Some days there is so much information to be shared about a topic that when I start to write, I end up with a great big mess. A jumble of ideas lead to other ideas, and to other ideas and next thing you know, I want to write a book. And having just finished a book with Deb Jones, I’m not ready – not even close – to taking up a new book topic.
Today’s head jumble has been caused by the topic of Motivation. I can’t find a way to start since there is so much to say, so I’m not even going to try. Instead, I’ll break the topic up into tiny pieces over time and try to post over the next few months.
For now, I’ve decided to start at the end and work back towards the beginning.
Here is a video of Lyra working on her positions over two days – there are four options for cues – sit and down (verbal) ; sit and down (hand signal).
http://www.youtube.com/watch?v=uk-i1qv6qXQ (August 27 and 28)
http://www.youtube.com/watch?v=qb4Wjwrfwcs (August 29th)
Criteria varied over the three lessons, but included a combination of verbal commands, hand signals, and increased distance/position from the gate.
Consequences also varied. A minor consequence was a delay of a few seconds before repeating the cue. A major consequence was not allowing her into the pool area for about ten seconds – you can see how she reacts to being left out on the first video.
I have included a random assortment of successes and failures on the videos. I can tell you that she went from about a 10% success rate to about 70% over the three sessions. We also increased distance from about three feet to ten or fifteen feet. In a twenty minute session, she probably spent 80% of that in the pool area and 20% working with me.
For Lyra, “pool” training is about much more than impulse control; it’s about motivation. Lyra cannot open that pool gate. If she wants to enter, she has to go through me – whatever I ask her to do. I will increase my expectations and range of cues slowly and over time. I will also vary where she performs the behavior. I hope that I will have her working in the training yard – for a chance to swim- by the time this swim season is over. That gives me about six weeks. If we don’t make it then that’s fine too. She’s a young dog and her responses to training will set the pace.
Until last week, I had nothing; nothing that mattered so much to Lyra that she would keep trying. Withholding classic toy and food rewards didn’t matter to her, which makes training a very delicate thing indeed. Now I have something that she cares about, and I can control access to that thing. That is the definition of a motivator – something your dog wants and that they will work to get.
This is an example of using a ‘life’ reward rather than a “classic’ reward. I won’t call it pre-mack, because ALL R+ dog training is pre-mack, so it adds nothing to the conversation. The pre-mack principle simply says that if an organism has to do a behavior to get something they want, it makes the behavior more likely to happen. So…if you have to eat your veggies to get ice cream, it makes the act of eating veggies more likely. If you have to sit to get a piece of liver, it makes sitting more likely. If you have to play tug to get a cookie, it makes tugging more likely. Pre-mack is also called grandmother’s law: If you eat your veggies you get a cookie. It is worth noting, however, that there is nothing in the pre-mack principle that says the dog (or child) will be more likely to ENJOY the behavior that they must perform to earn the preferred motivator. Indeed the opposite seems to be more likely in studies of children who are bribed to to eat. (there is a message in there but I wont’ get into it right now)
Now I have something that matters so much that withholding the reward drives frustration to the point that I’m able to insist on effort. Lyra is trying hard, probably for the first time in her life. Most people get that easily by withholding a cookie or a toy, but Lyra doesn’t care enough about cookies or toys to tolerate any frustration over their loss.
I can teach most behaviors even with a low value motivator, which is how Lyra has learned most of her cues. But driving precision – high level precision – requires a dog that will stay in the game even when they have to put out real effort, and I may have a way to get that now.
Will this get us in the ring? No. Getting behaviors and ring readiness are different concepts, though they are linked. I’m not there yet; first I need a seriously motivated dog that is driven to learn what I want to teach. I’ll deal with the ring readiness issues later.
I hope to do several posts on the topic of Motivation. It is deeply intertwined with many other concepts that appear to be misunderstood, so be patient with me. Let’s see what I can come up with over the next several months. And anyway, there’s some ‘on the job’ learning going on here – Lyra’s job is to force me to look in new directions, and she’s proving to be quite good at that.
If you’re past the stage of teaching behaviors and want to prepare your dog to compete, then check out my new online class starting in October. It’s called “Bridging the Gap; Reducing reinforcers, Proofing and Generalization”. And it’s about….Reducing reinforcers, proofing and generalization:). I have so much to say about this topic that it’s a two part class – sign up and you’ll take the class from October to February – with a two week break in the middle. You get to pay twice as much as for the other classes at the academy because it will run for two sessions. If that is something you need, then check it out here: http://www.fenzidogsportsacademy.com/index.php/courses/203
here’s a simple explanation of Premack: http://en.wikipedia.org/wiki/Premack_principle
Hi Denise, I just wanted to say that how much I enjoyed your seminars at MasterPeace in the spring, and your on-line courses. I wanted to comment on “motivation” in general with my dog. I was having some trouble when you came east in the spring, and had been struggling to find a way to make my dog happy in the ring. Between you and my great teachers at MasterPeace, we’re back on track, and I have my happy, bouncy partner back at trials. But today, the best thing of all happened. I went to leave the ring twice during different parts of my lesson, and he stopped at the ring gate, and said clearly, “I want to stay in here and play the game!” That was worth WAY more than any ribbon or title to me!
wow; that’s awesome!
Lyra and my GSD are VERY similar with regards to simply not being INTENSELY motivated by either food or toys of any variety. And like you, I found one thing that finally does make her eyes bug out of her head….for Lyra, it’s the pool; for my dog, it’s the hose. She will work endlessly in high drive for a chance to run to the hose and bite the water stream.
I hope I can learn from you where I made errors, because I tried to keep that intensity when the hose isn’t available and the results are only so-so. I built strong secondary reinforcers that came to predict the hose (i.e. recall leads to tugging which then leads to release to the hose). Then I put the hose on an intermittent schedule (sometimes the tugging is followed by the hose…..sometimes it’s not and we go back to training), so she would persist when she didn’t get the hose every time. At least my reinforcer is a bit more portable than yours, lol, so I would take the hose to show n go’s so she would learn that it IS available even when we aren’t in the back yard. It helps, but the tension and intensity in her simply isn’t the same as it is when the hose is readily available. I’ll be watching with great interest to see how you “transfer the value”, as Susan Garrett likes to say, from your pool reinforcer.
What you’re talking about is trial preparation and I’m only using the pool to teach higher quality behaviors. The pool won’t get me around the challenges of trial competition – I’ll have to work on that separately. Then again, that challenge exists regardless of a dog’s preferred motivator – food and toys also do not go into the obedience ring, and seven minutes over many trials is a long time to convince a dog to hang in there.
If your preferred event is agility you’ll be in much better shape because the activity is more likely to be self reinforcing for many dogs, and the time to the reward is extremely short – usually seconds at the end of the run. In obedience, its is likely to be a minute or so from the end of the final exercise to the reward outside the ring.
I don’t believe in transfer of value – only that you can create a delayed or intermittent reward schedule or that some dogs discover that they really do enjoy the activity itself (jumping, retrieving, tugging, eating vegetables, etc.). I’ve seen dogs force fetched who absolutely love to retrieve – that’s surely not transfer of value; it’s a dog who “woke up” to the fun of the activity. I see that with my own dogs when I shape the retrieve; at first they take the dumbbell to earn the cookie. And then one day I throw that dumbbell and suddenly they realize…it’s an object to fetch! And that’s when they love it – food no longer required. There’s no transfer of value at all in that scenario.
I DO believe in Conditioned Emotional Responses (CER), so we probably want to do everything in our power to make sure that dogs find training time a whole lot of fun and minimally stressful – because we take that CER into competition with us.
It’s true there’s nothing in Premack that says that the subject will come to enjoy the less-likely behaviour more, but I’m wondering if you think that will never happen? It just seems to me that by constantly pairing something ‘less good’ with something better, you would get at least some ‘transfer of value’ from one activity to the other. How much probably depends on the subject and the two activities you are using. I’d be interested in learning more about the children who were bribed to eat.
I believe in a Conditioned Emotional Response (CER) which I mention below. And for sure, some dogs can learn to enjoy an activity because it is paired with another positive activity. BUT there are risks to pairing so I use it as a very last resort – too complicated a topic to take up here.
“there are risks to pairing so I use it as a very last resort”. Could you expand on that thought – I don’t understand. Are you speaking specifically about aversive activities being paired with appetitive ones (i.e. forced retrieve paired with ball play) or do you mean there are risks with all paired activites, even appetitive ones?
It’s the “I use it (pairings) as a very last resort” that has me confused. If you don’t want to explain yourself, could you at least point us to a reference, like a chapter in a textbook, or a dvd lecture where this thought is discussed. It doesn’t seem to make sense in light of your own comment that all +R training follows the Premack principle, which are pairings.
There is no research on it because scientists don’t study it. I’ll hit it in a blog. The issue is one of intrinsic vs. extrinsic motivation. If you reward something that is already gratifying, you REMOVE the value of the activity itself and it’s hard to get back because it becomes a deprivation issue (what they REALLY want) as opposed to a motivation issue (what you offered that’s not highly motivating).
That does not apply if what you are working with is not (or rarely) motivating.
When you pair tug with food you are asking for trouble because tug is normally intrinsically rewarding. If you pair food with sit you’ll be safe, because the opportunity to “sit” is rarely motivating to a dog.
That’s the best I can do here. Motivation is a complicated topic. Really really complicated. And deep. Which I love, but not in a blog format.
best place to discuss this is really on FB – go to R+ comp obedience list. As a rule I really don’t respond to comments on my blog – it’s a time thing. But when I have time I respond on FB.
The mention of bribing children reminded me that, despite how many of us postitive trainers would insist that the way we train is *not* bribing, Premack is all about bribing – the reward is out there, plain as day to see, before the dog performs the behaviour. I LOVE Premack, but wouldn’t hesitate to tell someone that the way I train is not bribing. Oops! 😉
It CAN be bribing, for sure, but it doesn’t have to be, IMO. Premack can pertain to scenarios like asking the dog to do a stand for exam with a stranger (which they find stressful) and following it with the opportunity to heel (which is a highly reinforced, pleasant, well-known behavior for the dog so it’s much preferred by the dog). It doesn’t have to be an obvious food bowl on the counter that we let our dogs see (ok fluffy, see what you’ll get if you do what I ask”) or a ball in our hand that predicts play if they cooperate.
all pairing has potential risks – too big a topic to explain here.
I agree, Lee. I was thinking about it last night and realized I only tend to label something I’m doing as Premack when I’m using it in the bribing way. Latest example is me asking my Vizsla to take a dead bird in her mouth for the opportunity to hunt for more live ones. The field is right there in front of her when I’m doing this.
One last question (promise!). Why do you say ‘okay’ and ‘good girl’ each time she doesn’t respond to your cue?
Hi Denise, I enjoyed this post. I’m also curious why you are releasing Lyra when she doesn’t respond to your cues. I will usually give a different cue to get some behavior and then release from that, but now I’m wondering if I should be releasing even if my boy doesn’t perform the indicated behavior? Thanks!
She didn’t do it so I release her to start over or try something else. I have to release her or she may think she can still get it right. She can’t – the opportunity is over until I give a new cue.