Genetics and the Social Behaviour of the Dog

John Paul Scott, John L. Fuller

• Publisher: University of Chicago Press
• Year: 2012
• ISBN:
  9780226190822
• Topic: Biowissenschaften
• Subtopic: Wissenschaft Allgemein

The classic study of dog behavior gathered into one volume. Based on twenty years of research at the Jackson Laboratory, this is the single most important and comprehensive reference work on the behavior of dogs ever complied."Genetics and the Social Behavior of the Dog is one of the most important texts on canine behavior published to date. Anyone interested in breeding, training, or canine behavior must own this book."—Wayne Hunthausen, D.V.M., Director of Animal Behavior Consultations"This pioneering research on dog behavioral genetics is a timeless classic for all serious students of ethology and canine behavior."—Dr. Michael Fox, Senior Advisor to the President, The Humane Society of the United States"A major authoritative work.... Immensely rewarding reading for anyone concerned with dog-breeding."— Times Literary Supplement "The last comprehensive study [of dog behavior] was concluded more than thirty years ago, when John Paul Scott and John L. Fuller published their seminal work Genetics and the Social Behavior of the Dog." —Mark Derr, The Atlantic Monthly "Genetics and the Social Behavior of the Dog is essential reading for anyone involved in the breeding of dogs. No breeder can afford to ignore the principles of proper socialization first discovered and articulated in this landmark study."-The Monks of New Skete, authors of How to Be Your Dog's Best Friend and the video series Raising Your Dog with the Monks of New Skete.

PART I

THE DEVELOPMENT OF BASIC BEHAVIOR PATTERNS

CHAPTER 1

A SCHOOL FOR DOGS

For many years we have had psychologists and zoologists working side by side in our behavior laboratories. When we first brought them together, we often got stereotyped reactions to any new discovery in the field of behavior. A zoologist would say, “What a wonderful example of instinct,” and a psychologist would counter, “My, what a remarkable example of learning!” This happened over and over again in the early stages of our work, and it illustrates the confusion which once existed, and may still persist, regarding the role of heredity in relation to behavior.

In the more remote past, scientists such as John B. Watson took the view that heredity had little importance and that the behavior of a human infant could be molded in any direction by learning. Others, following Sigmund Freud, held that human behavior was deeply rooted in instincts whose suppression could result in serious mental disorders. The vast majority of psychologists, psychiatrists, and other practical workers in the field of human behavior had no definite information about either the rapidly developing science of genetics or the ways in which genetic variation could affect behavior directly. Nor was the clinical observation of human behavior much help. True, many children reared in poor family backgrounds later developed inadequate and maladjusted behavior, but some of them were apparently unaffected, and still others worked harder than most people to compensate for their early experience. Was their ability to rise above environment due to heredity or to some unknown difference in early experience?

The question “What does heredity do to behavior?” is therefore a basic one, both theoretically and practically. There are many different ways of trying to answer it. One can either make comparisons between as many as possible of the million or so species of animals in the world or concentrate on one, as we did. Scientists working with birds have discovered many stereotyped behavior patterns which are largely organized and governed by heredity. We ourselves chose a mammal, partly because mammals are more closely related to human beings than are birds, and partly because the job had not yet been done. Among mammals, we chose the dog because it shows one of the basic hereditary characteristics of human behavior: a high degree of individual variability. We reasoned that if there were any animal most likely to show important effects of heredity on behavior, it was the dog, and that these effects could be assumed to be at least somewhat similar to those which appear in human beings.

To put this in more enthusiastic terms, the dog is a veritable genetic gold mine. Besides the enormous differences between breeds, all sorts of individual differences appear at the stroke of the geneticist’s pickaxe, in this case the technique of mating two closely related animals. Anyone who wishes to understand a human behavior trait or hereditary disease can usually find the corresponding condition in dogs with very little effort. Dogs are timid or confident, peaceful or aggressive, and may be born with undershot jaws, club feet, or hemophilia (Dawson, 1937; Burns, 1952; Fuller, 1960).

Although it is true that a dog is a large, long-lived, and relatively expensive animal to keep, it is a good deal smaller, shorter-lived, and less expensive than some other mammals. Dogs breed more rapidly than primates or even the common herd animals, except for pigs. Dogs can pass through one complete generation within a year. A mature female continues to reproduce for five to seven years, bearing her young about twice a year in litters averaging four or five animals. Nevertheless, it requires very careful planning to get the maximum amount of information from an experiment with dogs, and one initial mistake can ruin years of effort.

Therefore, before our work had gone beyond the early planning stage, we asked a number of leading geneticists and psychologists in the United States to attend a research planning conference at Bar Harbor. Many of these men and women were doing related research, either on animals or on people. The chairman was Dr. Robert M. Yerkes, one of the pioneer workers on heredity and behavior, the founder of the Yerkes Laboratory of Primate Biology, and at that time the recognized dean of animal behaviorists in the United States. Equally distinguished workers made up the rest of the group. Dr. Frank Beach served as secretary, organizing the proceedings into a coherent whole. For three days we met with them on Mount Desert Island and discussed the direction which the research should take. Much of our final plan resulted from this conference (Scott and Beach, 1947).

At the start we had certain guiding principles and signposts derived from what had been discovered about other aspects of heredity. We knew that behavior itself is not inherited. In fact, the only things which are biologically inherited from the parents are the egg and sperm. We also knew that the basic factors of heredity are genes, which are nothing more than large organic molecules arranged in larger bodies called chromosomes. Even these are only parts of the nucleus of a living cell. The genes themselves can act only by controlling the synthesis of proteins, some of which, the enzymes, modify other chemical or physiological reactions, usually in the cells themselves. All this is a long way from the activity of a whole organism, or behavior, and the concrete fact of whether a dog bites you or wags its tail.

One of the primary questions we had to answer was whether heredity could produce an important effect upon behavior in a higher animal or whether it simply set the stage for behavioral activity which was then guided and molded by other causes. The question makes a great deal of sense in relation to the concept of levels of organization (see Table 17.1). Heredity stands at the most basic level, and therefore we can assume it to be very important. But assumptions are only the beginning of scientific inquiry, and the question still remains, What exactly can a factor operating on a low level of organization do to a phenomenon operating on a higher level? For behavior is primarily a phenomenon existing on the organismic or psychological level (Wright, 1953).

At the same time we realized that we could not neglect factors operating on still other levels. A major trend in animal behavior research in the last thirty years has been to demonstrate the importance of the effects of social organization upon behavior. The activity of an individual is understandable only in terms of his relationships to the members of a larger group. Our work was therefore designed in relation to levels of organization and can be so described.

In addition, we had the successes and mistakes of our predecessors to guide us. Professor Tryon (1930–41, 1963) of the Psychology Department of the University of California had previously done extensive and meticulous research on heredity and behavior in rats. He devised a mechanical maze operating in such a way that the experimenter could not influence the results consciously or unconsciously and selected different strains of rats from those which learned the maze most quickly or slowly. These differences persisted for generation after generation, indicating that heredity could produce an important effect upon behavior. It appeared that selection had affected intelligence alone, but when a student of his, Searle (1949), tested these rats on different sorts of mazes, he came to the conclusion that one of the principal reasons why the two strains of rats differed in performance was that the slow running or “dull” rats were afraid of the mechanical maze. Instead of selecting purely for intellectual ability, Tryon had also selected for a difference in emotional reaction.

C. S. Hall (1941) had shown by his selection experiments on the defecation rates of rats taken from their cages and placed in an open space that heredity could alter another kind of emotionality. His work, and that of Searle, suggested that emotional capacities were highly variable and could produce important effects upon performance. From these previous experiments we knew that hereditary emotional differences might have important effects on behavior which at first glance did not appear to be emotional. We also realized that with such a long-lived and expensive animal as the dog, we would seldom be able to duplicate our experiments. In other words, there would be very little going back to take a second look. We would have to measure as many important characteristics as we could while the experiment was in progress.

Besides the work on other animals, there had been an earlier major attempt to work with heredity in dogs. The late Professor C. R. Stockard of the Cornell University Medical School had begun with the idea (a promising one at the time) that many of the differences between breeds of dogs were caused by hereditary disorders in the ductless glands (1941). He did, indeed, find breed differences in the endocrine glands, but anatomical peculiarities in the same breeds, such as dwarfism and shortleggedness, turned out not to be caused by glands but by hereditary factors working in other ways. Professor Stockard died before he could change the project to allow for this. Since his original experiment was not designed to test a Mendelian hypothesis, the results were inconclusive, and we decided to set up our experiment in the broadest possible way, realizing that we could not predict the important results in advance and that if we chose too narrow an approach we might end with nothing. Both from this viewpoint and the simple statistics of life insurance tables, we knew it was imperative that this project be a cooperative venture, not a one-man show, and in this book we shall try to recognize the innumerable contributions made by co-operators and co-workers in the years covered by the experiment.

Our over-all experimental design was to systematically vary the genetic constitution of the dogs while keeping all other factors as constant as possible. The dependent variable was behavior, but since behavior occurs only in reaction to stimulation, we had not only the problem of designing a uniform system of test situations which would measure the dog’s reactions to specified changes in his social and physical environment, but also that of keeping all other environmental factors as constant as possible. Equally important was the task of keeping the dog’s internal environment constant by seeing that each animal was adequately nourished and free of disease.

GENETIC METHODS

Choice of breeds.—Genetic variation was the primary variable in our experiment: the supposed cause which might or might not affect behavior. We therefore divided the experiment into two phases. In the first we made a survey of the different dog breeds in order to select those which showed the biggest behavioral differences. Thanks to the energy of Dr. C. C. Little, we had a wide variety from which to choose, ranging from the Great Pyrenees, a white dog roughly resembling a Saint Bernard, to the tiny Chihuahuas. We soon eliminated the abnormal sizes at both ends of the scale, realizing that the large breeds were too expensive for long experiments and that the toy or dwarf breeds were unsuitable because of their low fertility. Furthermore, if we chose dogs of approximately the same size, we could use the same apparatus for all.

Certain breeds were discarded early. Both dachshunds and Scottish terriers have a reputation for being stubborn and self-willed, but those we studied showed no trace of this as puppies, being uniformly friendly and demonstrative animals, not unlike beagles and cockers in this respect. We would obviously have to wait too long for the behavioral peculiarities of these breeds to develop.

We further decided to concentrate on breeds with normal physique. Although it is true that the shortleggedness of the dachshund, for example, has an effect upon its behavior, this does not require an elaborate scientific experiment for proof. We preferred to work with the more subtle and perhaps more important ways in which heredity affects behavior.

We finally selected five breeds for intensive study: basenjis, beagles, American cocker spaniels, Shetland sheep dogs (shelties), and wire-haired fox terriers. As it turned out, these were representatives of the major groups of dogs as recognized by dog breeders and omitted only the toy and non-sporting breeds.

In the first phase in the experiment, we developed our tests and raised all five breeds in the same environment, measuring their similarities and differences. We also studied the development of their behavior by making daily observations from birth up to 16 weeks of age. Wherever possible, we cross-fostered puppies between the breeds in an effort to see whether the maternal environment as well as biological heredity was affecting behavior. We also experimented with an even more drastic shift in the social environment by taking certain puppies out of the kennels and rearing them in homes. This not only gave us a chance to study the effect of maternal environment, but to see whether or not our kennel-reared dogs were reasonably normal compared to dogs reared outside. Unfortunately, we were able to do this only on a very few animals, partly because of the limited number of friends and fellow workers on whom we could impose, and partly because of the danger of disease transmission. Anticipating a little, we can say that Gyp, the basenji; George, the beagle; and Silver, the Shetland sheep dog, did not do strikingly better or worse on most of the tests, although they developed very different relationships with dogs and people.

A Mendelian cross.—The second phase of the experiment was to make a Mendelian cross between two of the breeds (Fig. 1.1). Preliminary results indicated that the American cocker spaniel and the African basenji showed a great many differences in behavior. On the one hand we had a gentle and pampered house pet descended from bird dogs, and on the other a breed which had recently come from the rough-and-tumble conditions of African village life. Furthermore, there was good evidence that the basenjis, having been first brought over from Africa in 1937, had never been crossed with the European breeds, whereas the cocker (as well as the other three) had originally come from the British Isles, where there had been abundant opportunity for crossing in the past and considerable historical evidence that this had actually taken place. These two breeds had been made...