Chapter 1
Breed, Behavior and Temperament
What do we mean by "temperament"? We may think of terms such as loyal, congenial, affectionate, regal, aristocratic, tranquil, eager to please, shy, happy-go-lucky, attentive, alert, self-confident, energetic, obedient, sweet, rollicking, courageous, intelligent, gentle, upstanding and quiet. On the other side of the coin, we may think of words such as aggressive, mean, stupid, lazy, quarrelsome, aloof and so on.
These terms, normally used to describe a dog's behavior by everyday folk and by the American Kennel Club, are useful to the extent that they enable us to predict the behavior of a dog so labeled. For example, you would expect a shy dog to avoid strangers and novel objects, an affectionate dog to engage in a lot of nuzzling and licking, an obedient dog to train easily and retain its skills a long time. A rollicking dog would be one that does a lot of jumping and running about. A quiet dog is one that probably doesn't bark or howl very much. An attentive or alert dog is one that notices subtle changes in its environment or in its owners.
The problem, of course, is that there is no generally accepted definition for any of these potentially useful words. Even though we have just defined them my way, the average person may not agree with me or simply proceed to use the words any old way that pleases him or her at the moment.
In addition, these words have not been objectified by identifying, through a repeatable set of operations, the behaviors that go along with them. Since there is no standard set of operational definitions, we are free to use the words any way we choose, and we usually do. The result is labeling chaos.
To add to this chaos, we also use words like regal, aristocratic, congenial, courageous, upstanding, happy-go-lucky, sweet, quarrelsome and so on. These words do not reflect any definite canine behavior. I have never witnessed a quarrel among dogs, nor even a heated discussion, but I have seen a dog bark and fight with other dogs. Upstanding could mean standing erect and upright, but it is probably used in the sense of morally right and honest. You know, the kind of dog you would vote for.
There is a further problem in using such terms to describe temperament in dogs. Because all the previously mentioned terms are frequently used to describe human behavior too, we can fall into the trap of "anthropomorphizing" (i.e., attributing human characteristics to) the looks, temperament and behavior of dogs. This is an error of logic. It is best to remember that even though an English Bulldog may remind us of Winston Churchill, it would not really make an effective Prime Minister of England.
Canine Behavior
In spite of the fact that people frequently treat their pets like humans and think of their dog's behavior in human terms, dogs are canines. They are a member of the family Canidae, which includes wolves, coyotes, foxes, jackals and some obscure canids from Asia, Africa and South America. Our pets are also members of the genus Canis and the species familiaris. Table 1 provides a list of our dogs' family relatives.
Our pet dog, Canis familiaris, has roots that can be traced back some forty million years. It is commonly believed that the domestic dog, Canis familiaris, was developed from the Eurasian wolf, Canis lupus, beginning some twelve thousand years ago as the people who captured and domesticated wolf puppies started the practice of selective breeding. Each culture probably had a concept of what would be the most desirable behavioral and physical characteristics of the captured canines and mated those animals which exhibited these traits. Voilà! Twelve thousand years later we have 123 or more distinct breeds in the United States (500 throughout the world) varying in size from the Chihuahua, which can be 4 to 6 inches at the shoulder and weigh less than six pounds, to the Mastiff, which can be as much as 34 inches at the shoulder and weigh over 200 pounds.
Only recently has the study of animal behavior in general, and that of Canis familiaris in particular, come under scientific scrutiny. Researchers in this area most frequently are interested in identifying the genetic determinants of behavior and temperament (instinctual behavior) and separating such behavior from behaviors and traits that develop as a result of reinforced practice (learned behavior).
Around 1950, Dr. John Paul Scott and Dr. John L. Fuller started what turned out to be a fifteen-year project on the genetics of canine social behavior. Some of their observations will be incorporated herein. For our purposes, the important conclusion to be drawn from this work is that canine temperament was reliably shown to vary with breeds, at least in the five breeds these researchers tested. It's a pretty safe inference that the rest of the breeds vary systematically in temperament as well.
There are certain behaviors that have developed over millions of years of evolutionary trial and error. Some of these behaviors are common to the whole family of Canidae and can be called family-typical behaviors. The set of behaviors common to the genus Canis are genus-typical behaviors. The set of behaviors common to the species Canis familiaris are species-typical behaviors. The set of behaviors common to specific breeds of dogs are breed-typical behaviors. Finally, each individual dog may exhibit more or less of these family, genus, species and breed typical behaviors in its individual-typical behavior. Thus I define temperament, in part, as the degree to which an individual dog exhibits these individual typical behaviors.
Table 2 gives some examples of family typical and genus typical facial expressions in canids.
You can see from Table 2 that some facial expressions are common to all canids. These are family-typical behaviors. For example, all canids show a submissive grin by pulling their lips way back. This facial expression seems to indicate to other canids and even to some other species, "You're the master; I don't want to fight, and I'm not going to run away. So don't be aggressive toward me; let's be friends and I'll obey you." It's amazing what a smile can say.
A genus-typical behavior would be common only to dogs, wolves and coyotes. Submissive rooting is genus-typical. Your dog does this when he comes up to you and nuzzles his head under your hand or arm and pushes up or forward. Some people would call this a kind of cuddling behavior. It also occurs when your dog puts his paw on your hand or lap. Your dog is probably saying through this behavior, "Pay attention to me, scratch my head, pet me."
I have included another column in Table 2 labeled "Primate Counterpart." I did this to show how we as humans can sometimes understand what our dogs are trying to communicate. Human beings, Homo sapiens, are primates. We share some common behavioral characteristics with our simian cousins, the Great Apes, Orangutans and Chimps. Interestingly, these primate-typical behaviors are sometimes similar to our canine comrades' genus-typical behaviors. For example, smiling in people probably means about the same thing as the submissive grin in dogs. They both say, "I'm friendly; I'm not going to be a threat; in fact, I acknowledge your dominance over me."
This is probably why humans have been able to cohabit with canines for so long. We are both pack or social animals and share some social signals. This is also why humans have the unfortunate habit of anthropomorphizing canine behavior. We interpret the behavior of dogs as if it were human and give dog behavior human labels. Sometimes our interpretations are on the mark, especially when there is a congruence between human and canine social signals. This lulls us into believing that our interpretations will be always right. This conclusion, of course, is in error, because there are an equal number of social signals we don't share with dogs. For the most part, men don't define the boundaries of their territory by lifting their legs and urinating on the nearest tree or bush. Thus, it would be better for us to avoid anthropomorphizing our dog's temperamental characteristics. Rather, we need to categorize a dog's temperament along dimensions of canine genus- and species-typical behavior.
Dimensions of Temperament
Categorizing a dog's temperament along dimensions has two advantages over the use of terms like quarrelsome, congenial, loyal, etc., that I mentioned earlier. First, it provides "evolutionary validity" for our definition of temperament. That is, it connects the definition to the body of knowledge that has been accumulated on the evolutionary and genetic determinants of canine behavior. Secondly, a consideration of "dimen...
