Cover image for Evolution's eye : a systems view of the biology-culture divide
Evolution's eye : a systems view of the biology-culture divide
Oyama, Susan.
Personal Author:
Publication Information:
Durham [N.C.] : Duke University Press, 2000.
Physical Description:
274 pages : illustrations ; 25 cm.

Format :


Call Number
Material Type
Home Location
Item Holds
BF713 .O93 2000 Adult Non-Fiction Central Closed Stacks

On Order



In recent decades, Susan Oyama and her colleagues in the burgeoning field of developmental systems theory have rejected the determinism inherent in the nature/nurture debate, arguing that behavior cannot be reduced to distinct biological or environmental causes. In Evolution's Eye Oyama elaborates on her pioneering work on developmental systems by spelling out that work's implications for the fields of evolutionary theory, developmental and social psychology, feminism, and epistemology. Her approach profoundly alters our understanding of the biological processes of development and evolution and the interrelationships between them.
While acknowledging that, in an uncertain world, it is easy to "blame it on the genes," Oyama claims that the renewed trend toward genetic determinism colors the way we think about everything from human evolution to sexual orientation and personal responsibility. She presents instead a view that focuses on how a wide variety of developmental factors interact in the multileveled developmental systems that give rise to organisms. Shifting attention away from genes and the environment as causes for behavior, she convincingly shows the benefits that come from thinking about life processes in terms of developmental systems that produce, sustain, and change living beings over both developmental and evolutionary time.
Providing a genuine alternative to genetic and environmental determinism, as well as to unsuccessful compromises with which others have tried to replace them, Evolution's Eye will fascinate students and scholars who work in the fields of evolution, psychology, human biology, and philosophy of science. Feminists and others who seek a more complex view of human nature will find her work especially congenial.

Author Notes

Susan Oyama is Professor of Psychology at the John Jay College of Criminal Justice and the Graduate School and University Center, City University of New York. Her book The Ontogeny of Information is also published by Duke University Press.

Reviews 2

Publisher's Weekly Review

Evolution's everywhere these days, and some of its most strenuous public explainers like to make claims about genes and human nature: often they say they can show how the first shape the second. Oyama (The Ontogeny of Information) wants to complicate that picture. Her subtle and sometimes abstruse study of recent concepts in biology and social science--concepts like "evolution," "development," "phenotype," "construction" and "competition"--aims to displace models of selfish genes with models of competing and interacting processes: these processes, working at every level, can improve our explanations of how populations and (especially) people grow, differ and change. Oyama's developmental systems theory draws on the newish field of "science studies" (in which philosophers and sociologists look at the assumptions and logic of scientific disciplines), on biologists' critiques of their field (among them Richard Lewontin and Evelyn Fox Keller) and on bits of literary theory. A professor of psychology at New York's John Jay College of Criminal Justice and at the CUNY Graduate Center, Oyama writes for a highbrow audience, though one spread across many disciplines. Her prose can sound too academic or drably general: she hopes, for example, "to adopt a thoroughgoing interactive constructivism with respect to both developmental and evolutionary processes." What she means is that she wants to think--and to get us to think--about how culture, environment and genetic programming are constantly "talking to" one another, and how it's their interaction that creates us. It's a worthy goal, and one her book should advance. Illus. (May) (c) Copyright PWxyz, LLC. All rights reserved

Choice Review

Oyama (John Jay College, CUNY) presents and argues for a new resolution to the nature/nurture debate. Although scholars commonly accept the notion that behavior is not the product of one factor or the other, but rather an interaction of both, Oyama contends that the trend has been to determine how much of a trait can be attributed to either genetics or experience. She advocates a systems approach; using numerous examples, she illustrates that although researchers and theorists claim to have accepted that behavioral and physical phenotypes are reflective of an interaction, this is not the case. Oyama argues elegantly and provides myriad examples to illustrate the lack of an appreciation for the reciprocal relationship that exists between genetics and the environment. She does cause one to consider one's beliefs and to re-evaluate how one defines and discusses inheritance, but undergraduates will not need an entire book on the subject. Recommended for graduate students, researchers, and faculty seeking a fresh approach to this issue. M. Pilati; Rio Hondo College



Chapter One Transmission and Construction: Levels and the Problem of Heredity The conventional view of evolution involves two mistaken ideas. One, the idea that traits are "transmitted" in heredity, rests on notions of genetic programming that are ultimately quite preformationist, A second idea, what I call "developmental dualism," holds that there are two kinds of developmental processes, one controlled primarily from the inside and another more open to external forces; it both supports and is supported by the notion of trait transmission. The theory of evolution seems, in fact, to require a distinction between features that develop "under the aegis of the genes" (Konner 1982:157) and those that are shaped by the environment--hence the apparent need for dual developmental processes.     The concepts of trait transmission and developmental duality are linked by a way of thinking about the role of genes and environment in ontogeny that ensures that we will continue to find ways to carve up the living world into innate and acquired portions, no matter how vociferously we declare the distinction to be obsolete. This in turn ensures not only a degree of conceptual incoherence in our science, but also continued difficulty in synthesizing our knowledge of development with our understanding of evolution. Finally, our lack of clarity on these issues, because of their deep involvement in the old and tangled nature-nurture complex, encourages further confusion when scholars make pronouncements on the role of evolution in shaping our fundamental nature (Konner 1982; Midgley 1980; E. O. Wilson 1978). I will argue that evolution only seems to require these two ideas. In their place I offer an alternative way of looking at development and the succession of life cycles we call evolution. This alternative way requires no distinction between genetically transmitted traits and traits that develop in some other manner; the distinction becomes unnecessary when we relinquish the conviction that traits can be transmitted at all. Nor does it require us to divide psychological characteristics into those that are "biologically based" and those that are due to culture or "phenotypic plasticity" (as in Fishbein 1976:40). What it does require is a conception of development as construction, not as printout of a preexisting code.     Even though the distinction between the innate and the acquired has been under attack for decades (Anastasi 1958; P. Bateson 1983; Beach 1955; Hinde 19698; Klopfer 1969; Lehrman 1953, 1970; Oyama 1979, 1982; Schneirla 1966), and even though it is routinely dismissed and ridiculed in the scientific literature (Alland 1973:14-15; Konner 1982:80-89), it continues to appear in new guises. The very people who pronounce it obsolete manage, in the next breath, to distinguish between a character that is a "genetic property" and one that is only "an environmentally produced analogue" (Alland 1973:18, on phenocopies, discussed in chapter 2), or between "genetically determined fixed action patterns" and patterns in which "innate factors" play but a minor role (Konner 1982: 20-22, 186). Vocabulary and styles of description shift, but the conviction remains that some developmental courses are more controlled by the genes than others.     The reasons for this are many; some have to do with our philosophical traditions, and some with our attachment to certain analytical techniques. Intellectual inertia and the politics and sociology of academic life also help perpetuate dichotomous ways of describing development. In seeking to make our ideas intelligible to students and colleagues, we express them in familiar terms, thus reinforcing old conceptual structures. (Even when we seek to present new structures, we are often heard in terms of old ones.) And in justifying our efforts to others, we may point to our work's relevance for accepted paradigms and problems, thus further legitimating traditional modes of thought. Since dichotomies are unpopular these days, having been largely replaced by degrees of biological constraint, open and closed programs (Lorenz 1977; Mayr 1976a), and the like, it is important to realize that such fuzzing of distinctions does not alter the conceptual framework (see chapter 3). It allows people to dodge charges of simplistic determinism without having to change their ways of thinking.     Another potent factor hindering our attempts to transcend the opposition (or even interaction) between nature and nurture is the resurgence of interest in evolutionary questions, among biologists and others. This could have been, and to a limited extent has been, an impetus to serious conceptual reformulation. All too often, though, it has led to the resurrection of largely unreconstructed notions of nature, for reasons both scientific and nonscientific. Our theories are not isolated from what transpires outside the academy, and we must confront the interplay between scholars' ways of thinking and the concerns of the larger society.     Part of the attention attracted by evolutionary studies is prompted by the explosion of theory and research in the field itself. Beginning in the 1970s, we have witnessed several decades of impressive discoveries in molecular biology. Another aspect of the rush to adopt an "evolutionary perspective" is probably a pendulum swing back from the environmental determinism that held sway in the social sciences for many years. In addition, however, there appears to be a more general wish, not limited to academics, for a substantive "biological" bedrock. There seems, finally, to be a hope that an immutable natural "core" will stabilize an increasingly relativistic and uncertain world. Hence the attractiveness of the "ultimate" explanations offered by evolutionary theorists.     These reasons, however, are neither independent nor sufficient to explain the burgeoning interest in evolutionary issues. For now, I want only to point out that they tend to rest on, and ultimately to preserve, largely unanalyzed conceptions of "biological bases," not the least of which is the assumption that "biology" gives us a set of (largely) universal, unlearned, unchangeable, and inevitable traits which were formed by natural selection and which define our fundamental nature. (Context-dependent, facultative variation introduces a wrinkle into this formulation, but it is not a wrinkle that challenges most people's ideas about developmental dualism; see my comment on quadrifurcation below.) It is just this complex of ideas about biology and evolution that needs to be untangled. Although that task requires much more space than is available here, the notion of trait transmission via genetically directed development is certainly an important part of the complex. Natural Selection and Genetically Programmed Traits Natural selection, we are told, cannot work on acquired characters, for these are not passed on to offspring; only inherited traits are transmitted. The appearance and differential proliferation of novel genes or gene combinations, coding for new aspects of the phenotype, is seen to be the essence of evolution: Evolutionary processes are said to produce genetic programs for development if one is speaking of the individual organism, or changes in gene pools if the focus is on the population level. Thus, no matter how much one protests that all development requires both an adequate genome and an adequate environment and that all forms are jointly determined by genotype and developmental context, and no matter how often one points out that genetically determined variance is not the same as genetically guided development, the form of certain traits continues somehow to be placed in the genes. A paper by Scarr and McCartney (1983) provides a striking example of this. Writing in the service of an allegedly evolutionary perspective, the authors dismiss the opposition between genes and environment and emphasize the inability of analyses of variance to explain the combination of genotype and environment in development. They are quite happy, however, to speak of genetic programs controlling maturation, to contrast genetic and environmental transmission of characters, and, most emphatically, to speak of the genotype as driving development (p. 428). (They do all this, ironically, on the basis of analyses of variance.)     What has emerged from the confrontation between evolutionary and developmental concerns is a kind of credo that informs our thinking even when it remains implicit. Sydney Brenner (cited in Lewin 1984:1327), who some years ago embarked on the ultimate molecular analysis of a developmental program, offers an explicit statement: "The total explanation of all organisms resides within them, and you feel there has to be a grammar in it somewhere. Ultimately, the organism must be explicable in terms of its genes, simply because evolution has come about through alterations in DNA." In exploring the relationship between embryology and evolution, R. A. Raff and T. C. Kaufman (1983:234) similarly take as their "basic tenet" the assumption that the "genes control ontogeny." This is the case, they say, because evolution consists of genetic changes. Having accepted the definition of evolution as evolution of genes, these developmentalists have no choice but to conceptualize development of species characters as explained by, caused by, and guided by those genes.     Brenner's case is especially instructive. He set out to describe in complete detail the development of Caenorhabditis elegans , a remarkable worm that seems to fit the metaphor of a preset computer program mechanically producing an inevitable result. This worm is quite unusual in the precision of its ontogeny; it always has 959 cells, and the history of each of those cells can be mapped with great confidence. In spite of the prodigious knowledge Brenner and his colleagues have acquired about the ontogeny of this creature, however, or perhaps because of it, he now disavows the notion of a program, even warning against loose metaphorical use of the term.     My own inclination is to applaud this repudiation of his original aim. Brenner looked development in the face and reported what he saw: A multitude of events influence each other, set the stage for each other, and run off in improbable sequences, but there was no "genetic program." Cell lineages are unpredictable and complex; by the time one has given a "rule" for such a lineage, one has virtually described the sequence, step by step. There seems to be no logic. The paper by Lewin from which I take this material is entitled "Why Is Development So Illogical?" We cannot, unfortunately, explore the interesting question of what would have constituted a program--presumably a "logical" description--for Brenner's group. After all, we can describe many biological processes in computer language, and we can produce marvelous simulations. But such simulations must provide for much, much more than DNA sequences and their products. More to the point, a program amounts to a description, and we can write these descriptions for all sorts of things, including "acquired" characters. These latter programs may include information on DNA sequences as well. How, then, can the concept of a program help us distinguish between the innate and the acquired?     In any case, Brenner has not lost faith in the power of reductive analysis. In an intriguing (and, to this observer, frustrating) coda, he asserts that he is now convinced that the unit of development is the cell. Because one can presumably understand how the "genes get hold of the cell," this retreat up one level evidently allows him to continue to see development as explicable in terms of genes, even though he pushed this premise to its limit and found it lacking. Having given up genetic programs, he now speaks of internal representations and descriptions. In doing so he is like many other workers who have been faced with the contradictions and inadequacies of traditional notions of genetic forms and have tried to resolve them, not by seriously altering their concepts, but by making the forms in the genome more abstract: not noses in the genes, but instructions for noses, or potential for noses, or symbolic descriptions of them. This solves nothing. At most, it communicates a certain discomfort with the preformationism implicit in traditional formulations, but it does not alleviate the problem; it just makes the wording vaguer. Developmental Dualism and the Proliferation of Levels Since so many scientists accept the idea that the body is constructed by genetically guided processes, it is not surprising that they look for parallels to the development of body structure when they turn to psychological and social levels. Some of our mental, even communal, life, they reason, is phylogenetically derived, just as our bodies are, and is therefore attributable to the same kinds of genetically guided development. But insofar as cruder forms of nature-nurture distinctions imply an invariant biological base with environmentally wrought variations, they are incapable of supporting this complication. A consequence is that motivation, thought, perception, and behavior (and eventually culture) are divided into pseudolevels. And so scientists fret about how to identify the inherited aspects of the various human temperaments (Plomin and Rowe 1978; by "inherited," by the way, they do not simply mean "showing heritable variation in a particular population"), how to distinguish biological from cultural universals (Harris 1983), and so on. Both variable and universal features, that is, are somehow divided into those that are (mostly) formed from within and those that are (mostly) sculpted from without. The dichotomous has become quadrifurcate.     But what could it mean to say that behavior of any kind is "phylogenetically derived"? We are told that the behavior is "programmed" or in some other way "in the genes." If we are unhappy with this description, we are hastily assured that no trait, morphological or behavioral, is actually in the genes, but rather that information or rules are encoded in the DNA. Our imaginary interlocutor (imaginary but, I suspect, easily recognized) also assures us that evolved behavior may well be the result of "interaction" between nature and nurture, a mixture of biology and learning, and that nothing is due entirely to the genes or the environment. And while body structure is "directly" encoded, furthermore, gene-behavior links are "indirect." Any uneasiness we may have about genetic determinism is supposedly laid to rest as our friend assures us that usually the genes do not implacably command, but rather predispose.     If we are uncommonly mistrustful or skeptical, however, we are not mollified. Our doubts are fed by our friend's statements that such evolutionarily shaped behavior constitutes the very core of our being (if we are the species in question); that it is normal, if not desirable; and that it is altered or suppressed only at considerable psychic and social cost (Lumsden and Wilson 1981:358-360; Midgley 1980:75-76, 186). Believing that a particular kind of social behavior has evolved, the friend implies, does not necessarily entail its endorsement, only the sober admission that it is in our nature and is therefore a force to be contended with, to be accepted or ameliorated, even to be vigorously combated within the perhaps narrow limits allowed by the genes, but surely not to be underestimated (Konner 1982:206).     The use of terms like genetically transmitted for such features, however, is profoundly misleading. It is just what leads people to proliferate levels within levels as they search for more acceptable ways to express their belief that some aspects of mind, like most aspects of body, are somehow in the chromosomes and can thus be passed on with them. For differential reproduction to alter a gene pool, however, all that is needed is reliable genotype-phenotype correlations; and these, in turn, require not genetic "programs" for development but a reliable succession of organism-environment complexes--of developmental systems that repeatedly reconstitute themselves. Trait Transmission and Trait Construction Viktor Hamburger (1980) asserts that developmental studies have not been fully included in the grand neo-Darwinian synthesis. He gives several reasons, one being evolutionists' preoccupation with trait transmission across generations, a preoccupation that is at odds with developmentalists' interest in the elaboration of traits during the course of each life cycle. Developmentalists, he also says, tend to dislike the preformationist implications of transmission genetics. Raff and Kaufman (1983:20-22) describe the estrangement that developed between embryologists and geneticists early in the twentieth century and, like Hamburger, allude to conceptual differences that have hindered the synthesis of the two traditions. The dilemma is how to have an evolutionary perspective without being trapped with some untenable notion of genetically directed development and, by implication, an equally untenable notion of environmentally directed development to complement it. The reason circumlocutions fail to extricate workers from this dilemma is that they are just that, circumlocutions: indirect ways of saying that at least some of ontogeny is explicable in terms of macromolecular plans. No matter how much their own observations contradict this assumption, these scientists are committed to it because they do not question the definition of evolution as evolution of genes.     But the conception of evolution as change in gene pools or genotypes is an exceedingly abstract one--and, more serious, an incomplete one. The role of gene frequencies in reflecting some changes in developmental trajectories is undisputed. The importance of genes in development and as sources of phenotypic variation is also undisputed. What is misguided is the attempt to place phenotypic form, or any kind of "representation" of that form, in those genes. It amounts to an attempt to explain a dynamic multilevel phenomenon by a lower-level entity. (For some discussion of levels in evolution, see Sober 1984.) It is not only reductive, it is incomplete even at the molecular level because it ignores the rest of the interactive complex necessary for gene function: precursors, organelles, and other aspects of cell structure and chemistry. Given that many believe the primary problem of developmental biology is the differential activation of genes at different times and in different tissues, and given that it is the rest of the interactive complex that must in large part account for differential activation , this partial account is hard to justify.     To focus on gene pools or modal genotypes as the essence of evolution is to lose sight of the organism, of the successive life cycles whose distribution and reproductive fortunes are reflected in those gene frequencies. As a gene pool changes, however, so do many of the levels of environment needed to create and sustain those life cycles, and they must be included in any account of ontogeny and phylogeny. They are implicated in even the most basic and reliable of species forms. The environment, that is, must be seen not only as a source of phenotypic variance, but as fundamental to species character, and not only as passive support, but as equal partner with the genes in giving rise to living beings. Vital patterns are the result of interactive systems at many levels.     What is transmitted between generations is not traits, or blueprints or symbolic representations of traits, but developmental means (or resources , or interactants ). These means include genes, the cellular machinery necessary for their functioning, and the larger developmental context, which may include a maternal reproductive system, parental care, or other interaction with conspecifics, as well as relations with other aspects of the animate and inanimate worlds. This context, which is actually a system of partially nested contexts, changes with time, partly as a result of the developmental processes themselves. Differential gene transcription in diverse tissues is a result of this emerging system, as are interactions at the organ and organism levels. Developmental means are transmitted in the sense of being made available during reproduction and ontogeny. Often they are the very products of ontogeny, but they are no less crucial to further development for not having been present at conception.     Bodies and minds are constructed, not transmitted. Even what is generally described as culturally transmitted is not simply transferred in a lump from one being to another. Organisms, including people, often influence the development of other organisms; these influences, however, are interpreted and used by their recipient in ways that are complex, ill understood, and very much a function of the of the organism's developmental state and surrounding conditions. Developmental state, in turn, is the synthesis of past developmental events. We all know the frustration of seeing others construe our words and deeds in ways that seem mistaken, even perverse. What have we transmitted? That which was intended? That which was understood? Haven't we rather entered into interactions whose outcomes are only partially a function of our contributions? The Construction of Levels I admit that the term construction is problematic in some ways. For one, it often implies an acting subject. While this is perfectly appropriate when we are speaking of knowing subjects construing their worlds and their own actions, it is inappropriate at other levels. We find, for example, descriptions of development or behavior in which genes "foresee" contingencies, "direct" events, "recognize" stimuli, "select" or "program" outcomes. This is construction seen as revelation (Monod 1971:87), as fulfillment of a plan, and it is precisely the construction of ontogenetic "construction" that gets us into so much trouble. My meaning is quite different. Construction, in my view, does not require a subject or subject surrogate, which is often what the gene amounts to in accounts of development (Tobach's 1972 cryptanthroparion ; Weiss's 1969 anthropomorphic principle, or my own 1982 homunculoid gene ), but it does require multilevel systems constructing themselves, bringing about the conditions for their own further change. I see interactive construction, then, not as the fulfillment of a plan, and not necessarily as the activity of a knowing subject, but as a developmental phenomenon that is not explicable by only one set of its constituents or by only one of its levels. Developmental Systems Developmental systems evolve, generating one life cycle after another. One aspect of their evolution is the shifting constitution of the gene pool. But the full phenomenon is a succession of organisms, variable in some ways and constant in others, in their changing environments. In order to put these organisms back into an increasingly gene-centered conception of evolution and development, we sometimes need to step back from the gene level, however fascinating its processes, and reconstitute the rest of the systems that repeatedly make the living world. This is finally what students of development must do , however they may describe what they do. Whether or not one speaks of genes "getting hold of" cells, that is, one must finally describe not only intracellular processes but also relations among cells, as well as the ways these relations influence and are influenced by higher-level processes, including organism-environment interactions.     The benefits of seeing evolution as a succession of developmental systems are several. First, it restores the organism to us at a time when some biologists seem to be intent on analyzing it out of existence (Dawkins 1982). Second, it allows us to investigate and appreciate the role of the genes in development without turning them into wise little homunculi. Third, it affords us a way out of the multiple versions of the nature-nurture impasse while still allowing us to join our developmental concerns with our evolutionary ones. It does this not by giving us yet another way to make the distinction between genetic and environmental traits (as opposed to genetic and environmental sources of variation), but by showing how all features, species-typical or variable, morphological or behavioral, are produced by a complex of interactants, some exceedingly constant, and some variable. This frees us from the bogey of dual developmental processes and from the implicit preformationism and reductionism of all notions of preexisting plans in the genome and restores to us the living world in all its layered complexity without diminishing our ability to turn our analytical skills on any part of that world as we attempt to understand it and to make our way through it. Copyright © 2000 Duke University Press. All rights reserved.

Table of Contents

Part 1 Looking at Development and Evolution Transmission and Construction: Levels and the Problem of Heredity
What Does the Phenocopy Copy? Originals and Fakes in Biology
Ontogeny and the Central Dogma: Do We Need the Concept of Genetic Programming in Order to Have an Evolutionary Perspective?
Stasis, Development, and Heredity: Models of Stability and Change
The Accidental Chordate: Contingency in Developmental Systems
Part 2 Looking at Ourselves
Essentialism, Women, and War: Protesting Too Much, Protesting Too Little
The Conceptualization of Nature: Nature as Design
Bodies and Minds: Dualism in Evolutionary Theory
How Shall I Name Thee? The Construction of Natural Selves
Evolutionary and Developmental Formation: Politics of the Boundary