Misuse and Abuse of Punishment

Punishment and other forms of aversive control (e.g., aversive counterconditioning and negative reinforcement) can be humane and effective behavioral tools in the hands of competent trainers, but noncontingent (after the event) punishment and excessive physical punishment or brutalization (e.g., beating, hanging, or kicking) have no legitimate place in the armamentarium of professional trainers. That such methods exist today and are employed in the name of dog training is a blemish on the profession.

Noncontingent Punishment

Perhaps the most frequently misused form of aversive control is noncontingent punishment. Procedures involving such treatment are often recommended for the control of behavior problems that occur while the owner is away from home. Unfortunately, this abuse of punishment has been defended by a number of highly regarded authors (Koehler, 1962; Benjamin, 1985; Evans, 1991). The influence of this popular literature is compounded by many dog owners honestly believing that their dog's misbehavior is motivated by spiteful intentions.

"Spite" and Pseudoguilt

Dog owners who believe that their dog's misbehavior is motivated by spite point to the dog's appearance of guilt as proof of a premeditated purpose underlying the dog's undesirable behavior. The dog's guilty appearance during homecomings suggests to them that the dog knows and is behaving in a way calculated to somehow injure them. This rationalization provides a basis (at least in their minds) for the delivery of harsh punishment long after the behavior has occurred. Such treatment is targeted against the dog's bad attitude and the dog's need for discipline. The owner's urge to hurt the dog in such cases is rarely constructive but rather the outcome of an angry reaction to the presence of a soiled area or destroyed personal belonging—anger and frustration that is subsequently directed in the form of physical abuse toward the dog.

When such abusive treatment fails (as it inevitably does), the owner may interpret the failure as recalcitrance on the dog's part and point to growing levels of guilt on homecomings as additional evidence of such an interpretation, thereby justifying an ever-escalating cycle of abusive interaction. Konrad Most long ago repudiated this faulty interpretation, arguing that the dog may never know the reasons for punishment but only learn that some modes of behavior result in aversive outcomes:

It has to be constantly borne in mind that the animal can never learn the reason for a disagreeable experience, but only that certain modes of behavior result in disagreeable experiences. (1910/1955:17)

Later, he stresses, regarding the dog's appearance of guilt,

The "guilty conscience" is caused simply and solely by the so-called fear inspired by the menacing noises and gestures of the human being. In fact, the dog's "conscience" is quite "clear." Such fear is always aroused in the dog by hostile behavior on the part of its master. For, as a rule, the animal has had it knocked into his memory from puppyhood that hostile human attitudes are accompanied, or quickly followed, by some disagreeable experience. But the cause of fear in the presence of the master is never awareness in the dog of any present, let alone any past, behavior to which the man objected. (1910/1955:72)

Although it is impossible to know for sure, dogs probably do not reflect much on the past or future significance of their behavior. "Every dog," as Hans Tossutti (1942) once noted, "considers his acts as right."

Instead of worrying about the past or future significance of what they do, dogs are content with the here and now, living in a perpetual present where time flows like the Heraclitean river into which "we step and do not step." Although dogs can encode experiences and retrieve memories, they are most likely unable to form conceptual constructs and symbolic representations of events from which to deduce causal inferences about the distant past or future. Consequently, appealing to a canine ability to extrapolate from a present consequence to a past action does not help to explain the dog's appearance of guilt. Although a dog may be able to associate the presence of a destroyed item with the owner's anger, it is unlikely that the culpable action is directly influenced by the owner's disapproval or abusive efforts. Unfortunately, however, the owner reads the dog's guilt as if it was related to a remote action present in the dog's mind at the time of punishment. Dogs do not appear to have such cognitive abilities. To dogs, threats of future punishment are as useless and meaningless as punishment is for long past actions. Actually, most of what we do and value as humans is probably lost on dogs. William James offers a bit of sobering analysis regarding the situation:

Our dogs, for example, are in our human life but not of it. They witness hourly the outward body of events whose inner meaning cannot, by any possible operation, be revealed to their intelligence—events in which they themselves often play the cardinal part. My terrier bites a teasing boy, for example, and the father demands damages. The dog may be present at every step of the negotiation, and see the money paid, without an inkling of what it all means, without a suspicion that it has anything to do with him; and he never can know in his natural dog's life. (1896/1956:57-58)

The Persistent Belief that Noncontingent Punishment Works

Another factor contributing to the popularity of noncontingent punishment is the appearance that it somehow works. Since noncon-tingent punishment is often directly associated with the object or area where the offending behavior took place, any appear ance of effectiveness is probably due to the influence of aversive counterconditioning. In other words, the ostensible benefit of such treatment is not due to the remote suppression of the unwanted behavior, but rather such methods probably work by indirectly conditioning fear toward the object or location where punishment took place in the past. One of the most repugnant examples of noncontingent punishment in the dog-training literature illustrates this effect:

If you come home and find your dog has dug a hole, fill the hole brimful of water. With the training collar and leash, bring the dog to the hole and shove his nose into the water; hold him there until he is sure he's drowning. If your dog is of any size, you may get all of the action of a cowboy bull-dogging a steer. Stay with it. I've had elderly ladies who'd had their fill of ruined flower beds dunk some mighty big dogs. A great many dogs will associate this horrible experience with the hole they dug. ... It is not necessary to "catch the dog in the act" in any of the above instances of correction. Be consistent in your corrections and your dog will come to find the smell of freshly dug earth quite repugnant. (Koehler, 1962:200)

Pressing a dog's nose into water is irrelevant to digging per se, but, as the author points out, the terrifying sensation of drowning causes the dog to acquire a repugnance to the smell of soil, to say nothing of how it affects the dog's attitude toward the owner. instead of suppressing the tendency to dig, chew, or eliminate in the owner's absence, such extreme methods cause the dog to avoid the item or place where aversive stimulation took place. Along with Koehler, Benjamin (1985) and Evans (1991), using much more restrained aversives, also emphasize the need to present evidence or proof to the dog to make the "disciplinary" event effective. Such treatment does nothing to deter destructive behavior or inappropriate elimination, but it may instill a fear of the object, place, or person associated with "punishment."

Although aversive counterconditioning has a useful place in dog training, such variants as the aforementioned method are ill-conceived and excessive. One concern about the method is that dogs may learn to associate aversive stimulation, not only with the surrounding area or object, but with the abusive owner applying it. Because of this risk, aversive counterconditioning is best carried out through remote means utilizing booby traps and other procedures by which the object or area itself appears to deliver the aver-sive stimulus. Such methods require comparatively mild aversives, with far less risk of producing side effects, while at the same time promising a much greater likelihood of success.

Interpreting Pseudoguilt if dogs are unable to connect punishment with the behavior occurring in the remote past, what causes their appearance of guilt? A frequently cited analysis of guilty behavior interprets guilt as a ritualized submission display aimed at avoiding noncontingent punishment (Borchelt and Voith, 1985). This theory holds that pseudoguilt is maintained by a triadic structure of conditioned associations involving three components: (1) evidence of a destroyed object or soiled area, (2) the presence of the owner, and (3) a history of previous punishment under similar conditions in the past. Many anecdotal reports support this sort of interpretation. For example, it is not uncommon for an adult dog who is kept with a puppy to show guilt when the owner returns home, especially if the puppy happens to eliminate during the owner's absence. it is the adult dog who exhibits guilt, even though the puppy's action was responsible for the offending mess. There are other potential causes of pseudoguilt that ought to be investigated. One possibility is that emotional cues current at the time of the unwanted behavior persist until the occurrence of remote aversive stimulation. These internal emotional cues may subsequently predict pending punishment. Whatever the cause, pseudoguilt is most likely not due to a lingering bad conscience over a past deed.

Negative Side Effects of Noncontingent Punishment

Noncontingent punishment is often harsh and sustained, with the dog often being beaten immediately after homecomings. Most normal dogs are very enthusiastic about greeting their owners after a long separation. The active emotions are intensely affiliative, and the dog naturally seeks reciprocation— that is, the expectant dog anticipates an equally friendly reply. Instead, its affectionate efforts are met with an unexpected and aggressive assault. The result is a collision of violently opposed and conflicted emotions, a situation structurally similar to the procedures used to induce experimental neurosis in the laboratory. As will be seen in the following chapter, from the perspective of experimental neurosis, the collision of opposing and mutually incompatible emotional reactions predispose dogs to develop neurotic conflict. Because of the intensity of the emotions involved, coupled with the inescapable character of the stimulation, the potential for serious side effects is extremely high.

Adult dogs exhibiting separational distress frequently develop a number of persistent behavior problems such as barking, destructive behavior, and inappropriate elimination whenever they are left alone. This group of dogs is at a particularly high risk of becoming the hapless target of abusive and escalating brutalization as part of their "reform." That such treatment is harmful should be obvious, but it is commonly employed on the recommendation of authors such as Koehler (1962), who interpret separation-related behavior as deriving from sullen vengefulness— a condition that must be tortured out of a dog's character through repeated "spankings." This fraudulent view reinforces the popular interpretation of such behavior, which erringly implicates spite as its primary cause, but Koehler takes matters to an all time low in the following useless and cruel prescription for the "revenge piddler":

For the grown dog who was reliable in the house and then backslides, the method of correction differs somewhat. in this group of "backsliders" we have the "revenge piddlers." This dog protests being alone by messing on the floor, and often in the middle of the bed. The first step of correction is to confine the dog closely in a part of the house when you go away, so that he is constantly reminded of his obligation. The fact that he once was reliable in the house is proof that the dog knows right from wrong, and leaves you no other course than to punish him sufficiently to convince him that the satisfaction of his wrong-doing is not worth the consequences. If the punishment is not severe enough, some of these "backsliders" will think they're winning and will continue to mess in the house. An indelible impression can sometimes be made by giving the dog a hard spanking, of long duration, then leaving him tied by the mess he's made so you can come back at twenty-minute intervals and punish him again for the same thing. In most cases, the dog that deliberately does this disagreeable thing cannot be made reliable by the light spanking that some owners seem to think is adequate punishment. It will be better for your dog, as well as the house, if you really pour it on. (1962:196)

There is no reasonable behavioral justification for this form of mental and physical abuse, but, every single day across America, hundreds of frustrated dog owners are carrying out similar rituals of confusion and cruelty in the name of dog training. After several weeks or months of such abusive interaction, besides irreparably damaging the owner-dog relationship, such treatment inevitably results in the elaboration of more serious behavior problems.

The Need for Close Temporal Contiguity

A brief review of basic learning principles will help to underscore the importance of response-dependent punishment. As has already been repeatedly emphasized, learning depends on the timely and regular presentation of relevant stimuli. This holds equally true for both classical and instrumental types of learning. In the case of classical conditioning, the CS (e.g., whistle) must immediately precede the US (e.g., food) for a conditioned association between the CS and US to be established. Similarly, in instrumental learning, reinforcers and punishers must closely follow upon the emission of the target behavior. The behavior-modifying effects of reinforcement and punishment are both significantly diminished to the extent that their delivery is delayed or delivered independently of the oc currence of the target behavior. In the case of punishment, effective use depends on its prompt delivery whenever the unwanted behavior occurs. Under these experimentally established constraints, "punishment" occurring long after the event is a wasted effort that unnecessarily exposes dogs to aversive stimulation. Such interactive punishment serves no purpose, other than providing owners with an outlet to discharge anger and frustration.

Hitting and Slapping: Okay?

The routine hitting and slapping of puppies and dogs are also inappropriate forms of punishment, especially when they are delivered on a noncontingent basis. Sensitive dogs exposed to such treatment may develop a negative expectation about hands moving abruptly or startlingly in their direction. voith and Borchelt have noted a significant correlation between abusive house-training measures and an increased incidence of fear-related aggression in adult dogs:

Direct physical punishment from the owner, even if the dog is "caught in the act" can lead to fearful and defensive behaviors. Punishment unrelated to "the act" results in even more-intense defensive reactions to being approached, reached for, or touched by a person. Although dominance aggression is the most commonly diagnosed behavior problem presented to animal behaviorists, fear-induced aggression is probably the most common type of aggression among pet dogs. Fearfully aggressive housedogs almost invariably have a history of difficulties in housetraining and were inappropriately and unpredictably punished by the owners. (1996:176)

Under conditions of heightened distress or even momentary distraction, the startling approach of a child or stranger with outstretched hands may be interpreted as a threat, resulting in a preemptive attack aimed at controlling it. Further, the transition from a smack on the rear or chops to Koehler's method is one of degrees, not kind. Physical sorts of punishment rarely yield lasting suppression of behavior, unless they are delivered strongly. This characteristic often causes the "spanking" to escalate gradually into a peri odic beating. Ethical trainers and behaviorists should draw the line firmly and exclude all forms of corporal punishment from routine training, except as might be needed in the case of self-defense.

Finally, uncontrollable painful stimulation occurring under some social circumstances may simultaneously elicit fear and anger as unconditioned responses. Where fear and anger are elicited together by the threat of inescapable pain (e.g., inappropriate physical punishment), the possibility of lasting irritability, vigilance, anxiety, and lowered thresholds for aggression may occur in the presence of the punishing agent. Under the influence of such abusive handling occurring early in a dog's life, fear and anger may become motivationally linked together as a conditioned response to pain or threat of pain and, over the course of the animal's development, "incubate" until the dog reaches maturity, by which time a highly intractable aggression problem may express itself.

How To Train Your Puppy

How To Train Your Puppy

Getting a new puppy is a fun and interesting time. You probably went to a breeder or pet store or maybe just saw an ad on the Internet or the newspaper, for puppies, and decided just to check it out. Before you knew it those little eyes and fluffy puppy fur had your heart melting and you were headed home with him or her in your arms. If you are like most new pet owners you had visions of playing fetch with your dog, of watching him frolic at the lake, and of cuddling up on cold nights.

Get My Free Ebook


Responses

  • cesare
    What is noncontingent punishment behavior example?
    6 years ago

Post a comment