KEY PSYCHOLOGICAL ISSUES IN THE STUDY OF RELIGION

The current article considers the role of scientific (experimental) psychology in the study of religion and argues that many of the questions central to the history, sociology and anthropology of religion are often psychological and hence require the use of appropriate psychological methods. Psychological study of religion differs from those other disciplines by virtue of its (a) definition of religion (in terms of individual mental states rather than culturally transmitted teachings and socially acquired behaviours), (b) methods of research (designed to elicit and examine relevant mental states), and (c) explanatory aims (concerned with the origin and development of specific cognitive events). Whilst the distinction between individual and social origin of concepts is central to psychological accounts of religion, non-psychological accounts of religion actually dwell on an interaction between the two. It is further argued that some of the key issues in the study of religion -origin of religious concepts, core religious beliefs, and universality of religious beliefs -can be most adequately tackled within the framework of cognitivedevelopmental psychology. Possible explanations are suggested for hitherto insufficient involvement of those psychological approaches in the study of religion.


INTRODUCTION: PSYCHOLOGY AND RELIGIOUS STUDIES
The fact that psychology is not commonly included among the disciplines of religious studies (e.g., Combermere, 1990) might be construed as implying that there are no distinctly psychological issues in the study of religion. Yet, many of the questions discussed in the history, anthropology, or sociology of religion could only be fully answered by using psychological research methods. In the current article, references to psychology are made primarily from the viewpoint of scientific or experimental psychology rather than psychoanalysis, transpersonal, phenomenological or other types of psychology including any form of psychotherapy. Furthermore, within scientific psychology, the emphasis is on cognitive and developmental psychology, primarily because of the relevance of those approaches to the issues addressed in this article but, also, because of a relative neglect of these areas of psychology in religious studies compared to the approaches mentioned above.
The relevance of psychology to the study of religion can be illustrated by its implicit presence in the works of eminent religious scholars. Eliade, for example, contends that it is the task of history of religion to arrive at "general considerations on the religious behaviour of man" (1959, p. 89), disregarding the fact that human behaviour as such is a topic for psychological research. Smart (1987) similarly assigns to history the essentially psychological task of identifying "recurrent patterns of religious thought [...] and experience that can be found cross-culturally" whereas psychology, in his view, explores "timeless patterns or types of religious experience" (p. 571). Yet the only substantial difference between the historian, on the one hand, and the psychologist, on the other, is that the historian looks for "recurrent patterns" in religious texts whilst the psychologist seeks to identify them in the thought and behaviour of actually existing persons. Needless to say, to understand religion as a human phenomenon fully, the work of the historian and the psychologist must be seen as complementary.
Sociology of religion likewise frequently implies psychological explanations. This is entirely appropriate given that any explanation of human behaviour in groups and societies presupposes some understanding of the individual processes. For this reason, Durkheim's view (in Morris, 1987) that religion is a culturally constructed phenomenon and thus constitutes a sociological topic rather than something that can be explained by reference to psychological factors is puzzling. Although religion undoubtedly is a social institution, whose many aspects can only be acquired through social life of the individual, religion is first and foremost a mental experience that need not have any manifest social expressions. For instance, in prayer, religious meditation, or contemplation, as distinctly religious events, there may be no observable behaviours that could be identified as uniquely associated with those mental states even when they take place in the presence of other people.
Psychological explanations are also present in anthropological accounts of religion, as Nisbet (1987) implies when stipulating that anthropologists locate religion in (a) psychic states, (b) ritual acts, or (c) awe of celestial bodies and terrestrial phe-nomena. Psychologists cannot overlook that all three sources clearly pertain to the mind. Thus, to clarify the role of each, including our reasons for performing rituals, psychological methods of research are needed. Significantly, some contemporary anthropologists (Boyer, 1994;Guthrie, 1980) acknowledge the limitations of nonpsychological theories purporting to explain origin of religion and cogently argue for a psychological theory of religion. They recognise that the study of religion cannot advance without a contribution of psychology simply because psychological questions are different from those asked in the other departments of religious studies. Consider the following examples. When we ask why religious beliefs are both similar and different across cultures; whether anything in human nature can account for those similarities; whether people from diverse religious traditions adopt certain religious beliefs under the same conditions; what experiences, if any, differentiate prayer from worship, and the like, in all such instances we are asking primarily psychological questions. Psychological explanations are therefore not an alternative approach to the study of religion but one that is fundamental and necessary. Put simply, psychological explanations deal with the more basic units of religious phenomena (i.e., at the level of the individual) than those addressed by the other disciplines studying religion (e.g., at the level of history, culture, or social groups). This is not to say that psychological methods imply or necessarily lead to reductionist accounts; rather, any complexities can be better accounted for by recognising the different levels of enquiry contained within them.
Psychological research can therefore make a positive and, indeed, vital contribution to religious studies by showing how the human psychological constitution underpins religious phenomena and why, in spite of the vast religious-cultural differences, humans from different cultures arrive at some highly similar beliefs of a religious kind. And vice versa: why, in spite of their major cognitive-psychological similarities as well as similar cultural influences, human beings often adopt different religious beliefs and practices. Information of this kind would, de facto, fulfil the historian's ideal of identifying "recurrent patterns of religious thought" (Smart, 1987) and "formulating general considerations on the religious behaviour of man" (Eliade, 1959). In contrast, studying the content of religious beliefs characteristic of different traditions or recorded in human artefacts with the aim of uncovering something about human nature (Smart, 1990) would do so only indirectly and tentatively.
Our brief review of the main tasks that distinguish psychological from the historical, sociological and anthropological approaches to religion illustrates the importance of delineating clearly the questions and methods of enquiry specific to each discipline. In particular, it is in the interests of constructive multidisciplinary enquiry to maintain a distinction between religious beliefs characteristic of institutions (i.e., public, such as those available in texts) and those of the individual (i.e., personal or private). Both kinds of belief are of interest and merit proper explanation but the methods of enquiry appropriate to each kind are different and their respective findings will have different theoretical implications.

RELIGIOUS VERSUS PSYCHOLOGICAL DEFINITION OF RELIGION
The confounding between psychological and non-psychological issues in religious studies undoubtedly stems from insufficiently precise definitions of the term "religion". Two connotations of this term are particularly relevant because they imply different methods of investigation and have a direct bearing on the distinction between what is taught, on the one hand, and how ordinary individuals assimilate the material taught, on the other. According to the first connotation, "religion" (or "religious belief") refers to a set of teachings or doctrines that make up a particular system of religious ideas such as, for example, Buddhism, Christianity, Islam, and other religions. Such beliefs are often formulations by the founders of particular religious systems or reinterpretations of those formulations by later theologians. This sense of religion is doctrinal-theoretical and as such requires textual, or hermeneutic, methods of investigation. Thus, for example, to explain the beliefs of religious founders and leaders such as the Buddha, Luther, or Muhammad, scholars have to rely on texts in order to provide coherent accounts of ideas expounded in those texts. Mainly historians and theologians are engaged in such work. When, however, we begin to ask how human beings in general practice such structured beliefs and organise their social life in accordance with them, we adopt anthropological and sociological methods of investigation.
The second sense of "religion" pertains to the beliefs, emotions, and behaviours of ordinary human beings, which arise in response to, or are about, some specific religious teachings and doctrines as well as from one's experience in the world that might give rise to concepts of a religious kind. This sense of religion requires methods of research by which we can investigate directly the experience of the individual rather than the group. Put simply, if our aim is to explain the beliefs of actually existing, ordinary individuals, who are situated within Buddhist, Christian, Islamic and other religious cultures, our interests lie in the domain of psychology.
Psychological definition of religion thus cannot ignore the relationship between what is taught, on the one hand, and how ordinary individuals assimilate the material taught, on the other; rather, such a distinction is a major concern in psychological research. In religious studies, by contrast, no corresponding distinction is made between doctrines and traditions, on the one hand, and people's beliefs about the matters expressed in doctrines, on the other. Rather, the prevailing phenomenological definitions of religion allow textual scholars constantly to shift from one sense to another. For instance, they may start with the aim of explaining what the Buddha or Jesus or Muhammad meant in the respective texts. But in the process of doing so, textual scholars of religion manifestly rely on their own experience or that of other people, whether actual or assumed, and generalise from this experience to a wider population. Some degree of confounding between psychological and nonpsychological issues and procedures is understandable in view of our intrinsic inter-est in the nature of our own mind; nevertheless, explanatory goals require that they should be clearly differentiated.

RELIGIOUS SCHOLARS AND PSYCHOLOGY
Seeing that the nature of religious experience as such is a psychological topic, it is paradoxical that most of the thinking and writing about it has come from nonpsychologists. One of the main reasons for this anomaly is an outdated conception of psychology held by many scholars of religion. Three examples of such misconception may suffice. The first is a widespread notion that psychology as virtually synonymous with psychoanalysis. Freud and Jung are the most often cited sources in any literature on religion where psychological factors are discussed (e.g., Segal, 1989Segal, , 2000. Yet, because psychoanalysis is concerned with affective and unconscious motives in human behaviour rather than cognitive processes, this approach could not even in principle explain the nature of religious beliefs (i.e., cognitive mental states), notably those of a basic kind. Secondly, and closely linked to the above, is a limited understanding of modern scientific psychology and the nature of its methods. Accordingly, religious scholars question the possibility of objective psychological study of a "religious phenomenon" on the grounds that religion as such is a subjective (i.e., mental) phenomenon (e.g., Smart, 1987). The fact is, however, that modern psychologists have been quite successful at designing methods that allow objective study of mental states (i.e., through experimental research and under laboratory conditions). Because other researchers can replicate those procedures, they cannot be "subjective". For instance, modern psychology has developed methods for investigating abstract concepts in preverbal infants, logical inferences in children and adults, hypothetical reasoning, development of both scientific and philosophical abstract concepts (i.e., concepts of unobservable entities), to mention some of the more challenging areas of cognitive-developmental research. There is, therefore, no obstacle that necessarily precludes the use of the same methods for studying religious concepts provided that the ontological distinction between the concept, on the one hand, and its referent, on the other, can be maintained (Petrovich, 1994). Yet, the tendency to conflate the concept of something (i.e., mental state) and the concept's referent (i.e., the thing itself) has led to the view that religion can never be explained by psychological study. Thus Eliade's claim that psychology "misses the element of the sacred" (1959, p. 21). Whilst it is true that a religious concept refers to something non-empirical (i.e., "holy other", "sacred" or God), the concept thereof is an empirical event (mental state) and the conditions of its occurrence belong in this world. Consequently, the empirical features of this concept are a matter for psychological research whilst its ontological referent (transcendent realm) is manifestly not but constitutes a topic normally pursued by theologians and other scholars of religion.

PSYCHOLOGISTS AND RELIGION
The weak interaction between scientific psychology and study of religion is only partly explained by the misconceptions about psychology among religious scholars. We also need to consider why experimental psychologists have tended to ignore religion as a domain of human cognition. A number of factors may have contributed to this. To begin with, standard psychological education does not prepare researchers for tackling specifically religious issues so that few psychology graduates have clear ideas for research in this domain. Further, the terms "religion" and "religious experience", even when understood as mental states, are too broad to be seen as translatable into operational definitions (which are an essential requirement in psychological research). For instance, the term "religious experience" is commonly used in the literature to refer to conversion, prayer, mysticism, and worship, without, however, indicating any criteria in terms of which those experiences can be differentiated (or, what common underlying characteristic they share). Most empirical psychologists wouldn't know where to begin when faced with such broad phenomena. Finally, the terms "religion" and "religious experience" are typically used to refer to doctrinally committed and practising adults who declare themselves to be religious (e.g., Brown, 1988;Beit-Hallahmi & Argyle, 1997). Those who do not make such a declaration -children, unchurched adults, undecided -are presumed to be non-religious. There is no doubt that study of well-defined religious groups is relevant to psychology and considerable research involving such groups has been done by social psychologists (e.g., Argyle & Beit-Hallahmi, 1975). But there is a clear need to account for some of the less well-defined forms of religiousness also, those that have no visible behavioural correlates and hence require methods capable of eliciting any implicit, or incipient, religious mental states. In other words, the nature of human populations considered by religious studies, on the one hand, and modern psychology, on the other, is one of the more salient differences between the two approaches to the study of religion. Thus psychology presupposes that the whole human population (i.e., its representative samples) can be compared with respect to any characteristic that can be observed or measured, and that only systematic comparisons of this kind permit generalisations. In the case of religion as a psychological variable, this implies that every living individual can be identified as more or less capable of acquiring religion rather than being totally devoid of any of its components. By contrast, the traditional disciplines of religious studies focus on already selected groups (e.g., practicing Christians or Muslims) and dwell on the specific cultural and social differences among them, thus arriving at findings of a more limited relevance to the general population.

COGNITIVE PSYCHOLOGY OF RELIGION: PROBLEMS AND PROSPECTS
To apply the statistical model of the normal distribution to the study of religious belief requires that data should be obtained not just from those who declare themselves to be religious but also those who may see themselves as outside this definition. For people often misunderstand their position when they perform imaginary comparisons with the rest of the population. They may say that they are agnostic or even atheist yet hold the same basic views about the transcendent as those who consider themselves to be religious. A recent example that illustrates my point is the novelist Jim Crace who was described by The Daily Telegraph (September, 1999) as a "dogmatic, unrelenting, and hard-nosed" non-believer yet admitted that he shared with religious individuals openness to transcendence and spiritualism. In order to find out if indeed most atheists and agnostics agree with religious believers on the basic idea of transcendence, we would need to examine a large section of the population in order to consider it representative of the diversity of views encompassed by Crace's assumption. Such a task is distinctly psychological in that it requires the use of proper sampling procedures; testing subjects under clearly defined and systematically controlled conditions; obtaining sufficiently large data sets; and, finally, applying appropriate statistical techniques in data analysis. The more commonly used survey is a useful technique for gathering preliminary information about certain questions; however, it is not suited to finding answers to specific questions (i.e., testing hypotheses).
In short, many of the problems in the study of religion mentioned above could be overcome by a greater involvement of cognitive as well as developmental psychology. To illustrate the relevance of those psychological approaches, we look at three issues that have been singled out as of central importance in historical, anthropological, and sociological studies of religion yet are fundamentally psychological. These include (1) origin of religion; (2) core concepts in religion; and (3) universality of religious beliefs.
(1) Origin of religion is an issue of vital importance in any attempt to explain religious phenomena adequately. Whilst it is proper to consider how religion originated in human society or in particular geographical contexts, the more basic question is how it originates in individual human development. We simply do not have the information needed to reconstruct reliably the earliest forms of human society and to explain accurately how religion first appeared as a social institution. By contrast, the availability of actually existing humans of different ages and backgrounds provides us with opportunities to investigate many important questions pertaining to the origin of religion. The distinction between how ideas originate in individual development, on the one hand, and in human history, on the other, as pertinent to the study of religion is rarely made in non-psychological sources. Freud's theory is the most widely cited "psychological" account of the origin of religion although it actu-ally deals with how religion began in human history (Totem and Taboo), which Freud assumed to be a pattern that replicates itself in individual development. Freud's ideas later influenced several developmental accounts of religion in childhood (e.g., Bovet, 1928;Elkind, 1964;Piaget, 1925;Sully, 1903), all of which take for granted his assumption that the "filial sentiment" is the source of personal religion, namely, that parents are our initial deities.
A non-psychological theory of the origin of religion that seems to be consistent with certain psychological principles is the "intellectualist" theory (Skorupski, 1976). According to this theory, religious beliefs are plausible hypotheses so that even primitive religion can be said to be rational. Modern psychology substantially corroborates emergence of rationality in early development, as findings about children's scientific and philosophical theories and hypotheses indicate (e.g., Wellman & Gelman, 1998). To verify Skorupski's intellectualist theory of religion, however, we need to test empirically a number of psychological questions, which he himself did not envisage but sought to explain religion as a sociological topic entirely.
If it is indeed most appropriate to tackle the issue of origin of religion by examining individual development, the question arises when humans begin to acquire religious beliefs. To answer this question adequately, our earlier distinction between beliefs as doctrines, on the one hand, and beliefs as mental states of ordinary people, on the other, is of crucial importance. Doctrinal beliefs are transmitted through culture and thus need not be representative of people's everyday (often spontaneous) religious understanding. Spontaneous (i.e., "untutored") beliefs are of special interest in psychology because they can be triggered by some maturational (i.e., natural or innate) mechanisms which determine how and when we acquire such beliefs (Petrovich, 2000). The distinction between innate and acquired concepts is an old question that modern psychology has inherited from philosophy and which continues to be in the centre of modern developmental research (e.g., Elman et al., 1996;Karmiloff-Smith, 1991). Innately prepared concepts not only occur spontaneously in human development but are also basic or simple, unlike concepts that are transmitted through instruction. Interestingly, the relevance of innateness to religion was noticed by Otto (1923Otto ( /1979, who argued that the first and central task of studying religious behaviour was to establish the development of rudimentary religious concepts, in particular that of the numinous (i.e., the holy), which task he rightly saw to belong in psychology.
In short, of all the disciplines involved in the study of religion, psychology alone is explicitly concerned with conceptual development from childhood to adulthood and allows us to examine separately the spontaneous or natural component of religious thought and experience from the doctrinal or culturally transmitted one. One way of achieving this is through cross-cultural research whereby we can arrive at theories that explain different aspects of human religious disposition. In addition to the question of how people acquire religious beliefs, psychology can contribute to a better understanding of why people abandon religious beliefs or replace them with alternative beliefs. Further, psychology can ask whether the same factors cause similar changes in belief at all stages in individual development; whether the same factors are operational in all cultures, including those where religion and science have had a pattern of interaction different from that in the West (e.g., Japan); and other relevant questions. Joint application of the historical-comparative method and comparative-psychological research would thus significantly strengthen our explanatory framework for answering many of the perennial questions about religion as a universal phenomenon in human experience.
(2) The controversy regarding the existence of any core religious beliefs or concepts has been prominent in religious studies. Scholars from a variety of backgrounds have persuasively argued their particular stance on this issue and opposing schools of thought have emerged. There is, however, a major difference in the understanding of what constitutes "core" religious concepts in religious studies, on the one hand, and in psychology, on the other. Whilst scholars of religion debate whether or not there are any common doctrinal beliefs across different religious traditions, the current psychological view is that core (or foundational) concepts are those very basic conceptual categories that underpin our learning and knowledge acquisition in a particular domain. Examples of psychological core concepts include object, space, time, and quantity. Such concepts are thought to emerge spontaneously and early in development, are simple, and likely to be found in all cultures (e.g., Wellman & Gelman, 1998). Moreover, core concepts are few in number and psychologists have so far examined mainly those that occur in the physical domain. The issue whether any concepts or beliefs constitute a core component of religion itself is a researchable question and as such ought to be verified empirically. The ongoing controversy regarding the existence of such concepts will continue as long as no distinction is made between religious beliefs of ordinary individuals, especially in their early development, and those beliefs that are recorded in religious and historical texts, usually as a result of a prolonged scholarly effort.
(3) Universality of religious beliefs is an issue that is closely linked with that of core beliefs and has consequently been an important topic in comparative religion. The comparative approach as a historical method aims to detect similarities and differences between doctrines and rituals characteristic of diverse traditions in order to determine any universally occurring patterns of religious thought in the history of religious ideas. As already mentioned in this article, historical accounts include some psychological theorising as well. In other words, when using the comparative approach, historians of religion do not adhere strictly to textual analysis but often adopt an implicitly empirical, i.e., psychological, approach. Their actual approach fluctuates between explaining the text, as a set of formalised beliefs, on the one hand, and construing a wider meaning of those beliefs for ordinary individuals in response to such doctrines, on the other. It is important to recognise, however, that the hermeneutic approach, although challenging and creative for students of religious traditions, is insufficient to show if there are any universal religious beliefs in the human population. For texts are typically produced by sustained intellectual effort of a small number of specialists and as such need not be representative of the beliefs of ordinary individuals. To establish whether members of the human population universally hold any particular religious beliefs, we also need methods of research that are used in cognitive and developmental psychology when dealing with large numbers of participants.
The possibility that some religious beliefs are universal (e.g., basic belief in a non-anthropomorphic God as creator of the natural world) seems to have a stronger empirical foundation than could be inferred from religious texts (Barrett, 2001;Petrovich, 1997Petrovich, , 1999. Some of the initial findings of research into early religious understanding are consistent with other areas of developmental research which suggest that there are cognitive universals in a number of domains of human knowledge (e.g., R. Gelman, 1990;Spelke, 1988;Wellman & S. Gelman, 1998). Search for invariants in human religious cognition requires data from children and adults alike as well as comparisons between individuals from different cultures (Petrovich, 2000). Such evidence is relevant because it contributes towards establishing criteria or empirical conditions under which humans show a reliable tendency to think in terms of some basic religious categories such as God or the transcendent. It is a mainstream psychological position that human behaviour is not infinitely variable (Broadbent, 1961) but can be reliably linked to certain empirical conditions, including those that give rise to abstract concepts where religious concepts belong.