It is common for scholars sympathetic to religion – from Friedrich Schleiermacher to the present – to describe the modern critics of religious experience as reductionist. By this they mean these critics do not respect religion for what it is; they rather ‘attempt to assimilate religion to nonreligious phenomena’,1 say, primitive science or metaphysics. Needless to say, these scholars consider this incorrect.

Into this scene strides Wayne Proudfoot. In his award-winning book Religious Experience2 he sets out to scrutinize this accusation and rehabilitate the modern critics. He dismisses this accusation as a protective strategy to render religious claims immune to rational critique. His rehabilitation program is based on a distinction he draws between two kinds of reduction – descriptive and explanatory. This distinction has so caught the imagination of philosophers that it has found its way into textbooks.3 On the other hand, the book is also controversial; needless to say, its claims are problematic for believers.

The purpose of my paper is not to offer a comprehensive critique of his book, but only of his methodology, or the lack thereof, where methodology is understood as an integrated set of tools used for inquiry. I argue that although he proposes description and explanation as appropriate instruments for the study of religious experience, the two do not constitute a methodology, because his descriptions play no role in his explanations; rather, both go their separate ways. I argue that explanation must be guided by description. Moreover, I distinguish three kinds of inquiry, each calling for a different kind of description. The kind of description Proudfoot recommends is appropriate for genetic inquiry into how particular experiences came about, but not for a conceptual inquiry into the constitutive elements of religious experience. He knows lots of trees (particular experiences), but misses the wood (the nature of religious experience). Thus my title.

The paper is divided into three parts. First is a brief presentation of Proudfoot's thesis. A second part analyses three different kinds of inquiry; it shows that genetic and epistemic explanations are logically independent and irreducible one to another. These different types of inquiry call for distinct types of description, and these descriptions in turn guide the explanation that will be considered satisfactory. The context-sensitivity of descriptions, and their relation to explanation is dealt with in the final part. Let us start with Proudfoot's contention.


Proudfoot says the charge of reductionism levelled against modern critics is based on an improper conclusion drawn from an important insight in the modern hermeneutic tradition. This insight is that an experience (or an emotion, intention, action, or belief) must be described from the point of view of the person who has the experience.4

To understand a ritual practice I must grasp the concepts that are presupposed and the set of cultural conventions in which it is embedded. Anger can be specified only by reference to the subject's beliefs about the object of his anger, and fear only by citing the relevant beliefs about the present danger.5

In describing an experience, therefore, the perspective of the subject has the last word; the subject must be able to recognize a given description as faithful to his or her experience. A description that cannot be so recognised would be an unsatisfactory description.

To describe the experience of a mystic by reference only to alpha waves, altered heart rate, and changes in bodily temperature is to misdescribe it. To characterize the experience of a Hindu mystic in terms drawn from the Christian tradition is to misidentify it.6

He takes Charles Peirce to task for suggesting that there is really no difference in the meaning of the Lord's Supper between Catholics who believe in the doctrine of transubstantiation and Protestants who do not.7 Description of an experience, emotion or belief, then, must be done from an insider's perspective.

Having described an experience to the satisfaction of the subject, the reporter withdraws and the scientist then takes over. The task of the scientist is to explain, and in this task he or she is not bound by the constraints of the reporter. Explanation may be done in terms that do not receive the approval of the subject. This sort of explanatory reduction is ‘perfectly justifiable’.8 For example, historians may offer explanations of past events by employing concepts like socialization, ideology, means of production, feudal economy and the like; these are concepts which do not (in most cases) characterize the perspective of the subjects, but this poses no problem for explanation. In other words, the subject's identifying description need not be considered normative for the purpose of explanation. Proudfoot claims that failure to distinguish these two kinds of reduction, one legitimate and the other illegitimate, is to engage in a ‘protective strategy’ that precludes critical inquiry into religion.9‘To require that any explanation of a religious experience be one that would be endorsed by the subject is to block inquiry into the character of this experience’.10

Proudfoot's epistemological concerns are legitimate: the critical function of epistemology should not be sacrificed in the name of hermeneutics and language games, where the insider's perspective is always accorded the last word. At the same time, he does not wish to sacrifice the insights of hermeneutics. At first sight it would seem that he has managed to do justice to both the participant view of the insider as well as the critical view of the outsider by marking this distinction between descriptive and explanatory reduction. When we look more closely at his work, however, we find that there is no coherent methodology operative, despite his employing his considerable analytical acumen ‘to distinguish between descriptive, analytical, explanatory, and evocative elements in the accounts of religious experience’.11 On the face of it, description and explanation seem to be two moments together forming his methodology, with the expectation that they will be closely aligned one with the other; however, no such connection is maintained. His descriptions and explanations are uncoupled, conceptually disconnected, and logically unrelated to one another. I attribute this to Proudfoot's failure to distinguish different kinds of inquiry, and to specify the type of description (and explanation) appropriate to each.


Although the concept of explanation is central to Proudfoot, nowhere does he explain what he means by it. The fact that he talks about ‘competing explanations’ and ‘inference to the best explanation’ in the singular, however, suggests that he thinks of explanations as all one kind of thing; one explanation may be better or worse than another, but to engage in explanation is to engage in a single kind of activity. This is not, however, the case. There are at least three different types of activity that go under the label of explanation, all of which are relevant to Proudfoot's discussion of religious experience.12 I shall call them epistemic, genetic and hermeneutic explanations.

Let us consider an ordinary situation. A friend invites a boy to a party. There the boy sees a girl, their eyes meet, hearts flutter, and there is a quickening of the heartbeat. He sees signals from her side too and comes to the conclusion that they are in love with each other. When he shares it with a friend of a more sceptical bent of mind the friend asks him to explain his claim. In response the boy points to evidence that he takes as signs of the girl's interest in him. The friend asks again, ‘Are you sure it really is love and not just infatuation?’ Another time when the question of love between them is no longer in doubt, someone asks him: ‘Tell me, how did all this begin?’, and he explains that it happened because of an invitation to a party from a friend.

This is an example that uses all three types of explanations. It involves an experience, an experience we call ‘falling in love’; this event of falling in love finds an expression in the statement, ‘We are in love’. This expression could be true or false.13 When the sceptical friend asks for an explanation of the assertion, he is demanding evidence to substantiate the claim. Various factors could be pointed to in favour of (or against) its truth; this is epistemic explanation. Another term for it is epistemic justification. When he is further asked, however, if it really is love, and not just infatuation, this calls for another type of response. What is at stake here is something more fundamental than evidence: it calls for a conceptual clarification; I call this hermeneutic explanation. In the third place when he is asked about how it all began, this calls for yet another kind of response; it calls for a description of the conditions that led to the event of their falling in love. In this sense, the invitation to the party given by the friend serves an explanation; this is genetic explanation.

(i) Genetic and Epistemic Explanations

Let us make these distinctions a little sharper by focusing first on the contrast between epistemic and genetic explanations. There are at least three important differences between them. First, in epistemic explanation the object of justification is a proposition or a belief or an hypothesis (e.g., ‘We are in love’, ‘I see a tree’, ‘The bridge collapsed due to the floods’). It is key to note that it is not the event or the happening that is being explained, but a belief. The belief that is the object of epistemic justification may or may not arise from the experience; the important point is that the object of epistemic explanation is a belief and not an event. By contrast, in genetic explanation what is being explained is the event or the happening and not an expression or belief. Secondly, the purpose of epistemic explanation is to determine the truth or falsity of the belief concerned; in contrast, the purpose of genetic explanation is to discover how a particular event came about, its genesis. A third difference concerns the answer required in each case. Epistemic explanation calls for evidence or warrant for the belief concerned, whereas the genetic explanation calls for details regarding the background material or the conditions leading up to the occurrence of that event. Its concern is not with the truth of propositions, but with genesis of an event.

Let us pay more attention to the second point. Numerous examples can be given to show that genetic conditions of beliefs have no direct bearing on their truth. Here is one of the more famous from the history of science. The circumstances leading to Archimedes's discovery of the principle of buoyancy is legendary. The king asked him to determine if his new crown was made of solid gold or whether the goldsmith had cheated him by mixing silver with the gold. The scientist was at loss as to how to solve the problem because it had to be done without damaging the crown. Burdened with this commission he goes for his bath, and the moment he notices the water overflowing while getting into the bath tub, the answer strikes him in a flash. This is the genesis of the discovery of the principle of buoyancy. To explain the event, the circumstances leading to it (the order of the king, the bath etc.) are important; but it is also clear that as far as the truth of the principle is concerned, these genetic circumstances are irrelevant.

Or consider the following landmark case14 in the history of medical science: Ignaz Semmelweis, working as a medical doctor in Vienna General Hospital noticed that a large number of women who delivered their babies in one of the Maternity Divisions of the hospital died of ‘childbed fever’ (Puerperal Fever). A number of factors about the case puzzled him, including the fact that the death rate was far higher in the First Maternity Division where medical students worked than in the Second Division where ordinary midwives took care of the women. Different possible explanations and solutions were tried out and nothing seemed to work. It is then that a colleague of Semmelweis began to develop symptoms similar to those of the women suffering from childbed fever and in a few days he died. The major difference was that while the women developed the symptoms after childbirth, his colleague developed the symptoms after getting a small slash while performing an autopsy. This leads him to suspect that the death of his colleague was caused by blood poisoning or the introduction of ‘cadaveric matter’ into the blood stream while performing the autopsy. This prompts him to make a brilliant guess that the cause of childbed fever was the same. Since the medical students who attended to the women in the First Division, unlike the midwives in the Second Division, often came to their maternity duty after performing autopsy on dead bodies without cleaning their hands properly, they were the carriers of infection. He tests out this hypothesis by instructing the medical students to properly disinfect their hands prior to their examination of the women and it produced dramatic results. Thus his hypothesis was confirmed.

What is important for us is to note the fact that the discovery or genesis of the final solution is the accidental death of a doctor in the hospital; it gives us a genetic explanation of how Semmelweis comes to suspect that ‘childbed fever’ is caused by infection carried by the medical students. However the accident is not taken as ‘proof’ or evidence for this hypothesis. The hypothesis is confirmed only on the basis of the positive outcome of the follow up of that insight. Genetic explanation is therefore clearly different from epistemic explanation. The two being so different, one cannot be said to constitute a rival or alternative to the other.

The relevance of this distinction is that all the examples Proudfoot gives to illustrate the legitimacy of explanatory reduction are genetic explanations. After recounting the experiences of Stephen Bradley and Mrs. Jonathan Edwards as narrated by William James, Proudfoot says,

Bradley's experience might be explained in terms of the conflicts of early adolescence and that of Sarah Edwards as a consequence of her life with Mr. Edwards. No reference need to be made to God or to Christ in the construction of these explanations.15

Or again: ‘For Bradley, we would need to know something about Methodist revivalism in early nineteenth-century New England, about the particular meeting he attended earlier in the evening, and about the events in his life up to that moment.’16 If we follow the same logic, Semmelweis's discovery is explained by the accidental death of his colleague but it would tell us nothing about the truth of that insight. These are genetic points that would tell us why an event took place, but not about the truth of the belief occasioned by that event.

(ii) Exception?

It is not that Proudfoot is unaware of the difference between genetic and epistemic explanations. He agrees that ‘beliefs can be ordinarily assessed without regard to the origin or cause of the experience that gave rise to those beliefs.’ But he thinks that ‘an exception must be made in the case of beliefs that include claims about the cause of an experience’.17 Perceptual beliefs, beliefs that identify an emotional state (like love, anger or fear), and beliefs arising from religious experiences, he believes are fit candidates for this exceptional treatment.

He defends this point with examples from perception. One example is a case of seeing a tree about thirty yards away. After receiving further information I realize that what I saw is only a reflection of a tree in a large mirror. Another example is becoming frightened upon seeing a bear ahead on the trail. Sensing my fear, a friend points out that it is not a bear but only a log, and my fear subsequently subsides.18 There are several points to be noted here. First, a conceptual inquiry into perception reveals that a causal connection with an experienced object is constitutive of perception.19 In the two examples, the description involves a reference to my seeing a tree and a bear respectively. But once I realize that there is no causal link between the event of my seeming to see a tree (or bear) and such an object, I am no longer justified in maintaining such a belief. The reason is that the concept of perception carries with it this causal link. This is the conceptual point.

A second epistemic point is about the role of causation in justification. The presumed causal link between the perceptual experience and its object bestows a noetic quality to the experience, such that the subject considers the experience to be authoritative and true. It is also clear, however, that the causal link (in any particular experience) is more of a presumption than the result of empirical inquiry. Hume taught us long ago that a causal link cannot be empirically established. The role of the presumed causal link in justification is therefore more negative than positive: those instances where the causal antecedents are suspect on independent grounds should be held suspect as instances of perception. But until I have reason to believe otherwise, i.e., until I discover that what I thought I saw was really only a reflection in the mirror, my experience of ‘seeing a tree’ provides justification for my belief that I am seeing a tree. For this reason this kind of justification is held to be prima facie: it is provisional or fallible.20

One could now ask a further question: how do I discover that what I took to be an instance of seeing a tree was actually not so, but only seeing the image of a tree? Obviously, we do it by relying on another perception, or the report of another perception. Justification of perceptual beliefs, therefore, involves an epistemic circularity.21 It is by relying on one perception that we correct or justify another perception. There is another question that needs to be probed immediately: what is it about the nature of perception that enables me to conclude from my information that I was seeing an image in the mirror (or a log) to saying I really did not see a tree (or a bear), but ONLY an image (or a log)? What enables me to do this is another constitutive element of perception: the objects of ordinary human perception are space-time entities, such that both a physical bear and a physical tree can not occupy the same space at the same time; in other words, the two possible causes of my experience (bear and a log) are mutually exclusive. If one is shown to be the case, the other must be excluded. It is this conceptual truth about perception that enables us to ignore or overlook the distinction between genetic and epistemic explanation, such that if the genesis is attributed to one cause, the other is automatically excluded.

If religious experiences are to be given the same sort of exception, it will have to be shown that similar space-time considerations also hold for them. In the case of Bradley's experience, for example, Proudfoot would have to show that the action of God or Christ on the one hand, and the effects of adolescent conflicts or the Methodist revivalism on the other are mutually exclusive, as with the experience of a bear or a log. Until this is done, even if we have admitted an exception in the latter perceptual case, there is no reason to accord the same treatment to religious experiences. To be fair to Proudfoot, he does try to show they are exclusive by exploring the concept of miracle as treated by Hume and applying it to religious causation. But this works only if (1) it is the case that Hume's understanding of miracles as excluding natural laws is the way religious believers understand them and (2) all religious experiences constitute such miraculous happenings. Neither of these, however, should be considered true. Even if we let Hume's definition of miracle go unchallenged, there is still no basis for identifying religious experience exclusively with the miraculous (i.e., events that exclude natural causes). This applies all the more to the experiences of Bradley and Edwards because both were Christians who believed in a God who, though not subject to natural laws, is immanent in nature and therefore works in and through nature. It is clear then that religious and natural causes are not mutually exclusive in the manner in which the log and the bear exclude one another as the cause of a perceptual experience. Even if we agree with Proudfoot that the distinction between genetic and epistemic explanation does not hold in matters of perceptual experience, therefore, there is no reason to extend this exception to religious experiences.

(iii) Concepts and Justification

What is clear in the above considerations is that our understanding of perception plays an important role in the justification of perceptual beliefs. Two characteristic features of the concept of perception came to prominence: (1) perception includes the belief that there is a causal link between the event of perception and the object perceived as its object; (2) such causal objects are mutually exclusive, such that if the cause is found to be a log it cannot be a bear. Both these beliefs (along with many others) constitute our understanding of the nature of perception, and such constitutive features as a class are taken for granted when we are engaged in justifying individual perceptual beliefs. It is these taken-for-granted beliefs that form the basis for justifying individual perceptual beliefs. When an individual perceptual belief is held not to be warranted, these taken-for-granted beliefs about perception are not abandoned.

This leads us to the Wittgensteinian insight that all epistemic explanation is done within a system of beliefs. ‘All testing, all confirmation and disconfirmation of an hypothesis takes place already within a system’, says Wittgenstein.22 Norman Malcolm makes it more explicit: ‘Within a language-game there is justification and lack of justification, evidence and proof, mistakes and groundless opinions, good and bad reasoning, correct measurements and incorrect ones.’23 Justification of individual perceptual beliefs is done within a system of inter-related of beliefs we already take for granted about perception as a class. So too with religious experiences: only against the background of a system of inter-related beliefs about religion can we justify any belief arising from a religious experience. It is therefore imperative that epistemic inquiry into the veracity of individual perceptual or religious experiences be preceded by a conceptual inquiry into the nature of perception or religion; understanding must precede justification. Conceptual inquiry must be considered a third kind of inquiry distinct from both genetic and epistemic.

(iv) Conceptual Inquiry

Genetic and epistemic inquiries are common; they are routinely undertaken by specialists and laypersons alike. This is not so for conceptual inquiry. Concepts are more routinely used than made the subject of explicit inquiry. Every time we identify an experience as an instance of ‘perception’, ‘love’ and so on – in fact, every time we use language – we are using concepts. Sometimes rather than use concepts, however, we subject the concepts themselves to scrutiny. In the earlier example of falling in love, we saw that besides the epistemic and genetic inquiries, there was also the question about the appropriateness of the use of the concept of ‘love’. Similarly, when Proudfoot asks the question, ‘What are the conditions under which we identify an experience as perception’24 he is engaging in a conceptual inquiry about perception. So is his discussion concerning causal connection, as well as our observation that causal objects of perception are spatio-temporal and hence mutually exclusive, etc.; these are claims made about the concept of perception. Our understanding of concepts is made up of a number of such implicit beliefs, which may be correct or not, or simply confused. Looking into such matters and making such beliefs explicit is the task of a conceptual inquiry, and it must be carried out in a coherent fashion. A conceptual inquiry into perception, for example, brings our implicit beliefs about perception into a coherent whole, such that they form a system of beliefs within which epistemic explanation of a particular case of perception can take place. It provides a considered view of what constitutes perception, such that this can then be applied not only to this experience or that, but to a whole range of experiences that are candidates to be called ‘perceptual experience’. The same should be said about other concepts like ‘religion’, ‘virtue’, ‘knowledge’, ‘love’, etc.

A conceptual inquiry is, in some sense, more fundamental than both genetic and epistemic inquiries. The subject matter of a genetic inquiry is the genesis of an event or happening. It begins by bringing the event concerned under a certain concept, and then goes on to examine the circumstances leading to the event. A genetic inquiry is therefore an empirical affair. A conceptual inquiry, however, is not an empirical matter. A conceptual inquiry is also related to epistemic inquiry, as we have seen; if justification requires evidence for determining the truth of a belief, it is an understanding of the concepts concerned that tell us what counts or does not count as evidence. It is within an already taken-for-granted system of beliefs that all justification is carried out, as Wittgenstein pointed out.

Having found that justification is done within an already taken-for-granted system of beliefs, Wittgensteinians (and sometimes Wittgenstein himself) seem to say that this system itself is beyond justification.25 This is a questionable claim and obviously not one that would be endorsed by Proudfoot because the whole point of his argument is that interpretation and language games must not be used for making religious claims immune to critical inquiry. But in our eagerness to subject language games to scrutiny, we must not forget the Wittgensteinian insight that a conceptual inquiry is a different type of activity than an empirical one; the standards we apply for justifying empirical beliefs are not to be applied to the standard itself.

If conceptual inquiry is a different type of activity, we would expect that it would require a different type of explanation than either genetic or epistemic explanations. I will call it hermeneutic explanation. One may raise a terminological objection at this point, however: Granted that there are different types of inquiry, is it appropriate to use the term ‘explanation’ for all three different kinds of activity?26 The term is most often used in empirical inquiry dealing with the occurrence of an event or happening, and the term ‘justification’ is used in the context of inquiring into the truth of propositions. Sometimes the two overlap, however.27 Moreover, it is clear from Proudfoot's examples and arguments that when he is looking for the explanation, he is looking for a justification. It is thus clear that we can talk about genetic and epistemic explanations. The term ‘hermeneutic explanation’ can be problematic, however; it might even sound like an oxymoron, when we consider that Wilhem Dilthey conceived hermeneutics as the method that is appropriate to the human sciences, in opposition to the type he saw appropriate to the natural sciences. More recent philosophers (e.g., Karl Otto-Apel), however, have questioned such a sharp dichotomy between hermeneutics and explanation; moreover, we saw that Proudfoot's own attempt is to bring the two together. Since as we have seen his manner of accommodating both is problematic, however, I suggest that hermeneutics itself be considered a kind of explanation. To argue that this is indeed a type of explanation we will have to examine the relationship between description and explanation. This will be done in the next part, and we will then be able to see that hermeneutics is indeed a kind of explanation. The point I have been trying to drive home for the moment is a simple one: not all inquiry is of one type, and not all explanations are competitors for the same ‘throne’, so to speak, a throne that goes by the name of the ‘best explanation’. Thus, if there are different types of inquiry, what will count as the best explanation or not will depend on the kind of inquiry. This is the next point. To establish this latter point we must begin by looking at the different types of description each kind of inquiry calls for and presupposes.


It is important is to note that the subject matter of a conceptual inquiry is not a particular instance of it. ‘Perception’ is not the same as the experience of perceiving a tree or a bear or a reflection in a mirror, and yet would include them all. ‘Love’ is not the same as what existed between Romeo and Juliet or between Ram and Laxman, or between the prodigal son and his father, although it would include them all. A genetic inquiry, in contrast, is about the genesis of a particular event or happening. The event could be as varied as a particular experience of perception (of a tree, a bear, an image in a mirror, etc.), a particular instance getting angry, as when Ram found that his wife was abducted, a particular experience of falling in love, as when Romeo met Juliet, and so on. The event could even be the origin of an idea or a hypothesis, as when Archimedes discovered the buoyancy principle, or when the heliocentric idea occurred to Galileo. But in each case, the subject matter of a genetic inquiry is a particular event; not so the subject matter of a conceptual inquiry.

(i) Different Descriptions

When the subject matter of an inquiry is different, one would expect the descriptions also to be different. The descriptions relevant to a conceptual inquiry would be the characteristic marks of the concept being inquired into, rather than what is characteristic of a particular event. Familiarity with descriptions of particular events is needed for finding the characteristics that are common to the whole class; but those descriptions of particular events do not constitute the descriptions appropriate to a conceptual inquiry. The presumed causal connection between a perceptual experience and the object perceived, for example, is one characteristic mark of perception as such, and not merely of this perception or that. This is not the only one, however; there are others. I have pointed to the spatio-temporal character of perceptual objects as another characteristic feature. Here are other examples: It is a common experience of human perceivers that perception requires our attention, such that even something presented to us may go unperceived for lack of attention. On the other hand, it is not the case that we see only that to which we pay attention; sometimes we notice something even if we are not paying attention (e.g. hearing a loud noise, for example). These are phenomena regarding perceptual experience that can be described in some detail, but none would be descriptions of individual events.28 The point is to show that a conceptual inquiry calls for a different type of description than a genetic inquiry. Our descriptions, in short, are context-sensitive: depending on the kind of inquiry, the descriptions vary. Genetic inquiry calls for one kind of description; conceptual inquiry for another. When we look at Proudfoot's descriptions, however, we notice that all of them – whether from perception or from religion – are instances of particular events or happenings. Such descriptions are appropriate for a genetic inquiry as to how that particular event has come about, but they are hardly appropriate for a conceptual inquiry. It is a case of missing the wood for the trees.

The same can be said of religious experiences. A conceptual inquiry into religious experience calls for a description of the characteristic marks of religious experiences as such, and not a description of the individual events or experiences. Proudfoot does consider some of the more commonly accepted characteristics, such as ineffability and noetic quality; however, he takes them up one by one and discards them. For example, he rules out ineffability as a characteristic mark of religious experience because, according to him, it is prescriptive of experience rather than descriptive of it;29 it is said to constitute an experience rather than to describe it.30 If this is reason enough to rule out ineffability as a characteristic mark of mystical experiences (as he does), the same can be said about the causal connection that he discusses in the context of perception. We noted that the causal link is a constitutive element of perceptual experiences, such that where this connection is seen to be absent, the concept of perception is not applicable. Proudfoot, of course, would not consider the causal link as a description. This raises two important questions: (1) What qualifies as a description? (2) How are descriptions and prescriptions related to one another?

(ii) Descriptions and Prescriptions

What, then, may count as a description? A description, Proudfoot has told us, must be acceptable to the insider, the subject of the experience. This is correct. The only difference is that the insider in a genetic inquiry is an individual, whereas the insider in a conceptual inquiry is any proficient user of the concept. This change in the meaning of insider devolves from our recognition of the context-sensitivity of descriptions. In a genetic inquiry it is the occurrence of a particular event that needs to be explained, and when this event is an experience, the insider is the individual who has this experience. In the case of a conceptual inquiry, however, the insider is not merely an individual but any proficient user of that concept. All those who use correctly the concept of perception, for example, understand (or can be brought to understand in the course of the description) that any of the phenomena described above (only as examples) are characteristic marks of perceptual experiences. Where such agreement is lacking, it would be improper to talk of a phenomenological description. What counts as a description, therefore, depends on the consent of the insider. We thus concur with Proudfoot that the main contribution from the hermeneutic tradition is that descriptions be done from an insider's perspective; otherwise it would constitute a descriptive reduction. The difference, however, is that for him the insider can only be an individual, because he does not distinguish adequately between a genetic inquiry and a conceptual inquiry.

Once it is conceded that what qualifies anything to be a description is its acceptability to an insider, ineffability cannot be ruled out as a description, as Proudfoot does, especially in light of the fact that it is one of the characteristics on which there is almost universal agreement among users of the concept of mysticism. Besides ineffability, which is commonly recognized as a characteristic mark of mystical experiences, there are other characteristics of religious experiences. One of the most famous descriptions is the one given by Rudolf Otto, according to which the ‘object’ of religious experience is the ‘wholly Other’, in the sense that it is unlike anything experienced in our ordinary state of consciousness.31 Since my purpose is not to provide a substantive theory of religious experience, there is no need to go into further descriptions.

To consider the second question, how are descriptions and prescriptions related to one another? The answer can be discovered in Proudfoot's own practice. On being told by a friend that what I see is only a reflection of a tree in a mirror and not a real tree, my experience is disqualified as a case of perception. What enables Proudfoot to say this is the fact of the causal connection, and the exclusive nature of such a causal connection, as we have noted. However, these are the precise characters that enter into a description in a conceptual inquiry regarding perception; therefore, those characters that are found to be constitutive of a concept seem to become prescriptive for an epistemic inquiry. This only spells out the Wittgensteinian insight we noted earlier concerning the distinction between justification, which is always done within a system of beliefs, and the system itself. Ineffability, in other words, is a description when the task is a conceptual inquiry regarding mysticism; but it becomes prescriptive for an epistemic inquiry.

(iii) Hermeneutic Explanation and Reduction

Once the description is done, explanation can begin, as Proudfoot tells us. But there is an important difference: what is being explained in a conceptual inquiry is not a particular experience; that would be appropriate for a genetic inquiry. What is being explained here is religious experience as such, or perceptual experiences in as much as they are perceptual. Explanation in the context of a conceptual inquiry means giving a coherent account of the various phenomena described. Since no reduction is permissible, in the sense that they must be acceptable to an insider, it is typically the case that the descriptions are not a coherent lot. At least their cogency may not be apparent; therein lies the need for explanation. Take for example two of the possible descriptions about perception to which I have alluded: the need for attention on the one hand, and also the fact that sometimes we perceive something even when we are not paying conscious attention. Or consider two of the common descriptions found in religious experiences: On the one hand the object of experience is held to be ‘wholly Other’ and on the other, there are descriptions that identify Brahman and Atman, God and the soul. Clearly, it is not an easy task to uncover a coherence of these descriptions; yet the task of a hermeneutic explanation is to hold such apparently contradictory descriptions together.

For this precise reason, it may become necessary to perform some reduction in the process of hermeneutic explanation. This brings us to Proudfoot's point about explanatory reduction. I am suggesting that there is a place for explanatory reduction in conceptual inquiry and hermeneutic explanation; but such reduction must come only after description, and is done only to provide a coherent account of the descriptions. The result of such reduction would be significantly different, however, from the outcome of Proudfoot's investigation, because the explanations are directly linked to the descriptions; they are actually guided by the descriptions. With Proudfoot, not only is the context-sensitivity and specificity of descriptions missing, there is no logical link between description and explanation; his descriptions play no role in his explanations. His insistence that descriptions be done from the insider's perspective functions merely as a grandstanding claim for superiority and psychological sop to the insider (which for him is the individual subject of the experience). All of this is different in hermeneutic explanation: the insider is not an individual, the description is not of a single event of experience, and the descriptions guide the explanation. Since explanations are guided by descriptions made from an insider's perspective, there is no possibility of wholesale reduction in hermeneutic explanation. The sort of wholesale reductionism engaged in by Proudfoot, where all religious experience is explained purely in naturalistic terms, would not be possible. On the other hand, within the broader understanding of religion it allows for, some of the described phenomena may have to be explained away to bring about a coherent understanding. Thus, when explanatory reduction is done, it is done from within the insider's perspective in a broad sense.

Since the purpose of explanatory reduction in hermeneutic explanation is to provide a coherent account of the varied ‘inside’ accounts, if a particular reduction is accused of going too far by any insider (proficient user of that concept), the insider concerned can be challenged to provide an alternative that does better justice to the descriptions for the other users of the concept. Thus, while there is certainly room for criticism, mutual correction and reduction within the religious realm, there is no place for the reduction of religion to something else. This, I suggest, is the method by which we can arrive at the best explanation, not only for religion, but also for other concepts like ‘perception’, ‘virtue’, ‘knowledge’ and so forth. In effect such explanations provide the basis for philosophical theories, and the various descriptions also supply the criteria for judging the adequacy of such theories.

Having investigated the different types of inquiry, the different kinds of descriptions required in the different types of inquiry, and the link between description and explanation, we are now in a better position to decide whether what I have called hermeneutic explanation really deserves to be called an ‘explanation’. It is recognized that, except for a brief period when logical positivism reigned supreme, there is no philosophical consensus regarding what should be counted as an explanation. One of the main ways in which ‘explanation’ is understood today is in terms of ‘unification of the phenomena’, or ‘the best way of systematizing our knowledge’.32 Understood in this manner, hermeneutic explanation clearly qualifies to be called explanation, as it unifies the various descriptions relevant to a concept. It could be called explication, as long as explication is understood as the type of explanation appropriate to conceptual inquiry. The advantage in calling it ‘explanation’ is that this will enable us to perceive the continuity between hermeneutical and epistemic explanation, rather than dichotomize them; both may be considered explanations with their own distinctive identity and role in human inquiry. While such a usage recognizes the continuity, it still refuses to confuse one with the other. This is the point of the next subsection.

(iv) Hermeneutics and Epistemology

Here we must ask whether and how Proudfoot's legitimate concern can be safeguarded. He is concerned that the insider perspective of hermeneutics and language games not be used as a protective strategy to preclude critical inquiry. This concern is safeguarded if we realize that a conceptual inquiry is neither a substitute for empirical inquiry, nor immune to the findings of empirical research. First of all, to arrive at the descriptions appropriate for a conceptual inquiry, one must have sufficient familiarity with the empirical data. For example, a conceptual inquiry into religious experience does not require a detailed description of the peculiarities of the experience of St. Paul or Stephen Bradley or Ramana Maharshi. Therefore, genetic considerations such as the possible role of Methodist revivalism in early nineteenth-century New England, the particular meeting Bradley attended earlier in the evening, his adolescent conflicts, and so on would be irrelevant for a conceptual inquiry. On the other hand, no adequate description of the phenomena of religious experience can be done without knowledge of particular instances of religious experience, which is of course an empirical matter. It is a vast exposure to particular cases that makes possible the extraction of common characteristics required for conceptual inquiry. Such descriptions will then guide our hermeneutic explanations.

This is not the only point where our methodology touches the external world, however. Once hermeneutic explanation is done, conceptual inquiry goes into the background, and epistemic explanation takes over. Epistemic explanation, we have seen, is dependent on the findings of conceptual inquiry to accomplish its task. The former provides the norms, the standard or the language game within which particular claims to knowledge are tested as warranted or not. However, it is not concepts or language games, but the world that makes our beliefs true. It is not the circle of insiders, but the outside world that makes something true. Conceptual inquiry provides the standards, but whether these standards are met in the actual world is determined by the epistemic inquiry. If hermeneutic explanation has a constructivist orientation, epistemic explanation has a realist orientation; if hermeneutic explanation is a matter of understanding, epistemic explanation is a matter of looking for the warrant to see if what is understood is also true or supported by the way the world is. The insider view of a hermeneutic explanation, therefore, is ‘sandwiched’ by the outside world from both directions: on one side, by a knowledge of particular cases that feed into the descriptions, and on the other side through justification of what is understood.


In conclusion, Proudfoot's study of religious experience suffers from a serious methodological flaw. He does not recognize the autonomy of different types of inquiry and different kinds of explanations. As a result, the type of description he gives is appropriate to genetic inquiries, but not for understanding the concept of religion from the inside. Further, having missed out on religious understanding, he cannot be expected to do justice to religious truth. By contrast, we have recognised genetic, conceptual and epistemic investigations as different in kind, each requiring a different type of description and explanation. What is advocated here is therefore a more complex methodology that makes room for descriptions, hermeneutics, and justification. With these in place, we also acknowledge there is a place for explanatory reduction, but this would be part of hermeneutics and meant to provide a coherent theoretical account of concepts like ‘religion’, ‘perception’, ‘love’ and the like.33


  1. 1 Wayne Proudfoot, Religious Experience (Berkley and Los Angeles: University of California Press, 1985), p. xiv.

  2. 2 Ibid., It received the American Academy of Religion Award in 1986.

  3. 3 James C. Livingston, Anatomy of the Sacred: An Introduction to Religion, 3rd ed. (New Jersey: Prentice-Hall, 1998), pp. 43–44. Books that provide readings in philosophy of religion, often has a section on this too.

  4. 4 Proudfoot, Religious Experience, p. 47.

  5. 5 Ibid., p. 72.

  6. 6 Ibid., pp. 196–7.

  7. 7 Ibid., p. 67.

  8. 8 Ibid., p. 197.

  9. 9 Ibid., p. xvi.

  10. 10 Ibid., p. 200.

  11. 11 Ibid., p. xv.

  12. 12 Unlike Peter Lipton's famous book, Inference to the Best Explanation (London and New York: Routledge, 1991 and 2004, revised edition), that is primarily a work in Philosophy of Science, Proudfoot's discussion covers a wide range of philosophical issues that make it absolutely necessary to make these distinctions.

  13. 13 Not all expressions can be true or false. He might break out into a song and dance, which is an expression arising from the experience of love, but not true or false.

  14. 14 Carl Gustav Hempel, Philosophy of Natural Science (Englewood Cliffs, N.J.,: Prentice-Hall, 1966), p. 3ff.

  15. 15 Proudfoot, Religious Experience, p. 195.

  16. 16 Ibid., p. 223.

  17. 17 Ibid., p. 176.

  18. 18 Ibid., p. 217.

  19. 19 Ibid., p. 114.

  20. 20 William P. Alston, ‘Perceptual Knowledge,’ in The Blackwell Guide to Epistemology, ed. John Greco and Ernest Sosa (Malden, MA: Blackwell Publishers, 1999), p. 223.

  21. 21 This point has been repeatedly made by William Alston in his writings. See, for example, William Alston, Epistemic Justification (Ithaca and London: Cornell University Press, 1989), p. chapter 12.

  22. 22 Ludwig Wittgenstein, G. E. M. Anscombe, and G. H. von Wright, On Certainty, trans. Denis Paul and G. E. M. Anscombe (New York,: Harper, 1969), p. 105.

  23. 23 Norman Malcolm, ‘The Groundless Belief,’ in Reason and Religion, ed. Stuart C. Brown (Ithaca, NY: Cornell University Press, 1977), p. 151.

  24. 24 Proudfoot, Religious Experience, pp. 176–77.

  25. 25 Malcolm, ‘The Groundless Belief’. There are also certain passages in Wittgenstein's own writings that seem to suggest this. See for example, Ludwig Wittgenstein, Philosophical Investigations, trans. G.E.M. Anscombe (Oxford: Basil Blackwell, 1958), p. IIxi, pp. 200, 217, 654.

  26. 26 I am grateful to Prof. Winfried Löffler of Innsbruck for raising this question.

  27. 27 See, Karuvelil George, ‘Rationality of Mysticism: A Methodological Proposal’, Journal of Dharma 30: no. 4 (2005), p. 426.

  28. 28 Some of these descriptions relevant to perception can be found in George Karuvelil, ‘Knowledge and Religious Conciousness’, Journal of Indian Council of Philosophical Research XI: no. 1 (1993).

  29. 29 Proudfoot, Religious Experience, p. 127.

  30. 30 Ibid., p. 125.

  31. 31 Rudolf Otto, The Idea of the Holy: An Inquiry into the Non-Rational Factor in the Idea of the Divine and Its Relation to the Rational, trans. John W. Harvey (Oxford: OUP, 1923; reprint, 1936).

  32. 32 Philip Kitcher, ‘Explanation,’ in The Concise Routledge Encyclopedia of Philosophy (London and New York: Routledge, 2000). See the third way of understanding explanation cited by the author.

  33. 33 An earlier version of this paper was presented at the joint JDV-Innsbruck Conference held at Innsbruck in May 2007 and will be published in its proceedings.