There are lots of great posts and discussions on social media these days about reinforcement based training. But there are some that absolutely baffle me. Seriously. Trainers who are well known and well established almost 20 years ago. There are new techniques and “systems” of training that seek to simplify or make use of a basic component of this kind of training that I had just assumed every trainer in the positive training community are posting about their surprising new and deep insights about training. Things that I learned and baked into my training and used as a standard part of their training. How did people manage to skip past the basics when learning about reinforcement based training?
It’s been almost two decades. I learned about Mark & Reward training back in 2001 because of a failing relationship with our young Belgian Tervuren. I jumped in with both feet and read everything that I could get my hands on. Positive reinforcement training, behavioural science, Clicker Training, and lots more. Books, magazines, videos (there was no YouTube back then), and Internet discussion forums all provided lots of great information that I still use to this day. But of all the things I have learned in the last 18 years, by far the most important is that basic process of communicating with my dog in training.
It’s not complicated. Just three basic steps; watch my dog until they do the behaviour I want, mark that behaviour in some way so the dog knows that’s what I wanted, and reward the dog for doing that behaviour. Sounds simple, right? Well, yes and no. As I quickly learned, it matters a great deal how you do each of those steps. How I set my dog up to offer me the behaviour I’m looking for matters. How and when I mark a correct behaviour matters. And how I reward my dog matters, perhaps most of all.
Ok. So maybe a little more detail but still not complicated. I also learned some very important “ground rules” for this kind of training. Rules that would help me set the stage for successful training with my dog. Simple things like finding out what my dog really likes – preferred foods, preferred games, etc. – and using the things she liked best in training. Setting easy, achievable goals and making sure my dog was successful much more often than she failed. Keeping training sessions short, distraction-free, and, most importantly, FUN!
I couldn’t pinpoint one particular source for these “ground rules” but they have been as valuable to me as the basic mechanics of the training itself. It seemed like every article or book I read hammered home the same basic messages. Keep it easy, keep it successful, keep it short, keep it fun. I learned that there are good reasons for all of these rules. They all help the learner to learn and the trainer to stay focused on good training.
Watch the dog – If you’re going to mark and reward behaviours, you just have to watch the dog. You can’t see what you need to mark if you’re not watching. But watch long enough and closely enough and you see a lot more than the behaviour you want. Did the dog do it slow or fast? Were they hesitant or enthusiastic? Are they focused or distracted? Do they need more help or have they understood this and need more of a challenge? There is a lot more going on than just the simple “yes or no” or did they “Sit!”
Make the dog successful – Nobody wants to play a game that they lose most of the time. Sure, we all like a challenge but not winning enough is demoralizing and just plain not fun. One of the basics I learned was that keeping my dog successful was more important than getting the behaviour I wanted. What that most often meant was that if my dog wasn’t giving me the behaviour I wanted, I was probably asking for too much too soon. It was on me to make it easier. Sometimes that meant asking for less of the behaviour or just a part of it. Sometimes it meant that we should shift to something my dog knew well so she could have lots of success for a bit. No matter what the strategy, the goal was always to make my dog feel like a winner. Like a competent learner who could overcome any challenge because she succeeds so often.
Make it easy and fun – One of the best pieces of advice I got back when I was just learning is that you should make training a “game.” Training should be fun! It should almost feel like a special treat where we get to play that fun game with lots of new things and lots of rewards. I try to pick times when my dog is alert and wants to engage with me. I plan my training ahead of time so that we can keep our sessions short, fast-paced, and successful. If I’ve done things correctly (at least as I learned this kind of training), my dog should be eager to play the “training game” with me as often as she can. After all, it’s a party just for her!
So what has me so baffled? Well, things like getting your dog’s “consent” to do husbandry behaviours. Why would I teach my dog to offer me a particular behaviour that says “I’m ok with this” when I would know that she is fine with it from the body language that she is showing me? How do I know she is fine? Because I learned to watch my dog from basic training mechanics. I’m supposed to be looking for signs of frustration or confusion or discomfort in addition to the behaviour I want. I should know what my dog looks like when she says “I’m not sure about this”, “No thank you”, and “Oh hell no!” And I want our non-training times to be successful and fun for her as well so that she looks forward to engaging with me any time. So I need to make it easier or do something else when she says “no.” Pushing ahead with training when my dog is uncomfortable is a recipe for failure. Why would pushing ahead with nail trimming or putting on a harness be any different? If I have done my work as a good Mark & Reward trainer, the foundations for my dog accepting her harness or nail trimming should already be there. She knows she can trust that I won’t push her beyond her tolerance and she knows she will be rewarded. It’s how we do our other training together, after all.
Another confusing concept for me is “Start Button Behaviours.” As described in some posts on the Internet, “Start Button Behaviours” are behaviours that the dog will offer to let the handler know that they are ok proceeding with whatever is going on. To me, this looks very much like the “consent” issue. This is another place where just watching my dog for her responses to my actions should be sufficient. Perhaps this is just a clever way of encouraging people to “listen” to their dog’s behaviours. But, to me, this is such a basic part of Mark & Reward training that I don’t know why you would need to call it out.
A third thing floating around that confuses me is the idea of giving the dog a “choice.” Way back when I was first starting to learn Mark and Reward training with a clicker, one of the things that was emphasized over and over was that we are marking the choice our dog makes to do the behaviour. It isn’t enough for the dog to accidentally sit, we should be able to see them actually decide that putting their butt on the floor is what they think is earning the reward. So, as I learned it, the dog ALWAYS has a choice. The process of training is really just finding a way to influence the choices the dog makes by rewarding those things that we want.
Another aspect of the “choice” conversation has been around having multiple different rewards available so that when my dog has earned a reward, I can let her choose what she wants most. It might be a choice between food or a toy. It might be a choice of different kinds of food. It might even be a choice to go out or play a game. What I find confusing about this is that, if I have planned my training correctly, I should already have a reward that I know my dog values in mind before we start. I shouldn’t have to offer a variety of options as rewards if I know what my dog likes because I have observed her reactions to training in the past. It would be foolish for me to bring kibble to a training session where I want to keep my dog engaged when she is learning something new. It just wouldn’t hold her interest. I should know what my dog likes and bring the best stuff to the training party if I want to be successful.
Now, having said all of that, it’s possible that many of these “new” concepts are just trainers trying to make these aspects of training more easily understandable to those who are new to Mark and Reward training. But I never found them difficult in the first place. In fact, I don’t really understand how you can be successful with this kind of training if you don’t learn and start practicing these things at the very beginning of your training. They just seem so integral to good training to me that I can’t imagine I would have had the success I had without them.
I continue to be baffled. Time moves on but the basic fundamentals of reinforcement based training haven’t changed in over 70 years. If we learn to do these fundamentals well and we understand and apply them in the time tested ways, then there doesn’t seem to be a need for new techniques and systems. The old ones still work as brilliantly as they ever did. I wonder if maybe we have forgotten that.