Author Archives: Melissa Alexander

No Reward Markers (NRMs)

by Melissa C. Alexander

The following article was originally printed in Teaching Dogs magazine.

Humans are notoriously verbal creatures. We love to talk, and we do so automatically, even when the person we’re talking to doesn’t speak our language, can’t hear what we’re saying, or even when the “person” isn’t a person at all.

We’re so used to verbal instructions that the ideas of stilling the chatter during early learning and adding the cue after our dogs are offering the behavior seem ridiculous. The clicker, however, we understand. The clicker may not be verbal, per se, but it gives clear, audible feedback that the dog did what we want. It seems only logical then that we need another signal to tell our dogs to try something else when they offer something we don’t want.

This signal is called a “No Reward Marker,” or NRM. NRMs are intended to be a verbal cue for extinction, not a punisher, so people attempt to say them in the most neutral tone of voice possible. “Uh-uh,” said quietly and calmly, is a common NRM. In a training session, the trainer would either click or use the NRM after each rep to let the dog know whether his behavior was correct or not.

Although they sound logical, NRMs are not without problems—and controversy. The biggest problem is that NRMs may not be as neutral as we want them to be. Imagine yourself on a game show…

Your goal is to solve a puzzle to win the prize. You know that every right choice will get you closer to the grand prize. After a right choice, a bell sounds. After a wrong choice, a buzzer sounds. How do you feel about that buzzer? Is it neutral? Is it simply telling you “wrong choice,” or is it increasing your stress level? Would it be any different if the sound were a beautiful harp chord instead of a buzzer? I doubt it. Now imagine that you weren’t given any feedback for wrong choices except the lack of the “right choice” bell. Would you be as frazzled by the problem-solving process?

Some dogs will take NRMs neutrally, simply as information offered. Others will view it as a punisher. In the short term you might not be able to tell how an individual dog views the NRM. The effects may be clear only over the long-term, as rate of emitted behavior falls off and the dog becomes reluctant to experiment with new behavior.

You may be lucky enough to have a dog that takes NRMs neutrally, but that may not be true of your next dog. Another problem with NRMs is that they’re habitual. Once you, as trainer, are in the habit of giving an NRM, you will do so nearly automatically. It can be very, very difficult to break that habit if you need to work with a dog who finds them punishing. NRMs are also habitual for the dogs. They learn to rely on them, expect them. If you get into a situation where you can’t give that expected feedback, the dog can become confused and anxious.

Fortunately, NRMs aren’t required for training. Though they make sense to us verbal humans, the reality is that you can communicate the same information simply by withholding the click. The click means “You did what I want!” No click, then, means “Try something else.” By clicking and reinforcing the choices you like and ignoring and not reinforcing the choices you don’t like, you allow positive reinforcement and extinction to work together in a powerful way. Reinforcement makes those behaviors stronger and more likely to occur. Extinction makes the other behaviors weaker and less likely to occur.

The process for training without an NRM is simple. At the beginning of the session, set your criteria—decide exactly what will make you click. Each and every time the dog achieves that criteria, click and reinforce. If the dog offers anything else, including a sub-par version of the goal behavior, simply do nothing but reset for another repetition.

The bottom line is that NRMs, though logical, add an unnecessary level of complexity to training. Keep training simple for you and for your dogs: forego NRMs and stick to basic positive reinforcement and extinction.



by Melissa C. Alexander

The following article was originally printed in Teaching Dogs magazine.

Generalization is the ability to apply a concept to a situation different from the one it was initially learned in. Humans do this quite easily and quite naturally. For example, when you learned to write, you didn’t have to relearn the process when you went from school to home, changed from notebook paper to poster board, or switched from pencils to ballpoint pens. Generalization is “big picture.”

Discrimination, by contrast, is the ability to focus on the smaller picture -– the details. Humans generalize more easily than they discriminate. Police offices, for example, spend hours and hours honing their observation skills. Dogs, however, are master discriminators. “Sit” doesn’t necessarily mean “put your bum on the ground” to a dog. With improper generalization, sit may mean “Put your bum on the ground directly in front of mom when she is in the kitchen standing next to counter wearing a bait bag and holding a clicker and cookie.” Now that’s discrimination!

Generalization is considerably more challenging for dogs. Except for aversives, which they generalize easily (though frequently inappropriately) as an instinctive survival mechanism, dogs must work as hard to learn to generalize as humans must work to discriminate.

When dog trainers speak of generalizing a behavior, their goal is to teach the dog that a cue and its associated behavior apply in more than one environment. This process includes more than practicing the behavior in more than one location, however. Let’s review the example I gave above: With improper generalization, sit may mean “Put your bum on the ground directly in front of mom when she is in the kitchen standing next to counter wearing a bait bag and holding a clicker and cookie.” There’s a lot more than location to generalize there.

The key to generalization is variability. Unless you want your naturally-discriminating dog to conclude that something in the environment is a necessary element of a behavior, you must make sure that nothing but the true cues (called discriminative stimuli) remain consistent during training.


  • The physical location, including position within the room. Practice at angles to walls and furniture occasionally.
  • The dog’s position relative to you. How many people have accidentally taught their dog that “sit” means “sit in front of me”? Vary this from the beginning, using physical boundaries, if necessary, to help generalize the concept.
  • Your physical position. Unless your position will be consistent during performance, practice the behavior while you are standing, sitting, lying down, kneeling, or doing yoga. Wave your arms around. Stand on one leg. Hop in place.
  • Presence of food, clicker, bait bag, and other tools. Tools used for training will ultimately need to be faded from the picture. Be creative in your placement of and choice of reinforcers. Vary the routine you have set up that means “training time.” Condition multiple conditioned reinforcers.

Though it sounds like a lot to remember, the good news is that generalization is habitual. Once your dog has generalized a few behaviors, he will begin to generalize others very, very quickly.



by Melissa C. Alexander

The following article was originally printed in Teaching Dogs magazine.

Extinction is the cessation of reinforcement. When you shape behavior, you start by reinforcing a tiny bit of behavior. When you make your criteria harder, you stop reinforcing the lesser offerings. Because these previously reinforced behaviors are no longer being reinforced, the dog stops offering them, choosing instead to offer the new behavior that is being reinforced.

Extinction is not punishment. Punishment is an event. When you punish, you either add something (positive punishment) or take something away (negative punishment) in order to suppress a behavior. Extinction is a “non event.” You didn’t add or take away – you simply did nothing. Let’s look at an example.

Every day, a person drops a cookie outside a fence for the dog that lives there. One day, however, the cookie lands a bit further from the fence than normal. The dog tries to reach under, but he can’t quite reach it. He tries to dig under, but the ground is rocky. He tries to climb over, but the fence is too high. He tries to go around, but the gate is locked. He will try harder and harder – stretching to reach further, contorting his body for an extra millimeter. He will try a running start to jump higher. But if he never succeeds, he will eventually give up and quit trying. The next time he’s there, he might try again, but he’ll give up more quickly.

This was an example of extinction. The fence didn’t shock him when he reached under or tried to climb. The man didn’t snatch the cookie and take it away. The behaviors he tried simply didn’t work. When one thing didn’t work, he changed his behavior. He tried again. He tried harder. He tried something different. These are characteristics of extinction bursts – characteristics we count on when we increase our criteria during shaping. After he stopped trying to get the cookie, he occasionally tried again. That’s another characteristic of extinction called spontaneous recovery. Had he gotten the cookie during spontaneous recovery, the behavior would have bounced back, very strong. However, with continued lack of reinforcement, the behavior went away again.

So how do you use extinction? When you train, define the specific criterion you are looking for in your session. For example, when you begin training sit, you would reward the dog for putting his bum on the ground. If your dog offers anything else – for example, if he lies down or backs up – ignore it.

Once your dog is reliably offering sits, you’re ready to increase your criteria. Instead of just any sit, you decide to reward only tucked sits – sits where the back feet are drawn toward the stationary front feet. Suddenly some of the sits you used to reward aren’t being reinforced anymore. The dog has to experiment to figure out what will work. This experimentation is part of extinction – try again, try harder, try something different. Gradually he begins to offer more and more tucked sits. Because you aren’t reinforcing non-tucked sits anymore, those sits will extinguish.

Extinction can also be helpful in solving some problem behaviors. For example, a dog jumps on someone to get attention. If people suddenly don’t respond – literally become a statue – the dog will have to try something different to get the attention he wants. If the people instead reinforce sitting with lots of attention – and doggedly refuse to reinforce jumping – then the dog will begin to choose sitting for greetings instead.

However, for this to be true, the trainer must control the reinforcement. If a behavior is either self-rewarding or rewarded by the environment, extinction will fail unless the trainer can consistently offer a better reward. Barking is frequently a self-rewarded behavior. Waiting for a dog barking at passers-by to simply get bored from “lack of reinforcement” is an exercise in futility. When dealing with self-rewarding or environmentally-rewarded behaviors, a combination of management, positive reinforcement, and negative punishment are an alternative solution.

Comparison of extinction and negative punishment:

  • Both decrease the occurrence of a behavior over time.
  • Negative punishment is an EVENT — the actual removal of something that causes the decrease in behavior.
  • Extinction is a “NON-EVENT.” It is lack of reinforcement. Instead of getting something good to strengthen the behavior, or having something added or taken away to suppress it, nothing happens. It just “doesn’t work” to get the desired reinforcement any more.
  • Both are frustrating to the learner. The level of frustration varies from learner to learner, situation to situation.
  • Extinction causes the dog to change his behavior – try again, try harder, try something different. Punishment, on the other hand, suppresses behavior causing the dog to do offer less and to restrict offerings to what he is sure will result in reinforcement.
  • Extinction can cause a decrease in rate of response. However, this is most likely to happen if the overall rate of reinforcement has also fallen, such as when the trainer lumps criteria or increases too quickly.
  • Extinction is effective only when the behavior is in no way self-reinforcing to the learner.

A Case for GLs

by Melissa C. Alexander

I love Suzanne Clothier — just bought her new book — but she wrote an article about head collars that I strongly disagree with. I think head halters are a wonderful device for management, for safety, and for getting a reinforceable behavior in die-hard pullers. Do I pop them on every dog? No. I much prefer to teach a dog to walk on a flat collar or harness from the beginning. My dogs have never worn them, but I own one, and if Pax goes through a particularly out-of-control adolescent period, I won’t hesitate to condition and use it temporarily.

Let’s look at some issues…

  • Management. The head collar itself doesn’t teach the dog a darn thing. It can, however, be used to prevent a dog from dragging you all over the place (undoing your training) until you’ve taught him to walk on a loose leash.
  • A head halter is not a magic cure for a pulling dog. Dogs can, and will, learn to pull on a head halter, just as they learn to pull on a choke or prong. It should be used in conjunction with a training program — reinforcing for desired behavior — even if you never plan to wean the dog off of the head halter.
  • Safety. If a person is young, small, elderly, fragile, or disabled, a head halter can be an extra level of safety, sometimes making the difference between having and not having a dog. Even a very calm dog can bolt unexpectedly, pulling someone down. I’ll never forget when my Newf caught sight of my husband and bolted. His leash was tied to a belt loop on my jeans. Fortunately, the loop broke. Had it been around my waist, I think he’d have broken my back — or at least injured me badly.
  • The head halter can help set you up to succeed. In order to train a dog to walk on a loose leash, you have to have reinforceable behavior — and you have to be able to maintain a high rate of reinforcement. (Failure to do so is probably the biggest mistake I see, head collar or no head collar.) With a distracted dog or a confirmed puller, how can you do that? With a head halter, you can circle the dog to regain lost attention. Plus, dogs tend to be more subdued when wearing the halter, also giving you more opportunities to reinforce.
  • “Leading the dog by the nose.” You’re not — no more than you’re leading by the neck with a regular collar. Ideally, a leash, no matter what it’s attached to, is a tether for safety, and the dog learns to walk next you (as opposed to learning “don’t pull”). If the dog is manically pulling, it’s learning exactly what you don’t want it to learn. The head halter makes it easier for you to keep your dog from undoing your training. You don’t use a head halter to hold the dog in position — he should still be learning to walk on a loose leash.
  • I’ve never heard of any scientific study that compared a lifetime of results of prongs versus anything but choke chains. I *do* know that the GL company aggressively tracks rumors of injuries resulting from use of their products and actively solicits customers to report injuries. The fears about potential problems simply have not been borne out in reality. If you put a head halter on a flexi, and the dog were to bolt to the end, sure, there’s a risk of injury. Don’t do that. Use it correctly on a short leash, and if you need to regain attention, pull *sideways* to circle him, not up and back.

And because it was brought up, let’s talk about whether the GL is aversive. There are two separate issues that people raise — and they are unrelated. One is “The GL is aversive because dogs don’t like to wear them / are subdued when wearing them.” The other is “The GL is an aversive because it decreases pulling.”

First, let’s address the “my dog hates to wear it” issue. Most dogs freak out when you first start using it. Most dogs react the same way when you begin using a collar and leash as well. Do they like head halters? I would guess that no, they don’t. So they probably are aversive.

BUT WAIT. What if you condition your dog to the head halter, associating it with all good things? Many, many people do that, and their dogs cease to react to the head halter. Obviously the halter is no longer an aversive. Have we committed a crime by initially exposing the pup to an aversive stimulus and then counter-conditioning the response? If so, we commit the same crime when we condition to a collar and leash — or to nail clippers or to grooming or to being restrained or to be intimately handled.

One other point — some time ago someone posted the abstract of a study that examined the physiological response to a head halter. The study found that even the dogs who rolled and pawed at the halter did not have the physiological responses associated with increased stress. Nor did the dogs who appeared to be “calmed” by the halter. (Conclusion: the pawing is not “stress” and the calming is not “shut down.”)

Second issue — “the GL reduces pulling.” As I already mentioned, the GL itself doesn’t really stop pulling. It makes pulling more difficult because the dog has lost leverage. It is much more difficult for an animal to pull against pressure on his head than it is against pressure on his neck and shoulders. (That’s why horses are worked in head halters, not collars.) There is pressure when the dog pulls, but frankly, I equate that with the pressure exerted by a regular collar.

Bottom line, in my opinion — the GL is a tool that can be used or misused, just like any other tool. I equate it much more to a flat collar than to a prong or choke, which work on pain, not leverage. No matter what, it doesn’t teach the dog to walk next to you on a loose lead anymore than a choke or prong does. And, minus reinforcement for correct behavior, many dogs learn to pull on it, just like many learn to pull on chokes and prongs.

I first recommend that people teach their dogs to walk on a flat collar or harness. However, if someone isn’t willing to do that, has a dog that outweighs the person who is walking it, or is elderly or physically disabled and could be injured if pulled over by an unexpected lunge, I’m going to continue recommending a properly-conditioned head halter. The head halter can be associated with good things so that the dog won’t dislike wearing it, doesn’t hurt the dog, doesn’t have the potential of misplaced aggression (as with prongs), and makes walking safer.


Creating a Nighttime Ritual

by Melissa C. Alexander

A list member with a new puppy is at her wit’s end because the puppy is crying for hours at night. Okay, first of all, realize that your pup is scared to death. She has never been away from her siblings and parents. It will take her some time to get used to being crated by herself at night. You did good by putting the crate in your bedroom. If you can, put it right alongside your bed — even up on the nightstand, so she can see and smell you, and you can dangle your fingers inside to soothe her.

Take her up to bed about 15 or 20 minutes before you want to go to sleep. Let her get her last-minute zoomies out, and then let her settle down on her own to sleep. When she is already asleep, pick her up and put her in her crate. Obviously, that will wake her again. Sit with her, and stroke her until she falls asleep again — presto! Bedtime with no fuss.

In the middle of the night, it’s not quite that easy. When she wakes, she’ll need to go potty. Pick her up and carry her outside — don’t talk to her or pet her or otherwise interact with her. Put her down, and as soon as she potties, pick her up and carry her back to her crate. Do not talk to her or pet her — she needs to be stimulated as little as possible, and she needs to learn that the middle of the night is *not* for playing.

She’ll wake up early. You have a choice then — you can get up early and stay up, or you can get up, feed her breakfast, and pop her in a second, larger crate with a marrow bone for another couple of hours while you go back to bed. I choose the latter. :-)

Be prepared for lack of sleep for the first month or so. It *does* pass, but it’ll pass on her terms, as her bladder grows and as she gets control of her bodily functions.