Imagery of ritual actions in religious dreaming: steps toward a theory and method of the cognition of dreamt ritual interaction

ABSTRACT

The research literature on religious dreaming demonstrates the commonness of supernatural agent (SA) imagery, as well as various ritualistic and divinatory uses of dream content in the waking state. Less well mapped is the related domain of dream imagery about rituals (DIR). This article aims to provide a survey and analysis of DIR items and interactions in line with dream and ritual theories, with case illustrations from dream reports based on ethnographic fieldwork among Hindus in Nepal. The analysis relies on an understanding of “ritual” based on their assumed “obvious aspects” and traits, and on theoretical assumptions drawn from cognitive and evolutionary accounts of ritualization in relation to prominent strands in dream research such as threat simulation theory, and social simulation theory. The aim is to develop a novel strategy for exploring imagined ritual traits in dreaming. This undertaking contributes to the methodology and theoretical understanding of religious dream cognition.

KEYWORDS:

• Dream simulation
• religious dreaming
• ritual imagery
• cognition
• cultural transmission

Introduction

As a universal experience and theme in human cultures, dreaming is often understood as a special event and ascribed special standing in the world’s religious traditions and practices (Bulkeley 2007; 2008; Jedrej and Shaw 1992; Lohmann 2003). A change in how dreams are experienced and used has been described in terms of monophasic versus polyphasic cultural traits (Laughlin 2011). Monophasic cultures, seen in much of the modern world and presumably ‘WEIRD’ societies (Henrich, Heine, and Norenzayan 2010), tend to lack concepts and schemas that encourage belief in any distinct value of dreams or need to give them attention, along with making a strong distinction between imagination and reality (Laughlin, McManus, and d’Aquili 1990; Laughlin 2011). The opposite condition is seen in polyphasic cultures, which have historically prevailed in most human societies. These have a seemingly hypercognized or rich availability of schemas for dealing with and interpreting experiences (Levy 1984). These environments are characterized by epistemic recognition of dreams as communicative signals, ontological assumptions that dream contents are evidence about reality rather than products of imagination, and ascription of high value to dream experiences (e.g., Bourguignon and Evascu 1977), perhaps comparable to that of drug/entheogene-, trance-, and ordeal-driven visions (e.g., Locke and Kelly 1985).

In early anthropological theory, dreaming was seen as a kind of primary and experiential source of religious beliefs (e.g., Tylor 1871). A cross-cultural survey of dreams from the ethnographic literature demonstrates the importance of cultural traits that relate dreams to religious systems (D’Andrade 1961) and widespread use of dreams to contact or gain control over supernatural powers (agents). D’Andrade’s analysis further shows that anxiety about being alone, demands for self-reliance, and isolation correlated with a preoccupation with dreams containing supernatural agents or forces (1961, 320, 328). Relatedly, in various cultural and religious traditions, nightmares are reported in connection with initiations and conversion ordeals (Bulkeley 2007). Intense emotional dream experiences trigger a sense of ‘realness’ of imagery that is seen in involuntary and intrusive encounters with supernatural entities, as evidenced by nightmares, so-called ‘big dreams’ (encounters with SAs of importance in the tradition) or ‘visitation dreams’ (encounters with ancestors, the dead, relatives and loved ones). These types of dreams provide experiential salience relating to danger, threats and challenges, as well as to the dreamer’s concerns about the future in relation to prognosticatory content of dreams.

Accordingly, dreams are precursors to divination and oracles (e.g., Bulkeley 2008; cf. Patton 2004) and are used as ‘validation’ to support religious beliefs (Bulkeley 2008). Ritualized dream practices – in the waking state – relating to dream incubation, divination, dream interpretation and prognostication of future events have existed in polyphasic contexts within many religious traditions, ancient civilizations, and traditional societies (e.g., Bourguignon 1972; Bulkeley 2008; Hughes 2000; Jedrej and Shaw 1992; Laughlin 2011; Lincoln 1935; Tedlock 1987). Apprehensive emotions, belief in supernatural agents in the dream state, and various ritualistic and cultural uses of such dreams in the waking state are commonly reported in these studies. In this way, the literature implies some form of interaction with various supernatural agents.

Less is known about the ways that ritual interactions are presented in dreams.

Generally, dream processes ‘simulate’ various types of behaviour. A subcase of this would be dreamers’ experiences of performing a ritual behaviour that is not actualized in the waking state. This makes dreams an interesting case of isolating cognitive representations about rituals and reutilization behaviours. Contrary to metacognitive representations – where people ponder or communicate about (the meaning of) rituals – this is direct cognitive simulation of ritual performance (of self or others) during the dream state. The present analysis starts from the presuppositions that dream processes draw upon evolved simulation capacities, that ritualistic dream imagery is a frequent and somewhat unmapped terrain in so-called religious dreaming, and that these conditions may even generally constitute transmission factors in the cultural communication of dream representations in the waking state.

Prominent theories in recent dream research, as well cognitive and evolutionary theories on religion and cultural transmission, however, would present distinct explanatory accounts of what to expect regarding the function and features of ritual interaction occurring in dream imagery. However, the term ‘ritual’ does not refer to a single type of phenomenon, and seems to constitute a polythetic category (e.g., Needham 1975) based on family resemblance. Producing one single (explanatory) theory of ‘ritual’ is thus not feasible. Nevertheless, there are obvious behavioural traits in cases of ‘rituals’ (Rappaport 1979), and these recurrences can be modelled and methodologically applied according to their evolved cognitive functions (Boyer and Liénard 2020).

Anthropologically speaking, ‘religion’ in the present context is not a unitary category (e.g., Bloch 2008; Boyer 2013; Sperber 2018), but is better understood as a collection of more basic and recurrent traits and cultural expressions not all of which co-occur in all social environments. These elements – and their recurrence and relationships – can be studied separately, making them more amenable to scientific investigation (e.g., Atran 2002; Boyer 2005; Sørensen 2005). This can be described as a kind of ‘piecemeal’ (Barrett 2007) or ‘building-blocks’ (Taves 2011) approach to religious phenomena. The term ‘religious’ dreaming is therefore used here as shorthand for a conglomerate of disparate parts, phenomena, processes, and causes. In relation to dreaming and dreams, these parts are first and foremost certain beliefs, values, and practices, and in particular, as is suggested and addressed in this article, ritual practices.

Consequently, one aim of this article is to develop a theory and methodology that can guide the exploratory analysis of dream imagery about ritual (DIR). A second aim is to apply in the analysis of DIR the recognition that many/most ‘rituals’ refer to various constellations of (a) scripted, normative interaction (SNI); (b) signalling of affiliation and social cohesion (SASC); (c) magical causation and reckoning (MCR); and (d) ritualized behaviour (RB) (Boyer and Liénard 2020). A third aim is to map and align DIR with evolutionarily grounded theories of dream simulation. These aims are finally applied to, and exemplified by, dream narratives from ethnographic fieldwork in Nepal. One caveat is needed. It is not the aim of the present article to offer full-blown statistics, such as intercoder reliability and multi regression factors related to the data and method. This highly relevant undertaking will be the topic of another publication.

Some crucial neuro-cognitive processes in simulation theories of dreaming

Dreams function as novel but realistic simulations of waking (social) life, drawing on an assemblage of negative and positive affects, entities, motivations, and scenarios (e.g., Blagrove, Hale, et al. 2019) to test predicted and hypothetical outcomes. It has been demonstrated that dream processes contain various inexplicable images and modalities of cognition that enable prospective modelling and counterfactual past and future states of affairs (McNamara et al. 2002). More specifically, such processes are related to the state of REM dreaming of prospective coding (Llewellyn 2016) and encoding of episodic memories (Llewellyn 2013). Dreaming is involved in the origination and elaboration of goals and values/desires, and in the content of daydreaming, which may be crucial to episodic prospection (McNamara and Bulkeley 2015; Szpunar 2010) of probable states of affairs and outcomes, and research by Wamsley explicitly connects dream simulation to episodic prospection (2022).

That dreaming and waking consciousness operate as a kind of ‘world simulation’ and ‘realistic simulation’ (Foulkes 1985; cf. Llinás and Paré 1991) is a rather widely accepted view in dream research, despite technical differences (Domhoff 2007; Foulkes 1985; Hobson 2009; Hobson and Friston 2012). As stated by Tart the concept of ‘world simulation’ proposes that the same brain processes govern dreaming and the waking state (1987), a stance that has been elaborated in a full-scale theory of dreaming and consciousness in Revonsuo (2006). Addressing the alleged phenomenological core of world-simulation in dreams, the principle of ‘immersive spatiotemporal hallucination’ has been suggested (Windt 2010) and been related to self-model of dreaming (Windt 2015). Besides realistic simulation, dream simulation and the concept of ‘virtual reality’ has been connected with empirical dream research (Revonsuo 1995).

Three major approaches to dream simulation (following the overview in Revonsuo, Tuominen, and Valli 2016) will be addressed below: Threat Simulation Theory (TST) and Social Simulation Theory (SST), which are functional and adaptation models, and the non-functional Continuity Hypothesis (CH). However, there undoubtedly are numerous other theories dealing with the content and functions of dreaming. A basic assumption of the continuity hypothesis is that dream content reflects or is causally influenced by (social and cultural) processes in the waking consciousness (e.g., Schredl and Hofmann 2003). For example, Muslims are expected to dream about the Hajj pilgrimage more than Roman Catholics or atheists, and people in hunter-gatherer societies, where people interact intensively with nature and wildlife, will have a higher proportion of animal imagery in their dreams than people in modern societies. While the CH suggests that dreams simulate whatever is imported from the waking consciousness, it lacks a solid theory of why some traits are selected and imported for dream simulation rather than others (Revonsuo, Tuominen, and Valli 2016).

Other simulation accounts, however, propose more thorough and parsimonious models that presume the operation of active selection biases in the specific information and emotions occurring in dreams. The pervasiveness of prospective vigilance in dreaming is modelled by threat simulation theory (TST) (Revonsuo 2000; Valli and Revonsuo 2009). According to TST, dream cognitions seen during REM and even perhaps in NREM sleep are capacity evolved, in relation to direct dangers to adaptive fitness, for the simulation of threatening events, rehearsal of threat perception and threat avoidance skills to augment vigilance perceptions of threats in waking life (Revonsuo 2000; Valli and Revonsuo 2009). TST claims in this way that the evolved function is to run through the various motor brain mechanisms, as well as the emotional, cognitive, and perceptual mechanisms involved in detection and avoidance of threat, to enhance prompt and accurate responses in relevant situations in the waking state.

Nightmares concerned with themes about survival threats, aggression, and accidents are omnipresent and provide support for the TST account. Indeed, themes of ‘being attacked’ or threatened by enemies, wild animals, strangers, or monstrous creatures are very common dream themes, and studies points to the occurrence of various threats in 60-75% of dream reports (Domhoff 1996; Hall and Van de Castle 1966, for detailed statistics and empirical evidence, see Valli and Revonsuo 2009). That is, in questionnaire studies, if respondents are asked to select from lists of prototypical dream themes and what themes they remember having experienced, the topic of ‘being chased or attacked’ (most typically by wild animals or male strangers) is the most cross-culturally widespread and typically remembered dream topic.

TST resembles other proposed cognitive proclivities such as a ‘hypersensitive agency detection device’ (HADD) (Barrett 2004), and threat-detection psychology (e.g., Boyer and Bergstrom 2011). These all include the appraisal and detection of prospective fitness challenges relating to predation, social intrusion, contagion, social offence, and threat to offspring. Consequently, such adaptations would not only have evolved from direct fitness threats in natural selection, but also from pressures from social and sexual selection. Hence it should be expected that adaptive functions of dreaming have developed in relation to the simulation and mapping of social interactions and complexities (e.g., Brereton 2000; Humphrey 2000; McNamara 2004; Nielsen and Germain 2000).

In social simulation theory (SST) (Revonsuo, Tuominen, and Valli 2016; Tuominen, Stenberg, et al. 2019), dream contents are modelled as simulations of frequent cardinal social challenges and needs in waking life situations related to social actors and characteristics, bonds, interactions and connections. SST suggests that specific evolved functions operate in the rehearsal and simulation of social perceptions and scenarios with emotionally neutral and positive dreams complementing those that are described for TST (above) and operate in nightmares. Both social and threat simulations may co-occur in real dreams, but are dedicated to their specific domains of simulation and thus may operate in complementary ways (Revonsuo, Tuominen, and Valli 2016, 14).

Evidence of social simulation is demonstrated in the literature, and studies describe at least one social trait being present in over 83.5% of dream reports (Tuominen, Revonsuo, and Valli 2019), or alternatively that only 6.5% of dream reports are about non-social simulation (Domhoff and Schneider 2018). REM sleep promotes social bonding dreams (McNamara 1996), and comparatively more social actors appear in dreams than reports from waking states (McNamara et al. 2005). Social simulation in dreams has also been abundantly demonstrated in studies about social perception, mindreading and theory of mind attribution (ToM) in dreaming (e.g., Kahn and Hobson 2005; McNamara et al. 2007).

Dream simulation, the issue of diminished dream agency, and supernatural agents

A remarkable dimension of dream imagery is the tendency to reconfigure the locus of the dreamer and dream agency. Nightmares are known to exhibit a much greater decrease in the sense of agency in the dreamer than non-nightmarish dreams (McNamara et al. 2018). While REM-sleep dreams are full of action-filled imagery of a self as acting, REM nightmares significantly appear in association with diminished agency in the dreamer – and this tendency leads to an increased attribution of a causal agentive role to other special dream characters, in particular ‘superhuman/supernatural’ entities (McNamara 2016; McNamara et al. 2018; McNamara and Bulkeley 2015). Such social dream simulation of external agents in relation to the dreamer locus has been explained as a general neural pruning of predictive model failure (e.g., Bucci and Grasso 2017; Hobson and Friston 2012; cf. Tuominen and Valli 2019). Specifically, during REM dreaming, a downregulation of a self-model occurs in relation to activation and intentions which causes predictive model gaps. This in turn instigates various simulations of counterfactual and possible scenarios (Wamsley 2022), as well as attributions of agency to ego-external agents to correct for the predictive failure (McNamara and Bulkeley 2015).

As research by McNamara and colleagues suggests, dream imagery about supernatural agents (SA) reliably appears in unpleasant dreams and nightmares in association with diminished agency in the dreamer (McNamara et al. 2018; McNamara and Bulkeley 2015). Accordingly, when dreamers dreamt about SA in unpleasant dreams and nightmares, 90% of the dreams were associated with diminished agency in the dreamer relating to victimization, physical threats of aggression, and harmful intentions (from SA) towards the dreamer. The sense of diminished agency (e.g., threats, being preyed upon, being eaten) of the dreamer in SA (and counterintuitive) dreams has also been shown to the decisive predictor for engaging in divinatory communication in recent studies (Nordin 2023). Additionally, REM sleep related to sleep paralysis and generalized atonia may arguably affect dream content and sense of agency (McNamara et al. 2018, 432), which makes sense in relation to themes of nightmares or unpleasant dreams about drowning, being mute, being stuck, being unable to move or run and the like. More generally, the topic of dream/nightmare-related imagery about agency restriction has been related to a general proposal about ‘decentring’ schemata and tendencies in a variety of allegedly religious cognitions relating to mystical experiences, spirit possession, and ritualized and meditative practices (McNamara 2009; McNamara et al. 2015). Such a decentring schema involves in a significant way experiences of diminished or lost sense of self, agency and control, as well as fear, anxiety, disorientation and so on (for technical details, see McNamara et al. 2015).

Diminished dream agency, ritualistic behaviour, and compensatory control (theory)

Research also suggests a connection between experience of a lack of (internal and external) control and supernatural agency attribution, which provides compensatory control (Hoogeveen et al. 2018; Kay et al. 2010). According to this theory (CCT), religious beliefs can be entertained in conditions of anxiety during experiences of confusion, randomness, and uncertainty (cf. Boyer 2001, 203–228) to boost a sense of compensatory control over the environment. Indeed, it is the compensatory control provided by religious beliefs, rather than its being comforting in general, that is decisive (Hoogeveen et al. 2018). Reestablishing a sense of control in circumstances of experienced uncertainty motivates behaviour. Legare and Souza (2014) demonstrate that perceived uncertainty and feelings of lack of control increase perceptions of ritual efficacy (see below on ‘ritualized behaviour’ and cf. Lang, Krátký, and Xygalatas 2020). Rituals seemingly do not have actual effects in relation to their instrumental results and physical causation (Humphrey and Laidlaw 1994; Legare and Herrmann 2013). A reasonable suggestion is that rituals operate as means to assert an illusion of increased control by entertaining stipulated and sanctioned actions that mimic causal efficacy (Legare and Souza 2014). Although CCT research is based on religious cognition in the waking state, it provides hints and informed suggestions about how dream imagery about rituals operates and its neurocognitive underpinnings. Perhaps dreaming can provide functions for coping by simulating threats and social challenges, repeating traumatic experience, and simulating repetitious ritualized behaviour (cf. Türcke 2013).

Despite the pervasive manifestation of action-filled experience and a dream self as acting in the dreamer’s REM-sleep dreams (e.g., in Valli and Revonsuo 2009, 54% of dreamers with a nightmare report instigating active behavioural response), it can be hypothesized that in dreaming, particularly during nightmarish REM sleep, there is significant production of imagery of diminished sense of agency (as discussed above). Diminished agency and control, and sense of emotional threats, fear and terror could enhance simulated rehearsal of social action prototypes, such as rituals that mimic causal efficacy and control.

This may be especially likely in circumstances involving ritualized behaviours that mimic or draw upon systems for precaution and threat-detection and heightening of safety appraisals. Noteworthily, the specific tendency toward diminished agency and fear (of threats) in dreams highlighted here also resembles models of ritual behaviour that specifically point to modification of agency in many ‘rituals’ (for further discussion, see McNamara et al. 2018, 437; Nordin 2011a, 233–235). Indeed, all variants of ‘ritual’ interaction can and often seem to be related to such precaution or threat systems, as will be further discussed below.

Consequently, types of dreamed ritual sequences (according to the model presented below) also in several cases manifest the function of social simulation, which relies less or only indirectly on threat simulation (e.g., contagion). These dreamt ritual sequences manifest more or less agent/agency properties on a spectrum, and the actual frequency and distribution of these threats need much more thorough empirical exploration. If dreamed imagery of ritual is coded according to magical causation and ritual form theory (Lawson and McCauley 1990; McCauley and Lawson 2002), then the cognition of the situation is partly based on threat and contagion, but also partly on the modalities of being acted upon, being a patient or acted upon by supernatural agents. This can be related to various forms of evolved vigilance to threats relating to appeasement etc. (Cantor 2009) discussed below in the context of ritualistic behaviour.

The elusiveness and variants of ‘ritual'

Rituals have been described in all cultures and historical eras, and yet the exact nature, scope, and cohesiveness of their nature (if any) are disputed by scholars in anthropology, psychology, and the study of religion, even if such disagreement is fairly absent among ‘ritual practitioners’. There is an immense body of literature on ‘ritual’ (for some overviews, see Bell 1992; 1997; Grimes 1996; 2006; 2013; Kreinath, Hartung, and Deschner 2004; Lawson and McCauley 1990). Intellectualist and symbolist theories were two influential approaches to ritual (for an overview, see Lawson and McCauley 1990). The former held that (religious) beliefs determined ritual behaviours that in themselves were instrumental and rational, but were based on false or flawed (religious) assumptions about the world (e.g., Frazer [1922] 1963; Tylor 1871). Symbolist theories understood ritual not as a defect of instrumental behaviour, but as a symbolic expression of hidden nonmanifest meaning with specific social or psychological functions (e.g., Douglas [1966] 1984; Durkheim [1912] 1995; Geertz [1973] 1993; Turner 1967). On a related note, the notion that dreams express symbolic meanings, such as hidden wishes, is a popular one (e.g., Freud 1913), and raises the question of how the alleged symbolism of rituals interacts with dream symbolism? However, the model according to which the patient in analysis somehow unlocks and decodes the symbolic functions and expressions of dreams is rather outdated in neuro-cognitive research today, and other models of dream functions (e.g., simulation) are more prevalent (Bulkeley 1997; Foulkes 1985; Hobson 1994; Revonsuo 2000; Solms 1997). This is not to say that the assumption of dream symbolism is not widespread as a cultural model (cf. Tedlock 1987).

Intellectualist and symbolistic accounts run into difficulty when attempting to explain the odd features of ritualistic behaviours by either dissolving them into flawed instrumental behaviours or conceiving them as another form of symbolic communication. This raises the questions of why seemingly rational people would resort to (flawed) ritual action, and why people would engage in ritual symbolic communication rather than more efficient modes of engagement. Relatedly, the assumption of some inherent ritual meaning would seem like an obvious or trivial starting point in various classical symbolic approaches (e.g., Douglas [1966] 1984; Durkheim [1912] 1995; Geertz [1973] 1993; Turner 1967; for an overview, see Lawson and McCauley 1990). These approaches tend to overemphasize symbolic aspects of ritual communication – viewing it almost as a kind of text (e.g., Geertz [1973] 1993). If this were the case, however, ritual symbolic transmission would be less communicatively successful than other forms of social interaction (Sørensen and Nielbo 2019), and thus would be maladaptive from the perspective of natural and cultural selection. On the contrary, ritual communication actually hampers semantic redundancy and thus limits conceivable symbolic meaning (Bloch 1974), while attempts to decipher the ‘ritual meaning’ risk degenerating into a cryptology that conflates ritual behaviours with various deep messages to be decoded (Sperber 1975).

Consequently, rituals were seen as pure, goal-less and inherently meaningless behaviours (Staal 1979). In a related way, anthropologist Roy Rappaport stressed the importance of explaining the unique traits that differentiate ritual actions from other actions (1979). To the question of how humans would cognize such distinct ritual actions sequences, a formal model was suggested centred around the ‘acceptance’ and ‘efficacy’ of beliefs held by ritual performers and observers (Lawson and McCauley 1990). One crucial cognitive ingredient in this ‘ritual action representation system’ is that religious rituals are characterized by direct or indirect inclusion of some culturally specific supernatural/human agent(s). Although the theory represents an ambitious account of ritual and how beliefs in ritual efficacy may be construed, the enigma of ritualization remains, and its predictions are not conclusively supported by empirical research (e.g., Malley and Barrett 2003). Apart from this caveat, the connection between magic and causal agency will here be treated as one ingredient of ‘ritual’.

Variations and traits in the imagery of dreamed ‘ritual' interaction

The label ‘ritual’ in lay and scholarly parlance is often applied to different examples based on similarities of family resemblance (Boyer and Liénard 2020). This description derives from the fact that anthropologists have long observed that ‘ritual’ is another case of polythetic categories (Needham 1975; Sperber 2018) fractured into family resemblances of slightly different traits and phenomena (cf. Barrett 2017; Taves 2011). This observation is thus well established in the recent research literature on cognitive and evolutionary theories of religion and culture. ‘Ritual’ denotes such widely differing assemblages of behaviours and traits that it is not feasible to produce any ‘grand theory’ with credible explanatory value. What is possible, and indeed provides a fruitful understanding, is to conduct an analysis of the various building blocks of social interactions labelled as ‘ritual’. According to this approach, the term ‘ritual’ points to different clusters of phenomena and representations that (a) have no necessary co-occurrence in cultural distribution; and (b) co-occur to different degrees. As modelled by Boyer and Liénard (2020), this assembly of ritual variants concerns scripted, normative interaction (SNI), signalling of affiliation and social cohesion (SASC), magical causation and reckoning (MCR), and ritualized behaviour (RB).

By applying and modifying a typology of these predominant ‘ritual’ features – their cognitive and social mechanisms and effects – the present analysis aims to discern the frequency and distribution of dream imagery about ritual (DIR). Any sense of ‘ritual’ – as the previous analysis suggests – can concern (a) ‘ritualistic’ use of dreaming and its content (e.g., as in dream incubation or divination), (b) dream imagery about ritual interaction and ritual items, and (c) the interface between (a) and (b) – ritual dream imagery and ritual engagement with ritual dream imagery. Additionally, there is the question relating to (b) and the encouragement of cultural transmission of ritual activities in the waking state. The present analysis will primarily be concerned with topic (b) – what kinds of dream imagery can be found about ritualistic items according to the given typological model of ritual ingredients. These tendencies in ritual behaviour and beliefs, and their cognitive and evolutionary backdrop, also suggest strong predispositions for, and probabilities of, these representations being reconstructed and actualized as attractors among the immense variety of alternatives in cultural transmission (e.g., Claidière and Sperber 2007; Scott-Phillips, Blancke, and Heintz 2018).

The stance adopted here is a sense of ‘creativity’ based on the general observation that dreaming and sleep can spur novelty and innovation (e.g., Walker 2017), and that the cognitive dynamics of cultural transmission are governed in crucial ways by inferential processes (e.g., Boyer 2009; Sperber and Hirschfeld 2004), not simply by copying or ‘down loading’, as in a ‘fax-model’ (e.g., Quinn 1992) of acquisition and transmission. REM-sleep dreaming may foster the solving of problems in the waking state and fuel creativity, as shown by research findings on the origin of artistic expressions in music and literature (Walker 2017). Dreams and dream-sharing communication foster a sense of personal insight (Blagrove, Hale, et al. 2019; Balgrove, Edwards, et al. 2019; Edwards et al. 2015), where insight is the sudden and unexpected understanding of an incomprehensible concept or problem by recognizing and integrating new connections between domains of knowledge (Zander, Öllinger, and Volz 2016). Evidently, dream imagery of ritual variants is relevant for cultural transmission and communication in the waking state that may provide novelty in religious and supernaturalistic imagination as well as nonreligious contexts, although this complex topic is beyond the scope of this article.

Detailed model of conceivable traits in dreaming of ‘rituals'

Scripted, normative interaction (SNI)

The research literature (e.g., Kuper 1947) points to various social procedural interactions guided by specific scripts (Schank and Abelson 1977) for behaviour. These interactions tend to be associated with intuitions that if the scripted rules are not followed properly or are violated, then the action will not be an instance of that categorized action type. Scripted and categorized interactions tend to be labelled (or named) and normative, such that participants know that there are specific procedures defining the interactions, and that others are also likely to understand the actions in a similar way. Ethnographic descriptions describe degrees of compulsoriness or voluntariness in following ritual procedures (e.g., Gellner 1992). As scripted and named interactions, these behaviours tend to express normative stances, in so far as people expect each other to follow the given rules (e.g., Bicchieri 2005). This would suggest that pressure from an evolved norm-psychology geared toward social coordination of interaction (Bicchieri 2005) and developed during infancy (Rakoczy and Schmidt 2013) underpins cultural transmission (Göckeritz, Schmidt, and Tomasello 2014) and enforces conformity with the scripted interactions seen in ‘rituals’.

Signalling of affiliation and social coalition (SASC)

Scripted and named interactions can be extremely social and communicative, with obvious potential to serve social functions, and can have pragmatic effects such as signalling affiliation, cohesion, identity, and commitment (e.g., Watson-Jones and Legare 2016; cf. Atran 2002; 2011). Such ritual signalling can serve purposes of deference (e.g., Bloch 2004), stipulation (Humphrey and Laidlaw 1994), or commentary on order, rank, and relations in social interaction. The signalling contributes to the construal of a mutual socio-cognitive framework or common knowledge (Chwe 2001) based on similar representations of the event among the ritual participants, the creation of strong shared feelings, or experiences of group cohesiveness and altruistic loyalties toward ingroup members of a community (e.g., Reddish et al. 2016; Whitehouse and Lanman 2014). The fact that scripted and named interactions have implications for signalling social unity and connection makes good sense in the light of human adaptations to group living (e.g., Legare and Wen 2014), in the attitudes connected to the dynamics of ‘us/them’ or ‘in-group’ versus ‘outgroup’ (e.g., Hornsey 2008), as well as in forming and sustaining cooperative alliances (Yamagishi and Mifune 2009). An evolved psychology for coalitional cognition and interaction is likely to be devoted to several of the highly complex tasks related to managing in- and between-group dynamics (e.g., Boyer, Firat, and van Leeuwen 2015; Pietraszewski 2016). Communicating and signalling commitment with scripted public behaviours is thus a crucial coalitional mechanism for ingroup affiliation. Costly displays of ritual signalling (e.g., Sosis 2006), such as undergoing ordeals like long pilgrimages (Nordin 2011b), convey in this perspective information about commitment to other members in a group, or gestures towards rival coalitions. It is also likely that these forms of coalitional psychology contribute additional factors of stability and attraction in cultural transmission of these forms of ‘ritual’ interactions (Boyer and Liénard 2020).

Magical causation and reckoning (MCR)

Many forms of scripted and named interactions that are categorized as ‘rituals’ employ behaviours, objects and reasoning based on ‘magical’ influences, effects or connections, such as touching, seeing, consuming, or coming into contact with sacred water or other substances believed to provide blessings, cures or protection, or with objects used to prepare for future rituals (e.g., Nordin 2009; Siegal et al. 2001). Classical accounts in the history of anthropology modelled various relations of magical causality in terms of similarity, proximity, or contiguity (e.g., Frazer [1922] 1963; Mauss 1972; Tylor 1871; cf. Nemoroff and Rozin 2000; Siegal et al. 2001; Sørensen 2007). These interactions and connections entail notions about causal and/or physical efficacy, although the detailed workings of the process are often opaque to the practitioners. Some people may have elaborate theories regarding such causal efficacy; however these are rationalizations and are not crucial for belief in magical connections (Boyer and Liénard 2020, 3). The strength and cultural pervasiveness of magical beliefs rather seem to draw upon the sense of naturalness that is evoked by compelling intuitive expectations about disease transmission, contagion, sickness, and pathogens (cf. Nemoroff and Rozin 2000; Nordin 2009; Siegal et al. 2001). These expectations involve assumptions about invisible vectors that can be transmitted in various ways between humans and are equally dangerous or infectious regardless of the dose or degree of exposure.

One form of magical (superhuman) causation is involved in the way human understanding of actions informs the conceptualization of ritual behaviour, beliefs in their efficacy, and how they are thought to work (Lawson and McCauley 1990; McCauley and Lawson 2002; cf. Malley 2019; Malley and Barrett 2003). According to McCauley and Lawson’s theory, believers assume that (supernatural) agents are responsible for the efficacy of religious rituals, and that the supernatural agent operates either as the actor or patient in ritual action (2002). This theory entails that people assume that the efficacy of magical causation is due to superhuman agency SA in rituals, which is activated in different forms. For example, the SAs are acting via a priestly official in special agent rituals (many forms of initiation, marriage, cremations, thread ceremonies) that cannot be repeated. In repeatable special patient rituals, the action is done to the SA, (thanking, cleaning, honouring, preparing, etc.). Finally, in repeatable special instrument rituals, the believer may receive blessings through various magical connections to an SA or SA object (for details, see Malley 2019; McCauley and Lawson 2002). Ritual form theory suggests specific predictions about why SA dreams receive special ritual status. This is because it models a notion of supernatural immediacy that implies structurally basic and central rituals in any cultural and religious system (Lawson and McCauley 1990, 126). In this way, basic rituals can enable subsequent ones (McCauley and Lawson 2002, 33).

Ritualized behaviour (RB)

One phenomenon that seems typical of, though not universal or exclusive to ‘rituals’, comprises the features of ritualized behaviour referred to as ‘obvious aspects of ritual’ (Rappaport 1979). These characteristics have been discussed in detail in recent research (Boyer and Liénard 2006; Dulaney and Fiske 1994; Fiske and Haslam 1997; Liénard and Boyer 2006; and see Sørensen and Nielbo 2019; for overviews, see Whitehouse 2022; Xygalatas 2022). These features includes traits such as compulsion (an urge to perform the action, often without a clear idea of why); scriptedness (a strong motivation to perform the action in a stereotypical, way and order, and without necessarily having discursive reasons for the scripted performance); goal-demotion (where the series of actions can be connected to a stated overarching goal, but the segments of the action cannot individually be represented as leading to the goal); and repetition (a strong motivation to repeat actions or sequences, often without a clear representation of why it is needed). These traits have been explained as by-products of cognition dedicated to ‘hazard precaution’ (Boyer and Liénard 2006; Liénard and Boyer 2006), an adaptation that evolved for detecting and responding to a range of diffuse and hard-to-monitor potential threats, such as contamination and contagion, harm to offspring, intrusion by strangers and social misdemeanour (Boyer and Liénard 2020). The abundance of these vague threats makes them difficult to fully discern and take measures against. This condition activates a threat-detection system with its associated ritualized behaviours. When such threat-detection psychology is overactivated and interrupted, it can result in specific conditions such as phobias, PTSD, and especially obsessive-compulsive disorder (OCD) with symptoms related to ritualized, compulsive, repetitive, scripted behaviour (Dulaney and Fiske 1994; Fiske and Haslam 1997), as well as biased and intrusive cognitions and rumination about potential dangers (e.g., about impending doom or supernatural justice). Under these conditions, a failure to increase one’s sense of security with precautionary behaviours leads to overactivation and pathological repetition of OCD rituals (Szechtman and Woody 2004). OCD rituals have in similar way been summarized as an overactivation of harm prevention (Del Giudice 2018). Hence, ritualized behaviour arises from a ‘hazard precaution system’ and ‘action parsing system’ that is geared toward dividing the continuum of behaviour into meaningful parts. When the hazard precaution system is activated, it triggers applicable action segments with matching evolved precautionary behaviours. Simultaneously, the system sequences and splits the behaviour into minor parts with attention focused on the properties of actions at a lower level (hence obsession with following distinct action sequences and avoiding ‘wrong’ motions). When no salient ‘stop’ signal is manifest, uncertainty triggers repetition of the action sequence and this tendency may eventually swamp the working memory and bring about a provisional relief of anxiety. Obsessive-compulsive themes are common in cross cultural perspective and the evolved cognitive proclivities that underpin such precautionary behaviours are very likely strong attraction factors in the distribution of cultural and religious ritual behaviour.

The presence of behavioural traits such as compulsion, invariance, and repetition is why ritual has been metaphorically described as a tunnel into which one plunges and only can pass through by not turning to either side (Bloch 1974).

Related forms of imagery about inhibited behaviour that occur in dreams and REM sleep tend to manifest a model of demoted self and/or agency loss in the dreamer (McNamara 2016; McNamara and Bulkeley 2015). Also, nightmares, purportedly ‘mystical’ dreams and sleep paralysis are reported from many cultures (De Sá and Mota-Rolim 2016; Hinton, Hufford, and Kirmayer 2005; Sharpless and Doghramji 2015). There is a strong connection between the diminished agency of the dreamer’s model of self and or agency during REM sleep states and an elevated ascription of the causal-agentive function to other dream characters – especially supernatural agents (McNamara et al. 2018; McNamara and Bulkeley 2015). In accordance with predominant predictive coding approaches, cognitive brain processing aims to prune redundancy and prediction errors by positing external causal agents (McNamara and Bulkeley 2015). In sleep paralysis, which mainly occurs while falling asleep or waking up (e.g., Mahowald, Cramer Bornemann, and Schenck 2011) immobility is accompanied by intense anxiety and often fantastic hypnagogic hallucinations or delusions (e.g., Dahlitz and Parkes 1993), often including a sense of a threatful presence and intrusion (Cheyne, Rueffer, and Newby-Clark 1999), which in many cultures and folk traditions are construed as various supernatural assaults (De Sá and Mota-Rolim 2016; Hinton, Hufford, and Kirmayer 2005). Arguably, these tendencies in sleep-paralysis cognition resemble some processes in nightmares involving supernatural imagery although the latter seem to involve both modalities of flight/ fight or diminished agency on a continuum. In any event, paralysis in this context may be related to the activation of evolved systems of threat psychology, and in the present case specifically in the involuntary, automatic response of adaptive tonic immobility seen during situations of severe threats where escape or fighting is not possible, or is too costly, such as among rape victims or prey animals (Buss 2021; Marx et al. 2008). Additionally, according to Cantor, not just fight or flight is activated in vigilance and risk assessment but routinely a more complex complement of different responses such as avoidance, attentive immobility, withdrawal, aggressive defence, appeasement, and tonic immobility (2009). Interestingly, Cantor suggests that appeasement is a primary defence in conspecific encounters with socially more dominant individuals that involves submission, pacification, and conciliation. However, depending on the type of supernatural agents (high status deities, demigods, spirits, soul, ghosts, demons) that are involved in the dream imagery of ritual behaviour, the behaviour should be related to either avoidance, attentive immobility, withdrawal, aggressive defence, appeasement, or tonic immobility in threat and/or social simulation. For example, appeasement strategies of ritualized imagery may be significant in social simulation of ‘visitation dreams’ and dreams about high-status deities but not in dreams about ghosts, demons, and supernatural animals. Avoidance, attentive immobility, withdrawal, aggressive defence, and tonic immobility are likely to be predominant in prototypical nightmarish dreams with monstrous and threatening entities.

Dream simulation and imagery of ritual variants

If the dream process simulates ritual imagery, then assuming the stances adopted here it should be based on either SST or TST simulation, and manifested as some ritual variants such as SNI, SASC, MCR and RB. As the theoretically derived descriptions of ritual variants suggest, the morphologies of these items are themselves generated by evolved precautionary psychological systems dedicated to managing fitness challenges in social life and survival threats, just as SST and TST propose. This preliminary layout suggests that dreams that produce content with SNI and/or SASC imagery are appearances that derive from overarching social simulation dedicated to issues of social perception, interaction, and in-group inclusion (Tuominen, Stenberg, et al. 2019; Tuominen, Revonsuo, and Valli 2019). Dream content with MCR and RB imagery would be generated by systems dedicated to threat simulation, particularly in relation to predation, intrusion, contagion, and social offence, as well as threats to offspring (cf. threat-categorization in Boyer and Liénard 2020; Revonsuo, Tuominen, and Valli 2016). In actual dream processes, social and threat simulation are blended or occur in rapid sequence (cf. Revonsuo, Tuominen, and Valli 2016), and such a modal dynamic is also to be expected in the appearance of ritual variants. For example, a dream report may contain SNI through reference to a stereotyped named action (e.g., puja or darshan), or reference to purification, blessings, or sacred objects (manifesting MCR) and/or imagery of specific action sequences (manifesting RB). The occurrence of ritual variants such as SNI, SASC, MCR and RB in dream imagery can thus be nested, implied, or manifest, a fact that may suggest metaphorical manifestations of information (Antrobus 1977; Blagrove, Hale, et al. 2019; Lakoff 1993). More to the point, such dreams can have indexical properties (e.g., smoke as a sign of fire) in the semiotic sense. For example, cognition of indexical relation formation (e.g., Frøkjær Sørensen 2021) suggests that humans establish such associations on the basis of intuitive causal reasoning and associative and cultural learning. Dreams of a ritual artefact may similarly be indexical for named and scripted interaction and a specific ritualized behaviour, and such a behaviour can have indexical properties of magical causation, etc.

Imagery of dreamt ritual – coding and exemplification

Previous sections have surveyed the dream simulation approaches and the varieties of ‘ritual’ by emphasizing a composite model. It was further suggested that – although the research primarily highlights topics related to the appearance of various supernatural entities – religious dream imagery is likely to entail fragments of social interactions. This is due to the predictions of SST and the highly social content of dreaming; the highly social, communicative and interactive nature of religious phenomena; and the prevalence of religious traits and supernaturalism in dream imagery. However, what traits of ritual are conceivable, and what is their prevalence and distribution in dreams? If dreams contain imagery about ritual interactions, they will be variants of scripted, normative interaction; signalling of affiliation and social cohesion; magical causation and reckoning; or ritualized behaviour. Such imagery should also be seen as resulting from dedicated systems of evolved cognition and emotions underpinning rituals as well as social threat simulation in the dream process. Behaviour connected to ‘ritual’ may, but must not, activate the presence of all cognitive mechanisms relating to signalling (SNI), cooperative coordination (SASC), magical belief (MCR) or ritualized behaviour (RB), and these can co-occur to various overlapping degrees (Boyer and Liénard 2020). For example, in many cases dream reports concern designated social interaction types, some of which have normative effects, and most of which involve supernatural agents. These nested layers involve the mechanisms of signalling, coordination, magical beliefs, and ritual behaviour to differing degrees in the currently activated dream simulation.

The final section of this article takes the analysis a step further and offers a brief description of how dream imagery about ritual interaction is manifested, using examples from anthropological research among Hindus in Nepal. This research consisted of 61 interviews with 39 male (63.9%) and 22 female (36.1%) subjects. (For basic information about the setting, conditions, layout and limitations of this research, see Nordin 2020; Nordin and Bjälkebring 2019; 2021).

In this dream data, imagery about rituals (DIR) was either manifested or implied, and the occurrence and frequency of the modes are indicated below. Ritual dream imagery and information of any sort occur in 90% of the dream reports (55 informants out of 61) and refer to directly manifested or implied items in the domains of SNI, SASC MCR and RB. In 18 dream reports (29.5%) SA(s) commanded the dreamer to conduct rituals, while only one case consisted of a human commanding this conduct (in 42 cases, 68.9%, there was no information).

Coding SNI traits in dream imagery and results

The analyses tracked the occurrence of explicitly stated names or labelled action/interaction-sequences uttered in or implied by action descriptions in the dream reports (e.g., puja, Aarthi, tirtha yatra, bali, animal sacrifice, etc.). Markers of SNI traits were evident in 45 dream reports (73.2%) and describe either one single action type (e.g., chanting [5], darshan [4], puja [5], tirtha yatra [1], bali [3], barta [2], snan [3], cremation [1], prayer [3], or other gestures – subordination, greeting, bowing [7]) or a mixture of several named ritualized action prototypes in the imagery (e.g., procedures such as snan or darshan, or offerings and gifts, such as tikka, prasad, mala). The named category of darshan, an exchange of gazes with supernatural agents, was the most commonly manifested type (in total 14, 23%).

Coding SASC traits in the dreams and results

In seeking to identify SASC traits, the focus was on descriptions of dreamt social interaction that indicate relations to family, social group and order (e.g., through comments and allusions). Unless the opposite was manifested in the dream, participation in various SNIs was coded as indicating SASC, in particular social rituals such as initiations (e.g., receiving a ‘sacred thread’ and taking Hindu vows), weddings, cremations, and ancestral cults. Regarding the question of whether the dream imagery implied signalling of coalition and cohesion, this could be identified in 48 cases (78.8%). The type that was most common, occurring in 42 cases (68.9%), concerned imagery about named, scripted and normative actions (see the section on SNI) in combination with social signalling and interaction between humans and SAs. Again, the presence of SA imagery stands for the bulk of dreamt social agent interactions (46 cases, or 75.4%), which indicates a high frequency of social dream simulation relating to cooperation and coalition.

Coding MCR traits in dream reports and results

This coding procedure aimed at identifying dream reports containing references to objects and substances with a magical function, or to interactions that have magical effects. MCR traits were coded in dream reports referring to explicit or implicit interaction with an SA, via a priest, such as in special agent rituals (e.g., marriage, cremations, thread ceremonies, etc.), special patient rituals (e.g., thanking, cleaning, honouring, preparing, often labelled as puja, aarthi, and shraddha, etc.), or special instrument rituals, blessings by sight (darshan), touch or proximity to sacred objects (murti/mandir water) or immersion and consumption of substances (water jaal, or food, candies, flowers, fruit prasad). These are all magical items with peculiar causal characteristics and ‘essences’ that were alluded to or explicitly labelled (murti statues, mandir temples, saligram sacred fossils, prasad sacred food, etc.).

The present dream reports implied or manifested indications of both specific magical objects and types of ritual forms that implied causal functions of SAs. First, 48 dream reports (78.7%) referred to images of some type of ritual objects, instruments and/or substances, while 13 cases (21.3%) lacked any information. The most common MCR objects in dream imagery were temples mandir (18, 29.5%); (religious) statues murti (6, 9.8%); combinations of temples and statues (9, 14.8%); combinations of sacred objects such as a ritual thread and mala, tikka, prasad (5, 8.2%), and other types and combinations of objects (6, 9.8%).

When the data analysis connected magical objects to cognitive models of ritual action forms in the dream narratives (61 in total) the results indicated that special agent rituals occurred in five cases (8.2%), special patient rituals in 21 cases (34.4%), and special instrument rituals in 12 cases (19.7%), while a mixture of both special patient and instrument rituals occurred in 12 cases (19.7%), and a mixture of special agent and patient rituals occurred in two cases (3.3%). There was no information in nine cases (14.7%). According to these figures, the most common form of dream simulation was special patient ritual imagery such as various kinds of thanking, cleaning, honouring, and preparing for SAs (33 cases, 54.1%), followed by special instrument ritual imagery, especially blessings from SAs (24 cases, 39.4%). When coding for blessed objects and substances (cf. Nordin 2009), 30 reports (49.2%) referred to specified or unspecified blessings. Twenty-five interviews (41%) referred to specific named blessings such as asirbad, tikka, mala, and darshan.

As previously noted, darshan – ‘divine gaze exchange’ – is a culturally widespread idea that redemptive blessings are received through the gaze (cf. Eck 1985; Gell 1998). It was most commonly implicit, with reference to statements such as ‘I saw’ this or that (SA). Accordingly, darshan implies that the ‘worshipper’ passively receives the blessing, while a supernatural agent ‘transmits’ it. A common cultural metaphor about dreams in South Asian contexts is indicated by the use of the expression ‘to see a dream’ rather than ‘to have a dream’ (Young 1999, 29). Additionally, dreams are ‘given to’ the dreamer, which makes it clear that a dream has an external source – some supernatural agency or demonic influence – rather than originating from within the person (Young 1999, 29). Such polyphasic beliefs (Laughlin 2011) would be further examples of culturally widespread notions of realism in relation to religious dreaming (cf. Meyer and Shore 2001). In sum, since SA agents in dreams are construed as the immediate and acting party, and the dream is often held to be a ritualized sequence, this can be a specific ritual cognition – explaining why and how SA dreams gain their allure and potential religious and ritual use. More research is needed to confirm these suggestions, however.

Ritualized behaviour: coding RB traits in dream reports

These items concerned explicitly stated or implied descriptions in the dream reports, particularly with reference to (a) rigid, repetitive, redundant, goal-demoted, and scripted action sequences; (b) the dreamer being unable or finding it difficult to act, move and speak (i.e., paralysis), having to escape, being chased, being acted upon (e.g., being eaten), and a lack of reference to the dreamer in the imagery.

The frequency of omitted self /agency models in this (Nordin 2023) and general dream research is well established. In the present study, such imagery co-occurred with the dreamer as the prime agent in 12 (19.7%) of 61 cases. In 17 reports (27.9%) one or several SAs were the prime agents; while in 25 reports (41%), both the dreamers and SAs were agents. In two reports, other humans than the dreamer were the agents. These sketchy figures suggest that in this sample social dream simulation primarily concerned the dreamer and or SA(s).

Ritualized behaviour (RB) was manifest in the dream reports in 23% of cases, while it was much more frequent (67.2%) when implied by reference to stipulated named interactions (SNI) that often designate ritualized behaviour. The instances thus mostly consisted of metonymic or indexical references to culturally labelled ritual action-types in the dream reports. Another phenomenon that can be related to imagined dream ritual behaviour is the occurrence of omitted self /agency model in the reports. The presence of nightmares in the reports from this study was a predictor of the occurrence of omitted self-agency models, and this latter imagery was a predictor of divinatory practices in the waking state (Nordin 2023).

Discussion and conclusion

Research on religious dream imagery has usually examined how various types of supernatural agents and forces are manifested. These can be seen in so-called ‘big dreams’ (encounters with important deities in the religious tradition) or ‘visitation dreams’ (encounters with ancestors, the dead, relatives, and loved ones). The importance of dreams is evident in all the world’s religious traditions (e.g., Bulkeley 2007; 2008; Doniger and Bulkley 1993), and scholars in early anthropology, for example, have even suggested that dreaming serves as the primary source of religion by providing the experiential basis for religious ideas (Tylor 1871).

The present article contributes to the study of religion and to subfields in the cognitive and evolutionary study of religion and the anthropology of religion, by analysing a key feature of religion, rituals, and in particular the understudied topic of dream imagery about ritual interactions. The contributions thus seek to fill gaps in the scientific and scholarly understanding of the processes, modes, and prevalence of dream imagery about ritual. The article points to fruitful studies of dream imagery about ritual within the study of religious phenomena that go beyond the scope of supernatural or superhuman entities – something that is of interest to most empirical research and theoretical stances on religious phenomena. A prominent stance on rituals and religion is that outlined by Boyer and Liénard (among others), based on a naturalistic research programme, strong theories and empirical research. This approach was methodologically employed in the present study as (1) a typology for developing a coding scheme for analysing religious dreams, and (2) as a mid-range theory that can integrate cutting-edge research on dream simulation with the cognitive and evolutionary accounts of rituals. Arguably, the presented approach to ritual is clearly suited to be used and tested in the present study, because it relies on widely accepted and well-informed explanatory theories and a typology and ontology of ‘ritual phenomena’. It is assumed here that religious dreams typically refer to supernatural agent or entities (which are typically minimally counterintuitive), and that these dreams may exhibit one or more ritual features.

What is the evolutionary and cognitive rationale for dream imagery about specific ritual behaviours and interactions? This article presented a theoretical model and methodology suggesting that dream imagery about rituals should be understood as produced by (a) basic dream simulation theory relating to social life and threats, which are directly rehearsed in specific (b) ritual variants, which are also underpinned by evolved cognition dedicated to managing a range of fitness challenges. While argument (a) offers explanatory suggestions of the level of dream simulation, (b) offers both an established explanation of ritual variants (that may occur in dream simulation processes) and a theory-driven method for the mapping and analysis of items that potentially can be coded as ‘ritual’ in dream reports and imagery. The final section of the article provided an exemplification of how this way of reasoning can be applied to actual dream data, and as a preliminary result this survey suggests high frequencies (above 50%) of ritual variants in the dream reports.

The present discussion points to the issue of whether (religious) dreams of ritual imagery always appeared with reference to supernatural agents, or at least did so when the coding system was tested on the given dataset. Dreamed ritual features often, but not always, include imagery of SAs. When SA imagery is included in dreams of ritual action, the ritual features may be implied or manifest. According to the ritual theory adopted here, there is no necessary connection between SAs and normative interaction (SNI), signalling of affiliation and social cohesion (SASC), magical causation and reckoning (MCR), and ritualized behaviour (RB). An exception is the subcase of magical causation developed in McCauley and Lawson’s theory, where believers (in theory) assume that (supernatural) agents are responsible for the efficacy of religious rituals (2002). These ritual variants also occur in social situations and interactions with no connection to supernatural agent concepts.

As a mere empirical observation from the present ethnographic information suggests, the reference to SA imagery as a directly manifest appearance or as an indirect and implied interaction was very common in the dream narratives seen from the various ritual variants in the present data. Whether the frequency and distribution of dream imagery of SA and ritual variants in the present example is subject to sampling errors, culturally unique factors and the like, further cross-cultural research would provide a more nuanced picture.

To sum up. This article not only argues for a theoretically informed analysis, but also takes steps toward creating a method and an empirical coding system for studying ritual dream imagery. But what can we learn from this? The empirical coding exercise presented here suggests that religious dreams are highly structured by principles of ritual action, and that these appear to be related to deeper propensities for simulation that are continuous with, and related to, religious practice in the waking state. The use and refinement of such a coding schema allows for research into the religious, social and cultural relevance of specific types of ritual interactions in dreamers’ dream experiences. Although the statistics and frequencies presented here are limited, and the dream coding scheme and its use are open to further refinement, the study also reveals certain tendencies in the structure and distribution of dream imagery. Such information is of value for testing hypotheses about religious dreams and ritual theories, and for carrying out precise analysis of dream communication and the interface between dreaming and the cultural environment in which it is situated.

Acknowledgements

I would like to thank the two anonymous reviewers for their invaluable and encouraging comments on an earlier version of this article.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

Andreas Nordin

Andreas Nordin is an Associate Professor of Religious Studies, and a Lecturer in the Department of Cultural Sciences, University of Gothenburg. His primary areas of research are cognitive and evolutionary anthropology, cultural transmission, dreaming, moral notions, honour and reputation, religious cognition, and the cognitive science of religion.