San Diego — Both sides agree the seeds of the dispute were planted in 2000 when the County Board of Supervisors announced that the three county animal shelters -- in Carlsbad, Linda Vista, and Bonita -- would be "no-kill" shelters within five years. "No kill," at least for public shelters that can't turn away animals, does not mean a complete cessation of euthanizing animals. It means no "adoptable" animals will be killed. The announcement resulted in the summer 2003 hiring of Marlena Young, an animal-behavior specialist formerly of the Kansas City Humane Society. She is trained in a method of dog temperament testing known as the Sternberg Method. Young's hiring and her subsequent implementation of the temperament testing touched off a round of infighting and firings among staff, management, and volunteers at the shelters.

Cynthia Starkovsky volunteered at the central shelter for 12 years. She leads a loose group of people in the local animal shelter and animal rescue fields who are unhappy with the changes. "At the top of our list," Starkovsky says, "is our objection to the imposing of these very harsh tests on older, very young, sick, injured, and very frightened animals. I would say that is the core issue. A collateral issue also is, since they imposed this system, they are now recharacterizing euthanasia in new ways, declaring animals untreatable for behavior or medical reasons. There are other issues, too, such as the fact that the Department of Animal Services has retaliated against any volunteers and rescue groups who have attempted to speak up on behalf of these animals simply by firing [the volunteers]."

Starkovsky, who was asked not to come back to the shelter after protesting against the test, complains that the county's new system fixes something that wasn't broken. "Historically," she explains, "animals at public shelters were evaluated by a very democratic system. There was input from volunteers, kennel staff, even members of the public as to the adoptability -- in terms of behavior -- of an animal. Now, one evaluator basically has displaced that democratic method of deciding which dogs live and die, which is going to be helped and which not."

Marlena Young believes there's a danger in being too democratic when deciding a dog's future. "We actually have a form that is called a Behavioral Awareness Form that volunteers can fill out and give to the staff at that shelter to review, and they can write anything on there that they want. The problem is that they want to excuse behaviors that we feel are legitimate. In other words, if a dog snaps at you, people tend to go, 'Oh well, it was just...' and we say, 'No, we want to know what the trigger was and if this is going to happen again.' We need to know more about that behavior. We don't want to ever excuse that away because it may mean that somebody is going to get hurt. There was a Doberman in New Jersey, in the last few months, that was adopted out after 87 days in a shelter where he never showed any aggression. The staff liked this dog, it was great, they adopted him out. Ten days later, he killed a woman."

Young was hired, she says, "because the county decided that they wanted to gather more information about each animal and make a decision based on that in-depth information. The only way that you can do that is to get the animal out of the kennel and do an assessment of its behavior. They hired me and said, 'We want to have some sort of a behavior-evaluation system with which we can begin to separate out these dogs not only for our purposes of moving animals through the shelter but also to protect the public from dangerous animals.' "

"Dogs that Marlena Young cleared," Starkovsky responds, "have actually bitten adopters and have been returned to the shelter and euthanized. So are her judgments any better?"

No standardized system has ever governed the animal-sheltering field. Individual shelters have always developed their own protocols for euthanizing and adopting out dogs. The Sternberg Method, in which Young was trained, is the first system to spread across the country. It was developed in a private shelter in Upstate New York by a woman named Sue Sternberg. "Her motivation," Young explains, "was that she was sending out dogs that were coming back having bitten. She wanted to stop that, so she decided we need to learn as much as we can about dog behavior by watching dog-to-dog interaction. Because dogs speak body language with the way they hold their body, tail, mouth, and eyes, and the way they move and turn and all that. So because dogs speak the same language, we watch what they are doing with one another. Then we apply that to what we are seeing when we bring these dogs in here."

Young, an energetic woman about 40 years old with short brown hair, demonstrates an evaluation in her office at the central shelter. While her trained assistant, Jenny, takes notes, she first stands silently for a minute with the dog on a leash. Later, Young explains that she wants to see if the dog is interested in making a personal connection and receiving affection from her. But the Dalmatian is indifferent. Next she begins to talk to him in a high, sweet voice and reaches down to pet the dog, who accepts one pet but shrugs off the second. Young then sits in a chair, and the dog puts his head on her lap. This time, he willingly receives affection from her. He even accepts the full-body hug Young gives him, though with noticeable tension in his body.

Young stands up and places her left hand under the dog's chin. With her right she attempts to lift his upper lip and inspect the teeth. The point is not the teeth but his reaction, which is to pull away forcefully though with no attendant growling or snapping. Young tries another 10 to 15 times with the same results. Next she tries to interest the dog in toys, but he is unmoved. He sniffs at the door while Young offers him a doggie treat. He does accept a rawhide chew but not enthusiastically. When Young reaches to take it away, he makes no attempt to resist.

Next she gives him a bowl full of canned dog food, which he begins to scarf. Using a lifelike rubber hand and arm -- about the size of a seven-year-old's -- Young grabs the bowl and pulls it away from the eating dog. This particular part of the test angers Starkovsky and other opponents to the test. "You give nice food to a dog who may have been eating out of garbage cans," she says, "or who may be sick, or may be old and have never tasted such good food, then you yank it away from him. How would you expect it to respond?"

Our Dalmatian, despite clearly enjoying the food, doesn't react to its removal in the slightest. "That is something we like to see," Young explains. "I don't care if he eats a little faster. I don't care if he eats in dog gulps, which is a little bit of defensive learning, but it is so minor that it isn't even worth talking about. If, when I reached in, he would freeze, and he might even reach up to bite the hand, we need to know that. Because dogs are in situations all the time where they get ahold of something off the counter, or they have grabbed something that the kids have left on the table, and you reach for it, and if the dog has learning problems..."

Young continues the evaluation, first yelling suddenly, "HEY!" The dog snaps to attention and looks toward her but nothing more. Next, Jenny leaves the room and, after a pause, knocks on the door and then enters. The Dalmatian curiously wags his tail and stares at Jenny, who is talking in a high sugary voice to the dog. Suddenly, Jenny turns and bolts out the door. The sudden movement arouses the dog, but he doesn't give chase.

Next, Young sits in a chair cradling a lifelike baby doll. The dog comes over, sniffs, and returns to his lying position on the floor. Then Marlena puts a realistic mechanical cat on the floor. The cat walks and makes slightly phony-sounding cat noises. The dog looks but doesn't move toward the cat.

Though she will not give the Dalmatian a final score until he's had a dog-to-dog test -- meaning, he will be put in the company of another dog (or dogs) to see how he behaves -- Young gives him a tentative score of two on a one-to-four scale, one being perfect, four being probably not adoptable.

A small minority of the dogs that come through the county shelters receive the temperament test. "Probably 5 percent of them get temperament tested," Young explains. "We are talking about 30,000 dogs that come through these three shelters. I'm training some of the employees, but right now I am the only one [who is qualified to perform the test]. And it takes two people to do it -- an observer and a handler -- so that takes two people away from their work if they are going to be doing this. Our resources are very limited in terms of which animals we can test. So we do have a preliminary behavior grade. Every day, a supervising animal-care attendant and a lieutenant walk through the kennels and look at animals that have just completed their hold period -- when they come in stray, they have to be held for four business days. We have a sort of in-kennel review test that we do for these animals, and we assign them a grade. It is either A, B, C, or D. They do very specific things, such as stand at the front of the kennel and offer a hand and see what the dog does and things like that. And they're looking for very specific reactions. The A grade would be a dog that is very friendly: comes up to the front of the kennel, ears back, eyes kind of squinty, tail low, maybe stepping a little bit, wagging his tail. And the average public recognizes those things. They are in and out of here so quick that I never see them. The grade B dogs may be at first suspicious, or apprehensive, or maybe even a little fearful, and then when you soften your body position, they react, they recover, and they come up real soft. They melt. Those dogs go out quickly. It is the ones that grade C that we concern ourselves with. They are standing; usually their body is a little taller and stiffer; their tail is usually high; they may make direct eye contact with you; they may lean forward a little bit. You offer a hand, and they jump over it, and they kind of look at you, but they are not interacting with you. Even when you soften your body position, it is still the same with them. They probably just want out. Those dogs, while they are not mean, and they are not showing us aggression in the kennel, we know there may be some issues in there. So those are the dogs that we concentrate on looking at, the ones that grade C in the kennel. The D dogs are the ones that have no recovery, that are either showing us aggression with threats -- like growling or snarling or snapping -- or, they are hidden in the back of the kennel, and when we begin to approach them, then they begin to threaten us. So they are unsafe to test."

Grade D dogs will, in all likelihood, be euthanized. That's a fact that angers Starkovsky. "Any dog who is trembling in the least, they don't even test it; it is just labeled passive-aggressive and it is killed. But all of us in the rescue business, we removed those kinds of dogs regularly, no problem at all. As soon as the dog is out and in a private home, it calms down. All it needs is a quiet environment and some affection."

Other operators of animal-rescue groups contacted for this story expressed concern over Young's system but refused to comment for the record, some for fear of being banned from the county shelters. "Rescue people are really angry," Starkovsky says, "because sometimes you will go in and try to rescue the animal, and they will say, 'Oh no, you can't have it; it has been graded level 4. We are going to euthanize it; that is it.' If you beg and plead enough, maybe you will get the dog, but then maybe you won't. I went to bat for a couple of dogs, and they just said, 'No, absolutely not, we are going to kill it.' It's almost like genocide."