Материал: The Routledge Companion to the Study of Religion by John Hinnells

Religion and cognitionâ 531â

Religious ideas

If ordinary human actions are predicated as ‘religious’ by claims to superhuman agents, then the basis for such claims must themselves be accounted for. The anthropologist Pascal Boyer has argued that representations of superhuman agency, documented from virtually every human society, are readily and easily produced by our ordinary cognitive equipment and are, consequently, as ‘natural’ as are the actions they predicate (Boyer 2001). For example, ‘agents’ as self-motivating, intentional objects in the world are readily distinguished from inanimate objects even by infants. Further, a surprisingly large number – perhaps a majority – of all young children claim to have imaginary friends. Such individuals, like those who are generally able to recognize but still enjoy fantasy (as fantasy) as adults tend to be particularly capable in their social understandings and abilities (Taylor 1999) – an attribute often associated with religion (e.g. Durkheim 1915). This innate – or at least developmentally early – ability to detect and/or imagine agency is generalized as a tendency to represent all objects in our environment anthropomorphically, i.e. in terms of human features and attributes, and, likewise, all occurrences in our environments as intentional (Guthrie 1993). Anthropomorphic representation is such an exquisitely tuned feature of our cognitive processing that we tend to conclude that there is agency all around us (which of course there is) but even when no agent may actually be present (e.g. faces in the clouds, bumps in the night, etc.). There is, of course, a survival advantage for any organism to be able to react reflexively to ambiguous information from its environment, such as a fleeting perception of movement, since this information may indicate the presence of a predator or foe. Even if it turns out, upon reflection, that the inferred presence was that of a friend or even incorrect – a blowing in the wind, as it were – ‘it is better to be safe’, the old adage holds, ‘than sorry’.

There is, in other words, little cognitive difference between imputing agency to unambiguous data and to ambiguous data, such as movement, intention, certain shapes, etc., especially when those data are deemed to be potentially relevant for our lives, as a possible indication of the presence of danger, for example. And if otherwise ambiguous events are judged potentially significant for our lives, again whether those effects are positive or negative, it is also ‘natural’ to conclude that they may have been intentionally instigated. The human brain, of course, seeks explanations for such intuitive responses to environmental cues and they typically become rather quickly judged as fact or fiction. For others, such intuitive responses provide the occasions for metarepresentation, reflection and intellectual ‘rationalization’.

The category of agency belongs to what cognitivists refer to as our ‘intuitive ontology’, that is to say, to our ordinary expectations about the world. Thus, when any information is represented in terms of or as the effects of agency – whether actual or presumed, whether superhuman or not – a great deal of information is inferred from these expectations apart from any specifically learned knowledge. Such ordinary agency expectations include, for example, intentionality, self-movement, some form of metabolism, response to external stimuli, etc. In the absence of complete information, such ontological categories as ‘agency’ are sometimes ‘violated’. A common example is ghosts which are generally represented in terms of ordinary agents – they act and react in terms of ordinary sensory stimuli such as light, sound, smell, touch, they exist in time and hold memories of the past, they communicate and can be communicated with, etc. However, they are also represented as possessing a few characteristics that violate our ordinary expectations about agents, such as being capable of invisibility or of passing through physical barriers. Whereas such claims about ghosts violate

532â Religions in the modern world

ordinary expectations about agents, they are not so excessive as to be judged bizarre, like the Godzilla of Japanese film (even though they might still be enjoyed as popular diversion), or simply dismissed, at least by most (as are, for example, claims to the existence of aliens who abduct earthlings for titillating sexual experimentation). Rather the more minimalist violations, exemplified by ghosts, prove to be attention-grabbing and, consequently, highly memorable and readily transmissible while being, at the same time, ordinary enough to be readily understood and, thereby, easily accepted (Boyer 2001). Most of the Christian Bible, for example, contains a collection of rather mundane stories – genealogies, family intrigues, accounts of kings and battles, insightful but unexceptional teachings, etc. – rendered attractive and memorable, however, by their association with claims to miraculous ‘acts of God’ (or of His Son). It is these extraordinary acts that reportedly attracted the attention of the first Christians (e.g. The Gospel according to John 2:11; 6:12; Acts of the Apostles 2:22; the Epistles to the Romans 15:18–19 and to the Hebrews 2:3– 4) and that retain the attention of modern Christians, many of whom admit never having read most of the ‘ordinary’ portions of the Bible at all.

In addition to agency, cognitivists also refer to intuitive categories of ‘substances’ or ‘physical objects’, both natural and man-made, of ‘animals’ and ‘plants’ (Boyer 2001; Atran 2002). By investing any of these ordinary categories with some qualities that defy expectations, attention is drawn to the information embedded in or associated with them and that information tends, thereby, to be considered more valuable than others in the marketplace of possible human ideas and, consequently, selected for and transmitted.

Religious persistence

The original effects to which superhuman agency have been attributed often prove to be historically inaccessible or, if known, of little significance – that is to say, any number of ambiguous possibilities can evoke conclusions about and representations of superhuman intentionality. Is the hearing of voices, for example, to be interpreted as a divine call or as an auditory illusion? Are feelings of exaltation an indication of spirituality or of having a particularly good day? Whereas naturalistic explanations for such experiences garner little attention, their interpretations as ‘religious’ prove to be attention grabbing and, once introduced, are readily transmitted in predictable ways. Cognitivists are interested in these modes of transmission.

The cognitive anthropologist Harvey Whitehouse has identified two divergent modes of religious transmission which he terms ‘imagistic’ and ‘doctrinal’ (Whitehouse 2004). The ‘imagistic mode of religiosity’ does not refer, in Whitehouse’s description, to religious traditions that trade in images – a trait of virtually all religions. Rather, ‘imagistic’ is Whitehouse’s designation for a mode of religiosity whereby religious knowledge is transmitted through infrequently performed rituals that – like Lawson’s and McCauley’s special agent rituals

– are rendered especially memorable through intense sensory pageantry and heightened emotionality. The dramatic, often traumatic, character of these rituals typically occasions a spontaneous exegesis of that experience by its participants as well as an enduring cohesion among them that is occasioned by the shared rigors of the ritual regimen and that result in closely-bound, face-to-face communities.

In contrast to the spontaneous exegetical reflections evoked by the emotionally salient rituals of the imagistic modality, religious knowledge in the doctrinal mode is formulated as a coherent set of shared beliefs or doctrines maintained by a strong, hierarchically organized

Religion and cognitionâ 533â

leadership. Such coherently formulated sets of orthodox teachings allow for their ready transmission by authorized teachers and missionaries and for the relatively faithful reception and retention of such routinized catechetical instruction by followers. This widespread distribution of and adherence to a shared corpus of religious knowledge is characteristic of large, imagined communities in which individual affinities are mostly anonymous – large Protestant denominations, for example. While this doctrinal modality may be found in nonliterate contexts, it is more often characteristic of literate societies or of those influenced by them.

The two modes of religious transmission proposed by Whitehouse rely on and are constrained by different systems of memory that are invoked by the different forms of ritual practice. The catechetical instruction in and the repetitive reinforcement of beliefs that are characteristic of the doctrinal mode of religiosity – and that are reminiscent of the encoding of memory associated with the repetitive character of special patient rituals

–become encoded in the explicit memory system as generalized schemas of knowledge. The personalized experiences and exegeses characteristic of the imagistic mode are, on the other hand, encoded in the episodic or autobiographical memory system, the contents of which are only recalled when presented with stimuli associated with an individual’s own participation in a particular event. This remembered material is organized (and transmitted) in terms of those personal associations and not in terms of any shared largescale belief system.

A particularly salient type of episodic memory, sometimes referred to as ‘flash-bulb’ memory, often results from participation in an especially traumatic or consequential event. This effect is exemplified by the abrupt and overwhelming emotional experiences that are a feature of many initiation rites both ancient and modern, e.g. initiations into a number of tribal societies, the Hellenistic mystery cults, criminal organizations or revolutionary cells, contemporary pseudo-religious fraternal groups or ‘fundamentalist’ religious factions,. Such events tend to create especially strong memories that, while incomplete, nevertheless accurately retain many details.

The cognitive anthropologist Dan Sperber has emphasized that the transmission of religious knowledge – like that of any knowledge – is from mind to mind. Such transmission inevitably involves transformations by which ‘remembered’ traditions are, at the same time, the consequence of constructive cognitive processes (Sperber 1996). This transformative inevitability is illustrated, at a non-profound level, by the children’s game known variously as ‘Chinese Whispers’ or ‘Telephone’ in which a message that is transmitted from one person to another around the room becomes transformed, sometimes radically, by the time it reaches the final participant. On the other hand, messages which resonate with innate cognitive proclivities are attracted to those mental structures and result in a certain stabilization of knowledge that may be shared and become public and that we know as ‘culture’. Any knowledge significant enough to become stabilized and publicly shared may also be considered significant enough to be inscribed and conserved in material culture as well. Such inscribings

–from the first flint tools to writing itself – provide cultural way stations for continuing mnemonic and reflexive traditions of cultural transmission and exegesis.

534â Religions in the modern world

The significance of cognitive science for the study of religion

What exactly can a cognitive science of religion contribute to the study of religion that has otherwise been lacking? Cognitive science cannot, of course, explain all religious data. While, for example, cognitive science has little to say about the meanings claimed for specific cultural constructions, it can explain the ubiquity of religion among virtually all human societies, past and present. It can offer naturalistic explanations for recurring patterns that have long been noted among the diversities of religious expressions. It can offer explanations for the modes of conservation and transmission employed by those particular constructions and for individual commitments to them. And it can express these explanations with some precision in ways that may be assessed from the wealth of ethnographic and historical data controlled by scholars of religion. For example, the previously discussed cognitive theories of religious behavior, of religious ideas and of religious persistence have all been, and continue to be, systematically assessed by anthropologists, archaeologists and historians. Results of this research to date broadly confirm the predictions of cognitive theories of religion (Barrett 2004; Whitehouse 2004; Whitehouse and Laidlaw 2004; Whitehouse and Martin 2004).

In addition to proposing specifically cognitive theories of religion, the cognitive sciences can also contribute to three issues in the larger study of religion. They can help to stipulate the kinds of data that might be included – and excluded – from such an area of study, they can provide a framework for organizing and evaluating the history of religions, and they can offer a non-ethnocentric basis for comparing religions.

Defining ‘religion’

A comprehensive definition of religion – and consequently the focus and scope of its study – has long been debated. Proposals for such a definition have ranged from those with parochial (theological or confessional) biases, to those with a universalizing but still quasi-religious (‘sacred’ or ‘spiritual’) basis, to those shaped, however unintentionally, by Western conceptual categories (such as philosophical dualism) and/or political policies (colonialism). On the other hand, some functionalist definitions of religion (such as ‘ultimate concern’) are so broad as to include virtually anything and exclude nothing. Some recent scholars have even conceded defeat in the definitional endeavor and advocate collapsing the study of religion into that of culture(s), posing then, of course, the even more daunting task of defining ‘culture’, or they simply disregard the issue altogether and pursue their work without any explicit delineation for the data their work might include. As with conceptions of ‘religion’ as a natural kind that might provide an object for natural selection, there is, of course, no natural ‘thing’ as ‘religion’ in the world for which a ‘correct’ definition might be agreed. ‘Religion’ is, however, no less susceptible to definition as an analytic category than are other domains of culture such as ‘economics’ or ‘politics’. Analytic categories, in contrast to categories whose contents refer to natural kinds in the world, must be theoretically stipulated in a clear and explicit manner (which is not to say that referential categories don’t present their own theoretical problems) and, consequently, be subject to assessment of their validity and utility rather than simply being idiosyncratically asserted or confessed.

What counts as ‘religious’ data can be stipulated from a cognitivist perspective, as we have seen, as those ordinary behaviors and mental representations that are, however, legitimated

Religion and cognitionâ 535â

by claims to the authority of superhuman agents. This definition, adapted from E. B. Tylor’s classic ‘minimum definition of religion’,1 has the advantage of stipulating what religion is not. Ideologies such as Marxism or world views such as Freudianism, for example, are excluded from considerations as religion, as are those patterned, repetitive human acts characteristic of such sports as football and often analyzed as ‘religious’ ritual. Whatever the functional similarities to religious ideas and practices that may be exhibited by such cultural expressions, they make no claim upon superhuman agency.

Some may object that ‘legitimating claims to the authority of superhuman agency’ as a stipulation for what counts as ‘religion’ excludes certain forms of ‘atheistic’ religious thought, such as Buddhist, Taoist or Confucian. However, anyone with minimal experience ‘in the field’– even as tourists – will recognize that the actual practices of the overwhelming majority of participants in such traditions involve an acknowledgment of and devotion to superhuman agency. Such positions of ‘religious atheism’ are espoused by a very small number of intellectuals in these traditions. In fact, cognitivists have demonstrated that a dissonance between intellectual formulations and actual practice is a common feature of religions (Barrett 2004; Slone 2004). Intellectual representation of a deity as omniscient, for example, does not negate a confessor’s impulse regularly to convey information to that deity through prayer.

Further, the minimum definition of religion differentiates ‘religious’ behaviors and concerns from those associated with other social systems, whether or not such differentiations are made emically. Since ‘religious’ and ‘political’ systems, for example, both claim legitimacy by appeal to the power of authority, the one is often embedded in the other, as was the case with ancient Greece and Rome. And whereas ‘religious’ systems are often virtually identified with moral traditions, especially by biblically-based religions, this is not necessarily the case for other traditions, as is again exemplified by ancient Greece where representations of deities exhibited wide moral latitude in contrast to the ethical authority of the philosophers. Rather, evolutionary biologists have argued that distributions of power and systems of morality are elaborations upon and codifications of evolved behavioral tendencies, such as relationships of dominance-submission, reciprocal altruism, a concern and ability to detect cheaters, etc.

In addition to Tylor’s minimal definition of religion, an additional ‘Durkheimian’ caveat stipulates that religious representations are those that are ‘costly’ (in terms of resources, time, labor, cognitive effort, etc.).2 When, in other words, social elaborations of ordinary concerns and behaviors are legitimated by costly appeals to superhuman authority, we may consider them to be ‘religious’

Still, questions for further research are posed by stipulating ‘religion’ as costly claims to the authority of superhuman agency. How are those superhuman agents deemed to have certain attributes worthy of that costly behavior within a particular culture to be distinguished from the proliferation of superhuman agents that are ‘naturally’ produced by the mind but held within that same culture to be insignificant? How are gods to be distinguished from figures of folklore such as fairies or trolls? How are the ‘true’ (culturally accepted) deities to be distinguished from ‘false’ gods, from newly revealed deities or from those imported from another cultural context? Are there cognitive predispositions for such valuations and differentiations or are they simply a matter of sociohistorical construction? But how then do we explain those constructions?

Nevertheless, the stipulation of religious data as those that are legitimated by claims of superhuman agency which result in costly behavior emphasizes that the study of religion requires no privileged approach or method but rather is the study of ordinary human