Panpsychism and Parsimony

by John J. Shepherd

John J. Shepherd is Lecturer in Theology and Religion at Weymouth College of Education, England. He received his Ph.D. (religious studies) from the University of Lancaster.

The following article appeared in Process Studies, pp. 3-10, Vol. 4, Number 1, Spring, 1974. Process Studies is published quarterly by the Center for Process Studies, 1325 N. College Ave., Claremont, CA 91711. Used by permission. This material was prepared for Religion Online by Ted and Winnie Brock.


Dr. Shepherd holds that Whiteheadian panpsychism, from the argument of parsimony, is unwarranted, but also that it is actually incompatible with what it seems responsible to take to be facts about a physical world, and should therefore be deemed false.

It is sometimes urged that Whitehead presents doctrines without giving supporting reasons. I wish here to consider one important defense of psychicalism, what might be called "the argument from parsimony," which may be approached via some remarks of Whitehead’s on "Philosophic Method" in Adventures of Ideas. I shall argue both that the defense fails and also that it fails in a way which actually leads to rejecting panpsychism as false.

The relevant quotation on philosophic method is as follows. When a general idea has been obtained, it should not be arbitrarily limited to the topic of its origination. In framing a philosophic scheme, each metaphysical notion should be given the widest extension of which it seems capable. It is only in this way that the true adjustment of ideas can be explored. More important even than Occam’s doctrine of parsimony -- if it be not another aspect of the same -- is this doctrine that the scope of a metaphysical principle should not be limited otherwise than by the necessity of its meaning. (AI 237)

Similarly Hartshorne writes: "By metaphysical system, I mean one in which the attempt is made to generalize all ideas to the fullest possible extent" (LP 219).

Now few people would quarrel with the view that a general idea should not arbitrarily be limited to the topic of its origination. Yet after issuing this warning, perhaps Whitehead over-redresses the balance? His subsequent plea is that we should extend the general idea as far as we conceivably may, and this procedure is apparently equally arbitrary. The extension may or may not be warranted, and the only means of reaching any kind of rational decision is surely to appeal to the available evidence.

Nor, it might be pressed, is this objection avoided by appealing to Whitehead’s explicitly adopted method of metaphysical construction, the method of imaginative generalization (of which his plea here as in effect an expression). For even if the "first requisite" of speculative philosophy is to "proceed by the method of generalization so that certainly there is some application," Whitehead goes on to say that "the test of some success is application beyond the immediate origin" (PR 8). and this invites the question of what is the test of application beyond the immediate origin; and the answer must be more than "conceivability." It is necessary to distinguish between a speculative but tenable hypothesis because to some degree authenticated, on the one hand, and conceivably plausible but wholly unwarranted fantasy. Consequently an appeal to evidence is ineluctable. By all means let observation be directed by theory and let the relevance of evidence be dictated by theory (AI 221) -- but evidence there must be.

Yet none of this is decisive. Whitehead refers to Occam’s doctrine of parsimony, and here is the crux. The world may or may not be as tidy as it is often possible to conceive it to be; but it is part, and surely a legitimate part, of our inheritance from the successes of science, quite apart from anything else, that given two otherwise satisfactory interpretations of which one is simpler then that simpler one is preferable even if there is no means of judging -- by appeal to evidence, say -- that it is not, in respect of its greater simplicity, mere plausible fantasy. There is an important sense therefore in which the test of application beyond the immediate origin may not have to be anything more than "conceivability," in which the stretching of an idea to its limit and applying it may not be arbitrary. For if it is granted that our awareness of ourselves leaves us no option but to admit the existence of experiencing entities, and if it be further granted that it is conceivable that supposedly vacuous matter be constituted by occasions of experience, then, arguably, preference should be given to the panpsychist doctrine on the ground that it is a more economical theory than any kind of dualism which posits an additional, radically different kind of entity, the unit of vacuous actuality. In effect, as Hartshorne at one point urges, "the panpsychist challenges his opponents to indicate any reason for admitting an additional factor of insentient matter to explain the world. Can the challenge be met?" (BH 173f). It may be added that Hartshorne’s overall claim on behalf of the Whiteheadian system in this respect is very large indeed. "Causality, substance, memory, perception, temporal succession modality, are all but modulations of one principle of creative synthetic experiencing, feeding entirely upon its own prior products. This I regard as the most powerful metaphysical generalization ever accomplished" (CSPM 107).

Nevertheless although the appeal to parsimony may be right in principle, in practice it is less successful than Whiteheadians suppose. With regard to causality, for example, the supposed simplification is offset by having to introduce a primordial actual entity, God, to intervene virtually continuously to keep the process going by supplying each novel occasion with its initial aim. Now although this primordial actual entity may itself be interpreted in terms of the principle of creative synthetic experiencing, the fact remains that simplicity of principle is gained at the expense of multiplication of entities, and the latter offends Occam’s principle as surely as does multiplicity of principles.

A similar objection might be lodged against the Whiteheadian interpretation of substance. (The objection is mentioned here for the sake of completeness in response to Hartshorne’s overall claim above, but this is not the place to try to substantiate it.) If anything may be regarded as a substance in a Whiteheadian context, the self may be. Yet since the account of substance offered is of a route of occasions or a personally ordered society of selves rather than of, say, a changing continuant, the objection may be raised that this atomization of substance involves an enormous multiplication of entities which in the case of selves at least goes well beyond necessity -- though this would have to be shown. Moreover, it is arguable that the Whiteheadian analysis of the self is morally unacceptable; for insofar as each momentary self is held responsible for the actions of its predecessors in the personally ordered society of which it is a member, the Whiteheadian analysis involves the Old Testament practice of visiting the sins of the fathers on the sons. It is interesting to note in this respect that in Whitehead’s judgment the Jews "conceived one of the most immoral Gods ever imagined" and that he endorses Thomas Hardy’s remark in Tess of the D’Urbervilles "But although to visit the sins of the fathers upon the children may be a morality good enough for divinities, it is scorned by average human nature."1 Hartshorne, of course, claims that the Whiteheadian view of the self is the only fully moral view since it alone enables complete obedience to the Great Commandment to love one’s neighbor as oneself. For: "To love oneself as identical with oneself and the other as not identical with oneself is not, whatever else it may be, to love the neighbor as oneself" (CSPM 200). Yet as E. L. Mascall neatly observes, "The other side of this view . . . must presumably be that I am just as altruistic in promoting my own future welfare as in promoting that of other people!" 2

Hartshorne refers too to temporal succession. It is perhaps just worth mentioning in passing that, according to a recent notable analysis, adoption of the Whiteheadian view creates apparently insuperable obstacles for the theory of time being used successfully in physics.3

The difficulty on which I wish to focus attention here, however, arises in connection with perception. We perceive an apparently material, spatially extended world. Can Whiteheadian panpsychism account for this? Naturally it undertakes to do so, but with what measure of success?

Let us first be clear about a possible misconception of Whitehead’s doctrine -- a misconception perhaps shared indeed by D. H. Lawrence. (The misconception hardly needs exposing for readers of Process Studies, but reference to it serves to bring into focus the basic point at issue.) In chapter sixteen of Lady Chatterley’s Lover the following passage is quoted: "The universe shows us two aspects: on the one side it is physically wasting, on the other side it is spiritually ascending. It is thus passing with a slowness, inconceivable in our measures of time, to new creative conditions, amid which the physical world, as we at present know it, will be represented by a ripple barely to be distinguished from non-entity" (RM 160). It would seem that Lawrence interprets this as a vision and prediction of a nonphysical cosmos in place of the "mixed" mental and physical cosmos in which, as White-head (on this view) concedes, we find ourselves. Now as V. Lowe observes regarding Lawrence’s interpretation, "in terms of Whitehead’s system, this is a complete mistake, since the physical and the mental are universal features of every actual entity, including God" (UW 108). Yet this way of puffing the objection reinforces the view that for White-head the cosmos and indeed all its ultimate constituents are "mixed." Johnson too writes, in similar vein: "Every actual entity has a physical and a mental pole -- i.e., every actual entity is composed of physical and mental activities" (WTR 39). Moreover from Whitehead’s own pen we read:

The dualism in the later Platonic dialogues between the Platonic ‘souls’ and the Platonic ‘physical’ nature, the dualism between the Cartesian ‘thinking substances’ and the Cartesian ‘extended substances’. . . all these kindred dualisms are here found within each occasion of actuality. . . . The universe is dual because each final actuality is both physical and mental. (AI 190)

In the light of such remarks Lawrence and others may be forgiven for supposing that Whitehead’s panpsychism takes the form of some kind of "double aspect" atomism. Moreover outside a Whiteheadian context panpsychism is not infrequently defined in "double aspect" terms, that is to say, it is defined as the doctrine that entities or individuals possess mental properties as well as physical properties. Now the claim that could appropriately be advanced on behalf of this form of panpsychism in respect of considerations of parsimony would be that it does with fewer kinds of entity what it would be vain to do with more -- that it does with one kind of entity what the dualist only achieves with two. Yet against this claim it could be urged that it vainly does with more mental entities what may equally satisfactorily be done with fewer -- that the panpsychist multiplies beyond necessity the features considered generic to all entities. Thus this dispute is not one which an appeal to Occam alone can settle.

It must be emphasized, however, that in a Whiteheadian context this interpretation of panpsychism is a misconception -- though not one that Lowe and Johnson in fact share. Nevertheless Johnson may again mislead when, after pointing out that every actual entity is composed of physical and mental activities, he adds that in Religion in the Making there is a suggestion that some occasions are not bipolar, for Whitehead speaks of "mental occasions" and "physical occasions" (RM 102f, 118). The distinction is not at all one between mind and matter, however, as Whitehead’s reference to "a physical occasion of blind perceptivity" makes clear (RM 103). The physical pole of an occasion is constituted by physical prehensions or feelings (strictly the notion of feelings excludes that of negative prehensions, but that is irrelevant here). Moreover, to the extent that their being "physical" prehensions is defined in terms of their being reactions to concrete occasions rather than to eternal objects (in which case they would be "conceptual" and form the mental pole) (PR 35, 366), they are not "physical" in the sense of "material" at all. Dualism of mind and matter is replaced by duality of quality of experiencing corresponding to the actual occasion/eternal object duality of what is experienced. Again Hartshorne, having conceded that, relatively speaking, molecules are mindless, denies that they (or their constituents) are totally lacking in experience and continues: "the panpsychist . . . further will not admit that the lower degrees of awareness are due to the dilution of mind by its mixture with increasing doses of another something, matter" (BH 170). For him, as for Whitehead too, dualism of mind and matter is replaced by a gradation of higher and lower degrees of experience.

Yet now the question becomes pressing: what account of the properties generally regarded as material is furnished by this analysis? There are indeed only two possibilities. Either they are said to he mental properties after all; or, if this should prove impossible, or implausible, they are said to be only apparent rather than real properties.

Now the latter alternative is intended to apply to the property of total vacuousness, for example; but not to all properties. According to Hartshorne,

"Mind" and "matter" are not two ultimately different sorts of entity but, rather, two ways of describing a reality that has many levels of organization. The "mind" way I take to be more final and inclusive, so that my position is the opposite of materialism. However, I recognize that the material mode of description is that part of the complete mode which is capable of scientific precision and that, accordingly, "methodological materialism" or the restriction of attention to this mode, is a natural bias among scientists. (LP 217)

Methodological materialism abstracts from the complete fact -- and in Whitehead’s view "wherever a vicious dualism appears, it is by reason of mistaking an abstraction for a final concrete fact" (AI 190). Yet the abstraction is not unmitigated falsification -- it focuses too on genuine attributes of what are essentially experiencing entities even though they are not recognized as such. Among these attributes is spatial extendedness.

Here indeed is a test case. The problem is illustrated by Descartes’ division of the world into mind and extension; it is natural to treat the two as irreducible opposites. Yet Whiteheadian philosophy, following Leibniz, interprets the latter in terms of the former. After describing an actual entity as "the enjoyment of a certain quantum of physical time," Whitehead continues: ‘There is a spatial element in the quantum as well as a temporal element. Thus the quantum is an extensive region" (PR 434). Elsewhere we read that "every actual entity in the temporal world is to be credited with a spatial volume for its perspective standpoint" (PR 105). Hartshorne too writes:

The Cartesian division of events into extended and material, and inextended and psychic, is based on no clear evidence on either side of the division. Nothing shows that the psychical events, the experiences, are point-like, or that they are nowhere; and nothing shows that the extended events are simply without feeling, or are merely material (spatio-temporal). (LP 201)

Moreover, in some cases we are aware that experiences are not pointlike -- the case of images, for example; "and obviously if there were not a spatial pattern in some of our sensations, we should never have come to the idea of space at all" (BH 174f).

I must stress that it is this property of spatial extendedness claimed for the individual occasions which I am taking as the test case for the ability of the Whiteheadian analysis to furnish an account of properties generally regarded as material. In common parlance we do not distinguish between this property’s being material and its being physical, but we do distinguish between its being both or either on the one hand and a property’s being mental on the other. It is with this latter contrast that I am concerned. It is thus not necessary for my argument for me to introduce more technical parts of the Whiteheadian scheme such as the difference between a physical entity (an occasion dominated by physical prehensions) and a material entity (a society of physical occasions) and to distinguish correspondingly between the physical properties of individual occasions and the material properties of societies of occasions. These distinctions are unnecessary for I am concerned with something more basic than they, namely the property of spatial extendedness common to all occasions. Again, I do not need to say exactly what the difference is supposed to be between mental and material or physical properties as commonly understood, for my argument will turn on a particular example of the contrast which should be clear and unproblematic. Moreover, since the general, nontechnical use of the term "mental" is all that is required for the argument, it is not necessary to introduce technical Whiteheadian distinctions such as that between "mental" and "subjective."

What then is to be said about the "mentalist" interpretation of physical space offered by Whitehead and Hartshorne? (Any differences between their positions are also unessential for present purposes.) It is not wholly implausible, yet it is, or so I now argue, unacceptable. Consider in this respect a form of that materialism which in Hartshorne’s view "is so plainly untrue that it is hardly necessary to discuss it" (BH 165). This judgment might he justified with regard to behaviorism, but it is virtually worthless with regard to central-state materialism. Now if the latter is to be rejected, it must be because mental events firstly are considered not to be mere appearances and secondly are considered to possess properties which are incompatible with physical properties. Moreover a prima facie case can be made out for incompatible spatial properties. It is said for example that thoughts do not occur at a place in the body at all, not even in the brain; more importantly it is pointed out that the spatial properties of images do not coincide with spatial properties of any part of the brain and that therefore the images must be in some sort of mental space. Yet the central-state materialist replies that images are not really spatial entities of any kind at all except perhaps in appearance -- they are the way in which certain brain processes appear to the material system which we call a person and whose brain it is. Now in my view this is unsatisfactory. Yet the essential consideration for present purposes is that whether or not central-state materialism is true it would seem from a panpsychist or nonmaterialist standpoint to be a necessary condition of its being true that such things as images be only appearances (since otherwise there is the problem of incompatible spatial properties).

Conversely, the panpsychist or nonmaterialist must claim either that both the mental images with their spatiotemporal pattern and the brain with its different and incompatible spatiotemporal pattern are real, or that the images are real and the brain itself only an appearance. Granted this, however, the panpsychist generalizing in Whiteheadian fashion from the case in question must argue either that both mental and physical properties (Or the two kinds of spatiotemporal extensiveness enjoyed by mental images and physical brains respectively) are real and that all entities have both kinds, or that mental properties are real but physical ones only apparent. The former alternative is the non-Whiteheadian double aspect panpsychism; yet the latter and sole other alternative is non-Whiteheadian also, for the Whiteheadian view is that the properties which science abstracts from occasions and embodies in its theories are genuine as far as they go -- though they do not go far enough. It

fully admits the reality of the entities referred to in physics . . . but asks only, What are these individuals like? To the answer -- They are as the physicist describes them and that is all that can be said. -- panpsychism replies that so far as all matters of detail are concerned this is correct, since philosophy has no jurisdiction over questions of detail, which belong entirely to the special sciences, but that there are some questions of principle which in the present state of the special sciences are likely to be forgotten. At present, at any rate, physics describes the mere spatiotemporal outline of things, but says nothing about the qualitative stuff by which these outlines are filled in to constitute realities. (BH 177f, cf. 121)

However, if the argument advanced here is correct, what it shows is that this complementary view of physics and Whiteheadian panpsychism is untenable.

The conclusion is of considerable importance. Even if Whiteheadian philosophy is more of a generalization from human experience than from modern physics, it is inconceivable that it should repudiate the latter. Moreover if it did (assuming this to be possible in the framework of an overall Whiteheadian scheme), then it would itself be forcefully repudiated -- and not simply by physicists, for the material world of common sense as well as of physics would be drastically impugned. Thus examination of the argument from parsimony serves finally to suggest not merely that Whiteheadian panpsychism remains unwarranted, but also that it is actually incompatible with what it seems responsible to take to be facts about a physical world, and should therefore be deemed false. The process of basing the doctrine of panpsychism in an imaginative generalization from our experience as experiencing subjects cannot even get started because our experiencing or mental activity, at the very least with regard to mental images, has to be distinguished from material happenings in the brain, and while the former is not reducible to the latter, the latter can neither be constituted by the former nor wished away.



BH -- Hartshorne, Charles. Beyond Humanism: Essays in the Philosophy of Nature. Chicago: Willett, Clark & Co., 1937; and Lincoln: University of Nebraska Press, Bison Books, 1968.

CSPM -- Hartshorne, Charles. Creative Synthesis and Philosophic Method. London: SCM Press, 1970.

LP -- Hartshorne, Charles. The Logic of Perfection and Other Essays in Neoclassical Metaphysics. La Salle, Illinois: Open Court, 1962.

UW -- Lowe, Victor. Understanding Whitehead. Baltimore: The John Hopkins Press, 1962.

WTR -- Johnson, A. H. Whitehead’s Theory of Reality. New York: Dover Publications, 1962.



1 See Lucien Price, Dialogues of Alfred North Whitehead (New York: Mentor Books, 1956), p. 162.

2 E. L. Mascall, review of CSPM in the Journal of Theological Studies, 22 (April, 1971), 316.

3 See Adolf Grünbaum, Modern Science and Zeno’s Paradoxes (London: Allen and Unwin, 1968), especially pp. 53-65.