Dye, Hume on Religion, or the Case of the Missing Propensity

Hume on Religion, or the Case of the Missing Propensity
James Dye
Northern Illinois University

Hume's Natural History of Religion starts by separating the question of the justification of religious belief, "its foundation in reason," from the question of why religion exists as a pervasive cultural phenomenon, "its origin in human nature."¹ Basically, NHR takes as its task answering the latter question, as the Dialogues concerning Natural Religion tackles the former, although both works contain passages in which they poach on each other's respective territory. This paper examines Hume's methodology in NHR, first with respect to a couple of features of his general program of explanation and then with respect to specific explanatory factors to which he appeals. The latter are found to be inadequate for generating religious belief and an additional principle is proposed to remedy that deficiency.²

Religion and Rationality

Once embarked on the task of deriving religion from human nature, Hume goes far beyond that to which he is entitled should we grant him the initial distinction between the rational defense of a belief and its origin. He seems to be trying, as much as possible, to exclude reasoning from religion's origin altogether; and since humans are unquestionably reasoning animals, if not entirely rational ones, that move seems unjustified. This is all the more peculiar given what Hume seems to take as the essential feature of religion: "belief of invisible, intelligent power" (NHR, 309/21). Since he doesn't try to define 'religion,' it would be pointless to quarrel with this assumption at this point, although it would exclude some associations which are normally called 'religions' from actually being such. Suffice it to note that assigning a vanishingly small role for reasoning in the genesis of religion would be a more plausible move were religion primarily some characteristic feeling(s) or type of behavior rather than a belief, especially a belief about something not sensibly apparent and therefore, one might suppose, arrived at by some reasoning, however elementary.

Of course, religion, even as characterized by Hume, is certainly not the child of philosophical theology as that has been practiced since Plato; but he tries a bit too hard to paint all our remote ancestors in the same dreary colors he uses for "the vulgar" of his own age. Why could religion not begin in philosophy--crude, prehistoric, but philosophy nonetheless (or 'science' if one prefers to use that term for the attempt to satisfy our curiosity about natural events)? To begin thus would be contrary to human nature, since

The causes of such objects, as are quite familiar to us, never strike our attention or curiosity; and however extraordinary or surprising these objects in themselves, they are passed over, by the raw and ignorant multitude, without much examination or enquiry. (NHR, 311/24).

Hume does not make his view fully explicit, but it seems to go as follows. The wonder which initiates inquiry does not normally arise with respect to familiar things. The child's curiosity about everyday objects gradually fades as those objects become more and more familiar. Reactivating that childlike interest in ordinary objects and events after the faculties needed for advancing understanding have matured is not a natural development. It is artificial and must be cultivated; its production requires leisure and study. However, "a barbarous, necessitous animal . . . has no leisure to admire the regular face of nature" (NHR, 311/24). Consequently, archaic man has no curiosity whatsoever about the grand regularities of nature, and gets excited only about abnormal events, such as a "monstrous birth." All the regular events of nature seem "ordinary" to him and are no cause for wonder or religious feeling.

Ask him, whence that animal arose; he will tell you, from the copulation of its parents. And these, whence? From the copulation of theirs. A few removes satisfy his curiosity, and set the objects at such a distance, that he entirely loses sight of them. Imagine not, that he will so much as start the question, whence the first animal; much less, whence the whole system or united fabric of the universe arose. Or, if you start such a question to him, expect not, that he will employ his mind with any anxiety about a subject, so remote, so uninteresting, and which so much exceeds the bounds of his capacity. (NHR, 311/25)

Hume must have been fond of this argument, since he has Cleanthes make the same claim, with the same example, in Part III of the Dialogues. There the conclusion is that the "ignorant savage and barbarian" is simply stupid.³ This exclusion from archaic mentality of an impetus towards complete causal explanation seems to conflict with Hume's own claim in the Treatise that

Nothing is more curiously enquir'd after by the mind of man than the cause of every phenomenon: nor are we content with knowing the immediate causes, but push on our enquiries, till we arrive at the original and ultimate principle (T, Bk. 1, Pt. 4, §7, p. 266).

According to NHR, this persistent curiosity for ultimate causal explanation should have been located, not in "the mind of man" but 'the mind of modern man' or, perhaps more accurately, 'the philosophical mind.' Hume apparently regards early humanity as uniform in intelligence and curiosity; and that level of development is roughly the same as that of "the vulgar" of his own time. A passage from the Enquiry affirms a view similar to that of NHR, but applies to Hume's contemporaries rather than to prehistory or antiquity.

The generality of mankind never find any difficulty in accounting for the more common and familiar operations of nature; such as the descent of heavy bodies, the growth of plants, the generation of animals, or the nourishment of bodies by food: But suppose, that, in all these cases, they perceive the very force or energy of the cause, by which it is connected with its effect, and is for ever infallible in its operation. . . . It is only on the discovery of extraordinary phenomena, such as earthquakes, pestilence, and prodigies of any kind, that they find themselves at a loss to assign a proper cause, and to explain the manner, in which the effect is produced by it. It is usual for men, in such difficulties, to have recourse to some invisible intelligent principle, as the immediate cause of that event, which surprises them, and which, they think, cannot be accounted for from the common powers of nature. But philosophers, who carry their scrutiny a little farther, immediately perceive, that, even in the most familiar events, the energy of the cause is as unintelligible as in the most unusual, and that we only learn by experience the frequent Conjunction of objects, without being ever able to comprehend anything like Connexion between them. Here then, many philosophers think themselves obliged by reason to have recourse, on all occasions, to the same principle, which the vulgar never appeal to but in cases, that appear miraculous and supernatural. (E, §7, Pt. 1, p. 69)

Here failure to inquire after ultimate causes is not just a matter of a lack of interest, imagination, and sufficient leisure as NHR and the Dialogues have it, but depends on the absence of one particular philosophical insight: that causes are never immediately perceived. This circumstance widely pertains to modern humans, but presumably universally pertained to those who lived before this sceptical discovery.

Such disparagement of the ancients and the uneducated was a common enough viewpoint among Enlightenment thinkers, confident as many of them were of their superiority to all that had gone before; but it will not withstand much scrutiny. Hume is not on the cutting edge of the then growing recognition of the intimate connection between the practical and the theoretical, as exemplified in Diderot's Encyclopédie. Although the tie is arguably intrinsic in that "experimental philosophy" to which Hume, despite his scepticism, wished to ally himself,⁴ he manages to keep a rather Aristotelian view of inquiry into the general principles of things. Philosophical or scientific inquiry is far removed from the practical concerns of the hoi polloi. The issues it addresses are abstruse, the province of an intellectual elite of leisured gentlemen, and the vulgar are neither interested in the questions nor capable of appreciating good answers to them. The deities of our polytheistic predecessors supposedly arose from their concerns with the events of life, and for Hume they invariably scrutinized those events as "various and contrary," through "eyes still more disordered and astonished" (NHR, 315/28). His instances of the sorts of gods they imagined are drawn largely from the Roman pantheon, or from the Greek referred to by their Roman names. He does not trouble himself to deliberate about the meaning of the myths about these gods. That they might have some important function of cosmic significance is not even considered. The myths are but "fables" which seem to fall into one or more of three categories. They may be (1) careless and perverted accounts of remote historical events (§1, 312/25; §5, 327/39), in effect, bad history becoming religiously significant as it gains in marvelousness by the accretion of errors and exaggerated additions. They may be (2) "light, easy, and familiar," more playful than useful, narrative poems, entertaining accompaniments to a kind of "true poetical religion; if it had not rather too much levity for the graver kinds of poetry" (§12, 349/61). "Who could forbear smiling, when he thought of the loves of MARS and VENUS, or the amorous frolics of JUPITER and PAN?" (loc. cit.) Or, they may be (3) allegorical fictions, stories which have been invented to complement the character which a god of a given domain might be supposed to have. He acknowledges that the myths contain elements which, if taken allegorically, seem to be inconsistent and puzzling ("That Fear and Terror are the sons of MARS is just; but why by VENUS? That Harmony is the daughter of VENUS is regular; but why by MARS? That Sleep is the brother of Death is suitable; but why describe him as enamoured of one of the Graces?" [§5, 326/39]). Do such incongruities cause him to question the allegorical interpretation of these myths (as one should have, since the allegorical interpretation arises quite late in antiquity)? Not at all. Do they provide an occasion for trying to discern some unifying idea or theme which would resolve the puzzle, as one would certainly try to do were the inconsistency in an ancient philosophical treatise? Not at all. Since Hume has already decided that myths are "productions of ignorance and superstition," he simply classifies any paradoxical elements in them as "gross and palpable" mistakes (loc. cit.).

No intelligentsia among ancients

religion being distinctively human

This is not terribly coherent. Hume's archaic humans have no leisure for cosmogonical speculation but they have time for entertaining poetry. Although they have no desire to trace causes more than "a few removes," they apparently have an interest in very remote history, even if their historical practice is sloppy. They are preoccupied with the struggle to survive but not practical enough to attend to features of their experience which would assist them. They would indeed be stupid if they were not concerned with the great regularities (the migration of game, the flooding of rivers, the seasonal variations of weather, etc.) upon which their continued existence depended. To be more curious about such items as monstrous births than about the annual return of the sun or the signs of an impending storm borders on suicidal folly. Happily for us, our ancestors did not follow Hume's script. It would be unreasonable to expect an eighteenth-century philosopher, given that scientific study of the ancient myths did not then exist, to have even suspected that a case could be made for there being good science underlying the myths, as Giorgio de Santillana and Hertha von Dechend have done in their fascinating argument for celestial regularities being the primary concern of archaic science and for myths encoding rather detailed information about astronomical events.⁵ But Hume's depiction of the ancients is more of a caricature than it need have been. He totally ignores the predominance of agricultural and sky gods, the ancients' well-known interest in astronomy and astrology, and the fact that creation stories are among the oldest, not the latest, surviving myths.

So, at the very least, Hume's distinction between being motivated by "contemplation of the works of nature" and "a concern with regard to the events of life, and . . . the incessant hopes and fears, which actuate the human mind" (NHR, §2, 315/27) is artificial and unrealistic. He does mention that the ancients deified "sun, moon, and stars" (§5, 325/38) but does not trouble himself to wonder if this deification might have something to do with the regular functional utility of these celestial powers for human life. He observes that the ancients made even the gods subject to fate or destiny (§4, 324/37) but does not broach the question whether fate, then, might not be closer to the theist's god than any of those many other beings denominated 'gods.' On the same page he classifies the "mythology of the heathens" as like the "ancient EUROPEAN system of spiritual beings, excluding God and angels, and leaving only fairies and sprights." Again, they apparently have time for amusing caprice but not for serious inquiry. In sum, Hume is certainly correct in maintaining that archaic religion could not arise from "the pure love of truth"; but he gratuitously assumes that there is no utility in "enquiries concerning the frame of nature" (§2, 315/28). NHR begins with what seems to be an innocent enough distinction: there is a difference between the rational justification of religious belief and an explanation of the origin of religion. However, as Hume pursues the latter task, spinning his likely story about how religion arises and develops, the tale is throughout colored by a tint of disdain and contempt. He writes as if religion arose entirely from the triggering of non-rational passions. This is not very likely; nor can be it adhered to without a bit of cheating.

Original Instincts and Secondary Principles

Hume's explanation of the rise of religion rests on an appeal to a set of natural propensities. These are presumably the "first religious principles" he mentions in the Introduction, although he does nothing so obvious as providing a table of religious principles. He classifies them as "secondary" principles of human nature, as distinct from "an original instinct or primary impression of nature," because, he says, there are "some nations," although not many, among whom the belief in invisible intelligent powers is not found.⁶ Moreover, where it does occur, the actual object of the belief, i.e. the quantity and quality of the invisible power(s), differs from case to case. Thus the first principles of religion fail to satisfy two conditions Hume specifies as criteria for being primary principles of human nature: being absolutely universal and having "always a precise determinate object." Were it not for the fact that 'disposition' normally names a tendency acquired by some repeated practice, we might call them dispositions, innate rather than acquired dispositions, which may "easily be perverted by various accidents and causes" (NHR, 309/21). So an "original instinct" apparently never misfires, but those propensities which produce religious belief do. But if we look at Hume's list of products allegedly generated by some original instinct, we find "self-love, affection between the sexes, love of progeny, gratitude, resentment" (loc. cit.).⁷ Are we to believe that these are always expressed and that their manner of expression is never "perverted"? We might reasonably find these to be less varied in their expression than is religion; but that they differ in kind, as Hume supposes, seems much less credible.

One consideration which may have been at the back of Hume's mind is a very good reason for not postulating a religious instinct; but I am unable to prove from the text that it played a role in determining Hume's procedure. It is just that an account of religion in terms of a variety of natural tendencies is a better explanation than an account in terms of a single religious instinct. Certainly Hume, like most philosophers of his day, was quite aware of the explanatory frivolity of appealing to an "occult quality," i.e. a specific power in things whose only function is to produce that specific effect. In the Dialogues, Philo accuses Cleanthes and his fellow "anthropomorphites," as well as mechanistic materialists, of just this sort of obfuscation by pretending to explain cosmic order merely by postulating a principle of order, the one in mind and the other in matter. "These are only more learned and elaborate ways of confessing our ignorance," says Philo (D, Pt. IV, p. 162). Now to appeal to an instinct, say to love one's children, as the reason why people love their children, is just to invoke an occult quality; it is the sort of stab at an explanation one resorts to when one is unable to produce a more detailed account. That Hume does himself use the concept of instinct by no means obviates my conjecture, since he characterizes instincts as "not ascertained by argument or reflection."⁸ When he refers to instinct in the Dialogues it is to acknowledge that its sphere is that for which we cannot account (D, Pt. I, p. 134)⁹ or to cite it, along with reason, generation, and vegetation as principles whose effects "are all known to us from experience; but the principles themselves, and their manner of operation, are totally unknown" (D, Pt. VII, p. 178). So it just may be that Hume, in finding religion's source in a number of "secondary" principles rather than in some "original instinct," is not so much demoting religion as he is expressing an epistemic optimism about it that he must forego in the case of instinctual behavior. His famous concluding paragraph notwithstanding ("The whole is a riddle, an Ænigma, an inexplicable mystery. . . ."), Hume is considerably more sanguine about understanding religion than he is about understanding mother love.

Natural Propensities

What, then, are these natural, albeit secondary, principles? Since, as we have already noted, they are not explicitly enumerated, one must compile a list from a perusal of the text. Here, in what I take to be the order of their relative importance, are the items Hume designates as either 'tendencies' or 'propensities':

1. "The universal propensity to believe in invisible, intelligent power, if not an original instinct, being at least a general attendant of human nature, may be considered as a kind of mark or stamp, which the divine workman has set upon his work; and nothing surely can more dignify mankind, than to be thus selected from all other parts of the creation, and to bear the image or impression of the universal Creator" (§15, 362/75)¹⁰

2. "There is an universal tendency among mankind to conceive all beings like themselves, and to transfer to every object, those qualities, with which they are familiarly acquainted, and of which they are intimately conscious" (§3, 316/29)

3. "their propensity is equally strong [to the first propensity] to rest their attention on sensible, visible objects; and in order to reconcile these opposite inclinations, they are led to unite the invisible power with some visible object" (§5, 325/38)

4. a "propensity to adulation" (§6, 332/44) or a "continual propensity towards flattery and exaggeration" (§6, 332/45)

5. "a propensity in human nature, which leads into a system, that gives them some satisfaction" (§3, 316/29)

6. "It is remarkable, that the principles of religion have a kind of flux and reflux in the human mind, and that men have a natural tendency to rise from idolatry to theism, and to sink again from theism into idolatry" (§8, 334/46-7).

Clearly these are not all equally significant. The last can be immediately excised, since, according to the explanation that Hume gives of "flux and reflux" in §8, it is just the product of the interplay of two other propensities. The propensity to adulation leads to theism by a gradual escalation of terms of praise, until attributes such as 'infinite' and 'one and only' are reached; the propensity to attend to sensible objects produces a reversion to polytheism because the idea of the theistic god is too abstract and remote to satisfy religious consciousness. The status of the penultimate item is a tad less obvious. The context in which the reference occurs is this: Hume is making the point that the very uncertainty about how unknown causes operate would have made forming concepts of them so extremely difficult that our forebears would "would have abandoned so arduous an attempt, were it not for a propensity in human nature, which leads into a system, that gives them some satisfaction." The above-quoted formulation of the second propensity, to conceive things as possessing human traits, immediately follows. Although it may be barely possible that Hume has in mind some innate tendency to unite ideas into a coherent system, this would conflict with the general anti-intellectualism of his account and it is not further discussed.¹¹ The most plausible interpretation is that the "propensity in human nature" merely refers to our anthropomorphizing proclivity, which "leads into a system," by enabling us to conceive unknown causes on the model of quasi-human agents. The "system" is nothing other than polytheism, in which the unknown causes are identified as a community, the gods, modeled on human communities. Hume is not invoking a propensity to systematize; he is just saying that human beings would never have arrived at any system of belief about unknown causes were it not for the blind luck of their having this predilection to model other things on themselves. Therefore, the fifth item should also be excised from our catalog.

Of the remaining four propensities, the last, the propensity to flattery (F), serves as the principle which drives the evolution of religious belief from polytheism to monotheism, whereas only the first three are implicated in the initial formation of religious beliefs. Even these three principles are hardly of equal weight, since the propensity to believe in invisible intelligent power (I) is accorded such obvious precedence over the tendency to anthropomorphize (A) and the propensity to attend to sensible objects (S) as make them seem ancillary. I seems entirely responsible for making religious beliefs religious.

I don't think this can be counted as an explanatory virtue of I. Since Hume writes as if the belief produced by I were synonymous with religious belief, or at least a sufficient condition of religious belief, invoking I looks very much like an appeal to an occult quality. Perhaps a philosopher who relies on instincts and propensities cannot, in the end, avoid occult qualities; but it would be nice to avoid ones which directly produce the phenomenon currently under investigation. If one wishes to understand sleep, showing how it serves various organic needs is surely preferable to saying simply that it is the consequence of a dormative virtue. Similarly, if one wishes to understand religion, showing how it arises from certain general human traits is vastly superior to citing a specifically religious inclination in human nature. The most bothersome feature of I is that it is a tendency to posit intelligent invisible power. This seems terribly ad hoc, although a tendency just to posit invisible power seems a natural, and generally useful, extension of causal reasoning. Let's call this less specific principle I�. Technically, for Hume, all power is invisible; but the sense of 'invisible power' being used here is a correlate of Hume's references to "unknown causes"; it is not power in the sense of 'agency' but power in the sense of 'agent' that is posited. This clashes with the beginning of §4, where he speaks of it being distributed. If in our experience some characteristic event X always precedes some event Y, we may take X to be the cause of Y. If we experience some kinds of events whose production we are unable to assign to any experienced object, we may presume that they are produced by some invisible agent(s). The belief that causes really exist, and so may be found out, although they are not yet known, is a prerequisite not only for religion but for science. Hume would have done better had his premier propensity been more in the spirit of the rule Philo enunciates in the Dialogues: "That Nature does nothing in vain, is a maxim established in all the schools, merely from the contemplation of the works of Nature, without any religious purpose" (D, Pt. 12, p. 214).

But how then would distinctively religious belief arise? Probably through the conjunction of I� and A. On Hume's account, A is significantly redundant relative to I. Although he begins his explication of A with such examples as seeing "faces in the moon," ascribing "malice or good-will to every thing, that hurts or pleases us," and personifying "trees, mountains and streams," he subsequently specifies "sentiment and intelligence" and "thought and reason and passion" as among properties which accrue to the unknown causes through A (§3, 316/29). Dropping A would seem unwise, not only because it adds attributes commonly predicated of the gods, but also because it does seem to be a recognizable general tendency whose operation extends beyond the domain of religion. The person who talks to her pets or even plants need not, but might, adopt a distinctively religious attitude. The poet who personifies nature does not. In thinking of invisible powers at all, are we not constructing, in our imagination, a picture of the world derived from our experience of human agents?¹² In lieu of Hume's apparent partiality towards I, some attenuated application of A might plausibly be taken to be the foundation of either I or I�, and perhaps even of the idea of causation.

But, wait, if A is general enough to explain poetry and pet rocks, then will its addition to I� yield religion? At the very least it would seem that a specific kind and degree of anthropomorphic attribution would be required. Either I or I� conjoined to A would eliminate merely poetical anthropomorphism, since either requires reference to real causal agents. But one who took certain phenomena as evidence the earth is being visited by aliens from outer space, or one who was convinced by certain events that they were being observed and manipulated by sinister secret agents, would not, for that reason, be said to be religious, despite believing the unknown causes to be intelligent. In fact, although some of Hume's examples of anthropomorphization are clearly non-religious, he does not indicate precisely what makes an instance of such thinking religious. I believe, as I suggested above, that he is counting on this work having been accomplished by I; but the counter-examples above show that the attribution of intelligence is insufficient for this purpose. Is there some other feature of I which could be responsible for his making this assumption?

The intelligence requirement in I is egregious; but Hume may also be packing more into the invisibility condition of I than he makes explicit. Could he think we have a propensity to postulate not just invisible power but necessarily invisible power, power which can never be experienced or understood? He can write as if that were so; for example, in §3 he reminds us that

the true springs and causes of every event are entirely concealed from us; nor have we either sufficient wisdom to foresee, or power to prevent those ills, with which we are continually threatened. We hang in perpetual suspence between life and death, health and sickness, plenty and want; which are distributed amongst the human species by secret and unknown causes, whose operation is oft unexpected, and always unaccountable. (NHR §3, 316/29, my italics)

That all causes are unknown and their operation unaccountable is good Humean doctrine; but remember that the "savage and barbarian" regards most regular occurrences as natural consequences of their proximate antecedents. So the savage is not a Humean sceptic. All causation is mysterious, but it seems so only in special cases. "Convulsions in nature, disorders, prodigies, miracles, though the most opposite to the plan of a wise superintendent, impress mankind with the strongest sentiments of religion; the causes of events seeming then the most unknown and unaccountable" (329/42, my italics). My counterexamples are instances of phenomena whose supposed causes, although unknown, are in principle capable of being experienced in regular conjunction with the phenomena. Were they so experienced they would be accepted as the "natural causes" of the phenomena.¹³ They are non-religious cases because they don't seem to be unaccountable. A seemingly "unaccountable" cause, however, is presumed to lie outside the order of natural explanation, i.e. to be supernatural. Let A make this essentially mysterious power somewhat like us, and voila! a supernatural being or god.

Some support for this suggestion can be garnered from Hume's essay "Of Miracles," in which he cites "the strong propensity of mankind to the extraordinary and the marvellous" to explain why so many "forged miracles, and prophecies, and supernatural events" have not lacked gullible believers (EHU, §10, p. 118). The claim that archaic humanity was more interested in the unusual than in regularities is here put more radically than in NHR.

When we peruse the first histories of all nations, we are apt to imagine ourselves transported into some new world; where the whole frame of nature is disjointed, and every element performs its operations in a different manner, from what it does at present. Battles, revolutions, pestilence, famine, and death, are never the effect of those natural causes, which we experience. Prodigies, omens, oracles, judgments, quite obscure the few natural events, that are intermingled with them. But as the former grow thinner every page, in proportion as we advance nearer the enlightened ages, we soon learn, that there is nothing mysterious or supernatural in the case, but that all proceeds from the usual propensity of mankind towards the marvellous, and that, though this inclination may at intervals receive a check from sense and learning, it can never be thoroughly extirpated from human nature. (EHU §10, p. 119)

This principle is not, or is not only, a tendency to be impressed by uncommon data; it is a tendency to favor weird explanations--to prefer an amazing or mystifying account over a pedestrian one. Further, its domain seems more extensive than that of NHR's propensities. It covers not just singularly unusual events but "the whole frame of nature," and attributes unknown causes even to events whose "natural causes" are within our ken. This propensity "to the extraordinary and the marvellous" is not mentioned as such in NHR. If we try to derive this order of supernatural causation from the propensities named in NHR, we clearly cannot obtain it from A, since we are not supernatural. It obviously could not be produced by S, which turns attention back to sensible objects; and F functions only to consolidate power in a chief god after polytheism is well established. Therefore, if it is done by any of the named propensities it must be by I (or, in my revised version by I�), and "invisible" must imply 'supernatural.' This reinterpretation would sever the tie I earlier tried to establish between I� and the practical and scientific quests for causes. On this understanding I� is a tendency to astonish, to place events beyond natural explanation rather than to make them candidates for it.

That we have a propensity "towards the marvellous" is not wholly implausible, considering how we are fascinated by tabloid tales, mysterious or miraculous occurrences, cold fusion, and, indeed, the ancient myths. But since there are all degrees of weirdness, it would seem that gratifyingly bizarre explanations need not always invoke supernatural powers, although that might be the likely outcome of competitive mythopoetizing. Moreover, there is evidence against the assumption that Hume intends invisible power to be necessarily supernatural. He strongly suggests that the ancient gods are to be conceived as natural, rather than supernatural, agents.

If we examine, without prejudice, the ancient heathen mythology, as contained in the poets, we shall not discover in it any such monstrous absurdity, as we may at first be apt to apprehend. Where is the difficulty in conceiving, that the same powers or principles, whatever they were, which formed this visible world, men and animals, produced also a species of intelligent creatures, of more refined substance and greater authority than the rest? That these creatures may be capricious, revengeful, passionate, voluptuous, is easily conceived; nor is any circumstance more apt, among ourselves, to engender such vices, than the licence of absolute authority. And in short, the whole mythological system is so natural, that, in the vast variety of planets and worlds, contained in this universe, it seems more than probable, that, somewhere or other, it is really carried into execution. (NHR, §11, 340/53)

It seems the myths are not so marvelous, after all; they are science fiction or perhaps even, for some planet, prescient science fact. Furthermore, the word 'supernatural' does not even occur in NHR. So, it seems uncertain, at best, whether Hume in NHR conceives invisible power to be instrinsically supernatural. If he did, that part of the agenda of NHR which is concerned with showing, in the spirit of Bayle, that polytheism is in many respects a more reasonable and morally beneficial system than monotheism must have taken priority over that part concerned with uncovering the sufficient conditions for religious belief.¹⁴ In sum, I can find no way to generate specifically religious beliefs from the propensities Hume names without some revision or addition.

The Missing Propensity

To begin the exploration of the idea of adding a propensity more fundamental than A or I, consider the following characteristic assertion, which follows immediately after the claim in §3 that all causes are unaccountable:

These unknown causes, then, become the constant object of our hope and fear; and while the passions are kept in perpetual alarm by an anxious expectation of the events, the imagination is equally employed in forming ideas of those powers, on which we have so entire a dependance. (NHR §3, 316/29)

This feeling of hope and fear in the face of unknown causes is crucially important for generating religion.

The primary religion of mankind arises chiefly from an anxious fear of future events; and what ideas will naturally be entertained of invisible, unknown powers, while men lie under dismal apprehensions of any kind, may easily be conceived. (NHR §13, 352/65)

These emotions seem even to be requisite for activating I, for, says Hume,

Any of the human affections may lead us into the notion of invisible, intelligent power; hope as well as fear, gratitude as well as affliction: But if we examine our own hearts, or observe what passes around us, we shall find, that men are much oftener thrown on their knees by the melancholy than by the agreeable passions. . . . every disastrous accident alarms us, and sets us on enquiries concerning the principles whence it arose: Apprehensions spring up with regard to futurity: And the mind, sunk into diffidence, terror, and melancholy, has recourse to every method of appeasing those secret intelligent powers, on whom our fortune is supposed entirely to depend. (NHR §3, 318/31)

It is easy enough to understand why we become frightened because we already believe in gods, whom we may displease or anger, as Hume declares in his last sentence. But he also thinks that emotions, especially fear, in the presence of causal ignorance, initially "lead us into" the belief in invisible intelligent power. Why should hope or fear trigger this propensity? We think causally, and when the causes are unperceived cool-headed consistency should suffice for our postulating unknown causes. Why does being emotionally unhinged prompt the addition of anthropomorphic properties to those causes, including intelligence? In our citations he offers no better reason than that the imagination is overheated or morbid; but surely it is just as easy to imagine unintelligent, even inanimate, agents of destruction as intelligent ones. I (or I�) and A are supposedly activated by fear and anxiety about the future, but it is not obvious why they should be. Hume makes the religious imagination simply an irrational response to duress. But it is not hard-wired; it is a propensity, not an original instinct; so why have humans persisted in responding to events in ways which, upon reflection, they would have to regard as irrational and symptomatic of a shameful loss of self-control? Could not a better explanation can be given without abandoning a basically Humean system of thought?

A century before Hume Spinoza proposed an account for common religious beliefs which resembles Hume's in locating the initial impetus towards these beliefs in certain tendencies innate in human nature. Also like Hume, Spinoza takes ignorance of the true causes of events to be the fundamental necessary condition for the growth of such beliefs. However, he does not hypothesize any propensity like Hume's troublesome I. He does have an analogue of Hume's A propensity, inasmuch as he claims that when the final causes of events cannot be discerned "from external sources" men propose purposive agents with motives modeled after how they would have themselves justified those events had they been responsible for them.¹⁵ They do this because, says Spinoza, "As they look upon things as means, they cannot believe them to be self-created; . . . they are bound to believe in some ruler or rulers of the universe endowed with human freedom, who have arranged and adapted everything for human use" (Ethics, p. 76). Thus the work of Hume's I propensity gets done, albeit indirectly. Even this anthropomorphizing tendency differs from Hume's A propensity in being derivative rather than fundamental. The more basic feature of human nature from which it is derived is "that all have the desire to seek for what is useful to them, and that they are conscious of such desire" (Ethics, p. 75). They anthropomorphize nature because they wish to discover purposes conformable to their own interests within natural events. Spinoza takes this desire to procure one's own advantage to issue in a teleological interpretation of nature on a cosmic scale which Hume believes to be achieved only by sophisticated philosophical reflection. However, separated from this non-Humean consequence, would not something like this innate desire, and the consciousness of it, have provided Hume with the better explanation we seek?¹⁶

Suppose that humans, in common with all other animals, possess an instinct for self-preservation and that they seek pleasures and avoid pains.¹⁷ They naturally try to control their environment in a manner conducive to their well-being; in Spinoza's terms, they seek "what is useful to them." At those times that they are conscious of their actions they would have to believe that the control they seek to maintain is possible, even if not likely, since otherwise, to the extent that their actions were consciously directed, they would discontinue the pointless action. However bleak and capricious the situation, we have a propensity to believe, or if you prefer, to hope, that our environment may be controlled to a degree sufficient for our, at least minimal, welfare. Call this the C propensity. C is obviously generally useful and in no way specifically religious. Now there are only two techniques whereby environmental mastery can be sought: physical manipulation or psychological motivation.¹⁸ We learn both quite early in our lives; and in ordinary situations use whichever will get us what we need or desire. If I can reach the cookie jar myself, physical manipulation suffices; if not, I can request the cookie of my parents and, if refused, whimper until they relent. Now suppose that we, being scientifically unsophisticated, find ourselves facing some disturbing, even terrifying, phenomenon, perchance an unaccustomedly violent thunderstorm. What to do? So far as physical manipulation goes, not much can be done save taking the available defensive measures, say, scurrying into the best accessible shelter from the storm. The wind and lightning are clearly beyond our physical control; moreover, we, being scientifically unsophisticated, have not the foggiest idea of how they, in principle, operate. Persuasion is the only manipulative technique which, in the circumstances, has even the slightest possibility of success. There is nothing to lose in trying to negotiate with the storm or with whatever is producing it.

Let's assume that Hume is fundamentally correct in assigning fear a crucial role in the genesis of religion. That idea is both very ancient and confirmed by common experience. That we have, à la Hume, some irrational tendency to fall on our knees on such occasions, seems an entirely gratuitous hypothesis if we have a reason for doing so. Doubtless we would not be behaving that way if we were not frightened. But our terror is a function of our loss of environmental control. There may be a valid insight in the hackneyed dictum that there are no atheists in foxholes: religion is last-ditch thinking. We appeal to the gods only when we strongly desire control and the odds seem heavily against our gaining control by physical means. To do so is not to behave irrationally--desperately, yes, but not irrationally. We act that way because we have observed such behavior to be effective in other contexts, such as falling on one's knees before a king, patron, parent, or prospective lover. If nothing else can be done, then plead with the powers; just as if no other alternatives present themselves, you would plead with the enemy who has disarmed you and has you at his mercy. The basic techniques for reestablishing control via invisible intelligent powers are formally identical with techniques which have worked with perceptible "natural causes."

Note that C would apply not only to the instances of unpredictable misfortune which Hume stresses but equally well to the great cosmic regularities upon which our welfare intimately depends, and which I have complained of Hume's ignoring. Had Hume come up with this propensity, he could have tied the domain of the regular and the accidental together, made better sense of the myths, and avoided his implausible characterization of early humanity as thoroughly stupid and irrational. Moreover, C can be evolutionarily justified; those who hold it and act on it are more likely to survive than those who do not. The religious sceptic will demur, perhaps justifiably, at the idea that supplicating invisible intelligent entities enhances survival; but C is just the tendency to believe that something can be done; it does not specify what to do. To use a programming metaphor, Hume's I (or my I�) and A propensities are subroutines which get called by C only under certain conditions. Although the set of requisite conditions may vary with cultural background, they might normally include failure to either understand or acquire physical control of the phenomena, those being, as it were, higher priority subroutines. The entire survival value of C may conceivably reside in its calls to non-religious subroutines, although I suspect that crucial psychological support is often provided by religious practices. Nevertheless, if attempting natural control has priority, religion should diminish with its increasing success. There is general agreement that this indeed happens as scientific comprehension expands. Of course, there are circumstances under which even the scientifically sophisticated agent may call upon the gods. Acquiring understanding may take too long, or the best information may support the belief that the situation is genuinely hopeless; but a prayer in a pinch is easily produced.

Conclusion

Hume's account of the origins of religion is a rhetorical tour de force, but as a serious explanation it has a number of problems. Early man is deprived of curiosity, rationality, and common sense. For Hume, religion is superstition and his account is cast in terms designed to minimize the similarity between our thought and that of the savage. Once reason is pushed aside, he invokes an untidy assemblage of propensities. Some of these are dispensable; others overlap; one approximates being an occult virtue for religious belief. What Hume fails to do, which would have made his story far more cogent, is to show (1) that these propensities have some survival value (or, if this is too Darwinian, some general utility) and (2) that they are not unique to religious thought but have general application. I have tried to show how these remedies could be applied without harming the basic structure of his theory.