THE

ARGUMENT FROM DESIGN - A PIECE ABDUCTIVE REASONING*

OF

In the Phaedrus Socrates tells us that "every discourse ought to be a living creature, having a body of its own," in which one begins with the head, a definition which comprehends the "scattered particulars in one idea," and then proceeds to divide the body "according to the natural formation, where the joint is, not breaking any part as a bad carver might." (246B-265D) 1 In general I appreciate Socrates' wisdom, particularly when he advises that, "The speaker should define his several notions and so make his meaning clear." (Phaedrus 265D ) Lest this discourse turn out to be a dead two-headed monster, rather than a living creature, let me, therefore, first attempt to explain what I mean by the phrases, "the argument from design" and "abductive reasoning," as used in my title. And let me take the more difficult phrase first-abductive reasoning. When I characterize the argument from design as a piece of abductive reasoning, I am obviously borrowing the term "abductive reasoning" from C.S. Peirce. Peirce was convinced that scientific reasoning could not be cast adequately in the traditional deduction vs. induction mold, and he felt that there was a need to add a third type of inference, abduction, to the traditional two forms of inference. Despite the fact that he probably considered abductive inference his own unique contribution to the logic of scientific inquiry, Peirce left the notion of abductive inference somewhat undeveloped and he did not devote nearly the time or effort to developing the rules of abduction that he did in developing the rules of deduction and induction. And there are perhaps reasons for this, for as Peirce * D e l i v e r e d as the P r e s i d e n t i a l A d d r e s s to Society for P h i l o s o p h y of R e l i g i o n a t N e w Orleans meeting, t974 . 1 All references a r e to H o u s e , I92o ) .

The Dialogues of Plato, trans. B. J o w e t t ( N e w Y o r k , R a n d o m

66

I N T E R N A T I O N A L J O U R N A L F O R P H I L O S O P H Y OF R E L I G I O N

himself tells us at one point, " T h e abductive suggestion comes to us like a flash. It is an act of insight, although of extremely fallible insight." (5.181) 2 A n d it is difficult to give the formal rules for a "flash of insight". There are, however, a few" things that appear quite clear about abductive reasoning in Peirce's writings. First, the conclusion of an abductive inference is an hypothesis. In introducing the notion of abductive reasoning, Peirce is trying to give some systematic account of w h a t we might call "an educated guess" or "intelligent conjecturing," or whatever it is we do when we come up with an hypothesis. A n d this is a significant part of scientific reasoning, for it is, as Peirce suggests, the only way in which knowledge is expanded. But an hypothesis is not merely an isolated guess. Even to count as an hypothesis, a guess must at least account for some of the facts. (5-189) It would be strange to call a guess an hypothesis if it has no supporting reasons whatever. Therefore, for Peirce, an hypothesis must be the conchasion of an inference, but a special kind of inference. T h e second thing that appears clear about abductive reasoning in Peirce's writings is t h a t abductive reasoning is distinguishable from inductive reasoning. A b d u c t i o n and induction, according to Peirce, do have a c o m m o n feature; " b o t h lead to the acceptance of an hypothesis because observed facts are such as would necessarily or probably result as consequences of that hypothesis." (7.218) But given this c o m m o n feature, they are at opposite poles, they are the reverse of each other: " A b d u c t i o n makes its start from the facts, without at the outset having a n y theory in view, t h o u g h it is motivated by the feeling that a theory is needed to explain the facts. I n d u c t i o n makes its start from a hypothesis which seems to recomm e n d itself, w i t h o u t at the outset having a n y particular facts in view, though it feels the need of facts to support the theory." (7.128) As a result of this different direction and motivation, we would expect two different general forms of inference. All inductive inference, Peirce tells us, " m a y be regarded as the inference that t h r o u g h o u t a whole class a ratio will have the same value that it has in a r a n d o m sample of t h a t class." (6.526) Thus, in an inductive

2 .~lI references are to C.S. Peirce, Collected Papers Vols. I - V I , eds. Charles H a r t s h o r n a n d Paul Weiss, Vols. V I I - V I I I , ed. A r t h u r ,~r. Burks (Cambridge, H a r v a r d University Press, ~966).

A P I E C E OF A B D U C T I V E R E A S O N I N G

67

inference, we conclude "that facts similar to observed facts are true in cases not examined." (2.636) On the other hand, in an abductive inference, an hypothesis is suggested as an explanation of a certain set of observed facts. Here we conclude the possible existence of a fact, namely, the hypothesized fact which is quite different from anything observed. (2.636) The third thing that appears clear about abductive reasoning in Peirce's writings is that an hypothesis can be evaluated in two different ways. An hypothesis may be evaluated in terms of its being a good or a bad hypothesis; that is, one which is or is not worthy of consideration. This type of evaluation has nothing to do with its truth or falsity. A good hypothesis might very well be false. In fact, at one point Peirce anticipates Carl Popper by suggesting that "the best hypotheses, in the sense of the one most recommending itself to the inquirer, is the one which can be most readily refuted if it is false." (i. 12o) This type of evaluation, evaluation in terms of good or bad, has to do only with the selection out of the many possible hypotheses suggesting themselves, those worthy of serious consideration. Here Peirce suggests two criteria, clarity and simplicity. An hypothesis is clear to the degree that it suggests definite testable consequences. (I. ~2o) An hypothesis is the more simple one, if it is the "more facile and natural, the one that instinct suggests." (6.477) The other type of evaluation of an hypothesis concerns its truth or falsity and this evaluation is usually given in terms of its probability. And this, according to Peirce, is a problem for induction. The fourth thing about abductive reasoning that appears relatively clear in Peirce's writings is the general form of abductive inference. According to Peirce tile general form of abductive inference is the following: A. i) The surprising fact, C, is observed. 2) But if A were true, then C would be a matter of course. 3) Hence, there is good reason to suspect A is true. (5.I89) There might appear to be some discrepancy between this general form of abductive inference and what Peirce calls "the form of perceptual abduction" and also what might be called the form of a statistical hypothesis. These latter two types of inference have essentially the same form, so we need only consider one. Let's consider

68

INTERNATIONAL JOURNAL FOR PHILOSOPHY OF RELIGION

the form of perceptual abduction, which Peirce characterizes in the following way.

B. I) A well-recognized kind of object M has for its ordinary predicates p1, p2, p3, etc. 2) The suggesting object S, has the same predicates, p1, p2, pa, etc. 3) Hence, S is a kind of M. (8.64) In order to show that this discrepancy is more apparent than real, we can easily rewrite the form of perceptual abduction in this way: C. I) The object S is observed to have the properties P~, p2, p~, etc. 2) I f " S is a kind of M" is true, then S would have the properties p1, p.,, p3, etc. as a matter of course. 3) Hence, there is good reason to suspect "S is a kind of M" is true. Thus, Form C is just a special case of Form A. The same translation can easily be performed for Peirce's form of a statistical hypothesis. We have now succeeded in giving one head to our discourse, not in the form of a Socratic definition, but in terms of the general form of abductive inference: that is form A, which like a definition comprehends the many "scattered particulars in one idea." Let us now attempt to put the other head on our discourse, by explicating what we mean by "the argument from design." One of the more classical instances of the argument from design is in the opening pages of William Paley's Natural Theology. Paley, for example, writes: In crossing a heath, suppose I pitched my foot against a

stone, and were asked how the stone came to be there, I might possibly answer, that for anything I knew to the contrary, it might have lain there for e v e r . . . But suppose I found a watch upon the heath, and it should be inquired how the watch happened to be in that place, I should hardly think of the answer which I had before given - that, for anything I knew, the watch might have always been there...~ 3 William Paley, Natural Theology (New York, American Tract Society, N.d.), p. 9.

A P I E C E OF A B D U C T I V E R E A S O N I N G

69

Then Paley raises the question, "Yet, why should not this answer serve for the watch, as well as the stone?" And the answer, he takes to be obvious: "When we come to inspect the watch, we perceive (what we could not discover in the stone) that its several parts are framed and put together for a purpose, e.g. that they are so formed and adjusted as to produce motion, and that motion so regulated as to point out the hour of the d a y . . . " ~ So, Paley concludes, "This mechanism being observed, the inference, we think, is inevitable, that the watch must have had a maker. ''5 But, Paley goes on to observe, "Every indication of contrivance, every manifestation of design, which existed in the watch, exists in the works of nature; with the difference, on the side of nature, of being greater and more, and that in a degree which exceeds all computation."G Thus we are led to conclude that the world had a maker. At the heart of this argument lies an analogy, which might be expressed in the following proportion: the watch = the world a watchmaker = a world maker The seeing of this analogy might be called, following Peirce's suggestion, the religious abductive insight. And using the term, "designer" rather than "maker", we can use Peirce's form of perceptual abduction to cast Paley's argument into the form of an inference as follows: D. I) A well-recognized kind of object requiring a designer (i.e. the watch) has for its ordinary properties, such things as its several parts are framed and put together for a purpose, etc. 2) The suggesting object, the world, has the same properties, etc. 3) Hence the world is a kind of object requiring a designer. If we abstract the form of Paley's argument from the particular well-recognized object, the watch, and the particular properties which he mentions, and at the same time rewrite the inference, as before, in the general form of abductive inference, we will have what

Ibid., p. 9. 5 Ibid., p. Io. Ibid., p. 20.

7~

I N T E R N A T I O N A L J O U R N A L FOR P H I L O S O P H Y OF R E L I G I O N

we might call the form of the argument from design. It would take the following form: E. I) The world is observed to have the properties, p,1p,2p, a etc. 2) If "The world is the kind of object requiring a designer" is true, then the world would have the properties, pl, p~, pa, etc. as a matter of course. 3) Hence, there is good reason to suspect "The world is the kind of object requiring a designer" is true. If the foregoing analysis is justifiable, then we have suceeded in producing the second head for our discourse. Although, again, it is not cast in the form of a Socratic definition, but in the fbrm of a general form for any argument from design, which does like a definition comprehend the many "'scattered particulars in one idea." In this case, however, we have gone a step further, for we have, at the same time, demonstrated that the argument from design is an instance of abductive reasoning. At this point we are justified in raising this question: What are the advantages in viewing the argument from design as a piece of abductive reasoning? Even though our foregoing analysis may be justified, unless we can find some distinct advantage in so treating it, our discourse becomes a somewhat trivial logical exercise. Although I do not wish to belittle logical exercises, [ would suggest that their importance lies in the significance of their application. In attempting to point out some of the advantages of viewing the argument from design as a piece of abductive reasoning I want to treat the problem against the background of what I consider to be the greatest attempt in philosphic literature to treat seriously the argument from design; namely, David Hume's Dialogues Concerning

Natural Religion. The first advantage in treating the argument from design as a piece of abductive reasoning that I would like to mention is this: To treat the argument from design as an abductive inference clearly distinguishes this argument from another argument which contains an extremely weak, if not unsound, inductive inference. In inductive inference, as we noted earlier, one infers, on the basis of an observed sample of the instances of a kind, something concerning all instances of this same kind. We can illustrate this form of inference fi'om Paley's own argument. When he raises the question, "Yet, why should not

A PIECE OF ABDUCTIVE REASONING

71

this answer serve for the watch, as well as the stone?" The explanation is this: He has observed a sample and in that sample he has observed that watches always have makers and stones do not, and on the basis of this sample he has inductively inferred that all watches have makers while stones do not. Now since the particular object found on the heath was a watch and not a stone, by a simple deductive inference, Paley concludes that this particular watch had a maker. Thus, in this argument there is an inductive inference based on a sample and with a universal conclusion, "All watches have makers," and a deductive inference, based on that inductive conclusion with the following conclusion, "This particular watch has a maker." Philo appears to be pointing out in the Dialogues that in order to produce this type of inductive-deductive argument concerning the world, either you must have observed a sample, consisting of m a n y worlds, and observed that all the worlds in that sample have designners, or you must assume that the world is the same kind of thing as a watch and its designer the same kind of thing as a watch designer. For, as Philo says, in speaking of just reasoners, "Unless the cases be exactly similar, they (the just reasoners) repose no perfect confidence in applying their past observation to any particular phenomenon". 7 Let us consider the first alternative. If we have observed many worlds and have observed that they all have designers, then we are entitled to inductively infer that all worlds have designers. Then by simple instantiation we can deductively infer from that inductive conclusion that this particular world has a designer. But this appears to be a most unlikely procedure, for, as Philo challenges Cleanthes : "Have worlds ever been formed under your eye, and have you had leisure to observe the whole progress of the phenomenon, from the first appearance of order to its final consummation? If you have, then cite your experience and deliver your theory. ''8 When, however, we examine Philo's second alternative, we find that this alternative presupposes exactly what is at stake in the argument from design. For here one must assume that the world is the same kind of thing as a watch, and that God is the same kind of

David Hume, Dialogues Concerning Natural Religion, ed. H e n r y D. Aiken (New" York, Hafner Publishing Co., i969) , p. 2~.

8 Ibid., p. 25.

72

I N T E R N A T I O N A L J O U R N A L FOR P H I L O S O P H Y OF RELIGION

thing as a watch designer, and consequently include them in the class of things under investigation. From an examination of a sample of this class, one could inductively infer that all the members of this class have designers. Then from that conclusion one could subsequently infer deductively, by instantiation, that this particular world has a designer. But this is possible only upon the assumption that the world is the kind of thing which requires a designer, and this is precisely what we have previously called the conclusion of the argument from design. We can easily see this difficulty if we contrast Philo's and J o h n Stuart Mill's reactions to the argument from design. Philo, on the one hand, is " s c a n d a l i z e d . . . with this resemblance which is asserted between the deity and h u m a n creatures, and must conceive it to imply such a degradation of the Supreme Being as no sound theist could endure. ''9 This may be feigned and questionable piety, but the point is clear: The cases are not exactly similar, therefore the second alternative fails. Mill, on the other hand, does grant that analogies are an "inferior kind of inductive evidence 'u~ but in the case of the argument from design, he tells us that the circumstances in which it is alleged that the world resembles the works of man are such that the "argument is not one of mere a n a l o g y . . . It is an inductive argument. ''n Thus, Philo tells us that the circumstances are such that the world designer, on the one hand, and the watch designer, on the other, are such that they cannot be considered instances of the same kind. While Mill, on the other hand, tells us that the circumstances are such that they can be considered instances of the same kind. But is not Mill's assumption here precisely what is at stake in the argument from design; that is, "The world is the kind of object requiring a designer."? To treat the argument from design as a piece of abductive reasoning clearly distinguishes the argument from design from the type of inductive-deductive argument which, in the case of the world and its designer has this apparent circularity. For if the argument from design is a piece of abductive reasoning, then we are not, as in an inductive argument, inferring "that facts

g Ibid., p. 20.

10 J o h n Stuart Mill, Theism, ed. Richard Taylor (New York, Liberal Arts Press, i957) , p. 28. 11 Ibid., p. 29.

A PIECE

OF ABDUCTIVE

REASONING

73

similar to observed facts are true in cases not examined," but rather we are inferring the existence of a fact quite different from anything observed. As a consequence, Philo's dilemma is inapplicable when the a r g u m e n t from design is taken to be an abductive inference. The second advantage in treating the a r g u m e n t from design as a piece of abductive reasoning that I would like to suggest is this: To treat the a r g u m e n t from design as a piece o f a b d u c t i v e reasoning gives an equally rational basis for different forms of the divine hypothesis other than that of traditional theism; and in so doing, it helps to give a rational account of the history of religion. It is one of Philo's roles in the Dialogues to point out the fact that there are alternative hypotheses. And it is this fact that there are alternative inferences, and consequently, alternative hypotheses, each perhaps equally justified abductively, that contributes to Philo's own skepticism. Philo, for example, is quick to point out that one could just as well infer that the world had m a n y designers: A great n u m b e r of men join in building a house or ship, in rearing a city, in framing a c o m m o n w e a l t h ; w h y not several deities combine in contriving and framing a world? . . . if foolish, such vicious creatures as m a n can yet often unite in framing and executing one plan, how m u c h more those deities or demons, w h o m we m a y suppose several degrees more perfect. 1~ Given our c o n t e m p o r a r y culture of mass production, this hypothesis of m a n y deities might be the hypothesis which would more readily come to mind today. And Paley, himself, even recognized this, for he at one point writes, when speaking of the watch found on the heath, "there must have e x i s t e d . . , an artificer or artificers who formed i t . . . , , ~ 3 But Philo, no doubt with a sense of humor, goes even further. W h e n he considers merely the characteristics of design in the world a host of alternative hypotheses come to mind: This w o r l d . . , was only the first rude essay of some infant deity who afterwards a b a n d o n e d it, ashamed of his lame performance; it is the work only of some dependent, inferior deity, and is the object of derision to his superiors; it is the 1~ H u m e , p p . 3 9 - 4 o. 13 P a l e y , p . I o.

74

I N T E R N A T I O N A L J O U R N A L FOR P H I L O S O P H Y OF R E L I G I O N

production of old age and dotage in some superannuated deity, and ever since his death has run on at adventures, from the first impulse and active force which it received from him. 1~ This last hypothesis we have seen resurrected recently by the God-isdead theologians. Laying such contemporary theological allusions aside, the point I wish to make here is that if these alternatives and other hypotheses which have played a role in the history of religion can be considered as conclusions to abductive inferences, then there is no need to seek the source of religious beliefs concerning the existence of a deity or deities in the emotions, whether it be fear or the more immature emotions suggested by Freud. Perhaps the emotions, cultural circumstances, or what have you, may cause one hypothesis rather than another to suggest itself to the h u m a n mind. Yet the fact that the divine hypothesis, in one form or another, does suggest itself to the h u m a n mind appears to be quite natural and rational. Also, the history of religious beliefs about the deity or deities might be viewed as the formulation, evaluation, and reformulation of the divine hypothesis in the light of subsequent personal and communal experience. And this brings us to the third advantage of treating the argument from design as a piece of abductive reasoning. The third advantage which I would like to suggest is this: To treat the argument from design as a piece of abductive reasoning allows us to use the dual criteria for evaluating an hypothesis in evaluating any form of the divine hypothesis. In other words, we might evaluate any one form of the divine hypothesis in terms of whether or not it is a good hypothesis; that is, whether or not it is worthy of serious consideration. Some forms of the divine hypothesis might be ruled out on this score alone. For example, one is hard pressed to know how some forms of the divine hypothesis are to be tested. As such, they would be ruled out, in Peirce's terminology, as unclear. And no doubt this is the major difficulty with most forms of the divine hypothesis. But history bears testimony to the fact that the divine hypothesis, as such, is simple, in Peirce's sense of the word ; for few hypotheses have come more natural to the h u m a n mind.

14 H u m e , p. 4I.

A P I E C E OF A B D U C T I V E R E A S O N I N G

75

The evaluation of any form of the divine hypothesis, in terms of truth and falsity, however, depends upon the criterion of clarity; that is, how it is to be tested. But this, as I suggested earlier, was perhaps the major difficulty with most forms of the divine hypothesis. And this difficulty leads me to the fourth advantage of treating the argument from design as a piece of abductive reasoning. I would like to suggest that treating the argument from design as a piece of abductive reasoning gives us a fruitful and systematic way of treating the theistic form of the divine hypothesis, and consequently, this form of the argument from design, in its relationship to the other classical arguments for the existence of God. Too frequently, and almost universally in philosophy textbooks, the classical arguments for the existence of God are treated as isolated arguments, each lending some support, however weak, to one theistic conclusion. The analogy which comes to mind here is that the various isolated arguments support the theistic conclusion in much the same way as the various columns of a Greek temple support the architrave of a temple. There is also the tendency in philosophy textbooks to point out the weakness in each argument so that we are left with the following alternatives : Either we feel that despite the weaknesses of the various isolated arguments, together they have sufficient strength to support the conclusion; or, on the other hand, we feel that each argument is so weak that taken together they do not have sufficient strength to support the conclusion. I would like to suggest, however, first, that this analogy of columns supporting the architrave of a building is a faulty analogy for the arguments for the existence of God; and second if it were an appropriate analogy, four or five bad arguments for the existence of God are no stronger than one bad argument. Let us return, then, to a consideration of the argument from design as a piece ofabductive reasoning, and let us suppose that the conclusion, or hypothesis, under consideration is the theistic hypothesis, and let us raise the question of the testability of that hypothesis. It seems to me that it is precisely at this point that the ontological argument becomes relative to the argument from design - that is, if we take seriously the modal character of the classical ontological arguments. And if one does not take seriously the modal character, for example, St. Anselm's classical form of the ontological argument, then I think one is hard pressed to distinguish the logical form of his

76

INTERNATIONAL JOURNAL FOR PHILOSOPHY OF RELIGION

particular argument from the logical form of the various cosmological arguments. And when I speak of taking the modal nature of the ontological argument seriously, I have in mind what St. Anselm wrote Gaunilo that he had proved with reference to God; namely, "if such a being can be conceived to exist, necessarily it does exist. ''~5 In order for the ontological argument to be sound, the existence of God must be both possible and necessary - and it is these two conditions which I would like to suggest as criteria for deciding the truth value of the theistic hypothesis. Let me explain by offering an analogy: Suppose I proposed a geometric hypothesis, a possible theorem, in Euclidean geometry. How" would we decide its truth value? First, it must be a possible theorem; that is, it must be formulatable in Euclidean geometry. Second, it must be necessary; that is, it must be provable from the axioms of Euclidean geometry. W h a t I am suggesting is that the modal character of the ontological argument sets the same kind of criteria for the theistic hypothesis. And it is at this point that the various forms of the cosmological argument become relevant to the argument from design. Let me first state my thesis and then I shall attempt to explain it and justify it. My thesis is this: the various forms of the cosmological argument are attempts to demonstrate the possibility and necessity of the existence of God along the lines suggested by the analogy with the Euclidean hypothesis above. If we examine the logical structure, for example, of the first four of St. Thomas' Five Ways, we shall find that they all concern a set of entities ordered by some given relation. In the first way we have the relation "x moves y," and the set of entities comprising the field of that relation. In the second way we have the relation, "x is the efficient cause of y," and the field of that relation; in the third, "x is contingent upon y," and the field of that relation; and in the fourth, "x is better, or more noble than y," and the field of that relation. In each way we have a definite description specifying a unique individual within the general order created by the given relation. In the first way, we have the first unmoved mover ; in the second, the first efficient cause; in the third, the first which is contingent upon no other; and in the fourth, the best, or the most 15 St. Anselm, Proslogium; Afonologium; etc., ed. Sidney Norton Deane (La Salle, Ill., Open Court, I958), p. t54.

A PIECE OF ABDUCTIVE

REASONING

77

noble. In each case, the proof is an attempt to show that the general order created by the given relation necessitates the existence of the unique individual specified by the definite description. And in each case the way terminates with a statement something like this: "And this everyone understands to be God." In short, to specify a unique individual which can function as a referent for the name "God", within a general order of things created by a given relation and necessitated by that general order of things is to demonstrate "the possibility" and "the necessity" of the existence of such a being. It should be noted in this connection that the earlier steps in St. Anselm's own argument perform the same function as a cosmological argument. He has the relation, "x is greater than y," and the set of entities ordered by that relation. The definite description which specifies the unique individual, in this case, is "that than which nothing greater can be conceived." And this, he tells us, even the fool understands to be God. St. Anselm, likewise, attempts to show that the order generated by the given relation necessitates that this particular unique individual exists in fact as well as in the mind. Thus this particular segment of St. Anselm's argument functions analogously to a cosmological argument and has the same logical form. Thus, while the modal character of the ontological arguments gives us the criterion of testability for the theistic hypothesis, the cosmological argument suggests the method for testing the theistic hypothesis; that is, the text for the theistic hypothesis is the construction of a cosmological argument. The critical nature of the strategy which I have outlined here can best be seen against the background of the Dialogues. For it is the strategy which I have outlined for testing the theistic hypothesis that is explicitly denied in the Dialogues; not by Philo this time, but by Cleanthes. When Demea proposes a form of the cosmological argument, Cleanthes interjects what might be termed an anti-ontological argument in order to prove the impossibility of any cosmological argument. If Cleanthes' argument is accepted, then the theistic hypothesis remains an untested hypothesis, and perhaps even an untestable hypothesis. And Philo is then justified in drawing the conclusion that "the whole of natural t h e o l o g y . . , resolves itself into one simple, though somewhat ambiguous, at least undefined, proposition, That the cause or causes of order in the universe probably bear

78

INTERNATIONAL JOURNAL FOR PHILOSOPHY OF RELIGION

some remote analog), to human intelligence." 16 In short, we are left with nothing more t h a n what we earlier called the religious a b d u c t i v e insight. B e y o n d that we c a n n o t go. But we need to note that the anti-ontological a r g u m e n t offered by Cleanthes rests solely u p o n one very critical premise: " W h a t e v e r we conceive as existent, we can conceive as non-existent. ''17 This premise merely asserts t h a t the existence of no being can be both possible and necessary. A n d yet this is precisely w h a t the m o d a l c h a r a c t e r of the ontological argum e n t requires as criteria for testing the theistic hypothesis; that is, that the being be both possible and necessary. D e m e a ' s religious piety caused him to be scandalized by the highly tentative n a t u r e of the conclusion of the a r g u m e n t from design and it u r g e d him to seek a m o r e "simple and sublime" proof. I do not know w h e t h e r or not this is an exemplification of true religious piety. But this I do know': O u r logical sense should cause us to be scandalized by the u n s u p p o r t e d status of Cleanthes' critical premise, " W h a t e v e r we can conceive as existent, we can conceive as non-existent," and it should urge us to d e m a n d some form ofjustification for this critical premise. BOWMAN g . C L A R K E

University of Georgia

a6 Hume, p. 94. 17 Hume, p. 58.