THE INEFFABILITY THEME

If one reads a bit in the works which provide raw materials for philosophy of religion, or much in philosophy of religion itself, one finds suggestions to the effect that human reason is inherently limited and that this is in some way religiously important. The issue is a complex one, partly because of the variety of limitations suggested, and partly because of the variety of reasons for which the suggestions are made. In what follows, I will roughly characterize, and then discuss, one sort of approach to the issue, one line of reasoning which concludes that human reason is inherently limited in some identifiable way. I will not pursue all the nuances of this line of reasoning, and while I will quote important figures (and some less important ones) I do so in this instance more to have pegs on which to hang characterizations than in terms of claiming exact exegesis of the essays from which I quote--this not because I do not respect or value careful historical work (in fact, I do), but because it seems to me appropriate to paint sometimes with a broad stroke. It is sometimes said that reason is a dangerous ally for religion; one might put the thesis of this paper by saying that lack of reason is a more dangerous ally still. So far as I can see, then, one can (and philosophers and theologians and mystics do) argue that there are interesting inherent limitations to human rational capacities, and that these limitations are religiously important (that certain central religious doctrines are called into question if these limitations are not present). They have developed this theme, in part, by claiming that human nature itself places inherent epistemic limitations on human beings. Knowing no defensible and philosophically interesting version of an essence of religion, I will limit myself here to considering this issue as it relates to monotheism, though analogous questions arise in other religious traditions. I begin by offering a single example of the sort of claim in question. While it is a claim about the nature of God, discussion of it will quickly lead to a consideration of the line of reasoning which begins, not with divine, but with human nature.

210

INTERNATIONAL JOURNAL FOR PHILOSOPHY OF RELIGION

Reformed theology holds that God can be known, but that it is impossible for man to have a knowledge of him that is exhaustive and perfect in every way. To have such a knowledge of God would be equivalent to comprehending him and this is entirely out of the question: It is impossible that the finite comprehend the infinite.~ It is not clear, to me at least, exactly what to make of these claims. Assuming that God exists, it is no doubt true that no human person knows all the truths about him that there are to know. But of course the same thing is true for such garden variety objects a~s my shoelaces and the door to my office. So this sort of lack of knowledge that is exhaustive and perfect in every way is not something that marks off divine from non-divine objects of knowledge. Perhaps the idea is that there are an infinite number of truths about God, but not an infinite number of truths about shoelaces and doors. I'm not sure that this is the case, but even if it is, it is also true that there are an infinite number of truths about the number one--truths, e.g., of the form "n is greater than, and a direct or indirect successor in the number series to, one" (where an indirect successor to one is a successor to a successor to one, or a successor to a successor of a successor of one, or ... and so on). So once again this will not mark out a distinction between divine and non-divine objects of knowledge. And even if it did, somehow the sheer cardinality of truths seems not to be of any particular theological or religious or philosophical interest that I can see. Thus it seems puzzling as to why such stress should be put on the point. Iohn Scotus Erigena in his On the Division of Nature, Book I, refers with approval to Gregory the Theologian who said that "no substance or essence, whether of a visible or invisible creature, can be comprehended as to what it is by the intellect or reason," 2 so I suppose they too would agree that God is not distinguished from a shoelace or a door by His being other than completely comprehensible and they not, though they ground this claim in a thesis about essence which I will not explore here. Erigena, however, has a theoretical context in I L. Berkhof, Systematic Theology (Grand Rapids: Eerdmann's Publishing Co., 1953), p. 30. A. Hymann and 1. Walsh, eds., Philosophy in The Middle Ages (Indianapolis: Hackett Publishing Co., 1973), p. 136.

THE I N E F F A B I L I T Y THEME

211

which knowledge of God and knowledge of other things is allegedly different in a more interesting way, and it is a simplified version of this context that I wish to consider. Erigena adopts the notion of a chain of being which rests on "the orders and differentiae of the natures of creatures, which beginning from the most pre-eminent intellectual power placed closest to God, descends as far as the extremity of the rational and irrational creature, that is, to speak more plainly, from the most sublime angel down to the extreme part of the rational and irrational soul, namely the nutritive and augmentative life. That general part of the soul which nourishes and augments the body is the lowest." 8 Without descending quite so far as Erigena does, one can suggest an interesting reason as to why God is incomprehensible. Take simply the distinction between Creator and creature (itself admittedly a rather complex affair) as constituting two "levels of being." Then add thereto an epistemic claim: (El) No being of level N can comprehend a being of level N + 1. This is just what Erigena seems to say when he is explaining how something can be such that It is and It is not can be true of it. For this reason...every order of the rational and intellectual creature is said to be and not to be. For it is insofar as it is known by superiors or by itself, and it is not insofar as it does not permit itself to be comprehended by inferiors? Whatever else the passage may mean, it seems at least to sanction (El). Of course, ever~ if one is willing to sanction as many as two levels of being, or grant that there are two kinds of being such that a, or the member of the one kind is in relevant ways greater than a, or the, member of the other kind, this will not of itself explain why one should accept (El). For that, it will be useful to have something like Russell's principle of acquaintance: ...every word that you understand must either have a nominal definition in terms of words having ostensive definitions, or must itself have an ostensive definition; and ostensive definitions, as

Loc. cit. 4 Loc. cit.

212

INTERNATIONAL IOURNAL FOR PHILOSOPHY OF RELIGION

appears from the process by which they are effected, are only possible in relation to events that have occurred to y o u ) Along with it, one will need a doctrine of acquaintance of this or a similar sort: Knowledge of things, which is of the kind we call knowledge by acquaintance, is essentially simpler than any knowledge of truth, and logically independent of knowledge of truths .... We shall say that we have acquaintance with anything of which we are directly aware, without the intermediary of any process of inference or any knowledge of truths. Thus in the presence of my table I am acquainted with the sense data that make up the appearance of my table its color, shape, hardness, smoothness, etc.; all these are things of which I am immediately conscious when I am seeing and touching my table. The particular shade of color I am seeing may have many things said about i t - - I may say that it is brown, that it is rather dark, and so on. But such statements, though they make me know truths about the color, do not make me know the color itself any better than I did before: so far as concerns knowledge of the color itself, as opposed to knowledge of truths about it, I know the color perfectly and completely when I see it, and no further knowledge of it itself is even theoreticaUy possible. 6 Then one can add to (El) something like: (E2) For any being of level N, it is not the case that it has direct acquaintance with any being of a level higher than N, and it is the case that any description that such a being can give of a being of a level higher than N must be given in terms of words which have ostensive definitions, that is, words that gain their meaning from the fact that they designate states or properties or events or processes or entities with which a being of level N can be acquaintained. ~ Bertrand Russell, Human Knowledge: Its Scope and Limits (New York: Simon and Schuster, 1948), p. 186. e Bertrand Russell, The Problems ot Philosophy (London: Oxford University Press, (London: Oxford University Press, 1912), p. 46, 47. r Understanding "it is not the case that it has any direct acquaintance with any being of any level higher then N" to preclude that a being of level N has direct acquaintance with any being, property of a being, event occurring to a being, state of a being, etc. which dwells at level N + 1.

THE INEFFABILITY THEME

215

And then perhaps one need but add something like: (E3) No one-or-more membered set S of items with which a being of level N can be acquainted is such that the terms which distributively or collectively designate S give to a being of level N the ability to describe the essence of a being of level N + 1. Whether or not (El-E3) suffice as accurate if partial accounts of how he got there or not, something like (E3) does seem to express Erigena's position, for he says that "names...are referred from the creature to the creator through a certain divine metaphor" and (although he says he will not discuss the matter in great detail) does affirm that "the ineffable (divine) nature can be properly signified by no word, no name, that is by any sensible sound, (and) by no thing signified...he is called superessential, more than truth, more than wisdom, and similar things." s Now of course the "great chain of being" perspective has received a good deal of critical attention, and it is nothing like obvious that it is correct, or even coherent. Later, I will discuss the notion of a perfection, which is (so far as I can see) central to that perspective. But for now I will not pursue this matter; let us stick to the assumption that there are two levels of being at least, Creator and creature, and forge ahead. Then, too, the notion of ostensive definition, and in particular the notion that one can begin to construct a language by simply baptising items or objects of experience, has been criticised, notably by Wittgenstein, roughly on the grounds that ostensive definition far from being a simple and unsophisticated matter, already presupposes a complex structure of language and conventions which make meaning-giving-pointing possible. But that too I shall not pursue. For present purposes, it suffices, I think, to note that the perspective (El-E3) characterize is not radical enough to yield its conclusions--particularly the conclusion that the divine nature or essence, if such there be, cannot be described by mere creatures. Consider, for example, two rather different passages that are relevant to this matter, one from Anselm and one from Hume. In his Reply to Gaunilo Anselm says:

s Hyman and Walsh, op. cit., p. 141, 143.

214

INTERNATIONAL IOURNAL FOR PHILOSOPHY OF RELIGION

..., you say that upon hearing "that-than-which-a-greater-cannotbe-thought" you cannot think of it as a real object known either generically or specifically or have it in your mind, on the grounds that you neither know the thing itself nor can you form an idea of it from other things similar to it. But obviously this is not so. For since everything that is less good is similar In so far as it is good to that which is more good, it Is evident to every rational mind that, mounting from the less good to the more good, we can from those things than which something greater can be thought conjecture a great deal about that-than-which-a-greatercannot-be-thought. Who, for example, cannot think of this (even if he does not believe that what he thinks of actually exists) namely, that if something that has a beginning and end is good, that which, although it has had a beginning, does not, however, have an end, is much better? And just as this latter is better than the former, so also that which has neither beginning nor end is better again than this, even if it passes always from the past through the present to the future. Again, whether something of this kind actually exists or not, that which does not lack anything at all, nor is forced to change or move, is very much better still. Cannot this be thought? Or can we think of something greater than this? Or is not this precisely to form an idea of that-thanwhich-a-greater-cannot-be-thought from those things than which a greater can be thought? There is, then, a way by which one can form an idea of "that-than.which-a-greater-cannot-be-thought." In this way, therefore, the Fool who does not accept the sacred authority (of Revelation) can easily be refuted if he denies that he can form an idea from other things of "'that-than-which-agreater-cannot-be-thought". In Section II, paragraph 14 of the Enquiry Concerning Human

Understanding Hume writes: when we analyze our thoughts or ideas, however compounded or sublime, we always find that they resolve themselves into such simple ideas as were copied from a precedent feeling or sentiment. Even those ideas which, at first view, seem the most wide of this origin are found, upon a nearer scrutiny, to be derived from it. The idea of God, as meaning an infinitely intelligent, wise, and good Being, arises from reflecting on the operations of our 9 Ibid., p. 160.

THE I N E F F A B I L I T Y THEME

215

own mind, and augmenting without limit those qualities of goodness and wisdom?~ Given this scope for "augmenting" our ideas, no one need complain that even a Human theory of the origin of our ideas unduly restricts their ultimate scope. The problem for the line of reasoning rehearsed above, then, seems to be this: even given a theory of meaning for which all our descriptive or nonsyncategorematic terms designate objects of sensory or introspective experience, or an abstractionist, empiricist theory of meaning, there is no guarantee that we cannot "augment" or sophisticate those concepts in unforeseen ways so that they apply far beyon0 the range of all possible sensory experience. An Aristotle may tell us that "to gain light on things imperceptible, we must use the evidence of perceptible things" (a remark he makes in Book Two of the Nichomachean Ethics), but he and his followers are able to discourse of essences and powers and unmoved movers. To put the difficulty in modern dress, it was for a time a familiar move in philosophy to painstakingly analyse the real or imagined circumstances under which a word is learned and take that setting as its home base or paradigm case of application (by itself, an appropriate enough move), and then to claim that any meaningful use of the word must be very similar to that on whose occasion it was learned--a tactic that ignored the fact that (as Julius Weinberg used to remark) a person's having learned a language was best tested by his or her ability to apply its terms in circumstances very different from those in which the user learned them. Once a person has even modest conceptual furnishings, there is no telling what he or she may make of them. Suppose then, it is true that: (E2) For any being of level N, it is not the case that it has direct acquaintance with any being of a level higher than N, and it is the case that any description that such a being can give of a level higher than N must be given in terms of words which have ostensive definitions; that is, words that gain their meaning from the fact that they designate states or properties or to Edited by L.A. Selby-Bigge (London: Oxford U. Press, 1962 edition), p. 19. Originally published in 1777.

216

INTERNATIONAL IOURNAL FOR PHILOSOPHY OF RELIGION

events or processes or entities with which a being of level N can be acquainted. Nonetheless, so long as an N-type being is sentient, a being of level N in particular, a human being named, say, Ralph--can note that he knows that 2 and 2 are four but is unsure as to whether the Republic is an early, or a middle, or a late Platonic dialogue. He is able to note that he knows some things, and does not know others. So long as the concept of omniscience is not incoherent, he may be able to move from the concept of knowing some things but not others to that of knowing all there is to know. In a similar fashion, he may be able to move from the notion of being able to do some things but not others, which does apply to him, to the notion of being able to do anything not involving a contradiction, which does not. He may, as Hume has it, reflect "on the operations of his own mind, and augment without limit." Or, following Anselm's lead, he may move from the notion of something which has a beginning and an end to notions of things which have no end but a beginning, and things which have no beginning but an end, and things which have neither beginning nor end, but are everlasting. If it is coherent, he may even be able to frame the notion of a being that exists but is eternal, having no temporal properties at all. Nothing in (E2) prevents this. Nothing in (El) prevents this either, for (El) simply precludes Ralph from comprehending (i.e., knowing everything about) anything at level N + 1. But, as we have seen, he is equally precluded from comprehending anything at level N. Apparently, then, something more is needed besides (El) and (E2) in order to get (E3). It would suffice to add: (E4) No being at level N can have any concepts that apply to any being at level N q- 1. But a little reflection on (E4) will show that it is something less than obviously true, and indeed something no theist can cheerfully embrace. Obviously, (E4) is intended to be true of beings at level N + 1 : it says that no N -k 1 being is conceptually accessible to any being at level N. Ralph, then, as our paradigmatic N-level being, cannot know that (E4) is true, or even so much as entertain (E4), for in order to do so he would have to know a truth about N § 1 level beings. But to know a truth about some entity E is just to know that some propo-

THE INEFFABILITY THEME

217

sition P is true of E. And that, in turn, or something much like it, is what it is to be in possession of some concept or other that applies to E. More carefully, on one plausible account, for someone to have a concept of x is for someone to know what is true of any x's there may be. I have the concept of a triangle if I know that if x is a triangle, then x is closed, plane, and three-sided. Differences in accounts of what should be included in human nature differ with respect to what is true if (or since) there are human beings. There will, of course, be different degrees of being clear about a concept, this being a function of different degrees of cognizance of what is true of any item that concept is to apply to. Plainly enough, on this account of what it is to have a concept, one cannot have a concept of x without knowing truths about x. The somewhat more familiar view that to have a concept is to be able to use a word properly has, I think, much the same consequences. Central to the relevant notion of 'Ralph's using word w properly' are the notions 'using w of x when x is w is true' and 'Ralph's knowing when it is true that w applies to x'. A purely behavioristic account of what it is to use a word properly of course will want to translate out at least reference to knowledge, but the great difficulties of behavioristic accounts of mentalistic terms are too well known to require discussion here. Or consider the account of concepts which says that for a person to have a concept C is for that person to be able to coherently complete a schema of the form '... is C' if C is a predicate concept or a schema of the form 'A C-item is ...' if C is a subject concept. The details of how to distinguish subject from predicate concepts is notoriously complex and difficult, and is happily not our problem. If the distinction is made at all, the account in question will once again make having a concept a matter of knowing what can and cannot be coherently said (and so what can and cannot be true) of items to which a concept applies. These attempts to analyse what it is to have a concept make appeal to a connection between a person's having a concept and that person's knowing some truths. They differ in the details of how to state the connection. Each owes its core of plausibility to that connection,

218

INTERNATIONAL IOURNAL FOR PHILOSOPHY OF RELIGION

and I think it is clear that any account which does not make that connection central will be less plausible than those just canvassed. 11 It seems, then, that in order for Ralph (or, of course, any other N-level being) to know that (E4) is true, (E4) must be false, since then Ralph will have concepts applicable to N § 1-1evel beings after all. Nor will it be easy to repair (E4) so as to escape this difficulty. Consider, for example: (E4a) No being of level N can have any observation concepts that apply to any being at level N + 1. Here an observation concept is any concept C such that C applies to x entails x is observable (i.e., can be seen, heard, felt, smelled or tasted). But of course it is part of traditional theology that God has no observable properties, and thus that no observation concepts apply to Him. (E4a) is coherent, but not radical enough to support the ineffability theme. Perhaps matters will be furthered by noting some other formulations that, for obvious reasons, will not do, and follow them up by a formulation that escapes the defect of its predecessors. 1~ Consider, then: (E4b) No concept applies to a being of level N + 1. (E4c) No concepts other than existence-entailed ones apply to a being of level N § 1. (E4d) No concepts other than experience-entailed ones aoply to a being of level N + 1. Now (E4b) cannot be true of a being of level N § 1, for if it is then of course concepts (those requisite to (E4b)) are true of such beings, and hence (E4b) is false after all. In sum, (E4b) is self-defeating. An existence-entailed concept is any concept C such that x exists entails C applies to x. The concepts having properties and having only consistent properties are presumably such concepts. For no such concepts to apply to beings of level N + 1 is taaatamount to there being no such beings; the same goes, of course, for no concepts at all applying to n Cf. Russell's comment to the effect that "it would be rash to assume that h u m a n beings ever, in fact, have acquaintance with things without at the same time knowing some truth about them." B. Russell, Problems oJ Philosophy, p. 46. 1~ I have discussed the difficulties of these formulations more fully in "Some Varieties of Ineffability," International Journal ]or Philosophy ot Religion, Vol. 6, No. 3 (Fail, 1975), 13. 167-179.

THE INEFFABILITY THEME

219

beings of level N + 1. So (E4c) is, if true at all, only vacuously so-true, not because N + l-level beings are (at least for N-level beings) ineffable, but because N + 1-1evel beings do not exist. So neither (E4b) or (E4c) can be true of anything. Turning to E4d), we have the notion of an experience-entailed concept. This can, of course, be developed in more than one way; for example, along the lines of: (1) C is an experience-entailed concept if and only if C applies to x entails x can be experienced or else: (2) C is an experience-entailed concept if and only if C applies to x entails x has been experienced. Now the theistic tradition will of course not want to deny that a being of level N + 1--eg., God--cannot be experienced at all. It will wish to allow that God can be self-conscious, that a Beatific Vision or some other post-mortem experience of God is possible, and even that some numinous experiences (say, those of Moses, Isaiah, and Paul) have been what they seemed to their subjects to be, namely experiences of God. So the tradition will not wish to deny that experience-entailed concepts apply to the Deity. Hence it will reject (E4d). What is required, then, to further the program under review is some set of claims about human nature which will allow one to restrict the range of concepts humans can have while leading to note of the inelegancies of (E4a), (E4b), (E4c) and (E4d). Perhaps the doctrines regarding human nature and its epistemic furnishings provided by the following remarks will provide the requisite claims. Meister Eckhardt writes that: Whatever the soul does, it does through agents. It understands by means of intelligence. If it remembers, it does so by means of memory. If it is to love~ the will must be used and thus it acts always through agents and not within its own essence. Its results are achieved through an intermediary. The power of sight can be effectuated only through the eyes, for otherwise the soul has no means of vision. It is the same with the other senses. They are effected through intermediaries. In Being, however there is no action and, therefore, there is none in the soul's essence. The soul's agents, by which it acts,

220

INTERNATIONALJOURNALFOR PHILOSOPHYOF RELIGION are derived from the core of the soul. In that core is the central silence, the pure peace, and abode of the heavenly birth, the place for this event: the utterance of God's word. By nature the core of the soul is sensitive to nothing but the divine Being, unmediated. Here God enters the soul with all he has and not in part. He enters the soul through its core and nothing may touch that core except God himself. No creature enters it, for creatures must stay outside in God's agents, from whence the soul receives ideas, behind which it has withdrawn as if to take shelter. When the agents of the soul contact creatures, they take and make likeness of them and bear them back again into the self. It is by means of these ideas that the soul knows about external creatures. Creatures cannot approach the soul except in this way and the soul cannot get at creatures, except, on its own initiative, it first conceives ideas of them. Thus the soul gets at things by means of ideas and the idea is an entity created by the soul's agents. Be it a stone, or a rose, or a person, or whatever it is that is to be known, first an idea is taken, and then absorbed, and in this way the soul connects with the phenomenal world... When all the agents (of the soul) are withdrawn from action and ideation, then this word is spoken. Thus he said "Out of the silence, a secret word is spoken to me." The more you can withdraw the agents of your soul and forget things and the ideas you have received hitherto, the nearer you are to (hearing this word) and the more sensitive to it you will b e P

And W. T. Stace claims that: What are the fundamental characteristics or elements of our ordinary consciousness? We may think of it as being like a building with three floors. The ground floor consists of physical sensations--sights, sounds, smells, tastes, touch sensations, and organic sensations. The second floor consists of images, which we tend to think of as mental copies of sensations. The third floor is the level of the intellect, which is the faculty of concepts. On this floor we find abstract thinking and reasoning processes. This account of the mind may be open to cavil. Some philosophers think that colors, sounds, and so on, are properly called "sensations"; others that images are not "copies" of sensations. These is Meister Eckhardt: A Modern Translation,

Harper and Row, 1941), p. 96-99.

Tr. R.B. B1akney (New York:

THE INEFFABILITY THEME

221

fine points, however, need not seriously concern us. Our account is sufficiently clear to indicate what we are referring to when we speak of sensations, images, and concepts as being the fundamental elements of the cognitive aspects of our ordinary consciousness. Arising out of these basic cognitive elements and dependent upon them are emotions, desires, and volitions. In order to have a name for it we may call this whole structure-including sensations, images, concepts, and their attendant desires, emotions, and volitions---our sensory-intellectual consciousness. Now the mystical consciousness is quite different from this. It is not merely that it involves different kinds of sensation, thought, or feeling. We are told that some insects or animals can perceive ultraviolet color and infrared color; and that some animals can hear sounds which are inaudible to us; even that some creatures may have a sixth sense quite different from any of our five senses. These are all, no doubt, kinds of sensations different from any we have. But they are still sensations. And the mystical consciousness is destitute of any sensations at all. Nor does it contain any concepts or thoughts. It is not a sensory-intellectual consciousness at all. Accordingly, it cannot be described or analyzed in terms of any of the elements of the sensory-intellectual consciousness, with which it is wholly incommensurable? 4 Something like the following theses seem a not unfair presentation of some at least of the themes these and similar passages suggest: (1) Direct acquaintance is passive; the perceiver, as it were, contributes nothing to the experience; his or her conceptual and experiential baggage is laid aside for the moment and does not affect the experience. (2) Direct acquaintance is aconceptual; it is a matter of having experiences without applying concepts. (3) Whatever is conceptual is verbal or verbalizable, and conversely. So what we are directly acquainted with, we cannot describe, or even conceive. (4) Concepts are ideas, and so are representations of something else, and hence are intermediaries between the thinker and the object of thought; thus all conceptual knowledge involves inference. 14 W.T. Stace, The Teachings o] the Mystics (New York: New American Library, 1960), p. 12, 13.

222

INTERNATIONAL JOURNAL FOR PHILOSOPHY OF RELIGION

(5) In direct acquaintance, at least, we experience things as they are. Now some of these claims we have, in effect, discussed. (2) and (3), for example, will lead directly to the embarrassments of (E4a) throught (E4c). Nor is the representative theory of cognition expressed by (4) very plausible. It was, after all, the basis for 'the way of ideas' in, say, Descartes and Locke and it led directly, not away from, but right into, scepticism, as Berkeley nicely demonstrated. But claims (1) and (2), concerning the passivity, and consequent reliability, of direct acquaintance are more promising. Nor, as we shall see, are the other claims without their point. Perhaps I should say that I am by no means sure that Stace, or even Eckhardt, would wholly approve of my use of their remarks, and I certainly do not ascribe to them what is to follow. Nonetheless, their summary of a perspective they share with others (many of whom would look more kindly on what is to come) will be useful for our purposes. What (1) claims, in effect, is that in direct acquaintance, one is passive in the sense of doing no interpreting of what one encounters. One experiences, and one can describe what one experiences (so far as claim (1) alone is concerned), but one will not thereby be interpreting what one experiences. A difficulty of (1), of course, is that the passive/active distinction, or the description/interpretation distinction to which it is here tantamount, is so hard to make out clearly. Another is the fact pointed to by Broad's perceptive remarks to the effect that "a Roman Catholic mystic may have visions of the Virgin and the saints, whilst a Protestant mystic pretty certainly will not" l~----a remark, of course, that provides a pattern on which one can ring the changes by using examples of traditions more distinct than the varieties of Christian tradition. Nor, of course, is claim (2) without its problems. Perhaps the basic one can be put in terms of a dilemma. If no concepts are involved in direct acquaintance, their direct acquaintance involves no beliefs at all? 6 If beliefs, and so concepts, are involved then in any religiously committive case, it is possible that a mistake be made. 1~ C.D. Broad, Religion, Philosophy and Psychical Research tLondon: Routledge and Kegan Paul, 1953), p. 193. ~6 This is argued more fully in the present author's "Some Varieties of Ineffability," " O n Windowless Experiences," Christian Scholar's Review, Vol. 4, No. 4

THE I N E F F A B I L I T Y THEME

223

I think this can be seen briefly if we note three things. 17 The first is that the plausible cases of incorrigible knowledge (which does not involve necessary truths) are cases in which the private states of the subject are in question. If I cannot be wrong about whether or not I am in pain at time t, provided I am introspectively attentive at t, this is because (or at least is reflected in the fact that) S believes S is in pain at t entails S is in pain at t; if I cannot be wrong about whether I exist or not at t so long as I believe that I do, this is because (or at least is reflected in the fact that) S believes at t that S exists at t entails S exists at t. Generally, S's belief that P is true is incorrigible if and only if S believes that P entails P is true. The second is that when it is claimed that someone has incorrigible knowledge, not that something is the case, but of some non-propositional item, this is true only if there is some proposition P which is true of that item and which the person in question incorrigibly knows. Schematically, S has incorrigible knowledge o] (non-propositional) item 0 entails There is some proposition P such that P is true of 0 and S incorrigibly believes that P. The third is that cases which are religiously committive---cases in which there are experiences taken to be veridical and to confirm or provide the truth conditions of religious claims--are not merely matters of one's own private momentary states. The reason for this is that religious claims characteristically range over other things besides such states, if they range over private states at all. Thus religiously-relevant direct acquaintances will not be incorrigible acquaintances. Indeed, in principle and in fact, one can be very mistaken about what one directly or non-inferentially experiences. What would make any-one think otherwise? One directly relevant answer to this question comes straight from the remaining important claim of the passages quoted from Stace and Eckhardt, but so far ignored, namely: (6) There are two types of concepts, intellectual-sensory and spiritual, the former being abstracted from sensory and intro(1975), p. 311-318; and the critique of the varieties of (E4) above. See also Russell's remark, cited in f.n. 11; and George Mavrodes, Belief in God (New York: Random House, 1973). lr This is argued more fully in the present author's "Some Varieties of SelfAuthentication," Sophia (forthcoming), and "Religious Experience and Rational Appraisal," Religous Studies, Vol. 10, No. 2 (June, 1974), p. 172-187.

224

INTERNATIONAL IOURNAL FOR PHILOSOPHY OF RELIGION

spective experience and reflection on such experience and the the latter being obtained by having mystical experience. N~w putting things this way does take some liberties with the passages quoted above. I think, however, that the justificaion for putting (6) in the fashion in which it appears is not far to seek. Authors in the sort of tradition Eckhardt and Stace represent tend to deny that concepts apply to mystical experience because they have a particular view of what a concept is. Roughly, they view concepts as "intermediaries" between a knower and an item known (Eckhardt), as invented by us (Eckhardt), as connected to the way things s e e m (Eckhardt), as tied to images and perhaps copies of images (Stace), as linked to "ordinary" consciousness (Stace), and in fact as being applicable in a really adequate way only to physical objects and the properties of such objects. Philosophers who think of concepts as having their 'home base' in application to middle-size observable objects and tend to think of other applications as logically dependent on, if not actually metaphorical extensions of, such 'home base' uses agree with the Eckhardt-Stace line, even if they deny that there are any concepts of the sort (6) calls "spiritual." In ways that I shall not try to document here, radically empiricist theories of concepts seem to me to be shared by many of those who prize, and many of those who despise, talk of mystical experience. So it is not surprising that it is so often said that concepts do not apply to such experiences, both by those who go on to say how important such experiences are, and also by those who think that such experiences are at best pathological. The fact that such theories of concepts have been, and continue to be, held in spite of the fact that the very concepts requisite to the theories (e.g., of a concept, of a theory, or of meaning) belie the theories themselves is perhaps partially explained by two factors: that the 'syncategorematic' concepts of logic and mathematics are assumed, and whatever concepts one needs to frame and defend radically empiricist theories are assumed to be available from this bountiful source, and that the metaphor "having ... before the mind" allows one to fill in the blank by a description either of an image or of some propositional content. Still, those who prize 'non-ordinary' experiences notoriously say a good deal about them. This in itself suggests that they have in mind a theory of concepts quite consistent with that sketched earlier in this paper. Some among them make up special terms for concepts

THE INEFFABILITY THEME

225

(in that earlier sense of concepts) which apply to 'non-ordinary' experiences. Otto, who shared some version of a radically empiricist theory of concepts with Stace and Eckhardt, spoke of 'ideograms' which (rather than concepts per se) made religious language possible. 18 Eckhardt talks about "a secret word" and Stace about a "mystical consciousness.., wholly incommensurable" f r o m the ordinary. But given our critique of the varieties of (E4), it seems both inevitable and maximally charitable to parse this alleged incommensurability in terms of two distinct types of concepts rather than rendering the theory self-defeating, or even coining terms unnecessarily. The idea, then, is that concepts come in two varieties--sensory and spiritual---and that the latter are attainable by human beings only under certain special conditions. This two-sorts-of-concept line does take some account of claims (2) through (4). Insofar as sensory concepts are in question, mystical experiences are (at least for the perspective currently under review) a conceptual, nonverbalizable (in discourse framed for descriptions of rocks and rhinos), and non-inferential (where for an experience to be inferential is for it to be a direct experience of one thing (e.g., an image) from which the existence and properties of another thing (e.g., a physical object) is inferred). Nor need one espouse the claims that direct acquaintances are utterly uninterpretive and incorrigibly reliable in order for one to accept (6). Implicit in all of this, of course, is an analysis of human nature in which human beings have a sort of epistemic duality: they are able to see, hear, feel, smell, and taste on the one hand, and are able to experience God (for Otto) or a monistic Absolute (for Eckhardt) or an undifferentiated reality (for Stace). And apparently (at least for Stace and Eckhardt) such epistemic capacities are quite distinct (animals can have one, and angels the other, without having the human counter-part) and their knowledge-products are incommensurable. Putting (6) in the fashion we have gives promise of avoiding the embarrassment of having no answer to Descartes' pointed question that if one has no concepts applicable to God, then how can one say:

18 Rudolph Otto, The Idea o/ the Holy, John Harvey, tr. (New York: Oxford U. Press, 1958), p. 30. (First published in 1917.) Cf. the present author's Basic Issues in the Philosophy of Religion (Boston: Allyn and Bacon, 1971), Ch. 4, for a discussion of Otto.

226

INTERNATIONAL JOURNAL FOR PHILOSOPHY OF RELIGION

that God is infinite and incomprehensible, and that He cannot be represented by our imagination? How could he affirm that these attributes belonged to Him, and countless others which express His greatness to us, unless he had the idea of Him? It must be agreed, then, that we have the idea of God, and that we cannot fail to know what this idea is, nor what is meant by it; because without this we could not know anything at all about God. It would be no good saying that we believe that God exists, and that some attribute or perfection belongs to Him; this would say nothing because it would have no meaning to our mind. Nothing could be more impious or impertinent? * One thereby escapes the pointed critique of Berkhof when he writes that: If man were left absolutely in the dark respecting the being of God, it would be impossible for him to assume a religious attitude. There could be no reverence, no piety, no fear of God, no worshipful service.~ We asked a few paragraphs back what might make one deny that direct acquaintance with an item might be terribly misleading, and by now an answer is to hand. While the application of sensory concepts to an item of sensory acquaintance can indeed be inappropriate, the application of a spiritual concept to an item of mystical acquaintance (so long as one makes no inferences, at any rate) cannot. Unfortunately, another query arises: why can't an application of a spirritual concept be mistaken, even in direct-acquaintance cases? I know of no satisfactory answer. That spiritual concepts, or religious language games, or theology, be incommensurate with, or conceptually unrelated to, such mundane affairs as wars, committee meetings, and sibling quarrels has a perennial attraction. If religious claims can be irrelevant to matters of history and science, then no scientific or historical claim can have the inelegant consequence of falsifying a religious claim. Of course, incommensurability has its price. If God loves us is incommensurate with regard to moral claims, it provides no support for W e ought to lg A. Kenny, ed. and tr., Descartes' Philosophical Letters (London: Oxford or the Clarendon Press, 1970), p. 106. 80 L. Berkhof, op. cit., p. 30.

THE INEFFABILITY THEME

227

love one another. If God is in Christ and God called Abraham and God spoke to Mohammed are incommensurate with historical claims, then they could be true even if Jesus, Abraham, and Mohammed never so much as existed. If God created the world is incommensurable with respect to scientific and common sense matters such as the existence of order and the habitability of earth, God has nothing to do with the possibility of knowledge or the provision of daily bread. With incommensurability goes irrelevance. But I think this is not the deepest problem with (6) and the context that goes with it. That honor must go to the concept, not the consequences, of incommensurability. For if the theological claims noted above do not have the entailments we mentioned, or entailments much like them, their own meaning is opaque. It is not merely that such propositions as God is Lord of Nature and History or God is Maker o~ Heaven and Earth contain syntactical or syncategorematic terms such as 'is' and 'and'; one could so interpret (6) as to allow that there be a third sort of concept which underlies both sensory and spiritual ones, or which connects one concept of one type with another of the same type. Nor will the intent of (6) be affected if such underlying and/or connecting concepts 'externally' connect concepts of one alleged type with those of the other. Just as it is true that if Jones is fat, then Jones is either fat or else a prime number, so, for all a supporter of (6) need care, it may be true that if God is holy, then God is either holy or else bright red. The trouble will come only if concepts of the one sort are 'non-externally' related to concepts of the other sort. This way of putting the matter makes use of the notion of concepts being related 'externally' to one another, and it is not lucid what that amounts to. Perhaps we can define it as follows: where A and B are monadic predicative concepts, ~1 A and B are related non-externally if and only if either x is A entails x is B or conversely. Thus the concept of being red is related non-externally to the concepts of being extended, being shaped, and being visible-in-principle, but externally to the concepts of being holy, being prime, and being wise. There are at least two problems with this way of defining 'external'.

2~ "Monadic predicative concept" is of course short-hand for "concept of a monadic, or one-place, predicate."

228

INTERNATIONAL JOURNAL FOR PHILOSOPHY OF RELIGION

One is that if x is A entails x is not C, and C is also a monadic predicative concept, then it would seem as if the concepts being A and being C are non-artificially or non-externally related. Thus x is red does entail x is not a number. For it is, I suppose, a necessary truth that a number has no color, and if a proposition plus a necessary truth entail another proposition, then according to one standard rule of inference the first proposition by itself entails the second, n But this is easily remedied: A and B are externally related monadic predicative concepts if and only if none of the following entailments holds: (a) (b) (c) (d)

x x x x

is A is B is A is B

entails entails entails entails

x x It It

is B is A is false that x is B is false that x is A.

(And two concepts are related non-externally unless they are externally related, though I will hereafter ignore this feature of the definition.) The other objection can be put along these lines. Consider the concept being a painting and the concept being beautiful; x is red and x is a Painting, and x is red and x is beautiful, are related to one another in ways that satisfy (a) through (d). But x is a painting entails x has some color property or other, and being red is a determinate with respect to being a color property. Further, x is beautiful entails x has some observable property or other, and being red is a determinate with respect to being an observable property. So being red is not arbitrarily or externally related to certain concepts of which it is a determinate, and certain concepts of which it is a determinate are not arbitrarily or externally related to being a painting and being beautiful. Hence being red is not arbitraily or externally related t,,~ being a painting and being beautiful, though it is a consequence of the definition of 'external' under review that it is arbitrarily or externally related to these other concepts. This leads to the natural if cumbersome additions that: (e) There is no concept C such that x is A and x is C entail either x is B, or else entail It is false that x is B ~2 This objection, in its present form at least, does of course assume that "A number has no color" is a necessary truth--and so not nonsense and (perhaps) not synthetic a priori.

THE I N E F F A B I L I T Y THEME

229

(f) There is no concept C such that x is B and x is C either entail x is A, or else entail It is false that x is A. But even with all these qualifications, it seems to me clear that socalled spiritual concepts are non-externally related to so-called sensory, or intellectual-sensory, ones. Consider God is holy. Holiness is, for Otto at least, a concept one can have--or have in its full and rich version, at any rate--only if one has had a numinous experience; it is not a concept applicable to 'ordinary' experiences. At least, it seems as good a candidate for being such a concept as any. But for God to be holy is not merely a matter of God's being set aside from other things; holiness is (for Otto, and I think for the tradition he is describing) an intrinsic property of God, and not merely expressive of a relationship He has to other things. But holiness in this sense is closely related to being u n i q u e ( t h e r e does not merely happen to be just one God) and righteous ("God is of purer eyes than to behold inquiry"). Now no being can be holy in this sense without being sentient. But the concept of sentience is one that is not syncategorematic (it does not fall among the concepts of logic and mathematics, but is a semantic and not merely a syntactic notion) and is sensory. So at least one sensory concept is non-externally related to a spiritual concept; God is holy entails God is sentient. (Even a negative theologian will admit that God is holy entails It is not ]alse that God is sentient.) But this example can be multiplied again and again; consider the concepts being a spirit, being worship-worthy, being changeless, being a necessary being, being simple, and so on, much less the religiously thicker notions connected with creation, redemption, and final judgment. The supposed separation between sensory and spiritual concepts simply cannot be made. The result of all this is that even if (6) is true, what is really needed is not merely (6) but something like: (7) There are no (non-external) conceptual connections between intellectual-sensory concepts and spiritual concepts.

One could require that C too be a monadic predicative concept, but this requirement may be itself arbitrary, and in any case being beautiful and being a painting seem in fact to be monadic predicative concepts. (The notion of a "monadic predicative concept" is perhaps itself not very lucid, but that is another matter.) I have also assumed what seems to me quite justified, namely that being external is a transitive relation.

230

INTERNATIONAL IOURNAL FOR PHILOSOPHY OF RELIGION

In the case of all the candidates for spiritual concepts I can think of, (7) is simply false. And without (7), (6) will be of no help to the ineffability theme. We have now tried to develop the ineffability theme in two ways. For one, (E3)--the claim that no one-or-more membered set S of items with which a being of level N can be acquainted is such that the terms which distributively or collectively designate S give to a being of level N the ability to describe a being of level N + 1--was required. In order to get (E3), some one or more of the varieties of (E4) was necessary; but these varieties were one and all found wanting. For the other, (6) and (7) were necessary, and (7) seems flagrantly false. Of course the defender of the ineffability claim may well not feel daunted by all this. He or she may simply reply that (6) is not the right way to go, but insist that no concepts at all, or at any rate none that human beings can have, apply to God. But the consequence of none of the concepts we have applying to God is simply that God does not exist (or subsist, if one wants to claim that existing beings are necessarily spatio-temporal). (It is one thing to say that concepts which we do not have apply to God that is true, no doubt, of doors and shoelaces too---but quite another to say that only such concepts apply to Him.) I have failed, then, to find any version of the ineffability theme, based on claims about concepts and/or about human nature, which is coherent, much less cogent. Unless there are better versions of the theme, the consequence is not that ineffability is very mysterious, but rather that there is in fact no ineffability theme at all, but only confusion and nonsense. I am very much inclined to think that this is exactly the truth of the matter. Those who think otherwise are invited to state the theme coherently. There is one additional consideration I should mention though I do not know how much or little of a role it may have played in developing the perspectives characterized above. One can understand the word 'cow' without having to be one. One can understand what the words 'bereaved' or 'retired' or 'elected' or 'married' mean without being bereaved or retired or elected or married. But there is something, nonetheless, that you do not understand (I suppose) without having been there yourself. And it is surely true that however well one understood the word 'God' or the doctrine of the Trinity, one would not thereby know what it was like to "be there"----to be an omnicom-

THE INEFFABILITY THEME

231

petent Creator and Povidence; e.g., presumably even when St. Teresa claimed that: Our Lord made me comprehend in what way it is that One God can be in three Persons. He made me see it so clearly that I remained as extremely surprised as I was comforted..? 5 she did not suppose herself to have gained insight, if I may use the phrase, into the Trinitarian phenomenology. Each person sees the world differently, in one sense, from every other--that much (without relativistic or sceptical consequences) seems to me true. And this is especially true, if one of the persons in question is divine. But it is no part of my purpose to develop this theme here, nor do I suppose that it would blossom into a new, more cogent basis for the ineffability theme. Keith E. YANDELL University o/Wisconsin-Madison

Cited in William lames, The Varieties of Religious Experience (New York: New American Library, 1958 reprint), p. 320.