Many years ago, when I was a child, I was terrified of deep water. Both my parents were concerned that I might drown if I fell accidentally into deep water. I therefore had to learn to swim.
Dogs often have to face their fears: bicycles, strangers, children, dogs, skateboards and so on.
Many treatment options (and acronyms) exist. Boiled down to the basics, you can punish, reinforce, create associations and break them.
Many people feel that humane rehabilitation gives the dog choice. The idea being that choice is empowering. Limiting choice is somehow less effective and perhaps even less humane.
Wait one minute. My parents did not give me any choice where it came to my fear of water. I had to overcome it. If I had been asked, “Do you want to learn to swim in the deep water?” I would probably have screamed “no” and ran.
Dogs don’t get to choose whether to participate in rehabilitation programs either. It is entirely plausible that dogs would refuse to consent to rehabilitation.
Adult humans are different and they do get to choose. For example, you can choose to face a fear of snakes or not. It is your choice.
However, just as my parents decided that I must get over my fear of deep water, a dog’s owner decides if a dog will do rehabilitation exercises. There is no choice here for the dog.
The human also chooses the techniques to be used on the dog. Some programs claim it is best to let the dog choose when to approach and when to retreat. Is this wise?
Let’s return to the lake for a moment. I enjoyed playing in shallow water. Going to the lake was fun and I ran enthusiastically toward the water. I knew that I could leave the water at any time. Which incidentally, did not make me like deep water. I just avoided it.
It is very common for lakes to have areas that drop off sharply. Should the sand break away beneath your feet, it can send you plummeting into deep water. One minute you are happy. The next you are in a fit of terror. This is not necessarily a momentary setback. Fear can quickly escalate and spread. Shallow water becomes unpredictable and potentially unsafe.
The wiggle room between safe and dangerous is so small – one tiny step. Giving choice to someone who cannot see the bigger picture may create threshold roulette.
Similarly, in dog training, a dog can be given the choice to approach scary things. Much like the child that runs into the lake, the dog can get quite close to that line that separates okay from definitely not okay. Reading body language is not going to help because the dog is legitimately okay until it is not.
These are the dogs that happily walk past other animals on walks, but lose control when unexpectedly sniffed. It’s the animal that can happily nudge the hand of a child and walk away for a treat – only to blow up if the child fidgets or squeals. Like sand crumbling beneath your feet in the lake, the dog learns that situations that previously were enjoyable are now unpredictable.
The notion that the dog should be in control is flawed. Dogs cannot foresee problems and may not anticipate surprises.
That’s not to say a dog should never have choice. If my dog startles when it sees a garden gnome, I am confident the gnome is not going to start dancing about. I can give my dog more choice because the dog’s threshold isn’t sitting on a behavioural sandbar.
However, scary things such as other dogs, children, people, bicycles and skateboards are not fully predictable. No matter the skill of the trainer, a decoy dog may bark unexpectedly. Children fidget. Adults fail to follow direction. Skateboards may inadvertently flip and bang. The closer the dog is to the precipice, the more likely you will find yourself playing threshold roulette. When luck runs out, your only choice becomes restraining the dog or letting it flee. Neither of those is desirable.
Different strategies aim for different thresholds. For example, classical conditioning aims to keep the dog well below threshold. Done correctly, the dog may even be unaware that rehabilitation is taking place. There is plenty of wiggle room for errors.
By contrast, negative reinforcement aims to reward the dog by relieving discomfort or pain. The dog is intentionally placed much closer to edge. The closer the dog is to the edge, the more likely that you’re going to lose your roulette game. Positive punishment lets the dog go over the edge, disciplining the dog after it has gone off the deep end.
I did overcome my fear of water. My parents devised a most clever rehab program. They took a life jacket and divided the foam core into strips. I swam with this life jacket. I had fun. I felt safe. I could not slip into deep water.
Little by little, as my ability to swim improved. Unbeknownst to me they removed one strip of foam at a time. Eventually, I was swimming with a life jacket that did not float. When the time was right, I found out that the life jacket was useless. I could swim and rather enjoyed it.
The point being that this program was:
- Under threshold.
- Stress free.
- Aversive free.
- Avoided threshold roulette.
- I was given no choice. In fact, I was blissfully unaware – as classical conditioning should be.
In the end, I learned to swim, overcame the fear and even took up scuba diving.
Taking away choice is not always a bad thing, especially if it avoids creating threshold roulette. Taking away choice can be humane, effective and thoughtful. Moreover, you may even find that the dog learns to love – rather than tolerate – the things it once feared. The bigger concern is how do you keep a dog truly under threshold?
awesomedogs 
Great article and the example is wonderful, please update the blog, has amazing information
Hey, I‘m having trouble finding an email address. Can you email me back to ask you a question?
–Shaye
shayewalsh1@gmail.com
All my contact information can be found at http://www.awesomedogs.ca.
My favorite point of this post, among many (wow, shifting sands underfoot; great metaphor/comparison) is that classical conditioning can be going on without the subject even being aware of it. Now THAT’s non intrusive.
Beautifully written. Thank you for giving me a different perspective on this. Very thoughtful. One thought I have is this: sometimes using food for counter conditioning actually masks the true threshold and you do still get into roulette territory. On the other hand, when done well, sometimes when not using food, but instead focusing on choice-points forces the trainer to really be more sensitive to threshold. But I certainly see the merits of both perspectives, as well as have seen many dogs respond equally well to a combination of approaches.
I think in some cases, using food can change the threshold. So yes, you have to be sensitive to it. But that also means that if you don’t use food, you have to be sensitive to the effect adding food to the mix at a later date might have.
So, yes I would agree that food vs no food affect threshold. Would be nice if we could ask the dogs the way one would ask a human patient … no?
The discussion starts to sound a little familiar. A person never needs to use pain or discomfort, either physical or emotional to train or rehab a dog. It is easy to talk about physical pain, that seems obvious to some degree. Emotional discomfort is a little harder to discern it seems, given the uncertainty about being able to recognize where threshold is for your dog. A certain amount of emotional discomfort does seem to be acceptable for some people who are also practitioners of force free training. The arguments are similar – aversive situations happen naturally, my dog has control over the level of aversive he experiences, the dog knows what to do, it worked when counter conditioning didn’t, dog isn’t food motivated, my dog is happy and loves me, I would never hurt my dog…………..
The bottom line for me is that my dog depends on me to keep him safe, well fed, housed and active. Most of what he needs for a good life will come from me. I feel it is so important that my dogs trusts that I won’t hurt him. If I put him in a situation where he feels uncomfortable, or unsafe, and I push him even a little in that direction, how is that building trust that I won’t bring him harm? I don’t want to systematically and intentionally cause my dog to hurt physically or emotionally.
My dog will still eat if he is near threshold or even at threshold. In fact, yesterday he was barking at a person on a moped from inside the car, he still ate the cheese bits I offered him. He just barked in between eating the cheese bits. We were in a car stopped at a red light, so we were kind of stuck.
I use my intuition as a barometer also, if I am feeling good and comfortable and at ease, then probably my dog is as well. It is important to be in that place. It is important for the dog to feel really calm and at ease. Why go in between calm and peaceful and over threshold? We may still be defining threshold, and it may take some practice to recognize it in your own dog, but I think it is possible. Just because we can’t define 100% what under threshold is, is not a reason to push the dog to near threshold. I taught my dog to take treats gently and he does. You barely feel his mouth when he takes a treat, he doesn’t grab for it. If, when we are out on a walk, he starts catching the tips of my fingers or taking the treats quickly, I know he isn’t 100% calm and relaxed and that is a good enough indicator for me that we need to back off.
Threshold roulette. Interesting. This would also rule out any type of therapy that sets as it’s basic principle of starting under threshold. For although one could track this using neurobiology, we don’t. We still have to use observations to judge if our point of contact with the aversive stimulus is under “threshold”. But which threshold is the proper one?
We dog trainers have devised a cleaver threshold, that doesn’t exist in the literature. And it’s completely re-definable according to what you want to do. Some people like Jean Donaldson write, that you know you’re under threshold, when the dog eats. If the dog refuses to eat, the dog is “over threshold”. But she doesn’t identify the threshold. So what thresholds are there in the literature? Is it possible for a dog to eat and still “be over threshold”? Only if you use some other lower threshold. But is THAT the proper one?
Here are the thresholds I found while researching for a talk i gave on this subject:
1) Absolute threshold: the lowest level at which a stimulus can be detected.
2) Recognition threshold: the level at which a stimulus can not only be detected but also recognized.
3) Differential threshold: the level at which an increase in a detected stimulus can be perceived.
4) interest/fear threshold – the point where interest in a subject can change to fear of the subject
5) Terminal threshold: the level beyond which a stimulus is no longer detected.
I’m sure there are more, but in the psychology and behavior literature, I didn’t find any other relevant ones.
Even these were not brought into conjunction into any observable behavior, let alone are they identified by trainers as THE “threshold” they use.
Some will say that Threshold 1 is the right one for an effective behavior change in a fear situation. But … how do you know the dog has detected a stimulus without recognizing what it is. Is that even possible? How could it be an aversive stimulus that needs to be turned into a joyful one, if the dog doesn’t recognize it?
Some will say it’s Nr. 2 – ok, but if it’s recognized, won’t the dog go overboard? Maybe, but only if the distance is still ok, that it’s not quite too intense. But then we’ve already perhaps gotten into the range of Nr. 3, which already puts us “over threshold” for Nr. 2. Mostly because how are we supposed to know if the dog recognizes it but isn’t bothered by it? And how to tell the difference between detecting something it doesn’t recognize and recognizing it at th next levle?
Others still will say it’s Nr. 3 – this sounds great, but it wold appear to me, that they can be 100 of these if the stimulus is recognizable at 100 meters, for every meter can be another level of stimulus. To the dog. Will the dog show us all these levels she feels? How do you KNOW this?
Number 5 doesn’t really concern us, because it seems to be at the complete opposite end of the spectrum, going away.
So that leaves us with Nr. 4 – On the face of it, this should be easy. Going towards the stimulus is interest and going away is fear. Except when it’s not. Except when charging the stimulus is not a distance decreasing behavior but rather a distance increasing behavior based up learning, that “When I charge barking and snapping, that scary dog goes away.”
Ok, so how could we determine, if interest (under threshold) is what’s going on and not fear (over threshold). Body language, position, facial expression, other signs such as lip licks. We’ve all seen the flowing charts of the “Calming Signals”, some conceived in one way some in another. But I was told in no uncertain terms recently, that basing a scientifically proven and effective protocol on such signs which have not been studied, written up and published after a process of peer review is also not proper. And because of this, although these very signs are used in countless studies and books for the quantization of the results of scientists in their studies, for example as visible signs of stress in various methods of training to mention just one, we don’t REALLY know if that lip lick is a sign of stress or a fly on the nose. How hard is a hard stare as opposed to a soft stare as opposed to watching a baseball game going on behind the stimulus.
At that point it also doesn’t matter what protocol you use. A D&CC is dependent upon starting at and keeping a dog “under threshold”. Are we SURE that dog IS under THRESHOLD, under which one, is it the right one, how do we know it?
One more thing … research is beginning to show, that D&CC is fine, but continually needs to be refreshed, otherwise it, like other respondent trained behaviors suffers from fade out if the associations are not themselves always present (think about suddenly not following the click with a reinforcer) and a spontaneous recurrence can occur. Training so-called coping strategies, which allow the subject to look, analyze and decide how to react to the stimulus have shown themselves to be as or more resilient to the presentation of the same stimulus over time and in different situations.
Also something to think about. No psychological procedure, no medical procedure and no medicine has ever had a 100% effectiveness rate. None. Ever. While the poor application of same will certainly not help it’s effectiveness, the perfect application is also no guarantee that it will work (or stay “worked”).
I didn’t rule out any therapy. I pointed out that the closer you chose to move toward being over threshold, the more likely you’ll trigger threshold roulette.
I think we owe it to our dogs to work at a place of the least aversion possible. Of course, there will always be those who are willing to add or remove aversives simply because it is faster and invent all kinds of justifications for doing so.
I personally do not think aversives are faster. Each technique has the potential for fallout (I’d say even R+), some more than others. So, if I look at the average time to fix a problem over multiple dogs, I find some techniques end up taking way too much time to fix all the problems that rehab created.
And to me, I’m not a fan of a dog that appears fixed but may be suppressed. If that dog gets cornered – and you can debate how that shouldn’t happen all you want – life isn’t predictable…I don’t want a dog that seems okay but really isn’t. When cornered is unable to handle things. To me, if a dog is at that point there is much more work to be done.
um …. over which threshold? Not the made-up dog trainers’ threshold. Over which actual recognized in behavioral psychology threshold? And how do you determine this?
If you have a particular trainer’s made up threshold in mind, then perhaps mention the name and I’ll tag them into this?
To me, I think a defined threshold, even if made up at least is defined. Otherwise it’s at risk of being just another made up buzz word. Like positive means one thing to “some trainers” and it can mean another to another.
The act of saying that x or y technique is under threshold really doesn’t mean very much otherwise does it?
1) Trainers usually refer to threshold as being between socially acceptable and socially unacceptable behavior.
2) Some will refer to how to determine if a dog is over threshold if the dog can take food.
3) Others will say that as soon as the dog shows signs of stress (distress or eustress?) the dog is over threshold
4) others say three but give 2 as a way of telling that stress.
None of these are actual thresholds mentioned on any literature. The only thresholds I’ve found mentioned are:
Absolute threshold: the lowest level at which a stimulus can be detected.
Recognition threshold: the level at which a stimulus can not only be detected but also recognized.
Differential threshold: the level at which an increase in a detected stimulus can be perceived.
interest/fear threshold: threshold between interest and fear: increase in the rate of informational input arouse interest, until the rate exceeds the interest-fear threshold.
Terminal threshold: the level beyond which a stimulus is no longer detected.
What we sometimes see as fixation is an attempt by the dog to gather more information about that “thing” in the distance. The more information the dog can get at a safe distance, the less scary it becomes. When the dog herself learns to look, gather information and either approach more OR break off, the dog is becoming empowered to control her own environment. Empowerment training is not new and is an answer to the decades old problem of “Return of Fear” that is often encountered in pure respondent (desensitization and respondent counterconditioning) protocols. One of the first to document this and start the research ball rolling was S. Rachman, The University of British Co/umbia, “THE RETURN OF FEAR: REVIEW AND PROSPECT” 1989. This was also explained in a lecture by Dr. LeDoux: https://www.youtube.com/watch?v=9_IIgXWdF-w
P.S. – the empowerment therapy is therefore started at the Absolute Threshold and the DOG determines where the interest/fear threshold is, in that the owner has the dog on a loose, long leash and only serves as a brake if she feels the dog may through his own moved towards the trigger cross that interest/fear threshold.
Now here is where it gets REALLY interesting:
Counterconditioning in the treatment of spider phobia:
eects on disgust, fear and valence
Peter J. de Jong*, Ingrid Vorage, Marcel A. van den Hout
Netherlands, 2000
Methods
“2.2.1. BAT
A behavioural approach test (BAT) was used to assess the approach of a medium sized house spider. The spider was placed in a glass jar on a table. A pencil and a plastic washing-up bowl were also placed on this table. The participants were instructed as follows: “To get an impression of how far you dare to approach a spider, I will ask you to perform a number of steps. You are free to refuse each step, you are not required to force yourself. But, you should do your very best so that we get an impression of how far you dare to go. Do you have any questions concerning this procedure?” The participant was instructed to perform each step
following verbal instructions given by the assistant, who remained in the corner of the room.
There were 8 steps:
(1) walk towards the spider as near as you can;
(2) touch the jar;
(3) open the jar;
(4) take the jar in your hands;
(5) touch the spider with the pencil;
(6) put the spider in the washing bowl;
(7) touch the spider with a finger;
(8) let the spider walk over your hands.
After each instruction, the assistant asked the participant whether she was willing to carry out the step or not. When participants refused, the instructions describing the step were repeated.
To get a positive rating, participants had to start with the step immediately after the
instruction and had to perform it successfully within 1 minute. The assistant neither
encouraged nor praised the participant.”
Sound familiar? 8-9 years before this was structurally applied in a similar manner (but not exactly the same) to dog training and … with a similar name.
Exposure therapy compared to classical conditioning with candy and preferred music. And yet, classical conditioning was just as effective. Guess people should stop slamming the classical conditioning as not effective.
That being said, I don’t see anyone increasing distance between trials. So it’s not BAT. So not sure how it’s relevant.
Plenty of research out there demonstrating the use of desensitization and counter conditioning offers excellent results.
If someone wants to make the case that a technique is more effective/faster/free from the risk of R-, then it is up to them to have supporting peer reviewed evidence available. NO ONE should have to buy a course to get access to research/proof.
And yet other researchers raise serious concerns with Ledoux’s work, making the case for maladaptive coping strategies. When those same concerns are raised by dog trainers, they are yelled at – called dogmatic – and other insults.
Well, if other researchers can raise the red flag on risks of this strategy, then I think their warnings should be heard. Any behaviour that facilitates escape/avoidance, and is maintained through R- is considered a maladaptive coping strategy. It’s marked by a quick decrease in anxiety, that over the long-term supports the phobia.
http://www.jneurosci.org/content/33/9/3815/reply
Ledoux himself offered the following practical advice/example to his readers…..
“Finally, the individual who cannot bear to go back to work should be encouraged to do something else instead but not sit at home and brood. If going to work is too painful, the person might go shopping or visit a friend. The more involved and formal procedure of instructed desensitization to the actual fearful context can be undertaken at a later time. ”
Ledoux says to desensitize. He’s also a big fan of developing pharmaceuticals. I have no problem with medication if appropriately used. Just not sure we should be interpreting his words as anything more than what they are. Ledoux says desensitize, in the meantime, don’t go to work if work is too anxiety provoking. Go shopping.
See, all this angry ranting against people who scratch their head and say, “But wait, R- comes bundled with side effects” on various threads is pretty derogatory. We know that R- in anxiety tends to come bundled with maladaptive coping strategies.
Ledoux warned about them too. There is nothing radical or dogmatic in pointing that out.
I think the biggest onus on those who want to support BAT as an alternative to CC&D/A DRI/R+ is to show not just that it works (we KNOW it works, so does P+, P- and R+), but to show it is MORE efficacious than CC&D/A DRI/R+. I have seen nothing to indicate it is.
If we can achieve the same results using a less invasive method, it behooves us to do so. To do any less is a disservice to the dogs. Anyone who is TRULY dedicated to being force free would not go there.
There are lots of pretenders out there and we need to have a strong ability to analyses to keep from falling prey to those claiming to be force free while still using coercive methods.
Thank you, Yvette, for pointing out that methodology should be supported by peer reviewed research, available to scholars without paying for a course. Lately, with regard to literature, I’ve also thought we need to be more careful, even, about which journals we rely on;-) I respect that you have cited your sources clearly and make cogent arguments.
Kim, kudos to you for pointing out the obvious – force free trainers DO attempt to use the least invasive methods, not to us but to the dogs!
Interestingly, the criticism of how one determines thresholds, or that there are multiple or different thresholds, is a straw man. All that is necessary to conduct classical counter-conditioning is that the dog notice the stimulus and not be reactive to it. He can “gather information” while the trainer is feeding him, since there’s no imperative for any behavior on the part of the dog (if there was, we’d be conducting OC, not CC).
I’m glad this post got active again – it gave me a giggle, too – the “behavioural approach test” acronym seems such a coincidence now, doesn’t it? Think I should add it to our glossary on the RD page?
Swimming takes a physical skill to prevent drowning, so you teach in shallow water (which is what the buoyancy things are doing really). There was no fear of water, if there was that fear would be treated with DS
There is a fear of drowning, a very realistic fear, i.e. if you can’t swim, or can’t swim well, you drown
Another analogy….
Fear of heights will be done with a supportive person, whether child (having no choice) or adult. It will be done using DS, you will be given the OK to move away if it gets too much (looking at dog body language for dogs) and beforehand will be taught to relax, (food, sniffing, play etc. assist relaxation during breaks)
This is totally different to fear of tightrope walking, which involves a physical skill
If a child is made to tight rope walk, I suspect, stabilizers will be used, something thicker to walk across, gradually getting thinner, and the walk on/rope only just off the floor. You wouldn’t have the rope at full height, to go on a bit then off, you would learn the balancing skill 1st. Same learning to ride a bike
So; Fear of a something needing a physical skill is totally different to fear of objects, esp. if the skill may lead to death
Fear of dogs/humans/spiders/heights etc. are all done by systematic DS, there is always some form of –R involved as you will need to take breaks, during these breakes there will be relief, (-R) There is no physical skill needing to be taught
Note; The parents didn’t feed her to learn to swim, they made it so that it was no different to being in shallow water This is not a good analogy as it involves a physical skill, without that skill 1st there most definitely will be death or injury.
Fair enough. I do tend to separate the ability to stay afloat from the ability to swim in any circumstances. Just because I can “swim” in shallow water to me doesn’t mean that I can swim. But, let’s just set that aside and look at another analogy.
I used to be scared of snakes.
I had a son. My son loves snakes.
We’d go to the pet store and they had snakes. He wanted to look at the snakes. I absolutely did not.
If the snake was in a locked cage, I could go near. If the cage was open or part of a demonstration, I did not want to be near.
You can’t let a young child wander off unsupervised. Wanting to be a fair mom what do you do?
“You go look at the snakes. I will stand way back here where I cannot see the snakes but can see the child.”
The joy on my kid’s face is the cookie.
I am feeling no stress because I am far away. There is no aversive because I’ve backed away to that point. (Don’t know anyone who has choice that chooses to put themselves in the position of experiencing an aversive without a very good reason. I question the logic that a dog will choose to push to discomfort in order to experience R-. It’s a bit like saying I love the feel of the air conditioning, so I’ll sit in a hot car for a while because turning on the air feels so good.
I am not reinforced for anything (leaving, twirling my hair (calming signal?), looking at snakes…nothing.)
I am not punished for anything (why? I’m not flipping out.)
Without even realizing it, I enjoyed watching him so much that I found myself taking pictures and getting closer. Surprisingly, I petted the big pythons and rather enjoyed it.
Who knew? No choice in rehab. I really had zero desire to get over my fear of snakes. No fear. No aversive. Fear of snakes gone. Which is a good thing because I now live in the country and snakes are all over the place.
No skill involved. You do not need the application of an aversive to create a conditioned emotional response. There is a sweet spot where awareness is not aversive.
You saw someone else enjoying being close to snakes. I think this is what helped you overcome your fear at a distance. You were also distracted up to a point (taking photos). Was there more going on to get to the point during that one encounter where you finished up by touching a python? There was no need to overcome your fear of snakes, you’re right, at that time but if you hadn’t have had that chance to overcome your fear, the fact that you then moved to the country where snakes are all over the place would have meant you either had to overcome them or not go out if your phobia/fear was so strong. How would you have overcome your fear after moving? How would you have helped yourself or what therapy would have been available to you? How do you think the therapy would have looked? A process of desensitization and communicating to your therapist? It should certainly not be run whereby you’d be working to avoid the aversive object. This would be inhumane. I can’t think of any modern and ethical behaviour modification therapy that would set the process up like this.
I wouldn’t say I was distracted. I’d say I was having fun taking pictures.
As an adult, you have true choice. If snake phobias are that bad, then the adult has a decision. “Don’t move to the country, near a corn field.”
True choice means you can entirely avoid the situation.
That’s right. People and snakes don’t normally mix, but dogs and dogs and dogs and humans usually do. To spend a life away from them because you think your dog would choose that way is ridiculous and short sighted. I have met people who think their dog only needs them and does not need to interact with anyone else or any other dog, ever. I find that a bit sinister. There is no true choice in any behaviour modification protocol, I agree with that but as social creatures, if dogs could talk, I am sure they would say ‘I want to be better. Please help me with the social situations I find difficult’; We choose to arrange to go to ‘growly dog’ or ‘reactive rover’ classes, we choose to do set ups. It’s what happens during that time which counts.
Exactly! I do not like taking Huck to the edge and pushing his boundaries. It makes me feel very uncomfortable and I don’t like seeing the stress in him. It just doesn’t feel right, even if it works. I know things are going well when I am having fun and feel at ease and most importantly, feeling capable. When Huck sees a dog, bike or kid and does the happy dance and looks over at me right away, it’s a good feeling. But if he is tense, if I am tense, it isn’t fun.
“Threshold roulette” – brilliant concept, and how interesting that you chose to blog on it. It seems we’ve both been mulling this issue lately, as I posted this today on my Facebook timeline: “Environmental control is an interesting concept. Everyone likes to control their environment, even dogs perhaps, but here’s a question for you. If you were afraid of snakes, would that make you go in search of snakes so that you could run screaming away from one yet again? My thought is that you wouldn’t, rather, you’d run away from one, and not really want another one to appear.”
I like to keep my dogs safe from things they fear, and influence them, in the least invasive ways possible, to change their minds, rather than just their behavior. I plan to share this blog with others because the points you make are so closely aligned with how I perceive a safer world for dogs should be.
How do you ever know for sure if your dog knows he has control over a situation? Or how do you know he will use that control the way you think he should?
You don’t know. And if you use desensitization and counter-conditioning to change a dog’s emotional state with regard to a situation or stimulus, it doesn’t matter, because the dog isn’t engaged in a three part contingency where he is expected to exhibit any control or offer any behavior. You are just saying to the dog, “If this, then that.” If you see a guy in a baseball hat, you get roast beef. As long as the dog is capable of eating the roast beef, and likes roast beef, the association will be made (provided there is decent inter-trial latency and the actual trigger stimulus is the association the dog makes). The dog will eventually come to know that guys in baseball hats mean the roast beef party is going to start.
IMO, too many people are too willing to over-complicate this process to the point where they are actually working much harder than they have to in order to accomplish behavior modification.
Two part contingency works perfectly for *seeing* triggers on a leash if it’s done correctly, as you say. Not so fluid and natural when the person might want to move on a stage and join an off leash group walk… with people wearing baseball caps. Movement, greetings, retreatings (behaviours) then come into play.