Behavior modification 'rules' need clearly defined for owners

Behavior modification 'rules' need clearly defined for owners

Feb 01, 2001

This case involves a 10-month-old, male, castrated Shih Tzu that has been biting the clients since shortly after being adopted from the humane shelter at 4 months of age.


Both of the clients owned dogs before, but they had been "dog-free" for a while when they decided to add this pup to their household. When the clients went to the humane shelter, this pup had just been brought in because the elderly lady with whom he had started his life had to enter a care center.

Until approximately 7 months of age, the pup was problem-free. He accompanies the husband to his university office multiple times per week where a multitude of graduate and undergraduate students fuss over, walk and play with the pup.

When not at the university, the pup usually accompanies his people or is at home with them. While seldom left truly "alone", he is regularly left for short periods in the car or at the husband's office. In addition, their adult children, all of whom adore the dog and play energetically with him, frequently visit the couple.

In fact, using the long hallway in the house, the sons have invented a game called "dog bowling" that the pup actually loves.

Beginning at about 6-7 months of age, the wife noticed that the dog was becoming a little less compliant when she wanted him to do things for her such as moving or ceasing to solicit play.

The first bite occurred at 7 months when the wife was grooming him (something she did regularly) and trimming his face (something she did less regularly).

As she moved to hold the pup's head still to scissor over his eyes the pup stiffened and made a very deep growl as he moved to bite her hand.

Since the growl was actually a part of the bite sequence, the wife had very little warning before the pup bit her, drawing two small drops of blood.

Although the injury was extremely minor, the wife was horrified and tremendously emotionally hurt. It was obvious from his rigid stance and direct, dark stare that the pup wanted no more to do with her so the grooming session was abandoned. The husband and wife explained the behavior as the wife had accidentally and unknowingly hurt the dog. This explanation would have sufficed except for two further developments.

First, the next time the wife tried to groom the dog, he stiffened and vigorously vocally resisted. The wife had been reading a lot of dog training books and so felt that the appropriate response was to force the dog on his side or back ("dominance down" or "alpha roll.")

The dog became incensed, like a mad dervish, and the wife abandoned both the restraint strategy and all attempts to groom the dog.

Next, the husband, who desperately wanted the dog to sleep with and cuddle with them, was growled at when he moved to cuddle and kiss the pup ­ who was sleeping on the bed but at its foot ­ upon awakening.

Thinking that he had misheard, the husband further pursued a cuddle session, only to be rebuffed by a more intense growl and baring of teeth. The pattern was the same on ensuing days until the husband lost his temper and hit the dog (for which he felt and feels truly awful and embarrassed), which resulted in the dog lunging at him.

Meanwhile, life at the office was not all that copacetic either. The pup, who had been picked up and tickled and cuddled by virtually everyone who had met him in his life, had begun to sometimes growl at people who reached to pet him or put on his collar. More confusing still was a change in the pattern of his aggressive behavior: he would solicit attention from someone and paw to be picked up, but once in someone's lap would growl at him or her if petted or otherwise manipulated. It was at this point that the clients decided (a) this made no sense and (b) the disciplinary tactics recommended in their training books made him worse.

The behavioral evaluation

The pup was in the wife's arms in the waiting room, where he was avidly watching all the other animals. When placed on the floor to walk to the scale, he growled at a perfectly sweet and friendly Rottweiler in passing. The husband admonished the pup that such behaviors were one day going to land him in trouble.

Once in the exam room, the pup cheerfully explored everything and everyone, barked at his reflection in the one-way glass, and stood up on everyone's knees to sniff them. After watching, but not interacting with the dog for about 20 minutes while soliciting additional history regarding play (e.g., "dog bowling") and watching the pup's response to other dogs, I picked the pup up and put him in my lap when he solicited me. He made himself comfortable and sat alertly with his back turned towards me. After a few minutes, I began to slowly but firmly pet him along the dorsum. I had not reached his tail when he began to growl. Further touching elicited a stare and a snarl. When I moved my hands away from him, he stared at me once more and jumped from my lap. He then turned towards me, sat in front of me, and stared at me. One of the things that made the snarl so surprising to the clients and students was that I had already really manipulated the pup while he walked around the room; I'd held his head and looked in his eyes, I had pushed on his hips and neck, each time allowing him to walk away.

Within about five minutes the pup lost interest in me and moved to the husband who began to play roughly with him. This play involved turning the pup on his back and spinning him (an expanded version of which is involved in "dog bowling"), smacking at and around the dog's head and on the floor, allowing the dog to bite at hands and hold on. While this is truly rough and inappropriate play, in truth, it is the people (the males in this family) who initiated, taught and reinforced the physically fierce behavior. What stunned the students was that the dog not only loved this form of play, but also showed no signs of any aggressive behavior.

Finally, when left alone in the room and observed from behind one-way glass, the pup started to cry immediately, paced and whined, after about five minutes sat down and stared at the door.

Physical and laboratory evaluation

The physical and laboratory exams were both non-remarkable and the students had no trouble getting blood or examining the dog.


This dog has some mild attention-seeking behavior and mild dominance aggression. He may be developing some separation anxiety. His people were diagnosed with inappropriate play behavior.


The clients were advised that since the aggression was so minor that they would likely be able to reverse it easily.

All behavior modification plans have three fundamental steps:

* Cease all punishment and "correction."

* Avoid all circumstances known to provoke the dog.

* Encourage the dog to defer to all people in all situations, by sitting and looking to the people for cues about the appropriateness of his or her behavior, and reward him or her when this is done. This step means that the clients have to be patient, consistent, reliable and humane.

The logic of these steps is especially clear in cases of dominance aggression. Dominance aggression is best understood as an anxiety disorder that is about control or access in social situations involving humans. Dominance aggression is not ­ as the common dogma states ­ about a dog who is trying to dominate a human and needs to be shown "who is boss." This myth has caused generations of dogs to suffer abuse and inhumane treatment with the goal of forcing them to physically "submit" to people. In this outdated view, the dog's behavior is a variant of "normal" that can be "corrected" with force. Genetic, familial and neurochemical data all support that dominance aggression, although poorly named, can be broadly understood as an anxiety disorder involving the serotoninergic system. These dogs are not normal dogs that are "willful" and misbehaving; they are abnormal dogs that are almost always uncertain about the appropriateness of their response in some classes of interactions with humans. Hence, their need for controlthe only way they can manage their uncertainty is by using what appears (to humans) to be very odd and out of context "control" behaviors. Accordingly, these dogs provoke people to "test" and deform the system and to obtain information about how each specific individual is likely to react. The threats are not about the people's status, they are about the dog's uncertainty, and represent the only way that these dogs, who have difficulty sitting back and learning from context, can get information about what to expect.

Accordingly, it should be no surprise that "punishing" or "correcting" these dogs makes them worse. Think of it this way ­ if the dog was testing to see if you were a threat, use of any of the forceful, outdated and inhumane techniques involving biting, "dominance downs," "alpha rolls" and "hanging" from a choke chain will absolutely convince an anxious dog that you are a threat. You will be bitten, and the dog will likely be killed because of a human misunderstanding. The pup in this case responded according to plan when his people "disciplined" him. They, however, realized that the forceful techniques were not living up to their billing and sought other help.

Because the challenges and the deformations of the social system used by dominantly aggressive dogs actually form a "rule" for interaction, we should be able to treat these dogs by substituting another "rule." This is exactly what the behavior mod does. The reason that the pup was never aggressive during the rough play was because-no matter how inappropriate- the play "rule" was absolutely clear. The client's must now make all their other rules for interaction equally clear.

The first step of the behavior mod is the passive Protocol for Deference. Simply, the dog must sit calmly and look at their people for anything they want; food, love, attention, walks. For dogs sitting is a deferential behavior that communicates that they are willing to consider the judgment of others. Sitting also acts as a "stop" signal; you cannot simultaneously be sitting calmly and bite someone. Because this is a behavioral sequence used this way by dogs toward other dogs, we can use it with little risk of ambiguity. Once the dog solicits input from the people they can pet or play with the dog, reward him or her with praise, or even call him or her onto their laps. At the first sign of any further threat, all the client has to do is withdraw attention. Although behaviorists frequently recommend complex behavior mod programs to their clients, in mild cases the aforementioned triad may totally stop the problematic behavior. Raising all puppies this way is also likely to prevent the development of many behavior problems. Pups as young as 5 weeks can and should learn to sit for people for anything they want. Puppies have no problems learning to defer if their people are consistent with their requests. Encouraging consistency from clients, however, can be a challenge.

Intuitively, avoidance of all provocative situations (i.e., in this dog's case, hugging and kissing first thing in the morning when he's in the bed) makes sense from the clients' viewpoint because they don't wish to be bitten. This is fine, but the more important reason for avoidance has to do with how learning occurs.

The more often you successfully repeat a behavior, the better you will learn that behavior. At the molecular level, this process is called long-term potentiation (LTP), and it actually changes the proteins that comprise synaptic receptors. So every time the dog repeats inappropriate behaviors-those behaviors are being encoded at the molecular level. One of the reasons that the tricyclic antidepressants (TCAs) and selective serotonin re-uptake inhibitors (SSRIs) are so successful when used in combination with a behavior modification program is that they act through this same LTP pathway to change receptors at the molecular level.

The clients readily agreed to the above and were extremely happy to see how nicely the pup could sit and attend for them. The clients had hinted that the pup might also be anxious or worried sometimes when left alone, so we watched from behind one-way glass when the clients left the pet. The pup instantly cried and barked and spent the first few minutes pacing before settling down in front of the door to watch through the doggie window. While these behaviors are not sufficient to diagnose separation anxiety, they could be early warning signs.

Accordingly, we emphasized how important altering this dog's anxiety level was. In addition to the behavior mod, we added a low dose of the TCA amitriptyline (Elavil; 5 mg po q. 12h or 1 mg/kg po Q. 12h) to the treatment regimen. The drug was added, not because it was absolutely mandatory in this case, but the client had personal experience with anti-anxiety medications and wanted the dog to have every chance at the speediest recovery and at a full and happy life. Could the dog have improved without drugs? Of course, however, the strongest finding of a double blind placebo-controlled study evaluating the efficacy of clomipramine for the treatment of separation anxiety was that the drug increased the rate at which animals improved when combined with passive behavior modification, compared with the behavior mod, alone. If learning matters -and in a young pup it does-then drugs are useful to increase the rate at which the newly acquired, relaxed behaviors replace the old. TCA and selective serotonin re-uptake inhibitors have so few side effects compared with other commonly used veterinary medications (e.g., phenobarbital, prednisolone, insulin) that the benefits far outweigh the risks here.

Remaining issue

The remaining issue to be discussed was the client's behavior. The rough play not only encouraged the dog to be "mouthy," but also placed the dog at extreme risk for injury. Once we gave the clients some idea of the risks of luxated patellas and proptosed eyeballs, they were extremely willing to readjust their play behaviors and use only toys. This is a very good thing, and this is a lucky puppy. All will go well.

Dr. Overall heads the Behavior Clinic at the University of Pennsylvania. She is certified by the Animal Behavior Society as an Applied Animal Behaviorist. Overall received her VMD degree from the University of Pennsylvania in 1983. She completed a residency in behavioral medicine at the University of Pennsylvania in 1989. She has spoken nationally and authored numerous articles on behavioral medicine and lizard behavioral ecology. Overall is author of a book entitled, "Clinical Behavioral Medicine for the Small Animal" by Mosby.


What's your question? Send your behavior-related questions to: DVM Newsmagazine, 7500 Old Oak Blvd., Cleveland, OH 44130. Your questions will be answered by Dr. Overall in upcoming columns.