Some people still don't get it.
Bribes and lures are not the same thing. Not the same thing! Bribing is not a training method, but a weak management strategy and one I would generally not recommend because it can really become a crutch, and just as likely, the part of the antecedent cue of the training scenario.
An example of bribing: you are ready to leave the park but cannot catch your dog, she's having too much fun and knows that if she comes to you the leash will be snapped on and the fun-park-time will end (how to avoid that is a story for another day). You've been calling her to come for ten minutes to no avail. You then produce a tasty morsel of chicken and seductively coo "CHICKEN, LUCY, CHIIIIIICKEN!" while dangling and waving a greasy morsel from your fingers. Lucy comes close enough to you to grab the chicken-y goodness and you quickly snap the leash on while she's in range. Mission accomplished? Sure, you get to go home now. Training the recall? Absolutely not. Why? Because you had to produce the chicken before the cue and then GAVE YOUR DOG THE CHICKEN. Plus the dog only comes once chicken has been produced even though you’ve had her for three years. You can almost never catch her unless you’ve got some chicken on you.
So what the heck is a lure?
A lure is a training tool that prompts the action you would like to see more of in your subject (the dog for instance) and, as a tool or training implement, must be phased out within the first training session in order to also avoid it becoming a long-term part of the antecedent cue. Ah, haaaa.... this is where it gets confusing.
Isn't that what the chicken did in the park scenario above?
It is easy to tell the difference between a bribe and a REWARD because a bribe is produced before the action while a reward is produced after the requested action has been completed, as a reward, duh.
But what separates a bribe from a LURE? People seem to have much more trouble with this one. Is there a difference? If so, where do they part company?
In short a lure is a tool, and a bribe is a cheap payoff. The difference is in the delivery. Another difference is the amount of time one relies on the tool to prompt the behavior they are trying to reinforce.
A lure is not meant to also be given to the dog as a reward. Also, a lure must be phased out and replaced by the cue within the first training session.
This may be easier to understand if we use something other than food as the reward (something I highly recommend you do with your dog as much as possible). Or, alternately don't use food as the lure, not as easy, but possible.
Park scenario II: You just adopted a new dog from the shelter; you have no personal relationship or training history as of yet. You take your new dog to an enclosed park because you haven't yet worked on training a recall. You call your dog to come and the LURE her to do so by becoming interesting. You can do this by clapping, running backwards and jolly talking, waving a favorite toy. When the dog comes over to you smile, grab her collar, and reward her for coming with a bit of greasy chicken and release her to go play some more. She learns when you call her to come it's win/win for her much of the time as she gets a tasty morsel and gets to continue sniffing and playing. Voila! You are on your way, training a recall. You prompted your dog to come over to you after giving a yet unknown and unreinforced cue and she learned that good things happen when I approach my new owner.
The problem is that so many people fail to use a lure only as a lure and not also as the reward and so many people fail to phase out the lure right away. Lure reward training done wrong may become bribery if poorly executed but is not in-and-of itself bribery. So the difference is very clear, however there is potential for misuse or over-reliance to occur, as with any training tool or sequence. Any too or training implement that remains on the handler beyond the very earliest teaching phase has become a crutch.
Don't get me wrong, it is important to continue to reward a dog for performing favorably throughout his or her lifetime, but you shouldn't have to continue to use training devices for basic behaviors into old age.
So that is the difference between lures and bribes as I understand it, but really, I'd like to hear from Dr. Ian Dunbar on this topic. He is largely responsible for popularizing lure/reward training and detractors often call this method bribing. I know Ian still believes in this particular method, perhaps he'll be kind enough to shed some light on the subject from his point of view and we can get the word right from the horses mouth, or fingers in this case....
Ian, are you out there?
- kelly gorman dunbar's blog
- login or register to post comments
- 145 reads
- back to the dog blog
- more by kelly gorman dunbar
Hi Kelly,I have a question
Hi Kelly,
I have a question about something you said. You said "a lure is not meant to also be given to the dog as a reward."
But on the lure-reward section of this site, it gives the following example: "For example: 1. Say, “Sit,” 2. Lure the dog to sit by moving a food lure upwards in front of the dog’s nose, so that 3. As the dog raises his head to follow the food, he compensates by lowering his rump to the ground and sits — the desired Response and so, 4. Reward the dog with a scratch behind his ear, by throwing Tennis Tug ball to retrieve, or simply just give him the food."
The last sentence: "...or simply just give him the food," implies that it is fine to use the lure as the reward. So I just wanted your opinion on that. I think the two critical errors people make are not transitioning from a lure to a cue (hand signal, verbal) and not incorporating many differerent rewards (i.e. they become dependent on one or two). However, I've never had any problems using the lure as the reward, as long as neither of the afforementioned critical errors are made.
Lastly, there are some people who assume it's a bribe basically any time food (or other rewards) are being used. But as I say, you cannot bribe a dog (or a human) to do something they don't know how to do (i.e. is the behavior in the dog's repertoire) and therefore lures are used to initially generate the behavior frequently enough to put it on cue.
Thanks,
Jordan
Just one more thing I
Just one more thing I noticed in regards to "a lure is not meant to also be given to the dog as a reward." For example, in the video where Dr. Dunbar is teaching Hugo to sit (http://www.dogstardaily.com/videos/sit), he is using the food to lure Hugo into the sit position, and then also using the same food as the reward once Hugo is sitting. The way I interpret this is that the lure IS being used as a reward. That being said, I see no problem with this as long as one executes the rules of lure-reward training properly.
Phasing out Lures and Rewards
Hi Jordan
Thank you for your post. Going back and rereading the Training Textbook, I see the problem. All of the information is there but in the wrong order. We start off talking about the four steps of Lure/Reward Training: 1. Request. 2. Lure. 3. Response. 4. Reward, but only later (http://www.dogstardaily.com/training/quantum-leaps) talk about the stages of training, most notably phasing out lures and phasing out external rewards. Thus, more emphasis is given to using food as a lure rather than phasing it out. Phasing out food lures and rewards is illustrated by other videos of Hugo in the Quantum Leaps section.
After reading your post, it is so obvious to me now, why people cannot train their dogs to a high level of reliability and precision via lure/reward training. They simply do not understand the process because I haven’t written clearly enough. (I can’t seem to get away from writing sequentially, rather than hitting all the most important points up front.)
Basically, lure/reward training isn’t working well for people because they are unclear: 1. How to phase out lures (within just 6–12 repetitions. 2. How to phase out training rewards and first replace them with Life Rewards and then eventually with internal rewards (self-motivation). 3. How to enforce responses without fear or force.
I will write some blogs on these issues and then amend the Training Textbook accordingly.
Mega woofs to you for a thought-provoking post.
Thanks for the response Dr.
Thanks for the response Dr. Dunbar. I agree that people have trouble with the phasing out part--which of course leads to "But he'll only do it if he sees that I have the food on me." Furthermore, the funny thing about life rewards is that people are using them everyday--the only problem is that they don't realize it and therefore end up using them to reinforce undesired behavior (such as opening a door after the dog scratches at it instead of opening the door after the dog sits). So I think one crucial thing is for people to understand that there are hundreds of life-reward opportunities in a day--the question is are they intentionally giving them (after the behavior of course) contingent on desired behavior, or unintentionally giving them non-contingent on the dog's behavior. The latter of course will lead to lots of unintentional variable reinforcement. Again, thanks for taking the time to respond.
Jordan Rothman, CTC
www.keencanines.net
Lure-Rewards
So how does the use of clicker training fit into your system or does it?
Paul Anderson
My take on it...
When teaching a dog what you want or how to do it, it's a lure.
Once you give it to him, it's a reward.
If the dog fully understands a command, but you still use a lure, then it becomes a bribe.
Definitions aside, I think the important thing is that you phase out lures and external rewards as the dog learns what you want, and begins the habit of following.
I still like to reward my dog often with a "Good dog" or a rub. I follow each food reward with a "Good dog" and a rub, so she knows they all mean she did a good job. Then when I phase out the foods, I can express my approval verbally and physically - but I don't have to give her anything.
Here are some thoughts: Can you reward your dog too consistently, so that she always expects some kind of reward? Should rewards be phased out completely, so that the dog obeys for the sake of obeying?
I'd guess intermittent rewards are best. My feeling is that a dog shouldn't expect a reward, but you should still keep her feeling good about working in general by occassionally (or even frequently) giving her some kind of praise/reward.
What do you guys think?