sums up to me how SATS addresses calmness. By identifying our horses' body parts, by helping them to understand how they feel when they are both alert and relaxed, we can help them to understand what calm feels like. Further, we can help them practise achieving the state until it becomes reinforcing for them too.
My final point relates to semantics. I believe that it is important to try to understand things scientifically in order to provide some objective framework against which to analyse our training interactions. However, not only does science not have all the answers, whether something is X or Y will not influence the ultimate training outcome and remains both unimportant and unknown by our horses. The danger with feeling beholden to science is that it may stifle our interactions with our horses by robbing us of our instincts and creativity. For those people reading this who use the science to their advantage and are content with their training relationships with their horses, I have great respect for you and I will seek to learn more from you. For those of you who like me either encountered issues with their training or merely who feel that their training has become rather unidimensional, I would encourage you to liberate yourselves to go out and experiment!