Archaeology of religious change: introduction

This issue explores archaeology's contribution to the study of religious change, transmission, interaction and reception. While the study of how certain religious traditions move into new areas and relate to pre-existing religious, cultural, political and economic structures has been dominated by sociology, anthropology and comparative religion, archaeology has made significant contributions to the field. The aim of this volume is to bring together recent field-based research on the material correlates of religious change. Of particular interest are those studies which look beyond the traditional ritual-based focus of religious change, to its wider economic, political or ‘practical’ ramifications.

The resulting papers encompass a broad chronological and geographical scope, ranging from the fifth millennium BC to the sixteenth century AD, and including case studies from Australia, the Indian subcontinent, South America, Scandinavia, Spain and northern England. Eight out of a total of ten papers deal with three of the major ‘religions of the book’, Christianity, Buddhism and Islam, and their interaction with pre-existing traditions; the remaining two deal with the origins of prehistoric religions in northern Europe (Bradley and Numara), while Eeckhout focuses on early Peruvian traditions prior to European contact.

Religion as ‘worldview’

Many of the contributors have problematized the narrowness of the term ‘religion’ as an etic, and culturally specific, concept which fails to encompass adequately the diverse levels of human existence into which its influence extends. Graham et al. stress that prior to European contact the Maya had no word for a separate entity that was equivalent to such a term and that, rather than ‘religion’, we should speak of the Maya ‘worldview’: religious change is actually change in worldview, which in this context includes decisions over warfare and the treatment of the dead. They argue that many aspects of Christian ritual were not too unfamiliar to the Mayas, but that what was novel to them was the Spanish attitude towards warfare and in particular the religious sanctioning of killing ‘under the umbrella of war’. The fundamental contradictions between Christian and Maya concepts of victory, the former being based on mass killing, the latter on the taking of captives, led to the Spanish maintaining the upper hand as the Mayas were unable to ‘win’ without compromising their fundamental worldview. The eventual changes to Maya worldview were thus intricately linked with changes in their warfare tactics as well as burial customs, both of which came to reflect Christian/Spanish traditions.

What we mean by religion is perhaps easier to define in relation to the three ‘founder’ religions, which figure in some way or other in most of the studies here: Christianity (Andren, Graham et al., Lund, Äikäs and Salmi, Perring, Hiscock), Buddhism (Shaw) and Islam (Carvajal), all of which can be framed within a specific set of textually sanctioned theological precepts, with a clearly defined origin in both time and space, but, even in such contexts, the ramifications of transmission are not always as ‘religious’ or, indeed, comprehensive as one might expect. The consequent emphases on the ‘social’ or ‘practical’ outcomes of religious change, and the varying and often unexpected outcomes of ‘conversion’ are discussed in more detail below. Further, the term ‘religion’ does not always find its direct equivalent across the three examples that figure here. In India, for example, although the Sanskrit term yoga (related to the proto-Indo European ‘yewg’ = to join, from which the English ‘yolk’ is also derived) is close enough to the root meaning of ‘religion’ (Latin ‘religare’ = to bind, as discussed by Graham et al. in this volume), the intertwined concepts of dharma (literally ‘law’) and jati (caste) are the main shaping factors for individual and collective ‘worldview’ and ‘identity’. As discussed by Shaw, the Buddha's reworking of orthodox notions of dharma flew in the face of orthodox ‘religious’ frameworks that had hitherto upheld the fabric of society, but his message was also a deeply social and political one that for the first time allowed for self-elective group membership that operated outside the confines of caste and family. It is not only for those traditions, such as the Maya (Graham), that had no term for a separate religious sphere that we need to acknowledge the fluidity between various dimensions of human existence, and the dangers of defining major periods in world history according to neat religious categories such as Muslim, Buddhist, Hindu, Christian and so on.

Lund, in her study of Christianization in Viking Age Scandinavia, argues that religious change did not refer only to ritual action but also to more general changes in worldviews, ‘cognition and mentalities’, and more particularly to changing attitudes towards personhood and the relationship between the living and the dead. For example, the shift from fragmentary burials to complete inhumation reflected not only religious conversion, but also new attitudes towards the integrity of the whole body and objects in burials as well as in life. Further, while such changes in burial customs reflected the theological impact of Christianization, they were also the results of the adaptation and transformation, rather than obliteration, of Viking Age customs, with very little evidence for the direct transplantation of customs from Christian Britain.

Carvajal, in his study of Islamicization in the Vega of Granada, argues for a social definition of Islam, which acknowledges not just the religious or theological aspect of conversion but also the broader changes that led to the transformation of social practices and the development of an Islamic ‘culture’ that transcends purely religious elements.

The shortcomings of an etic approach to religion are also highlighted in Äikäs and Salmi's study of the interaction between Christianity and Saami worldviews in Fennoscandia between the eleventh and twenty-first centuries. They challenge received models of religion that do not incorporate varying notions of worldview and that present religion as a closed entity disconnected from other aspects of life.

Conversion, syncretism or something else: how ‘successful’ can religious change be?

The idea that religious change constitutes a clear-cut and comprehensive replacement of pre-existing beliefs and ritual practices is challenged by many of the papers in this volume. While syncretism has often been posited as a nuanced alternative to ‘conversion’, the concept of syncretism is problematized by Äikäs and Salmi on account of the fact that it usually refers to an amalgamation of religious traditions based on the Western idea of religion as a separate sphere of life. Moreover they argue that a form of syncretism was part of Saami tradition through long-distance trade from at least six hundred years prior to the formal establishment of Christian missions during the seventeenth century. They argue that in the post-Christianization era the prevailing model wasnot syncretism but rather ‘dual religious participation’: Christianity did not replace old Saami ways nor did it merge with them. Rather, the Saami worshipped both old and new gods for different reasons, resulting in a ‘double consciousness’ of religious practices. Similarly, through the ‘transformation’ of the old custom of leaving gifts at offering stones, older customs, the Saami now offered gifts to the church, in order to receive help with certain social problems, but they also continued to leave offerings at the neighbouring Saami site. Here we can draw parallels with anthropological accounts of Christianization in Nigeria where a ‘medical’ analogy has beenused to describe the way in which indigenous gods continued to be worshipped for specific aims (Peel 1968; also Goody 1975): just because one takes a new cough medicine, it does not mean one has to stop taking traditional medicines for other ailments.

Shaw's study represents a variant of the ‘medical’ analogy, her argument being that Buddhism in early India was not overtly concerned with conversion. Being a monastic tradition, it operated within a two-tier religious structure whereby the laity's ritual identity often remained unaffected despite its financial support of the monastery. Despite canonical attempts to stress the ‘ritual’ incentives for such exchange networks, there were clearly additional motives, including access to agrarian, medical and educational resources which, in addition to being instruments for lay patronage, offered practical means for the alleviation of suffering, a quest that formed the central tenet of the Buddhist worldview.

Lund also challenges the idea of a full-blown, comprehensive application of the Christian European package to the Scandinavian way of life, arguing that there was never a one-way progression to Christianity but rather a gradual ‘transformation’ of worldviews as represented in changing burial customs.

Monumental v. ‘natural’ sites

Despite the long enduring Cartesian polarization between ‘natural’ and ‘cultural’ zones in the landscape, the importance of ‘natural places’ for understanding long-term continuities is now generally accepted in archaeological circles, with some suggestions that certain British Neolithic megaliths represented ‘monumentalizations’ of natural locales already established during earlier periods (Bradley 1999; Tilley 1994, 1996). Since the 1990s, a growing level of conflict between the interests of governmental organizations and indigenous groups in Australia and the USA has led to the rise of new legislative measures to protect places of cultural significance which lack a ‘monumental’ dimension (Layton 1999; Smith 1999). The recognition of such ‘associated landscapes’ by the World Heritage Committee in 1993 (Carmichael et al. 1994) was a much-belated step towards addressing the deep-seated division between archaeological and ‘local’ perceptions of the landscape.

In this issue, Äikäs and Salmi's study of Saami offering stones in Fennoscandia offers a useful example of how human activity at such ‘natural places’ can be examined archaeologically, through the excavation of offering deposits, chemical analysis of offering stones and reference to literary and oral evidence. Without the latter the cultural significance of such sites is easy to miss. Bradley and Numara in their study of religious change in Bronze Age northern Europe also focus on ‘natural’ rocks in the landscape, although endowed with more obvious archaeological significance by virtue of their associated carvings, the content of which they compare with motifs on bronze vessels. Shaw, in her study of Buddhist propagation in central India, also draws on ‘natural’ sites in the landscape such as rock-shelters and topographical features to understand the ways in which Buddhist ‘geography’ unfolded in time and space.

Static v. dynamic religions

The traditional polarization between ‘unchanging’ and dynamic religions is a recurrent theme in this issue. Eeckhout, with reference to coastal Peru, draws on Lévi-Strauss's theory of warm v. cold traditions to challenge the view of a straightforward polarization between unchanging native traditions and dynamic traditions associated with textually attested religions. Similarly, Äikäs and Salmi challenge the idea of an unchanging, static Saami religion prior to the arrival of Christian missions in the seventeenth century. Lund also discusses the issue of diversity versus homoegeneity, arguing that Viking Age worldviews as expressed through burial customs did not exist in an unchanging vacuum before the formal introduction of Christianity in the eleventh century. Rather, influences from Christian Britain and Europe were discernible in preceding periods while the diversity which is evident during this earlier period continues in post-Christianization periods.

Hiscock also presents evidence in Southern Australia for continuous change in religious expression over the last millennium, thus challenging the traditional view of Aborigine religion as stable, with the suggestion that a similar degree of dynamism applied also to ancient Australia. Drawing on evidence for axe production together with nineteenth-century European accounts, he argues that Aborigine religion was characterized by transformation and reconfiguration, in response to changing social conditions, through the constant reworking of myth and ritual whereby myths that were created in response to new rituals endured even when rituals were transformed, thus giving the impression of religious continuity.

Similar misrepresentations have arisen from models that assume a clear-cut process of religious change based on the polarization between the ‘Great’ and ‘Little’ tradition, an idea first coined by Robert Redfield (1956) and Milton Singer (1972) and applied to the Indian context by McKim Marriott (1955). This model discounts the idea that cultural landscapes are a product of multi-layered societies and characterized by dialectical relationships between the sacred geography of the ‘Great’ tradition and the local sacred landscape made up of various cult spots related to gods of locality. Another problem, especially in India, is that scholars have tended to situate the ‘Great’, textually sanctioned tradition within urban ‘sacred centres’, while the ‘Little’ tradition, which is seen as the parochialized or diminished version of the ‘Great’ is restricted to the village, traditionally the domain of the anthropologist (Srinivas 1967).

Shaw, in this volume, problematizes the clear-cut polarization between the ‘local’ and the pan-national, arguing for example that the idea of the serpent cult as the passive ‘other’ of Buddhism is as much an etic construct of the Buddhist tradition as a reflection of local reality.

Similar themes are explored by Andren who queries whether Christianization should really be taken as the first ‘Europeanization’ of Scandinavia or whether, on the contrary, it was the much earlier and longer-term incorporation of European traits that made the eventual transition to Christianity possible. Thus, according to Andren, the various dynamics in the ritual landscape are tightly bound up with issues of identity such as ‘Pagan–Christian and Scandinavian–European’.

Continuity v. change

The degree to which ritual continuity can be surmised from the appearance of archaeological continuity is discussed in varying degrees in several of the studies here (Eeckhout, Äikäs and Salmi, Andren, Shaw). During the 1970s there was a tendency in archaeology to link the two as clear-cut synonymous processes, as exemplified by early interpretations of sites such as Yeavering (Hope-Taylor 1977: 249) where an unbroken continuity of ritual tradition from the Bronze Age to the Anglo-Saxon period was envisaged, due to the fact that Neolithic causewayed enclosures had been built over by hill forts, which in turn had been replaced by Romano-Celtic temples. The conclusion was that the over-arching ritual significance of such places endured despite political or economic change. We may refer here to Glyn Daniel's statement: ‘I find it difficult to envisage why there should be a Christian occupation of some megalithic sites, unless a real tradition of their importance as special and sacred sites was carried through the period of the Bronze Age and early Iron age of barbarian Europe and into historic times’ (1972: 59).

Such a picture has had its critics, and in the words of Bradley, ‘the proponents of ritual continuity are forced to make imaginative leaps across impossibly long periods of prehistoric time, and they do so in order to support a model which is difficult to sustain as archaeological theory’ (1987; 15). A similar critique had been applied to ‘archetypal’ Eliadian models of sacred places which drew on Rudolph Otto's (1923) idea of mysterium tremendum, the sense of awe-inspiring mystery which arises between people and the ‘spirit of place’ (numen loci), or what the phenomenological geographer, Tuan (1974), called ‘topophilia’ which endures irrespective of changes in space and time. More historically and politically contextualised models which were also reactions to normative, structuralist accounts of sacred space showed influence from Foucauldian models of multiple power structures, as well as Bourdieu's (1977) theory of habitus, whereby people interact non-discursively with their environment through long processes of enculturation. Another influence was Giddens's (1979) structuration theory, which presented structures as both the ‘medium and outcome of action’. Giddens drew heavily on Hagerstrand's ‘Time Geography’, the concept of the fusion of time and space in the form of locales which, being the focus of repeated congregation, give meaning and structure to everyday life, and on Husserl's idea of the retention of past events in the present. The influence of these wider strands of social theory is evident in later studies of the ‘temporality’ or constant ‘becoming’ of archaeological landscapes (Ingold 1993; Thomas 1991) as well as the more phenomenological approaches of Barrett (1990) and Bradley (1991a). The latter studies, also referred to as ‘archaeologies of action’, stressed the interactive relationship between people and buildings, as opposed to earlier structuralist theories which highlighted the inherent and ‘essential’ meanings of the built environment. Of particular concern was the way in which the physical layout of ritual sites helped to maintain behavioural and ideological regularity, especially though the control and restriction of vision and bodily movement. More recently, archaeologists have focused on the visual dimension, seeking in particular to identify linkages between ancient sites through the element of intervisibility (e.g. Tilley 1994).

In contrast to earlier approaches, we now find long sequences of occupation and apparent ‘continuity’ being presented as ‘masks’ for a situation of continuous change in ritual significance and ‘worldview’, with an emphasis on changing spatial orientation and astronomical alignment (Bradley 1991a: 1987). A similar ‘phenomenological’ approach is taken by Bradley and Numara in this volume through a spatial analysis of rock art sites and bronze vessel find spots and their relationship to the sea and features in the landscape, together with analyses of the thematic content of the rock art and bronze imagery. Shaw also offers a kind of ‘historical phenomenology’ (as called for, for example, by Thomas 1991) by examining visual linkages between sites as mechanisms for ‘presencing’ the Buddha and his dharma in central India, drawing also on textual sanctions within the Buddhist tradition. Further historically grounded parallels to the phenomenological approach are found in Äikäs and Salmi's study of animal offerings excavated at Saami stones in Finland with particular animal species being correlated with specific zones in the pastoral landscape.

Having established that ritual continuity cannot be assumed from sites with long cultic sequences, we must distinguish between what Evans (1985: 90) calls ‘culturally cumulative’ (passive) continuity and ‘intentional’ (active) continuity, which may involve the appropriation of the past, or what Hobsbawm (1983) calls the ‘invention of tradition’: ‘a set of practices, normally governed by overtly or tacitly accepted rules and of a ritual or symbolic nature, which seek to inculcate certain values and norms of behaviour by repetition, which automatically implies continuity with the past.’ This latter type of ‘continuity’ may indicate precisely the opposite phenomenon; that of ritual hiatus, for it is well known that the need to recreate the past, or to compose what has been called ‘prestigious but fictitious genealogies’, often arises at times of crisis and change, as illustrated at Stonehenge during the mid-first millennium BC (Bradley 1991b). It is in this respect that archaeologists can benefit from access to a textual tradition which often sheds important light on such processes of appropriation.

We should also distinguish between what Michell (1982: 107) calls ‘high’ analyses of the past, which draw on continuities and similarities with the present, in order to support the type of appropriation described above, and ‘low’ analyses, which do so by distinguishing themselves from an ‘inferior’ past. This is important, since the reuse of a monument, far from professing a common recognition of the site's sacrality, may in fact reflect outright desecration, as in the case of Chinese Tibet where Buddhist Mani stones have been built into public lavatories. On another level, the construction of a mosque or church over a pre-existing sacred building can be more an expression of divinely sanctioned political domination than a reflection of either a ‘low’ or ‘high’ analysis of the past. Recent studies of Islamicization in India, for example (Eaton 2001) have suggested that the appropriation of temple wealth was a greater driving factor behind temple destruction than the more commonly cited Islamic revulsion at idolatry. It is interesting to note that the less violent method of religious appropriation, whereby certain aspects of the past are assimilated in order to facilitate conversion, also conforms to a ‘low’ analysis, although its success depends on posing as a ‘high’ analysis.

However, what Voss (1987) aptly terms the ‘folklore’ of prehistory need not always involve political or religious manipulation, but may arise through a spontaneous interplay between historical ‘fact’ and popular imagination. Here, Pryor's (1995) model of ‘establishment’ and ‘respect’ rather than ‘use’ and ‘abandonment’, is useful, as it reminds us that, although a locale may fall from the ‘archaeological limelight’, new inhabitants may continue to interact with and reinterpret it within their own cultural framework. There may therefore be no such thing as discontinuity, but, although archaeology needs to align itself sympathetically with ‘alternative’ histories, Michell's (1982: 102) assertion that ‘the one source of information that remains constant and coherent is the folklore record’ obviously does not fit with recognition of archaeology's own role in the ‘folklore’ of the past (Shaw 2000).

Any discussion of ritual continuity necessarily requires some acknowledgement of different kinds of time, as distinguished by Leach's (1961) ‘primitive/cyclical’, ‘historical/sequential’ or ‘magical/ritual time’. It has been suggested, for example, that the proponents of ritual continuity may have misunderstood the nature of ritual time, which by definition is often connected with the past. Since ‘ritual time dissolves the distinction between past and present on which historical time depends’, we cannot approach ritual continuity in the same way as settlement continuity, as we may not be comparing ‘like with like’ (Bradley 1987: 14–15).

It may be precisely because historical time is made irrelevant by religious practitioners, since the ‘eternal’ or ‘universal’ cannot be reduced to a particular temporal and geographical field, that so many ritual theories are historically or politically decontextualized. However, Bloch (1977) points out that human time is not ‘monolithic’, and, indeed, pilgrimage often involves a mediation between timelessness and time-bound sacredness, thus combining the qualities of both Eliadian ritual and Durkheimian social time. In fact, what makes ritual time such an ideal tool for the ‘invention of tradition’ is that, while arising from everyday experience, it gains its authority by alluding to being other worldly. In Bloch's words, it ‘takes on a mystified nature consisting of concepts and categories of time and person divorced from everyday experience and where inequality [is seen] as an inevitable part of an ordered system’.

In summary then, it should be stressed that no single type of continuity is ever the overarching norm. Rather, a sacred landscape is likely to be made of a complex tapestry of ‘continuities’ which may include elements of the ‘real’ endurance of place-bound sacrality and recognition of its inherent ‘power’, preserved in the material or temporal sphere or abstracted into oral history; the invention of antiquity in order to legitimize a religious or political discourse; the assimilation of ancient cult spots as part of the conversion process; all intermingled with the discourse of archaeology itself.

Returning to this volume the question of why certain places endure and others are abandoned is central to Eeckhout's study of change and permanence in coastal Peru based on the case study of Pachacamac which maintained its religious significance over a millennium despite various instances of political, cultural and environmental upheaval. Rather than being abandoned, as might be expected based on broader regional patterns, the site's enduring religious reputation led to it becoming a pan-Andean pilgrimage centre, with its belief system and its material correlates being constantly adapted and modified.

Äikäs and Salmi examine why churches were built at, or in close proximity to, older Saami offering places, arguing for a combination of factors, including appropriation, syncretism and ‘dual consciousness’. Perring in her study of the impact of the Reformation on the secular cathedral of York Minster, draws on Giddens’ theory of ‘ontological security’, the idea that old customs are retained for the sake of well-being, to explain the particular relationship between change and continuity in York. Andren deals with what he calls the ‘classic’ issue of whether Christian churches appropriated earlier pagan sites in Scandinavia, dwelling on the question as to whether we are dealing with the continuity of cultic association or with political power. Finally, Shaw examines the issue of cultic continuity or transformation as a possible factor in the placing of Buddhist hilltop monastic sites in central India, considering also the way in which the stupa and the relic cult were central to the development of a new economic framework based on the monastic control of land and natural resources.

Agency and mortuary archaeology

Bourdieu's theory of agency and habitus has also figured prominently in the interpretation of site- and object-based data, especially with respect to mortuary analysis which has long served as a focus for the archaeology of religion. Several of the papers in this volume contend that material culture is not just a symbol of something else, in this case religion, but rather an actual medium of change. Graham et al., for example, suggest that changes in burial customs were not just indicators of changing religious beliefs, but rather that, as discussed earlier, they reflected and mediated changing attitudes towards warfare and killing. Lund also challenges the idea that religious change can be read in a direct way from observable changes in burial custom, and proposes that, rather than using changes in orientation or grave goods as a kind of checklist for distinguishing between Viking and Christian burials, such changes need to be viewed as both reflections and instruments of changing attitudes towards personage and the body (also discussed earlier). Perring, in her study of the impact of the Reformation on the ordering of space at York Minster, argues that changes in material culture not only reflected but constructed social and political change that transcended the ritual and theological spheres. Shaw also draws on elements of mortuary archaeology in her discussion of the Buddhist stupa and the relic cult, both of which challenged orthodox conceptions of purity and pollution and whose spread was central to the creation of a pan-Indian Buddhist geography, worldview and identity.

Sacred versus secular landscapes

There has been a general recognition, in both anthropological and archaeological circles, of the arbitrariness by which spaces and landscapes are divided into ‘ritual’ and ‘profane’ or ‘domestic’ spheres. In archaeology, this polarity was until recently particularly firmly entrenched, because of the subject's traditional site-based focus of enquiry and the general lack of investigation into the trajectory of ritual into other, less tangible areas of experience. In anthropology, the prevalence of Turnerian models meant that ritual was often presented as operating in a depersonalized, detemporalized present, disconnected from wider social or political currents. In later anthropological literature there is a general consensus that ritual should be viewed as just one of a number of competing discourses operating through a network of social practices (Asad 1983). Archaeologists have also recognized the need to treat ritual not as a unique kind of action, but rather as an integral ‘aspect of action’ (Lane 1986). In many respects, the bringing together of both spheres of life has been most successful within regional landscape projects, which, as pointed out by Andrew Sherratt (1996), are ‘simultaneously exercises in settlement archaeology and investigations of ritual landscapes’. This is not only because a comprehensive understanding of regional history requires an investigation of both types of sites, but also because the distinction between ritual and secular monuments may depend on differential rates of archaeological preservation or whether the decision to ‘monumentalize’ was directed towards the ritual or secular domain.

In this issue attempts to redress the ritual versus profane polarization in the landscape include Shaw's study of the relationship between Buddhist monasteries, settlements and land use in central India and Perring's study of the impact of the Reformation on the organization of the secular cathedral of York Minster, which looks beyond the reconceptualization of sacred space to the wider social landscape.

Pilgrimage

Developments in the anthropology of pilgrimage, generally framed as a reaction to traditional Durkheimian and structuralist/archetypal models of ritual, have offered useful lessons to archaeologists of religion by highlighting the ‘negotiated’ and multi-tiered nature of sacred spaces and landscapes. For example, both Durkheim's (1965 [1912]) theory of pilgrimage as a social mechanism for integration and cohesion and Turner's (1973) view of pilgrimage as ‘anti-structure’, that is as a subversion rather than reflection of social order, have been criticized for overlooking the interplay between varying or contradictory discourses within a single pilgrimage ‘field’ (Coleman and Elsner 1994; Eade and Sallnow 1991). It has been argued that the pilgrimage field should be regarded as a ‘religious void’, which provides an arena for divergent interpretations of the sacred beneath the umbrella of the official discourse (Eade and Sallnow 1991: 15). Many of these concerns have crossed over into the archaeology of pilgrimage (Stopford 1994, and other papers in same issue), which until recently has dominated the archaeological study of religious change particularly within historical contexts. Some of these themes are explored in the papers of Andren, and Äikäs and Salmi, while Eekhout argues that Pachacamac's enduring importance as a pan-Andean pilgrimage centre was due to its perceived symbolic, ritual significance rather than political or economic impact, but that it nevertheless fits with the Durkheimian model in keeping with its reflection of the aims of the state.