9

hzt J Phil Rel 15:13-30 (1984). Martinus Nifhoff Publishers, The Hague. Printed in the "Vetherlands.

THEISTIC EXPERIENCE AND THE DOCTRINE OF UNANIMITY

J. WILLIAM FORGIE University of California - San ta Barbara

We can distinguish two sorts of questions that can be raised about a perceptual, or perception-like, experience. We can ask: (1) is the experience veridical, as opposed to hallucinatory? But we can also ask: (2) what is the object of the experience, i.e., what is the experience - veridical or not - an experience off. Answers to these two questions are independent of one another. One can try to identify the object of his experience without knowing whether or not the experience is veridical. Alternatively, one can try to decide whether his experience is veridical or hallucinatory without knowing what the object of the experience is (i.e., what it is a veridical perception, or an hallucination, of). If I take one of my sense experiences to be veridical, and thus to constitute a perception of something in the world around me, the request to identify the object of that experience can be construed simply as the request to identify what it is I am perceiving. In that case my answer will normally be based not only on the content of the experience itself but also on various extra-experiential considerations. Suppose, for example, that Tim and Tom Tibbetts are identical twins living next door. I look over the fence and have a visual experience of someone who looks like one of the Tibbetts twins sitting in the yard. If I had to rely solely on the content of my experience I would have no basis to identify its object as Tim, and not Tom, Tibbetts, for an experience of one would be, by hypothesis, phenomenologically indistinguishable from an experience of the other. But I may know that Tom is out of town this week, and that, together with my reasonable belief that anyone in that yard looking like the young man I see is very probably either Tim or Tom, leads me confidently to identify the object of my experience as Tim Tibbetts. Let us call the collection of beliefs one has and items of knowledge one possesses - this epistemological baggage which one brings to a particular experience one's "epistemic base" relative to that experience. In the example just mentioned my identification of the object of my visual experience was based in part on the content of the experience and in part on various aspects of my epistemic base. Suppose, however, that I do not take a particular experience to be veridical. Perhaps I am even prepared to admit that it is hallucinatory. How do I identify its

14 object in this case? Here I might well simply ask myself what I would take that object to be if I were convinced that the experience was veridical. This would allow me again to base my identification both on the content of the experience and on aspects of my epistemic base. My identification would then involve essentially the same process as that used in the case of an experience which is taken as veridical. In the Tibbetts example~ even if I knew my experience was hallucinatory I would still identify the object as Tim Tibbetts. For I would pose to myself the question, "If I took this experience to be veridical how would I identify its object?", and I would have the same reasons for giving the same answer I gave previously when I did take it to be veridical. If one is asked to identify the object of an experience, whether or not he takes the experience to be veridical, I believe he would normally rely (as our examples so far illustrate) both on the content of the experience and on his epistemic base. But let us consider another possible procedure. I might try to base my identification of the object solely on an examination of the phenomenological content of the experience itself, eschewing any appeal to my epistemic base. As before I might try this approach whether or not I took the experience to be veridical. In either case I would ask myself: "Suppose this experience is veridical and so constitutes an accurate perception of something, X; then given only the phenomenological content of the experience itself, what would X have to be or what features would X have to have?" My aim here is to isolate those things which the content of the experience itself guarantees will be true of anything of which that experience could be an accurate perception. These things I will think of as the "phenomenologically guaranteed" truths about the object of the experience. And it is these to which I will want to restrict myself when I try to identify the object of an experience while relying only on its phenomenological content. If one identifies or describes the object of an experience mentioning only phenomenologically guaranteed truths about that object, let us say he produces a "phenomenological" account or description of the experience. And if claims of the form, "It's an experience of N" or "It's of something @," are phenomenological accounts of a particular experience we will say the experience is "phenomenologically of N" or "phenomenologically of something r A phenomenological account of an experience will normally differ from the identification one would make were he also allowed to rely on his epistemic base. When I use my epistemic base in the Tibbetts example I identify the object of my experience as Tim Tibbetts. But that experience would not be phenomenologically of Tim Tibbetts, i.e., "It's of Tim Tibbetts" would not be a phenomenological account of the experience, for nothing in the content of the experience guarantees that its object is Tim Tibbetts as opposed to Tom Tibbetts, or for that matter any of a number of other things - a third "look-alike," an appropriately made-up dummy, or even a cleverly devised hologram - an accurate perception of which could be phenomenologically indistinguishable from the experience in question. Phenomenological descriptions will have to play an essential role in any ultimate assessment of the Doctrine of Unanimity (DOU), the view that mystical ex-

15 periences had by prophets, saints and ordinary people from various religious traditions and at various times are essentially the same at the phenomenological level, and that the striking differences in the descriptions and identifications of the objects of those experiences simply represent different "interpretations" which are imposed on the experiences and which reflect the different religious and cultural backgrounds of the experiencer themselves) Stated in the language we have been using, the DOU claims that the different descriptions actually given of mystical experiences are due not to differences in the content of those experiences themselves, but instead to differences in the epistemic bases of those having the experiences, and that were those individuals to confine themselves to giving phenomenological descriptions of their several experiences, their accounts would be essentially the same. If we are to evaluate this doctrine it is clear that we need access to phenomenological descriptions of the relevant experiences. We need to be able to extract from a mystic's description those parts which serve as phenomenological accounts from those that incorporate the mystic's epistemic base. Otherwise we run the riks of mistakenly rejecting the DOU because we wrongly take different descriptions as reflecting different experiences. Alternatively we might mistakenly accept the doctrine, wrongly regarding different descriptions as due to different epistemic bases and not to different experiences. Sameness and differences in the mystics' descriptions show nothing one way or the other about the DOU so long as those descriptions involve the reporters' epistemic bases. But sameness and differences in phenomenological accounts will, respectively, confirm or refute the DOU.: Let us say that a particular mystical experience is a "theistic experience" just in case the claim that its object (or part of its object, if the experience is more complex) is God would constitute a phenomenological account of that experience. More briefly, a theistic experience is a mystical experience which is phenomenologically of God. In this paper I wish to discuss the challenge to the DOU posed by the alleged existence of theistic experience. This challenge can be simply stated as follows. We can plausibly suppose that there have occurred many mystical experiences that are not theistic. Consequently if theistic experiences have existed, the DOU is mistaken. Theistic experiences have indeed existed. Therefore, the DOU is mistaken. 3 I believe the DOU has nothing to fear from theistic experience, for I suspect such experiences are not possible. In the next section I will offer considerations in support of this suspicion. In a final section I will criticize a defense of the possibility of theistic experience offered by Nelson Pike.

I Even a casual acquaintance with the history of Christian mysticism can leave no doubt that there have occurred many experiences whose object has been taken by their possessors to be God. 4 Now of course defenders of the DOU will say that

16 in so taking their experiences the Christian mystics are invoking extra-experiential beliefs and not giving phenomenological descriptions, and so the experiences they are describing have not been shown to disconfirm the DOU itself. Such a response can seem dogmatic unless backed up by something more than an antecedent commitment to the DOU. But perhaps such backing can be provided. In this section I hope to bring out some considerations which suggest, at least prima facie, that "It's of God" could not be a phenomenological account of any experience, i.e., that theistic experiences are not possible. Whether theistic experiences are possible may depend on whether 'God' is a proper name of certain individual or a (disguised) definite description. If it is a proper name, then if an experience is to be phenomenologically of God, the content of the experience must guarantee that its object is a certain unique individual, the one named by 'God', and not any other. It must not be possible, that is, for the experience to constitute an accurate "perception ''s of some individual other than God. (My visual experience in the Tibbetts example is not phenomenologically of Tim Tibbetts because the experience could constitute an accurate perception of someone other than that particular individual, e.g., Tom Tibbetts.) On the other hand, if 'God' is a description, meaning (let us suppose) 'the all-powerful, allknowing, all-good creator of the heavens and the earth', then a theistic experience need only be phenomenologically of some individual or other - it does't matter which one - who satisfies that description. In this case it is required only that it not be possible that the experience constitute an accurate perception of something that fails to satisfy the description. Suppose 'God' is a name, i.e., that by 'God' we mean a specific individual and not just some being or other who satisfies a certain description. Then a theistic experience will have to be one which is phenomenologically of that specific individual. How is this possible? Indeed how could any exPerience, mystical or otherwise, be phenomenologically of an individual? Certainly no sense experience seems to have that feature. Nothing in a visual experience, for example, guarantees that its object is some particular individual (e.g., Tim Tibbetts) as opposed to another possible individual (Tom Tibbetts) who looks exactly the same. Similarly nothing in an auditory experience makes it phenomenologically of A (e.g., your telephone) as opposed to some B which might produce a phenomenologically indistinguishable sound. I believe analogous remarks hold for the remaining senses. So if some mystical experience really were phenomenologically of God (the individual), it would be strikingly different (in a way not frequently emphasized) from any sense experience. Let us use the label "UIP" ("uniquely instantiable property") for a property that can be instantiated by at most one possible individual. Now a necessary and sufficient condition for any experience being phenomenologicaUy of an individual is that it be phenomenologlcaUy of something having a UIP. If this condition is not met, if an experience is at best phenomenologically of something r where "being 0" is a property that could be possessed by more than one possible individual, then nothing in the content of that experience will guarantee that its object is any particular one of those possible individuals to the exclusion of the

17 others. So the condition is necessary. On the other hand if the condition is met, then if the experience constitutes an accurate perception of something it must be a perception of that unique individual who alone can have the UIP in question. The condition is sufficient. It is worth noting that many of the properties typically attributed to God are not UIPs. So even if an experience were phenomenologically of something having those traditional properties it would not follow that it was phenomenologically of an individual, let alone of God. Take "being omnipotent," "being omniscient," "being all-good" or "being the creator of the heavens and the earth." Even if these properties are instantiated in this world by a certain individual, A, there is another possible world in which they are instantiated by a numerically distinct individual, B. (We can suppose B also exists in the actual world but without the properties we are discussing.) These properties are thus not UIPs. This shows, incidentally, why it is important in our present discussion to distinguish between 'God' as a name and 'God' as a description. For suppose some experience is phenomenologicaUy of something answering the description, 'the omnipotent, omniscient, all-good creator of the heavens and the earth'. Then the experience would be phenomenologically of God if 'God' is an abbreviation of that description. But since the properties mentioned in the description are not UIPs (neither singly nor collectively) it will not follow that the experience is phenomenologically of some specific individual. 6 More particularly it will not follow that it is phenomenologically of God if 'God' is a name. It is not hard to find examples of UIPs. "Being Socrates," for example, can be instantiated only by Socrates. If that property is instantiated in any possible world it is instantiated in that world by Socrates. 7 Similarly, "being Tim Tibbetts" is instantiable only by Tim Tibbetts, the particular individual who happens to live next door. More to our present concern, if 'God' is a name, then the property "being God" is instantiable by only one possible individual. However even if a particular individual has a UIP it does not follow that an experience could be phenomenologically of that individual. After all, Tim Tibbetts has the UIP, "being Tim Tibbetts" but no sense experience is phenomenologicalty of that person (and it is not at all clear what other kind of experience might be). So if it is to be possible that an experience be phenomenologically of A it is not enough that A have a UIP. Something more is required. Perhaps we can see what that is if we once again consider sense experience. Sense experiences are not phenomenologically of individuals, but at best only of things that appear in certain ways. A visual experience, for example, is at best phenomenologically of something that looks a certain way. And auditory or tactile experiences are at best phenomenologically of things that sound or feel certain ways. But numerically distinct individuals can look the same, or sound or feel the same. What is necessarily unique to a particular individual does not include how it looks or sounds or feels. Perhaps it does include that individual's being constituted, either now, or at some earlier time, of precisely this set of molecules or, in the case of certain sorts of beings, of its having originated from precisely this cell or this

18 sperm and this egg. But numerically distinct molecules or eggs or sperm could also be phenomenologically indistinguishable. And so even if in seeing, or otherwise sensing, some individual we somehow also saw (etc.) its component molecules or the sperm and egg from which it originated, the experience still would not be phenomenologically of one particular individual as opposed to a different and phenomenologically indistinguishable individual whose molecules or whose originating sperm and egg al~peared the same. The point that emerges here can be put in either of the following two ways. (1) At best sense experiences are phenomenologically of things that appear in a certain way, but since properties of the form "being something that looks (sounds, feels, etc.) - or is capable of looking (etc.) - this way" are not UIPs, sense experiences are not phenomenologically of individuals. (2) If a sense experience is to be phenomenologically of an individual, it is not enough that that individual have a UIP. It must have a UIP of the form "being something which appears - or which is capable of appearing - in a certain way." It is because no object of sense experience seems to have a UIP of that form that no sense experience is phenomenologically of an individual. How does this bear on the possibility of theistic experience? Perhaps it is not relevant at all. One could always maintain that mystical experience is so radically unlike sense experience that any truths about the latter have no implications for the former. But if we accept the putative analogy between the two sorts of experience noted earlier s it does not seem unreasonable to think that the results of our discussion so far can be extended to mystical experience. In order for a mystical experience to be theistic what is required is not just that God have a UIP. What is required is that He have a UIP of the form "is capable of appearing in a certain way" (viz., the way displayed in that allegedly theistic experience). For if an experience were phenomenologically of something that appears in that way, and if it were possible that some individual other than God had the capacity to appear that way, then nothing in the content of that experience would guarantee that its object is God rather than the other individual, and the experience would therefore not be theistic. Now I know of no reason to think there are any UIPs of the form "is capable of appearing in a certain way," much less that any of them belongs to God. On the other hand I know of no argument to show that there are no such UIPs. As is typically the case in thinking about whether or not something is possible, one ultimately comes down to intuitions. My own intuitions suggest there are no such UIPs. Given any experience which is alleged to be phenomenologicaUy of God, it seems intuitively that there is a possible world in which the causal laws pertaining to the relations between possible objects of "perception" and the "perceivers" of those objects are such that some individual, not identical to God, is capable of appearing in just the way displayed in the experience in question. If so, then nothing in the content of an experience will render it phenomenologically of God as opposed to another possible individual. In the absence of considerations tending to show that this intuition is mistaken I am inclined therefore to think that if

19 'God' is a proper name then theistic experience is not possible. But suppose we treat 'God' as short for a definite description. Then a theistic experience need not be phenomenologically of an individual, but only of something answering a certain description. As mentioned earlier, a typical description might be 'the all-powerful, aU-knowing, all-good creator of the heavens and the earth'. Could an experience be phenomenologically of something having the properties mentioned in this description? It will be of help in trying to answer this question to consider causation, or causal efficacy. Might a sense experience be phenomenologically of one event causing another? Suppose we see one billiard ball roll into another, at which point the first stops and the second rolls away. We might naturally say we had seen event A (the first ball rolling to a certain point) cause event B (the second ball rolling away from that point). But suppose we had to confine ourselves to phenomenological descriptions. What would we say? Even if we allowed that our visual experience is phenomenologicaUy of one event succeeded by another, would we allow that it is phenomenologically of one event causing another? I should think not. Causation, regarded as something over and above mere sequence, does not seem to be "phenomenologicaUy presentable." There would seem to be a possible world in which an event phenomenologically indistinguishable from A is merely succeeded by, but does not cause, an event phenomenologicaUy indistinguishable from B and in which our visual experience while watching the two, merely sequentially related, events is itself phenomenologically indistinguishable from the one we have while watching A cause B. Similar considerations hold for cases where it is an agent, not an event, which is supposed to be the cause, whether the agent is thought to be the cause of some event (e.g., a chisel's moving, or a statue's coming to be) or of some substance (e.g. the statue itself). Suppose we perceive some agent, A, cause some event or substance, B. There would seem to be a possible world in which A exists, then later B occurs or comes to be without being caused by A 9 , and yet our perceptual experience of "A then B" is phenomenologically indistinguishable from that of A causing B. Again no experience would be phenomenologically of A causing B as opposed to A merely succeeded by B. 1~ I am suggesting that no experience is phenomenologically of something causing (or being the cause of) something else. The following consideration may help to make this suggestion plausible. Suppose that an experience is phenomenologicaUy of A and phenomenologically of B. 11 Now suppose it is claimed that there is something in the content of this experience which distinguishes it from one which is phenomenologically merely of A and B, or of A then B, that this experience is also phenomenologically of some third factor, X, where X is said to be A's causing B. It should then be asked: if the experience is phenomenologically of a third factor, what makes that factor causation rather then merely some additional element which either co-exists non-causally or occurs in a non-causal sequence with A and B? What makes the experience one of A causing B rather than one of three things, A, X and B, which merely co-exist or are merely related sequentially? Is there no

20 possible world in which A, X and B (or things phenomenologically indistinguishable from A, X and B) are not related causally but instead merely co-exist or occur in sequence and in which an experience of those three elements is phenomenologically indistinguishable from the experience supposed to be phenomenologically of A causing B? Suppose we are right in suggesting that causation, regarded as something more than mere sequence (or co-existence), is not phenomenologically preientable. Then there are difficulties in supposing that an experience could be phenomenologically of some being having any of the properties mentioned in the description, 'the allpowerful, all-knowing, all-good creator of the heavens and the earth'. Take "being all-powerful" and "being all-knowing." I assume that the former involves, as a minimum, having the power or ability to bring about certain states of affairs, and that the latter involves, also as a minimum, the having of many true beliefs. Now I take it that powers and beliefs are not, by themselves, perceivable aspects of those beings having them. In saying "by themselves" I mean to suggest that there is nothing about an agent, apart from its manifesting its power or expressing its belief, to which one could draw attention and say "this, which you now perceive, is the ability to do such and such or the belief that so and so." Apart from my doing something with my legs, nothing about me which anyone perceives can be said to be my ability to run a mile in under six minutes. (Perhaps my muscles enable me to run that fast, but it would be absurd to identtfy my ability with my muscles.) And apart from some expression (including verbal expression) of my belief that Nixon resigned the Presidency there is no perceivable aspect of myself of which it can be said "this is that belief about Nixon. ''12 The point here is not one about the limitations of the five senses or of human perceptual abilities - as though some sixth sense might be used to perceive powers and beliefs. It is intended as a point about powers and beliefs themselves: they are not the sorts of things which by themselves can be objects of perception. If the point here is correct, then it would seem that the best candidate for an experience which is phenomenologically of something having certain powers and beliefs is one which is phenomenologicaily of something manifesting those powers or expressing those beliefs. If there can be no experience which is phenomenologically of some power, or some belief, by itself (a claim which would seem to follow from the point made in the previous paragraph), perhaps an experience can be phenomenologicaily of something manifesting a power or expressing a belief. If so, then an experience itself could guarantee that its object is something manifesting, and hence possessiong, that power, and also something expressing, and so having, that belief. But here is where the earlier point about causation is important. If causation is not phenomenologically presentable then neither is agency. If some agent is manifesting a power or expressing a belief, that agent is causing something to happen, producing some state of affairs) 3 But if no experience is phenomenologically of someone's causing or producing a state of affairs (as opposed to that state of affairs simply co-existing with the agent or coming into existence while the agent is present), then no experience will be phenomenologicaUy of

21 someone manifesting a power or expressing a belief. So the best candidate for an experience which is phenomenologically of something having certain powers or beliefs turns out not to be up to the job. Similar sorts of considerations apply to the property, "being all-good." It is surely reasonable to suppose that an all-good being is, again at a minimum, a being who is disposed to perform, and/or who actually performs, certain (morally good) actions and someone who has certain (morally laudable) beliefs. We have already suggested that no experience is phenomenologically of someone having certain beliefs, nor of someone performing certain actions, i.e., causing certain things to happen. And a disposition, like a power, would seem not to be, by itself, a perceivable aspect of any being having it. It would at best be phenomenologically presentable only through its being manifested. But a being's manifesting a disposition to perform certain sorts of actions would involve agency and thus would not be phenomenologically presentable. So it is not easy to see how an experience could be phenomenologically of something which is all-good. Finally, consider the property "being the creator of the heavens and the earth." In claiming that God has this property one typically means to make a point about God's current role as a sustaining cause or His past role as an efficient cause. But if causation is not phenomenologically presentable then no experience will be phenomenologically of any being currently sustaining the heavens and the earth. 14 And it is even more difficult to see how an experience at the present time would be phenomenologically of a being who, in the past, was the efficient cause of something. Even if, per impossibile, we were witnesses of that act of creation at the time it occurred (or if, by some miracle, our experience in the present allowed us to "see" backward to that particular moment), our experience would not, for reasons already given, be phenomenologically of some being causing anything. And once that moment of creation has passed, what of any relevance would remain with that being such that an experience of it in the present would be phenomenologically of something which once caused something? I have been attempting, in this section, to cast doubt on the possibility of theistic experiences, whether 'God' is construed as the proper name of a particular individual or as an abbreviated definite description. Claims of the form, "The object of that experience is God," are not phenomenological accounts of the experiences they report. 15 This result squares nicely with the claim by proponents of the DOU, viz., that reports of experiences as being "of God" are not phenomenological accounts, but rather "interpretations" which reflect the epistemic bases of the reporters. The DOU should thus have nothing to fear from the alleged existence of theistic experience. Two final points of clarification may be helpful. First, if the considerations presented here are cogent it does not of course follow that God is never the object of any experience, nor that no one ever "perceives" God, nor that one is never justified in claiming to have "perceived" God - anymore than it would follow from the fact that no experience is phenomenologically of Tim Tibbetts that he is never the object of any visual experience, or that no one ever sees him, or that no

22 one is justified in claiming to see him. Second, nothing in the arguments I have suggested shows that the DOU is correct. Those mystical experiences which are mistakenly said to be theistic may still be phenomenologically different, in important ways, from other (e.g., so-called "monistic") mystical experiences.

II I want now to consider an intriguing defense of the possibility of theistic experience put forward by Nelson Pike. His defense is contained in a reply to a paper by Ninian Smart in which Smart defends the DOU against the views of Zaehner. Pike thinks Zaehner is correct in distinguishing "theistic" from other sorts of mystical experience) 6 Pike calls attention to the following report of an experience from James' Varie-

ties of Religious Experience: There was not a mere consciousness of something there, but fused in the central happiness of it, a startling awareness of some ineffable good. Not vague either, not like the emotional effect of some poem, or scene or blossom, of music, but the sure knowledge of the close presence of a sort of mighty person, and after it went, the memory persisted as the one perception of reality. Everything else might be a dream, but not that. 17 Pike then makes a number of comments about this report. First, he takes the report to be what we are calling a phenomenological account of an experience; the reporter, Pike states it, is mentioning only aspects of the content of the experience itself. (p. 146) Second, Pike claims that one could just as well describe the experience as one in which the reporter was "aware of Self in contact with God." This could simply be one way, Pike says, of stating that the reporter experienced himself in contact with an ineffably good and overwhelmingly powerful person. And it would still be merely a phenomenological account of that experience. As Pike puts it, "'self in contact with God' mentions nothing other than the explicit content of the experience." (p. 146) Then he says this: ...James's reporter tells us that he was aware of a good and mighty person. By what marks did he determine that what he was aware of was a person, or (given that he was aware of a person) that the person in question was good and mighty? For example, did the experience involve a visual image of a man having a kindly face and big muscles? If it did, then we know why he says that he was aware of a person who was good and mighty. Of course, Jarnes's reporter would probably reject this suggestion. He would probably say that his experience involved no (visual) image whatsoever. Then how did he know that he experienced a good and mighty person? I think this line of questioning is misguided. Suppose I report a dream in which I was riding a carousel and the President of the United States was riding the horse beside me. By what mark did I determine that there was a person riding beside me; or (given that I experienced a person riding beside me) that the person in question was

23 the President of the United States? I might well have no answer to this question. I might go on to say that in my dream I was so nervous about this frolic with the President that I could not bring myself to look at him. I had no visual image of a person bearing certain presidential features. I had no visual image of a person at all. That I was aware of the President riding beside me was an immediate datum of the dream. I do not need to itemize criteria by which I determined the content of my experience. So too in the case of James's reporter. One can be aware of himself in the presence of a good and mighty person without being aware of the identifying marks by which good and powerful persons are ordinarily recognized. (pp. 147148.) is The example Pike is discussing may be unfortunate in at least one respect. Since he allows that 'self in contact with God' would constitute a phenomenological account of the experience, he commits himself to regarding the experience as what we are calling a "theistic" one. But some defenders of the DOU (e.g. Smart) distinguish numinous and mystical experiences, maintaining that the DOU is true only of the latter. The experience Pike cites, however, might seem to be numinous, and so even if he were fight in regarding the experience as theistic it would not create problems for the DOU. Indeed Smart, in reply to Pike, regards the example as irrelevant to the DOU, and for just this reason. (p. 156) But the interesting feature here is what Pike does with the example. If we had before us an example of a mystical experience which was described with the words 'self in contact with God', and if we take the description as intended as a phenomenological account, we would have something directly relevant to the DOU. (In fact Pike actually goes on to discuss such an example from yon Ruysbroeck, pp. 148ff.) We could then ask about the relevant example just the sorts of questions Pike imagines being raised about the example from James. Since Pike's response is couched in terms of the example from James, I propose to stick with the James example, pretending that it is a relevant sort of theistic experience. Suppose you are convinced there can be no such thing as a theistic experience. Then when someone describes an experience as "of God" you will assume he must be basing his description in part on extra-experiential beliefs and not on the content of the experience alone. One possible way of exposing the description, "It's an experience of God," as not being a phenomenological account is to ask what particular aspects or features of the experience justify the description given. The expectation is that any phenomenologically presentable features then mentioned will be seen to fall short, by themselves, of justifying the description, "of God," thus showing that the reporter must have been supplementing the fragmentary experiential grounds with extra-experiential considerations culled from his espistemic base. 19 Pike attempts to blunt this line of attack. His response to questions like, "How, by what marks or features, do you know that the object is God, or all-good, or just a good and mighty person?" is to reject them. They wrongly presuppose that one can be aware that the object of an experience is a person, or good or powerful, only by being aware of features or "identifying marks" by which good and power-

24 ful persons are recognized. That this presupposition is mistaken is shown by the case of dreams. The content of Pike's dream included his riding a carousel next to the President of the United States. But Pike did not determine on the basis of features or marks that there was a person next to him or that that person was the President. That he was aware of the President riding next to him was "an immediate datum" of the dream. This example shows that in a dream one can be aware of being in the presence of such and such without being aware of any identifying marks of a such-and-such, and can therefore reject the question "How did you know it was a such-and-such you were in the presence of?." replying "By no marks; that I was in the presence of a such-and-such was just an immediate datum of the dream." Pike believes a similar sort of response is in principle open to James' reporter (and presumably to anyone else reporting an experience alleged to be theistic). When asked how he knew, by what marks he determined, that the object of his experience was a good and mighty person (or that it was God), he could reply, "By no marks; I was aware of myself in the presence of a good and mighty person without being aware of any identifying features by which good and powerful persons are ordinarily recognized; that the thing I was in the presence of was a person, and that it was good and mighty (or that the object of the experience was God), were immediate data of my experience." Such a response will enable the mystic to ward off the sort of sceptical doubts we might have about theistic experience. Instead of being led to give up the claim that "It's an experience of God" is a phenomenological account of the experience because he is brought to see that the features and marks he is relying on fail to guarantee that the object is God - the mystic reinforces his conviction that the experience is theistic with the claim that its being of God is an "immediate datum" of the experience itself. Let us say, following what appears to be Pike's usage, that some aspect, 0, is an immediate datum of N's experience just in case (1) the experience has 0 ; and (2) N knows (or, perhaps, can know) that that experience has r without determining that fact from features or marks. 2~ We will also say that one is employing "Pike's strategem' if he attempts to reject questions of the form "How did you know, by what marks or features did you determine, that that experience has aspect 0?" by replying "It's having 0 was simply an immediate datum of the experience." Now Pike's defense of the possibility of theistic experience depends on his making intelligible how something can be an "immediate datum" of an experience. Only in this way will he be able to explain how Pike's strategem can legitimately be used by James' reporter. Pike thinks he can do this by citing the example of dreams. I will suggest, however, that although Pike's strategem does seem to work for dreams, the dream model fails to make intelligible how something can be an immediate datum of a non-dream experience. If this is correct, then Pike's defense of theistic experience fails: he will not have shown how James' reporter can use Pike's strategem to overcome sceptical doubts about the object of his experience. Pike's strategem does seem intuitively attractive when considering dreams. -

25 To take another example of this, suppose I have a dream in which I have a veridical visual experience of Tim Tibbetts. Suppose that in the dream, as in real life, Tim and Tom are identical twins. How did I know, by what marks did I determine, that the object of the visual experience was Tim, and not Tom, Tibbetts? I might well have no answer to this question. I did not, in the dream, ascertain that Tom was out of town, nor did I have the belief that it is highly unlikely that there is a third look-alike, besides Tim and Tom, whom I might be perceiving. So I did not infer that the object was Tim in the way I might in real life. Nor did anyone in the dream tell me that I was seeing Tim Tibbetts. That the object of my visual experience was Tim Tibbetts was "an immediate datum" of the dream. Staying with this same example, how did I know (in the dream) that my experience of Tim Tibbetts was veridical? Here again I may not be able to point to marks or features in the situation - coherence, corroboration by other witnesses, or whatever else one might cite were the same question to arise in real life - from which I judged the experience veridical. That my experience was veridical was again just an immediate datum of my dream. Let us suppose these examples (as well as Pike's carousel example) exhibit a legitimate use of Pike's strategem. But what about its use outside dreams? Of course from the fact that it is possible for some state of affairs to obtain in a dream - e.g., being aware of being in the presence of something ~without being aware of any marks or features by which the presence of that thing is normally determined - it does not follow that that state of affairs is possible. "In my dream, p" or "It is possible that in someone's dream p is true" does not entail "P is possible." As Hobbes noted, 21 "In my dream, p" just means "I dreamed that p." And it is certainly possible to dream all sorts of things - e.g., that I squared the circle or discovered the largest positive integer - that are simply not possible states of affairs. So it would be a mistake to infer from the sort of possibility which Pike supposes obtains in his dream about the carousel, for example, to the legitimacy of using Pike's strategem in response to questions about non-dream experiences, including the experience enjoyed by James' reporter. This is just as well, for there is good reason to want to resist the use of Pike's strategem outside dreams. Consider some of the wildly implausible uses to which it can be put. We know that Pike would allow James' reporter to use that strategem to defend his claim that the object of his experience is a good and mighty person. But why stop there? Why not allow the same maneuver with ordinary sense experience? If I claim to have just had a veridical visual experience of Tim Tibbetts and am asked how I knew the object of the experience was Tim and not Tom, and how I knew the experience was veridical, why can I not make the same sort of move that can be plausible when discussing a dream, saying that it was an immediate datum of my experience that its object was Tim Tibbetts and also an immediate datum of the experience that it was veridical? Furthermore if James' reporter were asked why he supposes that his experience was veridical ("Everything else might be a dream, but not that."), why couldn't he use the same strategy Pike has him use in connection with his identification of the object, saying that not only

26 was it an immediate datum of his experience that its object was a good and mighty person, but also an immediate datum of the experience that it was veridical? But of course these uses of Pike's strategem are absurd. That the object of a visual experience is one and not the other of a pair of identical twins is just the sort of thing that cannot be an "immediate datum" of that experience. A visual experience of one is going to be indistinguishable in content from a visual experience of another. Any "datum" of the one experience is going to be a datum of the other. So there can be no datum which makes the experience one of Tim and not Tom Tibbetts. Furthermore whether or not a perceptual, or perception-like, experience is veridical is a matter not of some aspect of its content but instead, roughly speaking, a matter of the causal relations holding between the experience, whatever its content, and the thing allegedly perceived through it. Indeed, for any datum one might think could be added to an experience in order to insure its veridicality - from the inside, so to speak - it would still seem open to inquire whether that augmented experience was itself veridical, i.e., whether it bore the appropriate causal relations to that of which it is presumably a perception. Pike's strategem, applied to claims of veridicality, would amount to a novel defense of "self-authenticating" experiences. However once we allow James' reporter to use Pike's strategem in the way Pike himself advocates, it is hard to see how we can disallow its use in these other cases. Consequently I am inclined to resist the use of Pike's strategem in non-dream contexts. At the same time, that maneuver does seem to work for dreams. It would be helpful, therefore, to find a way of explaining why Pike's strategem works with dreams which does not commit us to accepting its possible use outside dreams. How is it that in his dream Pike can be aware of being next to the President without being aware of marks or features by which he determined that there was something next to him, that it was a person and that it was the President? The answer, I believe, is that the content of a dream is typically not exhausted by the content of the perceptual experiences occurring in that dream. In Pike's dream there may have occurred only a few perceptual experiences, e.g., the visual experience one might have if he were riding the carousal while looking straight ahead, and perhaps an auditory experience of children squealing and laughing. By hypothesis there was no relevant perceptual experience (e.g., looking to his left and seeing the adjacent rider, or hearing someone tell him who was on the next horse) which would inform him that the President was next to him. Yet in the dream he was aware of the President being there. In dreams certain things are not learned, let alone learned from the content of the perceptual experiences occurring in the dream. They are just understood, "given" if we like. 22 They are known in the dream without being known from something (e.g., some perceptual experience) in the dream. Such things might naturally be caUed "immediate data" of the dream. Pike's awareness that he was riding next to the President is one such immediate datum. That is why, when asked how, in the dream, he knew it was the President, or indeed a person, next to him he can say, "Not from features not from marks; I just knew it; it was given, something understood, an immediate datum of the

27 dream." Analogously that is also why in my dream I can have a visual experience and know both that its object is Tim Tibbetts and that it is veridical. These items of awareness may be just understood. They may be known in my dream without being known or determined from anything in that dream. So if I am asked how, in the dream, I knew the experience was of Tim, not Tom, and also how I knew it was veridical, I can reply, "I just knew these things; I didn't determine them to be tree from various features or marks; they were just given; they were immediate data of my dream." We can regard dreams as having, typically, both a phenomenological and a nonphenomenological content. The former includes the content of perception-like experiences which occur in the dream, and the latter comprises those things which are simply understood. So far as I can see the intuitive plausibility - and, more basically, even the intelligibility - of using Pike's strategem with dreams depends on their having this sort of non-phenomenological content. For to say that I~ is an immediate datum of a dream seems just to mean that ~ is one of the elements in the non-phenomenological content of that dream. Now I take it that the content of sense experiences and (let us temporarily suppose) of mystical experiences is exhausted by their phenomenological content. If something is part of the content of a visual experience, for example, it will have to be part of its phenomenological content, for that is all the content there is. There is nothing analogous here to those things which in dreams are simply understood or given. But in the supposedly instructive and paradigmatic case of dreams it is just these that are indispensable to our understanding how something can be an "immediate datum" of an experience. And so the dream model, which is all we have to go on here, does not make intelligible how something can be an immediate datum of a non-dream experience. That is why we can accept Pike's strategem as plausible with dreams but resist its use elsewhere. But suppose we are mistaken in assuming that mystical experiences have no content over and above their phenomenological content. 23 Then it may be possible, using the dream model as a guide, to explain how, for example, its being of a good and mighty person - or of God, or of self in contact with God - could be an immediate datum of the experience had by James' reporter. The explanation would involve (as the dream model suggests) regarding any such feature as something included in the supposed non-phenomenological content of that experience. This will be of no help to Pike, however. For he is concerned, let us remember, with phenomenological descriptions of mystical experiences, descriptions which mention nothing other than the phenomenological content of the experiences themselves. (Otherwise his remarks would have no relevance to the DOU.) His defense of theistic experience is supposed to be a defense of experiences which are phenomenologicaUy of God. So it will do him no good to explain (1) how "being of God" could be an immediate datum of a mystical experience, unless he also makes clear (2) how it could be an immediate datum while also being part of the phenomenological content of the experience in question. Even if mystical experiences, like dreams, have non-phenomenological content, the dream model

28 can at best help with the first o f these tasks, b u t n o t the second. I conclude that Pike's defense o f theistic experience is unsuccessful. The dream example does n o t make intelligible h o w J a m e s ' reporter can use Pike's strategem to overcome sceptical doubts a b o u t the object o f his experience. That example shows, at most, that something can be an i m m e d i a t e d a t u m o f an experience by being a part o f its n o n - p h e n o m e n o l o g i c a l content. But such a m o d e l has no useful application to experiences whose c o n t e n t is exhausted by their p h e n o m e n o l o g i c a l c o m p o n e n t s , nor - to speak m o r e directly to P i k e ' s purposes - to experiences whose p h e n o m e n o l o g i c a l c o n t e n t is our only concern.

NOTES 1. See, for example, Ninian Smart, "Interpretation and Mystical Experience," Religious Studies, Vol. 1, No. 1 (1965):75-87. 2. As I have introduced the notion, "phenomenological accounts" are given only of experiences which have objects. In claiming that such accounts are crucial to assessing the DOU I thus assume that mystical experiences, like sense experiences, have objects. This assumption does not seem unorthodox. See footnote 5 below. 3. R.C. Zaehner's attack on the DOU is based in part on the alleged existence of theistic experiences. See, for example, his Mysticism Sacred and Profane (Oxford, 1957), and Chapters X and XV of Concordant Discord (Oxford, 1970). 4. See, for example, the wealth of examples collected by Augustin Poulain in his The Graces o f lnterior Prayer, trans. Leonora L. Yorke Smith (London, 1921). 5. I use the term 'perception' here, although in scare quotes, because mystical experiences are frequently supposed to be like sense experiences in certain respects. They have phenomenological content (though presumably not necessarily the sorts of content, visual or tactile, etc., associated with the five physical senses); they have objects; and they are said to be possibly veridical or possibly hallucinatory - in short, they are thought to constitute possible "perceptions," or perceptionqike experiences, of something (e.g., God). In this paper I will assume that the two sorts of experiences are analogous in these ways. 6. Call this world "alpha." Now it might be objected that we know we are in alpha and hence that our experience, if veridical, is a perception of some individual in alpha. We may also know, the objection continues, that A is the only individual in alpha who has the several properties under discussion. Consequently our experience must be phenomenologically of A; it is not possible that our experience be an accurate perception of some individual other than A. This objection does not take seriously enough our focus on phenornenological accounts of experiences. Our knowledge that we are in alpha and that A is the only individual in alpha with the relevant properties is part of our epistemic base. It is not something apparent from the content of the experience itself. The content does not guarantee that we are in alpha, nor that our experience occurs in alpha, nor that A is the only individual in our world with the relevant properties. Consequently the content does not guarantee that the object is A and not some other possible possessor of those properties. If in fact we were in one of the worlds in which some individual other than A exists with all those properties, an accurate perception of that individual could be phenomenologically indistinguishable from the experience we are discussing. 7. See Alvin Plantinga, TheNature o f Necessity (Oxford, 1974), Chapters IV and V. 8. See note 5. 9. If events are identified in terms of their causes, then in the possible world we are considering it would not be B which A is said to precede but not cause. We would have to

29

10.

11.

12.

13.

14.

15.

16.

17. 18. 19. 20. 21. 22.

speak instead, as we did in the preceding paragraph, of "some event phenomenologically indistinguishable from B." However the point in the text could still be made. (For ease of exposition I wili assume throughout this section that events - and substances also - are not identified in terms of their causes.) We are here assuming that the cause exists prior to the effect. Similar arguments can be given for the case where cause and effect exist simultaneously, where for example A is a sustaining cause of B even though A never existed before B did. For ease of exposition, it will here be useful to ignore our earlier suggestion that no experience is phenomenologically of an individual. If that suggestion is correct, then no experience will be phenomenologically of the individual events we have labeled A and B. But we can safely ignore that point here since we are presently concerned only with whether any experience can be phenomenologically of something causing, as opposed to merely preceding, or co-existing with, something else. We are not concerned with whether some particular individual is phenomenologicaUy presentable, but with whether causation is. And in any case the point in the text can still be made without pretending that experiences can be phenomenologically of individual events. Here it might be objected that some version of the psycho-physical identity theory is surely correct, and so some perceivable physiological state might well be identified as my belief about Nixon. I believe that this objection is mistaken. But in any case the objection cannot be made when considering the beliefs of a presumably non-material being such as God. Since our concern is with alleged experiences of God we can therefore avoid the objection simply by restricting the point about beliefs to non-material beings. Instead of locating the causal relation as holding between the agent and the expression of belief, it may be more accurate to think of it holding between the belief itself and its expression. This will not affect the point being made in the text. For the difficulty in supposing that causal efficacy is phenomenologically presentable - whether the cause be the agent or the belief - remains. Moreover ff we did talk of an experience being phenomenologically of a belief causing its expression, we would involve ourselves in the additional difficulty of supposing that a belief itself is phenomenologicaUy presentable. If individuals are not phenomenologically presentable, there may be an additional problem about the suggestion that an experience might be phenomenologicaUy of a being who sustains the heavens and the earth. It should be noted that none of the arguments presented here depends on alleged difficulties with aspects of infinity that might be involved in the attributes, "being all-powerful, all-knowing and all-good." It is not being suggested (though it may be true) that theistic experiences are somehow suspect because they would have to be experiences which are phenomenologicaily of something which is inf'mite in one way or another. Even if 'God' were an abbreviation for the description, 'the powerful, knowing and good creator of the heavens and the earth' - with no "omni"-attdbutes mentioned - the arguments in the text would not be affected. The exchange between Pike and Smart (Paul Schmidt also participated in the symposium) can be found in Art, Mind and Religion, ed. by W.H. Capitan and D.D. Merrill (Pittsburgla, 1967), pp. 133-158. Further page references will appear in the text. William James, The Varieties of Religious Experience (New York, 1958), p. 63. Pike quotes this passage on page 146. Pike attributes the point in this paragraph to Richard Hensen. This kind of attack could be effectively used in connection with Pike's remarks about "big muscles" and "a kindly face." This is not very precise, but it is all Pike gives us. In any case it should be enough for present purposes. Thomas Hobbes, Leviathan, Chapter 32, in Works, ed. by W. Molesworth (London, 18391840). Perhaps they could be said to be part of the epistemic base one possesses within the dream.

30 23. I doubt that we have any understanding of the idea that mystical experiences have nonphenomenological content. Here is a possible way of looking at the situation. We can make sense of talk about the non-phenomenological content of a dream. In note 22 I suggested that such content can perhaps be regarded as comprising the epistemic base one possesses within that dream. We can then regard the dream as a single "experience" made up of: (a) the phenomenological content of the perception-like experiences occurring in the dream (its. phenomenological content); and (b) the epistemic base one possesses in the dream (its non-phenomenological content). But in a non-dream experience any relevant epistemic base is not thought of as something one possesses within that experience. It is thought of instead as something extrinsic to that experience and so not part of its content at all. Consequently the way in which dreams can be said to have nonphenomenological content provides no understanding of how non-dream experiences could. And I do not know of anything else that does. It might be that at the very moment that one has a mystical experience he also suddenly acquires the conviction that its object is such and such. But such a case is best described as one in which one has both a mystical experience and a sudden addition (perhaps even caused by the experience) to his (extrinsic) epistemic base. There is no reason to think one has an experience which includes "non-phenomenological" content.