What Does The Clicker Really Mean?

My Cattle Dog puppy, Ellie Rose, inspired this blog yesterday.  I was training her in agility and using both food and toys as a reinforcer.  I was clicking and treating for her mastering a difficult and new approximation of a behavior, and using the toy as a reinforcer to get speed and drive out of behaviors she had already learned.  She is crazy for the toy, and we were having a grand time playing together after each sequence.  Then, by mistake, I clicked her and offered her the toy. She sat down and flat out refused to take it until I gave her a treat, first.  Smart girl, she called me out on being sloppy.

When I click, I give my dog a food reward.  Every time, without fail, one to one, always.  I don’t offer my dog a toy, play, or other reinforcer after I click – I only offer food rewards after the click. Why is that so important? You have to remember that the click is a classically conditioned reinforcer, and to maintain the classically conditioned response you can’t “leave out” or mix up stimuli.  OK that’s pretty meaningless to most people, so lets break it down.

The way we “power up” a clicker, is to click, wait half a second, and treat.  We do this for many, many repetitions.  We are not trying to train our dogs to do anything, we are trying to create a classically conditioned response.  What that means is that, eventually, when the dog hears the clicker, he will have an involuntary physiological response to the click.  He will begin to relax  and drool, (his parasympathetic nervous system will kick in) just as if food were actually present, even if there is no food around.   This has huge advantages for a dog trainer.  Among other things, it’s a way of reaching in and controlling the way your dog feels – properly conditioned, a clicker makes your dog feel good, whether he wants to or not.  Furthermore, it can push  your dog back into the “thinking” part of the brain, rather than the limbic or flight or flight part of the brain.   This is where you want your dog to be when he’s learning new and complex tasks that require concentration and focus.  So, when we say a clicker is “powered up” what we really mean is we have successfully classically conditioned an involuntary physiological response to the clicker.

Now, to maintain that classically conditioned response, you have to pair the food with the click every time.  It’s not like training, where you can thin the ratio (gradually give less and less reinforcers) and get a higher response (longer duration and better performance).   Training is an intellectual function where the dog makes a choice about how to pursue or avoid consequences.  Conditioning is a different animal completely.  There are no choices, here, only conditioned involuntary responses – any time you fail to give the food in conjunction with the click, you weaken the conditioning.

So, what about clicking and giving a toy or offering play, or other reinforcers?  Sure, you can do it, but you will be weakening your nice conditioned response.  Lets think about what we are doing when we play with our dogs – we get his adrenaline pumping, kick him into play/prey drive, and try to rev up his performance.  It does not take much to realize that is at odds with the calming, thinking, state that the parasympathetic nervous system invokes when food is present.   For me, keeping the dog in the thinking brain is crucial for teaching him things,  As a general rule, only after the dog is fluent in any given behavior do I add “stimulating” reinforcers. You can’t classically condition the clicker to invoke two opposing physiological responses so I recommend, if you want to use toys or other reinforcers, use a different marker.   I use the word “yessss” for toys and the clicker is reserved for food.

Just a side note – you don’t actually have to give your dog the food each time you click. You can vary the reinforcer and create more drive by sometimes just letting him sniff the food, put his mouth on the food but not have it, give him ten pieces of food, a huge chunk of food, thowing the food, etc.  The bottom line it, food has to be involved, somehow.

This is a complicated topic and I’d love to hear your thoughts and questions on this!

 

Share
Facebook Twitter Email
This entry was posted in WPF. Bookmark the permalink.

15 Responses to What Does The Clicker Really Mean?

  1. Jane Killion says:

    Jan, great question. Remember, the definition of a reinforcer is something that increases the target behavior. The only way to know for sure if something is a reinforcer or a punisher is to measure the dogs behavior and see if it is increasing or decreasing after you give the dog that particular thing or do that particular thing to the dog. In my experience, sometimes presenting the food, even if the dog does not get to eat it, causes a huge spike in the target behavior thus is a reinforcer.

    I think this is a classic case where our assumptions get in the way. We think it’s “mean” not to give the dog the food, but perhaps from the dog’s view, he’s not “not getting” the food, he’s “getting closer” to getting the food.
    I go into detail on this in my Clean Run article – Consequences, Schmonesquencs: Understanding Reinforcers and Punishers – you can read it here – http://72.32.9.5/~janekilwa/extra/consequencesrev.pdf

  2. Jan says:

    In the ” side note” you wrote about let the dog sniffing the food. Isn’t that a negative punishment? (not getting a reïnforcer?)

  3. K.A. Krisko says:

    I just picked up my fourth rescued Australian Cattle Dog, an 18-month old blue male who was found stray. Very soon afterwards, I picked up your When Pigs Fly book! I’ve never clicker-trained one of my dogs before, but I could see this one is going to require all my resources. We’ve been clicking for 3 weeks, and I’m beginning to see some improvement. (The biggest thing is, although I had no intention of clicker training my 9 1/2 year old female, she is now clicker trained!).

    I’m using hot dog chunks, cheese balls, and pork rinds (none of which I eat!). I’m definitely beginning to see some improvement in behaviors I want (responds to name, will wait for release when I put down food), but so far I haven’t been able to replace the behaviors I don’t want (biting, jumping up, walking on counters/tables/washing machine/shelves, running away). Carter’s big reinforcer seems to be exporing new things. I’m trying to figure out how to use that as a reward???

    I realize it takes time…we’re doing 2-3 sessions per day, sometimes alone & sometimes with the other dog, and recall training while walking. I’m encouraged at the results so far – eventually I’d like to train him in agility, whether we ever compete or not…thanks for the book!

  4. karen mcnamara says:

    I was never a fan of clicker training but after studying your book I put the sessions to work with the lab puppies and my 4year old dalmatian. I could not believe how both of the dogs started working behaviors to get the clicker to make a sound. I even got Bailey the lab to pick up the box the first time we tried the exercise. We are now on heeling and the attention and loose lead work they are doing is amazing…. I am recommending it to all my daycare clients. It took Tipper the dalmatian 2 sessions to think through the box exercise but then dalmatians aren’t the most biddable creatures.

  5. Ellen says:

    I am new to all of this (never had a dog), and have read a jillion books, (including yours) trying to understand and train my reactive, active young Rhodesian/Shepherd rescued shelter dog. She’s very fast, very smart, and wants to please–and to boss around any dog near her. I had no idea how consistent I had to be in order for the training to progress–so I confused and frustrated her for months! It’s good to remember that click=food. Because she likes toys/tug so much, I was using that as an alternate reward when she didn’t immediately respond to food (even before a meal). That just muddied the picture. It got so that the clicker meant nothing, and I couldn’t get that head whipping around to me even under non-stressful situations. So I got back to basics, got someone to show me what I was doing wrong, and now I’m building progress again. It’s good to have clear instructions like this–very helpful. It’s also really helpful to pay someone to watch what you’re doing, and to correct you.

  6. Susan Ford says:

    I’m just beginning Clicker Training with my 6 month old BT; Doc. I had used Clickers with Gus, and he did have an involuntary physiological response to the click, just like you mentioned in the beginning paragraphs. During the last few weeks with Gus, when he’d have visits from our doggie friends, I’d click from another room, and Gus’s head would poke up just like a puppy. He loved to work. Doc, on the other hand will need all the Clicker Training I can do, as he’s very reactive with everything.

    Thanks for this site.
    Sue

  7. SarahMK says:

    Jane,

    I’m so pleased to see this article. I absolutely and only use food with the clicker. With my retired dog I had click=food, “yah”=tug with me, and “match”=tug with someone else.

    With my current dog I added one more marker: “et”=thrown ball. I could not agree with you more that the work of conditioning different markers for different classes of reinforcers is well worth the work when you see the results. Different handler/dog teams may have need for more or fewer markers. For example, I have a dog who launches at me for the tug (when he hears that marker). I don’t want him to launch at me if I am going to throw it, right? And it dramatically reduces the potential for confusion or disappointment during fast paced training – not only does the marker tell the dog what to expect next, it increases the likelihood that they are physiologically prepared to enjoy it! This very issue has been a topic of discussion on click-bite, a list for clicker trainers doing protection sports. This clarity with different markers for different classes of reinforcers is particularly helpful if you wish to reward a dog with food around a decoy.

    I was just reading in Kay Laurence’s newish Teaching With Reinforcement the idea to click the clicker with flat thumb (duller sounding) versus thumb tip (sharper sounding) depending on whether the treat will be tossed or whether the dog will be taking it from hand. Not the same thing, but on a similar road to removing confusion from the process of reinforcement.

    I know this is a long comment – I just think it is great that you are getting this very helpful idea out there… And I encourage any of you who are still on the fence after reading Jane’s article to give an alternate marker for toy reinforcers a try!

  8. Jane Killion says:

    Interesting, Jessica. So you don’t you notice a difference in the classically conditioned response to your various markers/rewards? Fascinating. I may be not be reading your post correctly, here, but it sounds as if you are mixing two conflicting stimuli while classically conditioning your marker. I’m curious to know – what classically conditioned (involuntary physiological) response are you getting to your verbal marker?

  9. Jane Killion says:

    Eileen, it’s something that I’ve been thinking about for years, but never formalized until recently. I name toys and use those names as markers – “ballie” is one that they all know.

  10. Jessica Obermiller says:

    I don’t use clickers, but I use very strongly trained verbal reinforcers much like clickers training. I just like the more personal feel of it.

    Using toys in training can achieve the same focus and use of the thinking side of the a dogs brain, just as well as treats. Especially in working breeds, like herding dogs, sporting dogs, etc. Toys are used in hunting training which builds on a dog’s natural instinct, and uses every bit of the thinking part of the brain! In fact, in this case toys would likely be the best motivator to hone a dog in to thinking and problem solving. Treats do have their place, and are very convenient for a fast, simple way of reinforcing. What it comes down to is finding the correct motivator for your dog and going with it. And if you do use a clicker for training you should be able to reinforce using whichever reward is the best fit for your dog.

    My lab is the same way: I had to train him to stop to take a treat, and then reinforced him with his toy. It’s a working dog thing 😉

    Just my two cents.

  11. Eileen Anderson says:

    So does anyone use discrete, separate markers for different classes of reinforcers?

    Thanks for the website and opportunity for discussion.

  12. Jane Killion says:

    Jeanne, absolutley. You raise a different point about which reinforcers to use which I agree with. As I say in my book, the customer is always right and the dog decides what’s reinforcing, not us :o). But whatever you choose to use to power up the clicker, I feel you should stay with that reinforcer or class of reinforcers to maintain the power of the conditioned reinforcer.

    You could just as easily power up the clicker on a toy, but you would be classically conditioning a different response. Instead of salivating and relaxing, you’d have a dog that amps up when he hears the clicker.

  13. Jane Killion says:

    Ah, Ayella, my plan worked :o). She’s hard to stop looking at and she’s even cuter in real life.

  14. Great article … but I have to admit I am totally distracted by how darn cute Ellie Rose is!!!

  15. Jeanne Shaw says:

    HMMMM….I agree, BUT one of my border collies will only work for play. I am building value for food through play (“you have to swallow the treat before I will give you the toy”) but really, his highest value reward is a fast tug. Does her learn more slowly? It IS slower training process becuase it just takes longer to play for a second than to swallow a treat, but I think you have to use what is MOST reinforcing to the animal and the animal gets to choose what that is.

Leave a Reply

Your email address will not be published. Required fields are marked *