What is Religion?

At first glance it seems as if a clear definition of‘religion’ is both desirable and necessary for the study of religion. A clear understanding of what religion is is needed if, in advance of historical and sociological enquiry, one is to know what is to be rightly included within the scope of study.

The search for definition is common enough in branches of the human sciences, but it faces enormous difficulties in the case of‘religion’. For the rest of this subsection we shall treat of the problems that arise in giving an initial, operational definition of‘religion’. These difficulties begin with disputes arising from the ordinary usage of ‘religion’ and from the conflicting purposes scholars may bring to the business of defining and studying religion.

The clues to defining ‘religion’ that come from the ordinary use of the word are best considered by separating the denotation and connotation of‘religion’. The denotation of ‘religion’ is given by listing all those systems of belief, or whatever, we would normally agree on calling religions or examples of religion.

Ordinary usage not being clear about the denotation of religion, and giving no sure guide even to the genus into which ‘religion’ fits, there is plenty of scope for scholars with different interests in the study of religions to resort to different operational definitions. The possibility of bringing different purposes to the definition of ‘religion’ reflects the fact of their being different kinds of reflection upon religion and different aims and purposes behind these modes of reflection. Thus a definition of‘religion’ may be influenced by avowedly theological purposes even though it wishes to treat ‘religion’ as a descriptive category. When Max Mueller defines religion as ‘A disposition which enables man to apprehend the Infinite under different names and varying guises’ (Mueller, 1893: 13) he is modifying an element present in the ordinary connotation of ‘religion’ (i.e. ‘religion is belief in God’) to suit a theological purpose behind his study, a purpose which becomes evident when he concludes that all religions place the human soul in the presence of God, despite their surface differences (see Mueller, 1893:192). Equally a definition could be advanced by a sociological theorist of religion to suit the purpose of showing that religion was a wholly social and human construct, as the following definition appears to do:

a religion is (1) a system of symbols which acts to (2) establish powerful, pervasive and long-lasting moods and motivations in men by (3) formulating concepts of a general order of existence (4) and clothing these conceptions with such an aura of factuality that (5) the moods and motivations seem uniquely realistic.

Given that different scholarly interests provide divergent ways of resolving the vagaries of ordinary usage, we must be clear as to the purpose of the study now in hand. Since this purpose is to provide a neutral survey of the broad sweep of mankind’s religious life, it seems to call for a definition of ‘religion’ which does not immediately imply any con­troversial claims about religion’s truth or origin and which preserves as broad a denotation for ‘religion’ as possible. The definition should allow us to make the necessary discrimination between the religious aspect of human life in history and, say, its economic or political aspects, whilst begging as few questions as possible over which particular forms of belief and practice illustrate this aspect. Boundary questions, such as those concerning the relation between religion and magic, can thus be commented on in the course of surveying tribal religions, rather than being settled arbitrarily in advance.

In producing a definition to meet the above purpose, the first requirement is to establish as broad a genus for ‘religion’ as possible. So we shall define ‘religion’ as a type of human institution. It is a complex to be found in human history having the following four important dimensions: the theoretical (e.g. beliefs, myths and doctrines), the practical (e.g. rites, prayers and moral codes), the sociological (e.g. churches, leaders and func­tionaries) and the experiential (e.g. emotions, visions and sentiments of all kinds). A refigion is an institution in human life showing a significant mixture of this complex of theoretical, practical, sociological and experiential dimensions (see Wach’s definition in Kitagama, 1967: 41 and Smart, 1971: 15ff.). It is an institution which may be further distinguished by three types of differentiae which qualify the dimensions making up its genus. These may be specified as the object of this complex, its goal and itsfunction. The object of the complex of dimensions that make up the institution that is religion is given in describing the content of its theoretical dimension and the focus of its practical and experiential ones.

A number of points now need to be made about this definition. We may see first that its very broadness provides a way of understanding the great variety of more specific definitions which have competed with one another in scholarly literature upon religion. For each of these rivals will usually be found to select only some of the dimensions of religion to provide its genus and to fasten upon some limited differentiae to fix the species to which religion belongs. Thus E.B. Tylor’s famous definition of ‘religion’ as ‘belief in spiritual beings’ (Tylor, 1903: 424) selects the theoretical element as the most important aspect of religion’s genus and differentiates it by a particular kind of belief. Religion becomes a kind of philosophy, which Tylor distinguishes from other types by the label ‘Anim­ism’ (a general view that the world is moved and controlled by superhuman and spiritual beings) and its other dimensions are then but secondary. By contrast the following definition from a recent author initially selects two of our dimensions to provide the genus and picks out the particular function of these dimensions to fix the species:

Religion, then, can be defined as a system of beliefs and practices by means of which a group of people struggles with these ultimate problems of human life.

Such a difference in selection by Tylor and Yinger leads to a difference in fixing religion’s denotation: Tylor’s definition will make us hesitate about including the ‘atheistic’ religions in the denotation, whilst Yinger’s will see the downgrading of belief in personal gods as unimportant in the light of the common function the beliefs and practices of theistic and atheistic religions share. These different definitions with their attendant consequences can now be seen to be selections from the dimensions which give religion its genus and from the various differentiae of these dimensions. Such selection can be motivated by a variety of reasons: the impressiveness of a certain aspect of the ordinary use of religion (or of a chosen example of religion) or the bringing of a specific purpose to the business of definition. No such reasons He behind our definition so it eschews such selection and remains as inclusive as possible.

The inclusiveness of our definition shows that it is in good measure vague. We allow in the first instance that a religion need not clearly show all four dimensions in developed form. We can thus accept that at the penumbra of religion there may be primitive, ancient religions which, as Robertson Smith suggested, display the theoretical dimension in but a weak fashion. We can equally accept the Society of Friends as practising a rehgion in which the sociological dimension has atrophied to a large extent and in which the ritual aspects of the practical dimension have all but disappeared. We look only for a ‘significant mix’ of these four dimensions. We have also allowed for a number of diverse differentiae so that we produce yet further species of our genus. Moreover many of these differentiae are themselves vague. Thus one of the most important differentia of religion relates to its object and is given by the great French sociologist Durkheim thus: ‘A religion is a unified system of beliefs and practices relative to sacred things..(Durkheim, 1976: 47). The advantage of this way of fixing the object of religion is that the object becomes broad enough to include religious realities (such as nirvana) which are none the less not divine. But how is the sacred to be defined itself? We may contrast sacred objects and realities with profane on the grounds the former are thought of as transcendent, extraordi­nary and beyond anything in normal experience, but a full understanding of the nature and range of the sacred can only come with an actual study of the complexity of beliefs about the sacred that can be found in religious history. It is impossible to give an adequate idea of this complexity in some simple formula (cf. Eliade, 1958: xii).

Vague though our definition is, it will still serve to pick out the religious aspect of human life in history from other aspects of that life. It is useful, for example, to be able to distinguish refigions in history from systems of belief we say are ‘merely philosophies’ even though they may have an object, goal and at least personal function which parallel those of religion proper. The demand that the theoretical dimension to religion be significantly linked with at least some of the other three enables us to do this. We might, on the other hand, come across institutions in human life which display the four dimensions of theoretical, practical, sociological and experiential but from which are absent any of the differentiae which mark out religion in the genus. In such institutions there would be no concern with the divine or the sacred, no attempt to rise beyond utilitarian goals and aims and no desire to function as the ultimate source of meaning or identity in life. Political institutions may thus show the dimensions of religion but lack any significant combination of the differentiae of central examples of religion. Some indeed regard the combination of science and technology in the mod­ern world as a significant instance of the displacement of religion by some­thing parallel but different to it. But though our definition has some obvious discriminating power, it will not, because it is vague, sort out the nicer boundary disputes which occur, say, in distinguishing religion from magic. For in some magical systems we may have elements of the four dimensions and at least one of the important differentiae of religion in the presence of belief in spiritual, superhuman beings. Yet other differentiae will be lacking. The goal and function of a system of magic may be entirely mundane and utilitarian, being concerned solely with the attainment of worldly and limited ends. Whether such a system is likened in the end to a crude kind of religion or a crude kind of science/technology (as Sir James Frazer famously does in the opening and closing chapters of The Golden Bough) will be left to a detailed study of its relation to central examples of religion.

We may summarise the above points by saying that our definition of religion serves to pick out religion as a recognisable aspect of human life in history but does so in an open-ended way. It serves the purposes of a broad enquiry into religion by giving it a general direction whilst not settling in advance important questions which can only be properly decided upon after detailed study has finished. This kind of operational definition of ‘religion’ works by listing the more important features which contribute to something’s being an instance of religion, but refuses to say exactly what combination of such features is essential for something’s being a religion and refuses to define in advance what each feature precisely means (cf. Alston, 1967: 142-3). To some this will seem like the abdication of the attempt to define ‘religion’ at all, since the operational definition given does not state in unambiguous terms the necessary and sufficient conditions for something’s being a religion. A statement of religion’s necessary and sufficient conditions would be a list of precise features such that, if an institution lacked them, it could not be classed as a religion, and if it did possess them, it could not fail to be a religion. All that our list of dimensions and differentiae of religion does is allow a number of different statements of sufficient conditions for some­thing’s being a religion to be produced, since we know that the possession of a significant number of these features in varying combinations will be enough to enable something to be a religion. But we neither close the list of possible combinations of religion-making features, nor do we attempt to define each feature precisely. But how may we suppose that ‘refigion’ has a meaning we can understand, if we cannot thus state necessary and sufficient conditions for the use of the term?

One answer to the above question would be this: ‘religion’ is a family-resemblance term and religions form a family. Ludwig Wittgenstein explained the family-resemblance idea of meaning in the fol­lowing way (he is discussing the words ‘language’ and ‘game’):

Instead of producing something common to all that we call language, I am saying that these phenomena have no one thing in common which makes us use the same word for all—but that they are all related to one another in many different ways. And it is because of this relationship, or these relationships, that we call them all ‘language’... Consider for example the proceedings that we call ‘games’... What is common to them all?—Don’t say ‘There must be something common, or they would not be called ‘games’ ’—but look and see whether there is anything common to them all.—For if you look at them you will not see something that is common to all, but similarities, relationships, and a whole series of them at that... we see a complicated network of similarities overlapping and criss-crossing: sometimes overall similarities, sometimes similarities of detail. I can think of no better expres­sion to characterise these similarities than ‘family resemblances’; for the various resemblances between members of a family: build, features, colour of eyes, gait, temperament, etc. etc. overlap and criss-cross in the same way.—And I shall say: ‘games’ form a family. (Wittgenstein, 1958: 31-2)

If the family-resemblance idea of meaning applies to ‘ religion’ then we should expect to find the following facts in our understanding of the word. 1. There will be a characteristic set of features found in examples of religion (such as those described above). 2. There will be no single set of such features to be found in each and every example of religion. 3. There will be no limits to set in advance to the kind of combinations of the characteristic features newly discovered or developing religions could be found to exemplify, nor will there be absolute limits to the additional features such new examples could add to the set. 4. The various examples of religion will be related to one another not therefore by them all sharing a single set of features but through a series of significant overlaps in the features they possess; there will be a network of relationships but no one set of things they all share. 5. The meaning of the word ‘religion’ will none the less be projectible, that is, having rehearsed the characteristic features of religion in an inclusive definition like that given above or having gained a knowledge of some central examples one will be able to say of newly-found examples whether they are religions or not.

The merit of accepting, at least initially, a family­resemblance meaning for ‘religion’ is that it agrees with our intuition that religion is an open-ended and ever-developing aspect of life. To think there social existence to be drawn in advance is to falsify our perception that religion will enmesh and intermingle with what surrounds it. Since, too, religion is part of human history it will be ever-changing and therefore it would be wrong not to be open to the possibility that new forms of religion await to be discovered, forms which indeed could extend our apprehension of what religion can be and amount to. The analogy with a family, an ever-developing unity of diverse yet organically-related elements, is more than apt to bring out these facts about religion. We should also be ready to extend the family-resemblance model to the prior picture we have of the particular religions.

This last point is worth stressing. Many believe that to be able to identify different religions and baptise them with names such as ‘Hinduism’ and ‘Christianity’ presupposes that there is some easily definable common element that distinguishes them from one another. But if we can legitimately distinguish a host of Indian forms of belief and worship by the term ‘Hinduism’, this is not because there is some essential unity among them that can be seen on the surface. We have, in fact, coined an umbrella term which refers to a mass of phenomena which does not aspire to be a unity and which is not united by any single item of belief, mode of worship etc. (see Smith, 1978: 65-6). The use of the single term to cover all these phenomena rests only on there being a significant family likeness between them of the sort Wittgenstein has described for us. Nor is this lack of unity simply a feature of the religious traditions of India. In Christianity there is evidently a greater desire to present the religion as a unity and its various sects will accordingly tend to present themselves as representatives of ‘the Church of Christ’. But even in this instance, the actual forms of religion to which we give the label ‘Christianity’ have shown, and continue to show, the most astonishing variety. Consider in this light the difference between the Christ­ian Unitarianism to be found in parts of Victorian England and the Catholic Christianity of the peoples of Latin America. Even an alleged common feature in all, such as ‘the preaching of Christ’ (or. the Gospel’ or *... the New Testament’) will be found to take different shapes in different examples of the religion and to bring with it no essential likeness. Again we will have to fall back upon a family-resemblance picture.

The reason for the variability within specific religions is the same as that for the variability of the general phenomenon, ‘religion’. To speak of‘Hinduism’ or ‘Christianity’ is to refer to aspects ofthe belief, practice, organisation and experience of human beings. These aspects of human life show the variability that human beings show. We might wish as students of human life that human beings believed and behaved more uniformly but they do not. There is limitless scope for novelty and for receptivity to the differing influences of time and place amongst human beings. It is the humanity of religion and the religions which commands that we begin our study with an open-ended definition of ‘religion’.