From moods to modules:

preliminary remarks for an evolutionary theory of mood phenomena

Dylan Evans

Department of Philosophy, Logic and Scientific Method

London School of Economics and Political Science

Houghton Street

London WC2A 2AE

d.evans@lse.ac.uk

Abstract:

In the past few decades, research in the psychology of emotion has benefited greatly from being located in a firm evolutionary framework. It is argued that research in the psychology of mood might attain equal rigour by taking a similar approach. An evolutionary framework for mood research would be based on evolutionary psychology, the main thesis of which is the Massive Modularity Hypothesis. Translating the folk-psychological language of moods into the scientific language of modules might clarify many theoretical questions and provide a sound basis for empirical research. It is argued that such an evolutionary approach would reveal mood to be a much more heterogeneous category than emotion. While the six basic emotions identified by Paul Ekman are probably each subserved by a single module, prototypical moods such as elation, depression, anxiety and irritability are likely to be subserved by a wide range of modules. An evolutionary approach to mood might therefore lead to the elimination of the concept of mood from scientific psychology altogether.

1. Introduction: evolutionary psychology and the massive modularity hypothesis

Evolutionary psychology is a relatively new research program that aims to map the human mind by combining cognitive psychology with evolutionary biology (Tooby and Cosmides, 1992). It is thoroughly cognitive in regarding the mind as an information-processing device that can be described in computational terms. Unlike most cognitive psychologists, however, who regard the mind as a general-purpose problem-solver, evolutionary psychologists argue that the mind consists of a large number of special-purpose devices, which are usually referred to as 'modules'. This view has since become known as the 'Massive Modularity Hypothesis' (Murphy and Stich, 1998; Samuels, 1998), after a phrase originally coined by Dan Sperber (Sperber, 1994).

Since Fodor introduced the term in 1983, the term 'module' is used - some might say misused or even abused - in an increasing number of works in cognitive psychology (Fodor, 1983). However, different theorists use the term in different ways, and rarely take the trouble to define it in any detail (Segal, 1996). Evolutionary psychologists have not been exempt from this lack of precision (Samuels, 1998). Nevertheless, it is possible to infer a coherent concept of modularity from the writings of various evolutionary psychologists. Following Samuels (1998), I will use the term 'Darwinian module' to refer to this concept.

A Darwinian module is a computational mechanism that is domain-specific, informationally-encapsulated; it is designed by natural selection as part of the universal species-typical design, and is thus innately specified/genetically determined. Darwinian modules have a typical ontogenetic pace and sequencing, and are often associated with characteristic breakdown patterns. They may also be associated with a specific neural or biochemical architecture, in which case they cannot be implemented by just any neural mechanism.

Those cognitive psychologists who argue that the mind has some degree of modularity usually restrict this to input and output systems. That is, they claim that input systems such as vision, and output systems (those responsible for direct muscular control), are likely to be modular, but that central systems (those responsible for 'higher' cognitive processes like reasoning) are nonmodular (Fodor, 1983). Evolutionary psychologists, on the other hand, argue that the whole mind, including the central processes, are likely to be modular (Sperber, 1994). If this is true, it has important consequences not just for our understanding of reasoning and other classically 'cognitive' processes, but also for our understanding of emotion and mood.

2. The modularity of emotion

Evolutionary theories of emotion have their roots in Darwin's own work. His 1872 book, The Expression of the Emotions in Man and Animals, traced many human facial expressions and bodily gestures to their putative origins in pre-human ancestors. For example, Darwin argued that the way that humans bare their teeth when scowling with anger was a vestige of a primitive agonistic display that can still be observed in modern chimpanzees (Darwin, 1872: 12). Despite the initial popularity of Darwin's book among the general public, it is only recently that it has had a similarly powerful impact on the work of professional psychologists. The neglect of Darwin's theory by psychologists throughout much of the twentieth century was closely connected with the culturalist theory that dominated psychological research in emotion during this time. According to this theory, the expression of emotion, and even the emotions themselves, were culturally specific, learned phenomena (La Barre, 1947). In the early 1970s, however, studies by Paul Ekman and colleagues began to provide strong evidence against the culturalist view. Studies of a remote preliterate culture in New Guinea strongly suggested that certain facial expressions were universally associated with a small range of 'basic emotions': surprise, joy, sadness, fear, anger and disgust (Ekman and Friesen, 1971). Ekman went on to argue that each of these basic emotions was characterised by a distinctive pattern of facial expressions and physiological changes, and was triggered by distinct kinds of event (Ekman, 1992). All the basic emotions tended to have a rapid and unbidden onset, and to last for seconds rather than minutes, hours or days. Ekman argued that the co-occurrence of this cluster of properties could only be explained by supposing that basic emotions are adaptations that evolved because they helped our ancestors to deal with 'fundamental life-tasks'(Ekman, 1992: 171).

As Paul Griffiths has argued, the properties that Ekman associates with the six basic emotions are among those that characterise Darwinian modules (Griffiths, 1990). Like modules, the operation of basic emotions is mandatory, fast, and informationally encapsulated. Like modules, the representations that govern basic emotions are often inaccessible to other mental processes. There is also increasing evidence that some basic emotions such as fear are 'hardwired'; that is, they are implemented by a specific neural architecture (LeDoux, 1998). Furthermore, the co-occurrence of these properties supports the contention that basic emotions are adaptations in exactly the same way that it supports the view that modules are adaptations. It is therefore plausible to identify basic emotions with a subset of Darwinian modules.

To be more precise, it is not the basic emotions themselves that should be identified with modules, but rather what Ekman calls the 'automatic appraisal mechanism' of the basic emotions. Ekman defines a basic emotion as a distinctive pattern of facial signals and physiological responses that is regularly associated with a distinctive kind of antecedent event. This analysis can appear rather behaviouristic, in that it links certain stimuli with certain responses without specifying the psychological mechanism that mediates the response. However, Ekman does not neglect the psychological dimension entirely, as is clear from his discussion of the mechanisms that detect and appraise emotionally relevant stimuli and coordinate the various facial and physiological changes that constitute the appropriate response to these stimuli. He terms these mechanisms 'automatic appraisal mechanisms' (AAM), and develops a basic model of how they work (Ekman, 1992: 185-87). Each emotion would be subserved by a distinct AAM, which would only attend to a particular range of stimuli. It is the AAM of a given emotion, rather than the emotion itself, that should be conceived of as a Darwinian module.

The claim that the AAMs associated with basic emotions are Darwinian modules involves more than the idea that basic emotions are subserved by psychological mechanisms that are domain specific, informationally encapsulated, hardwired, and so on. It also involves the claim that these mechanisms are adaptations. This is also consistent with Ekman's view of basic emotions. Similar arguments have also been advanced by a number of other theorists, including Robert Plutchik, R. S. Lazarus, Randolph Nesse, and John Tooby and Leda Cosmides (Plutchik, 1980; Nesse, 1990; Tooby and Cosmides, 1990; Lazarus, 1991). A crucial part of all these accounts consists of formulating hypotheses about the function of each of the various basic emotions. Such hypotheses posit that a given emotion evolved because it helped our ancestors to solve a particular problem. For example, all the theorists argue that fear evolved because it helped our ancestors to solve the problem of evading predators. Those of our ancestors who had the capacity for fear survived longer than, and thus out-reproduced, those who did not. This kind of hypothesis is often referred to as an 'adaptationist hypothesis' because it claims that natural selection was the major force in the evolution of a given trait rather than, say, genetic drift.

Translated into the terminology of evolutionary psychology, then, the work by Ekman and other evolutionary theorists of emotion suggests that there is a class of Darwinian modules whose function is to produce the characteristic patterns of facial expression and physiological change that we refer to as surprise, joy, sadness, fear, anger and disgust in response to certain situations. But what about other kinds of affective phenomena? Can the evolutionary model be extended to cover other emotions such as guilt, shame, and love? If so, might it extended even further, to cover moods such as anxiety, elation and depression?

The question of whether there are modules whose function is to produce what Griffiths has termed 'higher cognitive emotions' such as guilt, shame, and love has been explored elsewhere, and will not be addressed here (Griffiths, 1997: 120-22). What interests me is the question of whether there are modules whose function is to produce the kinds of thing that are commonly referred to as 'moods'. Before addressing this question, however, it will first be necessary to say a something about what moods are and how they differ from emotions.

3. What are moods?

Moods are commonly thought to share with emotions certain general properties of affective phenomena, but to differ from emotions in a number of specific parameters. For example, moods are usually thought of as relatively long-term states, lasting for hours, days or even weeks rather than minutes or seconds. Also, moods are generally thought to build up gradually as a result of many small incidents, in contrast with emotions, which are caused by a single powerful stimulus (Ekman, 1994; Parkinson, Totterdell et al., 1996: 6).

Not everyone is happy with this distinction. Randolph Nesse, for example, has argued that the distinction is not an important one, and claims that it is not routinely accepted in philosophy and even less accepted in psychology (Nesse, 1992). Nesse cites various authorities to support his contention, including Lazarus (Lazarus, 1991), though it is hardly necessary to go to such scholarly works to find support for his claim, as a brief perusal of any standard undergraduate textbook in the field suffices to show that mood is not a significant concept in contemporary psychology. In one standard text, the term 'mood' is not even listed in the index (Gross, 1996), while in another, the entry reads simply 'Moods: see Emotions' (Atkinson, Atkinson et al., 1996).

Nesse may be right about the tendency of contemporary psychology to downplay the mood-emotion distinction, but this does not mean that the distinction is invalid. The validity of the distinction will not be determined by a show of hands, but by conceptual advances and empirical research. It is only in the last two decades that mood has become a focus of attention among psychologists, and it is not surprising that this work does not yet constitute orthodox opinion nor appear in the standard textbooks.

One of the reasons why Nesse and others are sceptical of the validity of the mood-emotion distinction is that the distinguishing criteria most often cited are merely quantitative. If moods differ from emotions only in their duration and in the number and intensity of stimuli, then perhaps emotions and moods are, at bottom, the same thing, and this thing varies continuously along these parameters. In this case, dividing these continuous parameters in two would be an arbitrary decision. This kind of consideration has prompted a number of theorists to propose qualitative criteria for distinguishing moods from emotions. William Morris, for example, has argued that emotions provide consciousness with information about the environment while moods provide information about the self (Morris, 1992). More precisely, Morris argues that emotions tell the conscious system that 'something is going wrong or going well with the external situation, whereas mood tells us that our personal resources are insufficient for dealing with current demands or exceed what is required of us' (Parkinson, Totterdell et al., 1996: 7 emphasis in original). This distinction is problematic because whether something is going well or badly usually depends on the relationship between personal resources and environmental problems rather than on one or the other. If we are unhappy because we are hungry and have no food to eat, for example, this is because of a combination of low personal resources (low blood glucose) and an environmental deficit (no food is at hand).

Perhaps a more promising distinction is provided by Richard Davidson, who has proposed that emotions modulate or bias action, while moods modulate or bias cognition (Davidson, 1994). Davidson's distinction is reminiscent of the classic distinction made by Vincent Nowlis, who suggested that emotions are lower-order dispositions and moods are higher-order dispositions; emotions are dispositions to act in particular ways, while moods are dispositions to have particular emotions (Nowlis, 1963). Like Nowlis, Davidson implies that moods can alter the probability that particular emotions will be triggered, but unlike Nowlis he does not rule out the possibility that the causal arrow can point in the other direction too (emotions can also induce particular moods). Nor does Davidson's distinction limit the function of moods to their capacity to induce emotions, as Nowlis does. On the contrary, Davidson sees this capacity as part of a much broader function, which is to alter information-processing priorities and to shift modes of information-processing; 'mood will accentuate the accessibility of some and attenuate the accessibility of other cognitive contents and semantic networks', he claims (Davidson, 1994). It is only in virtue of this broader function that moods have the capacity to lower the threshold for arousing particular emotions. People in an irritable mood will become angry more readily than usual because they 'construe the world around them in a way that permits, if not calls for, an angry response' (Ekman, 1994: 57).

Ekman echoes Nowlis and Davidson in claiming that moods are dispositions to have certain emotions, though like Davidson he does not take this to be the defining or the only criterion for distinguishing moods from emotions (Ekman, 1994). He also claims that emotions have their own unique facial expression while moods do not, and that emotions that occur during a mood are more intense, less controlled and decay more slowly than those that do not.

Nico Frijda proposes a distinction between mood and emotion based on a distinction that Hume draws between the object of an emotion and its cause. Hume conceived of emotions as intentional states, meaning that they involved representations 'about' something, and claimed that the intentional object of emotion could differ from the cause of the emotion. For example, the cause of one's anger may be an insult, while the object of the anger might be the person who insulted you. Frijda proposes that moods differ from emotions in lacking an object. Moods are thus defined as 'noninentional affective states' (Frijda, 1994). This is reminiscent of Freud's distinction between fear and anxiety. Freud argued that fear is focused on a specific object while anxiety 'has a quality of indefiniteness and lack of object' (Freud, 1926: 325). This distinction has since become an accepted staple of psychiatric wisdom, in which it is now common to distinguish between phobias, which involve fears of specific objects, and anxiety states, which are 'not attributable to real danger or focussed on any particular stimulus (i.e. "free floating anxiety")' (Hughes, 1991: 48). Frijda's distinction is also consistent with the view of many theorists that moods are more 'diffuse' and 'global' than emotions (Morris, 1989).

As this brief survey of the literature reveals, there is no consensus about what moods are. There are a number of well-entrenched folk-psychological notions about moods, but these do not amount to a theory. The conceptual work on mood has consisted of little more than each researcher picking on one of these notions and pitching it against rival theorists who emphasise a different one. No one stops to ask whether these various aspects are produced by a single psychological mechanism or by several. Perhaps this is because researchers have lacked a good over-arching theory of the mental in which the question can be properly framed. I propose that evolutionary psychology provides such a theory, and has already proved its theoretical utility in explaining affective phenomena by relating Ekman's six basic emotions to six mental modules. The question is whether the same can now be done for mood phenomena.

4. The modularity of mood

Unlike basic emotions, moods do not seem like promising candidates for modular realisation. The very properties that are held to distinguish moods from emotions are precisely those that would indicate lack of modularity. For example, the 'global' effects of mood in altering information processing modes and memory across many different domains seems to argue against moods being inaccessible to other mental processes. Likewise, the idea that moods might be much less specific with regard to the nature of the stimulus suggests that they are more domain-general than is consistent with the notion of modularity. Moods are not always subject to conscious control, and so they do seem to have some degree of informational encapsulation, but there is evidence to suggest that moods are more easy to regulate than emotions, which implies that they are less encapsulated (and so less modular) than emotions. No one has found a specific neural structure associated with any particular mood; indeed, the only serious neurological theory of mood, the so-called 'catecholamine hypothesis', suggests that moods are associated with levels of neurotransmitters that are widely dispersed throughout the brain (Schildkraut, 1965; Panksepp, 1982).1

On the other hand, moods do show characteristic breakdown patterns, as is evident from the fact that psychiatrists have long recognised a class of disorders that they refer to as 'mood disorders'. These disorders are characterised by moods that are of greater intensity and longer duration than normal. However, the mere fact that some mental phenomenon exhibits a characteristic breakdown pattern is certainly not very strong evidence for the modularity of the phenomenon in question. Even the broader conception of modularity favoured by evolutionary psychology regards modularity as a cluster of properties; the possession of a single modular property is not sufficient to make the system modular, least of all the possession of a less important modular property such as having a characteristic breakdown pattern.

Nevertheless, attending to the breakdown patterns of various moods might suggest another strategy for investigating the links between moods and modules. For while there are, as yet, no evolutionary theories of normal mood, there are, as we have seen, a number of evolutionary theories of abnormal mood. These theories all presuppose an implicit theory of normal mood. This suggests that we might begin an investigation of the possible modular basis of normal mood by turning to the alleged modular basis of certain mood disorders. This is not as strange as it may sound. There have been many cases in which the first evidence for the existence of a given module came from a disorder. Indeed, this strategy forms the principal methodological tool of cognitive neuropsychology. It has also been prominent in evolutionary psychology. The theory of mind module, for example, the existence of which is supported by much non-clinical evidence now, was first postulated as part of a theory of autism in which it was argued that autistic symptoms result from a breakdown in this module (Baron-Cohen, 1995).

The social competition hypothesis of depression (Price, Sloman et al., 1994), the defection hypothesis of postnatal depression (Hagen, 1998; Watson and Andrews, 1998), and the reciprocal altruism hypothesis of dysthymia (McGuire, Fawzy et al., 1994) can all be construed as attributing depressive symptoms to the function or malfunction of certain hypothetical modules. The social competition hypothesis attributes depressive symptoms to the operation of a social comparison module. The defection hypothesis attributes the symptoms of postnatal depression to the operation of a module that evaluates the viability of newborn babies. The reciprocal altruism hypothesis attributes the symptoms of dysthymia to a malfunction in the modules associated with social exchange or to mistaken information in some non-proprietary store. Perhaps these three modules are all that we need to account for normal mood phenomena too.

In fact, it would be far too strong to claim that these three modules are all that we need to account for normal mood phenomena. For a start, there are many factors that cause low mood that do not seem to lie within the domain of any of the three modules in question. For example, bad weather and social isolation are both thought to cause low mood, yet neither of these factors is relevant to social comparison, the viability of newborn babies or reciprocal altruism. Furthermore, there are other moods apart from what I have called 'low' mood; euphoria, anxiety and irritability are usually regarded as moods as well (Ekman, 1994). While euphoria might conceivably be caused by the same modules as low mood, it seems far less likely that anxiety and irritability would also be due to these modules. Thus while the three modules discussed above may play an important part in some kinds of low mood and possibly also in some kinds of euphoria, it seems very likely that other modules are involved in other kinds of low mood and euphoria and still other modules in anxiety and irritability. It may be the case that mood phenomena also involve non-proprietary stores of information in addition to specific modules. This would seem likely in view of the allegedly 'global' and 'diffuse' effects of mood on cognitive processing.

All this suggests that 'mood' will turn out to be a massively heterogeneous category from the point of view of evolutionary psychology. It is not just the case that different moods will be subserved by different modules, as is the case with the various basic emotions. Rather, each mood may be subserved by a variety of different modules and non-proprietary stores, unlike, say, the emotion of fear, which may be subserved by a single predator-avoidance module.

5. Should the concept of mood be eliminated from scientific psychology?

As already noted, the folk-psychology distinction between emotion and mood is a phenomenological one. According to this distinction, moods are affective states that last longer than emotions and that are caused by a number of subliminal triggers rather than by a single explicit one. Can this phenomenological distinction be mapped on to any valid distinction at the level of underlying mechanisms?

This question is equivalent to asking whether the class of modules associated with basic emotions (the six AAM's) and the class of modules associated with mood phenomena are both natural kinds. The class of modules associated with basic emotions probably does constitute a natural kind. The six AAM's associated with the six basic emotions all produce, as their output, a specific facial expression and a specific suite of physiological changes. If no other modules do this (and there is no evidence as yet that any do), then the possession of these properties would be sufficient to make the six AAM's a natural kind. There would be, in other words, a motivated way of distinguishing between emotion modules and non-emotion modules in the human mind.

Whether or not the modules associated mood phenomena also constitutes a natural kind is far less clear. We have not yet come anywhere close to enumerating all the distinct modules associated with mood phenomena, so it is still too early to be able to say whether all and only these modules share any scientifically interesting properties over and above the property of being associated with mood phenomena. However, if one were to hazard a guess at the answer to this question in the light of the current meagre evidence, it would probably have to be negative. The sheer heterogeneity of the modules associated with mood phenomena argues against this class being a natural kind. Even if the modules associated with mood phenomena turned out to have certain properties in common, such as a comparatively high degree of connectivity with other modules (a property that would explain the comparatively 'diffuse' or 'global' nature of mood effects), it is unlikely that these would be the only modules with these properties. Other modules, not associated with mood phenomena, might also have high degrees of connectivity. In this case we would be faced with a dilemna. Should the concept of mood be expanded to cover the phenomena associated with these other highly connected modules, or should the concept of mood itself be abandoned as meaningless?

Debates about the status of folk psychology often pit elimination against validation as if all of folk psychology will have to undergo the same fate. Either every bit of folk psychology will be eliminated, or everything will be retained. This is clearly implausible. As the foregoing remarks on emotion and mood illustrate, a much more likely scenario will involve the validation and partial transformation of some folk-psychological concepts (such as emotion) and the elimination of others (such as mood). Evolutionary psychology offers a cogent outline of what a future scientific psychology will look like. It therefore offers a principled way of exploring the debate about the future of folk psychology, one that transcends the mere intuition-trading that has characterised this debate in the past.

Acknowledgements

Dominic Murphy and Stephen Stich gave me permission to quote from their unpublished paper, which will shortly be published in an edited collection of papers. Bruce Charlton, Nicholas Humphrey, Geoffrey Miller, Randolph Nesse, Keith Oatley, David Papineau, Dan Sperber, Lewis Wolpert and John Worrall all read and commented on a previous draft of this paper. Helena Cronin provided references and lots of inspiration.

References

Atkinson, R. L., R. C. Atkinson, et al. (1996). Hilgard's Introduction to Psychology. Twelfth Edition. Fort Worth, TX, Harcourt Brace College Publishers.

Baron-Cohen, S. (1995). Mindblindness: An Essay on Autism and Theory of Mind. Cambridge, MA & London, MIT Press.

Darwin, C. (1872). The Expression of the Emotions in Man and Animals. Chicago, University of Chicago Press, 1965.

Davidson, R. J. (1994). 'On emotion, mood, and related affective constructs' in The Nature of Emotion: Fundamental Questions. P. Ekman and R. J. Davidson. New York & Oxford, Oxford University Press: 51-55.

Ekman, P. (1992). "An argument for basic emotions." Cognition and Emotion 6(3/4): 169-200.

Ekman, P. (1994). 'Moods, emotions, and traits' in The Nature of Emotion: Fundamental Questions. P. Ekman and R. J. Davidson. New York & Oxford, Oxford University Press: 56-58.

Ekman, P. and W. V. Friesen (1971). "Constants across cultures in the face and emotion." Journal of Personality and Social Psychology 17(2): 124-29.

Fodor, J. A. (1983). The Modularity of Mind: An Essay on Faculty Psychology. Cambridge, Mass. and London, MIT Press.

Freud, S. (1926). Inhibitions, symptoms and anxiety. Harmondsworth, Penguin.

Frijda, N. H. (1994). 'Varieties of affect: emotions and episodes, moods, and sentiments' in The Nature of Emotion: Fundamental Questions. P. Ekman and R. J. Davidson. New York & Oxford, Oxford University Press: 59-67.

Griffiths, P. E. (1990). 'Modularity and the psychoevolutionary theory of emotion' in Mind and Cognition: An Anthology. W. G. Lycan. Oxford, Blackwell, 1999.: 516-529.

Griffiths, P. E. (1997). What Emotions Really Are: The Problem of Psychological Categories. Chicago & London, University of Chicago Press.

Gross, R. (1996). Psychology: The Science of Mind and Behaviour. Third Edition. London, Hodder & Stoughton.

Hagen, E. (1998) The defection hypothesis of depression, unpublished manuscript.

Hughes, J. (1991). An Outline of Modern Psychiatry: Third Edition. Chichester, John Wiley.

La Barre, W. (1947). "The cultural basis of emotions and gestures." Journal of Personality 16: 49-68.

Lazarus, R. S. (1991). Emotion and Adaptation. New York, Oxford University Press.

LeDoux, J. (1998). The Emotional Brain: The Mysterious Underpinnings of Emotional Life. London, Weidenfeld & Nicholson.

McGuire, M., F. Fawzy, et al. (1994). "Altruism and mental disorders." Ethology and Sociobiology 15: 299-321.

Morris, W. N. (1989). Mood: The Frame of Mind. New York, Springer-Verlag.

Morris, W. N. (1992). 'A functional analysis of the role of mood in affective systems' in Review of Personality and Social Psychology 13: Emotion. M. S. Clark. Newbury Park, CA, Sage: 256-293.

Murphy, D. and S. Stich (1998). "Darwin in the madhouse: Evolutionary psychology and the classification of mental disorders (unpublished manuscript)." .

Nesse, R. (1990). "Evolutionary explanations of emotions." Human Nature 1(3): 261-289.

Nesse, R. (1992) More on the mood-emotion distinction, Psycoloquy 92.3.7 http://www.ai.univie.ac.at/archives/Psycoloquy/1992.V3/0006.html.

Nowlis, V. (1963). 'The concept of mood' in Conflict and Creativity. S. M. Farber and R. H. L. Wilson. New York, McGraw-Hill.

Panksepp, J. (1982). "Towards a general psychobiological theory of emotion." Behavioral and Brain Sciences 5: 407-67.

Parkinson, B., P. Totterdell, et al. (1996). Changing Moods: The Psychology of Mood and Mood Regulation. London & New York, Longman.

Plutchik, R. (1980). Emotion: A Psychoevolutionary Synthesis. New York, Harper and Row.

Price, J., L. Sloman, et al. (1994). 'The social competition hypothesis of depression' in The Maladapted Mind: Classic Readings in Evolutionary Psychopathology. S. Baron-Cohen. Hove, Psychology Press: 241-253.

Samuels, R. (1998). "Evolutionary psychology and the massive modularity hypothesis." The British Journal for the Philosophy Science 49: 575-602.

Schildkraut, J. J. (1965). "Catecholamine hypothesis of affective disorders: a review of supporting evidence." American Journal of Psychiatry 122: 509-22.

Segal, G. (1996). 'The modularity of theory of mind' in Theories of Theories of Mind. P. Carruthers and P. Smith. Cambridge, Cambridge University Press: 141-157.

Sperber, D. (1994). 'The modularity of thought and the epidemiology of representations' in Mapping the Mind: Domain Specificity in Cognition and Culture. L. A. Hirschfeld and S. A. Gelman. Cambridge, Cambridge University Press: 39-67.

Tooby, J. and L. Cosmides (1990). "The past explains the present: emotional adaptations and the structure of ancestral environments." Ethology and Sociobiology 11: 375-424.

Tooby, J. and L. Cosmides (1992). 'The psychological foundations of culture' in The Adapted Mind: Evolutionary Psychology and the Generation of Culture. J. L. Barkow, L. Cosmides and J. Tooby. New York, Oxford University Press.

Watson, P. and P. Andrews (1998). An evolutionary theory of major depression. Tenth Annual Meeting of the Human Behavior and Evolution Society, Davis, CA, July 8-12, 1998.

Word count: 5014 words including captions and references.

1 Dan Sperber has pointed out that Darwinian modules need not always be instantiated by some fixed neuroanatomical structure, but may be realised in other ways, such as by some chemical pathway (personal communication). On this view, the catecholamine hypothesis would not count as evidence against the modularity of mood.

From moods to modules February 1999 Draft Version

Dylan Evans Please do not quote without permission

1

10