return to religion-online

Hartshorne and Creel on Impassibility

by George W. Shields

George W. Shields teaches philosophy at Indiana University Southeast, New Albany, Indiana. The following article appeared in Process Studies, pp. 44-59, Vol. 21, Number 1, Spring, 1992. 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.

In Divine Impassibility: An Essay in Philosophical Theology, Richard E. Creel has authored a systematic and sustained discussion of the time-honored issue of the impassibility of God. As James Keller has observed in a review for Process Studies (PS 15/4), not only is the work an important investigation of the topic as such, it is a highly original treatise in philosophical theology in its own right, which develops virtually an entire doctrine of God around the focus of a single issue. The work should be of special interest to close students of White-head and Hartshorne, for Creel is pointedly concerned throughout with contemporary process philosophy. Most importantly for my own particular interests, Creel’s work represents an important exchange between the perspectives of contemporary analytical theism and contemporary process theism. Indeed, I think that the future of the "process movement," at least in the English-speaking world, depends in some measure on its ability to engage the interests of analytical philosophers and natural theologians. Creel’s attempt at dialogue between the two traditions is thus worth careful consideration.

In this essay I will be specifically concerned with the topic of Professor Creel’s treatment of Charles Hartshorne’s view that God is passible in concrete acts of knowledge, will, and feeling. I will proceed by first presenting a general account of Creel’s concept of impassibility and his basic position. I will then go on in a second section to examine his crucial defense of the doctrine of God’s eternal knowledge of all possibility with a particular view toward the important Creel-Keller debate on that doctrine published in Process Studies (PS 12/4, 15/1). In a third and final section, I will examine Creel’s critique of Hartshorne s version of emotional passibilism. The net result will be a defense of the view that, while Creel’s critique of Hartshorne is certainly well-informed, thoughtful, and deeply searching, and while the issues at hand are some of the most difficult and tangled in the whole field of philosophical theology, I am not yet readily convinced that he has provided a satisfactory critique of or alternative to Hartshornean passibilism/impassibilism.


The topic of divine impassibility has a long history in Christian thought and culture. As Creel notes (DI I), the Post-Apostolic theologians employed the Greek term apathes in some of their assertions about God, taking this to mean, as for instance in Origen’s De Principiis, that God’s bliss cannot be altered and that God lacks all emotion other than bliss. Augustine once suggested that no one could be judged sane yet hold that God is affected by misery. However, many modem Christian theologians have pointed out that biblical theism, with its pervasive notions of God as loving and personal, stands in tension with such claims, for a loving person is not ordinarily conceived as a person unaffected or unmoved by that which is loved. This is not merely the view of modern process theologians. For instance, the British thinker A. M. Fairbairn once wrote that "theology has no falser idea than that of the impassibility of God," and theologian Douglas White denounced impassibility in strong language as "the greatest heresy that ever smirched Christianity" (cited at DI 2).

One of Creel’s seminal insights into this issue is that the long and vexed debate over impassibility has been unduly complicated by the fact that there are many senses of impassibility which are unconsciously conflated in the discourse of theologians. He discerns no less than eight semantically distinct usages of the term "impassible." This means that the question of impassibility is not the simple question, Is God passible or impassible?," as is sometimes presumed, but rather involves an affirmation or denial of impassibility with respect to each of its eight distinctive senses, making for the formal possibility of sixteen theological positions regarding the doctrine of impassibility. Fortunately, according to Creel, these eight senses can be grouped into four genera of senses. Consequently, God can be conceived as passible or impassible in (a) nature, (b) will, (c) knowledge, and (d) feeling. In what is to follow immediately, I will attempt to summarize Creel’s position with respect to each of these generic senses of impassibility in turn.

(a) The question of divine impassibility in nature is the least controversial of the four. Most thinkers who are usually considered passibilists, including Hartshorne, agree with those who are usually considered impassibilists that the divine nature or essence cannot be altered. Creel cites Hartshorne approvingly as writing in A Natural Theology for Our lime (DI 13, NTT 44):

If we abstract from God’s contingent qualities, with respect to the rest of his reality we can view classical theism as largely correct. Here indeed is the uncaused cause, impassible, immutable, and all the rest of it.

Of course, there may be and certainly are a number of disputes about precisely which attributes or predicates God possesses essentially or non-contingently rather than non-essentially or contingently. For example, some philosophical theologians regard God as essentially timeless while others regard God as essentially temporal. But few are willing to hold the view that once one has correctly fixed the notion of divine nature or essence, then God or anything external to God could alter that nature or essence.

(b) With respect to impassibility of will, Creel acknowledges the position of Hartshorne and Nelson Pike (in his God and Timelessness) that an impassible being who is timeless and immutable simpliciter cannot be impassible in will. For an impassible being in this sense cannot be said to have a will in any meaningful sense. The core of Hartshorne’s and Pike’s reasoning here is that to have a will or purpose or intention is to bring about something in time in which there is continual adaptation to a "continually changing present in order to bring about the achievement aimed at" (DI 15). Since, by definition, this cannot be said of an impassible being who is timeless and immutable simpliciter, then an impassible being cannot have a will or purpose or intention.

Creel’s own position is that, while he concurs that God is not timeless and immutable simpliciter, he also holds that the denial of timeless immutability simpliciter does not entail that God is passible in will. To see how he could claim this to be so, his position on divine knowledge needs to be introduced.

(c) Creel has been convinced by Hartshorne’s writings that God cannot know the future as actual until it becomes actual (DI 46, 62). In fact, he submits an interesting counterargument against the "neo-Molinist" position of Alvin Plantinga and Alfred Freddoso, who maintain that at any moment God knows all future contingent events in their full detail (DI 89-92)1 According to Creel, Molinism begs the question by assuming that God could know in advance of creaturely decisions which possible world the actual world is. That is to say, Molinism holds that, since God knows all possible worlds and thus all free responses to conditions in all possible worlds, and since the actual world is a member of the set of possible worlds, God knows creaturely decisions in advance. However, argues Creel, since free creatures are co-determinants with God in bringing about exactly which possible world is actualized, God logically cannot know which possible world the actual world is in advance of free creaturely decisions.

Yet, Creel is not convinced that Hartshorne is correct in his view of divine knowledge of possibility. Contrary to Hartshorne’s position that divine knowledge of possibility grows (in some sense) with the creative advance of actuality, Creel maintains that God eternally knows all possibilities. This claim is a crucial part of his argument and he devotes a key chapter entitled "Continuity, possibility, and omniscience" to vigorous defense of it (also see PS 15/1). On the assumption that God does not know in advance whether a given creature will decide to do X, God’s will with respect to X is nonetheless already decided, so that God will respond accordingly if X should become actual. This is possible because God can predecide responses to the possibles that God eternally knows. In this sense, then, God’s will is impassible; it could never be altered by any circumstances which might actually occur.

(d) The fourth and last sense of impassibility concerns feeling. Creel argues that God is impassible in feeling, yet it is coherent to hold that God is a loving person. In defense of this view, he argues that, if God has eternal knowledge of possibility (abbreviated by Creel as EKP) and is impassible in will, then God could and certainly would (given that the divine nature is love) predecide loving responses -- responses promoting the good of creatures -- to all possible circumstances and actions of free creatures. Furthermore, Creel argues that there are embarrassing and uneasy tensions in Hartshorne’s passibilist view that the aesthetic quality of the divine life -- God’s happiness -- depends on us, and yet, as Hartshorne insists (MVG 240), God has an infinite wealth of happiness. A more coherent picture of divine feeling, Creel holds, would be to maintain with classical theism that God is always and impassibly blissful, always has full knowledge of the creative advance of actuality and all possibility, but does not in any way physically prehend the feelings of creatures or have the feelings of creatures as if they were God’s very own. This is consistent with the notion that "God is love, because, Creel maintains, a person can coherently be said to love another without having the feelings of another as one’s own. In fact, Creel argues that profound loving is inhibited by dependence on emotion (see, e.g., DI 117-21).

In sum, according to Creel’s account, he holds with Hartshorne that God is impassible in nature and passible in knowledge of actuality (and immediate possibilities relevant to the becoming of actualities), but, contrary to Hartshorne, that God is impassible in will, knowledge of possibility, and feeling. In point of clarification, however, it ought to be said that, in light of Hartshorne’s concept of dipolarity, his position vis-à-vis Creel could be best phrased as follows: God is strictly impassible in nature and thus perfect adequacy of will, perfect adequacy of knowledge of actuality and possibility, and perfect adequacy of feeling, but is passible in the concrete or particular states of will, knowledge of actuality and possibility, and feeling, which instantiate God’s impassible transcendental

Finally, as a point of terminological clarification of Creel’s basic position, note that he sometimes speaks of divine volition in terms of responses to creaturely decision, but he also asserts that it is "perhaps more accurate" to speak of God’s eternal presponses, or if one is inclined to Boethian eternalism, indesponses of will (DI 23, 209). This means (on taking Creel’s own temporalistic position contra Boethius) that his deity in fact engages in two kinds of volitional acts -- eternal and impassible presponses to all possibles and secondary volitions in which God wills the already decided presponse as the proper response to an actual temporal occurrence.2


The key to Creel’s position is clearly the doctrine of EKP.3 For only if EKP is sound is Creel’s view of impassibility of divine will sound, as we have seen in the preceding exposition. But is EKP sound?

This is not an easy matter to decide. Indeed, Creel and Hartshorne both confess that their own positions on divine knowledge of possibility are in some ways "obscure and difficult" (DI 52f, CSPM 59). Despite this, I will here venture a response to Creel which maintains the following: (1) Hartshorne is in fact committed to the view that God has EKP in a very qualified sense, (2) this qualified EKP does not imply that God could eternally predecide responses to all possibles, (3) Creel’s position faces the horns of a dilemma -- either God predecides responses to all possibles by virtue of a formula or algorithm (which I argue is conceptually ill-formed on two counts) or God could exhaust a continuum (contrary to Creel’s acceptance of the Peirce-Hartshorne continuity thesis), and (4) Hartshorne’s position does not suffer from five defects alleged by Creel. In the proceeding I am indebted to the careful examination of Creel’s position given by Professor James A. Keller, but I will submit some qualifications and new points of contention, and I will attempt to help Keller do a better job by fleshing out suggestions and providing analogies and phenomenological or intuitive support.

Perhaps the best way to begin a discussion of the issue at hand is to examine Creel’s and Hartshorne’s view of the basic nature of possibility. Creel agrees that Hartshorne’s Peircean doctrine of possibility is correct (DI 36). The exact qualities achieved in the becoming of any two actualities produce a continuum of qualitative difference between them. But, as C. S. Peirce points out, a continuum is intrinsically infinite, meaning that a continuum is "something every part of which can be divided into any multitude whatsoever" (Collected Papers 3.569, cf. also 6.170). As such, a continuum is a determinable and contains no determinate points if we think of a determinate point as a well-ordered, unique, atomic individual distinct from all others. Because a continuum is such that any given part of it can be divided into any multitude whatsoever, a continuum cannot be actualized exhaustively in principle, i.e., in such a way that all the atomizations of quality inherent in the continuum could be exhaustively displayed as atomized units of quality. It follows from this that not even God could exhaust a continuum, for it is simply logically impossible to exhaust a continuum.

While Creel agrees fully with Hartshorne’s continuity thesis derived from Peirce, he denies that this entails that God cannot have eternal knowledge of all possibles apart from divine knowledge of actualities. The chief reason for this denial, according to Creel, is that by knowing a continuum of possibles God knows the possibility of each individual inherent in the continuum without knowing any possible individuals" (which he holds are unintelligible in light of Peirce’s analysis of continuity). The reason that God can know this about continuity is because we can know this about the continua of which we are aware. Creel writes (DI 43):

We are all familiar with the fact that we can take two sticks of equal length that are hinged at one end only and rotate them from a fully closed position to a fully open position, i.e., from an angle of 0o to an angle of 180o. If we close the sticks again, add an elastic band to the unhinged ends of the sticks, and then open them again from 0o to 180o, we will in the process have circumscribed the angularity of every possible isosceles triangle. I submit that when we realize this we will have understood what a Euclidean isosceles triangle is because we will have comprehended it as a continuum of possibilities.

The significance of this, according to Creel, is that by understanding a continuum of isosceles triangles, "we know all there is to be known about the possible relations among the angles of isosceles triangles" (DI 44). ‘Thus, by indexing a response to each possible situation that might arise, God is not indexing responses to each and every one of an infinite set of discrete individuals, but rather to the set of such individuals constituting a continuum. By knowing a continuum, God knows the kind of thing which could be atomized from a continuum.

We have been discussing continua in the plural, but Creel suggests that all possibles reside on a single continuum or "one grand range of potentiality" (DI 204). In fact, he devotes a very interesting chapter of his book to a defense of what he calls "the plenum," which is the entire domain of passive potentiality out of which God creates contingent actualities. I take it that the plenum is a single continuum of possibles, since Creel speaks of it as a unity, as "the dwelling-place of all possibilities" and as "something which passively contains the full range of logical possibilities" (DI 68, my italics). Indeed, as Keller rightly argues (PS 15/1: 291-92), Creel is compelled to place all possibles on a single continuum if he is to avoid the problem of nondenumerable groupings of things into eternal kinds of possibles, i.e., if the set of possibles is eternally divided into specific groups of possibles, then any arbitrary level of grouping specificity would be permissible and this tacitly affirms that the inexhaustible can be exhausted.

Keller also notes that, for Creel, God’s predecision for the whole continuum of possibles would be accomplished by fixing upon a "super-formula" as is modeled in the way that, say, the formula X=3y can specify an output for a nondenumerable set of inputs. But if this is what Creel’s view amounts to, namely, that God employs a super-formula or algorithm to predecide responses to all possibles, then I think he is rather clearly mistaken. At best I think it is unhelpful to suggest this notion as a way of making it clear how God could predecide responses to all possibles. One reason I say this is because algorithmic procedure will not hold even for the entire scope of mathematics. Recently, this point has been made rather nicely in Roger Penrose’s discussion of mathematical problems which are not "algorithm resolvable."4 Consider the Diophantine equations of classical Greek mathematics. For instance, how could an algorithm be produced which gives integer solutions for the following Diophantine set?:

[z3-y-1 = 0, yz2-2x-2 = 0,y2 2xz+z+l = 0]

In this case, we certainly can give a solution, namely, x=13, y=7. and z=2, but there is no generalizable way of getting the solution to this or any other arbitrary set of Diophantines. The Diophantines are "case by case," non-recursive,5 and irreducibly set specific. Of course, Creel could respond to this with an argument from ignorance: perhaps God, being an unattenuatedly omniscient mathematician (an assumption that Hartshorne would not reject), is somehow aware of an algorithm which generalizes the Diophantines and any other non-recursive domain of mathematical entities. But surely this would amount to special pleading, since we have no positive reason even to suggest that the Diophantines are anything other than non-recursive.

Another line of argument against Creel’s algorithm or super-formula view has been suggested by Keller. He writes, "I do not think that most process thinkers will find this [Creel’s algorithm] satisfactory, for they understand creaturely situations to be basically aesthetic constellations . . . and I suspect that the idea of a formula for giving aesthetic responses to every possible situation is impossible" (PS 15/1: 291). Keller does not develop this line of argument, but I think it could be done so (at least in outline) as follows: An important theme in Hartshorne’s metaphysical work is the problem of quality-structure duality. In a brief essay on Roy and Wilfrid Sellars, for instance, Hartshorne observes that Wilfrid Sellars6 has argued formidably for the position that mental states have a qualitative content (and it is phenomenologically ineluctable to say otherwise), which simply cannot be reduced to the structural-quantitative terms of neurophysical explanations of mental states (CAP 240). In fact, Hartshorne notes, structural notions such as geometrical figure have the logical status of abstract determinables, not concrete determinates. Any actual geometrical representation is never a mere geometrical object, say, a mere octagon, but rather, a white or green or blue octagon, etc. And it is naive to object that the color spectrum is reducible to mathematical form, since a color is "nothing more" than a quantifiable frequency of photon emission. For it can be quickly countered that a color is (or is not separable from) a color-sensation -- a psychophysiological event (somehow) with phenomenological content. To talk about a "quantifiable frequency of photon emission" is not to talk about a color, but to talk about that which gives rise to color-sensation.

Indeed, on this point of there being an irreducible phenomenological distinction between structure and quality, there is wide agreement among philosophers as disparate as the Sellarses, Russell, Whitehead, DeWitt Parker, and more recently R. M. Adams and R. G. Swinburne.7 The question which arises acutely for Creel’s position here is obviously: How could a divine algorithm or super-formula ever bridge the logical gap between structure and quality? In proffering such a suggestion I think Creel is tacitly committing the fallacy of misplaced concreteness. To borrow a phrase from F. H. Bradley, no "bloodless dance" of number or mathematical formulae could ever yield a quality (although the converse seems to be true -- number and geometry can be abstracted from a quality).

Now, Keller also attempts to argue against Creel’s position as follows: Creel himself admits that a nondenumerable continuum cannot be exhausted in principle and that even God cannot know as discrete all the potentia inherent in a continuum. But Keller holds that, if God cannot know as discrete all the potentia inherent in a continuum, then God cannot have predecided a response for each possible and therefore cannot have an impassible will. "For if something has not been brought to full conscious awareness [namely, a discrete possible], would not a decision about it be a decision about one knows not what?" (PS 15/1: 2).

I do not think the point of Keller’s rhetorical question follows or at least not exactly. I want to revise his question in this fashion: "If something has not been brought to full conscious awareness, would not a decision about it be a decision about one knows not somewhat?" This revision indicates the way in which I think Hartshorne’s view of divine knowledge of possibility ought to be interpreted. The extent to which God lacks full conscious awareness of possibles is the extent to which God cannot predecide responses to possibles. However, Hartshorne’s God has knowledge of all possibles in some way or another, because if such God did not this would be tantamount to saying that some possibles were not inherent in the abstract continua -- the cosmic variables -- that, according to Hartshorne’s explicit statements, God eternally knows (cf. BH, Ch. XIII). Thus, the situation of Hartshorne’s theory is more subtle than either Creel or Keller envisions: In one qualified manner of speaking, for Hartshorne, God has EKP and thus God’s knowledge of possibility is comprehensive. However, it does not follow that God’s possession of comprehensive EKP in this qualified sense entails that God could eternally predecide responses to all possibles. For despite this qualified EKP, God’s knowledge of possibility is ever increasingly fine-grained.

I wish to develop this interpretation of Hartshorne by way of first considering an aspect of his view that Creel finds particularly puzzling. In an often cited passage, Hartshorne complains against Whitehead’s theory of eternal objects that: "I do not believe that a determinate color is something haunting reality from all eternity, as it were, begging for instantiation" (CSPM 59). I might spell out what is troubling Creel here by noting that the following cosmological scenario is acceptable to Hartshorne in principle: Exactly eight cosmic epochs ago, or, say, some 350 billion years ago or what have you, the actual conditions for the color spectrum came into being and only then did God or any other entity have the foggiest notion of the imminent possibility of color. Does not the point of emergence here seem arbitrary and odd?

One Hartshornean response to this will no doubt be that the color continuum has category specificity and is thereby contingent and therefore ought not to be regarded as non-emergent in principle, whatever our subjective expectations. After all, one can conceive a possible world without color sensa, for such a world would not differ in this respect from historical states of this actual world in which presumably evolutionary process had not reached the stage of organisms with sensory organs capable of color sensation.8

What is more, if we accept Whitehead’s position on eternal objects, or any view which maintains or implies that discrete possibles are once and for all complete, then we have the following difficulty, which I find more intractable than any paradox incurred by the Peirce-Hartshorne continuity thesis and its rejection of eternal species of quality: The set of discrete possibles is absolutely, infinitely complete -- for Whitehead, they are forms of definiteness there for the prehension or envisagement even if they cannot all be consciously envisaged -- and this commits us to the view that there is a highest infinity in extra-mental reality. Cantor’s Big Omega is thus codified in the Whiteheadian domain of eternal objects, since such objects are definite, and thus countable, yet they admit of no additions. But most transfinite mathematicians reject the conceptual existence of Big Omega on grounds of a number of severe paradoxes,9 and any proponent of the theory of eternal objects must face such paradoxes. Hartshorne has been particularly concerned to reject Whitehead’s view of eternal objects, not only because of the powerful case for nominalism made by Quine, Goodman and others (CSPM 59), and Everett Hall’s arguments to the effect that the theory entails all the difficulties of extreme Platonism without gaining any benefits for Whitehead’s system,10 but because it is a self-referentially incoherent position for process theists. One cannot consistently accept the notion of a highest infinity (as implied in the doctrine of discrete eternal objects), while simultaneously rejecting (as do process theists) the Thomistic notion of God as esse ipsum subsistens, which in effect defines God as the realization of a highest infinity.11A thoroughly consistent process theism which understands itself ought thereby to reject any notion of an eternally complete set of discrete possibles.

But it would not follow from Hartshornean principles that, given any arbitrary color-empty possible world or world-state, God would have absolutely no awareness whatsoever of the possibility of the color continuum. God could have awareness of the possibility of the color continuum in a rough or approximate way because God, according to Hartshorne, would always have knowledge of the affective continuum -- the continuum of feeling -- under which the color continuum is subsumed. In his early essay on cosmic variables in Beyond Humanism (111-24), Hartshorne argues that the variables applicable to all possible situations are those whose concepts admit, not just an infinite range of instances, but an infinite scope, i.e., exemplification in every actual entity. The concept of feeling, he holds, is one of these variables. Thus, God always has knowledge of the affective continuum and some actual instances of it. Granted, to be sure, God could not know in advance of any actualizations of color sensa exactly what a color sensation would be, but God could always know it approximately and analogously, because God would always know some actual sensation or other which has an intrinsic connection to color sensation on Hartshorne’s theory of the affective continuum. Indeed, part of the significance of a continuum is that any instance of it shares something in common with every other instance. For example, coolness is analogous to blue as warmness is to red. In The Philosophy and Psychology of Sensation, Hartshorne makes this notion of intrinsic connection between sensory modes quite explicit as follows (PPS 6, my italics):

The type of relation existing between colors, whereby one is connected with or shades into another through intermediaries can be generalized so as to connect qualities from different senses...or from different elementary classes (e.g., secondary and tertiary qualities).

In a manner of speaking, then, for Hartshorne, the possibility of the color continuum has always existed, since it is an intrinsic part of the affective continuum which is always somehow actualized. Thus, according to my reading of Hartshorne’s theory, God does have some eternal knowledge of all possibility. The important qualification here is that just to the extent that possibles for future actualization are distant, ill-defined, and non-imminent, they are known vaguely and only through analogy with the creative advance of actuality.

But none of this entails that Creel is correct in holding that God’s knowledge of dense continua of possibles enables God to predecide responses to all possible circumstances. For I present him with the following dilemma:

(1) Either God knows all possibles through an algorithm for possibles or God knows all possibles as discrete. 12

(2) The arguments already submitted against the algorithm approach appear to exclude that alternative.

(3) The remaining alternative contradicts Creel’s own understanding of possibility as a continuum.

Thus, unless Hartshorne’s theory of divine cognition of possibility itself incurs obvious and irresolvable difficulties, it should then be asserted that Hartshorne’s position seems correct: if possibility is of the nature of a dense nondenumberable continuum, God cannot eternally know all possibles as discrete and God cannot predecide responses to all possibles as discrete.

Professor Creel will surely respond that Hartshorne’s theory does incur some rather obvious and apparently irresolvable difficulties. I discern at least five lines of argument (see DI 50-63):

(i) Hartshorne’s God has brought about an infinite past and has learned an infinite number of things. Consequently, Hartshorne’s God "has learned all that could be significantly learned" (DI 53). This "learned" knowledge is no different from EKP.

(ii) If God lacks EKP, then God’s action on the world must be conceived as experimental, partially ignorant, and based on trial and error. This is an unpalatable consequence for divine omnipotence.

(iii) God’s omnipotence implies that God has EKP.

(iv) Hartshorne’s view implies that God is eternally "recessively ignorant."

(v) Hartshorne’s view of the unrepeatability of actual instances of quality implies that the divine memory becomes increasingly distorted as tune advances.

I will respond to each of these objections in turn.

(i) Creel is correct to point out that, for Hartshorne, at any arbitrary point in the divine history, no matter how far we recede, God has already accumulated an infinite knowledge of actuality and possibility "grounded" by the creative advance of actuality. This is an inherently weird, but I do not think absurd, consequence of holding that God’s creative activity has no beginning. (Anyone who holds that God is essentially temporal, whether or not one also holds the process conviction that God is essentially creative, faces the same inherently weird consequence. For an essentially temporal God would always know, by definition, an infinitude of divine states [or a divine state] through an eternity of time.) But it simply does not follow from this that such a God at any point could not learn anything of significance about possibility. The reason is that, again, according to a widely shared intuition of mathematicians, any given transfinite is less than the absolute infinite (Cantor’s so-called Big Omega). Now, for Hartshorne, at any point at which a new actual occasion X completes its concrescence, a (partially) new kind of thing comes into being and into divine cognition. It is the kind of thing whose instances lie between the parameters set by the exact qualia of X and the exact qualia of all past actual occasions forming the set complement of X, and not just a new particular thing. So, even though God knows an infinite number of things, there is yet an infinite number of kinds of things which God at any arbitrary time will come to know, if any arbitrary transfinite is less than the absolute infinite. And surely "partially new kinds of things" are things of significance for God to know. Thus, I see Creel’s objection here as a non sequitur.

(ii) Creel is correct to suggest that Hartshorne’s God does not know entirely what will be the outcome of any given concrescence of an actual occasion, whether divine or non-divine. For Creel, this raises a serious puzzle

There is here, it seems to me, a serious question as to how God knows what he can bring about and how to bring about anything that has not yet existed. Does he use trial and error? Does he thrash about and accidentally discover what is possible and how to cause it?

As Keller points out, however, this is misleading (PS 15/1:11). For Hartshorne’s God has knowledge of the range of final outcomes of actual occasions, and consequently it is mistaken to assert of such God that it "lures into existence something he knows not what" (DI 55).

Perhaps an example will more clearly illustrate the force of Keller’s complaint. For the purposes of illustration, think of the process God’s agency as partly or weakly analogous to an artful archer aiming an arrow at a target. The archer has definite knowledge of parameters deemed "the bull’s eye zone." The archer definitely knows the parameters of string pressure, bow positioning and aim within which the arrow will strike the bull’s eye. (This aspect of the analogy is grounded in the process God’s absolutely perfect knowledge of the past.) But the actual outcome of the process, the exact location of the arrow’s strike, is not known by the archer. What is known by the archer prior to the outcome is that a bull’s eye strike will be achieved. Supposing all these conditions, should we then think of the archer, or by analogy God, as not knowing what she or he is doing or as "luring into existence what one knows not what"? Of course, the analogy here is only a weak one, since, on the process view, the target and the arrow are all partially self-moving, but it effectively makes its point, I think, about the inappropriateness of describing Hartshorne’s God as utterly out of control of things or as requiring experiment to achieve certain general yet specifically bounded aims.

(iii) Creel maintains that divine omnipotence entails divine EKP. God’s omnipotence, according to Creel, entails that God has "unlimited powers of thought" (DI 59). Now, Hume is celebrated for his thought-experiment in which he claims that human beings can imagine a shade of blue between two shades of blue given to perception, although they might never have actually perceived such an intermediate shade before attempting to imagine it. If this holds in the human case then should not God, unlimited in powers of imagination, be able to discriminate any shade of blue whatsoever? If this is so, then God has EKP.

Keller correctly argues, I think, that this proposal begs the question (PS 15/1: 9-10). His point is made in the context of discussion of a somewhat different variation on Creel’s theme that omnipotence entails EKP: Creel also holds that since God has the power to create any shade of blue and since God knows that this is within the divine power, God thereby knows the possibility of any shade of blue whatsoever at any arbitrary time. The problem here is that if God can create any shade of blue at any arbitrary time, what would preclude God’s cutting all points of a continuum? Indeed, in asserting that God knows any shade of blue whatsoever by imagining it as in the case of Hume’s missing blue, this just seems to be a verbally distinct way of asserting that an inexhaustible continuum could be exhausted, which Creel admits is logically impossible. Given Creel’s commitment to the continuity thesis, the reasoning here is internally inconsistent.

(iv) Creel correctly observes that Hartshorne’s position implies that God learns about possibility as time advances. But, if that is the case, then as we recede into the past God has less and less knowledge of possibility. Thus, at one point or another in divine history, Hartshorne’s God was as ignorant as a present day "clam" (PS 12/4: 224).

Now surely Hartshorne does not wish to hold that God could ever be as ignorant as a present day clam. Fortunately, this view does not follow from his position for three reasons. First of all, if my interpretation of Hartshorne is correct, his God does have qualified EKP. God eternally has some knowledge, always partly vague and analogue-dependent knowledge, of all potentiality. Presumably no clam or any other finite being could have such knowledge. Second, as Keller suggests (PS 15/1: 12), a clam or any other finite being is capable of only finite subranges of possibility, while Hartshorne’s God at any time always has knowledge of infinite actuality and infinite sub-ranges of possibility. Thirdly, Hartshorne’s God purportedly has all the knowledge of possibility that logically could be possessed by any being whatsoever at any arbitrary juncture. It is no fault of omniscience that there are logical limitations to what can be known inherent in the concepts of time and possibility as continuum. (In all fairness, Creel is quite cognizant of this point [DI 55], but he wants to try to make the case for the intelligibility of what he sees as a more "exalted" notion of divine knowledge.)

(v) Creel’s fifth line of argument is surely the weakest, for it is predicated on a clearly faulty interpretation of Hartshorne’s view.13 He begins with Hartshorne’s statement at CSPM 64 that "particular qualities in their absolute definiteness are irreducibly relational and historical." This means that each divine and non-divine actual occasion is unrepeatable in its absolute definiteness. But, if no actual occasion is unrepeatable, then Creel reasons that the actual occasions of divine memory are constantly changing and thus distorting the divine cognition of the past as time advances (PS 12/4: 228).

As Keller again rightly points out (PS 15/1: 13), this argument fails to distinguish between Hartshorne’s and Whitehead’s notion of the uniqueness of the subjective form of concrescence and the data of concrescence. In other words, for Hartshorne and Whitehead, each actual occasion reacts to its data in a unique way (Whitehead’s "subjective form" of occasions), but it does not follow from this (indeed it is explicitly rejected by both philosophers) that the data prehended by each actual occasion in its process of concrescence are as data subject to alteration in any way. This is the meaning of the process doctrine of the perishing of actual occasions. Once the actual occasion has completed its process of feeling and thus defined itself, it becomes being and will remain so eternally for both subsequent divine and non-divine prehension.

I conclude provisionally that, if my and Keller’s counterarguments are correct, Creel has not clearly established that Hartshorne’s position on divine knowledge of possibility suffers from the defects that he claims it suffers from. For all Creel has shown, Hartshorne’s view of divine knowledge of possibility is at least viable and not obviously incoherent or obviously religiously pernicious.


In agreement with the classical theistic tradition of Augustine, Anselm, and Aquinas, Professor Creel maintains that God should be regarded as an emotionally impassible being. To be sure God knows of our suffering, according to Creel. But this does not entail that God must feel our suffering such that God thereby suffers. God is eternally blissful and is completely emotionally unaffected by what goes on in the domain of creaturely actuality. In a chapter of DI entitled "Divine Impassibility in Feeling," Creel responds resourcefully and in detail to the arguments of theistic emotional passibilists. In this section, I will be concerned with only one of these arguments, namely, the argument of Charles Hartshorne that divine omniscience entails that God must be emotionally passible. I point out, however, that if Hartshorne’s position can be vindicated here, it would be sufficient for rejecting emotional impassibilism.

Hartshorne has maintained the view that to know fully and concretely is to feel fully and concretely. In fact, the great distance between God and creatures can be measured by the great difference between divine and creaturely feeling (MVG 163). Creel agrees with process philosophy that, it seems axiomatic that to know, intuit, or feel someone else’s feeling is to have that feeling oneself in some sense" (DI 129, my italics). If this is not admitted, it seems a "mystery" how one could know a feeling at all (DI 129). (Apparently then, Creel, like process philosophers, must reject Anthony Kenny’s account of omniscience and experience in which God can know "the information content of our perceptions without the hedonistic content." 14) However, Creel contends that God can have a feeling in some sense without being in the state of the feeling. This is because there is a certain "epistemic distance" in divine cognition of feeling in which there would be realization "that the feeling was not direct but indirect, not original but derivative, not primary but secondary" (DI 129). Indeed, Creel, I think quite rightly, regards the view that knowledge of a feeling is to be in the state of that feeling as "disastrous" for theology, for it would mean that God could feel stupid, nauseated, drunk, sexually excited, jealous, depressed, etc. This would obliterate any intelligible distinction between God and creature.

Hartshorne surely agrees with Creel that to feel is not to have the feelings of others as one’s very own. For example, he asks pointedly, "How can God feel the sadistic joy of a malicious man without being sadistic?" (CSPM 241). He answers in terms of Whitehead’s distinction between "subjective" and "objective form" of feeling (CSPM 241). God indeed feels the sadistic joy of the man (the "objective form"), but appropriates those feelings into a synthesis of feelings involving evaluation ("intellectual prehension"), resulting in a unique, partially reactive, subjective form of divine feeling.

The conclusion that Creel draws from the view that God can have feelings without being in the state of those feelings is, however, that "there is nothing in the nature of omniscience which requires that God be caused suffering or joy in his own life by virtue of knowing the sufferings and joys of others" (DI 131). I think this conclusion is much too strong as it stands. For it seems to me to suggest that God is simply not affected by creaturely feeling, and I want to argue that one can be affected by the feelings of others while still not having those feelings as one’s very own.

The point of Hartshorne’s talk about "participation in feeling" is that there can be experience whose fabric is shaped, influenced, affected by the feelings of others, yet is not identical to the feelings of others. I grant Creel, as surely would Hartshorne, that God is not affected in the same manner in which creatures are affected by knowing creaturely feeling, because this would be incompatible with certain features of the divinity of God. To use Hartshorne’s own example, in knowing the human experience of threat of death, God is not thereby put in a state of fear, because God is everlasting and thus cannot die and God knows this because God is omniscient (CSPM 263). But it does not follow from this that God is sheerly unaffected by or blissfully indifferent to human experiences of the threat of death.

Indeed, affective states seem to me to be highly complex, nuanced, and multi-layered. Their actual tertiary texture is composite yet identifiable as unified, as "mine." In fact, it is quite possible for precise qualities of feeling to be altered, while the gestalt, the essential or predominant character of the unity of feeling, is not altered. For example, as I write this paragraph, I feel at T "some itchiness on my skin and a slight stiffness of muscle, but basically absorbed in thought," and at T1 I feel, "decreased itchiness and increasing stiffness, but basically absorbed in thought." If this is a correct phenomenology of feeling, I see no reason why it could not serve as a model for understanding how God’s quality of feeling can be altered in precise elements of affect by knowing creaturely feeling, while there remains a gestalt of feeling which is constant and predominant.

Notwithstanding, it is important to notice that Hartshorne also adheres to the view that God is impassible in the categorically superior mode of happiness which attends the inner life of deity. Hartshorne writes (MVG 239f-40):

. . . since at all times God enjoys an infinite past, the wealth of happiness which he possesses is never less than infinite. Though not completely beyond tragedy or the possibility of increase in happiness (nor the risk of falling short of the maximal possible increase), yet is he superior to us in happiness, with a unique incomparable superiority, as the gap between the finite and infinite is unique.

There seem to be several reasons why Hartshorne would want to hold this position: (i) God’s happiness is always superior to creaturely happiness, because, according to Hartshorne’s metaphysical theory, God has always been creating some world or other and every world has some beauty as a whole (CSPM 289-90), and thus God has an infinite accumulation of experience of beauty. Creaturely happiness can only be constituted by a finite accumulation of aesthetic value, by definition of "creaturely." (ii) The divinity of God excludes a number of negative emotions such as fear of death, hatred and envy as the subjective form of divine feeling. For, on Hartshorne’s theory of divinity, God cannot die, the divine all-inclusiveness entails that God’s self-interest coincides with God’s interest in all others (precluding any "I against others" which is a necessary condition for hate), and God is categorically supreme and knowledgeable of this status and thus cannot be envious. (iii) God’s contemplation of the abstract essence of divinity is a unique source of happiness which God forever enjoys (PSG 162). Finally, as a significant corollary of (iii), (iv) the intention to enhance future experiences of creatures is itself an element of present harmony in the divine life (CSPM 308).

Creel is sensitive to this position and he finds it "more consistent" than other versions of theistic emotional passibilism (since Hartshorne admits that items of suffering are really items of suffering ingredient in divine feeling and are not to be "transformed away") (DI 139), but he is hardly satisfied with it. In particular, he points out (DI 136):

Hartshorne’s position is sensible as far as it goes, but there is an important omission. There seems to be no ascertainment of the weight that God’s enjoyment of his essence contributes to the total package of the quality of his experience at any moment. Does God’s essential happiness constitute 50% of his happiness at any given time, so that free creatures can affect only the other 50%? Or is the ratio somewhat different? 1% happiness from his essence and 99% happiness/unhappiness from others? Or vice versa? In short just how passible is God with respect to his happiness?

I surely grant Creel that I do not see how anyone could make a specific determination about the ratio between God’s essential happiness and creaturely contributions to that happiness. Moreover, I think it fair to say that, because of this, Hartshorne’s view is not entirely free of obscurity. I do not think, however, that it is by any means hopelessly obscure. The reason I say this is because of my just suggested phenomenology of feeling as complex and multi-layered. As I reflect on the aesthetic quality of my own life, I find that there have been relatively long stretches of time in which I could honestly say that I was happy, even intensely so, although during those times I was episodically physically ill, reacting to certain stresses, rather excessively cold, rather excessively warm, crying at movies, embarrassed, angry, hungry, thirsty, etc. The overall gestalt of happiness persisted despite this variety of episodic affective states. I do not at all know how to quantify the factors contributing to my overall sense of happiness, but I do know that, as a matter of phenomenological fact I have been in complex states of profound-happiness-cum-vicissitudes-of-living. Since I can make some sense out of this in my own case, I submit that some sense can be made out of Hartshorne’s God who is everlastingly, essentially happy, yet emotionally passible in the strict sense, a fortiori given the mentioned freedom from sources of unhappiness which Hartshorne’s God enjoys by virtue of divinity.

In conclusion, I find no insurmountable difficulty presented by Creel sufficient to prescind the neoclassical theistic position that divine omniscience entails that God is emotionally passible in precise elements of affect, yet impassible in the categorically superior mode of divine happiness. Notwithstanding this and the earlier verdict, process thinkers should be indebted to Creel for a penetrating and challenging examination of some of Hartshorne’s central doctrines.



BH -- Charles Hartshorne. Beyond Humanism: Essays in the Philosophy of Nature. Gloucester, MA: Peter Smith, 1975.

CAP -- Charles Hartshorne. Creativity in American Philosophy. Albany: State University of New York Press, 1984.

CSPM -- Charles Hartshorne. Creative Synthesis and Philosophic Method. LaSalle, IL: Open Court, 1970.

DI -- Richard E. Creel. Divine Impassibility: An Essay in Philosophical Theology. Cambridge: Cambridge University Press, 1986.

MVG -- Charles Hartshorne. Man’s Vision of God and the Logic of Theism. Hamden, CT: Archon Books, 1964.

NTT -- Charles Hartshorne. A Natural Theology for Our Time. LaSalle, IL: Open Court, 1970.

PPS -- Charles Hartshorne. The Philosophy and Psychology of Sensation. Chicago: The University of Chicago Press, 1934.

PS 12/4 -- Richard E. Creel, "Continuity, Possibility, and Omniscience," Process Studies 12/4 (Fall 1983): 209-231.

PS 15/1 -- James A. Keller, "Continuity, Possibility, and Omniscience: A Contrasting View," Process Studies 15/1 (Spring 1986): 1-18.

PS 15/4 -- James A. Keller, "Critical Studies and Reviews: Richard E. Creel’s Divine Impassibility," Process Studies 15/4 (Winter 1986): 290-296.

PSG -- Charles Hartshorne with William L. Reese. Philosophers Speak of God. Chicago: The University of Chicago Press, 1953.


1Process philosophers and theologians have been slow to respond to the now longstanding development of the middle knowledge perspective in contemporary analytical philosophy of religion. The only direct "process" response of which I am aware is Eugene Peters, "Divine Foreknowledge," Encounter 40 (Winter 1979): 31-34. For a quite thorough and I think essentially correct critique of middle knowledge see William Hasker, God, Time, and Knowledge (Cornell University Press, 1989), Ch. 2. Hasker recognizes the clear affinity between his position on divine knowledge and that of process theology, but he opts for Creel’s view of God’s knowledge of possibility (Hasker, 188-189). Also see Donald Wayne Viney’s well-written piece, "God Only Knows? Hartshorne and the Mechanics of Omniscience" in Hartshorne, Process Philosophy and Theology, ed. by R. Kane and S. Phillips (State University of New York Press, 1989), and my "Some Recent Philosophers and the Medieval-Renaissance Problem of Future Contingents" forthcoming in The Midwest Quarterly.

2 In a footnote at DI 209, Creel remarks that, "Leibniz objected correctly to the implication of occasionalism, and by anticipation modern process philosophy, that God is incessantly making billions of decisions every millisecond in response to what is happening in the world. Such a concept of God seems unnecessarily complex." However, since Creel’s God has two sets of volitions, namely, eternal presponses to an infinitude of possibles and efficiently causal volitions which match the presponse to actualities, I hardly see simplicity of volitional acts as a virtue of Creel’s model of deity! (I owe this point to David Griffin who was respondent to an earlier version of this article presented at the Center for Process Studies at Claremont. California.)

3EKP abbreviates "eternal knowledge of possibility."

4Roger Penrose, The Emperor’s New Mind: Concerning Computers, Minds and the Laws of Physics (Oxford University Press, 1989), p. 129.

5A set is recursively innumerable only in case the set can be generated by an algorithm.

6Wilfrid Sellars, "The Double-Knowledge Approach to the Mind-Body Problem," The New Scholasticism 45 (Spring 1972): 269-89.

7See Hartshorne’s references at CAP 240 and 243. Also see R. M. Adams, The Virtue of Faith (Oxford University Press, 1987), Ch. 15, and R. G. Swinburne’s "Arguments From Consciousness and Morality" in his The Existence of God (Oxford University Press, 1979), pp. 160-75.

8This assumes Hartshorne’s point against Whitehead, or rather an interpretation of Whitehead, that, say, molecules do not literally enjoy color sensa. Granted, there is some controversy as to whether Whitehead actually held that low-grade entities enjoy color sensa. For discussions of this issue see Hartshorne’s ‘The Interpretation of Whitehead (Reply to John W. Blyth)," The Philosophical Review 48 (July 1939): 415-23, and David Ray Griffin, "Hartshorne’s Differences from Whitehead" in Two Process Philosophers: Hartshorne’s Encounter With Whitehead, ed. by Lewis S. Ford (American Academy of Religion, 1973), pp. 40-45.

9Prof. Rudolf Rucker presents an Argument from Reflection for the inconceivability of the Absolute Infinite as the set of all ordinals and maintains a moderate form, of set-theoretic realism most amenable to Hartshorne’s view of general eternal objects. Infinite sets less than the Absolute Infinite share definite mathematical properties with the finite integers (e.g., regularity) and are subject to clear criteria for set-generation (unlike the Absolute). Thus, Cantor’s actual infinites can be accepted while rejecting the existence of the Absolute Infinite. See R. v. B. Rucker, Infinity and the Mind: The Science and Philosophy of the Infinite (Birkhauser Boston, 1982), pp. 206-11, 276-80.

10E W. Hall, "Of What Use Are Whitehead’s Eternal Objects?," in Alfred North Whitehead: Essays on His Philosophy (Prentice-Hall, 1963), pp. 102-16. Whiteheadians who support the doctrine of eternal objects disagree sharply. Although Professor Ford has more recently changed his mind on this issue, a good statement of reasons favoring Whitehead’s view can be found in Lewis S. Ford, "Whitehead’s Differences From Hartshorne" in Two Process Philosophers, pp. 58-65.

11 I take this view to be implied in the following passage: "Consider the phrase ‘greatest possible number.’ It, too, can be smoothly uttered; but does it say anything? It might be used to define infinity; but I am not aware of any mathematician who has thought this a good definition. There are in standard mathematics many infinities unequal to one another, but no highest infinity." Hartshorne, Omnipotence and Other Theological Mistakes (SUNY Ness, 1984), p. 7.

12Even if Creel is not compelled to place all possibles on a single continuum, the algorithm approach will fail for all non-recursive formal mathematical possibles and all possible aesthetic constellations, for the reasons already stated.

13This misunderstanding carries over into Creel’s discussion of Hartshorne’s emotional passibilism as I argue in Sec. III. But he also misunderstands the process position on the efficacy of actual occasions (DI 182). The concrescence of new actual occasions is initiated by past occasions, which are effective precisely because their subjective immediacy, their career of concrescence, has perished. Yet on something like Creel’s view that God sustains each individual (see DI 183-87), Hartshorne holds that God provides each actual occasion its total or essential object of prehension, thus constituting as well as altering each actual occasion (see his The Divine Relativity [Yale University Press, 1948], pp. 134-42, esp. 139.)

14Anthony Kenny, The God of the Philosophers (Oxford University Press, 1979), p. 32. Interestingly, Kenny defends this position with the notion that God’s knowledge is comparable to the knowledge that an architect has of a house by virtue of knowing its design. Kenny’s view is therefore much like Creel’s algorithm approach, which I have rejected.

Viewed 8805 times.