Français German Espagnol Portugese Chinese Japanese Arab Rusian Italian Norvegian Dutch Hebrew Polish Turkish Hindi
Accueil du site > 07 - Livre Sept : SOCIOLOGIE > 7-12 Sociobiology - A Caricature of Selection Theory

7-12 Sociobiology - A Caricature of Selection Theory

mardi 18 mars 2008, par Robert Paris

Sociobiology- A Caricature of Selection Theory

Richard C. Lewontin Darwin’s

theory of evolution by natural selection was devised by him from observations on the domestication of animals. The first chapter of the Origin of Species is devoted to a discussion of selective breeding and variation in domesticated animals and pigeons in particular. The human breeder had selected for particular traits in reproducing these domesticated species, and Darwin transferred this notion of selective breeding from the domain of domestication to the natural world. The analogous agent to human willful selection was "natural" selection, a concept that occurred to Darwin after reading Malthus’ essay on population. Natural selection would occur because organisms struggled to survive in a h6stile nature and only some of the many organisms born would live to reproduce. Darwin’s theory of evolution by natural selection, then, rests on three general principles which are unchallenged in their generality. I) There is variation in morphology, physiology and behavior among organisms belonging to the same species - the principle of variation. 2) There is a correlation between parents and offspring in phenotype so that relatives resemble each other more than do unrelated individuals
-  the principle of heredity. 3) Some phenotypes leave more offspring than others - the principle of natural selection. These three principles are sufficient to guarantee an evolutionary process. Provided there is variation among objects, that there is some temporal stability in this variation by some mechanism of heritabilit_ and that different sorts of objects leave different numbers of descendants in time, there must be evolu- tion•ary change in the composition of the population. So, rocks evolve by natural selection since they vary in hardness, split off rocks of equal hardness and have different rates of erosion and t_erefore of survival. So too automobiles evolve by natural selection as do soft drink containers. The system of explana- tion is so powerful that it can be applied to almost any situation and herein lies its weakness. A system of explanation that can potentially be used to explain any observations invites caricature and will be used in a crude and vulgar analogical way by ingeniouspeople. This is what happened to the system of Freudian psychology which was so all encompassing that it has been used to explain all of history, science and the arts. So too, the Darwinian theory has been vulgarized for the purpose of easy explanation of phenomena. The latest episode in this caricature of Darwinian explanation is the collection of theories speculations and observations about animal and human behavior that is called by its adherents "Sociobiology". Sociobiology is an attempt to explain all of anima_ and human behavior as the product of evolution by natural•selection. This includes not only the stereo- typed individual and group behavior of lower organisms, but all aspects of human social and individual activity that is within the normal human gambit. E.O. Wilson’s "Sociobiology, The New Synthesis" explicitly claims the arts, letters, music, ethics, history, economics, all the studies of the humanities and the social sciences are to be subsumed under sociobiology which will "biologize" these disciplines and give a new scientific basis both for understanding human society and for directing it in the future. An examination of the structure of sociobiology reveals, however, that it is not a science, but a system of unscientific specula- tion. The first element in the sociobiological argument is to describe the set of i phenotypes under investigation. This is done by making very general and very superficial characterizations of "human nature" by universalizing conventional wisdom. Thus in Wilson’s "Sociobiology’ I, we are told that "men would rather believe than know" and the people are extraordinarily easy to indoctrinate, indeed they seek it". Xenophobia, domination, entrepreneurship, territoriality, male dominance, are all said to be universals of human behavior and then provided with a biological explanation. The facts of history and of ethnography do not support the universality of these traits, but history is almost completely ignored by sociobiologists and exceptions to these generalizations in the ethnographic record are accounted for by redefinition. For example, it is stated the exceptions to the "rule ’’of genocidal warfare are only "temporary aberrations" or that the reason all human societies do not appear to be territorial is that "zoologists have been too narrow in their definition of territoriality". In some cases their claims are directly contradicted by the ethnographic record. For example, present day "primitive" hunter and gatherer societies do not engage in genocidal warfare, an invention of the modern state, but, on the contrary, engage in a kind of semi- ritual combat in which very few combatants are killed or wounded. In general, the description of human behavior by sociobiologists is remarkable for its lack of rigor and sophistication. There are three confusions that are common. The first is the error of arbitrary agglomeration. The totality of human behavior is broken up arbitrarily into units of description that may or may not be related to actual phenotypic entities corresponding to gene action and the action of selection. It is now known that the topology of the central nervous system corresponds rather poorly to the topology of central nervous function. For example, specific memories are stored diffusely rather than being located at specific points in the cerebrum. Even for ordinary morphology we are not certain what the units in evolution are. For example, is the hand a unit in evolution, or is a single finger, or a joint in that finger ? How are we to decide ? The difficulty of dividing up the description of behavior into appropriate units is far greater than for morphology. Yet sociobiology has never discussed this pro- blem. A second error is that of reification, in which mental constructs are mistaken for real objects. Are "entrepreneurship" or "dominance" or "altruism" real entities that evolve in time and have genetic influences or are they arbitrary ways of describing human social interactions that are conditioned by history and the particular form of society in which sociobiologists live ? Again, in morphology, such arbitrary constructs exist, as for example, the chin, which appears not to be an organ in evolution, but only a way of describing the relative positions of the alveolar and dentary jaw fields. Third, there is the error of conflation in which quite different phenomena are confused because they are given the same name. The most striking example is aggression, a term which is applied indiscriminately by sociobiologists to the interaction that occurs between individuals when one assaults another as the result of some individual insult or competition, and to the political struggle between nation states. As a result, there is a facile assignment of individual aggression as the cause of war. Yet everyone who has ever fought in a war knows that most people are there not because they feel aggressive toward others, but because they have been conscripted by the state in the interest of a political goal. Having established a description of the human behavioral phenotype, the next task of the sociobiologist is to ascribe a genetic basis or show a genetic influence on the characters described. It is here that sociobiological theory is weakest because it is here that we demand some experimental evidence and some rigorous concepts of quantitative genetics. Here sociobiologists try to have it both ways. For some traits it is claimed that "moderate heritabilities have been documented", so providing a basis for selection. Other traits, however, are said to be constant over the entire human species, so that the lack of genetic variance is a proof that they were fixed in the genes by previous selection. It appears then that no observation would contradict the claim of genetic determination of human behavior. But the understanding of the basic concepts of quantitative genetics seems to be particularly low. The claim of "moderate heritability" for a variety of human behavioral traits such as creativity, motor skills, dominance, personality traits, and so on, is based on the observations of parent-offspring correlations in these traits. Yet no separation of the environments of parent and offspring has been made so that the parent-offspring correlation is a total phenotypic correlation that cannot be resolved into a genetic and an environmental component. The central feature of human social organization is that it is organized around families and other groups of relatives so that genetic relationships also result in environ- mental similarity. Based on observed parent-offspring correlations, the highest "heritabilities" in U. S. populations are for religious affiliation and political party ! The fact is that we do not have a decent heritability estimate for any human behavioral trait because no one has succeeded in separating the genetic and environmental components of correlation in human populations. Sociobiologists speak vaguely of i0 per cent of human behavior being governed by genes, yet no geneticist would know how to interpret such a statement. Does it mean that all human behavioral traits have a i0 per cent heritability (broad or narrow ?) ? Often sociobiologists will simply postulate or assume genes with special and bizarre action in order to account for complex human behaviors. Thus, there are "conformer genes", genes for homosexuality, genes for altruism, all invented to make the theory fly. Clearly, sociobiology rests on a very shaky and uncertain genetic base. Having described human behavior in a conventional way, and having postulated genes wherever they are needed, the sociobiological theorist now proceeds to show how natural selection has established the trait. Again, no evidence is offered for the action of.natural selection nor can any be offered because we are entirely in the realm of past human evolutionary history. All that can be offered are imaginative speculations about how a trait might have conferred greater fitness on its carriers. This kind of speculation is easy although it may occasionally demand some ingenuity. Sociobiologists have been aided in their construction of modern Just So Stories by expanding the concept of natural selection. First, an attempt is made to explain a character by direct selective advantage to the indi- vidual possessing the trait. For example, a more aggressive person would get more food, if food were in short supply, and so survive better and leave more offsprinl Thus, genes for aggression and entrepreneurship would increase in frequency. Other traits seem maladaptive to the individual and his or her offspring. Homo- sexuality, for example. It is asserted (without evidence) that homosexuals have fewer offspring than heterosexuals (obviously true for complete homosexuals, but not necessarily for those showing mixed behavior). HOW is it that the genes for homosexuality (invented with no evidence) have not disappeared ? Because of kin selection, the second mode of selection invoked by sociobiologists. In kin selection, an individual’s genes are spread by sacrificing its own direct reproductive fitness in favor of relatives who may share the same genes. The homosexual, by helping to raise his or her sibs children may spread his or her genes according to this theory. In particular, one can sacrifice his own fitness for eight first-cousins, for example, and come out even. But what about human behavior that is directed toward increasing the fitnessl of non-related persons, pure altruism ? This is explained by the sociobiologists as a result of selection for reciprocal altruism. If an individual risks his or her life for an unrelated person, that person will remember and reciprocate in the future, thus propagating the genes of both. Then genes for such reciprocal behavior will increase, as will genes for cheating in this exchange. There is no conceivable behavior that could not be explained by recourse either to direct selection, kin selection or reciprocal altruism. The system is complete. There is one possibility of testing sociobiological theory quantitatively. Historical changes of major magnitude have occurred in the organization of human societies in remarkably short time periods, as for example the rise and fall of the Muslim Empire in a few centuries. But we know how raPid gene frequency changes can be and taking account of the fact that only four human generations pass per century, it would be hard to account for these dramatic historical changes geneti- cally. The sociobiologists take care of this chink in their armor by inventing a wholly new p_inciple, the multiplier effect, which states that an arbitrarily small genetic change in a species may be multiplied to an arbitrarily large pheno- typic effect. By allowing an arbitrary but unspecified multiplier effect, of course any observed phenotypic change in the species can be explained by small genetic changes. Thus the theory is made totally test-proof. There is no doubt that human behavior is, in some sense, the product of evolution. But the human species has reversed the relationship of organism and environment that dominates other species. For all living organisms, the environ- ment is more than an outside force that impinges on the organism, the organism coping in some passive way. The organism and the environment interpenetrate each other so that environment has a role in making the organism and the organism alters the environment. For lower animals, the effect of the organism on the environment is relatively small, but the human species is characterized by being a molder of the environment and therefore of its own evolution. The human species dominates the environment and not the reverse. Therefore the proper study of human evolu- tion and a proper understanding of human social organization cannot come from a facile and superficial analogy with natural history. It is up tO those of us whose domain of study is the exact science Of natural and artificial selection and population genetics to expose the pseudoscientific attempts to short_circuit a proper understanding of our very complex species. Above all, the human species must be understood in its uniquely human aspects. DR. RICHARD C. LEWONTIN
- A CARICATURE OF SELECTION T HEORY" ANDREW LEE : Is not part of the problem purely mathematical in nature  ? The rate of _0 or X is not defined. There are very wide variety of models •with 0 0 n parameters to fit n points. RICHARD C. LEWONTIN : Yes, that puts it in a nutshell. If there is said to be no phenotypic variance (the denominator) and no genotyplc variance (the numerator), then the problem of heritability becomes indeterminate. Moreover, the structure of sociobiological theory is such that it provides a few undetermined parameters which allow the fitting of any arbitrary data. The old saying is : "Give me three parameters and I can draw an elephant, give me four and I’llmake him walk." FRED SCHULTZ  : Are the TV networks’ mouthpieces for the Sociobiologlst or indirectly influenced by them (or are publishing concerns mentioned by you) ? Not asked directly at this time - are there any signs of the development of a "New Social Biology" based on genetic theory ? (this question is perhaps related to Gordon Dickerson’s question). RICHARD C. LEWONTIN : The media, including television, radio, newspapers and popular magazines have given a major play to sociobiology, especially just after Wilson’s book, "Sociobiology, The New Synthesis", appeared. There were full page advertisements in the New York Times (unheard of for a supposedly scientific work), major reviews and articles in many daily newspapers, an interview in People magazine, claims in Newsweek, Tim_____e, etc .... This makes part of a wave of "New Social Biology" or Social Darwinism which includes the theory that criminality is in the chromosomes, that social upheaval is the result of defective wiring in the central nervous system, that success in society is the result of genetically determined intelligence, etc.... The economist Paul Samuelson labelled this, and sociobiology in particular, as a new Social Darwinism in his article in Newsweek three years ago and in the interval many economists have shown him to be correct in his assessment by trying seriously to explain the American economy as being based in the genes. JAN S, GAVORA : Your talk makes us wonder whether you are arguing against Sociobiology because you really "underneath" believe in it ? My serious question is : How strong a stronghold sociobiology has and how much damage it is going to make in Genetics and Biology ? In my opinion, it does not have much chance among people who breed chickens or work experimentally, but it may be more dangerous in other areas. RICHARD Co LEWONTIN : I agree that sociobiologydoes not, ultimately, have much of a future in biology because it is not really a fruitful scientific theory. But it has taken hold in the social sciences and among those biologists who work in the "fuzzy" areas of evolutionary theory and animal behavior. That is why it is all the more important for people like us, breeders, population geneticists and statisticians, to insist on a rigorous level of scientific work consistent with the soundest methodological principles, and above all, publicly to correct errors where we see them. GORDON DICKERSON : You object to prostitution of legitimate genetics and evolutionary science in the pseudo science of sociobiology. This a defensive or negative approach to protect our science. However, defensive measures usually are less effective than offensive or positive ones. Do you have suggestions concerning positive approaches to "understanding realities of human nature, and incorporating such understanding the planning of more generally satisfactory social organization ? RICHARD C. LEWONTIN : I freely admit that the struggle against socio- biology is largely a defensive one. That is inevitable for us as practitioners of an exact science. There will continue to be attempts to use our science superficially tO support prejudices and to try to erect all-encompassing theories about the world. That is a position we necessarily find ourselves in exactly because we work in a rigorous experimental and theoretical discipline. The only positive contribution we can make to building a more satisfactory social organization is to insist on telling the truth about biology and especially about the real nature of interactions between genotype and environment inproducing pheno- type. My own guess is that there is not much in human biology that will constrain future human societies, but that is only my prejudice.

Répondre à cet article

SPIP | squelette | | Plan du site | Suivre la vie du site RSS 2.0