Literature
January 23, 2005
I got my copy of Jean Donaldson's new book, Fight: A Practical Guide to the Treatment of Dog-Dog Aggression, last week. Donaldson's discussion of the bite threshold in dogs, in her 1996 book, Culture Clash, is the most useful thing I have read for preventing dog bites. In Fight, Donaldson applies her behaviorist perspective to problems of dog-dog aggression: not only fighting, but also undesirable behaviors that are precursors to a full-blown fight, including fear and avoidance of other dogs, and pro-active lunging, barking, snapping and snarling. Donaldson's new book could be better organized, though the detailed table of contents is a great improvement over Culture Clash. And once again Donaldson could benefit from a good editor: whereas, in Culture Clash, her style was abrasive, here it is encumbered--and not just by the typos.
Anyone who has seen Strummer harass Mosley, a wonderful Golden Retriever at AAAS, will recognize Strummer in Donaldson's description of a "bully:"
Bullies . . . target certain dogs for harassment or attack. They rarely play normally with with these dogs yet play fine with others (80).
Indeed, Strummer's only "target" seems to be the lovable "Mo," who, at the park, wants nothing other than to play fetch with his person. Strummer likes to chase dogs who are chasing balls (frisbees, sticks...), and ordinarily the retriever and my
meta-retriever work it out, and both enjoy themselves (though they seem to be playing two different games). But when Strummer chases this particular Retriever, their interaction is completely different. She engages in all the contact behaviors she typically employs to solicit play (i.e. she doesn't play-bow, but she nips at Mosley's legs, neck and flanks and hits him with her signature "left jab"), while he ignores her and tries to focus on his tennis ball.
I, of course, don't enjoy watching Strummer be a bully: I distract her by calling her sharply and telling her to "play nice." When she persists in bullying poor Mosley, I put her on her leash and, if she fails to settle down and pulls on the leash, we leave the park. This is almost the regimen that Donaldson recommends for treating bullying behavior.
There is a disclaimer at the beginning of the book that advises,
the services of a competent professional trainer or applied behaviorist should be sought regarding the [book's] applicability with respect to your own dog.
Fortunately, Strummer's bullying of Mosley is not dangerous to the dogs or to the human handlers, so I am comfortable trying to implement Donaldson's treatment. Moreover, Strummer's "acquired bite inhibition" is excellent. Donaldson explains that
the most important prognostic factor in dog-dog aggression is the degree of acquired bite inhibition, which determines how much damage is done when the dog bites. This can range from bites that break bones, deeply puncture muscle and create massive tearing, to bites that leave saliva but no damage whatsoever (21).
Standard Poodles were bred for a soft mouth so that, as hunting companions, they can retrieve a felled duck without damaging the bird's flesh. Strummer's mouth is so soft that, when offered a treat, she will not close her mouth to hold it until you let go of it.
Donaldson recommends treating bullying with "negative punishment," the removal of a reinforcing stimulus.
Dogs that bully other dogs find both play and harassment/fighting to be reinforcing events, so timing them out for harassment and fighting is extremely effective at reducing these (82).
The method of applying negative punishment that Donaldson recommends to teach a dog to stop bullying looks like this:
1. Issue a Warning Cue
A. dog continues bullying
2a. Issue a Time-Out Cue
3a. Execute the Time-Out
B. dog stops bullying
2b. Acknowledge with a "Thank you"
3b. Continue monitoring the interaction
The command that the dog has to learn is the warning cue (Donaldson suggests "Enough," but "Play Nice!" should work just as well). The dog can either offer the desired behavior and stop harassing her target, or the dog can offer undesirable behavior and continue harassing her target. Once the dog is given the Time-Out cue (Donaldson suggests, "Too bad," or "You're out"), the dog has failed to offer the desired behavior and a Time-Out must be executed immediately.
The way this treatment works is that, if the Time-Out is executed properly, the dog learns the warning command, i.e. the dog is eventually able to reliably offer the desired behavior when the warning command is issued (scenario B). I'm not sure what purpose saying "Thank You" serves. Perhaps it merely addresses the human's need to respond to the dog's reaction, turning negative punishment into an opportunity for positive reinforcement.
In fact, teaching a dog not to bully by negative punishment is like teaching a dog the "Sit" command by positive reinforcement. The method of teaching a dog the "Sit" command by positive reinforcement (e.g. clicker training) looks like this:
1. Issue the command cue, "Sit"
In both cases (negative punishment and positive reinforcement), scenario B is what the dog is supposed to learn. Whereas by negative punishment, the dog learns to prevent something bad from happening (the punishment, administered as a Time-Out), in positive reinforcement, the dog learns to make something good happen (the reinforcement, administered as a treat.) The difference is that in teaching a dog the "Sit" command, you can shape the desired behavior before teaching the cue, so that the dog already knows how to sit when you start teaching her the command. In teaching a dog not to bully, you can't really shape the desired behavior with the target before teaching the warning cue.
UPDATE January 25, 2005: There is two feet of snow on the ground, and who did Strummer meet at Triple-A-S this afternoon? Mosley! ...whose tennis ball was aparently lost within twenty seconds of his arrival at the park. After Moe showed the little Strummer girl how to bound through the snow, they had a great time chasing each other! Ah, snow. It brings out the best in all of us...
Somerville is now safe from one aggressive dog. David Renna, Somerville's one-man Animal Control Department, confirmed this morning that Diamond, the unlicensed dog that mauled three people in an apartment in Homer Square (near Union Square) on the Columbus Day holiday, has been humanely euthanized at the North Shore Animal Hospital in Lynn. The Somerville Journal is running the story about Diamond, who "viciously attacked two kids and their mother" on Monday, October 11. The mother "was dog-sitting . . . for her downstairs neighbor . . . , who went out of town on business."
Renna said that when he arrived at the scene last week, he was able to calm the dog down so that he could transport her to the North Shore Animal Hospital, the facility that kenneled the dog for the ten days that the City is required to hold a dog that has not been claimed by its owner. But he said that, in his opinion, the dog was mentally instable, "crazy," due to inbreeding. He said that staff that worked in the kennel where Diamond was quarantined also reported that the dog was confrontational. A female, Diamond was fine with men, but she challenged women, almost as if she didn't distinguish between people and dogs. Although Animal Control had received no previous complaints about Diamond, Renna said that neighbors at the scene of the attack told him that they have had many problems with the dog, including the report that was printed in the Journal, that Diamond had bitten one of the children before.
Human-directed aggression in dogs is never acceptable. Diamond was a menace to society. Now that the problem of the aggressive dog has been taken care of, however, what concerns me is that the Journal article, rather than addressing the problem of an aggressive dog, insinuates that aggression was only to be expected because of the dog's breed. A San Francisco based organization of pit bull owners, rescuers, and supporters, Bad Rap--Bay Area Doglovers Responsible About Pitbulls--agrees with me on how to deal with aggressive dogs: "Pit bulls that do show aggressive behavior towards humans are not typical of the breed and should be humanely euthanized." A dog that is aggressive toward humans is not a typical dog. Dogs and humans have co-evolved over thousands of years to live together to our mutual benefit.
Rather than demonizing pit bulls, our community will be more safe from dog attacks when we understand aggression in dogs and prevent it from ever being directed toward humans. According to Jean Donaldson, in her book The Culture Clash, "the usual reason dogs bite 'without provocation' or 'for no reason' when they had never behaved aggressively before" is that "some novel combination of elements [specifically, 'risk factors'] pushes the dog higher than the elements on their own have ever pushed him previously." "Risk factors," are what Donaldson calls things that "bug" a specific dog. "Typical risk factors include: categories of people to which the dog is not socialized, hands and/or being touched, approach, presence of food bowl or other guarded resources and any discriminative stimulus for positive punishment (such as a choke collar or a strap used to beat the dog)." Donaldson gives the following example:
Hypothetical dog Zaphod has always been uncomfortable around strange men. His other major risk factor is that he freezes up on approaches to his food bowl. The owner has also noticed that he seems just a little bit more sensitive at night than during the day and not perfectly relaxed with hands or when approached. These last two, by the way, are in the profiles, to some extent, of most dogs. One day, Zaphod bites a man who approaches to pat him. The owner is completely floored as Zaphod has never bitten or even growled at anyone before, and there was no provocation on this occasion, from the owner's perspective. As you can see from his profile, however, Zaphod was a time-bomb which, unfortunately, went off (91-92).
Donaldson includes a graphic representation of Zaphod's profile, which shows how Zaphod's risk factors--"strange men," "hands" and "approach"--separately do not bug him enough to reach even his thresholds for growling or snapping. But Zaphod's growl, snap and bite thresholds are all at more or less the same level of provocation: the combination of elements--when a strange man approaches with his hand to pat the dog--bugs Zaphod beyond his growling and snapping thresholds and pushes him past his bite threshold.
Diamond's owner failed to understand his dog's risk factors and thus failed to prevent her from attacking and injuring people. Given that Diamond allegedly bit the ten-year-old girl last year, it was irresponsible of Diamond's owner to leave her in the custody of that child's family.
UPDATE January 23, 2005: Check out Jean Donaldson's new book about dog-dog aggression!
I like dogs. . . . They do not tell lies because they cannot talk.
. . .
I don't like proper novels, because they are lies about things
which didn't happen and they make me feel shaky and scared.
On the very first page of Mark Haddon?s The Curious Incident of the Dog in the Night-Time, Christopher John Francis Boone discoverers a body. It's not a murder mystery like Arthur Conan Doyle's, The Hound of the Baskervilles:
Two dogs were killed in The Hound of the Baskervilles, the hound itself and James Mortimer's spaniel, but . . . they weren't the victims of the murder, Sir Charles Baskerville was.
The comparison is made by our sleuth and narrator, Christopher, a 15-year-old with Asperger?s syndrome, who tells his story about writing a murder mystery while solving the murder. The victim of Christopher's murder mystery--the body on page one--is Wellington, a poodle: "not one of the small poodles that have hairstyles, but a big poodle." The weapon is a garden fork: "the points of the fork must have gone all the way through the dog and into the ground; because the fork had not fallen over."
Every good crime-fiction reader knows that to prosecute a suspect, you need to establish method, motive and opportunity. Method, in this case, is pretty obvious. To establish motive, Christopher resorts to the statistical evidence that "you are most likely to be murdered by a member of your own family on Christmas day." If his murderer's motive was to make Wellington's owner, Mrs. Shears, upset, he reasons, his prime suspect would be the one person he knew who didn't like Mrs. Shears: her husband, who had left her two years ago.
I want to take a moment to consider "opportunity" in this case. Given that Mrs. Shears's dog was murdered in the night-time in her yard, everyone in town that night had the opportunity to murder Wellington. And, given that her dog was murdered in her own yard, Mrs. Shears had the best opportunity of all.
A responsible dog owner Mrs. Shears is not. The only thing she is responsible for is her dog's death: Wellington was murdered because Mrs. Shears left him unattended in her fenced-in yard. Yet, like most evil masterminds in related genres, Mrs. Shears evades prosecution: she is not even on Christopher's list of suspects.
If all of the characters in the story were as responsible toward and respectful of other species as Christopher, their own inter-personal relationships wouldn't be so dysfunctional. Like the odyssey of Dorothy and her little dog, Toto--which, I have argued, would have been averted if Toto had had a responsible owner--if Mrs. Shears were a responsible dog owner, there would be no murder and no murder mystery in The Curious Incident of the Dog in the Night-Time.
Bad dog owner; good novel.
Today the movie, Benji: Off the Leash opens around the country, but last night the City of Somerville watched The Wizard of Oz. I can't help but think that, had Dorothy (Judy Garland) had her Cairn Terrier, Toto (Terry, handled by her trainer and manager, Carl Spitz), on a leash, the whole situation would have been averted.1
If Dorothy had had Toto on a Leash, he2 would not have gotten into Miss Gulch's (Margaret Hamilton) garden or chased her cat; Miss Gulch would not have been able to hit him with a rake, and most importantly, he would not have been able to bite Miss Gulch. Call me a wicked witch, but I agree with Miss Gulch: Off-leash, "that dog's a menace to the community". Sure, Auntie Em (Clara Blandick) says, "He's really gentle -- with gentle people, that is," but the fact is that communities have laws to protect people whether they're gentle or not.
As for Auntie Em, she may be "a Christian woman", but she's not an especially responsible guardian to Dorothy--and not because she sends Toto away with Miss Gulch to be destroyed. Contrary to Dorothy's plea--"He didn't know he was doing anything wrong. I'm the one that ought to be punished!"--Dorothy cannot be held responsible for Toto's transgressions any more than Toto can. Dorothy is a child, and, even for a child, no paragon of responsibility: Dorothy runs away from home. "While children can help with some age-appropriate responsibilities, pets require adult caretakers," advises the Partnership for Animal Welfare in Greenbelt, Maryland. The AKC also recommends that an adult take primary responsibility for a pet dog: "Involve your child in the dog?s day-to-day care, but be realistic about how much responsibility he or she can handle." In The Wizard of Oz, it is Auntie Em and Uncle Henry (Charley Grapewin), who are irresponsible.
For dogs the human world is probably a lot like Oz: "some of it [isn't] very nice... but most of it [is] beautiful". Whether your dog is in Oz, in Kansas, or in Somerville, he or she should be on-leash except when supervised in a place where there isn't any trouble. Like Dorothy, Somerville needs "some place where there isn't any trouble": Somerville needs safe and legal off-leash areas, where dogs are protected, both from suffering and from inflicting harm.
1I'm not the only one who thinks so. See "Are You a Good Owner Or a Bad Owner? A Pop Quiz From the Wiz" by Anne Leighton.
2Toto, like the new Benji, is played by a female.
