The sacred as a religious concept is inseparably linked with the linguistic conventions of Western societies, the roots of which go far back into the history of both Indo-European and Semitic cultures. Long before its conventionalised use and meanings within the Jewish, Christian and Islamic religious traditions, people have participated in sacred-making activities and processes of signification according to paradigms of thought created by their ethnic systems of belief within specific geographical limits. The sacred is not only a religious term, but also an anthropological constant, which has been used in various cultural contexts within various arenas of human discourse. Linguistically the sacred denotes that which has been set apart, but in religious vocabulary there is usually an ontological referent underlining its use and culture-specific semantics.

In my presentation I shall look at the notion of the sacred from a cognitive point of view. It is my conviction that scholars engaged in the scientific study of religion have blurred the boundaries of theological and scientific discourses in adopting the sacred as a superordinate category for religious worlds. From a cognitive perspective, the scholarly approach to the idea of the sacred does not entail metaphysical or religious questions about the nature of reality. Rather, the focus is on the cross-cultural regularities that have guided the perception of an object as sacred in a particular specific linguistic community in a particular geographical context. In this enterprise the student of religion has to become acquainted with the history of religions and with the multiple discourses in which terms denoting the «sacred» in various languages have been used. Most religious scholars have relied on their knowledge of the history of the concept in Hebrew, Greek and Latin sources, but have forgotten that the concept has a long history as a linguistic term in the vernacular of ethnic cultures and their languages; its root universally denotes «to cut», «to set apart», «to mark off (see Paden 1991; 1999; Lutzky 1993; Anttonen 1996).

In postmodern Western societies, there are millions of people who no longer accept the inherited religious traditions of their parents and ancestors as a grand theory for their lives. In Europe, in both Protestant and Catholic countries the number of people that are not affiliated with any institutional form of religion has grown (see Dobbelaere 1993). The old religious structures have become desacralised and new or unconventionally defined forms of religious sacralisation invented. People have greater intellectual and moral freedom to create their own «sacred» moments within their secular cosmology, by setting apart specific times, places, events and persons and marking their significance by specific symbolic means. Universally distributed forms of behaviour in which the idea of personal choice plays a crucial role, such as fasting, asceticism, celibacy, and in various forms of performance in the worlds of sport or art, can be comprehended in terms of the category of the sacred. These forms of behaviour are culturally based on the idea of using specific ritual strategies to mark one's physical and mental self as separate from the routines of everyday social life (Anttonen 1999a).

An analytical comprehension of their sacred-making quality, however, cannot be approached within the conceptual frames or paradigms of thought offered by established religious institutions. The present situation is comparable to prehistoric times before the rise of organised religions. From the viewpoint of the archaeology of the sacred, the earliest evidence of a human capacity to sacralise and confer religious meaning on various objects and phenomena in their natural and social environment can be read from the soil and from the landscape. Both archaeological findings and toponyms show that human beings have had cognitively fluid minds (see Mithen 1996) in setting apart marshes, springs, rapids, ponds, lakes, capes, bays, territories such as forests, wilderness, larger hills and mountains, or specific topographically anomalous sites, and have observed their special quality by specific ritual strategies. As an adjective, «sacred» has been used as an appellative designation for a place, for a specific topos, or for a specific period in reckoning time in order to mark a categorical boundary. The sacred has been used as an attribute whereby distinctions have been expressed between those things that possess a special cultural value and those that do not demand particular attention or specific rule-governed behaviour. In dealing with the theory of religion, the scholar with a social-scientific orientation needs a special explanatory perspective in order to display the logic governing cross-cultural regularities in setting something apart as sacred. There are denumerable entities - visible and non-visible, physical and non-physical - that have been perceived as sacred. Some of them refer explicitly to superhuman agents, while some are less overtly religious, for instance the marks such as the flag or soccer matches by which members of a nation-state enhance or preserve its value (see Smart 1985: 21-28; Bell 1997: 153-159).

In ethnographic study of religion the sacred is not as much a metaphysical enigma as an issue of epistemology. In an attempt to analyse various forms of manufacturing religious realities (cf. McCutcheon 1997; Pyysiainen 1996), the ethnographer of religion does not need to approach the discourses of religious persons from the explicit theological content of their speech acts. Rather, the ethnographer takes another route, starting from the actions, events and intentions of cultural agents in specific contexts as they make distinctions between spaces, mark them for specific uses, create visible and invisible boundaries, and establish cultural conventions of behaviour towards those boundaries (see e.g. Parkin 1991; Anttonen 1996: 39-43; 1999b).

The ethnographer occupies the driver's seat in explaining why people in all periods and places participate, in one form or another, in the cultural practices that «reasonable observers would agree is religion» (Rappaport 1999). In his posthumously published book Ritual and Religion in the Making of Humanity (1999), Rappaport provides the broadest possible context for explaining why people have spent so much time, energy, wealth and blood in building temples, supporting priests, sacrificing to gods and killing infidels (Ibid.: 1-2). His central question is the sacred, the numinous, the occult, the divine and their fusion into the Holy in ritual. In Rappaport's approach the sacred is just one constituent of the «total religious phenomenon» (1999: 24). The sacred signifies the discursive, logical, intelligible component of religion, which is implicated in the liturgy of ritual specialists, but which also appears in commandments and in the oaths and pledges of non-specialist participants (Ibid. 320; cf. Otto 1969: 143-145). In addition to the sacred, there is the numinous component, which denotes religion's non-discursive, affective, ineffable, inconceivable, mysterious, awesome qualities (Rappaport 1999: 23; see also Otto 1969: 36-37; Pyysiainen 1996: 51-57).

Rappaports' cybernetics of the holy owes to the scholars of religion who have held the category of the numinous to be the core element that characterises religious thought and behaviour. Rudolf Otto, Nathan Soderblom and Mircea Eliade have utilised the sacred or the holy as a category of feeling, and have conceptualised it as an inward sensation of the supramundane (überweltlich) that «evades precise formulation in words» (Otto 1969: 59).

Rappaport does not, however, appeal to sacrality to explain religion, or to religion to explain sacrality. For him sacrality is not the starting point of religion, but its end product, «a difference which makes difference» (1999: 402) as he quotes Gregory Bateson. Rappaport treats the sacred as a category the contents of which are ritually constituted as a response to maintain the adaptive flexibility of human social systems. He places the issue of sacrality in an evolutionary framework, but posits that «(n)either religion «as a whole» nor its elements will be reduced to functional or adaptive terms» (Ibid.: 2). He examines the effect of human language and increased conceptual capacities on the nature of ritual acts and objects and on adaptive flexibility. Towards the end of the book Rappaport's conclusion seems to be that the ultimate sacred postulates and their unquestionable status are losing significance. Human social systems are becoming less flexible, which mean an increasing unbalance in adaptive processes. For instance, the authority of political and religious representatives does not rest on sanctity to the same extent as it did in traditional societies; this means that regulatory structures in Western societies are no longer as flexible. Rappaport's ideas are worth keeping in mind in accounting for the emergence of oppositional social movements, both secular and religious, a topic currently in spotlight both in Europe and in the United States.

As a social-scientific category, the sacred has been semantically recontextualised as a taxonomic indicator (see Paden 1999: 166-167) for the analysis of both religious and secular systems of thought. In addition to Rappaport, anthropologist Mary Douglas has done remarkable work in developing Durkheimian notions of the sacred into risk analyses. Modifying the notion of the sacred as a collective representation whereby distinctions between notions of purity and impurity, the licit and the forbidden, have been made morally binding (see Durkheim 1995), Douglas has moved towards explaining the sacred on the basis of the cognitive mechanisms of the human mind. She posits that the idea of the sacred is based on the precariousness of the cultural categories guiding human thinking and behaviour. Douglas holds that the sacred is the universe in its dynamic aspect with inexplicable boundaries, because the reasons for any particular way of defining the sacred are embedded in the social consensus which it protects (Douglas 1978). Even though she has not aimed at creating an anthropological theory of the sacred, her analysis combines the positive and negative aspects of sacredness. She has treated the sacred as order, unity and integrity, and has also pointed out the necessity of paying attention to the taboo aspect of the sacred. Things that have been set apart as taboo, because of their impurity and contagiosity - the abject, as Julia Kristeva calls it - form the flip side of the sacred. In maintaining the socially legitimated order, individuals and social collectives create symbolic-cultural systems, or even neuroses if you like, by setting apart impure objects, substances, places or times, and emphasising their cognitive status by taboo norms and rules of avoidance (see Douglas 1989; cf. also Paden 1996; Sperber 1996; Parkin 1996).

But whence the notion of the sacred? And moreover, how can the notion of the sacred be viewed as an anthropological constant? In his cognitive theorising Pascal Boyer has argued that since religion is a cultural phenomenon, it is just as culture in general constrained by the human cognitive capacities. Every domain of the repertoire by which religion is characterised - ontology, morality, group or ethnic identity, religious action and private experience - includes the same cognitive constraints, which also characterise non-religious domains. Boyer's cognitive theory of religion is based on counter-intuitive assumptions in which expectations from one ontological category are transferred to another. If solid objects such as stones and artefacts have the psychological property of intention and free will characteristic to living human beings they violate inferential principles of intuitive ontology. As such they are more likely to be dealt with as representations belonging to the domain of religion (Boyer 1998; 1999; Pyysiainen 1999). Boyer posits that sufficiently «counter-intuitive» beliefs are adopted more easily and transmitted more effectively because they are more easily remembered and intuitively plausible (Boyer 1994, 1999). The notion of counter-intuitiveness can, however, be viewed as one instance in the broader discussion of anthropological theory-construction concerning human perception and the nature of the categories whereby reality is grasped. Violation of ontological expectations does not explain the attribution of sacredness to various «attention-grabbing» phenomena. Not all counter-intuitive representations of visible or non-visible entities whose biological or psychological properties become twisted in our minds have religious potential.

How, then, religious categories become generated and how do we develop the category of religion to fit with both the cognitive and the social approach to cultural materials? As Pascal Boyer has argued, religion is acquired through social interaction and is based on the cultural transmission of domain-specific principles. This takes place in a neutral mode, which means that there are no genetic constraints that limit or determine human capacity to learn those principles on which the intuitive ontology is built. As Boyer posits, religion is based on inferential capacities that evolve in human beings as a normal outcome of cognitive development (see Boyer 1998: 879). Boyer's argument is that for religious ideas to exist it only takes the regular inferential capacity. But to explain how specific religious traditions evolve from the basic cognitive properties of human mind, scholars of religion need to put counter-intuitiveness in objects and phenomena into wider theoretical frameworks in which the cognitive is connected with the cultural. In reference to my theory of the sacred, I am tempted to suggest that counter-intuitiveness in objects and phenomena need to be viewed as interfaces whereby the boundaries separating visible and invisible, physical or non-physical categories become represented and how they are ritually transcended. For being religious, counter-intuitive objects, persons and phenomena need to have a specific cognitive and normative status and become ritually manipulated in the controlled situation of ritual and in connection with culture- and context-specific values that are individually, socially, economically or politically significant.