For biologist Raymond Coppinger, who has been studying dogs and other canines for almost three decades, the question has been inescapable. Why do dogs bark?
In a nation of 52 million domesticated canines, it's of more than scientific interest. There is probably no neighborhood that has been spared ordeal by barking. And Mr. Coppinger, a biologist at Hampshire College in Amherst, Mass., notes that one of the drawbacks of his own research is the infernal racket.
But the more he probed, the more puzzling barking seemed. Not only do dogs waste a great deal of energy barking, but they bark at anything and everything. Unlike the vocalizations by other animals, dog barking seems to be more noise than communication.
After four years of working on canine behavior and vocalization, Mr. Coppinger and his colleague, linguist Mark Feinstein,
recently concluded that the answer is exactly what some people always suspected: Dogs bark for no good reason.
Basically, they argue, barking is just an accidental by-product of dogs' evolution.
The notion is controversial, however. Other researchers believe strongly that humans have bred many dogs deliberately to encourage barking.
The Hampshire researchers' theory is based on the belief, shared by a number of other experts, that dogs are wolves that never grow up. Although the issue has not been settled, it is likely that the wolf was the ancestor of the domestic dog. The two species, Canis lupus and Canis familiaris, are so similar genetically they can interbreed and produce viable, fertile offspring.
But wolves and dogs differ strikingly in their vocal behavior. Adult wolves and coyotes seldom bark, and when they do, it is brief and isolated. Unlike dogs, Mr. Coppinger and Mr. Feinstein note, wolves do not bark "in long, rhythmic stanzas." Wolf pups, on the other hand, bark repetitively.
Mr. Coppinger believes that dogs gradually domesticated themselves, as tamer wolves began to hang around the edges of human settlements for easy scavenging. Like barking, tameness another characteristic of juvenile wolves.
Gradually, Mr. Coppinger and Mr. Feinstein argue, these tamer, more juvenile wolves were transformed into dogs by evolutionary changes in the genes that control development to maturity. If the timing of the regulatory genes were slowed, dogs would retain for a lifetime the tameness and other characteristics typical of young wolves.
In their view, tameness was the trait that drove dog evolution. Barking, they hypothesize, is simply another adolescent trait that emerged coincidentally.
"Barking in dogs, we think, simply came along as a part of a whole package of changes wrought by a genetic alteration in the timing of the life cycle of the ancestral canid," they wrote in Smithsonian magazine. "Stuck in adolescence, the dog barks so much because barking is what a juvenile canid does."
As evidence, Mr. Coppinger and Mr. Feinstein cite a 30-year breeding experiment on foxes conducted by Soviet biologist Dmitry K. Belyaev. His goal was to breed docility into wild foxes that were raised for their fur in order to make them more manageable.
But after breeding the tamest animals for 20 generations, Mr. Belyaev found he had created what seemed like a new animal, a creature that didn't look or act like a fox. It approached humans rather than fleeing; it went into heat twice a year like dogs instead of only once; it had floppy ears and multicolored coats; and, according to Mr. Belyaev, it even sounded like a dog.
Some dog breeds, such as the Alaskan malamute, bark seldom, if ever. In an interview, Mr. Coppinger speculated that these dogs, which look and often act more like their wild relatives, might be less stalled in adolescence than other domesticated breeds. That speculation, is just a hypothesis "to be treated gingerly."
Another leading researcher, Benson Ginsburg of the University of Connecticut, agrees with the notion that dogs are developmentally retarded wolves, but he disputes the theory that barking is largely an accident of evolution.
Mr. Ginsburg specializes in the genetics of behavior, and in more than 30 years he has worked with 13 breeds of dogs, wolves, coyotes and coyote-beagle hybrids. He describes dogs as "genetic erector sets," whose parts can be assembled in an infinite variety of combinations, rather than having packages of traits that tend to appear together, as the Hampshire team hypothesizes.
Mr. Ginsburg argues that the cluster of changes in Mr. Belyaev's foxes were a product of inbreeding in a small population, rather than evidence that other traits are genetically linked to tameness or juvenilization.
Mr. Ginsburg says he does not know why dogs began to bark, but he is certain that "human breeding has shaped dogs' barking behavior.