Evolutionary psychology as computational theory in the cognitive sciences.

Scarcely a decade old, evolutionary psychology (EP) has developed a high profile within the cognitive sciences. EP's niche involves what David Marr (1982) called "computational theories:" functional descriptions of what information processing devices, including brains, are designed to do. Such theories constrain and inform the search for cognitive and neural descriptions of the device. EP is characterized by primary commitments to modularity of mind, the use of evolutionary biology's adaptationist program to generate hypotheses regarding mental modules, and the use of cognitive science's methods for testing such hypotheses. It is also characterized by a number of secondary commitments and positions on important issues that are not necessitated by the evolutionary approach. These are defensible within the broader fields of human behavior and cognition, and by tertiary commitments that are logically unrelated to the EP program but nonetheless characteristic of the field (i.e., anti-theistic biases and other str ong reductionisms). EP has much to offer. Combining the conceptual methods of evolutionary biology and the empirical methods of the cognitive sciences seems genuinely promising, but it's rhetoric is strong for an immature discipline and its reductionism pits it against disciplines with which it might productively coexist. The primary (and even secondary) commitments are not antithetical to an orthodox Christian faith, but the anti-theistic and anti-religious tendencies of some vocal protagonists represent a challenge.

Evolutionary Psychology (EP) is a relatively recent development in the cognitive sciences aimed at determining, by means of the methods of evolutionary biology and the cognitive sciences, the basic psychological functions common to all humans. Although psychology has a long history of Darwinian thinking (beginning with Wundt and James), EP protagonists suggest that what they have to offer is truly revolutionary. EP promises to deliver what psychology and the related brain and cognitive sciences have so far lacked: an overarching theoretical framework, or "computational theory" (Marr, 1982) that unifies the previously isolated subdisciplines, empirical observations, and methods.

EP is a controversial field, and not just on issues of science and theology. It also draws criticism from within the ranks of the social sciences, and sometimes even from evolutionary biology itself. Nevertheless, EP has become a prominent force in the landscape of the human sciences, and it behooves us to look beyond the rhetoric and examine its fundamental claims and logic. As scholars committed to pursuing truth in the cognitive and behavioral sciences, this provides us with a knowledge base required for intelligent dialog on the issue. As Christians committed to the doctrine of a personal, intelligent creator who sustains and governs His creation, it provides us with a fresh opportunity to clarify our thinking regarding the origins of and formation of life. (1)

The goal of this article is thus twofold. On the one hand, I wish to present the EP in a way that is palatable to those within EP, that is, to explore its scientific merits without misrepresentation or slander. Of course, this must include discussing EP's shortcomings (as I see them) as well as its contributions. On the other hand, I wish to explore EP's merits from within an orthodox Christian worldview, to ask whether it is even possible for Christians to embrace an evolutionary approach to the scientific questions surrounding cognition and behavior.

Before addressing these two goals directly, it is worth considering EP in light of phrenology, another controversial episode in the study of the mind. EP shares some interesting features with its ill-fated cousin, but distinguishes itself importantly by the interesting empirical research programs it has generated. In the first section of the article, I will compare EP and phrenology and review two such programs. In the second section, I will consider EP's understanding of itself in the larger web of disciplines engaged in the cognitive, neural, and behavioral sciences. In the final section, I will examine eight specific attributes and issues that define EP's current niche, including allegations of "Darwinian fundamentalism" and atheism.

PHRENOLOGY, MODULARITY, AND THE NEW FACULTY PSYCHOLOGY

Psychology's interest in decomposing the mind into its elementary components - one of EP's primary goals - predates its interest in evolution. Fifty years prior to the publication of Darwin's On the Origin of Species, Franz Josef Gall proposed a theory relating psychological function to brain structure that came to be known as phrenology. (2) Gall's theory assumed what we might today call "the modularity of mind:" the mind is composed of a variety of specific faculties, each of which operates more or less independently of the others. At least two things separated Gall's theory from preceding ones. Gall assumed that psychological functions were localized to circumscribed regions of the cerebral cortex, and that the physical size of these "mental organs" varied in proportion to use. He argued that one could correlate the protrusions of the skull, which themselves correlated with the size of the underlying "organ," with psychological characteristics. As it turns out (according to our current understanding of the brain, anyway), Gall seems to have been essentially right about localization of function, and essentially wrong about how one might go about investigating it. The story of the essential correctness of Gall's theory, it's dramatic failure, and the way in which its failure impeded the progress of the brain and cognitive sciences over the ensuing 100 or more years holds important lessons for us as we consider other new subdisciplines, including evolutionary psychology, the subject of this article.

Gall's program failed for at least three reasons: It rested on a faulty assumption about what sorts of psychological functions were localized, it employed an inadequate empirical methodology, and it was prematurely packaged and sold to those outside the field. The faulty assumption centered on the nature of the complexity of localized mental functions. Whereas phrenology proposed complex functions that appealed to the intuitions of Victorian Europeans, such as amativeness, friendship, hope, and suavity, the current view holds that discrete regions of brain tissue are responsible for more elementary functions, such as independent visual analysis of color, form, and motion, the control of muscle groups, and auditory sensation (e.g., see Kandel, Schwartz, & Jessel, 1991). The problems inherent in this assumption were compounded by an inadequate method. Not only are the bumps on the skull uncorrelated with the size of the underlying cortical regions, but, within a species, the size of the cortical regions seems t o have little to do with the quality or quantity of psychological attributes. The method simply left too much to the imagination of the investigator, and invited self-confirming hypotheses. Neither of these two problems would have in themselves proved fatal to Gall's program. More adequate methods were only a few decades away in the fields of physiology and clinical neurology, and such methods would likely have reigned in careless theorizing. What precipitated the failure of phrenology was the way in which it was brought to the public. Phrenology was, by any standards, an immature science-worthy of cautious optimism and empirical investigation, but not of strong rhetoric and confident application to clinical and other practical problems. Its adoption as the pop psychology of the day, the confidence of its primary protagonists, and the proliferation of quack practitioners ignited opposition and led to the debunking efforts of a number of serious scientists, such as Pierre Flourens, and to a backlash against fu nctional localization known as globalism (i.e., the idea that the entire brain contributes equally to any psychological function). Even as late as the 1950's, Karl Lashley (1929) was promoting an antilocalizationist "equipotentiality" position. The localizationist position was a sheep in wolves' clothing, and one wonders how much more quickly the neurosciences might have developed had the shepherds been able to chase after real foes.

Modularity and Adaptationism:

A New Faculty Psychology

Gall's program is an apt and convenient starting point for evaluating EP for a number of reasons. First, like Gall's notion that basic psychological faculties are localized to cortical areas, the most fundamental premises of EP seem reasonable and permissible within the bounds of contemporary science and Christian orthodoxy? (3) It seems reasonable, for example, to assume that humans have evolved from non-human ancestors according to mechanisms such as natural selection, and that the structure of the human brain was designed by means of such processes. (I am not arguing here that contrary views on evolution are unreasonable. I simply take as a starting point that, given current thought in theology and biology, theistic evolution is a viable alternative and has no greater theological and exegetical hurdles to clear than other accounts of creation. Of course, if you are not willing to grant this, then you probably have no reason to think seriously about EP to begin with.)

A second reason that warrants a comparison between EP and phrenology is the common focus on discovering the basic functional elements, or modules, of the mind (i.e., reverse engineering in EP terminology). Such faculty psychology has traditionally been an important part of the cognitive sciences. Cognitive psychologists have established a variety of methods for investigating putative modules, but the cognitive sciences have generally lacked a theoretical guide for hypothesizing specific modules to begin with. EP offers evolutionary biology's "adaptationist program" as a remedy for this shortcoming: Basic psychological functions correspond-to discrete brain systems that were adaptively useful for our Pleistocene ancestors. In this, EP differs from phrenology, which like much of modern cognitive science proposed specific modules rather arbitrarily. Nevertheless, EP seems similar to phrenology in another way. Whereas the modern cognitive and neural sciences have moved toward a relatively small number of simple i nformation processing functions as candidates for the basic modules of the mind and brain, EP proposes a large number of rather complex ones (e.g., modules for cheater detection, person-specific memory, child-care motivation, and sexual jealousy; Tooby & Cosmides, 2000).

Methods and Three Examples of Research in EP

A third comparison between EP and phrenology is that it proposes particular methods, and these can be evaluated with respect to their utility in achieving the goals of the program. Whereas phrenology proposed new and untried methods, however, EP borrows methods from two extant disciplines. From evolutionary biology, it borrows. the method of "adaptationism," based on Darwin's notion of natural selection, and from cognitive psychology it borrows a collection of methods for assessing performance on information processing tasks. If having the mental machinery to solve a particular information processing task would have been adaptive for our ancestors (i.e., if the hypothesized phenotype fits the formal definition of an adaptation; see Table 1), and if further (empirical) tests reveal that performance is enhanced in modern humans when information is presented in a format that would have been likely in the environment of evolutionary adaptedness (EEA; that is, the ancestral environment), then the mental machinery responsible for solving the task can be considered to be a bona fide mental module.

To illustrate this combination of methods, consider two examples from the EP literature. In both cases, the unique perspective and strategy of EP produces empirical results that move our understanding of psychology forward and challenge conventional psychological wisdom.

In the first example (reviewed by Cosmides & Tooby, 1992, 2000), Cosmides and colleagues suggest that the human cognitive architecture contains a mental module specialized for solving a small set of social information processing problems exemplified in the task of detecting cheaters.

This study is set against the backdrop of a familiar phenomenon in the human reasoning literature: When given logical rules of the sort if A then B, people often fail rather strikingly when asked to sample available data and determine whether or not the rule is true. This has been demonstrated using the Wason selection task (Wason, 1966). In the standard task, subjects are given some instance of the if A then B problem (e.g., if a person goes to Boston, then that person takes the subway). Subjects are then presented with an array of four cards, with each card showing the values A, not-A, B, and not-B (e.g., Boston, Maryland, subway, and cab, respectively). Each card represents a potential instance of the rule, with the unseen side of each card depicting information that would either corroborate or falsify the rule (e.g., the card marked Boston might show cab on its reverse, falsifying the rule if Boston then subway). The subject's task is to indicate the minimal set of cards that would need to be turned over in order to determine if the rule has been violated. Logically, only the A and not-B cards must be investigated; if the A card shows not-B on its reverse, or if the not-B card shows A on it's reverse, then the rule does not hold. In the standard task (when problems are descriptive rules, as in the example above), subjects perform poorly, with only about 25% of subjects choosing the correct cards.

Cosmides (e.g., Cosmides, 1989; Cosmides & Tooby, 1989, 1992) reasoned that such abstract logical rule-testing would have been of little adaptive value in the EEA, and so general-purpose circuitry for solving such problems could not possibly have been selected during the evolution of the human species (and therefore could not be assumed to constitute one of the basic mental modules of the modern human mind). On the other hand, there are domain-specific problems that fit the formal if A then B format and at least some of these problems are what biologists call adaptive problems. Such problems are evolutionarily long-standing, they are universal across the species, and anyone owning machinery capable of solving the problems would have had a reproductive advantage (statistically speaking) over those that did not. One such domain-specific problem involves reasoning about social exchange, in which conditional rules of the form if A then B are called social contracts and have the form "If person A provides the requ ested benefit to or meets the requirement of person or group B, then B will provide the rationed benefit to A" (Cosmides & Tooby, 2000, p. 1260). Based on an evolutionary analysis, Cosmides and colleagues predicted a number of "design features" (4) that could be empirically tested in the psychological laboratory. For example, machinery designed for use in negotiating social contracts would likely be specific to the domain of social contracts, would include inference procedures specialized for detecting cheaters, would not be much good at detecting altruists, etc.

Cosmides and colleagues tested subjects' reasoning about social contracts in the Wason task. The basic result is that subjects perform very well on social contract problems in comparison to conditional rules of other types, with 65% to 80% of subjects choosing the optimal solution. Thus, Cosmides argues, the human mind is not endowed with some general-purpose conditional reasoning capacity that functions equally well (or poorly) across all content domains and can be applied to situations of social contract. Because "what counts for cheating is so content dependent that a detection mechanism with a domain-general definition of violation would not be able to solve the problem of cheater detection," (Cosmides & Tooby, 2000, p. 1261), we have instead evolved a mental module specialized for this purpose.

A second example of research using the EP strategy is reviewed by Baron-Cohen (2000; see also Leslie, 2000). The adaptive problem involves "theory of mind" (ToM), and the proposal is that the human species (and probably no other) has evolved specialized information processing machinery for ascribing specific mental states (e.g., beliefs, desires, etc.) to appropriate objects. Such a hypothesis is closely related to the idea that there is a dedicated "folk psychology" module in the brain, and comports well with many Christian notions of human uniqueness in the domain of personal relationships (e.g., Brown, 1998).

Premack and Woodruff (1978) coined the term "theory of mind," noting that the ability to ascribe mental states to agents, including one's self, is a universal human trait shared with other species, if at all, in only a very limited way. As a result of its uniqueness to and universality across the human species, the ability to make such attributions suggests that it is a candidate for an adaptation, a mental module, if it can also be shown to increase reproductive fitness by solving an adaptive problem that was prevalent in the EEA and if further testing shows design evidence (i.e., that the mental machinery used in the task is specialized for this and not other problems).

For what adaptive problem could ToM be a solution? The answer seems to come in two parts. First, early humans entered what has been called "the cognitive niche" (Pinker, 1997; Tooby & DeVote, 1987). We universally form (or, according to EP, we are innately equipped with) intuitive theories about all sorts of things, including the biological and physical worlds, which were important for survival in the EEA. The second part of the answer is that the social world was an important feature in the EEA, and a module specialized for reasoning about the social world would have improved reproductive fitness by lengthening the lives of its possessors.

Baron-Cohen (2000) and others have suggested that the evolutionarily derived idea of ToM can shed light on the cognitive neuroscience of autism. Autistic individuals, they hypothesized, show a genetically determined selective deficit in the ToM module (Baron-Cohen, Leslie, & Frith, 1985). To test the hypothesis, Baron-Cohen and colleagues have tested autistic children and adults on a variety of tasks involving reasoning about people's mental states. Autistic children, for example, fail at tasks such as the false-belief task. In this task, a child is given a scenario such as the following. Sally has a marble that she places in a basket and covers, then departs; while she is gone, Ann removes the marble from the basket and places it in the box. The child's task is then to predict where Sally will look when she returns. The majority of non-autistic 4-year-olds have no problem in passing the test (presumably taking into account Sally's desire for the marble and her [now false] belief that it is still in the baske t). Autistic children of all mental ages typically fail. For example, Baron-Cohen et al. (1985) showed that 85% of normal 4-year-olds passed the test, but only 20% of 12-year-old autistic children (mean IQ of 82) did. The effect cannot simply be attributed to the lower IQ; in the same study, 86% of 10-year-old Down syndrome children (mean IQ of 64) also passed the test. These findings are consistent with the hypothesis that autistic children have a deficit in a ToM module. According to Baron-Cohen (2000), recent data on autistic children's reasoning about physical devices suggest that they are normal (perhaps even superior) on such tasks, indicating a dissociation between an intuitive physics module and an intuitive psychology module.

As exemplified in these two research programs, EP proposes that the mind is a collection of complex, domain-specific mental modules, and that the methods of evolutionary biology and the cognitive sciences are necessary and sufficient for discovering the nature of these modules. The proof of the pudding is in the eating, and so only time will tell whether or not the hypothesis of complex, domain-specific modules proves useful or true, and if so, whether or not adaptationism paired with standard methods of the cognitive sciences is the best way of determining the mind's essential architecture. In the mean time, we can ask how EP is trying to integrate itself with other sciences of mind and brain. In spite of its often arrogant exclusionary rhetoric, EP has actually made serious efforts to see itself as only one level of explanatory analysis. To this end, EP employs a levels-of-analysis framework due to Marr (1982), which is widespread and well accepted in the cognitive sciences and neurosciences (e.g., Cosmides & Tooby, 1995, 1997, 2000). In light of this framework, EP really does seem to have something to offer.

MARR'S LEVELS OF EXPLANATION AND THE INFORMATION PROCESSING FRAMEWORK

Marr (1982) suggested that before we can claim a complete understanding of any information processing device, we must have explanations at least three different levels. Because the human brain is, inter alia, an information processing device, Marr's levels of explanation apply to it as well as non-biological devices. To the extent that the mind is what the brain does, Marr's framework provides a means of investigating the mind. Furthermore, if an information processing device turns out to be complex in the sense that it is made up of some number of simpler, more basic information processing devices, then the framework applies to each of the simpler components, as well. To understand how EP sees its relationship to other behavioral sciences, we must understand Marr's framework.

Marr's proposed levels are shown in Table 2. The lowest level of abstraction, which Marr called "hardware implementation," is the familiar territory of neurobiology and physiological psychology (for the biological sciences) and electronic engineering (for the computer sciences). Examples of explanations at this level include the selective firing of neurons in the visual cortex for retinal images of particular shapes and orientations (Hubel & Weisel, 1968) and the discovery that the lateral nucleus of the amygdala is the cite of neuronal change that accompanies fear conditioning (LeDoux, Cicchetti, Xagoraris, & Romanski, 1990). The middle level of "representation and algorithm" is the bread and butter of cognitive psychology, and is exemplified by the early work on mental rotation by Shepard (1984, 1987). Shepard and Metzler's (1971) experiments, for example, revealed a lawful relationship between the angular difference of two identical geometrical figures and the time subjects took to confirm that the figures were in fact the same. From this, they argued that the information processing machinery of the human visual system operated on analog rather than propositional data ("representations") and that the operations ("algorithms") were similar to rotations performed on real physical objects. The point is that, for information processing devices, one important level of explanation includes specifying the nature of the input representation, the transformation or algorithm that operates on that representation, and the nature of the resulting output representation. For psychologically interesting functions (e.g., color constancy, face perception, or cheater detection), such explanations are rarely possible solely by recourse to the hardware components themselves, and thus this level of explanation does not reduce the explanations to the level of hardware implementation.

Of course, these two levels of explanation are linked, but Marr suggested that the coupling between the levels is loose, and that the progression of scientific discovery proceeds in a top-down fashion. For Marr, it seemed unlikely that detailed information about hardware would shed light on the algorithms. On the other hand, Marr suggested, explanations of the algorithm and representation (garnered from a psychophysics or cognitive laboratory) constrain the search for specific hardware mechanisms. It is noteworthy that opinion has changed in the twenty years since Mart's proposal. The field of the cognitive neurosciences, for example, is based largely on the assumption that bottom-up as well as top-down strategies are useful, neurobiological investigations both influencing and being influenced by cognitive and behavioral studies (e.g., see Churchland & Sejnowski, 1992, Chapter 2). Another link between these two levels involves the idea of modularity: One way of arriving at a candidate for a module is to isola te a brain region or circuit and assess its information processing contribution. To the extent that the structures can be identified on purely structural grounds, structure aids in the discovery of function.

EP is, of course, concerned with both neurobiological and cognitive levels of explanation (more so with the latter, in practice). Marr's third level, however, holds the key to understanding the rationale behind EP. Whereas hardware implementation and algorithm/representation levels are both concerned with the question of how something gets done, the level of computational theory concerns itself with precisely what is done and why. To provide a computational theory of an information processing device, in Marr's terms, is to provide a description of the problem the device solves and a rationale for why devices for solving such problems should exist at all. Explanations at this level might be less accessible to empirical testing than those at the lower levels, but they are critically important for constraining the search for algorithm and representation. The reason for this is that the computations (i.e., algorithms performed on representations) depend more on the computational problems they were designed to sol ve than on the hardware in which they were implemented. This seems particularly so when the hardware can realize a multitude of algorithms and representations, as is the case with devices built with either transistors or neurons. As a matter of practical discovery, then, the investigation of a computational device is much more likely to arrive at descriptions of the basic algorithms and representations if it starts by constructing a computational theory than if it starts by examining its elementary physical components. We cannot understand bird flight simply by studying individual feathers. By studying aerodynamics, however, the shape of their wings and feathers make sense. In a similar manner, we cannot understand why ganglion cells have the receptive fields they do by studying their anatomy and physiology without pursuing the mathematics of differential operators (i.e., an algorithmic explanation) and the recovery of object information from an optic array (i.e., computational theory).

Such theories, Marr argued, have been few and far between in the psychological sciences. As a rare example, he cites Gibson's (1966) ecological psychology, in which he described the problem of perception as that of recovering valid properties of the external world. Gibson's problem was that he underestimated the difficulty of detecting these "physical invariants" (which, Marr argues, does require computation, contra Gibson), and so Gibson saw his ideas as being in competition with those of psychophysicists and other computationally-minded psychologists working in perception. Instead, these are descriptions at two different levels. Similarly, Chomsky's (1965) theory of transformational grammar is also a true computational theory, describing the problem of language (e.g., syntactical decomposition) rather than algorithms and representations that could or do solve the problem.

It is noteworthy that Cosmides and Tooby, who have been so influential in initiating EP, rely on Marr's framework for making sense of their fledgling field. Indeed, it forms the organizing principle of their paper on the role of EP in the cognitive neurosciences (Cosmides & Tooby, 1995). In their EP primer, (5) Cosmides and Tooby (1997) propose three complementary levels of explanation that are based directly on Marr's levels (Figure 1). At the highest level of their hierarchy is the adaptive problem--if one knows what a biological information processing device was designed to do (i.e., what function a mental module served when it was selected during the environment of evolutionary adaptedness, or EEA), then one has a computational theory for the device in Marr's sense. This guides the search for the cognitive programs that solve the adaptive problems, and these in turn guide the search for the physiological basis. As a result of the empirical methods for investigating these lower two levels (collectively ref erred to as proximate-level explanations in evolutionary biology, according to Cosmides & Tooby, 1995) are already well established in the cognitive and neural sciences, it is at the topmost level that EP offers it services (i.e., evolutionary biology's ultimate-level explanations). As I understand the argument of Intelligent Design (e.g., Carlson, 2000), theology would properly replace adaptionism as the upper level of this hierarchy. Knowledge of God's nature and purposes constrains and informs our search for more proximal explanations in nature. For the evolutionary creationist, however, there is no reason to believe that theology must replace adaptive explanations. Rather, theological explanations complement adaptive ones, perhaps as a yet higher (truly ultimate) level.

EP involves exactly the sort of thing that Marr meant when he described the idea of a computational theory. Given the assumption that natural selection played a primary role in the evolution of the brain, it seems to be a very reasonable idea. Actually, according to the practitioners of EP, given that assumption, saying that one needs to know the adaptive problem is synonymous with saying that one needs to know the computational theory. Claims of sufficiency are quite different things than claims of necessity or utility. The latter are entailed by EP, but the former is not. Nevertheless, EP offers a method for constructing computational theories. Given some notion of what humans do (e.g., determining whether a conditional statement is true or not), we should investigate the particular constraints of the EEA involving problems of that type (e.g., negotiating social contracts), bring the problem into the cognitive psychology laboratory, and compare EEA-like versions with modern versions. If performance is super ior in the EEA-like version, and if a module for solving such an EEA-like version of the problem otherwise fits the definition of an adaptation (see Table 1), then add the module to the list of basic mental modules that make up the human mind. This, say those in EP, provides exactly what cognitive scientists and psychologists in other areas of basic research need: "theoretical guidance that is guided by something other than intuition" (Cosmides & Tooby, 2000, p. 1259).

EP's NICHE

If EP can offer a new theoretical framework for cognitive and neural sciences, and if that framework reasonably promises to guide the search for the fundamental functions of the mind, then it seems only fair to let it have its turn in the marketplace of ideas. To the extent that the possibility of theistic evolution is an open question, there is no reason in principle that Christians could not pursue (or at least cohabitate civilly) with EP. However, EP has itself evolved into more than its fundamental premise that careful evolutionary thought can guide psychological investigations, and it often seems to have goals that are more far-reaching than simply aiding in the search for fundamental functional units of behavior. (6) What is EP's niche in the complex of disciplines addressing human behavior and cognition? What are its unique or defining aspects? And, most importantly, which of these constitute central features, and which merely reflect the unrelated biases of the most vocal protagonists?

Adaptationism

Perhaps most importantly, EP relies heavily--some say exclusively--on Darwin's mechanism of natural selection. Because natural selection and its relationship to the idea of adaptation are central concepts in EP, and because they are often misunderstood, I will briefly describe them here before considering the claim of exclusivity.

Cosmides and Tooby (2000) define adaptions as "systems of properties crafted by natural selection to solve specific problems posed by the regularities of the physical, chemical, developmental, ecological, demographic, social, and informational environments encountered by ancestral populations during the course of a species' or population's evolution" (p. 1172). Moreover, as Cosmides and Tooby continue, one can recognize adaptations by "evidence of special design" (Williams, 1966) and by their complexity: They are mechanisms that are so well-engineered for the tasks they accomplish that the probability of their occurrence by chance can be ruled out.

The essence, if not the details, of Darwin's theory of natural selection is familiar territory for most. For sexually reproducing creatures, it works like this: (a) Within a species there is genetic and phenotypic variation (e.g., arising from random genetic mutation); (b) Phenotypes are characterized by various degrees of reproductive fitness. Depending on the match between phenotype and environment, individuals will have more or fewer surviving offspring; (c) As a result of individuals with higher fitness levels producing more offspring, genes responsible for those phenotypes become relatively more numerous in the population. Those phenotypic features that have genetic basis and increase reproductive fitness are called adaptations.

Few evolutionary biologists, it would appear, doubt the primacy of natural selection as a mechanism for constructing complex, well adapted organisms, but this is not to say that evolutionary biology has not changed in the nearly 150 years since the publication of Darwin's theory. Eldridge and Gould (1972), for example, have argued that the fossil record does not show the sort of gradual changes (i.e., phyletic gradualism) required by a strict interpretation of Darwin's theory, and have proposed ancillary processes that allow for long periods of stasis interrupted by geologically quick changes (i.e., punctuated equilibrium). Kauffman (1993) suggests that new developments in non-linear dynamics and "complexity theory" provide a basis for understanding how processes of self organization might account for complex, adaptive structures. In both of these cases (and others), it is still assumed that selection operates to weed out or fine tune nature's experiments. In light of these hypotheses, one might expect a psyc hology interested in evolutionary origins of behavior to adopt a plurality of evolutionary principles and methods.

Some critics of EP, most notably Gould and Eldridge, have argued that such plurality is conspicuously lacking; Gould suggests that EP's sole reliance on the adaptationist program makes it a sort of Darwinian fundamentalism, following patriarchs such as Richard Dawkins and John Maynard Smith (e.g., see Gould, 1997). This argument is usually articulated in such a way as to imply that EP needs to be more inclusive of other methods and principles in evolutionary biology. EP, of course, does employ a wide variety of methods at lower levels of Mart's hierarchy. The issue here is how one might fruitfully come up with "computational theories. At times it would appear that, according to El', the only acceptable way to construct such a theory is the direct and explicit application of the adaptionist program.

Those in EP have objected to the claim that they are fundamentalists in this way (e.g., see Tooby & Cosmides', 1997, response to Gould in the New York Review of Books), but it is hard to take their objections at face value in light of statements in the EP literature such as, "[t]he evolutionary process has two components: chance and natural selection. Natural selection is the only component of the evolutionary process that can introduce complex functional organization into a species' phenotype" (Cosmides & Tooby, 1997). Additional statements include: "[n]atural selection is the only explanation we have of how life can evolve" (Pinker, 1997, p. 155), and "[b]ecause there are no alternatives, we would almost have to accept natural selection as an explanation for life on this planet even if there were no evidence for it" (Pinker, 1997, p. 162).

Gould's fundamentalist label may be appropriate for other reasons, as well. It seems odd, for example, that EP should see itself as revolutionary, excluding psychologists working within a more general evolutionary framework from the ranks of the "true believers" (see also Rose, 2000, for a critique of EP from a social sciences perspective with additional comparisons to fundamentalist movements). Consider the statement of Cosmides and Tooby (1995) that the "cognitive sciences have been conducted as if Darwin never lived" (p. 1199). Work by people such as LeDoux (1996), Damasio (1994), and Ramachandran and Blakeslee (1998), for example, is clearly influenced by the very evolutionary concepts EP is proposing, even if natural selection does not provide the single unifying theme. Similarly, the history of psychology and the neurosciences is full of reference to Darwin's natural selection. Perhaps more careful attention to selection pressures and the like would serve these disciplines even better, but saying this i s a rather different thing than saying such notions have been entirely misrepresented or neglected.

Cosmides and Tooby (e.g., 1992) have suggested that almost all of the preceding work in psychology and the related disciplines has been built on what they call the Standard Social Science Model (SSSM), rhetoric that has been almost universally adopted by those in the EP ranks. The SSSM, they say, is characterized by ad hoc and intuition-based explanations, extreme empiricism (e.g., a tabula rasa view of the mind), reliance on a monolithic idea of mind and general-purpose mental faculties, adoption of a false nature-nurture dichotomy, and a deep commitment to the idea that the "contents of human minds are primarily (or entirely) free social constructs, and the social sciences are autonomous and disconnected from any evolutionary or psychological foundation" (Cosmides & Tooby, 1997). One of the themes of various papers in H. Rose and S. Rose's (2000) edited volume against EP is that the SSSM constructed by Tooby and Cosmides is a straw enemy.

In investigating social exchange and conditional reasoning, Cosmides and Tooby (2000) suggest that "the discipline of asking what adaptive information-processing problems our minds evolved to solve changes one's scientific intuitions" (p. 1261). Given an evolutionary starting point (in addition to whatever other starting points or worldviews one might hold, including evangelical Christianity), adding such a strategy to one's collection of "intuition pumps" seems immanently reasonable; adding such a strategy to the sciences of human behavior may even be necessary. It seems hard at this point, however, to imagine that the adaptationist program is necessary for all practitioners in the social and behavioral sciences, and even harder to imagine that it is sufficient for the discipline as a whole.

Modularity

As stated earlier, EP's claim that mind-brain modules are complex stands in contrast to current opinion in the neurosciences. Kandel et al. (1991, one of the standard graduate textbooks in the neurosciences), for example, say that the "functions localized to discrete regions of the brain are not complex faculties of mind, but elementary operations" (p. 15). One would expect, on this view, that functions as complex as detecting cheaters or forming theories of the mind would employ a large number of elementary operations, and would not be elementary themselves.

Although EP assumes modularity, it is not necessarily a theory of localization in the traditional sense. Whereas early localizationist theories proposed that mental functions lived in circumscribed brain regions (often in the cerebral cortex), the received view in the contemporary neurosciences is that the control of most complex functions is carried out by rather elaborate circuits, which comprise more elementary localized modules. In this case, though, the elementary components seem to be shared by a variety of complex tasks, and are therefore general purpose rather than domain specific. In such neural circuits, processing for any interesting task is distributed across a number of structures, but not in the indivisible sense of the aggregate field or equipotentiality views.

If complex functions are controlled by distributed neural circuits with more elementary components, and if the elements are shared by circuits involved in a number of domains, then it would be hard to justify EP's idea that nature selects complex, domain-specific modules. The point here is not that the complex-module view is wrong and that the distributed-elementary-components view is right, or even that the two views cannot be reconciled. The point is that this is still an open question.

Empirical Methods

Adaptationism in EP also seems to be haunted by potential methodological problems: If adaptionism is both the means of suggesting hypotheses about what counts as a module and the primary means of testing those hypotheses, EP is a circular argument. Adaptationism becomes the principle method not only for Marr's computational theory level of explanation, but also for the algorithmic level as well, and a mental module hypothesized on the basis of selective pressures in the EEA becomes a module confirmed on the basis of selective pressures in the EEA. Gould (2000), for example, alleges that the task of EP has become "a speculative search for reasons why a behavior that harms us now must once have originated for adaptive purposes," and that answers end up resembling Rudyard Kipling's "Just-So stories" (p. 119).

In principle, if not always in practice, EP survives these claims for at least two important reasons. First, recall that EP requires the empirical methods of cognitive psychology and the neurosciences. To the extent that these methods can confirm or falsify the existence of hypothesized modules, EP breaks the circularity of the adaptationist tautology. The second reason involves EP's claim of universality: psychological functions that are worth studying (i.e., those that evolved, via natural selection, as solutions to adaptive problems in the EEA) are the ones that are shared by all human beings. The reason for this is that all adaptations-including mental modules-are genetically specified, and in order for sexual reproduction to work, there must be a species-standard set of genes that allows for recombination. Universality so specified is a testable proposition for any particular hypothesized adaptation. If there are populations in which the module seems to be radically different or missing, then it cannot b e an adaptation.

Given a psychological function that meets these definitions of an adaptation, methods of the extant cognitive and neural sciences can be employed to further test the claim of modularity. Such tests involve, in adaptationist terminology, the search for evidence of special design (Williams, 1966; see Tooby and Cosmides, 2000, for discussion of design evidence specifically relating to EP). Natural selection produces species-typical structures that function, by definition, in ways that solve survival problems. Moreover, as Tooby and Cosmides (2000) maintain, "they are so well organized and such good engineering solutions to adaptive problems that a chance coordination between problem and solution is effectively ruled out as a counter-hypothesis" (p. 1172). Given a candidate adaptive problem and putative solution (i.e., for EP, a hypothesized mental module), standards for recognizing special design include economy, efficiency, complexity, precision, specialization, and reliability (Tooby & Cosmides, 2000). These m ay be tested in the cognitive psychology laboratory using standard methods, including analysis of error rates, reaction times, interference in dual-task designs, transfer and generalization, etc. Other useful analyses within the cognitive sciences include abstract task analysis via computational or mathematical models. At present, psychological and computational methods seem to be relatively weak tests for modularity as defined in the EP literature, but the convergence of multiple weak methods and analysis of plausible origins in a well-documented EEA seems promising for the EP program.

Neurobiological and neuropsychological methods can also be employed to test hypotheses derived using the adaptationist heuristic. Such methods are stronger than strictly using cognitive/behavioral tests, but modules so defined are not necessarily what EP is after. Traditional neuropsychological methods (e.g., lesion analysis, electrical recording, and stimulation) rely on strict localizationist assumptions, and modern neuropsychological techniques (e.g., functional imaging, such as SPECT and fMRI) are only now developing spatial and temporal resolution that would allow testing of distributed circuit hypotheses. Neurobiological methods typically involve smaller units of structural analysis than would be appropriate for the complex modules proposed by EP. Unless the proposed module admits of circumscribed localization or operates on a very small spatial scale (e.g., a cortical column), neuroscientific investigations can only aid as "bottom-up" methods in the search for algorithmic-level explanations, which migh t or might not turn out to be consistent with adaptationist claims at the computation. al theory level.

EP also fights an uphill methodological battle with respect to generating the adaptation hypotheses themselves. Whereas the fossil record and phylogenetic methods provide a wealth of data for many areas of biology, it is relatively impoverished regarding the EEAs for the sorts psychological functions typically addressed in EP literature. Gould (1997) states the problem well:

But how can we possibly know in derail what small bands of hunter-gatherers did in Africa two million years ago? These ancestors left some tools and bones, and paleoanthropologists can make some ingenious inferences from such evidence. But how can we possibly obtain the key information that would be required to show the validity of adaptive tales about an LEA: relations of kinship, social structures and sizes of groups, different activities of males and females, the roles of religion, symbolizing, storytelling, and a hundred other central aspects of human life that cannot be traced in fossils? (paragraph 31; also Gould, 2000, p. 120)

The limitations suggested in this section are not meant as a criticism of the EP program: methods in all disciplines have limitations, and such limitations do not necessarily preclude their usefulness. As ethnologist and EP critic Bateson (2000) observes, the suggestion that EP is a collection of Just-So stories is only partly justified: "Stories about current function are not about how the leopard got its spots, but what the spots do for the leopard now. That is a question testable by observation and experimentation" (p. 204). EP's challenge is to find that set of methods, independent of recourse to the EEA, that allows the confirmation or falsification of hypothesized mental modules, and current methods are a good start. The essential program, then, may hold considerable promise. Considering these limitations should, however, cause us to critically examine EP's rhetoric of premature triumphalism and distinctive preemptive claims (for such an examination, see Herrnstein Smith, 2000).

Physicalism 

Like most of their contemporaries in the cognitive and neural sciences, EP practitioners hold a materialist (i.e., physicalist) position that is at odds with the traditionally dualist anthropologies of Christianity. Since Descartes, the concepts of soul and mind have become inextricably linked, and the concept of a purely physical soul does not enjoy widespread acceptance in Christian orthodoxy.

It is hard to imagine how EP could work under any form of strong anthropological dualism in which mental phenomenas arise purely from a nonphysical soul substance. Of course, this does not preclude it working in a system of cosmological dualism in which the two types of things include (non-physical) God and (physical) creation (e.g., see Murphy, 1998). Nor does it preclude dualisms of the interactionist variety. Just because current spokespersons for EP reject cosmological and anthropological dualism does not mean that EP entails such rejections.

There are anthropological physicalist theories of human nature that comport well with both evolutionary positions and a wide variety of theistic worldviews. These are worth considering briefly here because the same central issue arises in both evolutionary theories of the origin of human nature and physicalist (neurobiological) theories of mind. How can purposive activity arise out of law-governed processes? In neurobiological theories of mind and behavior, we wonder how human purposes (e.g., I give my wife flowers because I love her, and wish her to know it) can be said to account for behavior if that same behavior is caused by physiological processes (e.g., neurotransmitter binding, action potentials, and the like). In evolutionary theories, we wonder how God's purposes (e.g., the redemption of creation) could possibly be responsible for the unfolding of history if that same history is caused by lawful evolutionary processes (e.g., natural selection, genetic drift, etc.).

Space does not permit a full exposition of physicalist theories of mind. (7) In all cases, human mental and physical behavior arises out of physical matter (such a position can be referred to as metaphysical reductionism), but such concepts as "emergent properties" and "downward causation" allow for mental and functional/behavioral explanations to retain ontological status, causal efficacy (avoiding ontological reductionism and causal reductionism, respectively, as well as epistemological reductionism; see below). Such non-reductionist anthropologies are similar to many recent views proposed by neuroscientists outside of the Christian literature, including Damasio (1994), LeDoux (1996), and Ramachandran and Blakeslee (1998). Such multilevel perspectives stand in contrast to reductionist theories in which consciousness, intentionality, etc. are mere epiphenomena with no ontological, causal, or epistemological status. (The latter view seems to be increasingly rare, but has been attributed to Crick, e.g., 1994). If contemporary neuroscience presents physical anthropologies that are non-reductionistic in the sense described here, and if such physicalist theories comport well with both evolutionary and Christian perspectives, then EP cannot be ruled out by Christian thinkers a priori on the basis of its physicalist assumptions.

Reductionism

Critiques of physicalist anthropologies typically center on issues of reductionism, and the defense of a physicalist position within a Christian worldview requires that phenomena, such as moral behavior, not be completely reduced to the laws of cellular physiology, chemistry, or physics. If they were (i.e., if there were no room for purposive action), it would be difficult to argue that a just God, or anyone else, could hold individual humans morally responsible for their actions. Discussions about evolutionary processes in Christian circles also seem to revolve around issues of reductionism: If scientific descriptions of evolutionary causes (e.g., genetic and selective) are the only possible levels of description, then it would be hard to argue that God works purposes of redemption, providence, etc. in creation. In both cases, however, there is no reason to believe that lawfulness at lower levels of explanation precludes genuine agency at higher levels.

The purposiveness of human behavior is obvious only from a subjective, first-person point of view; from a more objective standpoint, purposiveness in human agency must be inferred. Evolutionary psychologists, as with others in the cognitive sciences, see no compatibility problems between descriptions at neural, cognitive (intentional), and behavioral/functional levels (cf. Marr's three levels). However, what if there were no first-person perspectives? What if cognitive scientists were aliens (Marr-tians?), investigating the human species "from the outside"? Contemporary doctrine in the cognitive sciences holds that the analysis would change little -- we would still be justified in taking an "intentional stance" as well as assuming lawful causality on a neural level. This more objective viewpoint is precisely the position in which we find ourselves as we examine the evolution of the physical universe; one must imagine that the subjective viewpoint of God is rather different. As with human intentionality, the d iscovery of more proximal laws of evolution (e.g., natural selection) does not preclude the truth or utility of more ultimate levels of explanation (e.g., God's agency). Although evolutionary psychologists eschew neurobiological reductionism in favor of a multi-level analysis when it comes to questions of anthropology, many seem to have opted for a one-level evolutionary reductionism when it comes to questions of cosmology. Such a treatment hardly seems warranted within the EP framework, and is of course ruled out within any theistic worldview.

Claims of reductionism are often slippery things because the idea of reductionism can take on a variety of meanings. Consider, for example, the following hierarchy of reductionisms, listed in order from the weakest to the strongest (stronger claims entail weaker ones, but not the other way around). In all cases, the reduction relationship is between two particular things, where thing y reduces to component things x in the following possible senses, where components x are always physical.

1. Methodological reductionism claims that y results at least in part from x, and analysis of y into x is useful for understanding y.

2. Metaphysical (or ontological) reductionism claims that y results from x, and not from any additional nonphysical substance; facts about y are fixed by facts about x, that is, there is no additional metaphysical ingredient.

3. Causal reductionism claims that y is strictly determined by x in the additional sense that causation is strictly "bottom-up." Such a view may allow for some sorts of emergent properties (epiphenomena), but not for the "down-ward causation" allowable under weaker forms.

4. Epistemological reductionism claims that all explanations of y can be had in terms of explanations of x. Not only does physics fix the facts of the universe, but explanations in terms of physics suffice for all phenomena; "higher level" explanations are at best shorthand for more precise explanations at the lower level.

5. Eliminative materialism claims that only the components x exist at all; to talk about y is not just redundant (and therefore to be reduced), but false and misleading (and therefore to be eliminated).

Even the weakest reductionism is stronger than genuine non-reductionist claims. Such truly non-reductive views might include those relying strictly on downward causation (e.g., mind-body dualisms in the case of intentional human behavior, or occasionalist positions on God's agency in the natural world) and also those relying on external intervention (e.g., perhaps "spirit possession" or miraculous intervention by God). Before turning to the relevance of these concepts to the evaluation of EP, one additional philosophical term relating to agency and causation needs to be introduced: supervenience (e.g., Kim, 1993, 1996; Murphy, 1998). The notion of supervenience provides a plausible solution for the Christian adopting a physicalist position on the problem of mind, body, and soul; it also provides a partial solution for such people adopting an evolutionary explanation for the origin and diversity of life.

In the language of supervenience, description y supervenes on description x (or property y supervenes on property x) if and only if two objects cannot differ with respect to y without also differing with respect to x (but not the other way around; objects differing in x but not y form the case in which y is said to be multiply realizable, and this is important in some definitions of supervenience). Although there is not agreement on a precise definition, the general usage seems also to be that although both descriptions (or properties) are equally true, one does not reduce to the other in either a causal or epistemological sense, even though one may reduce in a metaphysical sense. Supervenience relationships have been proposed in moral philosophy, relating ethical judgments to descriptive judgments (Hare, 1952) and in philosophy of mind, relating mental to physical descriptions (Davidson, 1970; Kim, 1993). In some circumstances, the supervenience relationship may also be a reasonable way to talk about divine action within the natural world.

Consider the following scenarios in which a supervenience relationship might be more satisfactory than either reduction of one description to another or elimination of one description altogether:

1. In ethics: St. Francis' goodness consisted in the specific kind acts he performed, but it is also true that his goodness ran somewhat deeper than that (and, e.g., might well have been expressed in different acts under different circumstances). The description of St. Francis as "good" supervenes on descriptions of his specific actions.

2. In philosophy of mind: My giving flowers to my wife results from certain neural activity in my brain, but it is also true that it results from my intention to give her flowers. My having an intention to give her flowers supervenes on the description of certain neural activity in my brain.

3. In divine agency: The evolution of the human species results from natural selection, but it is also true that God created humans, sustaining and directing their development. The description of God's creation and sustenance of the human species supervenes on descriptions in terms of natural selection.

The difference between physicalist anthropologies and theistic evolution is that in the case of human behavior physicalist anthropologies state that agency is strictly emergent and any downward causation emanates from those same emergent properties. In the case of Christian theistic evolution, the divine agent cannot be merely an emergent property, but it must exist independently as well. God may indeed be immanent in creation (cf. Peacocke, 1993), infusing and sustaining nature in its very lawfulness, and we may well imagine that this is God's normal mode of effecting His purposes in the cosmos. Yet, God is also transcendent: If the world is God's body, as the immanence idea suggests, then evangelical Christians are dualists in this respect, and they must attribute to God an additional spirit substance from which special divine acts proceed. Murphy (1997) has proposed a model of divine action in the natural order in which God works (a) in a "bottom-up" fashion, immanent in all of nature at the level of quant um phenomena; (b) in special acts, as in the acts of inspiring scripture or in response to petitionary prayer; and, (c) in extraordinary acts (e.g., in the resurrection). The fact that God sometimes works in extraordinary ways does not mean, of course, that he does not often achieve his purposes through the natural laws that he designed and sustains. Those of us in the cognitive and neural sciences, who are accustomed to working at multiple levels of explanation that include both neural and behavioral laws and bona fide intentionality, should not be averse to such an account.

The implications of such views for evangelical Christians involved in the cognitive neurosciences and broader biological sciences are far-reaching. Adopting physicalist or evolutionary perspectives does not entail being an eliminativist about either human or divine agency; it does not even entail epistemological reductionism or, in the case of divine agency, metaphysical reductionism. If this is so, one need not reject EP on the grounds that it necessarily rules out either free will of the sort that could justify moral responsibility or divine agency in biological evolution. It does, of course, rule out certain forms of creationism in which geological time scales are precluded and species arise de novo by extraordinary divine acts. If EP is reductionist or eliminativist in practice, it is not so by necessity.

Probably the most common criticism of EP is that it is in fact reductionist in practice, and often eliminativist. With respect to divine action, for example, I have not found a single instance in self-proclaimed EP literature that is not eliminativist: God (or any divine entity) is simply assumed to be nonexistent. This in itself is not alarming; that God is seen to be irrelevant in the sciences and humanities is a shame, but it is not new news. What is disturbing is that where much of the secular academy is passively eliminativist with respect to spiritual reality, EP is actively so. Most claims of EP reductionism do not come from people who are spiritually minded in any way. Rather, claims of reductionism come both from within the natural sciences and from the social sciences and humanities.

EP's reductionism extends beyond issues of spirituality. Following Dawkins (e.g., 1976, 1986), genes seem to get primary or exclusive ontological priority for biological explanations: Many in EP agree with Dawkins in that organisms are "lumbering robots" or survival machines built by immortal genes motivated to self-replication. Thus the biological sciences, including experimental psychology, are reduced to the methods, principles, and explanatory modes of evolutionary biology. Even evolutionary biology seems to be a target of reductionism in EP, with the adaptationist program receiving nearly exclusive priority. In most cases, the reductionism does not seem to be merely metaphysical, but causal and epistemological. The philosopher Jerry Fodor has noted this strong sense of reductionism in two book reviews, one on mainline EP books by Pinker (1997) and Plotkin (1998; Fodor, 1998a), and the other on E. O. Wilson's (1998) closely related book Consilience (Fodor, 1998b). It may well be the case, says Fodor of Wi lson's thesis, that physics fixes all the facts there are (i.e., metaphysical reductionism), but this in no way implies that all the explanations that there are physical explanations (i.e., epistemological reductionism). The same applies with respect to biological creation if one replaces the term "physics" with "natural selection."

EP has also been strongly criticized regarding its treatment of the social sciences, arts, and humanities. As is the case with EP and the biological sciences, the reductionism is stronger than metaphysical, and is sometimes frankly eliminative. Pinker's (1997) final chapter on "The Meaning of Life" exemplifies EP's epistemological reductionism with respect to art, music, humor, philosophy, and religion. None of these are adaptive in the biological sense, and so cannot be considered as adaptations (i.e., basic mental modules). They must, then, according to EP logic, be "mere by-products," somehow engaging pleasure-circuits of the brain that evolved for genuinely adaptive purposes. Music, for example, "is crafted to tickle the sensitive spots of at least six of our mental faculties" (p. 534), including language, auditory scene analysis, emotional calls, habitat selection, and "something else." The problem here seems not to be that EP can offer a plausible account of the how such cultural phenomena arose under e volutionary pressures. The problem is that EP rhetoric (e.g., mere by-products," "incidental side effects") makes any other explanation of the phenomena secondary and uninteresting unless it is explicitly justified in evolutionary terms.

The EP literature is replete with references to the Standard Social Science Model (SSSM), a term introduced by Tooby and Cosmides (1992) to describe the pre-Darwinian set of assumptions that, they say, still guides the majority of work in the social sciences. The SSSM is strongly empiricist, focuses on the importance of individual experience and cultural factors, neglects biological constraints, relies on one or a few general-purpose learning mechanisms, and has as its goal to explain the variability of humanity in these terms. H. Rose (2000) develops the claim that EP has constructed a straw enemy in the SSSM and that EP has as its primary goal the colonization of the social sciences, establishing unity in the sciences by the reduction of all explanations of culture to principles of evolutionary biology.

The adaptationist reductionism of Pinker, Cosmides, Tooby, and others with regard to cultural phenomena is interesting but overstated. In the first place, if one takes their reductionism in the relatively weak sense of metaphysical reductionism, there is no reason to believe that a good number of episremologically irreducible explanations might coexist along side evolutionary ones. Second, the analysis of culture in terms of natural selection seems at best premature: As argued by Gould (e.g., 2000), Ayala (1998), and others, cultural phenomena seem to evolve according to Lamarkian means, with acquired characteristics (e.g., theories, belief systems, etc.) being passed from one generation to the next, not simply according to Darwinian means (i.e., natural selection).

Nativism 

At first glance, and perhaps even on closer inspection, the evolutionary approach would seem to be rather neutral on the nativism-empiricism (i.e., nature-nurture") argument. Indeed, EP literature often claims that the nature-nurture dichotomy is malformed. Of course, in order for EP to work, at least some structure must be passed on genetically, and such structure cannot be acquired during an organism's lifetime (biological evolution is not Lamarkian). Beyond this, however, one might imagine that EP would held the precise nature of our mental adaptations to be an empirical question. And so, for example, Cosmides and Tooby (1997) develop a very coherent "adaptationist perspective" on nature and nurture, distinguishing a number of nature-nurture dimensions that are sometimes conflated: what it means for something to be present at birth, the view of all phenotypes ("from molars to memory circuits") as a joint product of genes and environment, and the relationship between learning and innate structure. However, it is sometimes hard to take EP's claims of neutrality on the nature-nurture issue at face value.

Consider, for example, Pinker's statement that "[e]very part of the body, from the toenails to the cerebral cortex, takes on its particular shape and substance when its cells respond to some kind of information in its neighborhood that unlocks a different part of the genetic program" (Pinker, 1997, p. 35). The rhetoric here is clear: environmental information unlocks innately specified genetic programs.

The claim that EP is nativist hinges not only on its assumption of predetermined genetic structure, but also on its assumption of domain-specific mental modules. People who claim that we are equipped with a genetically specified general-purpose learning devise, a neural blank slate, are empiricists. To say, however, that we are equipped with a domain-specific inference system or "reasoning instinct" for detecting cheaters is to make a distinctly nativist claim. Not that this is necessarily bad; Fodor (1998a), for example, sees the rationalist/nativist aspect of EP as one of its only redeeming qualities. The point is that EP, as defined by the current literature, is strongly nativist, more so than is required by the adaptationist program. Given conceptual methods of evolutionary biology and the empirical methods of experimental psychology and the neurosciences, just how much the basic psychological functions are characterized by domain specificity and innate content is an empirical question.

Computationalism and Mentalism 

One of the important things to recognize about EP is its relationship to extant subdisciplines in psychology and related sciences. As a species of psychology, EP has clearly evolved from the cognitive psychology lineage. It embraces mentalism (i.e., the assumption that mental states such as beliefs, desires, precepts, plans, etc., are amenable to scientific investigation and are important links in the causal chain of human behavior), and computationalism (i.e., the assumption that mental operations involve algorithmic transformations of symbolic mental representations).

Although cognitive psychology's computational framework is standard fare for many psychologists these days, it certainly is not the only conceivable framework for understanding mind and behavior. One need not be a radical behaviorist to question the strictly computational approach (although many forms of radical behaviorism are alive and well in the psychological sciences today; see, e.g., Catania, 1998). Within the cognitive science community interesting and viable alternatives have been proposed, including the dynamical systems perspective (e.g., Port & van Gelder, 1995) and connectionist perspectives (e.g., McClelland & Rumelhart, 1986).

The point again is not which view is right, or whether some of the various alternative views might not be able to be reconciled with the computationalist, mentalist perspectives. The point is that the jury is still out, and (to use an evolutionary metaphor) the richness of diversity in today's perspectives on psychology forms the basis for selection to determine tomorrow's winners. An evolutionary psychology need not prematurely commit itself exclusively to the perspectives currently in fashion.

Religion

Many of the foregoing issues allow for a broad range of choices for Christian thinkers in the evangelical traditions, and I have tried to show that the defining aspects of EP fit, for the most part, within this range. There is, however, one apparently universal aspect of EP that cannot be reconciled with a Christian worldview: the outright rejection of even the most liberal theology. As is the case with many of the other issues, this is so in practice but not by necessity. This is another sense in which some in the EP enterprise might earn the "fundamentalist" label.

Pinker's (1997) book serves as a prime, but by no means the only, example of EP's anti-religion bias. The last chapter of the book is entitled "The Meaning of Life," and the final section of this chapter is devoted to religion. Where most of the rest of the book is a paragon of clear exposition and careful scholarship, this section could hardly be more offensive to religiously minded scholars of any ilk. He begins the section with a quote by H. L. Mencken: "The most common of all follies," says Pinker through Mencken, "is to believe passionately in the palpably not true. It is the chief occupation of mankind" (p. 554). Such palpable untruths include, says Pinker, the belief in a God or universal spirit and the belief that Jesus was raised from the dead.

On the one hand, Pinker treats religion similarly to art, music, and other "cultural" phenomena: they are "Sunday afternoon projects of little adaptive value" (1997, p. 524), assembled using the toolbox of innate psychological modules with which natural selection has equipped us. "Like the psychology of the arts," he says, "the psychology of religion has been muddied by scholar's attempts to exalt it while understanding it" (p. 555). On the other hand, whereas the arts may have some redeeming value as non-adaptive by-products, religion seems to have none at all. Perhaps this is because whereas he considers the arts to be merely "trivial and futile" (p. 521), religion is just plain false as well. He quotes Blaise Pascal as saying, "men never do evil so completely and cheerfully as when they do it from religious conviction" (p. 555). Given Pinker's amateurish theology and Pascal's lifelong dedication to the Christian faith, one is reminded of Kierkegaard's comment regarding Faust's reflection, "I now do see tha t we can nothing know." It is one thing, Kierkegaard suggests, for Faust to make this observation at the end of a life dedicated to the pursuit of knowledge; it would be quite another thing for a college freshman to use Faust's quote in justification of laziness. (8)

However, religion is even more than false, in Pinker's view; it is downright pernicious. "Religions_often serve the interests of the people who promulgate them," he says, suggesting that "priests are Wizards of Oz who use special effects, from sleight of hand and ventriloquism to sumptuous temples and cathedrals, to convince others that they are privy to forces of power and wonder" (1997, p.555). It is one thing to say that people are religious, and that is often in their interest to be so, but it is another thing to impute self-interest as the sole or primary conscious motive (note the strong reductionism).

Presumably of the practice of prayer, he asks, "[w]hat kind of mind would do something as useless as inventing ghosts and bribing them for good weather?" (1997, p. 556); it would appear that "[r]eligion is a desperate measure that people resort to when the stakes are high and they have exhausted the usual techniques for the causation of success" (p. 556).

Demonstrating his lack of experience with any serious scholarship by or about religious people, he resorts to ridicule to describe the lack of reflection on the part of believers: "They don't pause to wonder why a God who knows our intentions has to listen to our prayers, or how a God can both see into the future and care about how we choose to act. Compared to the mind-bending ideas of modern science, religious beliefs are notable for their lack of imagination (God is a jealous man; heaven and hell are places; souls are people who have sprouted wings)" (1997, p. 557).

Pinker's strong anti-religion rhetoric is certainly extreme amongst EP scholars, but his views seem to be characteristic of many others in the field. Consider, for example, a recent EP contribution in the psychology of religion by Kirkpatrick (1999). To his credit, Kirkpatrick acknowledges the irreducibility of religion to EP modes of explanation: "it would be foolish to think that theology, spirituality, or religion can be fully explained by EP or any other social science" (p. 943). Although this is refreshing, it is hard to reconcile with many of the statements in the rest of his paper.

Kirkpatrick approvingly notes that Pinker (1997) "points out that all religious and political movements attempt to subvert families in an attempt to redirect loyalties toward the larger group" (1999, p. 937), and that Crippen and Machalak (1989) "attempt to explain religion generally in terms of a 'hypertrophied kin recognition process"' (p. 937). Note the strong epistemological reductionism, particularly in the latter case. With respect to religious ethical systems concerning marriage, sex, and the like, he suggests that many cases such laws are probably best understood in terms of how they serve the interests of those in power" (p. 939; the italics are mine, and emphasize the epistemological reductionism). Considering anxiety attacks as non-adaptive by-products, reflecting a "misfiring or inappropriate activation of emotional mechanisms that ordinarily are adaptive," he suggests "that mystical experience eventually will be understood in a similar manner" (p.94l).

My complaint with Kirkpatrick's, Pinker's, and other EP accounts of religion is not due to their explicit lack of theistic references: most science is not theistic, and we've learned to live with it in chemistry, physics, biomedicine, etc. Nor is it that they offer and defend explanations of religious behavior and experience in terms of adaptive mental modules and other fitness-increasing traits, such as kin selection, attachment, power, status, and social exchange. These are typically empirical questions, and if I had to guess, I'd guess that such accounts will figure prominently in an accurate psychological theory of religion, including evangelical Christianity (and perhaps EP, as well). My complaint is that such accounts claim exclusive explanatory status-they assume epistemological reductionism with respect to explanations at other levels and eliminativism with respect to God.

There is nothing in the EP approach, or in the sciences more generally, that necessitates such strong reductionism. Echoing a claim common since Darwin's publication of Origins and borrowing terminology from Dawkins, Pinker suggests that Darwin "showed how organs of extreme perfection and complication, which justly excite our admiration, arise not from God's foresight, but from the evolution of replicators over immense spans of time" (p. 22). Evolution, we are asked to believe, implies atheism (or at least impotent deism), but where is the argument that supports this? The two premises from which atheism purportedly follows are, (a) natural selection sufficiently explains (9) the formative history of life, and (b) natural selection explanations preclude explanations based on miraculous divine intervention. (10) What conclusions do these premises warrant? God doesn't exist? That would be an egregious non-sequitur. There is no such thing as miraculous divine intervention? No. How about the formative history of l ife is outside of God's providence and divine action? Not even this follows from our premises. If it did, there could be no natural law accounts of anything without ruling out God's providence and sovereignty. As Christians in the sciences, we recognize God's hand in all of creation, whether by miraculous divine intervention or through some other mode of divine action.

CONCLUSION

EP's reductionist tendencies and its strong triumphal and preemptive rhetoric are offensive not only to Christians, but to many others in academia as well. For those of us in the psychological, cognitive, and neural sciences who are hopeful that evolutionary approaches (and even EP itself) might hold real promise, it is particularly frustrating to think that EP's offensive rhetoric might cause a backlash within our discipline that would postpone the discovery of its ultimate utility.

It is with this concern in mind that I offer a final comparison between EP and phrenology: Like Gall's phrenology, EP is being packaged in a way that could lead to its demise even if it's basic premises and methods turned out to be sound. Like its predecessor sociobiology EP is being packaged beyond the community of scientists in the relevant disciplines to a broader public, which seems eager to hear its message. Pinker's (1997) book, one of the most widely and popularly read pieces in EP, is an excellent example of rhetoric that is too strong for an immature discipline. Even the title, How the Mind Works, is a bit strong. Treating hypotheses--even really good ones--as facts might be good for selling books, but it is ultimately injurious to the pursuit of truth and knowledge. Ironically, this was well said by Darwin in The Descent of Man:

False facts are highly injurious to the progress of science for they often endure long; but false hypotheses do little harm, as everyone takes a salutary pleasure in proving their falseness; and when this is done, one path toward error is closed and the road to truth is often at the same time opened. (11)

AUTHOR

FIKES, THOMAS C. Address: Department of Psychology, Westmont College, 955 La Paz Road, Santa Barbara, CA 93108-1099. Title: Associate Professor of Psychology. Degrees: PhD, University of California, Santa Barbara. Specializations: Cognitive and behavioral neurosciences, brain mechanisms for implicit learning, control of voluntary movement.

(1.) It is basic doctrine of the Christian faith that God created the universe and everything in it, and that He sustains His creation. There are, of course, a variety of opinions about how this was and continues to be accomplished. These include, on the one hand, episodic creationists who maintain that God actively intervened in an otherwise relatively autonomous natural world during discrete episodes in time (e.g., during the first six literal days for young-earth creationists, or distributed over the first 14-plus billion years for intelligent design theorists). On the other hand, evolutionary creationists maintain that the relative autonomy of normal natural processes provides sufficient latitude for God to work his will in creation, and that natural selection, etc., are the processes by which he created the various forms of life, including humans, now inhabiting the planet. There are significant theological and scientific problems associated with each view, and I take it that there is ample room in the d iscussion between orthodox Christians for these (and perhaps other) views. See, for example, Carlson (2000) for charitable presentations and responses by proponents of the various views.

(2.) Actually, Gall referred to his theory as "organology"; only later did it go by the name "phrenology" (Rosensweig, 1996).

(3.) Because some maintain that atheism is a fundamental claim of any evolutionary theory, and since I consider my Christianity to be orthodox (a fortiori, I am not advocating atheism), the claim that "the most fundamental premises of EP seem reasonable and permissible" requires some clarification. For the purposes of the article, here are some of my relevant beliefs. First, humans evolved, and natural selection (as outlined by Darwin, 1859/1964) played an important (even primary) role. Ayala (1998) provides an overview of the timeline as currently accepted by evolutionary biologists, with, for example, the first hominids emerging approximately 4.4 million years ago and Homo sapiens arriving some 400,000 years ago (Eccles, 1989/1991, provides a similar timeline; both Ayala's and Eccles accounts echo a general consensus in the evolutionary disciplines and are developed within a Christian worldview). Second, one of the organs that evolved is the brain, and it is the brain's functioning that principally underlie s the control of behavior and "is" (in some sense) our mental life. Third, the brain is heterogeneous in structure, with each structure serving specific functions, and this heterogeneity allowed evolutionary selection of adaptive "functional modules" somewhat independently of each other. Fourth, I also believe that we exhibit genuine agency (e.g., that one can be held morally responsible for one's actions); in the Usage of Murphy (1998), descriptions of bona fide human agency supervene on descriptions of physical/physiological processes in a way that precludes strong reductionist accounts. I think these four beliefs are probably shared by the majority of my secular colleagues in the brain and cognitive sciences. Fifth, I believe that the purposes of a sovereign God also supervene on the lawful activity of all of creation, that God sometimes intervenes in acts of special or extraordinary divine action (for example, the virgin birth of Jesus), and that this creation "groans for redemption" and is inexorably mov ing in the direction of a "new heavens and a new earth." In short, I hold a fairly high view of evolutionary biology and the cognitive, neural, and behavioral sciences, and a much higher view of the central tenets of evangelical Christianity.

(4.) The use of the language of design is interesting here, and seems ubiquitous in the EP literature (Dennett, 1995, he even capitalizes the word when it occurs). The argument is that some biological devices are clearly designed in the sense that they are better-than-chance engineering solutions to specific problems. However, where the argument from design has typically been an argument for an intelligent designer (Paley, 1802; Behe, 1996; Meyer, 2000), contemporary evolutionary rhetoric commandeers the word and argues that evidence for design implicates natural selection. As I argue below, the word might fruitfully be used by evolutionary creationists in both senses, simultaneously and without contradiction.

(5.) Cosmides and Tooby's (1997) primer is available online (see references), and is an excellent introduction to the field from the EP perspective. For equally readable and informative criticisms, see Gould (1997) for an evolutionary perspective and H. Rose (2000) for a social sciences perspective.

(6.) In this sense EP is similar to its predecessor sociobiology and has been similarly criticized in the evangelical Christian literature (e.g., see Yancey, 1998, and Guelzo, 1998).

(7.) The interested reader may consult sources such as Mackay (e.g., 1991), Brown, Murphy, & Malony (1998; especially chapters by Murphy and Jeeves), and Peacocke (e.g., 1993) for physicalist anthropologies developed in explicitly Christian worldviews.

(8.) I am indebted to Bonhoeffer (1937/1963, p. 55) for Faust's quote and Kierkegaard's comment.

(9.) What counts as a sufficient explanation is important here, Perhaps, in keeping with the levels of explanation theme, it would be best to think of it as sufficient at a particular level. It does not mean that accounts at other levels (lower levels, such as molecular genetics or higher levels, such as non-miraculous divine action) are precluded. It is meant here in the same way we might say that gravity is a sufficient explanation for the trajectories of falling bodies or orbiting planets or that bacterial infection is a sufficient explanation for certain fevers.

(10.) This second premise is merely a metaphysical definition, separating relatively autonomous natural phenomena from direct miraculous intervention.

(11.) Quoted in Ramachandran, 1998, p. xvi.

REFERENCES

Ayala, F. J. (1998). Human nature: One evolutionist's view. In W. S. Brown, N. Murphy, & H. Malony (Eds.), Whatever happened to the soul: Scientific and theological portraits of human nature. Minneapolis, MN: Fortress.

Baron-Cohen, S. (2000). The cognitive neuroscience of autism: Evolutionary approaches. In M. Gazzaniga (Ed.), The new cognitive neurosciences (2nd ed.). Cambridge, MA: MIT Press.

Baron-Cohen, S., Leslie, A. M., & Frith, U. (1985). Does the autistic child have a "theory of mind"? Cognition, 21, 37-46.

Bateson, P. (2000). Taking the stink out of instinct. In H. Rose and S. Rose (Eds.), Alas, poor Darwin: Arguments against evolutionary psychology New York: Harmony Books.

Behe, M. J. (1996). Darwin's black box: The biochemical challenge to evolution. New York: Simon & Schuster.

Bonhoeffer, D. (1937/1963). The cost of discipleship. New York: Macmillan.

Brown, W. S. (1998). Cognitive contributions to soul. In W. S. Brown, N. Murphy, & H. N. Malony (Eds.), Whatever happened to the soul? Scientific and theological portraits of human nature. Minneapolis, MN: Fortress Press.

Brown, W. S., Murphy, N., & Malony, H. (Eds.). (1998). Whatever happened to the soul: Scientific and theological portraits of human nature. Minneapolis, MN: Fortress Press.

Carlson, R. F. (Ed.). (2000). Science & Christianity: Four views. Downers Grove, IL: InterVarsity Press.

Catania, A. C. (1998). Learning (4th ed.). Upper Saddle River, N.J.: Prentice Hall.

Chomsky, N. (1965). Aspects of the theory of syntax. Cambridge, MA: MIT Press.

Churchland, P. S. & Sejnowski, T. J. (1992). The computational brain. Cambridge, MA: MIT Press

Cosmides, L. (1989). The logic of social exchange: Has natural selection shaped how humans reason? Studies with the Wason selection task. Cognition, 31, 187-276.

Cosmides, L. & Tooby, J. (1989). Evolutionary psychology and the generation of culture, part II. Case study: A computational theory of social exchange. Ethology and Sociobiology 10, 51-97.

Cosmides, L. & Tooby, J. (1992). Cognitive adaptations for social exchange. In J. Barkow, L. Cosmides, & J. Tooby (Eds.), The adapted mind: Evolutionary psychology and the generation of culture. New York: Oxford University Press.

Cosmides, L. & Tooby, J. (1995). From function to structure: The role of evolutionary biology and computational theories in cognitive neuroscience. In M. Gazzaniga (Ed.), The cognitive neurosciences. Cambridge, MA: MIT Press.

Cosmides, L. & Tooby, J. (1997). Evolutionary psychology: A primer. Retrieved December, 2000 from the World Wide Web: http://www.psych.ucsb.edu/research/cep/primer.html

Cosmides, L. & Tooby, J. (2000). The cognitive neuroscience of social reasoning. In M. Gazzaniga (Ed.), The new cognitive neurosciences (2nd ed.). Cambridge, MA: MIT Press.

Crick, F. H. (1994). The astonishing hypothesis: The scientific search for the soul. New York: Simon & Schuster.

Crippen, T. & Machalak, R. (1989). The evolutionary foundations of the religious life. International Review of Sociology 3, 61-84.

Damasio, A. R. (1994). Descartes' error: Emotion, reason, and the human brain. New York: G. P. Putnam.

Davidson, D. (1970). Essays on actions and events. In L. Foster and J. W. Swanson (Eds.), Experience and theory. Boston: University of Massachusetts Press.

Darwin, C. (1859/1964). On the origin of species. Cambridge, MA: Harvard University Press.

Dawkins, R. (1976). The selfish gene. Oxford: Oxford University Press.

Dawkins, R. (1986). The blind watchmaker. Harlow, U.K.: Longman.

Dennett, D. C. (1987). The intentional stance. Cambridge, MA: MIT Press.

Dennett, D. C. (1995). Darwin's dangerous idea: Evolution and the meaning of life. New York: Simon & Schuster.

Eccles, J. C. (1989/1991). Evolution of the brain: Creation of the self. New York: Routledge.

Eldridge, N. & Gould, S.J. (1972). Punctuated equilibria: An alternative to phyletic gradualism. In T.J. M. Schopf (Ed.), Models of paleobiology (pp. 82-115). San Francisco: Freeman, Cooper.

Fodor, J. (1998a, January 15). The trouble with psychological Darwinism. London Review of Books, 20 (2). Retrieved December 19, 2000, from the World Wide Web: http://www.lrb.co.uk/v20n02/fodo2002.html

Fodor, J. (1998b, October 29). Look!. London Review of Books, 20 (21). Retrieved December 19, 2000, from the World Wide Web: http://www.lrb.co.uk/v20/n21/fodo2021.htm

Gibson, J. J. (1966). The senses considered as perceptual systems. Boston: Houghton Mifflin.

Gould, S. J. (1997, June 26). Evolution: The pleasures of pluralism. New York Review of Books. Retrieved January 26, 2001 from the World Wide Web: http://www.nybooks.com/nyrev/WWWarchdisPlay.cgi?l9970626047F

Gould, S. J. (2000). More things in heaven and earth. In H. Rose and S. Rose (Eds.), Alas, poor Darwin: Arguments against evolutionary psychology New York: Harmony Books.

Guelzo, A. C. (1998, January/February). Soulless, Books & Culture, 20-23.

Hare, R. M. (1952). The language of morals. Oxford: Clarendon Press.

Herrnstein Smith, B. (2000). Sewing up the mind: The claims of Evolutionary Psychology. In H. Rose and S. Rose (Eds.), Alas, poor Darwin: Arguments against evolutionary psychology. New York: Harmony Books.

Hubel, D. H. & Weisel, T. N. (1968). Receptive fields and functional architecture of monkey striate cortex. Journal of Physiology (London). 195, 215-243.

Kandel, E. R., Schwartz, J. H., & Jessell, T. M. (1991). Principles of neural science (3rd ed.). Norwalk, CT: Appleton & Lange,

Kauffman, S. A. (1993). The origins of order: Self-Organization and selection in evolution. New York: Oxford University Press

Kim, J. (1993). Supervenience and mind. Cambridge: Cambridge University Press.

Kim, J. (1996). Philosophy of mind. Boulder, CO: Westview Press.

Kirkpatrick, L. A. (1999). Toward an evolutionary psychology of religion and personality. Journal of Personality 67,921-952.

Lashley, K. S. (1929). Brain mechanisms and intelligence. Chicago: University of Chicago Press.

LeDoux, J. E. (1996). The emotional brain: The mysterious underpinnings of emotional life. New York: Simon & Schuster.

LeDoux, J. E., Cicehetti, P., Xagoraris, A., & Romanski, L. M. (1990). The lateral amygdaloid nucleus: Sensory interface of the amygdala in fear conditioning. Journal of Neuroscience, 10, 1062-1069.

Leslie, A. M. (2000). "Theory of mind" as a mechanism of selective attention. In M. Gazzaniga (Ed.), The new cognitive neurosciences (2nd ed.). Cambridge, MA: MIT Press.

Mackay, D. M. (1991). The open mind and other essays: A scientist in God's world. Leicester: InterVarsity.

Marr, D. (1982). Vision: A computational investigation into the human representation and processing of visual information. San Francisco: Freeman.

McClelland, J. L. & Rumelhart, D. E. (Eds.). (1986). Parallel distributed processing: Explorations in the microstructure of cognition (Vols. 1, 2). Cambridge, MA: MIT Press.

Meyer, S. C. (2000). Qualified agreement: Modern science and the return to the "God hypothesis." In R. F. Carlson (Ed.), Science & Christianity: Four views. Downers Grove, IL: InterVarsity Press.

Murphy, N. (1997). Divine action in the natural order: Butidan's ass and Schrodinger's cat. In R. J. Russel, N. Murphy, & A. R. Peacocke (Eds.), Chaos and complexity: Scientific perspectives on divine action. Vatican City State: Vatican Observatory Publications.

Murphy, N. (1998). Nonreductive physicalism: Philosophical issues. In W. S. Brown, N. Murphy, & H. N. Malony (Eds.), Whatever happened to the soul? Scientific and theological portraits of human nature. Minneapolis, MN: Fortress Press.

Paley, W. (1802). Natural theology; or, evidences of the existence and attributes of the deity collected from the appearances of nature. London: Printed for R. Faulder.

Peacocke, A. (1993). Theology for a scientific age: Being and becoming--natural. divind, and human, (2nd ed.). Minneapolis, MN: Fortress Press.

Pinker, S. (1997). How the mind works. New York: W. W. Norton & Co.

Plotkin, H. (1998). Evolution in mind: An introduction to evolutionary psychology Cambridge, MA: Harvard University Press.

Port, R. F., & van Gelder, T. (Eds.). (1995). Mind as motion: Explorations in the dynamics of cognition. Cambridge, MA: MIT Press.

Premack, D., and Woodruff, G. (1978). Does a chimpanzee have a theory of mind? Behavioral and Brain Sciences, 1, 512-526.

Ramachandran, V. S., & Blakeslee, S. (1998). Phantoms in the brain: Probing the mysteries of the human mind. New York: William Morrow.

Rose, H. (2000). Colonizing the social sciences? In H. Rose and S. Rose (Eds.), Alas, poor Darwin: Arguments against evolutionary psychology New York: Harmony Books.

Rose, H. & Rose, S. (2000). Alas, poor Darwin: Arguments against evolutionary psychology New York: Harmony Books.

Rosensweig, M. (1996). Aspects of the search for neural mechanisms of memory. Annual Review of Psychology 47, 1-32.

Shepard, R.N. (1984). Ecological constraints on internal representation: Resonant kinematics of perceiving, imaging, thinking and dreaming. Psychological Review, 91,417-447.

Shepard, R.N. (1987). Evolution of a mesh between principles of the mind and regularities in the world. In J. Dupre (Ed.), The latest on the best: Essays on evolution and optimality Cambridge, MA: MIT Press.

Shepard, R.N. & Metzler, J. (1971). Mental Rotation of Three-Dimensional Objects. Science, 171,701-703.

Tooby, J., & Cosmides, L. (1992). The psychological foundations of culture. In J. Barkow, L. Gosmides, & J. Tooby (Eds.), The adapted mind: Evolutionary psychology and the generation of culture. New York: Oxford University Press.

Tooby, J., & Cosmides, L. (1997). Letter to the Editor of The New York Review of Books on Stephen Jay Gould's Darwinian Fundamentalism (June 12, 1997), and Evolution: The pleasures of pluralism (June 26, 1997). Retrieved December 2000 from the World Wide Web: http://cogweb.english.ucsb.edu/Debate/CEP_Gould html.

Tooby, J. & Cosmides, L. (2000). Toward mapping the evolved functional organization of mind and brain. In M. Gazzaniga (Ed.), The new cognitive neurosciences (2nd ed.). Cambridge, MA: MIT Press.

Tooby, J. & DeVore, I. (1987). The reconstruction of hominid evolution through strategic modeling. In W G. Kinzey (Ed.), The evolution of human behavior: Primate models. Albany, NY: SUNY Press.

Wason, P. (1966). Reasoning. In B. M. Foss (Ed.), New horizons in psychology Harmondsworth: Penguin.

Williams, G. C. (1966). Adaptation and natural selection: A critique of some current evolutionary thought. Princeton: Princeton University Press.

Wilson, E. O. (1998). Conscilience: The unity of knowledge. London: Little Brown.

Yancey, P. (1998, January/February). Nietzsche was right. Books & Culture, 14-17.

Table 1

The Formal Properties of an Adaptation


An adaptation is:

* a cross-generationally recurring set of characteristics of the
  phenotype
* that is reliably manufactured over the developmental life history
  of the organism,
* according to instructions contained in its genetic specificiation,
* in interaction with stable and recurring features of the environment
  (i.e., it reliably develops normally when exposed to normal
  ontogenetic environments),
* whose genetic basis became established and organized in the species
  (or population) over evolutionary time, because
* the set of characteristics systematically interacted with stable and
  recurring features of the ancestral environment (the "adaptive
  problem"),
* in a way that systematically promoted the propagation of the genetic
  basis of the set of characteristics better than the alternative
  designs existing in the population during the period of selection.
  This promotion virtually always takes places through enhancing the
  reproduction of the relatives of the individual.


Note: Abridged from "Toward Mapping the Evolved Functional Organization
of Mind and Brain," by J. Tobby and L. Cosmides, 2000, The New Cognitive
Neurosciences, Table 80.1, p. 1173.
Table 2

The Three Levels at Which Any Machine Carrying out an Information
Processing Task Must be Understood


Computational theory

Representation and algorithm

Hardware implementation

What is the goal of the computation, why is it appropriate, and what
 is the logic of the strategy by which it can be carried our?

How can this computational theory be implemented? In particular, what is
 the representation for the input and output, and what is the algorithm
 for the transformation?

How can the representation and algorithm be realized physically?


Note. From D. Marr, 1982, Vision, Figure 1-4, p. 25.