|
Other Articles
• Dogs and Cats
• Let's Be Clear: Clicker Training vs. Training With
a Clicker
|
Clicker Training vs Training with a Clicker
by Gail Fisher
It’s so frustrating not having a common language! I
can never be sure when someone says they do “clicker training”
if they truly do until I ask some defining questions or see them train.
At seminars in which participants say they clicker train, sure as shootin’
as many as half are training-with-a-clicker, not clicker training. This
lack of clarity is frustrating for those who are clicker trainers, frustrating
for those who thought they were (and want to be!), and frustrating for
me as the presenter. Even as a participant at some “clicker”
seminars, I’ve been disappointed to find some highly respected, talented
people promoting themselves as clicker trainers, but are unquestionably
lure-reward-with-a-clicker trainers.
In an effort to differentiate between what is and is not
clicker training, Part I described four areas in which misunderstanding
generally occurs. While some trainers embrace some or even several of
the advantages clicker training affords, it’s important to understand
the profound yet subtle differences between approaches, to clarify precisely
what clicker training is, especially as it differs from our closest cousin
– lure-reward-with-a-clicker training. We cannot clearly communicate
what and who we are unless we reserve the term “clicker trainer”
for only those trainers who embrace the whole package.
Part I covered differences in “getting” behaviors and in timing
the reward. The other two areas in which important differences occur are
in delivering the treat to the dog and in how the trainer responds when
the dog doesn’t perform as desired.
Treat delivery
A huge advantage to using a behavioral marker such as a <click>
is the ability to manipulate treat delivery to greatly enhance and speed
training far more effectively than with any other approach. With lure-reward
training the dog must be in the lured or cued position when getting the
treat. For example, the dog is lured down and receives the treat in the
down position. In training stay, the dog must be in place to get the reward.
If he moves before getting the treat, the reward is withheld.
With clicker training, since <click> marks the moment
of success and ends the behavior, the dog doesn’t have to remain
in position to get the treat. A clicker trainer is able to select the
most effective treat delivery to speed the dog’s learning for each
and every behavior. Here’s a rundown of some of the different ways
using treat delivery following a behavior marker can greatly facilitate
and speed training:
• Re-set the behavior: Delivering the treat may be used to
get to “start-over” position. For example, when working on lie
down, the dog drops, the trainer clicks and either holds the treat far
enough away so the dog stands up, or tosses it a short distance requiring
him to get up to get it. Having eaten the treat, the dog is ready to do
it again. Rewarding in position adds an extra step of having to get the
dog out of position to repeat the behavior. In lure-reward training this
is often done by luring between two positions, such as luring
up then luring down again.
• Reward in place: A clicker trainer
can, of course, reward in position as well. For example, when working
on a duration exercise such as stay or settle, <click> marks the
end of the time frame. The dog may be rewarded in place, eliminating the
need to re-position the dog to start over. Rewarding in place can be helpful
when working on a positional behavior such as heeling, but it’s not
required. If the dog moves from position after the <click>, he still
gets his reward, as <click> ends the behavior.
• To affect demeanor: Treat delivery
can help increase animation or help calm an overly excited dog. For example,
after the <click>, rewarding in place or requiring little movement
of the dog may result in a calmer demeanor. On the other hand, skidding
a treat along the floor for the dog to chase may create more enthusiasm.
• Action is the behavior: When the behavior
is an action such as back-up or increase distance on a send-away, treat
delivery can bring the dog back to the starting point. Similar to re-setting
the behavior, the difference is that the trainer will <click> as
the dog is in motion as opposed to a completed behavior. For example,
while the dog is in the process of moving backwards, <click> tells
him that movement is correct, and when he returns to the handler for a
treat, he is ready for action again. Reinforcing movement is difficult
without a marker.
• Increase distance: The clicker trainer
can quickly build distance from the dog, for example when teaching a dog
to go to a target or when working on distance cues. Because the treat
doesn’t have to be right there at the moment of success, the trainer
doesn’t have to be either. With lure training, the treat must be
either strategically placed at the target (often a complicated maneuver
to set up properly), or distance can be achieved only slowly, in small
increments.
• Time: A behavioral marker buys time
between the behavior and the presentation of the reinforcer. For example,
when using a reward such as going outside to play, there may be a time
lag between the <click> to mark the behavior and getting to the
door, opening it, and letting the dog out. With the advantage of the behavioral
marker, the clicker trainer has a huge variety of activities to use as
reinforcers, mostly unavailable to the lure-reward trainer.
Non-compliance
The first time someone sees clicker training with an untrained
dog or puppy the gasps of disbelief and impressed “ahhhh”’s
heard during a demonstration at a workshop or seminar are universal. It’s
seemingly magic to see an untrained dog go from never having heard a <click>
to happily rushing over to a blanket and throwing himself down on it in
just a very few minutes. The first timer can’t help but be impressed
with the voodoo speed at which a dog starts offering a seemingly complex
chain of behaviors. They see it, and they believe it’s possible –
for the experienced trainer. But developing faith that this non-directed,
non-physical, non-controlled approach actually works for their dog? That’s
problematic for any beginner clicker trainer. It’s especially difficult
for someone with a background of luring, who is used to achieving automated
responses. While the robotic response is not “learning”, nonetheless
the instant gratification is highly reinforcing for the “lurer.”
In the beginning, it’s hard to have faith that this
clicker stuff really works – just be patient and let the clicker
do the communicating. It’s a natural human inclination to want to
relieve the stress of learning – to alleviate any confusion or potential
confusion the dog is feeling. This desire to actively help the dog is
the most insidious habit of human beings, especially the lure-reward crossover
trainer. When a dog seems confused, help him out. When a dog is frustrated,
help him. When the dog acts helpless, come to his aid. When the trainer’s
confidence in the method flies out the window, luring fills the void.
Faced with a dog that isn’t offering the behavior (whatever that
behavior may be), the trainer pulls out a treat and lures to show the
dog what to do. Then they try it again. Again, the dog fails to perform
the behavior. Out comes another treat, and again the dog is lured into
position. The trainer mistakenly believes that this gets the message across
– and it certainly does. But not the message the trainer is hoping
for. What the dog is learning is to do nothing until the trainer pulls
out a treat, then follow the treat to get the reward. The result is “reinforced
inactivity” or learned helplessness.
The dog’s inactivity is not a conscious decision. He
is not sitting there thinking “I know what to do, but I won’t
do it.” Rather his inactivity, his waiting behavior results in a
positive consequence.
As if that’s not bad enough, after luring a few times
in the mistaken belief that the problem was that the dog just needed to
be shown a few more times, the trainer makes a renewed commitment to try
shaping and waiting for behaviors. So she then tries to out-wait the dog
– to hold out longer in hopes the dog will offer the behavior. But
the dog doesn’t. After all, he’s already learned that he’s
getting rewarded for inactivity – waiting to follow the lure. When
the trainer holds out longer, but ultimately returns to a lure, she is
simply increasing the duration of the wait. Not only does this reinforce
waiting, it reinforces persistent waiting on a random schedule.
With or without a cue, giving in to the inclination to lure
after non-compliance is a slippery slope to what the trainer then calls
“stubborn.” “My dog refuses to obey me. He knows what I
want. But he just doesn’t do it.” Would a dog that knows what
we want, that knows he’s going to be rewarded handsomely for doing
it, still refuse to do it? He’s perfectly willing to comply when
lured. So since the dog is not unwilling to perform the behavior, is it
that the dog is “sticking it” to the trainer? Do we honestly
believe dogs do that? Of course not!*The answer is that the trainer has
trained persistent waiting.
I feel great compassion for trainers who fall into this
trap—after all, we all want to help our dogs. But it’s a false
reprieve. Helping the dog eliminates his ability to discover and learn
for himself – of far greater benefit both for the dog and for the
trainer! The good news is that as soon as a trainer recognizes what she’s
unintentionally doing and eliminates her inclination to help her dog –
taking the necessary time to out-wait the wait – training takes a
giant leap forward. And the thrill and excitement of seeing that light
bulb go off in the dog’s head is so well worth the wait.
The Semantic Plea!
So that it – the four areas of confusion that separate the clicker
trainer from other training approaches. While I am clearly an avowed clicker
trainer (a true believer – one might say a missionary), I have no
ill feelings for trainers who lure continuously to get a behavior. Or
who choose to mark the behavior simultaneously with presenting the treat.
Or who use the click as a “keep going” signal or “praise
marker.” Or who reward the dog only in position. Or who lure to help
the dog when he’s confused rather than using the clicker to allow
the dog to figure it out himself. I have nothing against people who use
clickers with other methods of training. Just please, please, please don’t
call what they do “clicker training.” It robs the rest of us
of being able to clearly communicate who we are and what we do. We are
clicker trainers.
* [footnote] Some dogs may perform
learned behaviors we might call “sticking it” to the trainer
– but that’s a trained response that comes from improper (unintentional)
training. It’s not something dogs do in the early stages of lure-reward-with-a-clicker
or even with clicker training.
|