In this article I argue that evolutionary psychological strategies for making inferences about present-day human psychology are methodologically unsound. Evolutionary psychology is committed to the view that the mind has an architecture that has been conserved since the Pleistocene, and that our psychology can be fruitfully understood in terms of the original, fitness-enhancing functions of these conserved psychological mechanisms. But for evolutionary psychological explanations to succeed, practitioners must be able to show that contemporary cognitive mechanisms correspond to those that were selected for in the environment of evolutionary adaptedness, that these present-day cognitive mechanisms are descended from the corresponding ancestral mechanisms, and that they have retained the functions of the ancestral mechanisms from which they are descended. I refer to the problem of demonstrating that these conditions obtain as “the matching problem,” argue that evolutionary psychology does not have the resources to address it, and conclude that evolutionary psychology, as it is currently understood, is therefore impossible.
Evolutionary psychologists believe that they have an inferential strategy that allows them to give accurate evolutionary explanations for contemporary human behavior. In this article, I call the strategy into question and argue that it is methodologically unsound. I argue that even if it is true that our prehistoric ancestors’ behavior was underwritten by an evolved modular psychology, and further, even if it is true that the contemporary human mind has a massively modular organization, this does not license the sorts of inferences that evolutionary psychologists make about the evolved psychology of contemporary humans.
The structure of the article is as follows. In the next section I discuss the historical and intellectual context from which evolutionary psychology emerged and go on to discuss some of the core theoretical commitments of the discipline. In the third section, I identify two inferential strategies used by evolutionary psychologists. In the fourth section, I set out a methodological problem confronting evolutionary psychological inference-making that I call the matching problem, and I identify explanatory criteria that need to be satisfied for the problem to be overcome. In the fifth section I introduce the notion of “strong vertical homology,” which refers to the relation that must be shown to exist between ancestral cognitive mechanisms and contemporary ones for the matching problem to be overcome. Then, in the next six sections, I provide reasons to think that the matching problem is insoluble and therefore that evolutionary psychology is not possible. In the twelfth section I illustrate the matching problem with an example from the evolutionary psychological literature, and in the following section I compare the pattern of evolutionary psychological inferences with the pattern of reasoning behind a plausible, non-hypothetical evolutionary biological inference. I conclude with some reflections about the role of evolutionary thinking for psychology. Finally, it is important to note that the article is addressed to any attempt to make evolutionary inferences about extant psychological mechanisms. While I use the idiom of modularity, the arguments apply more broadly than to just the massive modularity hypothesis.
Core Commitments of Evolutionary Psychology
Explaining human behavior in terms of its biological roots is not new. Even Darwin considered human behavior to be the result of evolutionary processes, and he forecast that evolutionary theory would one day provide new foundations for psychology (Darwin  2003; Richards 2003). However, a substantive evolutionary approach to explaining human behavior only came to fruition during the 1970s, with the discipline of sociobiology. Sociobiologists championed evolutionary theory’s explanatory power to give origin explanations of the social lives of animals, including human beings. Those theorists believed that human behaviors have been subject to natural selection, and that they are inherited genetically. But worries about the degree to which sociobiology could accommodate facts about behavioral flexibility led to its being overtaken by the newer discipline of evolutionary psychology, which focused on the evolutionary functions of the proximate psychological mechanisms that bring behaviors about rather than on the behaviors themselves (Crawford and Krebs 2008).
Evolutionary psychologists argued that the human mind is composed of cognitive structures that were designed by natural selection for performing highly specialized tasks: predator avoidance, mate selection, cheater detection, and so forth. The claim that the human mind contains specialized structures was first advanced by Fodor (1983), who argued that only the peripheral input systems are modular, and that mental operations such as belief fixation, inference-making, and so on are not performed by modules, but are performed by domain-general central processing. In contrast, evolutionary psychologists claim that much of central processing is also modular. This is known as the massive modularity hypothesis.
The cognitive structures postulated by evolutionary psychologists are supposed to have been sensitive to certain reproductively significant features in the environment of evolutionary adaptedness (EEA), a term used to refer to the environment(s) where ancestral organisms faced pressures of living and evolved adaptations in response to them. An organism is said to be in the EEA as long as it remains in an environment in which those challenges that gave rise to its adaptations persist. In the case of human beings, our present anatomy and physiology were almost entirely in place by approximately 200,000 years ago (Conroy 2005), and the environments where human beings now live are, for almost all of us, importantly different from the environments where our prehistoric ancestors evolved. However, according to evolutionary psychologists, although we are no longer in the EEA, our evolved, modular psychology remains adapted to the conditions that obtained in the distant past, hence the slogan, “Our modern skulls house a stone age mind” (Cosmides and Tooby 1997, p. 90). Since, in their view, our psychology is constituted by mental structures that were present a very long time ago, and that have changed very little if at all since that time, contemporary human behaviors are best interpreted in accord with their counterparts in the EEA. The cognitive structures underpinning our behavior are “very sophisticated computers, whose circuits are elegantly designed to solve the kinds of problems our ancestors routinely faced,” and thus, “Behavior in the present is generated by information-processing modules that exist because they solved adaptive problems in the past” (Cosmides and Tooby 1997, p. 91).
Arguments that the human mind is populated by very many special-purpose cognitive structures that are residues of our evolutionary history are theory-driven. They derive their conclusions from claims about constraints on the evolution of complex mental states, or from poverty-of-the-stimulus reasoning. The most popular of these arguments concerns constraints on evolvability. As Cosmides puts it:
The more important the adaptive problem, the more intensely selection should have specialized and improved the performance of the mechanism for solving it (Darwin 1859/1958; Williams 1966). Thus, the realization that the human mind evolved to accomplish adaptive ends indicates that natural selection would have produced special-purpose, domain-specific, mental algorithms – including rules of inference – for solving important and recurrent adaptive problems (such as learning a language; Chomsky 1975, 1980). (Cosmides 1989, p. 193)
When considering claims about massive modularity, it is important to appreciate that each of the components is thought to be quite fine-grained. This is a consequence of the principle, expressed in the quote from Cosmides above, that mental mechanisms are responses to specific adaptive challenges. This is well illustrated by Cosmides and Tooby’s celebrated study of social reasoning, which they interpret as demonstrating that the human mind is equipped with a module specifically devoted to the logic of social exchange (Tooby and Cosmides 1992; Cosmides and Tooby 2005). This is not a module for monitoring social relations generally. It is concerned with a particular sector of social life that is distinct from other domains, such as mating behavior and parental investment, which are assigned their own proprietary mechanisms (Buss 2014).
Evolutionary psychologists believe that these mechanisms are universal—that they are possessed by almost every member of our species, and that they transcend cultural and historical boundaries. This is why evidence of psychological uniformity across cultures is taken as evidence that these psychological dispositions are adaptations. And for these psychological structures to come under selection, and to be conserved, they must be genetically transmissible. As such, they are said to be built into the structure of the brain, rather than programmed into it by learning. Some version of nativism about cognitive architecture is therefore critical for evolutionary psychology.
The Inferential Strategies
Evolutionary psychologists use two inferential strategies (Buss 2014). One is to begin by identifying a behavior exhibited by contemporary humans and then set about identifying its function on the basis of its current effects. The other is to identify the recurrent adaptive challenges that were faced by our prehistoric ancestors, to speculate that these challenges acted as selection pressures that gave rise to specific psychological adaptations that have been conserved, and finally to extrapolate from these conjectures to explain the functions of features of present-day human psychology. In this context, it is important to distinguish between having a function and performing a function. A biological item has a function only if it underwent selection on account of its fitness-enhancing effects, and a biological item performs a function only if it produces the effects that accounted for its having undergone selection (effects which might not be fitness-enhancing in their current environment). Evolutionary psychologists do not always require that cognitive structures perform the functions that they have, because changing environmental contingencies can produce effects that they were not selected for producing (see Crawford 1998). In these cases, they may attribute a function to a mental structure on the basis of conjectures about the behaviors it would have produced in the EEA.
Both of these inferential strategies depend on practitioners being able to establish that the cognitive structures that they believe to be hardwired in modern human brains are the very same structures that were selected for in the EEA. So, for their program to work, evolutionary psychologists need to have some method for inferring which prehistoric structures are identical to which contemporary ones. Unless there is some reliable procedure for accomplishing this, it is difficult to see how one can establish that the psychological architectures of contemporary human beings, and the behaviors that they produce, have selected-for functions.
The Matching Problem
Evolutionary psychologists argue that behaviors in the present are caused by cognitive systems that operate today as they did in the past. In their view, each module was selected for because of its specific fitness-enhancing effects in the EEA, and each of them is domain-specific—that is, responsive only to the kinds of inputs for which they are adaptations.
Evolutionary psychological inferences are secure only if it is possible to determine that particular kinds of behavior are underwritten by particular structures. Further, these must have the evolved function of producing behaviors of just these kinds. If present-day human behaviors are caused by special-purpose cognitive structures, and that was also true of our stone age ancestors, and if there is a high degree of concordance between the structures populating the modern mind and those that populated the minds of our prehistoric ancestors, this would still fall short of securing evolutionary psychological inferences. This is because it might be the case that the similarities between prehistoric and modern cognitive architectures are due to ontogenetic processes—similar experiences producing similar functional differentiation in the brain. For a contemporary trait to be a strong vertical homolog of an ancestral trait, the contemporary trait must be of the same kind as the ancestral one. It must also have the same function as the ancestral one, and must be related by descent to that ancestral trait as part of a continuous reproductive lineage extending back to the EEA. Additionally, it must be the case that the contemporary trait and the ancestral trait are of the same kind and have the same function because the present-day trait is descended from the ancestral trait. In principle, it might be that a present-day trait and an ancestral trait are of the same kind and have the same function without one being descended from the other. If this is the case, then the architecture of the minds of present-day humans would resemble that of early humans without it being the case that this architecture was selected for and genetically transmitted through the generations. If the idea that mental structure can be acquired ontogenetically seems dubious, consider the area of the brain called “the visual word-form area” that is specialized for reading (it is a “reading module”). Written language emerged only around 3500 years ago (Woods 2010), so it is too recent for reading to have been selected for. This shows that cognitive mechanisms can be acquired by learning (Dehaene 2009; Dehaene and Cohen 2011; Heyes 2018; see also Buller and Hardcastle 2000). So, the condition specifying that the sameness relation must depend upon descent is essential for evolutionary psychological explanations. It follows from this that evolutionary psychological claims fail unless practitioners can show that mental structures underpinning present-day behaviors are structures that evolved in the EEA for the performance of adaptive tasks that it is still their function to perform. This is the matching problem.
To match ancestral structures with present-day structures in the way that is needed, three conditions must be satisfied.
One must determine that the function of a contemporary module is one that an ancestral module was selected for performing.
One must determine that the contemporary module has the same function as the ancestral one in virtue of its being descended from the ancestral module.
One must determine which ancestral structures are related to which contemporary ones in this way.
The matching problem poses a methodological challenge to evolutionary psychology. There seems to be a tacit assumption amongst evolutionary psychologists that the needed identities between contemporary psychological mechanisms and ancestral ones are so obvious that they do not need to be established. But this is not the case. Unless the challenge can be overcome, evolutionary psychological explanations fail. Put more strongly, if the matching problem cannot be solved, evolutionary psychology is impossible.
Strong Vertical Homology
Evolutionary psychological explanations presuppose that contemporary psychological phenomena are related to ancestral psychological phenomena in a homology-like way. I use the term “homology-like” because the relations that are supposed to obtain between ancestral and contemporary mechanisms are not homologous in the technical sense of the word. “Homology” pertains to similarities across taxa because of common ancestry. For example, bird wings and human arms are homologous to the extent that their structural similarities are the result of common descent from reptile forelimbs. But the sort of similarity that evolutionary psychologists wish to establish is similarity due to descent of contemporary phenotypes from ancestral ones within a single taxon (Homo sapiens or genus Homo). Since homology is standardly understood as a “horizontal” relation (across taxa), I dub the sort of similarity that evolutionary psychologists are concerned with “vertical homology.” Because it is possible for a phenotype to be descended from another phenotype without its function being conserved, I use the term “strong vertical homology” for cases in which a phenotype has the same function as an ancestral phenotype as a consequence of being descended from it.
Proximate and Ultimate Explanations
When evolutionary psychologists make claims about strong vertical homologies, they are offering what are known as “ultimate” explanations of psychological phenomena. The distinction between proximate and ultimate explanations in biology was first set out by Mayr (1961; and elaborated by others (e.g., Tinbergen 1963; Ariew 2003)). Proximate explanations pertain to events occurring to individual organisms during their lifetimes, in contrast to ultimate explanations, which explain features of organisms with reference to their evolutionary histories. Ultimate explanations are population-level explanations, rather than individual-level ones. They specify why it was, in the past, that some phenotypic feature proliferated through a population on account of having more often than not enhanced the fitness of those individuals having that phenotype.
One gives a proximate explanation of an organism’s behavior by citing causes operating within that organism’s lifetime that make a difference to the occurrence of that behavior. And one gives an ultimate explanation of a contemporary organism’s behavior by citing past effects of behaviors (behaviors that occurred in the EEA) that made a difference to its proliferation. These are ultimate explanations of contemporary behaviors only if the ancestral behavior is a strong vertical homolog of the contemporary behavior.
Consider primatologists’ explanations of the alarm calls made by vervet monkeys. Alarm calls are made in response to the presence of predators, and are proximately caused by factors like perceptions of predators, the causal connection between the perceptions and vocalizations, learning to sound the alarm calls correctly, the developmental processes that underpin their ability to sound the calls, and so forth. A proximate explanation of vervet alarm calls might cite any or all of these factors. In contrast, an ultimate explanation of the vervet alarm calls begins with the assumption that the behavior has a function that it was selected for performing. The idea is that ancestral vervet monkeys that made and responded appropriately to alarm calls reproduced more successfully than those that did not, because they were better able to evade predation. This was because the calls correlated enough of the time with the presence of predators and produced appropriate avoidance behaviors in those vervet monkeys that heard them. An additional assumption is that the mechanisms that produced the behavior were passed on genetically to vervet monkey offspring and eventually proliferated through the entire population.
There is some controversy in cases of animals that produce alarm calls about how the process of selection occurred. There may have been a direct fitness benefit to the caller (in the case of Belding’s ground squirrels, those that sound the alarm are less likely to fall prey to aerial predators—but not to terrestrial predators—than those that do not (Sherman 1985)), or it may be that the behavior was selected because it enhanced the caller’s inclusive fitness by promoting the fitness of kin, or there may have been some other process such as group selection at work.
When biologists give ultimate explanations of nonhuman animal behavior it is generally the case that those animals are still living in, or have been living in until very recently, an environment that is similar to the environment where the behavior evolved. In such cases, it is trivially true that the behavior under consideration as well as the proximate mechanisms underpinning that behavior are identical to a corresponding behavior and the mechanisms that underpinned it in the EEA. We can see this in the case of vervet monkeys. Whatever the nature of the selection process, it is reasonable to infer that because producing and consuming alarm calls enhances vervet fitness in the present by helping them to avoid predation, it was selected for this effect in the past. In contrast, applying this principle in the human case is problematic, because the circumstances of contemporary human life are, in many important respects, quite different from those in which our species evolved. This makes extrapolation from the present to the past particularly challenging. To succeed, ultimate explanations of human behavior must draw on some other method for establishing the existence of strong vertical homologies.
Evolutionary psychologists use two procedures to arrive at purported ultimate explanations. One is to identify the proximate psychological causes of a behavior, and then offer ultimate explanations of those proximate psychological causes based on assumptions about the EEA.Footnote 1 The other procedure begins with a hypothesis about selection pressures that were encountered in the EEA. Then, assuming that these pressures resulted in psychological adaptations, one posits the existence of hardwired, cognitive structures with the function of responding to those pressures. Finally, one proposes that this ancestral psychology has been conserved with its function in place, and that it is therefore strongly vertically homologous with the psychology of contemporary human beings.
Neither of these procedures addresses the matching problem. Both assume that present-day psychological structures are strongly homologous with ancestral ones. But this assumption is precisely what the procedures are supposed to demonstrate. The framework is such that evolutionary psychologists already presuppose that our present-day cognitive architecture was laid down in prehistory, and conclude from this that we have stone age minds in modern skulls.
Individuating Prehistoric Cognitive Mechanisms
We are all, of necessity, methodological behaviorists, because we can only infer psychological mechanisms from observations of behaviors. Even inferences about psychological mechanisms using procedures such as brain imaging or measurements of involuntary physiological indicators need ultimately to be based on correlations with behavior. So, knowledge of the mental mechanisms possessed by contemporary humans can only be acquired by making inferences from behaviors to the underlying mechanisms. Similarly, knowledge of the mechanisms that populated the minds of our prehistoric ancestors could only be gained by making inferences from their behavior to the psychological mechanisms. If it could be established in this way that the structure of the modern mind maps onto the structure of the Stone Age mind, then it might be possible to determine that they are strong vertical homologs. But this cannot be established, because we are not epistemically situated to make the right sort of inferences about the minds of our prehistoric ancestors. We have only fragmentary knowledge of the behavior of early humans, and what we do know is very general (they avoided predators, they reproduced, they hunted and gathered, they made tools, etc.). And, as I argued above, to infer prehistoric mental structures from their presence in modern-day humans assumes what needs to be shown.
Cognitive Mechanisms Cannot be Individuated by Adaptive Challenges
Some evolutionary psychologists would dispute the claim that we cannot know anything much about the cognitive architecture of prehistoric hominins. They argue that, although we do not have access to these individuals’ minds, we can “read off” ancestral mechanisms from the adaptive challenges that they faced (e.g., Buss 1995). The claim is that, for example, if predator evasion was an adaptive challenge due to the existence of big cats that preyed on humans, then natural selection would have produced a predator-evasion module, because humans did evade predators. However, this inferential strategy would work only if it were the case that all mental traits are adaptations, that adaptationist explanations are difficult to come by, and that adaptations are easily characterized (Gray et al. 2003). But there is no reason to assume that all mental traits are adaptations, just as there is no reason to think that all traits are adaptations (Lloyd 1999). We also know that adaptationist hypotheses are easily given (Gould and Lewontin 1979). Finally, the problem of how to characterize traits concerns what Sterelny and Griffiths (1999) call the “grain problem.” That is, any adaptive problem characterized in a coarse-grained way (for example, “predator evasion”) can be decomposed into a collection of finer-grained problems, which can themselves be decomposed into even finer-grained problems. As Sterelny and Griffiths remark, regarding a hypothesized module for mate choice:
The grain problem in evolutionary psychology challenges the idea that adaptations are explained by the problem to which the adapted trait is a solution. If (but only if) there is a single cognitive device that drives an organism’s behavior with respect to issues of mate choice, then mate choice is a single domain, and these are all different aspects of the same problem. It is not the existence of a single problem confronting the organism that explains the module, but the existence of the module that explains why we think of mate choice as a single problem. (Sterelny and Griffiths 1999, pp. 328–329)
Sterelny and Griffiths see this as a problem for evolutionary psychology because it introduces indeterminacy and arbitrariness into characterizations of mental structures. But it also raises an additional difficulty for resolving the matching problem, because if there is no fact of the matter about how psychological mechanisms are to be described, then there is no fact of the matter about how they are to be matched.
As I have shown, there are methodological obstacles to establishing that contemporary behaviors are produced by evolved cognitive structures. But even if these could somehow be overcome, there would still be obstacles to identifying the structures with particular kinds of behaviors that are hypothesized to have existed in the EEA. I mentioned above that mental structures can be individuated only by the behaviors that they produce, so it follows that their individuation depends upon the individuation of behaviors.
We ordinarily individuate behaviors by attributing intentions to agents. For example, we describe a certain behavior as “going to the kitchen” rather than “walking in the direction of Paris.” We do this because we have grounds for thinking that the former is how the agent represented her intention to herself. Evolutionary psychologists cannot use this method for individuating behaviors, because they offer subpersonal explanations of behavior (Heyes 2018) that are framed in terms of underlying computational mechanisms. Individuating a behavior by the intentions that produced it does not allow one to infer anything about the subpersonal mechanisms that produced that behavior. So, evolutionary psychologists need to have some other way of individuating behaviors that they can use to make inferences about the cognitive architecture of both contemporary and ancestral humans in order to establish the existence of strong vertical homologies.
There are three ways for doing this. One way is to individuate behaviors by their effects. Another way is to individuate them by their functions. And a third way is to individuate them by their causes. I will address each of these and show that, because of the explanatory demands of evolutionary psychology, none of these options are available to practitioners.
Let us say that a behavior is the same kind as an ancestral behavior because both of them produce the same effects. To do this, one would need to be situated in the right way epistemically with regard to the effects of ancestral behavior, but we are not. Even if this epistemic obstacle with respect to effects could be surmounted, difficulties remain. In order for the behaviors to count as the same, one needs to establish that the contemporary and ancestral causes of the behavior (structures) are strong vertical homologs. Evolutionary psychology does not have a procedure for surmounting this additional, and perhaps even more problematic, obstacle. Supposing that ancestral behaviors could be matched with contemporary ones on the basis of their effects, this would not license the inference that they have a common psychological cause.
The second option is to individuate behaviors by their selected-for functions. The function of a phenotypic trait is the effect of that trait on fitness in a critical mass of cases in the EEA. A present-day behavior is functionally identical with an ancestral behavior if the two behaviors share the same function. However, the functional criterion fails because of a problem of circularity. To individuate behaviors by their selected-for functions is the same as giving ultimate explanations of them. If one begins with the assumption that a contemporary behavior has a function, in the relevant sense of “function,” one has already assumed that this behavior has an ultimate explanation. Simply put, individuating behaviors on the basis of their functions illegitimately supposes that a behavior was selected for and then uses this supposition as evidence that the behavior was selected for.
Finally, one might individuate behaviors by their causes. This seems to be the option that evolutionary psychologists should pursue, because they need to be able to infer underlying psychology from behavioral effects in order to give ultimate explanations of present-day psychology. On this option, if two behaviors have the same proximate psychological causes then they belong to the same behavioral kind. Consider the claim that a present-day behavior is the same kind as a behavior in the EEA. If behaviors are individuated by their causes, then the contemporary behavior that one wishes to explain, and the behavior in the EEA by means of which one wishes to explain it, must have the same kind of causes. But this strategy fails to be explanatory, because it is circular; it relies on the principle that cognitive mechanisms are individuated by the behaviors that they bring about while these behaviors are individuated by the mechanisms that supposedly cause them.
Confounding Effects of Environments and Developmental Variation
I have already pointed out that if an organism is still in the EEA, or is in an environment that is relevantly similar to the EEA, it may be relatively straightforward to offer ultimate explanations of its behavior and psychology. That is because as long as the environmental contexts remain constant within certain parameters, the selected-for effects of a trait are likely to remain steady. But granting that, the effects of biological causes may also vary with varying environmental contingencies. Because natural selection operates on fitness-enhancing phenotypes, psychological structures can only be selected for if they make a difference to organisms’ reproductive success. This happens by producing behaviors that promote fitness in a certain sort of environment. A cognitive device operating outside the EEA might produce behaviors that differ significantly from the behaviors that it produced in the EEA.Footnote 2 And because of the possible confounding effects of environmental changes, a form of behavior that enhances fitness now need not have enhanced fitness in the past. Accordingly, it is illegitimate to infer from a behavior that contributes to reproductive success now, that it contributed to reproductive success in the ancestral past, and it is therefore illegitimate to infer that the cognitive device that produces the behavior now was selected for in the EEA.
Subtle environmental variations can lead to phenotypic differences (West-Eberhard 2003). For example, water fleas (Daphnia cucullata) that develop in an environment where there are chemicals indicating the presence of predators undergo dramatic morphological changes. They develop impressive, helmet-like structures on their necks and spines along their tails, and these effects persist across generations (Agrawal et al. 1999; Tolliran and Dodson 1999). Environmentally induced developmental changes affecting behavior have been observed in a number of species. For example, Bonnet macaques that have been raised in conditions where provisioning is inconsistent are more timid than those raised in environments where the macaques either do not ever have to forage extensively or always have to forage extensively. The infant monkeys raised in environments with variable foraging demands also show signs of depression of the sort normally observed only in maternally deprived primates, and as adolescents they are less inclined to engage in social play behavior. Studies show that inconsistent provisioning alters the development of the monkeys’ neural systems that mediate responses to stress (Coplan et al. 1996, 1998).
These examples illustrate that environmental variations can consequentially affect the expression of the underlying mechanisms, including the behavioral effects of psychological mechanisms. This is to be expected. Organisms are malleable. When we say that some feature of an organism was selected for, this is really a shorthand expression for a complex causal story involving genetics, developmental pathways, and chemical signaling systems. All of these processes are sensitive to environmentally contingent factors to a greater or lesser degree.
The impact of environments on development casts doubt on the claim that if a certain behavior is reliably produced by a certain causal mechanism in a present-day environment that is far removed (or perhaps even subtly removed) from the circumstances in which our prehistoric ancestors lived, then the presence of that same behavior (however individuated) justifies the inference that the behavior was produced by the same cause in the EEA.
I turn now to an example to illustrate how evolutionary psychology fails to address the matching problem. The study, entitled “Sex Differences in Perceptions of Infidelity: Men Often Assume the Worst” (Goetz and Causey 2009), appeared in a well-regarded peer-reviewed journal. The methodological deficits that it illustrates are typical of evolutionary psychological literature.
The authors begin with a claim about conditions obtaining in the EEA. They say that it would have been more biologically costly for men than for women to fail to detect a partner’s infidelity. They write:
Ancestral men…were susceptible to an additional and profound cost if they failed to detect a partner’s infidelity: cuckoldry—the unwitting investment of resources into genetically unrelated offspring. Cuckoldry was one of the most serious threats to fitness our male ancestors faced. Some of the costs associated with cuckoldry include misdirection of the male’s time, effort, and recourses to rearing a rival’s offspring, loss of time, effort, and resources the man spent attracting his partner, and reputational damage if such information becomes known to others. (Goetz and Causey 2009, p. 255)
The claims about the behavior of prehistoric humans go beyond paleoanthropological evidence. But setting this aside, even if these claims were well-founded, the inference made from them would still be unwarranted. Suppose that these claims about early humans are true. The authors infer from this that the costs associated with paternal uncertainty “provided selection pressure for an arsenal of anti-cuckoldry tactics in men” and that one of these may have been “evolved psychological mechanisms designed to overperceive the likelihood of their partner’s infidelity.” They further infer that contemporary men should be more suspicious of their partner’s future infidelity than contemporary women are, claiming:
Due to the costs associated with being cuckolded, men’s infidelity detection system may have been designed to overestimate the likelihood of their partner’s future infidelity. This overestimation bias would have generated behavior aimed at preventing infidelity, such as increased vigilance, mate guarding, and even affectionate behavior. (Goetz and Causey 2009, p. 258)
The study designed to test this hypothesis proceeded as follows. One hundred sixty-three male and female college students were asked four questions: (1) “How likely do you think it is that you will in the future have sexual intercourse with someone other than your current partner?” and (2) “Please indicate your agreement or disagreement with the following statement: ‘I will probably be sexually unfaithful to my partner.’” They were also asked (3) “How likely do you think it is that your current partner will in the future have sexual intercourse with someone other than you, while in a relationship with you?” and (4) “Please indicate your agreement or disagreement with the following statement: ‘My partner will probably be sexually unfaithful to me in the future.’” The authors concluded:
Using two independent samples, two different response formats, and two data collection methods, we found support for our hypothesis. Men’s perceptions of the likelihood of their partner’s future infidelities were greater than women’s. In conclusion, we found support for the hypothesis that men’s infidelity detection system should be designed to overestimate the likelihood of their partner’s infidelity. (Goetz and Causey 2009, p. 262)
Now, consider the inferential structure of this study. The authors begin with a theory-driven conjecture that in the EEA mate infidelity was costlier for males than it was for females. They also claim that those males who were good at safeguarding against their mates’ infidelity would have been more fit than those who were not able to guard against it as effectively. From this they conclude that selection favored a tendency for males to be more suspicious of their mates than females are. These ancestral behaviors (increased vigilance, mate guarding, and even affectionate behavior), they claim, are then ultimately explained as having the function of preventing their mates from conceiving offspring with other men. The behaviors were, they hypothesize, mediated by selected-for cognitive mechanisms.
Faced with evidence that 21st-century male American college students are more doubtful of the future sexual fidelity of their mates than their female counterparts are, the authors make two assumptions. The first is that male college students’ skepticism is caused by a hardwired, domain-specific cognitive module. The second is that this module existed in Pleistocene males, and produced behaviors of the same sort as the contemporary behavior. These assumptions are supposed to underwrite the conclusion that the sexual suspiciousness of contemporary males is caused by a cognitive mechanism with the evolutionary function of enhancing their fitness by preventing cuckoldry.
Goetz and Causey do not provide support for their claim that the psychological mechanism driving contemporary male sexual skepticism is a strong vertical homolog of the mechanism that drove (hypothesized) prehistoric anti-cuckoldry behavior. Consequently, their inferences about the evolutionary origin of the male students’ attitudes is unjustified. The authors also make assumptions about the computational structure of the ancestral module—namely, that it produced skepticism about the fidelity of mates. But this is not entailed by the hypothesis that there was selection in the EEA for a module with the function of guarding against female infidelity. Even granting that there is good evidence that the sexual skepticism of contemporary males is underwritten by a domain-specific module, this would not show that the contemporary module is a strong vertical homolog of an ancestral module with the function of preventing infidelity.
Contrasting Evolutionary Inferences
It might be thought that I am unfairly holding evolutionary psychology to a much higher epistemic and methodological standard than is normally used in evolutionary biology. But this is not the case. Evolutionary psychological inferences commonly fail to satisfy reasonable epistemic criteria.
When making evolutionary inferences about paradigmatically biological traits, biologists use experimental manipulations, comparative methods, the fossil record, and optimality models to determine that selection has taken place and that the items under consideration have retained their selected-for functions (Orzack and Sober 1994a, b, 1996, 2001; Kaplan 2002; Sober 2008).
Evolutionary psychologists are impeded by the fact that these methods are unavailable to them. Experimental manipulations of the sort used when studying other organisms are, in our case, usually ethically unacceptable or technically unachievable (Kaplan 2002). Comparative methods are not reliably informative, as there are no extant species that are closely related to Homo sapiens (we are separated from our nearest relatives, chimpanzees and bonobos, by around six million years of evolution) and the relevant behaviors are not generally highly conserved. The fossil record is also unproductive, as mental processes leave no unambiguous material evidence, and optimality modeling is problematic because the underdetermination of behavior by mental structures makes it difficult or impossible to apply an optimality calculus to the hypothesized structures.
Furthermore, evolutionary psychological hypotheses turn on inferences about hypothetical cognitive structures—the mental modules—for which there is a dearth of empirical support (see, for example, Samuels 1998, 2000; Buller and Hardcastle 2000; Currie and Sterelny 2000; Fodor 2000; Sterelny 2003; Buller 2005; Prinz 2006; Schultz 2008; Robbins 2009, 2013; Heyes 2014), and there is no evidence that the minds of our prehistoric ancestors possessed this sort of cognitive architecture.
To appreciate the differences between good evolutionary biological inferences and the inferences made in evolutionary psychological studies such as the one described in the previous section, consider a highly plausible evolutionary account—the claim that the eye-blink reflex (corneal reflex) was selected for as a mechanism to protect the eye from injury (Hall 1945). There are several converging lines of evidence that give substance to this claim. The first is that the reflex is highly conserved, as it is found in all mammalian taxa, and even in other taxa such as avians. Thus, comparative methods suggest that the reflex is under genetic control and that it was retained in the lineage because of its function. Second, it is clear that the eye-blink reflex protects the eye from injury in the taxa where it is found. Third, it is clear that the fitness of organisms that rely on vision would be impeded if there were not some mechanism for protecting the eye from injury. Fourth, the physiological mechanism underpinning the reflex is well understood. Fifth and finally, the reflex operates automatically, and is therefore mandatory. It is not “up to” the organism whose reflex it is. These five factors, taken together, support the claim that the eye-blink reflex in contemporary humans is strongly vertically homologous to the eye-blink reflex in earlier members of the lineage.
The study by Goetz and Causey, in common with other evolutionary psychological studies, meets neither these nor comparable evidential criteria. Comparative methods are not available for assessing the claim that male skepticism about female fidelity was selected for. Furthermore, it is not obvious that this attitude leads to fitness-enhancing behavior in the present, and that the fitness of human males would be diminished in its absence. Also, it has not been established that prehistoric mating involved exclusive pair-bonding rather than, for example, promiscuity within the group. There is no good evidence that males’ skepticism about their female partners’ fidelity is under genetic control, and there is no dedicated neurological mechanism known to mediate it (the module controlling the attitude is entirely hypothetical). Finally, that this is not an automatic response is evidenced by successful polyamorous relationships.
The central claim of evolutionary psychology is that human psychology has been conserved since the Stone Age. Genes, developmental mechanisms, anatomical structures, physiological processes, and behaviors are said to be conserved if they persist in a lineage. If it is true that that there is some sense in which human psychology has been conserved since the Pleistocene, this might be understood in several ways. One way is as a claim that the causal mechanisms that gave rise to behavior in the EEA have been conserved without the behavior that they produced having been conserved. Alternatively, it might be meant as a claim that behaviors prevalent in the EEA have been conserved without their causal mechanisms having been conserved. Or it might be meant as a claim that both causal mechanisms and behaviors have been conserved. To say that a behavior has been conserved is only to say that the behavior is “the same” as some ancestral behavior by some criterion or other. This weak sense of “conservation” cannot support evolutionary psychological claims, because it does not address the underlying psychology. So, for their project to succeed, the causal mechanisms underpinning the behavior need to have been conserved. In order to justify this commitment, evolutionary psychologists need an individuation criterion that allows them to infer sameness of psychological cause from sameness of behavioral effects, both now and in the EEA. In other words, the pattern of reasoning that would allow one to offer ultimate explanations for the psychological causes of behaviors depends on both psychological causes and behavioral effects being conserved.
Evolutionary psychologists’ claims about ancestral cognitive mechanisms are formulated on a theoretical supposition that recurring adaptive challenges were likely to give rise to mental adaptations in the EEA, and that these adaptations underpin contemporary behavior. If one could in fact establish that the minds of our ancestors possessed mechanisms that were responsible for certain sorts of behavior, this would still not secure evolutionary psychological claims. Knowing, for example, that ancestral females possessed a module that was sensitive to whatever attributes of potential mates were correlated with reproductive value would tell us nothing about inputs to which the module was tuned or the behaviors that bought it about. So, we would be unable to extrapolate from it to draw conclusions about the psychological mechanisms that regulate the mating behavior of contemporary women. And knowing that the mating behavior of contemporary women is regulated by a module would not provide a basis for inferring that this module is the same one that regulated the mating behavior of ancestral females.
These methodological problems prompt the question, “Is evolutionary psychology possible?” It is important to distinguish evolutionary psychological explanations of human behavior from evolutionary explanations of human behavior simpliciter. This is particularly important given that evolutionary psychologists often claim that those who reject evolutionary psychology but accept evolutionary theory are committed to a contradiction.Footnote 3 However, evolutionary theory does not entail nativism or massive modularity. One might reject the theoretical apparatus proposed by evolutionary psychologists while still embracing an evolutionary account of the human mind. As Lloyd remarks:
It is one of the most pernicious aspects of the present climate of discussion, that the situation is often set up as a forced choice between accepting the particular theories and oversimplified principles of evolutionary psychology, or retreating to a pre-Darwinian denial of the fact that we are evolved animals. (Lloyd 1999, p. 213)
Buller (2005) addresses this problem by distinguishing “evolutionary psychology” from “Evolutionary Psychology.”
The former is a field of inquiry, a loose confederation of research programs that vary widely in theoretical and methodological commitments and that are federated only by a commitment to ‘‘adopting an evolutionary perspective on human behavior and psychology’’ (Barrett et al. 2002, p. 1). The latter, Evolutionary Psychology, is a specific doctrinaire research program within this field of inquiry, a central doctrine of which is the so-called massive modularity hypothesis (MMH). (Buller 2005, p. 881)
If the arguments that I have presented in this article are sound, the methodological defects of Evolutionary Psychology are so profound as to be unrectifiable, and consequently Evolutionary Psychology is not possible. Evolutionary psychologists simply do not have the methodological resources to justify the claim that the psychological causes of contemporary behaviors are strong vertical homologs of the psychological causes of corresponding behaviors in the EEA. The result for evolutionary psychology (as contrasted with Evolutionary Psychology) is less clear. No one should contest that the human mind is a product of evolution, and evolution must therefore enter into an explanation of human psychology in some way.
Evolutionary Psychology rests on three pillars: the massive modularity hypothesis, the claim that modular structures evolved as adaptations to recurrent challenges in the EEA, and the tacit assumption that mental structures can be individuated and so license claims about strong vertical homologies. These three components, taken together, are inconsistent with the competing hypothesis that evolution fashioned the human mind as a domain-general or modestly modular learning system. On this view, the architecture of the human mind (whatever it turns out to be) was selected to be adaptive and malleable, rather than fixed and instinct-like. It supports a view of human nature that is less nativist than the version that is proffered by Evolutionary Psychologists. Importantly, this competing hypothesis is immune from the criticisms that I have developed here, chiefly because it does not reduce the mind to an array of domain-specific systems and require that these be homologs of ancestral systems. It is also well-supported by recent work in developmental biology indicating that it is the case that the behavioral repertoires (even of invertebrates) are often highly malleable and driven by learning (West-Eberhart 2003; Menzel and Benjamin 2013). However, a research program of this sort—one that restricts itself to scientifically justifiable claims about the phylogenetic origins of human psychology, and that gives developmental plasticity and learning their due—is unlikely to have much utility for explaining the specifics of human behavior biologically. It is likely to be less contentful than Evolutionary Psychology precisely because it makes no attempt to extend fine-grained adaptationist explanations to domains where they cannot be applied.
Because psychology can only be fitness-enhancing by producing fitness-enhancing behavioral effects, evolutionary psychologists must assume that ancestral psychology produced behaviors of the same sort as contemporary behaviors.
As mentioned earlier, evolutionary psychologists are often sensitive to this point (e.g., Crawford 1998) but do not recognize the problem that it poses for individuating modules.
A prime example is Anne Campbell’s frequently quoted remark that such people believe that “evolution stops at the neck” (Campbell 2002, p. 13).
Agrawal AA, Laforsch C, Tollrian R (1999) Transgenerational induction of defenses in animals and plants. Nature 401:60–63
Ariew A (2003) Ernst Mayr’s ‘proximate/ultimate’ distinction reconsidered and reconstructed. Biol Philos 18:553–565
Barrett L, Dunbar R, Lycett J (2002) Human evolutionary psychology. Princeton University Press, Princeton
Buller DJ (2005) Adapting minds: evolutionary psychology and the persistent quest for human nature. MIT Press, Cambridge
Buller DJ, Hardcastle VG (2000) Evolutionary psychology, meet developmental neurobiology: against promiscuous modularity. Brain Mind 1:307–325
Buss DM (1995) Evolutionary psychology: a new paradigm for psychological science. Psychol Inq 6:1–30
Buss DM (2014) Evolutionary psychology: the new science of the mind. Pearson, Boston
Campbell A (2002) A mind of her own: the evolutionary psychology of women. Oxford University Press, Oxford
Chomsky N (1975) Reflections on language. Random House, New York
Chomsky N (1980) Rules and representations. Columbia University Press, New York
Conroy GC (2005) Reconstructing human origins. Norton, New York
Coplan JD, Andrews MW, Rosenblum LA, Owens MJ, Friedman S, Gorman JM, Nemeroff CB (1996) Persistent elevations of cerebrospinal fluid concentrations of corticotropin-releasing factor in adult nonhuman primates exposed to early-life stressors: implications for the pathophysiology of mood and anxiety disorders. Proc Natl Acad Sci USA 93:1619–1623
Coplan JD, Teost RC, Owens MJ, Cooper TB, Gorman JM, Nemeroff CB, Rosenblum LA (1998) Cerebrospinal fluid concentrations of somatostatin and biogenic amines in grown primates reared by mothers exposed to manipulated foraging conditions. Arch Gen Psychiatry 55:473–477
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 (1997) Evolutionary psychology: a primer. In: Downes S, Machery E (eds) Arguing about human nature: contemporary debates (2013). Routledge, New York, pp 83–92
Cosmides L, Tooby J (2005) Neurocognitive adaptations designed for social exchange. In: Buss D (ed) The handbook of evolutionary psychology. Wiley, Hoboken, pp 3–87
Crawford C (1998) The theory of evolution in the study of human behavior: an introduction and overview. In: Crawford C, Krebs D (eds) Handbook of evolutionary psychology: ideas, issues, applications. Lawrence Erlbaum Associates, Mahwah, pp 3–42
Crawford C, Krebs D (2008) Foundations of evolutionary psychology. Lawrence Erlbaum Associates, New York
Currie G, Sterelny K (2000) How to think about the modularity of mind-reading. Philos Q 50:145–160
Darwin C (2003) On the origin of species. Signet Classics, New York
Dehaene S (2009) Reading in the brain. Viking, New York
Dehaene S, Cohen L (2011) The unique role of the visual word form area in reading. Trends Cognit Sci 15:254–262
Fodor JA (1983) Modularity of mind: an essay on faculty psychology. MIT Press, Cambridge
Fodor JA (2000) The mind doesn’t work that way: the scope and limits of computational psychology. MIT Press, Cambridge
Goetz D, Causey K (2009) Sex differences in perceptions of infidelity: men often assume the worst. Evol Psychol 7:253–263
Gould SJ, Lewontin RC (1979) The spandrels of San Marco and the Panglossian paradigm: a critique of the adaptationist programme. Proc R Soc B 205:581–598
Gray RD, Heaney M, Fairhall S (2003) Evolutionary psychology and the challenge of adaptive explanation. In: Sterelny K, Fitness J (eds) From mating to mentality: evaluating evolutionary psychology. Psychology Press, New York, pp 247–268
Hall A (1945) The origin and purposes of blinking. Br J Ophthalmol 29:445–467
Heyes CM (2014) False belief in infancy: a fresh look. Developmental Science 17:647–659
Heyes CM (2018) Cognitive gadgets: the cultural evolution of thinking. Harvard University Press, Cambridge
Kaplan JM (2002) Historical evidence and human adaptations. Philos Sci 69:294–304
Lloyd EA (1999) Evolutionary psychology: the burden of proof. Biol Philos 14:211–233
Mayr E (1961) Cause and effect in biology. Science 131:1501–1506
Menzel R, Benjamin PR (2013) Invertebrate learning and memory. Handbook of behavioral neuroscience, vol 22. Elsevier, New York
Orzack SH, Sober E (1994a) Optimality models and the test of adaptationism. Am Nat 143:361–380
Orzack SH, Sober E (1994b) How (not) to test an optimality model. Trends Ecol Evol 9:265–267
Orzack SH, Sober E (1996) How to formulate and test adaptationism. Am Nat 148:202–210
Orzack SH, Sober E (2001) Adaptationism and optimality. Cambridge University Press, New York
Prinz JJ (2006) Is the mind really modular? In: Stainton R (ed) Contemporary debates in cognitive science. Blackwell, Oxford, pp 22–36
Richards RJ (2003) Darwin on mind, morals, and emotion. In: Hodge J, Radick G (eds) The Cambridge companion to Darwin. Cambridge University Press, Cambridge, pp 96–119
Robbins P (2009) Modularity of mind. In: Zalta EN (ed) The Stanford encyclopedia of philosophy (summer 2010 edition). http://plato.stanford.edu/archives/sum2010/entries/modularity-mind/. Accessed 1 May 2019
Robbins P (2013) Modularity and mental architecture. Wiley Rev Cognit Sci 4:641–649
Samuels R (1998) Evolutionary psychology and the massive modularity hypothesis. Br J Philos Sci 49:575–602
Samuels R (2000) Massively modular minds: evolutionary psychology and cognitive architecture. In: Carruthers P, Chamberlain A (eds) Evolution and the human mind. Cambridge University Press, New York, pp 13–46
Schultz AW (2008) Structural flaws: massive modularity and the argument from design. Br J Philos Sci 59:733–743
Sherman PW (1985) Alarm calls of Belding’s ground squirrels to aerial predators: nepotism or self-preservation? Behav Ecol Sociobiol 17:313–323
Sober E (2008) Evidence and evolution: the logic behind the science. Cambridge University Press, Cambridge
Sterelny K (2003) Thought in a hostile world: the evolution of human cognition. Blackwell, Oxford
Sterelny K, Griffiths PE (1999) Sex and death: an introduction to philosophy of biology. University of Chicago Press, Chicago
Tinbergen N (1963) On aims and methods of ethology. Z Tierpsychol 20:410–433
Tolliran R, Dodson SI (1999) Inducible defenses in Cladocera: constraints, costs and multipredator environments. In: Tolliran R, Harvell CD (eds) The ecology and evolution of inducible defenses. Princeton University Press, Princeton, pp 177–202
Tooby J, Cosmides L (1992) Cognitive adaptations for social exchange. In: Barkow JH, Cosmides L, Tooby J (eds) The adapted mind: evolutionary psychology and the generation of culture. Oxford University Press, New York, pp 163–228
West-Eberhard MJ (2003) Developmental plasticity and evolution. Oxford University Press, New York
Williams GC (1966) Adaptation and natural selection. Princeton University Press, Princeton
Woods C (2010) Visible language: inventions of writing in the ancient middle east and beyond. The Oriental Institute, Chicago
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
About this article
Cite this article
Smith, S.E. Is Evolutionary Psychology Possible?. Biol Theory 15, 39–49 (2020). https://doi.org/10.1007/s13752-019-00336-4
- Cognitive structures
- Evolutionary psychology
- Matching problem