eileenanddogs

Category: Why Use Positive Reinforcement?

Let’s Talk about Using Aversives in Training. I’ll Go First!

Let’s Talk about Using Aversives in Training. I’ll Go First!

A speech balloon with the words, "This method is OK because..." in it.

Today’s post is about how people often justify the use of aversives. I’m going to use my own experience as an example.

  • I am going to present a description of an aversive method I used to use.
  • I am going to list many common justifications that could be offered as reasons why that method could be OK.
  • I’m going to describe the possible fallout from the method for the dog and for the handler.

Aversives

Paul Chance, Learning and Behavior, 7th edition, defined aversives as:

Stimuli the animal would avoid, given the option.

Continue reading “Let’s Talk about Using Aversives in Training. I’ll Go First!”
Just a Trick?

Just a Trick?

Zani's useful "Trick"
Zani’s useful “Trick”

“Crossing over” is a phrase dog trainers use to refer to the act of giving up training that uses aversives and changing over to training that uses principally positive reinforcement: becoming a Humane Hierarchy trainer, a force-free trainer, or a clicker trainer. (We have lots of phrases to describe ourselves.) Folks who have made this change (and those who never trained traditionally) will attest that this is more than just a different set of skills. It is a change of world view, and it runs counter to the emphasis on and acceptance of punishment in our culture. For many of us, it is not an easy thing to do. Social and technical support are both very important.

My friend Marge Rogers is a crossover trainer who crossed over with no local mentor, although she would credit her wonderful dog Chase, as well as books and internet resources. She wanted to change the way she trained and she needed to do it on her own.  She came from a competitive obedience background. She decided, brilliantly, to throw off everything she knew, put her obedience goals temporarily on hold, and train her dogs to do tricks.

Why Tricks?

Here’s what she told me:

  1. Teaching tricks improves mechanical skills like observation and timing.
  2. Teaching tricks helps trainers learn to create training plans and break down behavior (cognitive skills).
  3. It helps develop critical thinking skills. (How different are the skills for teaching dust the coffee table or blow bubbles in water than teaching drop on recall?)
  4. There is no pressure for the handler. Or the dog.
  5. Trick training encourages creative thinking and problem solving.
  6. Trick training give immediate feedback for the handler (via the dog’s behavior).
  7. There is no handler baggage.
  8. And the best reason for teaching tricks – you’re not burdened by the curse of knowledge for stuff you’ve never trained before.  No old habits to unlearn. In short: it’s the perfect way to become a better trainer.

P.S. You can make your own chicken camp.

The Result of Chicken Camp
The Result of Chicken Camp

Marge is referring to Bob Bailey’s well known chicken camps where trainers learn to hone their mechanical skills. This picture is the outcome of one of her personal “chicken camps,” where she taught her Rhodesian Ridgeback Pride a high leg lift to emulate taking a pee (he normally squatted to pee, by the way). She shaped that leg lift all the way up from a twitch.

Marge’s trick skills resulted in her fame as the “Ridgeback lady” on YouTube, who featured her Rhodesian Ridgebacks in videos such as these:

By the way, Ridgebacks have a reputation among traditional trainers as being an untrainable breed.

Finally!

Many was the time that Marge exhorted me to train tricks. I generally declined, saying that it’s all tricks (true, but perhaps evading her point a little bit), and that I had my hands full with polite pet behaviors and agility (also tricks!)

So a funny thing happened. Recently I broke down and trained my dogs a couple of tricks. It was supposed to be just for the heck of it, but two of the tricks immediately became very useful.

Marge says, “That figures!”

1) Sit Pretty. I’ve been teaching little Zani to “sit pretty.” We went slowly, so she could build up her abdominal muscles, but she really took to it. What’s a more classic “trick” that sitting up? Adorable but useless, right? But no sooner did we have a few seconds’ duration than it came in incredibly handy.

I’m teaching all my dogs to sit or stand on the bathroom scale by themselves. I thought I would have to manipulate the dogs’ feet a little bit so that I could see the readout. But Zani solved that problem by offering her “useless” trick.

Link to video for email subscribers

If I were Marge, though, I’d probably teach the dogs to curl their tails around as well, so they didn’t brace any of their weight on them if they were on the floor. That’s a little more than I have the patience for, though. I’ll just elevate the scale if I need to.

2) Leg weaves. I don’t remember why I decided to do this, but I taught Clara how to weave through my legs. Let me be frank: I think that is one of the silliest behaviors ever. Even when the most accomplished freestylers do it, it’s mostly a “yawn” from me.

But as soon as I taught Clara the rudiments, I discovered something. It’s fun! No wonder people do it. Clara and I both enjoyed it, although I’m sure we looked even dorkier than average. And no, I’m not sharing a video!

Two photos of a tan dog with a  black muzzle and tail pressing up against a woman's feet and legs. The woman is sitting in a chair and the dog is walking under her legs in one photo, and backed up and pressing into her feet in anther
Clara enjoying pressing against my feet and legs

The added benefit of this one is a little harder to describe, but no less real. Clara is a very “touchy” dog. She likes to lean against me, touch me, cuddle, and be as close as she can. So she loved the leg weaves. She got to be right “inside” my personal space. And darned if she didn’t make up a new game: she comes and weaves her way through my legs when I am sitting down, just for fun. Kind of like a very large, pushy cat. She clearly likes the sensation.

I couldn’t get a shot of the actual weaving when I was sitting down, but here she is walking under my leg and pressing against my foot. See how she is pushing toward me in both photos?

So Clara and I have not only discovered a new way to play one-on-one that needs no  toy or prop.  With a little finesse, I could even use it as a reinforcer. But right now, it’s just another way to have fun with my dog.

So thanks Marge, for urging me to train pure “tricks,” but they keep turning out to be useful! Or was that part of what you were trying to show me all along….?

Coming Up:

  • Punishment is not a Feeling
  • Why Counterconditioning Didn’t “Work”
  • How Skilled are You at Ignoring? (Extinction Part 2)
  • What if Respondent Learning Didn’t Work?

Eileenanddogs on YouTube

 

 

Goodie or Doodie? When a Classical Pairing Gets Stomped On

Goodie or Doodie? When a Classical Pairing Gets Stomped On

Priiiiiiingggggg!

I use WordPress.com to host this blog. It has a smartphone app. The app is most useful to me for checking statistics and getting notifications.

The app has a pleasant little sound effect. You can assign it to sound as a notification when different things happen on the blog. I didn’t understand the nuances when I first got it. Continue reading “Goodie or Doodie? When a Classical Pairing Gets Stomped On”

But I Want to Use All the Tools in the Tool Box!

But I Want to Use All the Tools in the Tool Box!

Have you heard the one about the toolbox?

This is the sixth in a series that details and rebuts fallacious arguments against force free training. Today’s erroneous argument is directly related to “But Every Dog is Different!” in that it claims that training without punishment (certain “tools”) is just too limiting. But since it centers on the potent metaphor of the toolbox, I’m treating it separately.

OK, about that metaphor. Naturally, I went and got a public domain graphic of a toolbox for an illustration. Here it is.

Toolbox
Credit: Wikimedia Commons

Thought experiment. Is this really a good metaphor for our methods when training dogs? Tools for cutting, prying, banging, twisting? Yipes!

This led me to do a little research. It may well be the great trainer and author Terry Ryan who originated the toolbox metaphor with her two books, “The Toolbox for Remodeling Your Problem Dog,” and “The Toolbox for Building a Great Family Dog.”

The blurb for the “Remodeling” book says,

The toolbox is a set of principles and practices you can use to analyze and address any behavior problem you encounter.

She goes all the way with the building metaphor in those books, with chapters on raw materials, building supplies, the foundation, etc. But I think something really interesting has happened.

Ms. Ryan’s metaphorical toolbox was a “set of principles and practices.” The metaphor caught on, but started to change a bit. The expression came to mean a set of training methods. Get the difference? That’s a little narrower and more concrete. Methods rather than principles. And finally, nowadays for many trainers, the tools they are referring to are often, well, physical tools. Gear including correction collars of various sorts.

I personally may be done with using that metaphor–if only I can think of something else!

Who’s Got a Bigger One?

But on with the discussion. First, I would wager that the average trainer who relies mainly on positive reinforcement and negative punishment already has a much bigger toolbox than someone who uses aversives. The aversive “tools” (for instance prong and shock collars) are pretty one-dimensional. Not that they can’t be used with more or less skill. Sure they can. But as I’ve mentioned before, it takes no special expertise or devotion to figure out how to hurt an animal. Countless Joes and Joannes on the street have figured out how to do it.

But trainers who seek to use positive reinforcement would already have a big toolbox to begin with, and if they found the dog difficult to motivate, the onus would be on themselves to expand it. They would be working hard to find all possible wholesome motivators for their dog, with the goal of getting reliable behavior with a happy dog, free of fear and threats. So this image of the big gleaming toolbox with certain wondrous tools “off limits” for force-free trainers is not accurate.

Please see “But Every Dog Is Different!” for an expansion of this point.

Are you back? Great!

See if you think the following logic holds.

(Almost) Everybody Has A Limit

I propose that the problem with a force or balanced trainer saying, “I don’t want to limit myself to only certain parts of the toolbox like you do”  is that in almost every case there is another trainer further down the line who can say that to them.

I put forth that almost every trainer knows about some aversive techniques they will not use. It is not only the force free trainers who ignore certain sections of the huge fictitious toolbox. Most trainers have their limits.

I would guess that most balanced trainers would not use Koehler’s method to remedy digging: filling the hole with water and holding the dog’s head under. Neither would most use a method I read about in a bird dog training book on teaching the dog to hold steady:

“You’ll need your checkcord, a choke or spike collar, and an assistant who is strong enough and willing to jerk a dog over backward with the checkcord at the proper moment…  When he hears the shot and sees the bird fall, the dog will break. Don’t say a word, and be sure your assistant remains silent but braces his feet and gives the dog the somersault of his life when he hits the end of the check cord.”–Ultimate Guide to Bird Dog Training, Jerome Robinson

You certainly could find people who still do these things.  A quick perusal of YouTube can show these methods and worse. But even for those trainers–there might be a method they wouldn’t employ.

Next time a trainer says they want “access to every tool in the toolbox,” try asking them whether there is any method in the world they wouldn’t use. If they need prompting, you could name some. The point is not to be aggressive about it.  The point is that you might get across that it is not at all “limiting” to avoid methods that don’t fit into your ethical stance. Especially when your “toolbox” is gloriously full and varied already.

If someone uses the toolbox metaphor to you, I think you could get in some really interesting discussions if you asked them why they don’t employ certain methods. Doesn’t the same metaphor apply? If it breaks down, then why are they saying it to you?

A Different Metaphor?

Like I said above, the toolbox metaphor is much too entrenched to fight. But for me, the more I think about it the less it fits. Besides the harshness of the idea of applying hardware to dog training, the whole tool thing fails to highlight the mutual learning that goes on between a trainer and a dog. I’m really not as interested in sculpting my dog into some ideal as much as I am interested in that magical partnership that is born when we learn together.

So I thought about it. I tossed aside “cornucopia,” “toybox,” and “treasure chest.” My new metaphor is a “bag of tricks.” And it’s a bag that my dogs and I can both open.

Photo credit: Kenneth Rivenes. Thanks Kenneth!  http://www.flickr.com/photos/sprengstoff72/3294941579/
Photo credit: Kenneth Rivenes. Thanks Kenneth! http://www.flickr.com/photos/sprengstoff72/3294941579/

This post is part of a series:

Thanks for reading! Coming up:

Eileenanddogs on YouTube

Force-Free Training and the Continuum Fallacy: Defining Ourselves

Force-Free Training and the Continuum Fallacy: Defining Ourselves

Although this post is about discussions and accusations about humane training, it doesn’t provide fodder for pithy sound bites or snappy answers. The whole point of it is why it can be difficult to explain succinctly our position as science-based, humane trainers in the face of opposition.  I hope it can be helpful for some folks. Gathering information, thinking this through, and writing about it has settled my nerves about a lot of things regarding the conflicts between trainers. Here we go.

Here’s something that force-free trainers hear a lot:

“There’s no such thing as force-free training because…”

  • “You use leashes and that’s force, the same or worse than a shock collar”
  • If your dog ran out into traffic you would grab him or pull on the leash”
  • “You all use force too, you’re just hypocrites about it”
  • “Harnesses are more cruel than prong collars”

Here Comes the Continuum Fallacy 

Color spectrum, from left to right (in order of frequency): red, orange, yellow, green, blue, indigo, violet ROY G BIV
There is no such thing as green because we can’t say precisely where the green starts. Really? (credit Wikimedia Commons)

I have previously written a bit on the continuum fallacy, in But What if Your Dog Runs Out Into Traffic? I wrote:

Likening pulling on a leash (in an emergency no less) with the habitual use of a shock collar to force a dog’s compliance is an example of the logical fallacy called the continuum fallacy… The continuum fallacy erroneously claims that because there is a range of possibilities between two extremes, there is no meaningful difference between them. In this case the extremes are pulling on a leash one time to remove a dog from danger, and using a negative reinforcement protocol with a  shock collar as a training method to teach them via force to come when called…

The examples cited above both employ negative reinforcement, or at least aversive pressure (we can’t really say if reinforcement occurred in the emergency situation since it’s a one trial example). Therefore there is a continuum of such usages between them.

More commonly the extremes cited are two types of training:  training based as much as possible on positive reinforcement (along with desensitization and classical conditioning), and training based almost entirely on negative reinforcement, positive punishment, and flooding, such as shock collar training.

You wouldn’t think there was any common ground between these two. But there is, or rather, there are intermediate states between them. Please bear with me if the idea offends you. My point is that they are absolutely different in essence, even though there exists the continuum.

So on the continuum,  next to the trainers who use shock exclusively are the ones who do use some food or play as positive reinforcement in addition to the shock used aversively. Next to them are the “balanced” trainers who combine positive reinforcement with “corrections.” Next to them are the ones who use a shock or prong collar for proofing only. Then the ones who use a shock collar in only one specific situation. And so on. (These could be split more finely of course.)

Going past the middle to the R+ paradigm side are the folks  who don’t intentionally use any aversive tools, but carry old habits (Eileen raises hand). We occasionally do something that is aversive to the dog, for instance, taking a step forward to apply pressure if a dog breaks a stay. We do this because of old habits or lack of knowledge of other ways, but no matter why, it’s still aversive to the dog.

I’m not going to describe every step from here on out. But we can travel farther and farther into R+ territory as other methods drop away.  But truthfully, most people don’t get to the point of never using negative reinforcement or negative punishment or extinction. As I’ve mentioned, that necessitates an almost godlike ability to predict every possible behavioral interaction if you live with your animal. And even if we consider only formal training sessions, it depends on the dedication and creativity of the trainer to unlearn our human punishment programming and get more and more fluent in humane methods.

Because of the infinite gradations between the two extremes, there are those who would argue you can’t make a distinction between them. They would be employing the continuum fallacy.  This link has a good definition and some nice examples of it.

One common application of the continuum fallacy is to claim that the concept the other party is describing does not even exist. Does that sound familiar? Punishment based trainers, particularly shock collar trainers, like to claim that there is “no such thing as force free training.” Since we use things like leashes and collars that are naturally agents of force (although we take pains to ameliorate that), and because some aversive situations are just going to occur in life, they claim that there can be no distinction, none, made between what we do and what they do with their specifically-designed-to-hurt tools. This argument is incorrect (especially when they throw in the straw man of “purely positive”), and a way of trying to talk us out of existence. I’ve written an entire post on it.

Now here’s where it gets really interesting. The continuum fallacy is connected to something called an “open concept.”  This really enlightened me about why it’s so hard for us in the force free community to come up with a single name for ourselves, and why we are repeatedly told we don’t exist. What we are trying to describe is a complex combination of a training philosophy, methods, and a mesh of practices. “The kind of trainers we are” is an open concept.

Open and Closed Concepts

So many things fell into place when I read about this.

A closed concept is something that can be exactly defined, such as a triangle. But many of the most important things in life can not be exactly defined. From “Open and Closed Concepts and the Continuum Fallacy” by Sandra LaFave:

An open concept is one for which the connotation cannot be precisely specified; rather, we recognize members of the class by their resemblance to paradigms of the concept.

Here are some examples. Vegetarian. Christian. Pacifist. Have you ever heard someone arguing about the definition of any of these or over who belongs to the group? I thought so. Yet the various individuals who identify with these terms can define their habits and belief systems beautifully, and they are often at the core of the person’s identity.

A portrait in pencil of Ludwig Wittgenstein. His face is angular and he looks intense and pensive.
Ludwig Wittgenstein worrying about open concepts (credit Wikimedia Commons)

The philosopher Wittgenstein (Austrian, 1889-1951) wrote about open concepts. His example was the concept of “game.” He advised the reader to think of different games and to try to think of what was common to them all. (My suggestion: use the examples of patty-cake, football (American or world), board games, and the often deadly games played in the Roman Coliseum, and try the exercise.) He wrote that you cannot identify one single characteristic common to all examples of games.

But that doesn’t negate the concept of game. He analyzed the similarities and differences in several types of games and concluded:

And the result of this examination is: we see a complicated network of similarities overlapping and criss-crossing; sometimes overall similarities, sometimes similarities of detail. –Wittgenstein  in his Philosophical Investigations as quoted by Donald Palmer in “Does the Center Hold?”, p. 394

Glupling Training

As I mentioned, I think the lack of a commonly agreed upon name by all in the community is one bit of evidence that “the kind of training we do” is an open concept. For that reason, for brevity, and to introduce a little levity into a heavy subject, I’m going to call force-free, science-based, humane, primarily positive reinforcement training “Glupling Training.”

It would be easier to live in a world where we could say, “If you do these five things, and don’t do these five things, then you are a bona fide Glupling trainer.” Nice clean line in the sand. But we don’t live in that world. Glupling training is a philosophy; a group of methods; a paradigm. I strive for it. I think most of you out there reading this are striving for it. I’ve got certain great trainers and thinkers in mind as my role models and perhaps you do too. But we need to acknowledge that the edges of the definition are not universally agreed upon.

For instance, within the Glupling community there are heated discussions about head halters and front attach harnesses, and whether these are OK as permanent solutions, temporary management aids during training, or never OK. People disagree about the use of Behavior Adjustment Training (BAT) and Constructional Aggression Treatment (CAT). Some people seek to be two-quadrant trainers (positive reinforcement and negative punishment). For some trainers negative reinforcement protocols are OK in general, others go case by case, still others try for “never.”  How about No Reward Markers? Or whether it’s OK to yelp as a training technique when a puppy bites you?

I have watched other groups in similar throes of self definition. Organic gardening discussion groups talk about whether the use of Bt (Bacillus thuringiensis) disqualifies someone from the group. “Childless by choice” people argue whether people who never had kids of their own but then marry people with children still “count.” There is discussion about whether the Americans with Disabilities Act applies to people who are or have been dependent on alcohol. This is just a completely typical situation with open concepts. I would even argue that these discussions can be healthy, as long as they don’t start to eat up your life.

The bigger troubles start with the people outside of the particular community who object to it. With regard to Glupling training, these are the folks who want force-based methods to be socially acceptable and so attack the Glupling paradigm. These folks have two main strategies.

  1. Some trot out straw men and the continuum fallacy and simply claim that Glupling training does not exist. These are the ones whom I describe in my blog post, “But Purely Positive is a LIE!
  2. Others jump onto the Glupling bandwagon and market themselves as Glupling trainers, prong or shock collars included. This method has the effect of diffusing the definition of Glupling training and confusing the public. You can find lots of folks on the Internet who salt the word “positive” throughout their website, even if they use physical dominance techniques, prong collars, or shock. These folks accomplish several things by adopting the term, “positive.” First, it is popular right now and it sounds wholesome and good. Second, they add to the confusion (some groups have actually created credentials and initialisms that are identical or similar to established organizations). Third, they help maintain the public’s confusion about the processes of operant learning, since “positive” in the behavioral sense absolutely does not equal “wholesome and good.” And fourth, as added by an early reader of this post, who would want to market themselves as someone who will throw things at your dog and yell “bah”? “Positive” sounds much nicer, doesn’t it?

One observation I have about these continuum fallacy arguments: it seems to always be the side with the less restrictive definition that is arguing that the other side doesn’t exist, not the other way around. Vegetarians never argue that omnivores don’t exist. Organic gardeners never argue that gardeners who use non-organic techniques don’t exist.

This disagreement is typified by a group of people (or an individual) seeking to distinguish themselves from others who are simultaneously trying to negate the distinction.

Static vs Dynamic

I said above that it would be nice to live in a world where Glupling training was easy to define. But actually…one of the hallmarks of Glupling trainers is that we are always using the science to find ways to be more humane, more fair, and better trainers for our animals. The research moves us forward.  So perhaps two of my (fictional) five things that might have defined Glupling training in 1998 are completely out of date in 2013. But that’s a good thing. Given a choice between an approach that is static and claims to know everything and be perfectly complete and definable, and one that allows room for growth and speculation and doesn’t claim to be perfect this very instant….well, you know which one I would choose.

By the way, that is one of the reasons I keep my hand in the discussions and arguments on the Internet. I learn stuff that way.

Conclusion

I have recently written a handful of posts with a deliberate intention of publishing talking points for Glupling trainers who are confronted by the same rhetoric from force-based trainers over and over. The posts are listed below.

I had hoped for this post to join that group, but I’m not sure how  helpful it is. It has been very helpful for me as I mentioned above because it has clarified some difficult things in my mind. Like, why do these fights keep going on and on? But this post is not the kind of thing a person can quote in an argument and say, “Hah! Read this! It proves my point!” Not even close.

But I will throw in some tips on dealing with the continuum fallacy when confronted with a version of it in debate. Dr. LaFave suggests a simple statement that even if there may be a continuum between extremes, the concepts at each end are meaningful. I mean, nobody really believes that black and white are the same because there are shades of gray in between. And she suggests that people who use words in eccentric ways (my example: like shock collar trainers who say the method is positive and force-free) should be called on to defend their eccentric usages of these terms and give good reasons for them.

And this from me: When I have encountered the continuum fallacy, in my observation it has not usually been an innocent misunderstanding. It is usually from someone who, in my opinion, is determined to obfuscate. If I find that to be the case, I will state my position once, if at all (and for the benefit of others who may be reading, and not with the hope of convincing the other person) and move on. In short, as my grandmother used to say, “Don’t argue with someone you have to educate.”

This post is part of a series:

Coming up:

Eileenanddogs on YouTube

But We Don’t Give Our Kids a Cookie Every Time they Tie Their Shoes!

But We Don’t Give Our Kids a Cookie Every Time they Tie Their Shoes!

Here’s another remark often addressed to reinforcement based trainers, sometimes in a mocking tone, sometimes seriously:

A woman's hand is suspended over a clear glass cookie jar. The jar is full of Vanilla Wafers, a small, disc shaped light brown cookie. The hand is holding a cookie (has just pulled one out of the jar). But we don’t give our kids a cookie every time they tie their shoes or pay them a nickel every time they say thank you!

The writer often further implies that to do that with children would be the worst sort of bribery, indulgence, and permissive parenting (and-by-the-way-it’s-responsible-for-all-the-current-evils-of-society). And we’re being just as weak willed when training our dogs!

But the “cookie” objection is so easy to address. Continue reading “But We Don’t Give Our Kids a Cookie Every Time they Tie Their Shoes!”

Crossing Over

Crossing Over

A small dog, a black and white rat terrier with very large ears that stand up, is running towards the side of a human (you can see only the human's pant leg. The dog's mouth is open, her foot is raised in mid stride, and she looks excited and happy.
Cricket performing a recall in a training session after I crossed over

I am a crossover trainer, and we tend to like to tell our stories.

“Crossing over,” and “crossover trainer” refer to a trainer who switched from punishment based, or mixed training that included aversives, to training that is centered around positive reinforcement and avoids force, pain, and actions intended to “dominate” the dog.

I have written the story of my transformation, and Ines Gaschot is graciously hosting it at her blog, “The Crossover Trainer.” She has lots of interesting crossover stories over there as well as many other lovely blog posts. I hope you’ll not only go over and read mine, but check out the rest of her blog and website too if you haven’t already.

So here is how to get to my crossover story, complete with an embarrassing picture of Summer and me graduating from our first obedience class in 2006. (She didn’t much like the mortarboard.)

Crossing Over: A Pet Owner’s Story

Thanks for reading!

Coming up:

Eileenanddogs on YouTube

Introduction to the Exercise Ball

Introduction to the Exercise Ball

Small black and rust colored hound dog is putting her front paw on a red exercise ball (peanut shaped). Her mouth is open, anticipating a treat.
Zani’s ready for a treat for foot targeting the peanut

I bought an exercise ball, a FitPAWS peanut, from CleanRun a couple of years ago. It’s a device to help dogs develop core strength and balance.

After seeing some YouTube videos and even a professional DVD that showed dogs and puppies being placed on exercise balls and held there while they were clearly stressed and uncomfortable, I decided to make a video showing how I introduced my dogs to the ball. We went comparatively slowly, over the course of a few days, with no force or pressure. I wanted my dogs to have a great association with the ball and no anxiety attached to it. So from the very beginning they always had a choice; they could walk away, jump off, take a break.

As is typical, giving them choice in the matter and building good associations made them absolutely fanatically fixated on getting on the ball! And once more, going slow turned out to be fast!

Small black and rust colored hound dog has both  her front paws on a red exercise ball (peanut shaped).
Zani has her whole front end up on the peanut!

You can see in the short video that I used a combination of shaping, targeting, and treat placement to get Zani happily on the ball in a few daily sessions. This method can be used to introduce a dog to all sorts of unfamiliar objects and equipment.

Small black and rust colored hound is standing on top of a red exercise ball (peanut shaped).
And she’s up on the peanut!

Zani’s a confident little dog and I probably could have done it all in one day, but 1) I wanted to take no risks of rushing her psychologically; and 2) we are dealing with a physical skill that builds muscles, and I didn’t want to overdo.

If you are considering getting an exercise ball for your dog, be sure and check it out with your vet. Also, size the ball correctly (CleanRun and the ball vendors such as  can help with that). I hope your dog enjoys it as much as Zani does.

I like easy ways (for me!) to exercise my dogs. Don’t forget flirt poles, too!

Link to the video for email subscribers.

Thanks for watching!

Coming up:

Eileenanddogs on YouTube

The Humane Hierarchy, Part 2 of 2: Examples

The Humane Hierarchy, Part 2 of 2: Examples

This is the second of two posts on Susan Friedman’s Humane Hierarchy. Here is Humane Hierarchy Part 1 in case you missed it.

In this part, I present examples of each of the methods listed in the Humane Hierarchy. My examples all center around crate training.

Here is the Humane Hierarchy again so we’ll have it handy.

A graphic that shows 6 levels of behavioral intervention, starting with the least invasive at the bottom, going to the most invasive at the top. The graphic looks like a road going straight ahead, with a right turn for each behavioral intervention. They are, in order: Health, Nutrition, and Physical Setting; Antecedent Arrangements; Positive Reinforcement; Differential Reinforcement of Alternative Behaviors; Extinction, Negative Reinforcement and Negative Punishment; Positive Punishment.
 

And here is a link to a different version of the Humane Hierarchy graphic that may be visually easier than the roadmap version.

Examples!

Remember, these interventions run from the least intrusive first, to the most intrusive last.

Intervention 1: Health, nutrition, and physical setting,  This means to check for physical reason for a behavior first, either a physical problem with the animal or something environmental that is affecting her.

Behavior 1: Dog just stands there when you ask her to go into her crate. Your old dog seems to have unlearned her crate behavior. Instead of going in eagerly when you cue it, she stands there licking her lips. She resists when you try to lead her in. You take her to the vet. It turns out that her vision is impaired. There is a glare coming off the stainless steel water bucket in her crate and it is scaring her. Your intervention: get a plastic bucket (and maybe a plastic crate).

Small black and white rat terrier with very big ears is lying down inside a wire crate with the door open.
My old dog Cricket in a crate

When considering a problem behavior, checking for a health-related reason should be the first step. This doesn’t apply only to old dogs, either!

Just think if you had tried to retrain the behavior, even with positive reinforcement. You would have had an apparently “stubborn” dog. Even worse, what if you had punished her?

Here is a beautiful video by Sonya Bevan of Dog Charming that shows some  “mis-behaviors” by dogs with some very interesting causes, including at least one that has to do with the physical environment: “There’s always a reason dogs do what they do.”

Intervention 2: Antecedent Arrangements. 

Antecedents are those stimuli, events or conditions that occur immediately before the behavior, which function to set the occasion for the animal to exhibit the behavior. — Susan Friedman.  A framework for solving behavior problems: Functional Assessment and Intervention Planning. Journal of Exotic Pet Medicine. 16,(1) 6-10.)

Cues are antecedents that we teach deliberately, but antecedents are happening in your dog’s life all the time. Antecedent arrangement means that sometimes you can deal with an animal’s unwanted behavior by changing what comes before it, rather than the consequences that come after it, as we do via the more familiar processes of  reinforcement and punishment.

Behavior 2: Puppy whines in crate. Your puppy’s crate is in the dog playroom. Your other dogs are loose in another part of the house. As part of the process of taking them outside when you get home, you let the other dogs into the playroom while the puppy is still crated. The puppy whines and screams in excitement when the others come in. Then you are in a quandary. Let the puppy out while he is whining? If so, you would probably reinforce it.  But what if he has to go to the bathroom?

The antecedent in this case is the entry of the other dogs. This precedes vocalizing by the puppy. The noise making might be OK in other circumstances, but whining and screaming in the crate cause problems. Here are three possible antecedent changes that could solve this problem:

  1. Complete elimination of the antecedent: Take the other dogs outside through another part of the house. Then go get the puppy separately to take him outside.
  2. Change puppy’s location during the antecedent: Let the puppy out first. Either take him outside, or let him be loose in the room when the older dogs come in. He may get excited and vocalize, but this doesn’t put you in the quandary that it does if he is in his crate.
  3. Change puppy’s location so he is no longer present for the former antecedent: Move the puppy to another part of the house and take the big dogs out through the dog playroom first, then release him to join them.

Any of these should solve this particular instance of whining in the crate without having to reinforce or punish anything, or train anything at all.

OK, here come the operant learning processes with which many of us are familiar. If you need a brush up, please see my blog post Operant Learning Illustrated by Examples, or go straight to my movie: Examples of the Four Procedures of Operant Learning.

Intervention 3: Positive Reinforcement. Something is added after a behavior, which results in the behavior happening more often.

Behavior 3: Dog goes in crate and stays there. This is something you want to teach your dog. To do so using positive reinforcement, you could use any of these three methods of training: luring, capturing, or shaping.

  • You could leave good stuff in there for him to find (luring).
  • If he went in there on his own, you could immediately mark and reinforce (capturing going in).
  • If he is in the crate and being quiet, you could drop him a treat or chewable as you go by (capturing quiet stay in crate).
  • You could play training games where you shape him to go into his crate from different areas of the room (shaping).

Brace yourself for inordinate cuteness in the video.

Link to the video on capturing crate behavior for email subscribers.

Intervention 4: Differential Reinforcement of Alternative Behaviors. This means a replacement behavior is (positively) reinforced while the unwanted behavior is extinguished (see extinction below).

Behavior 4: Your dog goes in her crate when visitors come (instead of leaping on them). This is something you want to teach. Your adolescent dog loves everybody and is thrilled when someone comes to the door. She jumps all over them. This is not your preferred way for her to greet visitors.

You start by training your dog to go to her crate using positive reinforcement, without visitors present. You train it really well until she is absolutely thrilled to go to her crate and runs top speed when cued.

Then teach her that the doorbell ringing is a cue for her to go to her crate. After this cue is very solid, you start practicing with people coming in, but not in real life yet. Use setups.

You will not get extinction of the jumping on people unless it ceases to be reinforced, so you will also take some management measures. For the beginning period you might keep an ex-pen around the inside of the doorway in case your dog makes a booboo and runs for the door like before. She still can’t get to the visitor and practice jumping.

For practice setups, you must train your visitors. You need them to absolutely ignore your dog if she does get to them and jump on them. This is the removal of the previous reinforcement for jumping up, which is generally human attention. But it’s best to try to avoid the situation entirely, because some dogs enjoy jumping even when the human is ignoring them.

A rule of thumb is that the reinforcement of the new behavior has to be more potent, or at least as potent, as the original reinforcement. So the finishing touch will be to teach your dog that after she has gone to her crate, she will sometimes be released to visit (if she enjoys that). She can calmly visit with the guests and get human attention as long as she has four feet on the floor. You will have to train that as well.

Here is an example of differential reinforcement of an alternative behavior where I taught Clara to lie down when I bent over, rather than mugging my face.

Intervention 5: Extinction, Negative Reinforcement, and Negative Punishment. Dr. Friedman does not give a hierarchical ranking order within these three. The degree of humaneness will depend on the application of each method and the individual animal.

Intervention 5a: Extinction. Extinction of a behavior occurs when the consequence that was previously reinforcing the behavior is permanently removed.

Behavior 5a: Puppy barks to be let out of crate at night. When you first got your puppy, sometimes when you were late letting him out to potty in the night he would give a little bark to wake you up. You would immediately get up and let him out to potty. As he got older you got tired of this. You were sure he really didn’t need to go. He would bark and you would stay in bed. So he barked longer. Finally when you couldn’t ignore it any longer, you would let him out. This has been going on for some time.

You get on the Internet to see how to get the dog to stop barking. Someone writes that you just have to outlast him. So the next night when he starts barking, you ignore him. And ignore and ignore and ignore. When he finally gives up and is quiet for a minute or two, you may let him out.

This scenario demonstrates the drawback of using extinction by itself. This situation is a mess. It’s horribly unfair to your dog, who may really need to go to the bathroom and is trying his best to tell you say in the way that was previously reinforced. His world has turned upside down and what used to work beautifully fails. Your dog has no clue now how to get out to potty. You waited until he was quiet to let him out, but you can’t use “being quiet” as a cue to be let out if he is quiet most of the night. Unless you want to start a behavior chain of: make noise, be quiet, get let out.

This is one of the reasons why using extinction alone is “farther down the road” than Differential Reinforcement of an Alternative Behavior. In that case, you are deliberately developing and reinforcing a new behavior to take the place of the old. The dog gets a big fat clue about what to do instead.

Intervention 5b: Negative reinforcement. Something is removed after a behavior, which results in the behavior happening more often.

Behavior 5b: Puppy stays in the crate. You are teaching your puppy to stay in the crate when you tell him to, without your closing the crate door. You put your puppy in and tell him to stay. He stays for a few seconds, then gets up and heads out the door. You get there first and keep walking forward, walking into his space and pushing him with body pressure until he backs up back into the crate.

Negative reinforcement uses an aversive, something the animal does not like. Because of that it can have fallout. My movie, Negative Reinforcement vs Positive Reinforcement, shows the difference in my dogs’ behavior when trained the same behavior with those two methods.

Intervention 5c: Negative punishment. Something is removed after a behavior, which results in the behavior happening less often.

Behavior 5c: Puppy whines in crate. Your puppy is in her crate. You enter the room and she starts to whine in excitement. (She has never done this before.) You immediately turn on a dime and leave the room.

Intervention 6: Positive punishment. Something is added after a behavior, which results in the behavior happening less often.

Behavior 6: Puppy runs out of the crate door when it is opened. Your pup has developed an unnerving habit of dashing out the crate door as soon as you open it. So you decide to show him who’s boss. You get a spray bottle of water and add some lemon juice. You walk up to the crate, open the door, and squirt him in the face as he tries to dash out.

A stuffed brown and white dog is positioned emerging from a dog crate. There is a hand and arm emerging into the photo from the other side. The hand is holding a squirt bottle and it is aimed at Feisty's face.
Feisty gets sprayed as she darts out of the crate

This demonstrates the many drawbacks of positive punishment. First, it may not be absolutely clear to the dog what he was squirted for. Looking out? Crossing the threhold? Whatever happened next?

Perhaps you haven’t even taught him the proper behavior that you do want, such as to sit quietly in the crate until released. So the next time you open the crate door, your dog may be afraid to come out at all. Or afraid whenever he sees the squirt bottle. His affection and trust for you may wane, since it was abundantly clear that it was you who were squirting him with the painful stuff. His anxiety level has probably shot up from the whole experience. What’s going to happen next time?

This scenario also illustrates what Dr. Friedman calls the “double whammy” of positive punishment. First, the dog didn’t get the consequence he was seeking: getting out of the crate. Second, he got squirted painfully in the eyes. And as Dr. Friedman wrote in the article that introduced the Humane Hierarchy,

Positive punishment is rarely necessary (or suggested by standards of best practice) when one has the requisite knowledge of behavior  change and teaching skills.

And she has kindly arranged a list for us of seven other things to try first!

Thanks for reading! Coming up:

Eileenanddogs on YouTube

The Humane Hierarchy, Part 1 of 2: Overview

The Humane Hierarchy, Part 1 of 2: Overview

I am a Humane Hierarchy trainer. That is the name of the roadmap I use to make ethical choices about the training methods I use. I’m going to describe the method in this post.

(Humane Hierarchy Part 2, which is now also published, comprises real world examples of all the methods in the Hierarchy.)

I don’t call myself a clicker trainer, although I have used one, nor do I call myself a force-free trainer, although that is certainly a goal, nor do I call myself “all positive,” since that could include positive reinforcement and punishment both. I do use the first two terms, along with several others, to refer casually to trainers who use those names and have similar goals to mine. The people who use these terms are part of my community.

But the Humane Hierarchy is a concept I love, and a name I take on for myself comfortably and with pride. And I was born a non-conformist, and throw off labels as fast as anyone can put them on me. But this one I’ll take. Because it’s a non-label of a label. You’ll see.

Susan Friedman, PhD, published “What’s Wrong with this Picture: When Effectiveness is Not Enough” in 2008, and in that article proposed the Humane Hierarchy. The article is about incorporating ethics into the choices we make when training animals, rather than considering only “what works.”

I have written about Dr. Friedman frequently. She is a behavior analyst and strong proponent of humane, ethical treatment of all animals. Here is my review of her course on Living and Learning with Animals, and here is her website, Behaviorworks.org. Be sure and check the free articles.

The Humane Hierarchy is not a set of “rules.” It is a general ranking of training methods, starting with the least intrusive for the animal and ending with the most intrusive. Least intrusive is defined as the procedure that leaves the animal with the most control over its outcomes. Any person who uses the Hierarchy as a guideline must inform herself about the species of animal she is working with and carefully observe the behavior of the individual animal, because different animals will respond differently to different methods.

Dr. Friedman takes behavioral intervention seriously. It is a large responsibility to intervene in the behavior of an animal, and her approach directs the user to consider the animal first: its needs, wants, likes and dislikes. What does the animal want, and how can we figure out if there is an acceptable method for it to get it? It’s only fair, since in all cases we are the ones with the keys to the cabinet, the cage, the car. But that’s a pretty radical concept for a lot of people.

So here is her new graphic of the Humane Hierarchy. To use it, think of a behavior of your animal that you might want to change. Then start at the bottom of the picture, in the little car, drive forward very slowly, and take every right turn. If the consideration on the side street is irrelevant or doesn’t work when tried with full information and skill, you can drive forward again and take the next right turn, or consult a trainer or colleague. Note the stop sign before you get to positive punishment!

A graphic that shows 6 levels of behavioral intervention, starting with the least invasive at the bottom, going to the most invasive at the top. The graphic looks like a road going straight ahead, with a right turn for each behavioral intervention. They are, in order: Health, Nutrition, and Physical Setting; Antecedent Arrangements; Positive Reinforcement; Differential Reinforcement of Alternative Behaviors; Extinction, Negative Reinforcement and Negative Punishment; Positive Punishment.
 

Here is a link to a different version of the Humane Hierarchy graphic that may be visually easier than the roadmap version.

In a previous post, “But Every Dog is Different!,” I hope I showed that the claims that trainers who avoid force are somehow employing a cookie cutter method or limiting themselves are wrong.  This graphic makes it explicit. But the speedbumps, caution sign, and stop sign warn us to take care as we reach the more intrusive actions. The path a person will take will be absolutely different with every animal she trains.

The fact that no procedure is ruled out does not mean that for me personally, and I dare say most people who use this roadmap, that certain commonly used tools are under consideration. If I ever did get to the positive punishment turnoff, unlikely in itself, you can be pretty sure I would not be strapping something around my dog’s neck to administer it. I would be consulting a professional who did not use such tools.

Someone who habitually shoots up to the end of the road with only a nod in the direction of the other turnoffs is showing their limitations. I don’t mean this in a snarky way. I mean that each turnoff and its method requires care, consideration, and often creativity to employ well. As an amateur, I know that the limitations of my skill level could further endanger an animal if I tried to employ some aversives. Even professionals I know consult with colleagues before going that far down the road.

Dr. Friedman describes employing a negative reinforcement protocol with a zoo animal after other methods were tried for a **year**. In hindsight, that may seem like too long to some people, since the goal was to get the animal to enter a more enriching environment. However, any method including aversive stimuli involves risks of fallout, and the keepers were unwilling to take those risks if they were unnecessary. As it turned out, the aversive method only had to be used once, and no fallout was perceived.

Note that positive reinforcement is third on the list. The first two considerations are new to lots of people, and discussed much too infrequently in my opinion. The magic of the Humane Hierarchy is on the “most humane” end in my opinion. There is so much to be learned there.

Part 2 of this post will give an example of every method on the map, all centering around a common theme: crate behavior. So come back to read about “antecedent arrangements” if you’ve never heard the term before. [1]Added note 1/7/14: I have a whole post about antecedent arrangements now.

Labeling

One of Dr. Friedman’s foci is that labels are not useful in observing and documenting behavior. “The dog is dominant” and “the parrot is acting hormonal” tell us nothing about actual behavior. One of the skills I am always working on, and which got greatly exercised when I took her class, was observing my dogs’ behavior and working on putting it accurately into words. That’s harder than it first seems! (Again I’ll refer to the great FaceBook group Observation Skills for Training Dogs. There’s the place to go practice.)

So even though I very much support the “Unlabel me!” campaign for our animals, as a writer I really struggled with some kind of term to refer to the type of training I do! I welcomed a term for my training approach. I sure didn’t need to write a paragraph about quadrants and force and aversives and management every time I refer to my training.

“Humane Hierarchy trainer” describes perfectly what I seek to do. It’s not a rubber stamp. I don’t have to qualify it or explain away anything. I just need to define it from time to time, since it is new to some folks. Thanks again, Dr. Friedman!

Closeup of the face of a caramel colored dog. Her eyes are squinted, her facial muscles are relaxed, and her ears are back. You can barely see a hand under her neck, petting her.
A commenter on one of my movies yesterday called Summer “dominant.” Is that a useful description for what she is doing here while I pet her?

Thanks for reading!

Proceed to Part 2 of the Humane Hierarchy (examples of each method)

Afternote, 5/29/13: Because of the comments of a reader, I realized that I did not mention something important. The hierarchy applies to operant learning only. If your dog is fearful or aggressive, you will almost certainly be using classical conditioning and desensitization techniques.  In those situations, no knowledgable trainer will ever recommend that you try any aversives. Although classical conditioning usually involves food, it is not the same as positive reinforcement because there is no contingency on the animal’s behavior. The goal is to elicit a respondent reaction that changes its emotional state. Thanks for helping me clarify!

Coming up:

Eileenanddogs on YouTube

 Copyright Eileen Anderson 2013

Notes

Notes
1 Added note 1/7/14: I have a whole post about antecedent arrangements now.
Theme: Overlay by Kaira Extra Text
Cape Town, South Africa