Methodological Integration in the Study of Religions

Era di mana segala sesuatu semakin terintegrasi seperti sekarang ini, ilmu pengetahuan dan pendekatan terhadap ilmu pengetahuan tidak dapat dipisahkan satu dengan yang lain. Bahkan, ilmu pengetahuan dan pendekatannya (metodologi) terintegrasi satu dengan yang lain. Integrasi metodologi itu juga terjadi dalam studi agama (religious studies). Dalam rangka mencapai pendekatan studi agama yang terintegrasi, sumbangan dari ilmu dan ahli-ahli lain di luar agama perlu diperhatikan. Termasuk pandangan mereka tentang agama. Art ike I ini bertujuan menjelaskan bahwa sebenarnya sebuah integrasi metodologi di dalam studi tentang agama, bukan merupakan sebuah hal yang sulit dilakukan. Misalnya pendekatan sejarah, psikologi, dan etnologi dalam dilakukan bersama tanpa harus dipisahkan.

These pages present a call for an integrated approach to the academic study of religions which does justice to its specificity, but without separating it artificially from other related avenues of research. For a discipline to reflect upon its methods is a normal part of academic endeavour, and this applies to the study of religions (or Relig-ionswissenschaft1) as much as to any other scientific research. This statement implies, and is intended to imply, that the study of religions may be regarded as a "discipline". "Religions" constitute a field of study and accordingly "the study of religion (or religions)" is a discipline. What is a discipline, that is, in the scientific sense? It is no more, and no less, than a methodically ordered approach to the study of a field. The field "religion(s)", no less than any other fields, requires a methodically ordered approach for its study. The methodically ordered approach, the discipline, takes on particular characteristics as required for the best study of the field. Consequently, the discipline of the study of religion(s) is not necessarily quite the same as the discipline required for the study of other fields, though it may be rather similar to the discipline required for the study of closely related fields.
The view of the field and the understanding of the discipline interact with each other. A stable methodological perspective corresponds to a stable view of the field. The destabilisation of either leads to the destabilisation of the other. However, an advance in methodology The German term (like its equivalent in various languages) has the advantage of including the element "science" in it, but the disadvantage of referring to religion in the singular. Care should be taken to avoid the term which puts the sciences into the plural, namely Religionswissenschaften, for this suggests on the one hand that "religion" is one, idealised entity, while on the other hand avoiding the strenuous task of being clear about what the appropriate science for its investigation is. may lead to a correction in the view of the field, and on the other hand, newly perceived or newly emergent features in the field may lead to pressures on currently held understandings of method. While openness to the recasting of perspectives is desirable, one may hope nevertheless for a certain, relative stability in the understanding of both field and discipline, for otherwise the critical interaction between individual investigators typical of a "science" cannot function at all. It is to be hoped that conferences on the subject of methodology in the study of religions, as famously held in Turku, contribute to the stabilisation process.' When there is relative stability, the discipline can be learned, practised, taught, corrected and developed.
The understanding that there is, and indeed must be such a process of methodological development and reflection does not imply that the study of religions has some one special method, unique to itself. At the same time the discipline of the study of religions requires its own particular gathering, or as we might better say, clustering, of methods. Though the methods at our disposal are in themselves known in the context of other disciplines, they are brought together in a particular way in order to facilitate the study of the precise field in question, namely religions. The resultant discipline is not quite the same as the disciplines required for the study of other fields, or of fields differently defined.
It is desirable to clarify, at this point in the argument, the nature of the specificity which the discipline requires and the reasons for which it should be affirmed. It arises firstly for the simple reason that there does not seem to be any other one, single discipline which could plausibly claim to be, alone and precisely, the discipline required for the study of religions. For example, "history" does not quite fit the requirements, because it does not usually include the methodological niceties of carrying out fieldwork among living people. Nor however does "sociology", because in general, quite correctly in its own terms, it subordinates the study of religious ideas and behaviour to wider questions about the nature and functioning of society. Such questions are of course valuable, but there are other questions of interest concerning religious idea-complexes, for example questions about their internal structure and dynamics, which are not necessarily "sociological" in nature. For analogous reasons the disciplines of anthropology, art history, archaeology, political science, and so on, also do not amount to just that discipline which is required, overall, for the study of religions. Unfortunately the use of the words "autonomy" or "autonomous" have sometimes been subject to misunderstanding or to misuse in this connection. This is because they have frequently been associated with an "essentialist" or "sui generis" view of religion as a unitary phenomenon, that is, with the idea that behind all the various religions there is some unifying essence which only specialists in religion can understand and which makes their study different in kind from the study of anything else. This position is by no means adopted here. Nor shall it even be discussed at this point, since such a view of religions is not relevant to the argument being advanced.3 It is quite a different matter to point out that none of the other disciplines currently practised in the human and social sciences specifically and adequately relate to the field of "religions". In some way or other they fail adequately to explore or elucidate the subject matter. Some do too little, and some, it might be said, do too much. This does not mean that the study of religions requires a special method which is unique to itself. What it does mean is that the right selection of available methods must be made and that these must be clustered together in a manner appropriate to the subject matter.
While it is necessary to realise that a specific clustering of methods is necessary to maintain and develop the discipline of the study of religions, it is not necessarily important to achieve complete agreement about what this clustering of methods should look like. Consequently there is no intention to offer a dogmatic statement about it here. Nevertheless, after clearing the way with some notes on present difficulties and the reasons for them, the following presentation will seek to show what such a clustering of methods might reasonably be expected to look like. The statement is formulated in what may appear to some to be disappointingly uncomplicated terms. However, this is intentional and is regarded here as an advantage. Simplicity is a strength, not a weakness. It is anticipated that those who are themselves engaged in the study of religions, in practice, will find it relatively easy to reach broad agreement along these lines. And indeed it is important, while continuing the methodological discussion within the discipline, that there should be a widely recognisable tradition of study which can be identified as "the study of religions" (or whatever formulation is preferred). Indeed, it may be maintained that to some extent there is already such a recognisable tradition of study, even if it is in need of greater crystallization.

Reasons for some present difficulties
Unfortunately, in spite of much attention to methodological questions in the study of religion there continues to be uncertainty, vagueness, and even irresponsibility in not a few quarters. Why is the methodological identity of the study of religion so widely misunderstood? There are various reasons.
First, it is deplorable that basic distinctions which ought to be easily understood continue to be slurred over or dismissed as trivial. A classic example of this is the difference between studying religious statements and making religious statements. It is remarkable, but true, that even today, after decades of methodological clarification, it is still necessary to make this distinction clear. Again and again, theologians appear who confidently assert that they are making statements which pertain to Religionswissenschaft, when they are in fact giving a religious analysis of some cultural situation. It is not surprising that other members of the public, even of academe, cannot take the trouble of making this distinction. However, as most real specialists in the study of religions would agree today, it is quite significant for the study of religion that it should not be identified with the making of religious statements. That would be a matter for theologians, Buddhist apologists, neo-shamans, and many others.
Second, there is a certain amount of intellectually obstinate compartmentalization furthered by the use of conventional phrases such as "comparative religion", "phenomenology of religion", "anthropology of religion", "psychology of religion" and so on. Though these are usually recognised to have a certain history, which is rehearsed from time to time, it is not so common to see them assessed conspectually and critically, with a view to their correlation, integration or abandonment as might be required. More commonly they are just listed as options which people may take up as they please. However if the field is regarded as coherent, then a greater degree of methodological coordination, or even integration, is intellectually desirable and ought therefore to be sought. For example "comparative religion" or "comparative study of religions" cannot really exist by itself. Nor can "ethnology of religion", in spite of the immensely valuable contribu-tions of those working at the interface between ethnology and the study of religions.4 Third, persons coming freshly to the subject often bring with them methodological perspectives which have been strongly formed in other disciplinary contexts. This is often enriching, but can also perpetuate mistaken assumptions and misunderstandings about the study of religions. Thus it sometimes happens that a person who has been trained as an anthropologist or ethnologist, and who goes on to specialise in religion, simply does not go to the trouble of acquiring a methodological orientation in the discipline of the study of religions. Humanly speaking, this may be acceptable in itself, depending on the case and the situation, but it becomes irresponsible when younger students, new to the subject, are told that the study of religions as such has no particular method. In such cases it appears that the researchers in question feel a professional need to continue to be identified above all as whatever they were before. Anthropologists, for example, once they have undergone their double intiation through field work and first publication, are sometimes a bit like boy scouts who have the saying "Once a scout, always a scout". The result is a failure to achieve "discipline identification"5 or integration with respect to specialised, or new fields of study such as "religions". A fourth reason for a certain amount of confusion is the development of serious methodological divergence as the result of an interest in new lines of thought which seem to make their own methodological claims. Sometimes new insights in a particular direction seem to demand to take over the methodological discussion entirely, while earlier gains are despised or forgotten. For example, because it is interesting to consider religion as a pattern of brain operations, we are tempted to regard cognitive science as the appropriate method for studying religions. If we are not careful, the need for fieldwork, for textual studies, and for disciplined comparison may then be forgotten. Putting it more generally, it is not infrequent for interesting figures such as Claude Lévy-Strauss or Michel Foucault to make the running, creating a bandwagon effect which disregards some of the everyday methodological requirements of the study of religions. The impact of various intellectual currents must surely be taken up 4 Phrases built on the pattern "ethnology of x" and equivalents in other languages such as "X-ethnologie" are easily framed but usually very imprecise in their meaning. 5 Although it may sound somewhat forbidding, this phrase (Pye 1991) refers to a normal and appropriate process in any discipline which is enriched by recruits from varied quarters. keenly by specialists in religion, as in the case of other disciplines, but at the same time it is necessary to work out carefully where the possibilities of integration lie. Otherwise tested and worthwhile methods will simply be scorned or forgotten in favour of a series of fashions.
Fifth, in recent years there has been an increasing recognition that the "history of religions" is not, and indeed never really was quite the same as "history" in a looser or more general sense. The adumbration within the field of history implied by the adjunct "of religions" implies an incipient theoretical horizon. It has therefore been asserted not infrequently that "history of religions" somehow brings along with it the systematic, comparative or typological study of religions. However, this is not enough. Simply to make this connection does not provide the methodological integration which we require. Moreover this stance deflects attention from the possibility of extremely valuable field research among the numerous religions open to direct study today. It is adopted, typically, by those who prefer to reject out of hand the methodological contributions of the various social sciences in favour of "the historico-philological" method. The approach also obscures the important point that "comparison" may be carried out both with respect to the internal characteristics of religion (leading to the typologies typical of the phenomenological school) and also with respect to functionalist explanations over the much wider range of sociological and psychological research. One cannot simply say that it is the "comparative" part of research which somehow makes the study of religions systematic and therefore scientific, or that this feature in itself makes it a distinctive discipline.
Sixth, the argument has moved forward in recent years. It has become widely accepted, contrary to the last mentioned trend, that "history of religions" can only stand in a full sense for the "study of religions" if the latter iself is also understood to be located within the overall range of the social and/or cultural sciences. Nevertheless these two major wings, the historico-philological (often with an emphasis on the study of texts) and the social-scientific, are still sometimes contrasted, as if inimical to each other. The recent debate over the name of the International Association for the History of Religions (IAHR), conducted during the years 1990-95, reflected these tensions, although it also had pragmatic aspects. In general it may be said that, because of their varying academic formation throughout the world, representatives of various trends in this discussion did not always find it easy to understand one another. This was the case even when some important positions were in fact shared, as in the contributions by Ugo Bianchi and Donald Wiebe.6 For all of these reasons uncertainty and lack of direction is often sensed by students and younger researchers. Nevertheless, it is argued here, considerable agreement can be perceived in the experience of specialists about how to go about studying religions. We will now turn, therefore, to strategic considerations for the development of an integrated methodology for the study of religions.

Strategic considerations
It is no longer sufficient simply to set out in a miscellaneous list, as has often been done, the apparently varied tasks of history, comparison, phenomenology, hermeneutics, sociology, psychology, phenomenology and so on. What is needed is to make the necessary effort to correlate and integrate clearly those features of academic (or in some languages "scientific") method which are particularly necessary in the study of religions. This will make truly inter-disciplinary discussions with specialists in other disciplines far more fruitful. What, then, are the key strands in a methodologically integrated study of religions? Without claiming finality, this paper will now continue by giving a broadly conceived answer to this question. Three focal points in the articulation of an integrated methodology for the disciplined study of religions will be first briefly mentioned and then treated in more detail below.
First is the relation between subject-matter and method. Certain methodological orientations arise out of the simplest available morphology of the subject-matter, namely in terms of four elementary aspects of religion. This amounts to an adumbration of the field to be studied rather than a pointed definition of the object of study. The four aspects to which attention is drawn are: the behavioural, the conceptual, the subjective and the social. This enumeration is reduced here to a form which is as simple as possible without gross omission, and further details and argumentation thereon may be found elsewhere (Pye 1972;Pye 1994). It will also be noticed that these four elementary aspects are enumerated at such a level of abstraction that they can also be discerned in other subject-matters, e.g. sport or politics. However as soon as the pattern is filled out with 6 Their statements, and other related contributions, are preserved in the informal IAHR bulletins between 1990 and 1995, when the discussion was taking place. an example (or "a case") of religion, certain methodological requirements emerge quite clearly which may not be applicable in quite the same way to all other fields of research. These will be explained below.
The second focus is the relation between sources and method. Sources are not the same as subject-matter. The subject-matter is a complex set of socio-cultural data for which sources provide evidence. The methodological question here is, therefore, how the sources in question should be studied. Thus the focus on sources gives rise to secondary methodological orientations arising out of the threefold nature of the primary sources available for study, namely written, oral and material sources. There is no unique method here which is particularly characteristic of the study of the religions. However there is a characteristic clustering of methods which arises out of the particular grouping of the sources which are relevant. As will be seen, one of the most important requirements in this regard is to achieve a coherent correlation between the "historico-philological" method and the methods typical of fieldwork in "living" and "oral" situations.
The third focal point lies in the methodological requirements of theory formation. It is necessary to distinguish between "theory" and "method", because an interest in new theories has often been mistaken for methodological advance. For example, a theory on gender relations in religion, or an interest in semiotics or cybernetics, does not necessarily imply an advance or a change in methodology as such. Admittedly, new theoretical positions may lead to some methodological adjustment. However there are two major aspects of method which contribute in particular degree to the development of categories and theories in the study of religions, namely: comparison and contextualisation. Since these are not exciting, like new theoretical approaches learned from elsewhere, they are sometimes neglected and scorned. Sometimes, too, they are over-emphasised. The main problem here, as a third step, is to correlate them appropriately with the requirements which emerge from the subject-matter and from the sources available.

Subject matter and method
These three focal points will now be explained in a little more detail. As indicated above, the enumeration of the conceptual, behavioural, social and subjective aspects of religion is regarded here as being the briefest possible indication of the subject matter which maintains a holistic view of it. That is, the enumeration enables us to think of the subject matter at once aspectually and conspectually. It may be that the same enumeration could be applied to other subject-matters, but the picture takes on colour in the study of religions when a particular religion or religions are regarded in this way. Any further delineation leads into increasingly complex questions of morphology and typology, about which differences of view might increasingly arise. However, the disagreements would be theoretical, not methodological. At the level of methodological reflection currently entertained it does not matter if views differ about the way in which morphological theories might be developed in greater detail, e.g. by listing more "dimensions", as done by Ninian Smart for example (Smart 1996). It should be noted therefore that, at this point, I am concerned only with the elementary methodological principles which arise from the simplest possible delineation of the subject matter.
The first requirement is that as far as possible, that is, as far as the available sources and research facilities permit, all four of these four elementary aspects should be considered in their integral relation to each other. Stating it negatively, for example, religious ideas should not be studied as if they had no relation whatever to religious behaviour. If this is done the researcher is likely to end by simply contributing to the further development of the religious tradition in question (as many pursuing "religious studies" in fact do). Similarly, the subjective aspects of religion cannot be completely separated from their conceptual accompaniment, a point which it has still seemed necessary to argue quite recently, and widely, in connection with mysticism.7 Or again, the social forms of religion should not be studied as if it does not really matter what the people involved think, feel or do. That is to say, the conceptual, subjective and behavioural aspects should be taken into account at the same time. For practical purposes a partial study may be undertaken, concentrating on one aspect by itself, but at least it should be recognised that the other aspects are latently relevant. In other words religion should be studied both aspectually and conspectually.
The second methodological requirement arising at this same level of analysis is that the poly-aspectual subject-matter should be studied, in the first instance, in terms of its integral meaning for the believers or participants in question. That is to say, it should be studied without reference to the value orientation or possible explanatory hypotheses of the researcher. If this is not attempted, the emergent See Mysticism and Religious Traditions (Katz 1983), a multi-authored work in which all contributors take this view. characterisation is very likely to be misleading in some significant respect. Naturally, it is perfectly legitimate, and indeed desirable, to proceed at a later point to questions of explanatory theory, and indeed later still into questions of truth and value which go beyond the task of the study of religions (Religionswissenschaft) as such. But in the first instance the study of religions should be recognitional, that is, the integral meaning of the subject-matter for the believers or participants in question should be recognised in its own right. It is this which requires to be elucidated and characterised in the first instance. Otherwise mistakes will surely be made which will vitiate any other forms of enquiry or debate. The word "recognitional" is newly coined to express this because of problems with other previously used terminology, as will be illustrated in the next paragraph. This feature of the necessary method includes a) elucidation and b) characterisation, two steps for which the wider discussion of hermeneutics is relevant.
This second methodological requirement has in fact been a commonplace in the study of religions since the emergence of the phenomenological tradition (in the study of religions), the term "bracketing" having become popular to express it. Unfortunately the point has often been obscured because it has been found necessary to reject other emphases found in the work of those who supposedly espoused it. In particular, it has been shown many times that leading representatives of the "phenomenological" school did not in fact proceed phenomenologically in this sense, or at least not consistently. Rather, they pushed and pulled their materials into more or less theological categories derived from or characteristic of Christianity. G. Van der Leeuw and Friedrich Heiler are prime examples of this.8 In spite of this deficit it is very important that specialists in the study of religions should continue to attempt to study them as systems which have meaning for their believers or participants. Previously I have tried to preserve at least the adverb "phenomenologically" to indicate this important methodological requirement. In view of the dense forest of potential misunderstandings, however, I have now decided to abandon it altogether. That is the reason for the introduction of the word "recognitional". Earlier, like others, I have usually stressed the importance of the "self-understanding" of the believers, and I believe that Jacques Waardenburg has been making a similar point by re-8 i It is hardly necessary to go into this in detail, but attention may be drawn to recent assessments of van der Leeuw (by Jacques Waardenburg) and of Heiler (by myself) in Axel Michaels Klassiker der Religionswissenschaft (1997). ferring to their "intentionality" (Waardenburg 1986: 241 ff.). However both of these terms focus a little too heavily on the conceptual aspect of religion. The term "recognitional" means that the researcher gives full recognition to the complex of experience covered by all four main aspects of religion for those who are involved in it.
This argument includes the idea that "specialists in the study of religions should continue to attempt to study them as systems which have meaning for their believers or participants", to repeat the phrasing already used. The word "attempt" is used deliberately here, for it is commonly held today that total objectivity or non-subjectivity simply cannot be achieved. This is not the place for a general discussion about the viability of a "value-free" science. However I firmly reject the oversimplified view that, because it is difficult to study religious systems in their own terms, this should not, and may not be attempted. To accept such a view would lead away from science into mere arbitrariness, and simply allow old prejudices to be replaced by new ones. However sophisticated the epistemological discussion becomes, there remains a difference between achieving a good elucidation and characterisation of the religion of a specified group of people, or getting it all wrong because one's own beliefs and values continually get in the way.
The third methodological requirement arising out of the subjectmatter as delineated above is that, even while proceeding on the one hand recognitionally, attention should also be given to the potential emergence of questions or insights which stand in tension to, or cut across, the self-understanding of the believers or participants. This tension increases with the move from elucidation towards characterisation and into explanation. As a result the tension arises for the following three reasons, which may amount to a particular characteristic of the methodology appropriate to the study of religion as a complex, but integrated enterprise. a) Within any one example studied, a structure may appear which is not apparent, or only partially apparent, to the believers or participants in question. The researcher's perception of this structure may therefore be more "correct" than that of the believers or participants (in so far as they are interested in the matter at all). At this point therefore the first degree of tension arises over against the idea (which used to be designated as "phenomenological") that the believers are "completely right" (Kristensen 1960: 14). b) The structure of any one religion may be rendered more visible as the result of comparative studies, that is, the as yet continuing, recognitional study of further cases. Though any one study in it-self will continue to be recognitional, the theoretical perspective resulting from comparative knowledge may not be visible to the believers or participants, and if it becomes visible it may not be acceptable. This is the second degree of tension. c) Finally, the intersections of any one of the four aspects set out above with related historical or socio-cultural contexts are likely to give rise to correlational reflections which require, and suggest, explanation in the stronger sense of the word. This is the normal task of those wider disciplines such as sociology or psychology which have a strongly explanational orientation. However it also applies to other contextual studies such as intellectual history, in so far as it includes the history of religious ideas as part of a much wider whole, or to contextual behavioural studies of different kinds such as research into the operations of the brain.

Sources and methods
The enumeration of the four basic aspects of religion has allowed and required us to make the first steps in the definition of the necessary methodology for their study. As noted earlier, further delineation of the subject matter in detail leads into questions of morphology and typology, and only secondarily into methodological questions. The next major step in the identification of a correct methodology lies elsewhere, namely in a general view of the sources. Again, the simplest possible view which does justice to the whole is preferred. This is as follows. The sources for the study of religions fall into three major groups: written sources, oral sources, material sources. "Material sources" here includes artifacts, buildings, non-verbal symbols, bodily positions and movements, etc.. The order "written, oral, material" reflects nothing more than the order in which they have, historically, come to be perceived as relevant. It could be reversed or jumbled. However the perception of the importance of all three is important, and does not always occur. For example it seems to be rather neglected in the collected essays of Kurt Rudolph (1992). While each of the three major classes of source has attracted its own methodological debates in the past (hermeneutics, problems of access, and so on), it is important today to correlate them in an integrated fashion. Successful correlation at this level will help to stabilise the discipline of the study of religions. Each of these three main kinds of source has leading characteristics which overlap with those of the others. Taken severally, the leading characteristics of the sources are as follows. (a) Written sources are linguistic, mainly historical, and only to a lesser extent field-based. (b) Oral sources are linguistic, mainly field-based, and only to a lesser extent historical. (c) Material sources are above all field-based and historical, and only in a derived or contextual sense linguistic. Thus, it will be seen that each of the three main types of source shares a leading characteristic with one of the other two, the common leading characteristics being as follows. Written and oral sources are preeminently linguistic, when compared with material sources. Oral and material sources are preeminently field-based, when compared with written sources. Material and written sources are preeminently historical, when compared with oral sources. That there can be a natural integration of these perspectives in the service of the study of religions can be illustrated in an easily conceived diagram.

OS Oral Source WS Written Source MS Material Source
A Linguistic knowledge required B Fieldwork required C participants/believers often not available Some of these relations will seem immediately obvious, but others may be less so. Oral sources are primarily field-based, and in further detail attract the modes of enquiry developed largely in social anthropology and sociology. They are only "historical" in cases where they have been gathered and elucidated in the past. In a weaker sense oral sources are also part of recent history. Material sources (buildings, ritual objects, bodily positions and movements, etc.) may be rather new, but most commonly they are part of a continuous history and may even be very old. Moreover, material sources are not themselves linguistic in character. (This aspect may be accentuated by referring to them negatively as "non-verbal sources".) Although they take on their meaning from contexts which have a linguistic aspect, which may be at least partly recoverable, it is methodologically important to draw in the appropriate methods of archaeology, numismatics, art history, and so on. It would be desirable to develop this sub-field of methodology further to take particular account of objects used mainly or only in religious contexts. There does not, as yet, seem to be any comprehensive name for this. Traditional terms like "iconography" only refer to a part of it. As to written sources, care is needed to perceive their full range. Written sources include both well-known texts, little known but formally impressive texts, inscriptions of many kinds and from many periods, and ephemeral texts. Such texts may be wholly, partly, or only indirectly religious in intention, a point which also applies to artifacts. The most important point which arises out of an integrated grasp of methodology at this level is as follows. It is evident that the "historical" or the "historical-philological" method is not enough by itself to meet the methodological requirements of the study of religions. It should be realised however, that the "extra" which is required does not arise simply because of the tradition of associating "comparative" studies with the history of religions. The "systematic" requirements of Religionswissenschaft are more far-reaching and strenuous than this. Such a view does not do justice to the requirements arising from the main groups of sources. In particular, it fails to integrate the methods drawn from history on the one hand and social anthropology on the other hand, even though both of these are widely recognised to be of great relevance to the study of religions. When the methods are appropriately clustered and integrated the study of religions is much the stronger.

Methodology and theory formation
This section of the argument will be stated with particular brevity because it is really a different subject and there is no intention here of moving into theory as such. The detailed development of typologies, for example, belongs to the realm of theoretical reflection rather than to methodology as such. It was noted earlier that it is necessary to distinguish between "theory" and "method", for the simple reason that an interest in new theories is often confused with methodological advance. However there are two aspects of method, or strands in the clustered methods which make up the discipline of the study of religions, which contribute in particular to the development of theoretical categories and models in the study of religion. These are, above all, comparison and contextualisation. Where do these methods, or aspects of method, belong in an integrated discipline for the study of religions?
In certain ways, comparison and contextualisation overlap with each other. A comparison may be developed on the basis of two or more cases of religion which are being studied recognitionally. That is to say, several religions or aspects of religions which have been effectively characterised can thereupon be compared. This will lead into the construction of thematic elucidatory categories (such as pilgrimage, tradition, mysticism) or of explanatory categories (such as syncretism) which are internal to religious systems. However, comparison may also be of great interest in the elaboration of explanatory hypotheses which correlate religious data with other social or psychological factors. This was massively exemplified by Max Weber, for example, whose work, while contextual, was also comparative. In summary, comparison is required both in the recognitional phase and in the explanatory phase of the study of religion. Contextualisation means, as may readily be understood, considering one or more of the aspects of a given case of religion in the setting of its historical, sociocultural and even biological context. While this may have an instructive value in the recognitional phase, it becomes much more prominent and is indeed indispensable in the fully explanatory phase. Contextualisation is ambiguous in the recognitional phase of study. It may be necessary for the elucidation of what believers mean. Incorrectly handled, however, it may lead imperceptibly but mercilessly away from the self-understanding of the believers or participants. Explanatory theories, on the other hand, quite correctly, only make sense in context.

Conclusion
The purpose of this argument was to illustrate in brief that methodological integration in the disciplined study of religions can be achieved with relative simplicity. Of course there is a continuous need for clarification and discussion at specific points. Strategically however, what is needed at the present time is not so much discus-sion of the detail as a clear focus on those features of academic or "scientific" method which are necessary and fruitful in the disciplined study of religions. This will make it easier to carry on worthwhile methodological and theoretical discussions with specialists in other fields which themselves require distinctive methodological orientations. If the detail is left aside, some important clarifications have emerged. It has been seen that there is a need for coordination and clustering of the various methods corresponding to the sources which are in fact available for scrutiny. The undogmatic perception of these sources leads in particular to the correlation of fieldwork methods with historical methods, and relativises the latter considerably. It has also been seen that the special character of the discipline does not lie merely in a cross-fertilisation between historical and comparative method. This popular correlation is a mis-match which does justice neither to the appropriate clustering of methods as related to sources, nor to the ways in which comparison is related to both recognitional and explanatory research Finally, it is important to insist that, at the level of greatest generalisation, the procedures for the study of religions, though open to refinement, are not arbitrary or optional. Elucidation and characterisation are not optional. In the academic study of religions they should precede explanation. Moreover elucidation and characterisation must also be "recognitional", as explained above. Neither religious ingression, for example in the form of theological debate, nor premature explanatory reductionism are acceptable in this phase of research. Again, the broad classification of sources is not really optional. The available sources cannot be pushed around on the basis of personal whim or university politics. There really are oral sources and material sources in the field as well as the well-known and less well-known written sources.
Methodological integration is envisaged here. The disciplined study of religion cannot be split down the middle, for example between history and ethnology, just because some people prefer to work with a certain kind of source material or prefer a certain kind of professional badge. It is an unduly easy alibi to say that the study of religion is "interdisciplinary", even if this is helpful in a preliminary way.9 All too often an emphasis on "interdisciplinarity" seems to suggest an openness to a variety of methods while it in fact allows the challenge of methodological reflection to be avoided. By contrast, as has been 9 C.f. Don Wiebe's criticism of "polymethodism" advanced during the 1997 Turku conference. seen above, the discipline of the study of religions both requires and can find its own specific methodological integration.