Negative or Positive Reinforcement

Just Dropped In (To See What Condition My Condition Was In)

- Kenny Rogers and The First Edition


Cute-Dog-Reading-About-Pavlov-Funny-Picture-There have been many arguments about whether to use negative or positive reinforcement.  As I’ve stated in the past, absolutes are absolutely ludicrous:  both negatives and positives are needed.  You can’t have one without the other.  Only by using both appropriately can you help your dog thrive. Unfortunately, most people don’t realize how easy it is to confuse your dog with the wrong kind of positive reinforcement.

Prime example:  A shelter dog named Simba.  Simba, for whatever reason, had been improperly socialized as a youngster, and grew into a young adult who exhibited dangerous behaviors.  As a puppy he should have learned what is appropriate and inappropriate play.  Not to jump.  Not to bite.  He should have learned moderation and self-control.  Typically this is learned from other pack members as a young pup (why do you think we don’t take puppies away from their family too early – they’re learning!).  Simba wasn’t a bad dog, nor was he aggressive in the slightest.  His problem was that he was very demanding.  He was a spoiled brat.

Simba was very willing to learn…for a price.  If Simba wanted to play, he would grab your clothing and drag you to the ground to wrestle.  If he wanted to go somewhere other than where you wanted during a walk, he would drag you there, sometimes by an arm or leg.  If you did something he didn’t like, he would nip you.  Hard.  Now I want you think about what would happen if a child were to engage in this sort of behavior.  Odds are, you’d give them negative reinforcement of some sort to let them know that this behavior is unacceptable.  Unfortunately in the shelter environment that Simba was in, they only believed in positive reinforcement.

At first it looked as if it were working.  If Simba started to pull on a walk, his handler would whip out some boiled chicken to coax him back into a polite pace (Simba would not listen for anything less than boiled chicken – no Milkbones here!).  If Simba jumped up and grabbed an arm or pant leg, he was bargained with:  release my arm and I’ll give you some chicken.  Again, it seemed to be working!

Now, some of you may be noticing a problem here.  See, Simba was an extremely intelligent dog.  He started to figure out the system.

“If I bite someone or become violent with them, they give me a treat! I’ve finally got this whole human thing figured out!”

What happens if you don’t have a treat?  This:

The only pic I can, for decencies' sake, I can publicly post

The only pic of the volunteer I can, for decencies’ sake, publicly post Quite a bite, huh?

One of the volunteers was attacked by Simba.  She literally had her shirt ripped off by him and was bitten several times on the torso area.  Again, Simba wasn’t what you’d call aggressive (I know…biting not aggressive?).  He had humans figured out:  he sat when told, he’d get chicken.  He released someone when he was playing, and he’d get some chicken.  Well, the human ran out of chicken when he had a hold of her.  In Simba’s mind, she didn’t keep up the end of the bargain!  So he did it again, and again, expecting the volunteer to finally figure out what he was telling her:  give me some chicken like you’re supposed to!

Simba’s story doesn’t have a happy ending.  He was eventually quarantined, and only select members of the shelter were allowed to work with him (still using only positive reinforcement).  Eventually, it was decided that he needed to be put down.  He was euthanized because nobody cared to tell him “no”.

So how could this have ended differently?  You’re probably wondering, didn’t I state in the first paragraph of this post that both negative and positive were needed?

Yes, but only done correctly.

Let me give you a different scenario.  My daughter, River (age 6) and my son, Eric (age 9) have quite a few things expected of them with regard to chores.  For example, Eric has to do dishes.  River is in charge of keeping the baseboards in the house clean. They are children, so they have to be reminded to do it (that’s why they’re called “kids” and not “adults”).  But I simply tell them to do it, and off they run and do it.  Their reward?  A hug and a thank you.  About once a week we go on a cleaning spree.  They are expected to help me clean for a couple hours.  I give them age-appropriate tasks, which they complete without putting up a fight or complaining.  If they need help, I give it to them.  But typically they don’t.  And typically, they do a great job.  Again, no complaining, and their reward is a thank you and praise.


Yeah, I don’t know why they’re cleaning in their PJ’s either.

Now, sometimes I when give them the mandatory thank you and praise, and throw in an extra.  Some money.  A trip to the zoo when we’re done.  An ice-cream treat. Once it was a Nintendo DS.  It’s not a reward for doing what I told them to do:  that’s expected!  It’s merely added to the “thank you” they receive.  And it is never presented in the “If you do X, I’ll give you Y” fashion.  They never find out about it until after the task is complete.

Are you seeing how this should be applied to dogs?  If a dog is biting me, I’ll give him a negative, and then they stop.  I will not reward the dog for respecting me.  I expect respect from a dog.  I give it in return.  But sometimes you can see a dog is really struggling, and comes through with good choices.  For example, walking with Sparta, and there’s a suicidal squirrel who runs directly in our path and decides to hang out (really…WTF squirrel!!!).  That’s a hard one.  Sparta has high prey drive.  Yes, I tell her to leave it, but it’s a struggle for her.  That’s where Touch, Talk, Treat comes in to play.

I have her conditioned.  Every time I give her a treat, or even her enrichment toy, she gets a gentle scratch behind the ears, as well as gentle praise.  Very soon she linked the Touch and the Talk to the Treat/Food.  Once you have that Pavlovian response going, you can give your dog a hard-core positive without the food.  So when she passes by that squirrel without making a ruckus, she gets Touch and Talk.  The Treat is implied, the same way “jelly” is implied if I say I’m making a peanut butter sandwhich because jelly and PB are always linked together.  Maybe she’ll get a treat later.  But the thing is, she doesn’t expect it.  It’s like the lottery:  you have to play to win.  Yes, occasionally I’ll have pocketed some treats to give her, but it’s not an expected.

The problem with Simba was that conditioning works both ways.  “We had a deal … I do *this* and you give me a treat when I stop.”  So who was wrong?  He kept up his end of the bargain.


Who’s being conditioned?

Positives are tied for the most useful thing in training…with negatives.  Eventually, proper use of both will shift the tide of things:  pretty soon you are only giving positives.  Good positives, given in the correct instances.  Sparta has not had a problem with squirrels in quite a while. Every so often she still get a treat for passing one.  She’s on the right path and doesn’t need to be guided towards it very much any more, so I can reward her for choosing well. Same with my kids.  We’re heading out for ice-cream right now.  They don’t know it yet.  But they did a great job, and (as usual) didn’t complain once.  They deserve a treat.

River and Eric at their favorite ice-cream shop.

River and Eric at their favorite ice-cream shop.


Keep calm and pilot onKerry Stack
Dawrin Dogs
Dog Training in Cleveland, Ohio

Leave a Reply

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

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>