?

Log in

No account? Create an account

Wed, Sep. 26th, 2007, 01:34 am
Flossing a dead horse: Some critiques to the “evolutionary psychology” research programme

 
The evolutionary psychology (herein refered to as “EP”) research programme was proposed by many psychologists – in special Leda Cosmides and John Tooby [cf. ref. 1] – as an “adaptationist” programme to study the evolution of psychological traits. This research programme spurred a lot of attention among scientists and non-scientists alike, and the attention it received from the media ended up creating an enormous unrest in the scientific community. On the one hand, psychologists tried to demonstrate that evolutionary psychologists were using evolutionary theory to attempt to eliminate competing theories within psychology without regard to evolutionary biology. According to them, the experiments led by evolutionary psychologists support a non-evolutionary psychological theory as strongly as an evolutionary one. On the other hand, evolutionary biologists shun the “adaptationist programme” as something that “regards natural selection as so powerful and the constraints upon it so few that direct production of adaptation through its operation becomes the primary cause of nearly all organic form, function and behaviour” [2]. In the middle, developmental biologists complain that EP has an almost preformationist vision of the mind [3,4], ignoring the epigenetic mechanisms that are, in fact, the selectable units of morphology that mediate the interactions between genotype and phenotype via evolutionary developmental mechanisms [5]. On this post, I will floss the dead horse and classify the body of critique to EP in 4 groups: evolutionary biology issues, neuroscience issues, developmental issues, and general epistemological issues.
 
A primer in evolutionary psychology
[EP] seeks to explain the evolved, functional characteristics of the human mind and thus their past and/or present adaptive value. This objective has been typically examined through the lens of an explanatory framework where special-purpose brain 'modules' are postulated to have been critical for hominid survival and reproductive success (...)[6]
 
Evolutionary psychologists have long advanced the view that the structure of human cognition reflects the ecological and environmental problems early humans confronted to their adaptive environment. This intimate relation between cognition and selective pressures has beeen thought to dictate an innately modular mental architecture. [4]
 
[T]he human psychological architecture contains many evolved mechanisms that are specialized for solving evolutionary long-enduring adaptive problems and (...) these mechanisms have content-specialized representational formats, procedures, cues, and so on [1]
 
The central tenets of EP are described in the above quotes. The adaptationist programme proposed by EP researchers states that the mind can be divided in modules, and those modules are specialized for solving adaptative problems that our Homo sapiens ancestores faced in the Pleistocene – what will from now on be called the “environment of evolutionary adaptedness” hypothesis, or EEA. The concept of module, here, is significantly different from the concept delineated in evolutionary biology (cf. refs. 7 and 8 for the EvoBiol definition of a module), and was derived from Fodor's [9] definition. Fodor delineated five criteria for a given psychological function to be modular; a module, according to his (strong ontological) definition:
1)      Present domain-specificity, ie, operates on, and present a computational architecture unique to, certain stimuli;
2)      Are linked to processing systems innately specified;
3)      Are defined by proprietary transduction systems;
4)      Cannot be affected by beliefs or be prevented to operate;
5)      Are informationally “encapsulated”, which means that they do not need to refer to external informations and/or other modules in order for its task to be completed;
6)      As a consequence, are fast-operating, and;
7)      Present “shallow”, i.e., non-conceptual, outputs.
8)      Are associated with particular neural structures; and
9)      Develop in accordance to a well-defined sequence.
These strong ontological prerequisites were thought by Fodor to apply to sensory information processing modules, which he believed to be the most low-level processing units of the mind. EP, however, proposed that the mind is “massively modular” [1,10,11]. In order to justify this hypothesis, EP theorists proposed another, ontologically weak, definition of module, in which the only necessary criterion to evaluate some function's status as a module is Fodor's fifth criterion – i.e., modules are informationally encapsulated functions. In this sense, a module is a dissociable functional component, a mental function that is relatively independent from other functions. It must be noted that, from the epistemological point of view, there is no necessary logical connection between any of Fodor's criteria; it does not follow, logically, from the fact that a function is “informationally encapsulated” that is is innately specified, or that it is associated with particular neural structures. That is why Fodor proposed nine criteria, and not only one.
The proposition made by EP is based on two premises: 1) there are domain-specific processes (modules); and 2) these processes are based on evolution. Cosmides [12], citing Darwin, Williams and Chomsky, states that
[t]he more important the adaptive problem, the more intensely selection should have specialized and improved the performance of the mechanism for solving it (...). 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 (...)).
She continues: it is clear that, while some mechanisms in the mind “are surely domain-general, these could not have produced fit behavior under Pleistocene era conditions (and therefore could not have been selected for) unless they were embedded in a constellation of specialized, content-dependent mechanisms”. This would surely be the case if the cognitive functions in question were fixed, under strong selection pressure, and have only one solution – which is not necessarily the case, one must add. As can be deprehended, the second premise (evolutionary origin of a module) is seen as derived from the first (existence of a module), but that is not the case; Cosmides uses the second premise as a claim “that evolutionary theory allows for or predicts her favored type of mechanism to discard or eliminate obvious competing hypotheses” [13]. This epistemological objection to EP will be developed further in its own section.
 
Evolutionary biologists vs evolutionary psychologists
Naturalistic models of cognition and other features of biological form must be consistent with the historical emergence of these features over time. Historical explanations that depend on the presence of unique sets of conditions presented in the correct sequences are not less satisfactory than explanations using general laws that apply across a wide range of initial conditions. In fact, the nature of biological systems provides reasons to expect historical explanations to be prevalent. Biological species are historical lineages capable of unlimited evolutionary change, not natural kinds of organisms, and so do not feature in traditional, universal laws of nature (...). The best candidates for traditional laws in evolutionary theory are ecological generalizations in which species and populations figure only as instances of ecological kinds, such as “primary producer” or “current occupant of patch” (...). But the output of processes governed by these ecological laws and generalizations is typically sensitive to the initial conditions of the process, so the resultant explanations are likely to be historical in nature (...). [14]
As the quote opening this section suggests, one important set of objections to EP comes from evolutionary biology. In short, the main objection is this: evolutionary function is an demanding epistemological endeavour, and its application requires considerable research. Whenever EP refers to EEA as an explanatory device, it is violating basic evolutionary biology methodological principles. The most severe is the assumption that one can infer historical and adaptative patterns through the observation of experimental data from one single species. The critical question that is made by evolutionary biologists to evolutionary psychologists is this: “Are there Pleistocene sources of current human cognitive adaptations?” By ignoring comparative data on brain and behavior, EP locks itself in an indefensable position, one that states that those cognitive adaptations appeared as a function of selective pressures in the Pleistocene, and were not inherited from common ancestors [6]. The lack of comparative phylogenetic analyses in EP, thus, produced some kind of “anthropocentrism” that denies the possibility of “primatocentrism” or “mammalocentrism” [15].
A second, but not less important, objection to EP is that natural selection is not the sole mechanism of evolution. There are many other evolutionary processes besides natural selection, including chance genetic sampling, various (ecological, developmental, etc) kinds of constraints on variation and development, and phylogenetic history, and, in some cases, those processes may be even more powerful than natural selection. Even though traditionally evolutionary biology focused on adaptations – i.e., those traits that have undergone primarily natural selection as the process that produced it –, not every trait is an evolutionary adaptation. The determination of the phylogenetic history of a trait can be estimated by several analytical techniques, such as those proposed by Blomberg et al. [16] and Gittleman and Kot [17], but, in order to achieve this analysis, the trait must be mapped in a cladogram. Even when comparative data are present, behavioral traits usually are more labile – i.e., present less phylogenetic signal – than other traits [16]. In the case of behavioral traits that are also prone to learning, or in the case of higher-level cognitive functions, it is very difficult to dissociate the phylogenetic component in the extant state of the trait from the a-historical component. Even though it is clear that the ability to learn may have evolved under selection, being an extremely good candidate for being an adaptation, in face of the difficulty to estimate how much of the trait is evolved and how much is ontogenetically (i.e., epigenetically and/or “learned”) it is not scientifically sound “to conclude that, because a given pattern of responses contributes to evolutionary success, then there is some 'organ' (or part of the brain) producing such a pattern, that is therefore an adaptation” [13]. Lloyd [13] argues that EP theorists did not understand the “theoretical and evidential standards” necessary for the investigation of the evolutionary history of structures and processes. She cites the following passages, from ref. 18, as a demonstration of this:
An organism's phenotypic structure can be thought of as a collection of 'design features' – micro-machines, such as the functional components of the eye or liver (...). Natural selection is a feedback process that 'chooses' among alternative designs on the basis of how well they function. By selecting designs on the basis of how well they solve adaptive problems, this process engineers a tight fit between the function of a device and its structure [18].
And
All traits that comprise species-typical designs can be partitioned into adaptations, which are present because they were selected for, by-products, which are present because they are causally coupled to traits that were selected for, and noise, which was injected by the stochastic components of evolution. Like other machines, only narrowly defined aspects of organisms fit together into function systems: most of the system is incidental to the functional properties. Unfortunately, some have misrepresented the well-supported claim that selection organizes organisms very functionally as the obviously false claim that all traits of organisms are functional – someting no sensible evolutionary biologist would ever maintain. Nevertheless, cognitive scientists need to recognize that while not everything in the designs of organisms it the product of selection, all complex functional organization is [18].
Lloyd is suspicious of those affirmations; to her, Tooby and Cosmides are paying “lip-service” to the variety of possible causes of phenotypic traits, but their actual practice uses another standards:
To show that an aspect of the phenotype is an adaptation to perform a particular function, one must show that it is particularly well designed for performing that function, and that it cannot be better explained as a by-product of some other adaptation or physical law [18].
That is, the “rule-of-thumb” for the identification of an adaptation is not taken from evolutionary biology, but rather from reverse engineering! EP theorists, in general, do not examine whether the function that is being analysed is, in fact, a well-defined biological trait; they do not check for phylogenetic signal; they do not examine whether variants of the phenotype are correlated with variants of fitness; and they do not demonstrate whether the trait in question is better explained as an adaptation or not. Rather, their approach reasons as follow:
1)      Any algorithm capable of solving the adaptative problems of cognition must have certain design features;
2)      Among these features is that the algorithms produce and operate on cost/benefit representations;
3)      We need, for survival, a cognitive mechanism that assesses the costs and benefits of different actions;
Therefore,
4)      Those cognitive mechanisms should be adaptative.
This logical fallacy is very common among EP theorists. I aggree with Lloyd [13] when she states that, even though EP theorists usually state that there are mechanisms other than natural selection that guide evolution, this is done primarily as a “lip-service” strategy to avoid criticism, and it is precisely this strategy that generates the theoretical confusion observed in EP.
 
Neuroscience objections to EP
In addition to evolutionary biological objections to EP, there are other objections in the midst of neuroscience itself against this research programme. This critique was best developed by Panksepp et al. [6, 19, 20], and concentrates on the fact that the formalisms proposed by EP are not “neuroscience-informed”. This is probably due to the very history of EP as an offshoot of cognitive science. In general, evolutionary psychologists tend to suggest that the underlying operations of the mind should be conceptualized as computations carried ou by the brain, but fail to constrain their theories with explicit neurobiological criteria. Krill et al. [21] cited a few fMRI studies using the Wason selection task to affirm that an “evolutionary cognitive neuroscience” (ECN) based on EP principles such as massive modularity is now viable. However, the conclusions of those studies can be reached whether or not the authors use EEA as an explanatory instrument or not! This, and the fact that cognitive neuroscience often suffers from the same category mistakes that EP has plenty (see below), makes evolutionary cognitive neuroscience a very complicated set of evidence in support of EP. It must be noted, though, that Krill et al. [21] affirm that,
as an explicit extension of evolutionary behavioral neurobiology, the ECN approach aims to understand a large spectrum of species differences and similarities, in addition to understanding the unique evolved capacities of the human mind and its neurobiology. Hence, the discipline is inherently comparative and should include investigations and comparisons within species and accross species.
Nonetheless, there is considerable resistance among evolutionary psychologists in assuming an organic perspective. Flombaum et al. [22] affirm that
neurobiological data cannot be used to bring a case against the thesis of psychological modularity (...). Data of this sort, although relevant to a discussion of the principles underlying modularity, are orthogonal to a discussion of whether or not modularity, in the psychological sense, exists in the first place
This kind of affirmation is simmetrical to neurobiologists' affirmation that, on the contrary, it is precisely the neurobiological data that will judge “whether or not modularity, in the psychological sense, exists in the first place”. Bolhuis and Macphail [23] reasoned that if a system is psychologically modular, it must have a modular neural substrate as well. Most EP theorists avoid this critique by proposing the “ontologically weak” definition of modularity delineated above [eg, 11]. As such, they reject neurobiological data, and instead rely on algorithms to describe their modules.
An algorithm is a formal procedure for any mathematical operation. While specific neural algorithms may underpin certain aspects of sensory/perceptual and motor/action, channel functions (...), the central neural substrates underlying psychobehavioral states are unlikely to be organized in a similar manner. Such functions are more fundamentally organic and analog, exerting global brain/mind influences. In other words, the brain can produce behavioral and mental phentypes via neuronal mechanisms that are markedly different in their constitution when compared with mathematical computations, whether manifested at the level of anatomical reorganizations and/or dynamic changes in the elements of individual synapses [6, emphasis in the original].
While an “algorithmic” approach in the cognitive sciences was acceptable when this discipline was born because neurobiological data was scarce, this is becoming more and more unnacceptable as modern neuroscience begins to provide a suitable foundation to think more clearly about the underlying organic basis of mental processes. Carefully validated animal models in psychopatology, for example, are beginning to reveal the neural underpinings of many mental processes, and there is an increasing preocupation in the area with the ethological and comparative aspects of construct validity [24,25]. Thus, the “comparative alternative to the model animal approach” envisioned by Preuss in cortical science [26] is also penetrating other areas of psychobiological research, elucidating the neural basis of mental processes in a evolutionary fashion without resorting to the algorithmic approach.
Premack [27] observed that at the same time that microarray and morphological studies reveal differences between human and non-human animals' brains, cognitive studies have found non-human animals to have abilities once thought unique to humans. This suggests a disparity between the evolution of the brain and evolution of the mind, a disparity that he deems to be illusory. In fact, he argues, cognitive research has simply not kept pace with neural research. The simple fact that a great deal of cognitive science papers use formal models – i.e., an “algorithmic” approach – and do not entrench those models in ontologically realistic neurobiological models is supportive for the thesis that a category mistake is being commited on the field of cognitive sciences. I suspect that EP inherited the same methodological problem.
 
The “black box” approach to genotype-phenotype relationships in EP: Developmental objections to evolutionary psychology
Lickliter and Honeycutt [28] generated a big fuss among psychologists with their paper “Developmental dynamics: Toward a biologically plausible evolutionary psychology”, probably because, differently than most critiques of EP, this paper was published in a Psychology journal. This fact alone corroborates their opinion that a critique of EP is “particularly important in that the assumptions of evolutionary psychology are often presented to psychologists having little background in or knowledge of genetics, embryology, developmental biology, or related disciplines contributing to modern evolutionary theory”. Their critique is very straightforward: contemporary evolutionary developmental biology postulates that the development of any individual organism is the consequence of a unique web of interactions among the genes it carries, among the complex molecular interactions within and across individual cells, and of the nature and sequence of physical, biological and social environments through which the organism passes during its development [5]. According to them, the primary assumptions of EP “center around the assumption that basic aspects of an organism, including its morphology, physiology, and psychology, are best understood as the products of its genes” [28]. At its core, this assumption assumes that phenotypes can be specified in advance of the organism's development [29], which carries in itself the seed of the assumption that evolutionary factors are ontologically prior to and more fundamental than developmental factors in directing phenotypic outcomes, thus creating a fallacious dichotomy between phylogeny and ontogeny. However, it is commonplace in contemporary developmental systems theory (DST) that gene expression is routinely modified by other genes, as well as by the local cellular and extracellular environments of the developing organism [for a nice review, see refs. 30 and 31]. It has been argued [5] that evolution occurs at (at least) three different levels: changes in gene frequency, the appearance of new characters, and the appearance, adaptation and radiation of new species, all of which rely on genetic change through time.
Changes in gene frequency operates at the population level through mutations, selection, drift, migration, and meiotic drive. The appearance of new characters and the appearance of adaptive radiation of species require alterations of ontogeny, even if the adaptation is behavioral or physiological (...). Changes which we think of as characterizing descent with modification – the origin of new phenotypes, gigantism, dwarfism, neoteny, paedomorphosis, mimicry, phenotypic plasticity, even speciation – are all changes that arise through the modification of developmental processes [5, emphasis added].
This can be best understood by figure 1, taken from Hall's figure 2. As can be deprehended, in the current model of trait evolution mutations introduce variation to epigenetic processes, eggs transmit maternal and zygotic genetic as well as epigenetic information from one generation to another, and epigenetic processes provides “hierarchical developmental processes that produce embryonic, larval and adult phenotypes” [5] upon which selection will operate, spreading the change and/or conservation through populations and individuals.
How can the “gene-centric” position of EP, together with its inflexible EEA postulates, be held in face of DST? According to Buller and Hardcastle [3], evolutionary psychologists
treat environmental factors as “triggers” that activate the development of a module in accordance with a “developmental program” that is coded in the genes (...). The picture is that genes encode the information for constructing a module, but that they await a cue from the environment telling them when to begin constructing the module (or, in the case of a conditional genetic program, that they await a cue from the environment telling them which of a limited number of “settings” to apply in constructing the module).
This is inconsistent with the observation that development creates a diffuse proliferation of brain connectivity, fixed in its final form by later brain activity, guided by interaction with the environment [32,33]. EP relies too strongly on its central tenets of natural selection of adaptations by means of genetic processes, ignoring the complex and rich processes by which the modification of developmental processes entail evolutionary changes.
 
Last but not least: General epistemological problems with EP
We have arrived at a point in which we can summarize the mistakes of EP. We will procede by demonstrating how an EP argument is built, and how mistakes in one level lead to mistakes in other levels.
Evolutionary psychologists begin with an extremely interesting proposition: a great deal of human behavior can be explained by its evolutionary history. For this, they begin with the following argument:
 
(1)   Behavioral trait X is linked to underlying mental process M1
 
which is justifiable from the point of view of contemporary philosophy of mind, as well as by most modern accounts in the behavioral sciences. A second assumption, though, is added to this first argument:
 
(2)   Mental process M1 is so important for survival that must have been selected for in some environment of evolutionary adaptedness.
 
This assumption is not, per se, true or false. Evolutionary biologists assess statements such as (2) by analysing if the trait M1 is a well-defined biological trait that presents significant phylogenetic signal, as well as if variants of the phenotype M1 are correlated with variants of fitness. Evolutionary psychologists, on the other side, employ a different reasoning that can be summarized as
 
(3)   We need, for survival, a cognitive mechanism that assesses the costs and benefits of different actions; therefore, those cognitive mechanisms should be adaptative.
 
Assumption (3) is an auxiliary assumption to the definition of assumption (2). This “reverse engineering” strategy is not compliant to sound evolutionary biological criteria for adaptation. Instead, it ignores the possibility that, even if the phenotype M1 is indeed an adaptation, it was selected for not in the evolutionary history of the trait as infered by a comparative approach, but in the environment of evolutionary adaptedness specific for the Homo sapiens lineage – most specifically, by the environmental pressures that existed in the Pleistocene. Also, evidences for this thesis is not gathered from paleoneurological data, but from experiments with contemporary humans. As such, the Pleistocene EEA pressures are inferred from common sense, not from paleontology.
By a leap of logic (I did not say “faith”, I swear), EP them postulates that, since those mechanisms are adaptative, then they could not possibly be domain-general; as such, M1 is taken to be domain-specific, which leads to
 
(4)   Evolution operates on mental modules, defined by their information encapsulation.
 
Again, no reference is made to the mechanisms that create those modules. In fact, indirect references are made to a central tenet of EP, which states that
 
(5)   Natural selection creates mental modules by positive selection of genes,
 
which is not necessarily true. Even if the trait M1 evolved as a module, it is not necessary that this process is guided by natural selection. As stated before, natural selection is but one mechanism of evolutionary change, and, in some traits, may not even be the chief process. Also, the lack of reference to developmental processes – or rather, the assumption that
 
(6)   Genes generate developmental programs,
 
an hypothesis not corroborated by contemporary evo-devo biology, as well as DST – renders the EP treatment of M1 doubtful. EP tries to avoid the criticism from evo-devo biologists by avoiding references to actual mechanisms, instead relying on yet another assumption:
 
(7)   M1, as well as it's evolution, is best characterized by formalisms instantiated in algorithms, independent of the underlying mechanisms.
 
This is also true to EP's approach to the neural mechanisms that underlie M1.
This series of phallacies and logical mistakes are enough to dismiss evolutionary psychology as is practiced. It does not mean, however, that the conclusion – behavioral trait X can be explained, in a good deal, by its evolutionary history – is doomed, even though premises 2-7 are false. One extreme caveat of EP, though, is probably underlying the other problems reviewed in this post. In defense of EP, one must add that this low-level mistake is not sole property of EP, but is a problem of all psychology. I am talking about here, obviously, about the category mistake problem, a more serious epistemological issue that was refered previously on other sections of this post. A category mistake represents a given instance of the world “as if they belonged to one logical type or category (or range of types or categories), when they actually belong to another” [34]. In short, a category mistake is made when one postulates, for example, that an epistemological category (such as a mental module) is actually an ontological category. In this sense, the premise that one given function M1 could be described as a module does not entail logically that this function is a module – only that the descriptive power of the “module” category suffices for the time being. Category mistakes usually appear whenever theoretical models are made ontologically independent [35], which seems to be the case with the concept of module as well as the concept of EEA. Every theoretical model has its assumptions, understood as the pressupositions that concern the structures and processes, causes and effects, and organizational and functional hierarchies found in its object. These sets of pressupositions can be formal as well as ontological; ontological assumptions, however, must meet the internal demands of the formal models as well as be consistent with prior ontological pressupositions closely connected to this model. EP, on the other hand, characterizes its formal assumptions as if they were ontological assumptions, in a clear example of a category mistake. As such, EP confounds both categories of assumptions, meeting the internal demands of the formalism, but not the prior ontological pressupositions that are connected to the formalism – such as the mechanisms by which evolutionary change occurs, or the neural underpinings of the variables it tries to model. As such, EP transforms the normative kinds that are its objects – the “mental modules” – to natural kinds, as if the formalism of a mental module had, in fact, ontological reality.
This category mistake is probably inherited from prior attempts to analyse behavior in evolutionary terms – namely, ethology, sociobiology and behavioral ecology. Griffiths [36] has argued that, while the concept of “innateness” is taken by some behavioral ecologists as irretrievably confused, other behavioral ecologists assume regard the same concept as refering to adaptations. This same mistake is observed on cognitive science and, finally, on EP – after all, the whole evolutionary psychology programme is an attempt to demonstrate that some behavioral traits, as well as their (putative) associated mental functions, are innately defined. According to Griffiths [36], three broad concepts are encapsulated by the innateness concept: developmental fixity, “species nature”, and intended outcome.
'Developmental fixity' means that the trait is in some sense 'hard to change': it is insensitive to environmental inputs in development; its development is or appears goal-directed, so that when prevented from developing in one way it develops in another; changing it disrupts or impairs development. 'Species nature' means that innate traits reflect what it is to be an organism of that kind, with consequent associations of typicality or universality. 'Intended outcome' means that innate traits are how the organism is meant to develop: to lack them is to be malformed; environments that disrupt them are 'bad rearing', to use Konrad Lorenz's term. This intentional or normative element of the innateness concept is today usually assimilated to the idea of design by natural selection: innate traits are those that the organism is designed to possess or which are programmed in its genes.
Since one cannot derive empirically one of these three concepts from another, a theoretical construct that conflates them is undesirable. EP, however, uses the concept of innateness as-is, and, thus, falls in the category mistakes delineated above, as well as the logical fallacies 2-7. Since the construct of innateness comprises three different concepts that are not linked empirically or logically, its use tends to give rise to illicit inferences from the presence of one biological property to the presence of the others. In the same sense that category mistakes in Psychology stems from the indiscriminate use of “folk psychological” terms [37], category mistakes in EP stems from the indiscriminate use of “folk biological” [38] terms in addition to the usual category mistakes of Psychology.
Can the EP programme be rescued? Many proposals were made. My favorite ones [39,40] simply ignore the whole programme and reconcile behavioral neuroscience with evolutionary neuroscience. As you can see, I am rather skeptical about the rescuing of the EP programme, rather than its total reformulation. Maybe one day I will write about that...
 
References
[1] Barkow J, Cosmides L, Tooby J (1992). The Adapted Mind: Evolutionary Psychology and the Generation of Culture. New York: Oxford University Press.
 
[2] Gould SJ, Lewontin RC (1979). The spandrels of San Marco and the Panglossian paradigm: A critique of the adaptationist programme. Proceedings of the Royal Society of London Series B 205: 581-598.
 
[3] Buller DJ, Hardcastle VG (2000). Evolutionary psychology, meet developmental neurobiology: Against promiscuous modularity. Brain and Mind 1: 307-325.
 
[4] Sarnecki J (2006). Developmental objections to evolutionary modularity. Biology and Philosophy X: XX-XX. http://dx.doi.org/10.1007/s10539-006-9029-7
 
[5] Hall BK (2003). Unlocking the black box between genotype and phenotype: Cell condensations as morphogenetic (modular) units. Biology and Philosophy 18: 319-247.
 
[6] Panksepp J, Moskal JR, Panksepp JB, Kroes RA (2002).Comparative approaches in evolutionary psychology: Molecular neuroscience meets the mind. Neuroendocrinology Letters 23: 105-115.
 
[7] Wagner GP (1995). The biological role of homologues: A building block hypothesis. N Jb. Geol. Paläont. Abh. 19: 279-288.
 
[8] Wagner GP (1996). Homologues, natural kinds and the evolution of modularity. American Zoologist 36: 36-43.
 
[9] Fodor J (1983). The Modularity of Mind. Cambridge: MIT Press.
 
[10] Carruthers P (2003). On Fodor's problem. Mind and Language 18: 502-523.
 
[11] Colheart M (1999) Modularity and cognition. Trends in Cognitive Sciences 3: 115-120.
 
[12] 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.
 
[13] Lloyd EA (1999). Evolutionary psychology: The burdens of proof. Biology and Philosophy 14: 211-233.
 
[14] Stotz KC, Griffiths PE ( ). Dancing in the dark: Evolutionary psychology and the argument from design.
 
[15] Keeley BL (2004). Anthropomorphism, primatomorphism, mammalomorphism: Understanding cross-species comparisons. Biology and Philosophy 19: 521-540.
 
[16] Blomberg SP, Garland Jr. T, Ives AR (2003). Testing for phylogenetic signal in comparative data: Behavioral traits are more labile. Evolution 57: 717-745.
 
[17] Gittleman JL, Kot M (1990). Adaptation: statistics and a null model for estimating phylogenetic effects. Systematic Zoology 39: 227-241.
 
[18] Cosmides L, Tooby J (1995). Beyond intuition and instinct blindness: Toward an evolutionary rigorous cognitive science. In: J Mehler, S Franck (Eds.), Cognition on Cognition, pp. 69-105. Cambridge: MIT Press.
 
[19] Panksepp J, Panksepp JB (2000). The sevens sins of evolutionary psychology. Evolution and Cognition 6: 108-131.
 
[20] Panksepp J, Panksepp JB (2001). A continuing critique of evolutionary psychology. Seven sins for seven sinners, plus or minus two. Evolution and Cognition 7: 56-80.
 
[21] Krill AL, Platek SM, Goetz AT (2007). Where evolutionary psychology meets cognitive neuroscience: A précis to “Evolutionary Cognitive Neuroscience”. Evolutionary Psychology 5: 232-256.
 
[22] Flombaum JI, Santos LR, Hauser MD (2002). Neuroecology and psychological modularity. Trends in Cognitive Sciences 6: 106-108.
 
[23] Bolhuis JJ, Macphail EM (2001). A critique of the neuroecology of learning and memory. Trends in Cognitive Sciences 5: 462-433.
 
[24] Rodgers RJ, Cao B-J, Dalvi A, Holmes A (1997) Animal models of anxiety: An ethological perspective.Brazilian Journal of Medical and Biological Research 30: 289-304.
 
[25] Blanchard DC, Sakai RR, McEwen B, Weiss SM, Blanchard RJ (1993). Subordination stress: Behavioral, brain, and neuroendocrine correlates. Behavioral Brain Research 58: 113-131.
 
[26] Preuss TM (2000). Taking the measure of diversity: Comparative alternatives to the model-animal paradigm in cortical neuroscience. Brain, Behavior and Evolution 55: 287-299.
 
[27] Premack D (2007). Human and animal cognition: Continuity and discontinuity. PNAS 104: 13861-13867. http://dx.doi.org/10.1073/pnas.0706147104
 
[28] Lickliter R, Honeycutt H (2003). Developmental dynamics: Toward a biologically plausible evolutionary psychology. Psychological Bulletin 129: 819-835. http://dx.doi.org/10.1037/0033-2909.129.6.819
 
[29] Lickliter R, Berry TD (1990). The phylogeny fallacy: Developmental psychology's misapplication of evolutionary theory. Developmental Review 10: 348-364.
 
[30] Davidson EH (2001). Genomic Regulatory Systems: Development and Evolution. New York: Academic Press.
 
[31] Prud'homme B, Gompel N, Carroll SB (2007). Emerging principles of regulatory evolution. PNAS 104: 8605-8612. http://dx.doi.org/10.1073/pnas.0700488104
 
[32] Rakic P, Bourgeois J-P, Echenhoff MF, Zaecevic N, Goldman-Rakic P (1986). Concurrent overproduction of synapses in diverse regions of the primate cerebral cortex. Science 232: 232-235.     
 
[33] Penn AA, Shatz CJ (1999). Brain waves and brain wiring: The role of endogenous and sensory-driven neural activity in development. Pediatric Research 45: 447-458.
 
[34] Ryle G (1949). The Concept of Mind. Middlesex: Penguin Books.
 
[35] Winther RG (2006). On the dangers of making scientific models ontologically independent: Taking Richard Levins' warnings seriously.Biology and Philosophy 21: 703-724. http://dx.doi.org/s10539-006-9053-7
 
[36] Griffiths PE ( ). What is innateness?
 
[37] Wittgenstein L (1953). Philosophical Investigations. Oxford: Basil Blackwell.
 
[38] Medin DL, Atran S (1999). Folkbiology. Cambridge: MIT Press.
 
[39] Bass AH (1998). Behavioral and evolutionary neurobiology: A pluralistic approach. American Zoologist 38: 97-107.
 
[40] Sherry DF (2006). Neuroecology. Annual Review of Psychology 57: 167-197. http://dx.doi.org/10.1146/annurev.psych.56.091103.070324
 
Edit: I was re-reading Futuyma's Evolutionary Biology last night and stumbled upon something that made me think. He says, bastically, that the main justification in a priori reasoning in terms of adaptation is not that the premise of a given trait being adaptative is true, but that it has heuristic value. I do agree with him in certain terms, but still do not think that assuming a pragmaticist epistemological stance suffices for the rescuing of that project.