Note: I have been working on this paper for 18 months. Today when I published it, I was unaware that Dr. Perone was the head of a recent task force that concluded that contingent electric skin shock of of a population that could include people with developmental disabilities, emotional disorders, and autistic-like behaviors could be part of an “ethically sound treatment program.” It casts his paper in a different light. I’m leaving my writeup published for now because I think we need these answers to what is an often quoted paper. Please don’t consider it in support of Dr. Perone in any way.
“The Negative Effects of Positive Reinforcement” by Dr. Michael Perone is a scholarly article some trainers like to use to muddy the waters about positive reinforcement training. They throw out Dr. Perone’s article title like a bogeyman and use it to defend aversive methods in dog training. That usually indicates they haven’t read it. It’s a thoughtful article and has some interesting things to consider, but it doesn’t say what they seem to think it does. Not even close.
Thank you to the readers who helped me with this paper. Any mistakes are my own.
Some terminology in behavior science is notoriously hard to get one’s head around. One of these terms is negative reinforcement. Not only is this learning process itself a challenge to understand, but the terminology itself is counterintuitive. Behavior scientists specialize in training, teaching, and learning, so naturally, if a term from their own field trips people up, they are going to analyze the problem. The terminology for negative reinforcement has already been changed once, in the 1950s to early 1960s. There has been more discussion since then. This post is about the article that started the more recent discussion, and how it is often misunderstood in the animal training community.
When I crossed over to training with positive reinforcement, I had no idea how much my behavior and even my belief system would need to change. I had to question my faith in some long-held cultural assumptions and learn to rely on scientific observation and analysis.
Crossing over was a lengthy process for me, and even now, after more than 10 years, I occasionally fall back onto old assumptions and behaviors. I wonder sometimes if I am the only one so vulnerable to cultural programming. But a quick look around social media says no, I’m probably not.
Once upon a time, there was a girl who decided to teach her dog some tricks. She figured out that if she gave her dog something he liked after he did something she liked, he was liable to do the thing again. So she taught him some simple tricks using food and play as reinforcement.
As she went along, her dog started finding playing training games lots of fun in and of themselves. But she still used food and play. He liked earning his “pay” and she liked giving it to him. She didn’t see any reason to stop.
This girl was unusual in that she didn’t try to tell her dog what to do in words. She realized what is not obvious to so many of us: he didn’t speak English. Things worked out just fine because he could generally discern from context and her gestures what she wanted to work on.
She used a little platform to teach him to pivot in a circle. He would put his front feet on the platform and walk around with his back feet and rotate. He got good at this and soon could spin in both directions. As soon as he saw the platform he would run over to it and start to pivot, although she could ask him to stop with a hand signal.
What do you see in this professional photo of Summer on an agility A-frame in a competition?
She’s so pretty in that photo, and running nicely, but you know what? She wasn’t happy.
Here are a couple more photos from that same trial.
Summer was not miserable. She was responsive and doing what I was asking her to do. (What a good girl!) But she was stressed. And she was not joyful. I can tell it from her face, which was drawn, even a bit grim. For some dogs, that particular look might just be focus. But for her, it shows unpleasant stress. Can you tell?
How Was She Trained?
Summer’s agility behaviors were trained with positive reinforcement. She was never forced onto the equipment, but was taught gradually and gently. She wasn’t scared of it. She was physically confident and generally enjoyed the activity. So why did she look grim in these trial photos? I can identify three reasons.
She was undertrained. She just didn’t have that much experience yet and wasn’t solid. The behaviors weren’t “can do it in her sleep” fluent.
She was stressed in the trial environment. It was outdoors, there were lots of dogs, it was muddy and rainy, and she didn’t have enough experience in challenging public situations.
This is actually the big one. I had trained her with positive reinforcement, but I had not sought out and used reinforcers that she was wildly crazy about.
Fixing the last one sent us well on our way to fixing all three.
Finding the Joy in Agility
At the time these photos were taken in 2008, I had recently found a new teacher. She helped me realize that even though Summer could perform most of the behaviors, and had even had some qualifying runs, I was trialing her too early.
So, we worked on Summer’s and my agility behaviors. We worked on her distractibility, especially her penchant for hunting turtles. We worked on my handling, so I could be consistent and clear. She showed me that almost anything Summer did (except to run after a turtle) was because I had cued it with my body. We cut down and practically eliminated the times when Summer might just run off after a bad cue of mine, both because my cues got better, and because Summer found it worthwhile to stick around even when I messed up.
At my teacher’s encouragement, I found very high-value treats that turned Summer’s attention on high. And we used a novel reinforcer—playing in the spray of a garden hose—as a reinforcer for a whole sequence.* The water play not only upped her excitement about agility in general, it was also great for proofing her performance. She learned that if she ran straight to the hose rather than following my signals, no water came out. But finish the sequence correctly, and there was a party with the hose. She loved it!
Transfer of Value
In my last blog post I described how I became a conditioned reinforcer to my dogs over the years through regular association with food and fun. The same thing happened with agility. All those good feelings associated with the high-value goodies, the fun, and the hose bled right over into agility behaviors.
Three years later, we competed again. We had practiced going to new environments. The fun of agility was so strong, and our behaviors were that much more fluent, that this is how she now looked in competition.
Summer came to love agility. She sprang from the start line when released. She ran fast and happy. She was an unlikely agility dog with her penchant for turtles and other prey. But she not only got good at it, she loved it. And I loved doing it with her. Even after I got Zani, who was young, physically apt and very responsive—running with Summer was always like coming home.
I thought about calling this post “Going Beyond Positive Reinforcement,” but I decided that was inaccurate. I didn’t need to go beyond it. The difference was just better positive reinforcement training. More thorough, more general, more thoughtful. And the result was joy.
If you want to see just how joyful, watch the following video. The first clip is from 2012, at a trial. Even though it was late in the day and I made some clumsy errors, she ran happy! The comparison in her demeanor from the previous competition is striking. It is followed by the best example of her speed I have on film: a run we did in 2014 at an agility field (with distractions). Finally, I show some messing around we did at home in 2016, just to share how delighted we were to be playing with one another. She was 10 years old then, and that winter was the last time shared the joy of agility together. (She passed away in August, 2017.)
My teacher, and other great trainers who have influenced me, have taught me to set the bar (ha-ha) high. It’s not enough that a dog can do the behaviors. It’s not enough that they can qualify. It’s not enough that they can get ribbons. It’s not enough that they are happy to get their treat at the end of the run or get to go explore the barn area at the fairgrounds.
What’s enough is getting the joy.
*If you allow your dog to play in water, especially with a hose, make sure she doesn’t ingest too much. Drinking too much water can be deadly.
Latent learning has a precise definition in learning theory and it’s not what many people think. It’s not magic learning that happens during downtime–at least not in the way people assume. It is not a sudden better performance after a break between training sessions. It’s not when everything suddenly comes together after we sleep on it.
Here’s the definition:
[Latent learning is] learning that occurs during non-reinforced trials but that remains unused until the introduction of a reinforcer provides an incentive for using it.–Lieberman, David A. Learning: Behavior and Cognition. Wadsworth/Thomson Learning, 1990.
Note that the definition includes nothing about making sudden cognitive leaps. If we are struggling with teaching our dog something and in the next session she has improved vastly–this does not fit the definition of latent learning.
One reason we can be sure it doesn’t fit is that when training we are regularly reinforcing the behaviors we want or reinforcing the closest approximation to them that we can get. Again, latent learning deals with “non-reinforced trials.”
The First Latent Learning Study
The study that prompted the definition and exploration of latent learning took place in 1930 by Tolman, Chace, and Honzik.[1]Tolman, Edward Chace, and Charles H. Honzik. “Introduction and removal of reward, and maze performance in rats.” University of California Publications in Psychology (1930). Rats were divided into three groups and the individuals in each group were put in a maze. The rats in Group 1 received a food reward when they reached the end of the maze. The rats in Group 2 never received food; they just were put in the maze and wandered freely for a certain amount of time for 10 days. The rats in Group 3 wandered the maze with no food for 10 days, then on the 11th day they started receiving a food reward for finishing the maze. It took them only one day to catch up to the Group 1 rate of running the maze. This was believed to show that they had been learning to navigate the maze during the period of no food, i.e., no reinforcement.
Stevenson demonstrated probable latent learning in humans in 1954. His experiment also dealt with remembering locations.[2]Stevenson, Harold W. “Latent learning in children.” Journal of experimental psychology 47.1 (1954): 17.
A real-life version of latent learning could go like this. Say I have no interest in bicycles or cycling. None. Nobody in my life does that. And say there is a bicycle repair shop in a little strip mall that I pass sometimes. If I notice that, there’s nothing in it for me. No reinforcement.
However, let’s say I have a new friend who is into cycling. She cycles to my house one day, and just as she arrives something goes wrong with her bike. She needs a repair. If at that moment I remember the location of that bike repair shop, that is latent learning. Learning about the location of the bike shop was not valuable earlier. There was no reinforcement available for it. To repeat the definition: The knowledge was “unused until the introduction of a reinforcer provided an incentive for using it.” In this scenario, the potential reinforcement is that I can help my friend.
What Should We Call the Other Thing?
OK, so if that’s latent learning, what should we call that thing that happens when we wait a little bit, then it all comes together? When everything gels and we, or our dogs, “get it”? It’s a great thing when it happens; no wonder we want a name for it!
Candidate #1 could be the so-called Eureka effect, where a perplexing problem becomes clear all at once in a flash of insight. But the focus on this term is not on the passage of time, except that a period of sleep is sometimes mentioned. Also, it’s not usually applied to animals.
Consolidation is the processes of stabilizing a memory trace after the initial acquisition.–The Human Memory
It involves converting something we know from short-term to long-term memory. It could contribute to fluency in knowledge and possibly tasks. It is even known to correlate with getting some sleep. I am pretty far out of my league here, but it seems like it could apply, for example, in something like cue recognition. It could account for a notable difference in correct cue responses from one session to the next. But I’m not sure whether that merits that dramatic change we are usually talking about when something all comes together.
Here’s a good review article if you want to read about memory consolidation: Memory–a century of consolidation.[3]McGaugh, James L. “Memory–a century of consolidation.” Science 287.5451 (2000): 248-251.
Candidate #3 could be that some dramatic improvements we observe are related to longer inter-session intervals. Since the early 20th century, learning and behavior researchers have been studying the effects of tinkering with the times between sessions of learning.[4]Shea, Charles H., et al. “Spacing practice sessions across days benefits the learning of motor skills.” Human movement science 19.5 (2000): 737-760. That time period is referred to as the inter-session interval (and yes, occasionally the time between is referred to as inter-session latency, just to build in some confusion). But I’m not aware of a zippy term for the advantages of a longer wait, although said advantages are common. Somehow, “benefit of a longer inter-session interval” isn’t sexy.
But What If There’s No Such Thing?
It gets more complex. There were later studies that countered the latent learning effect. There were researchers who argued strongly against it. They claimed that the rats in the maze without food were getting some type of reinforcement and that their behavior could be explained under standard principles of behaviorism. You can read about that point of view in this article:
After reading that article, I almost decided not to publish this post at all. But I still think it could be useful. I’ll let eager researchers make their own decisions.
So, in summation:
Latent learning has an official definition and it might not be what you thought.
There isn’t a sticky term for what you thought was latent learning, but I mention three possibilities.
Oh, and latent learning (as per the definition) might not exist anyway.
And if you think this turned out weird, check out my post on the (nonexistent) opposition reflex!
But Eileen, Language and Usage Are Always Changing!
Here’s the part where you can get after me for being stodgy or old fashioned. It could be that “latent learning” is on its way to becoming an acceptable term for a sudden improvement in performance after some downtime. I have seen one recent journal paper that uses the term that way.
I don’t know if popular usage will bleed into academia or not. But learning about the original definition turned me on to some pretty cool research, and I hope you enjoy it too.
This post started life as a rant about terminology on the Facebook group Canine Behavior Research Studies. Thank you to the people who contributed to the discussion there, particularly הדס כלבי ה, who suggested the term memory consolidation, and Sasha Lazareva, who brought up the “other” controversy about latent learning and cited the Jensen article mentioned below.
Tolman, Edward Chace, and Charles H. Honzik. “Introduction and removal of reward, and maze performance in rats.” University of California Publications in Psychology (1930).
People are rarely referring to choices between positive reinforcers when they write about their animals having a choice.
But here’s another thing that gets under my skin. These days it seems like many people who use the language of choice to describe their training are referring to the fact that they permit the animal to leave as relief from a difficult task. For instance, in a husbandry session, the dog may receive a food reinforcer for cooperative behavior. That constitutes positive reinforcement if we see cooperative behavior (usually staying still or focusing on something) increase or maintain. [1]This also applies to sessions of counterconditioning where the food is not contingent on behavior. I am setting that aside for now. The dog is allowed to leave as often as she wants. The session starts back up if she returns. The leaving constitutes negative reinforcement if we see leaving increase or maintain. But remember: escape is only a reinforcer if the activity is unpleasant.
Letting the dog leave is a good thing. But there is a big drawback if it is planned on as an expected response and built into a protocol.
Building escape behavior into a protocol can provide a disincentive to the human to make the process as pleasant for the dog as possible. Rather than working harder to create a situation where the dog doesn’t want to leave, the trainer can focus on saying that the dog is “empowered” by the ability to leave. On the contrary, some trainers, including myself, consider a dog repeatedly leaving as evidence that we have not worked hard enough at making the experience pleasant. It’s a failure, not a goal. It means we didn’t set up our antecedents and graduated exposures well enough.
Forced vs. Free Choice
I have written about forced and free choice before. Forced choice applies to our husbandry example. The dog can stick with the session and get food or another appetitive stimulus, or the dog can leave. Leaving usually leads to an environment that is bare of other positive reinforcers, or has very weak ones. We deliberately set things up that way as an incentive for the dog to stick with the session. There is no shame in that. Controlling other reinforcers is a part of positive reinforcement-based training. But bragging that escape offers the animal empowerment when the other option is bare of interesting activities is a bit strained.
Also, the presence of food can be coercive. The husbandry session may be unpleasant but the food quite good. Hence, the dog is putting up with discomfort to get the food. Again, sometimes we have to perform medical or husbandry tasks that are painful. But why start out that way if we don’t have to?
On the other hand, free choice is a choice between two appetitive stimuli: two good/fun/nice things. Two things the dog will work for. For instance, stay inside and be petted (for a dog who likes that) or go outside and play ball. Play with this toy, then that one. Dig in the yard or lie in the sunshine.
Is there a way to offer free choice between two appetitive stimuli in a husbandry session? Sure, and I tried it. My dogs LOVED it.
Summer watching to see if the Manners Minder will pay out
If You Really Want to Give the Dog a Free Choice…
…you have to stop controlling other options for reinforcement. Instead, offer another option. In my case, I set up for a husbandry session, but provided another reinforcement option in the form of a Manners Minder, an automated treat dispenser.
I loaded it with the same treats I was using and placed it a few feet away. I set it to eject treats on a variable interval schedule. My intention was for the Manners Minder’s rate of treat delivery and mine to be similar. It would eject treats every so often no matter what the dog was doing (no contingency from me). But the dog’s behavior of leaving the husbandry session could be positively reinforced.
I started a nail clipping session with the video camera running.
This unedited movie shows the very beginning, where Zani is still figuring out what the deal is. Is it OK for her to run to the Manners Minder in the middle of our session? (Yes.) Is there a good reason to return for nail clipping? (Yes, because there were gaps in the Manners Minder schedule.) Zani has a genius for optimization and was soon going back and forth.
I was super pleased that husbandry sessions are pleasant enough to her that Zani happily came back. If she hadn’t, that would be valuable information. It would mean I needed to work more on making husbandry pleasant for her. In the meantime, to get the job done, I could stack the deck a little in my favor via treat value or rate of reinforcement. I would have no problem with the ethics of that. In my opinion, it’s still far superior to the scenario where the dog’s only other option is escape to a boring room.
During my other dogs’ first sessions, I needed to call them back a few times. They both tended to get stuck in one place or another because of their reinforcement histories. Thinking it through, I don’t think calling them affects the balance of the two options much. The sound of the Manners Minder is a very strong cue that food is available. Likewise, my calling my dog is a strong cue for the same. I reinforced the dogs for coming back to me when I did so. They were free to leave again right away, but they usually stuck around for a nail clip or two, or until the Manners Minder produced another treat.
In the movie with Zani you can see me using the remote on the Manners Minder. I am turning the down-stay variable interval setting on and off. But in subsequent sessions (not filmed) I just set it and let it alone.
Choice Doesn’t Apply Only To Negative Reinforcement Protocols (Even Though That’s When You Often Hear About It)
One of the things that often gets lost in the discussions about choice is that we offer our dogs a choice every time we give a cue for a positively reinforced behavior. When I call my dog while she’s digging in the dirt in the yard, I have offered her a choice, whether I’m happy about that or not. And it’s a choice between two nice things. But this type of choice is often overlooked because the reason we train dogs is often to get them to do things we want. Offering a dog a choice between two appetitives can be inconvenient for the human. Whereas offering a dog a choice to leave an uncomfortable husbandry session doesn’t cost us much. We know the dog will probably come back because we are the source of R+ in the room. It seems pretty self-serving to me to promote choice primarily when it is easiest for us.
If a trainer or a protocol focuses on choice, ask questions. What are the choices? Ask the trainer or author to operationalize them. Are the choices antecedents or consequences? What will your animal be choosing between? The trainer should be able to tell you whether both of the choices lead to positive reinforcement, or if one leads to positive reinforcement and the other to negative reinforcement (escape).
Don’t Necessarily Try My Experiment at Home
This was an experiment. Our success with the dual reinforcement setup had a lot to do with the dogs’ history with me. Offering a powerful reinforcer for leaving a husbandry session could backfire if a dog didn’t have a strong reinforcement history for staying. I’m not necessarily recommending it. I wrote in another post about the down side of offering a dog between two positive reinforcers and how it can be tricky. That risk is very clear in my game with the Manners Minder.
Another issue is that the dual reinforcement setup as I presented it is not workable for procedures where the dog must stay still, perhaps as in a jugular blood draw. But that’s true for any method that allows the dog to leave. Most of us at some point also train the dog to stay still.
I tried this out because I was curious. I am publishing it because I want folks to see what it can look like for a dog to exercise free choice in a husbandry session. I’m continuing to do it because it makes toenail trims downright fun for my dogs.
Please share this post whenever you see someone suggest to an inexperienced trainer that she use sniffing as a reinforcer on walks.
Summer air sniffing
Hey. You know that thing that seems like the perfect solution to problems with your dog when walking on leash? It’s not a free ride after all.
Let’s say you are a beginner trainer struggling with teaching loose leash walking. Your dog is very tuned into the environment, and the thing he is most interested in is sniffing. So when you ask for advice on the Internet, it is pretty much guaranteed that several people are going to chime in with the same suggestion: use sniffing as the reinforcer!
It sounds perfect, right? Sniffing, i.e. access to odor, is a powerful motivator for most dogs and an obvious candidate to use for reinforcement. But unfortunately for us amateurs, it takes some finesse to use it. I know this from hard experience.
What’s the Problem?
The problem is that if you release a dog to sniff, you need a way of getting him back. He has already told you with his behavior that access to sniffing interesting odors is massively interesting, more so than the treats in your pocket. So if you let him sniff, then what? You have the same problem you had to begin with, only worse: your dog is all excited about the environment and can’t focus on the task of leash walking.
If your dog is a beginner at this, the problem is worse. You are asking him to learn a difficult skill, keeping the leash loose and walking with you in a connected way. But you are also letting him run around sniffing with you following behind on a sometimes tight leash. How do you keep the criteria clear?
Most of the time a non-fearful dog’s difficulty with loose leash walking boils down to some of the following:
The dog is too excited to focus on the difficult task of walking on a loose leash;
The dog didn’t get enough practice walking at his person’s side in less interesting environments; and/or
The item (food, toy) the trainer is trying to use as reinforcement doesn’t have enough value.
Zani is probably stress sniffing
(I am excluding fearful dogs, because their issues can be different. Sniffing may not be for the purpose of information or entertainment. It can be a displacement behavior, a sign of stress. In this post I’m discussing dogs who are not anxious, just intensely interested in the odors around them. Be sure you know which it is for your dog.)
Let’s see if releasing the dog to sniff will likely solve the above problems.
Dogs who are too excited to focus. If you get a few steps of nice walking on a leash from such a dog, and release him to sniff, will sniffing “get it out of his system” and allow him to come back to you, ready to focus and work? Probably not. If the dog couldn’t focus in the first place, allowing a sniff session will probably not fix that.
Dogs who haven’t gotten enough practice in less interesting environments. If he didn’t get enough practice in your training room or back yard, a sniffing session is not going to magically give him the skills to return to you and have good leash manners.
Dogs for whom the reinforcer is not sufficient. If your reinforcer was not strong enough to get the dog’s attention in the first place, how could it work to get him back when you have released him into happy sniff land?
What Happens When You Try It
Here’s what probably happens when you try the sniff thing without adequate background work. You get a few steps of nice walking, and then you release your dog to go sniff. You follow him around a bit. When you decide it’s time to move on, you say, “Let’s go, ” or a similar cue.
Your dog will keep sniffing. You say it again. Still sniffing. You finally pull on your dog’s leash to get him to come with you. Depending on how deeply interested he was in the smell, he may come now, or he may wait until you pull even harder.
If the environment is so fascinating, and you’ve cued him to go dive into it, how will you humanely get him back?
You might be one of the lucky few with a dog who thrives on doing stuff with his human. One or two of these pulls from you and he’ll get it that he is supposed to quit sniffing and come back to you, and will do so in the future. But if you had that kind of dog, you probably wouldn’t be losing him to the environment in the first place, right?
Pulling him to you if he fails to respond is employing an aversive. Depending on the dog, it can vary from mild to extreme.
In positive reinforcement-based training, cues are opportunities for reinforcement. We train so that there is no need for an “Or else.” And incidentally, pulling on a dog’s leash to enforce a “come” cue is exactly the behavior that was used in the graduate thesis from University of North Texas on so-called “poisoned cues.” The dog in the study responded in a completely different way to the cue that included the possibility of this type of forced compliance than to the cue that was trained with positive reinforcement only.
Pulling on your dog’s leash to get him to walk with you is aversive but not the end of the world. It probably happens to most trainers at one time or another. (It’s a good reason to attach the leash to a harness rather than a collar.) But if it happens most every time you cue the dog to go sniff, you are shooting your training in the foot. Dragging your dog around is one of the problems you were trying to solve to begin with.
The Solution: Practice, Practice, Practice
The solution is straightforward, if not exactly easy. You need to practice not only the loose leash walking in less stimulating environments, you also need to train the heck out of your “return to me” cue with positive reinforcement. And you will need to reinforce it on walks, at least some of the time. You can’t leave the food at home.
Elderly Cricket’s sniffing on walks was no problem
It can be difficult to emulate the real life scenario of an enticing odor. If you are in a low-distraction environment, there probably aren’t many novel odors there. So there’s a gap between practicing the cue when your dog is standing around with not much else to do, and practicing it when he is enticed by something.
I actually succeeded fairly well with this with Zani, my little hound mix, and now I’m practicing with Clara. Some of you will remember my post on when not to work on loose leash walking. I’m pleased to say that Clara handles being around people and dogs on walks so well now that we have “graduated” from counterconditioning and are working on loose leash walking.
Allowing my dog to get what she most wants is important to me. Here are some of the ways I worked on our “Let’s go” cue. I bridged the gap between low-distraction practice and real life situations in several ways. You can see these in the movie as well.
I started off by practicing indoors with Clara both on and off-leash. She learned to reorient to me when I cued, “Let’s go!” I often turned and ran after giving the cue, both to make it both more challenging and more fun.
I have a closed-off room in my house that contains dog food, chew toys, and other interesting stuff. Odor heaven. In training I allowed Clara to go in. After a moment I called her out and reinforced her generously for walking away with me.
I put synthetic rabbit pee on a piece of cardboard and taped it to the floor. Clara got a reward if she responded and left it at my cue. (You’ll see in the movie, though, that it was only enticing the first time. That was a lot of trouble to go to for just one repetition.)
I scattered kibble on the floor or ground, let Clara start eating it, then gave her the “Let’s Go!” cue and rewarded generously when she left the kibble behind. This is a good corollary to what we are asking dogs to do when we ask them to leave enticing smells.
And that leads to–Zen (“leave-it.”) Any kind of leave-it exercise is good practice for this.
Someone is sure to suggest sending the dog back to the odor or out to sniff again as a reward. That’s a great thing to do–sometimes. But you can’t use it as a reinforcer every time with most dogs. If you never reinforce at your side, you risk gradually sucking the value out of walking with you. Nope, most of us just can’t get away with leaving the food or toys at home.
If you watch dogs who are having a good time sniffing, they don’t just stand still and put their heads down to the ground for a moment. They follow where the odor takes them. If free to do so, many dash back and forth, run, walk, back up, make sudden u-turns, and stop just as suddenly. Our dogs are generally much faster and more agile than we are. As long as your dog is on leash you will be in the position of having to curtail some of the fun. How far into the neighbor’s yard may he go? How close can he get to their cat? You will have to assess how much these limitations frustrate your dog, and make sure that using sniffing as a reinforcer under these conditions is worth it to him.
Anybody have any sniffing odor as reinforcement stories? I’d love to hear about them.
Positive reinforcement-based training is subject to a lot of misunderstanding and misrepresentation. Many people genuinely don’t understand how it works, and others seem to deliberately misrepresent it. Some of these misunderstandings and misrepresentations are very “sticky.” Misunderstandings, straw men, myths—call them what you will, Continue reading “6 Myths about Positive Reinforcement-Based Training”
Thank you to Jennifer Titus of CARE for Reactive Dogs for editorial advice. All errors and awkward moments are mine alone.
Citing “stressed-out R+ dogs” in an argument is an old chestnut that comes around regularly. The writer usually describes a training session he or she witnessed where a dog being trained with positive reinforcement was exhibiting fear or stress. The goal of sharing this description generally seems to be to blur the real differences between training that is based on positive reinforcement (R+) and training that is based on escape, avoidance, and punishment. Sometimes it is a feeble attempt to argue with the ranking of methods in assessments such as the Humane Hierarchy.
Cherry-picking a moment out of any dog’s life to support a general point about methods is tempting but is not effective argument.
My dog Summer showing stress during an R+ training session. What can we therefore conclude about the learning process called positive reinforcement?
The “Stressed-Out” R+ Dog
So let’s consider the stressed-out dog in positive reinforcement training. What are some possible causes of stress in an R+ training session?
When using positive reinforcement, some metrics we use to assess the skill of the trainer and the effectiveness of the training are timing, criteria, and rate (or sometimes magnitude) of reinforcement. Let’s start our analysis there.
Bad timing can cause the dog some stress through lack of clarity. The trainer is marking and rewarding some incorrect behaviors while sometimes failing to reinforce some correct ones. If she cleans up her act and stops reinforcing the wrong stuff, the dog will go through an extinction process. Depending on the trainer’s skill, this can be stressful.
Raising criteria too fast means a higher failure rate. This can also cause some frustration. So while this is in an R+ training environment, what you have when you raise criteria too fast and the dog doesn’t do anything reinforceable is, again, an extinction problem.
If the rate of reinforcement is too low, you can actually put the desired behavior on extinction. So you may get a confused dog who starts throwing behaviors out of frustration, or a dog who will wander off and do something else more reinforcing, given the choice to do so.
Another stressor can be the use of negative punishment when the dog hasn’t learned the behavior. If the dog isn’t clear on how it can earn the reinforcer, it is frustrating to have it taken away contingently as it tries other things.
Note that none of the above errors is likely to hurt, scare, or startle the dog.
Two more types of stressors possible in an R+ training session are pressure of some type, and an accidental, momentary aversive. These two can indeed hurt, scare, or startle the dog, but are not linked to the positive reinforcement learning process.
What I’m calling pressure could consist of anything in the environment, setup, or even mannerisms of the trainer that the dog would like to escape from. Is something too loud? Is someone pressuring the dog with his or her body? Is the dog being kept too close to something she is scared of? This type of problem comes from the unwitting inclusion of an aversive stimulus.
Likewise, accidents happen, as they can in any training. A trainer might step on her dog’s tail during a stay, but again, this is an aversive accident, not an integral part of R+ training.
So our causes of stress are probably either technical mistakes on the trainer’s part or the presence of an unplanned or unrecognized aversive stimulus. Are these problems unique to positive reinforcement training? Absolutely not. They can happen in training based on aversives just as easily.
A Fair Comparison
Let’s compare apples with apples. Rather than focusing on the stressors in faulty positive reinforcement training, lets compare the net effect on the dog of R+ training vs. aversive-based training–with both done poorly. There is certainly no shortage of sloppy training done with aversive methods. I can find such a video on YouTube within a couple of minutes, and the trainer is often touting it as a success story.
So what happens to a dog being trained with escape/avoidance and punishment when the problems and errors I described above are present? Not only is the dog startled, hurt, intimidated, or at least irritated by the training itself, she will also be subjected to the additional stress resulting from trainer errors. Or she may experience aversives in addition to the ones the trainer is purposely using.
Here’s what it could look like.
Bad timing: Imagine popping a dog’s collar when she is heeling perfectly, in addition to popping her when she makes an error.
Changing criteria too fast: Imagine using duration shock to teach a dog to jump off a platform immediately after using it to teach her to jump on it.
Unplanned aversive stimulus: Imagine teaching stays using your hands to force a sound-sensitive dog to hold her position while a delivery truck with a no muffler drives by.
Those make the possible stressors in R+ training look rather like small potatoes, don’t they?
A Real-Life Example of the Results of R+ Training with Errors
I will be the guinea pig. I have a video of my own training that demonstrates many of the stressors I listed above.
In this popular video of mine that demonstrates lumping, I raise criteria too fast for Zani. She gets visibly frustrated. You can see it around 2:25 in particular. She plants herself in front of me in a sit and makes what I call the “terrier frustration noise.” A sharp exhale through her nose. I don’t blame her.
In addition to the training errors that are the subject of the video, there are more. I often mark late. I mark and reinforce improper behaviors, both when she targets my bare hand instead of the tape, or does a “drive-by” and doesn’t connect at all.
My rate of reinforcement is not bad, but there are a couple of times when Zani is going through extinction, trying other behaviors, where I might have interrupted her sooner, or marked something approaching the right behavior.
My reinforcement placement is not thoughtful. I am generally tossing the treat in order to reset Zani, but think how much faster she could have gotten to the wall if I had treated in that direction instead of away from it?
Another criterion issue is my poor choice of tape color. Gray, even metallic, is not a good contrast on a tan/yellow wall. Zani probably couldn’t see it well.
Interestingly, there is a subtle aversive stimulus in the session as well, and I think we can see the effects of it on Zani’s actions. The tape on the wall is in a tight area. I think her reluctance to enter that small area (in other words, an aversive setup) is one of the reasons she targets the desk multiple times instead of going for the tape. She is extremely pressure sensitive and I am asking her to go by me into a tight little space. She tries to avoid it.
So in one video, we have many of the problems I listed above.
But even with the errors in the training and the slightly aversive setup, Zani hung in there with me and was wagging her tail in the last section. She successfully learned the behavior I was teaching and got 24 tasty food treats in the three minutes of training time shown. Not a bad rate at all, considering that there were two dry spells and also that she was spending a fair amount of time chasing down treats.
So here is a thought experiment. Imagine that instead of what you saw in the video, I used aversive methods to get the targeting behavior from Zani. You can imagine a combination of physical manipulation and body pressure, or a shock collar. No food in the picture. (If you are imagining Zani falling to pieces, that’s about right.) Now add to that multiple errors of timing and criteria, and an unwise setup that creates a tight space. How is Zani doing now?
That is a much fairer comparison of the results of different training methods.
The Proper Rejoinder
Evoking the scenario of the stressed-out R+ dog in argument invites the following response:
It’s a good thing the dog was being trained with positive reinforcement then. Adding training errors and aversive situations to any protocol can cause stress. Think how much worse it would have been if the dog were being deliberately trained with aversives to start off with!
The real illogic of the comment in the title is that in most examples described it’s the addition of aversive stimuli that creates stress. Blaming stress that results from the accidental inclusion of aversive stimuli on the process of positive reinforcement training is not only illogical; it’s a cheap shot.
Conclusions from Examples
Drawing conclusions from examples is tricky, and can easily lead to the logical fallacy of “missing the point.”
A couple of the valid conclusions that can be drawn from the “stressed-out R+ dog” scenario are that some positive reinforcement trainers lack mechanical or observational skills, and that it is possible for other learning processes besides positive reinforcement to be going on when we are trying to train with R+.
What the scenario doesn’t support is the idea that there is some unknown dark side intrinsic to positive reinforcement training, or that there are characteristics of training methods that are immune to analysis through learning theory, or that stressors from lack of skill happen only in R+ training, or that training based on the use of aversive stimuli can make for a happier dog.
