Your continued exploration of this tricky topic has been indispensable in clarifying the crucial roles/functions of eternal objects and potentiality in Whitehead’s scheme, which has certainly refined my thinking. However, I remain unsatisfied with eternal objects and believe Whitehead’s scheme can survive largely intact by carefully reframing potentiality.
“Eternal objects, considered in abstraction from actualities, carry no coordinates, no metric, none of the pre-given structure that would make them a state space. Pure potentiality is not a dimension or a metrical order.”
As I understand it, your argument for the adequacy of eternal objects as a conception of pure potentiality (in spite of the commitments you affirm from Kauffman, Juarrero, and Jackson) rests heavily on your ability to insulate their definiteness from their position in some space, whether that be the extensive continuum, a Platonic realm of forms, or any other pre-given state-space. But as I argued in our recent exchange on Facebook, I think that effort falls flat within Whitehead’s scheme precisely because eternal objects are ordered by the primordial nature of God: “Such a primordial superject of creativity achieves, in its unity of satisfaction, the complete conceptual valuation of all eternal objects. This is the ultimate, basic adjustment of the togetherness of eternal objects on which creative order depends. It is the conceptual adjustment of all appetites in the form of aversions and adversions. It constitutes the meaning of relevance.” This positions eternal objects within a relevance landscape that licensed Whitehead to drop his category of conceptual reversion while still allowing concrescing occasions to conceptually prehend relevant eternal objects (e.g., the adjacent possible) and introduce novelty while maintaining coherence and intelligibility in the actual world process. It is true that this relevance landscape is not a metrical space (i.e., there is no meaningful distance function), but it is certainly a topological space that carries notions of adjacency and relational structure. And so the questions remain: Can the definiteness of eternal objects be disentangled from their relational position within this primordial valuation, which serves as a sort of pre-given state-space? What could we possibly mean by “definite” without the implicit notions of contrasts, boundaries, or properties provided by relational forms within some relational context, be it abstract or concrete?
“Whether graded relevance is historical construction all the way down, the result of coarse-graining with no ultimate ground, or an uncreated condition of evolutionary construction, is a question that will no doubt continue to linger. But I cannot myself make sense of the former possibility. “Effective relevance requires agency of comparison” (PR 31). Without God, possibilities that transcended realized temporal matter of fact would remain relatively nonexistent for a given finite occasion.”
I just finished expressing skepticism that one can insulate the definiteness of eternal objects from their position in a pre-defined state-space in Whitehead’s scheme. This is a direct consequence of Whitehead’s own answer to the question, “how, and in what sense, one unrealized eternal object can be more, or less, proximate to an eternal object in realized ingression” (PR 249), which was God’s primordial valuation/ordering of eternal objects. It seems to me that the perceived need for an agency of comparison arises precisely because eternal objects are first conceived as definite independently of their relational context. So, it seems your proposed justifications for releasing these two hostages may work against each other. The agency of God is invoked to explain how definite eternal objects could attain relevance to each other, but the achievement of that relevance landscape makes it difficult (if not impossible) to insulate their definiteness from their position within a pre-given state-space without treating definiteness as something that is intelligible independently of the very relational context that is supposed to render eternal objects relevant.
My paper-in-progress builds on these and related arguments to explore whether Whitehead's own relational insights can be extended into potentiality without appealing to eternal objects as pre-definite forms. I hope to have something worth sharing in the coming months.
I look forward to reading what you’re working on! These are all very live issues of interpretation, so I’m grateful for your sustained attention to my attempt to navigate them. For me, everything turns on whether God’s primordial valuation is a finished map that occasions just read off, or whether it constitutes a continually enriched gradation of intensive relevance, such that each novel occasion is not only met with a uniquely situationally graded lure, but contributes its decision back to tilt the relevance function for subsequent occasions.
Were it a finished map of adjacent relevance, occasions would be entirely drained of self-creative agency. They’d just be epiphenomena consulting the map, incapable of ever venturing off road. But Whitehead insists that God’s is not the only act of valuation, that every concrescence is a self-creating subject-superject (though divine concrescence is perhaps better described as a superject-subject). Relevance realization is always novelly enacted, secured not by pre-stated metrical proximities between EOs but by the fresh provision of an initial aim relevant to a once-occurrent actuality’s particular worldly situation. The PNG is the unconditioned ground of relevance, no way around that; but its completeness is the completeness of appetition, not of realization.
I grant (with Ben Snyder) that the eternal objects carry mereotopological relations of extensive connection. But I am trying at least to interpret extensive connection as, indeed, an order of extension, not of relevance. It tells you how regions may connect but it does not tell you which EOs will matter to particular actual occasions. Relevance is not a relation among EOs, after all, but between them and actualities.
My sense so far is that granting an envisaged mereotopology does not require granting a state space of pre-determined outcomes. What is primordially envisaged are the generic non-metrical requirements of togetherness, not the unprestatable achievements of value-experience that constitute the creative advance. The latter become part of divine concrescence only consequently.
Thanks, Matt. I really appreciate the thoughtful response. I think we're much closer than it might initially appear in how we understand the dynamic character of relevance. I completely agree that if God's primordial valuation were a finished map that occasions merely consulted, it would drain concrescence of genuine creativity. Your emphasis on each occasion contributing its decision back into the consequent nature of God and thereby continually enriching the relevance landscape resonates strongly with my own thinking about the emergence and evolution of dynamic value landscapes.
I suspect, though, that part of our remaining disagreement may come down to what each of us means by "relevance." In your reply, I take relevance primarily to mean the situational significance or lure that particular eternal objects have for a given concrescence. My concern has been a bit broader. I was also trying to get at the more abstract formal relations that make it meaningful to speak of one possibility as adjacent to another in the first place. That is why I brought up the primordial valuation as introducing a kind of topological ordering, even if not a metrical one. My question isn't whether God pre-determines the creative advance, but whether the very intelligibility and definiteness of eternal objects can really be separated from their position within that relational ordering.
Relatedly, I'd be interested to hear you say a little more about what you mean when you write that "the PNG is the unconditioned ground of relevance... its completeness is the completeness of appetition, not of realization." I find that a very suggestive way of putting it, but I'm not yet sure I understand what work "appetition" is doing there. Is primordial appetition itself already a form of valuation that orders eternal objects, or do you see it as something prior even to that?
I ask because I continue to find myself circling the same question: what exactly constitutes the definiteness of an eternal object for Whitehead and your reading of him? If its relevance is continually and creatively realized in relation to actual occasions (which I happily grant), then in what sense is its definiteness already there? Is definiteness itself intelligible apart from the relational ordering established in the primordial valuation, or is that ordering already ingredient in what it means for an eternal object to be definite? I suspect that question marks the deepest point at which my own more thoroughly relational/mereotopological conception of potentiality begins to diverge from Whitehead's.
I would push back against the claim summarized in your second-to-last paragraph. Just to cover some basic ground first to make sure we’re all on the same page: Complex eternal objects have structure, in the sense that they consist of a relation (a pattern) holding among constituent eternal objects. They are thus a way an actual entity may possibly be structurally characterized (i.e., characterizing a relation among said actual entity’s parts). Whitehead doesn’t really specify exactly how such a complex eternal object ingresses in a structured manner, and there’s a lot of debate over how structural universals are instantiated—so this is a contentious concept to appeal to in the philosophy of science (and one Whitehead scholars are not clear about), but it’s not an abnormal concept to appeal to. The key here, though, seems to be that such structural universals do exist for Whitehead. Since they are eternal objects, they exist as uncreated and eternal forms. However, for Whitehead himself the patterns involved in a complex eternal object do not seem to be exclusively extensive relations. Extensive relations characterize the structure of objective data. Whitehead also recognizes patterns that characterize subjective forms of feeling. There are thus structured ways that agents take on attitudes in response to felt data. One crucial passage is on page 233 of PR, where he discusses a subjective form involving its “qualitative pattern”: “the relative intensities of the qualitative elements in the qualitative pattern are among the relational factors which constitute that qualitative pattern. Also conversely, there are qualitative relations among the qualitative elements and they constitute an abstract qualitative pattern for the qualitative relations.” What exactly he has in mind is not spelled out: I could imagine at least relations comparing degree of intensity along the lines of “I believe this more than that” or “I enjoy this more than that.” So Whitehead at least recognizes complex eternal objects beyond those characterizing extensive phenomena. There’s a deeper issue, though, to even trying to argue that God—or any other such concrete acts of valuation—is capable of introducing structure among eternal objects in a way that did not exist even in abstract form before, which is that it is counter to the entire reason that Whitehead has universals: there is no explanation of what we mean by a subjective form having a property or relation other than that said subjective form ingresses an eternal object. If we feel the need to introduce a new way of explaining what it means for actual entities to have properties and relations, then economy suggests we don’t need the eternal objects to do any of that work at all (or else, we need more of an explanation of the difference requiring both kinds of explanation). Of course, there are still differences between universals that have not been instantiated, and a concrete particular that is instantiating a universal. For Whitehead, the latter is a feeling that has value, the former is not: the latter has an essence that is contingently created, whereas the former does not. I think that’s all that’s really needed for the relevant distinction.
Ben, these are some very helpful points, and I think they sharpen rather than weaken my concern about definiteness prior to (or independent of) relational structure. As you pointed out, once we start discussing complex eternal objects we very naturally find ourselves talking about multiple modalities of relation: metrical, extensive, qualitative, valuational, and likely others. One reason I've found topology and category theory to be helpful is that they provide conceptual resources for thinking about heterogeneous relational structures without reducing them to a single kind of relation. Whitehead already recognized that extension is topological rather than metrical, but he was writing at a moment in history when topology was just beginning to mature and decades before category theory provided an even richer language for relational organization more generally. I keep wondering whether those later conceptual developments might have sharpened and extended his relational insights into potentiality itself.
Regardless, for me the central question remains the same. If eternal objects include varied and increasingly complex patterns of extensive, qualitative, and other relations, then what exactly constitutes the definiteness of those relational patterns themselves? I want to push on whether definiteness can ultimately be made intelligible apart from relational organization. That's one of the directions I'm hoping to develop more carefully in my paper.
Sorry to be providing so much criticism of your Whitehead readings lately (I guess I have to fit them all in before you head off to the Whitehead conference). I think this is a good example of where contemporary philosophy of language and analytic metaphysics can help clarify Whitehead’s theory and its relevance to contemporary philosophy of science.
What would it mean for propositions to be “logical sentences”? Propositions the abstract objects ambiguously expressed by declarative sentences; this much is very familiar in analytic philosophy. It is also not distinctive of Whitehead to have propositions be “lures for feeling”: the analytic way of saying that would be, that propositions are the possible content of propositional attitudes.
Propositions being “hybrid” entities, composed of particular and abstract entities, is what is called a Russellian structural theory. The further consequence of this, that singular propositions contingently exist when they have contingent particulars as parts, is what is often called (following Plantinga) “existentialism.” This existentialism is the relevant thesis you’re utilizing here, though I’m skeptical of how you are utilizing it.
The crucial point is that, by Whitehead's reasoning, propositions are only nonexistent in a world when referring to particular actual entities that don’t exist in that world. The upshot is that all purely general propositions thus still exist and are meaningful/statable, as are higher-order propositions only naming properties and relations (insofar as we interpret these to be referring to eternal objects). Thus, whatever can’t be named directly could still be described (i.e., you could say, there is some x such that x has traits P, Q, and R and x is …).
Thus the problem with your example: “That feathers might become flight-worthy requires a world in which there are feathered reptiles of a certain body mass, a particular atmospheric density, etc.” This is a world we are capable of describing in terms of properties and relations that are themselves universals. Your example, “feathers are for flight”, does not really seem to be a statement naming any particular individuals. It is naming properties, namely, properties of an organism’s body parts (being a feather) and its actions (flying). Here, of course, is where your interpretation of traits as societies comes in, arguing that this type of description does, in fact, involve naming particulars, but to be blunt I do not think this argument makes sense.
Societies are concrete particulars for Whitehead. They are types of nexuus, where the members of the nexus are instantiating a common eternal object, the “defining characteristic.” I’d think biologists would generally want to be able to distinguish between a trait in general and a particular organism’s instantiation of said trait. Insofar as they are doing so, this is precisely the distinction between an eternal object and a society. Since we can talk of the convergent evolution of traits, we can talk about particular organisms sharing traits independently of their causal connection: a clear sign we are discussing properties being instantiated over distinct space-time events, and not a particular instantiation of a property at a particular space-time event. Much the same thing seems to apply to the types of behavior we can in turn consider a trait functioning for: these are _types_ of behavior we are describing in general, not particular instantiations of them.
The upshot of all of this, I believe, is that this approach is not really consonant with the kind of platonic realism of universals that Whitehead has. Once one admits universals into one's ontology, all of these traits and behaviors become describable in abstraction from any particular instances. You do not make it clear that we are losing anything relevant in our statements for a scientific theory when we move from discussing particular societies to discussing, in general, types of societies described in terms of their forms.
Whitehead uses the term “definiteness” primarily to refer to an attribute of actual entities: it is “the illustration of select eternal objects” (PR 25). An eternal object is thus a form of definiteness in the sense of being a form that contributes to an actual entity’s definiteness. Thus, contrary to what I take your reading to be, the eternal objects are the forms and not the definiteness. To put it even more specifically, the eternal objects necessarily are the forms that they are, whereas they contribute contingently to the definiteness of actual entities.
There is “pre-given structure” for eternal objects, since Whitehead has a theory of complex eternal objects, i.e., an eternal object that itself consists of a pattern of other eternal objects and may be ingressed as one complex structure. This is also familiar to more contemporary metaphysics as a theory of “structural universals.” Whitehead has eternal objects that serve the role of “Platonic mathematical forms”, which are thus the basis for his theory of extension and provide the geometrical structure involved in complex eternal objects. These are the regions and the binary relation of connection elaborated in his mereotopology: other simple eternal objects, sensa, can then be located at regions and thereby form into complex ordered forms. That the extensive continuum is in part derivative from the actual world does not mean the geometrical forms of structure are themselves contingently created, but that the actual world contingently ingresses some structure rather than others (out of a given realm of possible forms of structure).
There are thus many aspects of this argument I’d want to see given some more detailed or made more convincing—and not just in terms of being an accurate reading of Whitehead. What is the criterion for a proposition’s statability? How do propositions about traits fall under the conditions of thus not being statable? In what sense are eternal objects capable of being structured, and in what ways is said structure only contingently derived from the actual world?
No apology needed, Ben. You put me to work revising this post in a way that I feel improves its clarity immensely. So once again, I am indebted to your incisive critiques. You'll find your fingerprints all over my revision. Especially important was your demand for more clarity around the type/token distinction I had unhelpfully left submerged.
A few further reflections:
Any world we can imagine is describable in terms of universals. But as you know a general description lacking indicated actual subjects is not a proposition in Whitehead's sense. It tells no tales. What I am arguing are undeterminable in advance are novel propositions, the lure of new functions, not abstract descriptions.
Also, I just can't grant that Whitehead's account of propositions as "lures for feeling" is reducible without remainder to the possible contents of propositional attitudes. But this may be my ignorance. Are there analytic philosophers who accept that an oyster entertains propositions? Do any argue that propositions did pre-cognitive evolutionary work long before human judgers, even long before vertebrate nervous systems? I read some Millikan in undergrad, so I'm aware of teleosemantics; but I reject strong reprsentationalism for the same reason Whitehead rejects it (Whitehead subordinates representation to direct feeling rather than grounding feeling in representation). Seems to me that Whitehead's is not just a notational variant of analytic propositional attitudes but part of a radically new philosophy of nature.
Finally, regarding what we lose scientifically if we move from particular societies to generic types, I agree that, descriptively, nothing is lost. But this is part of my point. "Unprestatable" is Kauffman's term and it admittedly suggests the descriptive account. But I think he really means more broadly that biological novelty is unpredictable in the sense that the space of relevant possibilities cannot be proposed in advance, since it is generated by each exaptation event. So this is not a proposal about possible description but about actual generation. Re-description of an achieved function in general terms after the fact does not retroject a pre-established phase space before it.
"But as you know a general description lacking indicated actual subjects is not a proposition in Whitehead's sense." What you are not accounting for are general propositions, i.e., propositions that do not refer to any particular individuals but make statements in terms of some or all individuals. Whitehead is explicit that singular propositions that lack meaning for some entertaining subjects can be replaced by general propositions with a similar meaning that will be meaningful for said entertaining subjects. This is an insight taken from Russell's work on descriptions. This is what Whitehead is referring to in the following quote: "But there are propositions of a† more general type, for which ‘Caesar’ and ‘Rubicon’ have more generalized, vaguer meanings. In these vaguer meanings, ‘Caesar’ and ‘Rubicon’ indicate the entities, if any, located by any one member of a type of routes, starting from a certain type of inference and conjecture." I think this is a very important detail to understand for your thesis. I'm working on a paper more or less touching on this, hopefully I'll be able to share it sometime soon.
Philosophers are going to differ over what they would recognize as agents that genuinely take on intentional mental states and thus have propositional attitudes. I do not think this alters that Whitehead's theory of propositional feelings involves subjective forms playing the role that's typically identified as propositional attitudes, which is then the crucial thing to consider in terms of being able to refer to the relevant work already done in the philosophy of language.
I think you’re missing it and still treating possibility as something that requires its own ontology. That’s where I find myself taking a different path.
The central move of your essay seems to be:
Novelty requires propositions (“lures”), propositions require eternal objects, and eternal objects require God to grade relevance.
My question is: why assume possibility itself is an ontological inventory?
From my perspective, possibility isn’t a catalogue waiting to be selected. It emerges from what remains consistent with reality.
For example, you write:
“Error is the price we pay for progress.”
I’d put it differently:
Error is the price we pay for exploration under incomplete knowledge. Progress comes from calibration, not error itself.
Error isn’t creative. Calibration is.
Likewise, you argue that metaphysics requires eternal objects because we can think about unrealized possibilities. I’m not convinced the ability to negate requires an ontological realm of unreal entities. It requires only a system capable of representing states that are compatible or incompatible with reality. That seems a much weaker commitment.
I wonder whether this conflates possibility with value.
God need not be understood as maintaining an inventory of possibilities. God can instead be understood as the ground of the reality within which possibilities become meaningful. Reality itself constrains what is possible.
That’s why I wonder whether Whitehead, Lewis, and much of modal metaphysics begin one step too late. Rather than asking,
What is the ontology of possibility?
I’d ask:
What makes any possibility physically realizable while excluding infinitely many others?
To me, that’s a more fundamental question.
On that view, novelty isn’t reality retrieving something from a warehouse of eternal forms. It’s a genuinely new stable organization becoming possible because the evolving world has reached a configuration where that organization no longer violates the underlying constraints.
In other words:
Novelty is discovered, not stored. It is licensed by reality, not retrieved from a catalogue.
That’s why I wonder whether eternal objects are actually doing explanatory work, or simply relocating the mystery. If coherence, recurrence, and constraint are sufficient to explain why stable forms emerge and persist, what explanatory role remains that requires an ontology of eternal objects?
I am not arguing that possibility requires its own ontology. The ultimate category here is Creativity, a formless productivity or Ungrund (Böhme, Schelling), sheer indeterminacy rather than any stock of possible forms. Determinate actualities are not drawn from it as if thawed from a freezer but cut out of it by their own becoming. Whitehead's name for that self-cutting or decision by which an actuality is achieved is concrescence, of which God’s primordial valuation is the non-temporal instance, conditioning every temporal cut without preceding or determining it in advance.
As I wrote in the essay you’re replying to, eternal objects carry no coordinates, no metric, none of the pre-given structure that would make them a state space. Any coordinate/state spaces that come to be are built and rebuilt through actualizations, not imposed from above by pre-determined possibilities.
On my view, coherence and recurrence already presuppose the form of definiteness. To be inherited, recognized, contrasted, or tracked, something must be recurring. Definite potentials that which recurs. The point is not that a determinate form was stored somewhere before its emergence, but that emergent determinacy cannot be made intelligible as such unless there is some non-actual pole of definiteness involved. Note that the actual and potential are poles not separate worlds. There is one world, but the historic becoming of that world involves the finite actualization of infinite possibilities. Those possibilities are not “somewhere else.” They are in the infinite actuality, God; and precisely as you say, God is the ground within which possibilities become meaningful.
Regarding error as the price of progress, Whitehead is saying that non-conformal or false propositions, precisely because they are active throughout the living world and not merely the logical judgments of we speaking animals, can lure organismic feelings beyond conformity to the already actualized. “Error” is the risk internal to adventurous novel comparisons/contrasts. As far as I can tell, your idea of calibration works only if we are talking about an already constituted epistemic agent correcting its world- and/or self-model. I am after a more basic ontological question.
Your continued exploration of this tricky topic has been indispensable in clarifying the crucial roles/functions of eternal objects and potentiality in Whitehead’s scheme, which has certainly refined my thinking. However, I remain unsatisfied with eternal objects and believe Whitehead’s scheme can survive largely intact by carefully reframing potentiality.
“Eternal objects, considered in abstraction from actualities, carry no coordinates, no metric, none of the pre-given structure that would make them a state space. Pure potentiality is not a dimension or a metrical order.”
As I understand it, your argument for the adequacy of eternal objects as a conception of pure potentiality (in spite of the commitments you affirm from Kauffman, Juarrero, and Jackson) rests heavily on your ability to insulate their definiteness from their position in some space, whether that be the extensive continuum, a Platonic realm of forms, or any other pre-given state-space. But as I argued in our recent exchange on Facebook, I think that effort falls flat within Whitehead’s scheme precisely because eternal objects are ordered by the primordial nature of God: “Such a primordial superject of creativity achieves, in its unity of satisfaction, the complete conceptual valuation of all eternal objects. This is the ultimate, basic adjustment of the togetherness of eternal objects on which creative order depends. It is the conceptual adjustment of all appetites in the form of aversions and adversions. It constitutes the meaning of relevance.” This positions eternal objects within a relevance landscape that licensed Whitehead to drop his category of conceptual reversion while still allowing concrescing occasions to conceptually prehend relevant eternal objects (e.g., the adjacent possible) and introduce novelty while maintaining coherence and intelligibility in the actual world process. It is true that this relevance landscape is not a metrical space (i.e., there is no meaningful distance function), but it is certainly a topological space that carries notions of adjacency and relational structure. And so the questions remain: Can the definiteness of eternal objects be disentangled from their relational position within this primordial valuation, which serves as a sort of pre-given state-space? What could we possibly mean by “definite” without the implicit notions of contrasts, boundaries, or properties provided by relational forms within some relational context, be it abstract or concrete?
“Whether graded relevance is historical construction all the way down, the result of coarse-graining with no ultimate ground, or an uncreated condition of evolutionary construction, is a question that will no doubt continue to linger. But I cannot myself make sense of the former possibility. “Effective relevance requires agency of comparison” (PR 31). Without God, possibilities that transcended realized temporal matter of fact would remain relatively nonexistent for a given finite occasion.”
I just finished expressing skepticism that one can insulate the definiteness of eternal objects from their position in a pre-defined state-space in Whitehead’s scheme. This is a direct consequence of Whitehead’s own answer to the question, “how, and in what sense, one unrealized eternal object can be more, or less, proximate to an eternal object in realized ingression” (PR 249), which was God’s primordial valuation/ordering of eternal objects. It seems to me that the perceived need for an agency of comparison arises precisely because eternal objects are first conceived as definite independently of their relational context. So, it seems your proposed justifications for releasing these two hostages may work against each other. The agency of God is invoked to explain how definite eternal objects could attain relevance to each other, but the achievement of that relevance landscape makes it difficult (if not impossible) to insulate their definiteness from their position within a pre-given state-space without treating definiteness as something that is intelligible independently of the very relational context that is supposed to render eternal objects relevant.
My paper-in-progress builds on these and related arguments to explore whether Whitehead's own relational insights can be extended into potentiality without appealing to eternal objects as pre-definite forms. I hope to have something worth sharing in the coming months.
I look forward to reading what you’re working on! These are all very live issues of interpretation, so I’m grateful for your sustained attention to my attempt to navigate them. For me, everything turns on whether God’s primordial valuation is a finished map that occasions just read off, or whether it constitutes a continually enriched gradation of intensive relevance, such that each novel occasion is not only met with a uniquely situationally graded lure, but contributes its decision back to tilt the relevance function for subsequent occasions.
Were it a finished map of adjacent relevance, occasions would be entirely drained of self-creative agency. They’d just be epiphenomena consulting the map, incapable of ever venturing off road. But Whitehead insists that God’s is not the only act of valuation, that every concrescence is a self-creating subject-superject (though divine concrescence is perhaps better described as a superject-subject). Relevance realization is always novelly enacted, secured not by pre-stated metrical proximities between EOs but by the fresh provision of an initial aim relevant to a once-occurrent actuality’s particular worldly situation. The PNG is the unconditioned ground of relevance, no way around that; but its completeness is the completeness of appetition, not of realization.
I grant (with Ben Snyder) that the eternal objects carry mereotopological relations of extensive connection. But I am trying at least to interpret extensive connection as, indeed, an order of extension, not of relevance. It tells you how regions may connect but it does not tell you which EOs will matter to particular actual occasions. Relevance is not a relation among EOs, after all, but between them and actualities.
My sense so far is that granting an envisaged mereotopology does not require granting a state space of pre-determined outcomes. What is primordially envisaged are the generic non-metrical requirements of togetherness, not the unprestatable achievements of value-experience that constitute the creative advance. The latter become part of divine concrescence only consequently.
Thanks, Matt. I really appreciate the thoughtful response. I think we're much closer than it might initially appear in how we understand the dynamic character of relevance. I completely agree that if God's primordial valuation were a finished map that occasions merely consulted, it would drain concrescence of genuine creativity. Your emphasis on each occasion contributing its decision back into the consequent nature of God and thereby continually enriching the relevance landscape resonates strongly with my own thinking about the emergence and evolution of dynamic value landscapes.
I suspect, though, that part of our remaining disagreement may come down to what each of us means by "relevance." In your reply, I take relevance primarily to mean the situational significance or lure that particular eternal objects have for a given concrescence. My concern has been a bit broader. I was also trying to get at the more abstract formal relations that make it meaningful to speak of one possibility as adjacent to another in the first place. That is why I brought up the primordial valuation as introducing a kind of topological ordering, even if not a metrical one. My question isn't whether God pre-determines the creative advance, but whether the very intelligibility and definiteness of eternal objects can really be separated from their position within that relational ordering.
Relatedly, I'd be interested to hear you say a little more about what you mean when you write that "the PNG is the unconditioned ground of relevance... its completeness is the completeness of appetition, not of realization." I find that a very suggestive way of putting it, but I'm not yet sure I understand what work "appetition" is doing there. Is primordial appetition itself already a form of valuation that orders eternal objects, or do you see it as something prior even to that?
I ask because I continue to find myself circling the same question: what exactly constitutes the definiteness of an eternal object for Whitehead and your reading of him? If its relevance is continually and creatively realized in relation to actual occasions (which I happily grant), then in what sense is its definiteness already there? Is definiteness itself intelligible apart from the relational ordering established in the primordial valuation, or is that ordering already ingredient in what it means for an eternal object to be definite? I suspect that question marks the deepest point at which my own more thoroughly relational/mereotopological conception of potentiality begins to diverge from Whitehead's.
I would push back against the claim summarized in your second-to-last paragraph. Just to cover some basic ground first to make sure we’re all on the same page: Complex eternal objects have structure, in the sense that they consist of a relation (a pattern) holding among constituent eternal objects. They are thus a way an actual entity may possibly be structurally characterized (i.e., characterizing a relation among said actual entity’s parts). Whitehead doesn’t really specify exactly how such a complex eternal object ingresses in a structured manner, and there’s a lot of debate over how structural universals are instantiated—so this is a contentious concept to appeal to in the philosophy of science (and one Whitehead scholars are not clear about), but it’s not an abnormal concept to appeal to. The key here, though, seems to be that such structural universals do exist for Whitehead. Since they are eternal objects, they exist as uncreated and eternal forms. However, for Whitehead himself the patterns involved in a complex eternal object do not seem to be exclusively extensive relations. Extensive relations characterize the structure of objective data. Whitehead also recognizes patterns that characterize subjective forms of feeling. There are thus structured ways that agents take on attitudes in response to felt data. One crucial passage is on page 233 of PR, where he discusses a subjective form involving its “qualitative pattern”: “the relative intensities of the qualitative elements in the qualitative pattern are among the relational factors which constitute that qualitative pattern. Also conversely, there are qualitative relations among the qualitative elements and they constitute an abstract qualitative pattern for the qualitative relations.” What exactly he has in mind is not spelled out: I could imagine at least relations comparing degree of intensity along the lines of “I believe this more than that” or “I enjoy this more than that.” So Whitehead at least recognizes complex eternal objects beyond those characterizing extensive phenomena. There’s a deeper issue, though, to even trying to argue that God—or any other such concrete acts of valuation—is capable of introducing structure among eternal objects in a way that did not exist even in abstract form before, which is that it is counter to the entire reason that Whitehead has universals: there is no explanation of what we mean by a subjective form having a property or relation other than that said subjective form ingresses an eternal object. If we feel the need to introduce a new way of explaining what it means for actual entities to have properties and relations, then economy suggests we don’t need the eternal objects to do any of that work at all (or else, we need more of an explanation of the difference requiring both kinds of explanation). Of course, there are still differences between universals that have not been instantiated, and a concrete particular that is instantiating a universal. For Whitehead, the latter is a feeling that has value, the former is not: the latter has an essence that is contingently created, whereas the former does not. I think that’s all that’s really needed for the relevant distinction.
Ben, these are some very helpful points, and I think they sharpen rather than weaken my concern about definiteness prior to (or independent of) relational structure. As you pointed out, once we start discussing complex eternal objects we very naturally find ourselves talking about multiple modalities of relation: metrical, extensive, qualitative, valuational, and likely others. One reason I've found topology and category theory to be helpful is that they provide conceptual resources for thinking about heterogeneous relational structures without reducing them to a single kind of relation. Whitehead already recognized that extension is topological rather than metrical, but he was writing at a moment in history when topology was just beginning to mature and decades before category theory provided an even richer language for relational organization more generally. I keep wondering whether those later conceptual developments might have sharpened and extended his relational insights into potentiality itself.
Regardless, for me the central question remains the same. If eternal objects include varied and increasingly complex patterns of extensive, qualitative, and other relations, then what exactly constitutes the definiteness of those relational patterns themselves? I want to push on whether definiteness can ultimately be made intelligible apart from relational organization. That's one of the directions I'm hoping to develop more carefully in my paper.
Sorry to be providing so much criticism of your Whitehead readings lately (I guess I have to fit them all in before you head off to the Whitehead conference). I think this is a good example of where contemporary philosophy of language and analytic metaphysics can help clarify Whitehead’s theory and its relevance to contemporary philosophy of science.
What would it mean for propositions to be “logical sentences”? Propositions the abstract objects ambiguously expressed by declarative sentences; this much is very familiar in analytic philosophy. It is also not distinctive of Whitehead to have propositions be “lures for feeling”: the analytic way of saying that would be, that propositions are the possible content of propositional attitudes.
Propositions being “hybrid” entities, composed of particular and abstract entities, is what is called a Russellian structural theory. The further consequence of this, that singular propositions contingently exist when they have contingent particulars as parts, is what is often called (following Plantinga) “existentialism.” This existentialism is the relevant thesis you’re utilizing here, though I’m skeptical of how you are utilizing it.
The crucial point is that, by Whitehead's reasoning, propositions are only nonexistent in a world when referring to particular actual entities that don’t exist in that world. The upshot is that all purely general propositions thus still exist and are meaningful/statable, as are higher-order propositions only naming properties and relations (insofar as we interpret these to be referring to eternal objects). Thus, whatever can’t be named directly could still be described (i.e., you could say, there is some x such that x has traits P, Q, and R and x is …).
Thus the problem with your example: “That feathers might become flight-worthy requires a world in which there are feathered reptiles of a certain body mass, a particular atmospheric density, etc.” This is a world we are capable of describing in terms of properties and relations that are themselves universals. Your example, “feathers are for flight”, does not really seem to be a statement naming any particular individuals. It is naming properties, namely, properties of an organism’s body parts (being a feather) and its actions (flying). Here, of course, is where your interpretation of traits as societies comes in, arguing that this type of description does, in fact, involve naming particulars, but to be blunt I do not think this argument makes sense.
Societies are concrete particulars for Whitehead. They are types of nexuus, where the members of the nexus are instantiating a common eternal object, the “defining characteristic.” I’d think biologists would generally want to be able to distinguish between a trait in general and a particular organism’s instantiation of said trait. Insofar as they are doing so, this is precisely the distinction between an eternal object and a society. Since we can talk of the convergent evolution of traits, we can talk about particular organisms sharing traits independently of their causal connection: a clear sign we are discussing properties being instantiated over distinct space-time events, and not a particular instantiation of a property at a particular space-time event. Much the same thing seems to apply to the types of behavior we can in turn consider a trait functioning for: these are _types_ of behavior we are describing in general, not particular instantiations of them.
The upshot of all of this, I believe, is that this approach is not really consonant with the kind of platonic realism of universals that Whitehead has. Once one admits universals into one's ontology, all of these traits and behaviors become describable in abstraction from any particular instances. You do not make it clear that we are losing anything relevant in our statements for a scientific theory when we move from discussing particular societies to discussing, in general, types of societies described in terms of their forms.
Whitehead uses the term “definiteness” primarily to refer to an attribute of actual entities: it is “the illustration of select eternal objects” (PR 25). An eternal object is thus a form of definiteness in the sense of being a form that contributes to an actual entity’s definiteness. Thus, contrary to what I take your reading to be, the eternal objects are the forms and not the definiteness. To put it even more specifically, the eternal objects necessarily are the forms that they are, whereas they contribute contingently to the definiteness of actual entities.
There is “pre-given structure” for eternal objects, since Whitehead has a theory of complex eternal objects, i.e., an eternal object that itself consists of a pattern of other eternal objects and may be ingressed as one complex structure. This is also familiar to more contemporary metaphysics as a theory of “structural universals.” Whitehead has eternal objects that serve the role of “Platonic mathematical forms”, which are thus the basis for his theory of extension and provide the geometrical structure involved in complex eternal objects. These are the regions and the binary relation of connection elaborated in his mereotopology: other simple eternal objects, sensa, can then be located at regions and thereby form into complex ordered forms. That the extensive continuum is in part derivative from the actual world does not mean the geometrical forms of structure are themselves contingently created, but that the actual world contingently ingresses some structure rather than others (out of a given realm of possible forms of structure).
There are thus many aspects of this argument I’d want to see given some more detailed or made more convincing—and not just in terms of being an accurate reading of Whitehead. What is the criterion for a proposition’s statability? How do propositions about traits fall under the conditions of thus not being statable? In what sense are eternal objects capable of being structured, and in what ways is said structure only contingently derived from the actual world?
No apology needed, Ben. You put me to work revising this post in a way that I feel improves its clarity immensely. So once again, I am indebted to your incisive critiques. You'll find your fingerprints all over my revision. Especially important was your demand for more clarity around the type/token distinction I had unhelpfully left submerged.
A few further reflections:
Any world we can imagine is describable in terms of universals. But as you know a general description lacking indicated actual subjects is not a proposition in Whitehead's sense. It tells no tales. What I am arguing are undeterminable in advance are novel propositions, the lure of new functions, not abstract descriptions.
Also, I just can't grant that Whitehead's account of propositions as "lures for feeling" is reducible without remainder to the possible contents of propositional attitudes. But this may be my ignorance. Are there analytic philosophers who accept that an oyster entertains propositions? Do any argue that propositions did pre-cognitive evolutionary work long before human judgers, even long before vertebrate nervous systems? I read some Millikan in undergrad, so I'm aware of teleosemantics; but I reject strong reprsentationalism for the same reason Whitehead rejects it (Whitehead subordinates representation to direct feeling rather than grounding feeling in representation). Seems to me that Whitehead's is not just a notational variant of analytic propositional attitudes but part of a radically new philosophy of nature.
Finally, regarding what we lose scientifically if we move from particular societies to generic types, I agree that, descriptively, nothing is lost. But this is part of my point. "Unprestatable" is Kauffman's term and it admittedly suggests the descriptive account. But I think he really means more broadly that biological novelty is unpredictable in the sense that the space of relevant possibilities cannot be proposed in advance, since it is generated by each exaptation event. So this is not a proposal about possible description but about actual generation. Re-description of an achieved function in general terms after the fact does not retroject a pre-established phase space before it.
"But as you know a general description lacking indicated actual subjects is not a proposition in Whitehead's sense." What you are not accounting for are general propositions, i.e., propositions that do not refer to any particular individuals but make statements in terms of some or all individuals. Whitehead is explicit that singular propositions that lack meaning for some entertaining subjects can be replaced by general propositions with a similar meaning that will be meaningful for said entertaining subjects. This is an insight taken from Russell's work on descriptions. This is what Whitehead is referring to in the following quote: "But there are propositions of a† more general type, for which ‘Caesar’ and ‘Rubicon’ have more generalized, vaguer meanings. In these vaguer meanings, ‘Caesar’ and ‘Rubicon’ indicate the entities, if any, located by any one member of a type of routes, starting from a certain type of inference and conjecture." I think this is a very important detail to understand for your thesis. I'm working on a paper more or less touching on this, hopefully I'll be able to share it sometime soon.
Philosophers are going to differ over what they would recognize as agents that genuinely take on intentional mental states and thus have propositional attitudes. I do not think this alters that Whitehead's theory of propositional feelings involves subjective forms playing the role that's typically identified as propositional attitudes, which is then the crucial thing to consider in terms of being able to refer to the relevant work already done in the philosophy of language.
Nice essay but…
I think you’re missing it and still treating possibility as something that requires its own ontology. That’s where I find myself taking a different path.
The central move of your essay seems to be:
Novelty requires propositions (“lures”), propositions require eternal objects, and eternal objects require God to grade relevance.
My question is: why assume possibility itself is an ontological inventory?
From my perspective, possibility isn’t a catalogue waiting to be selected. It emerges from what remains consistent with reality.
For example, you write:
“Error is the price we pay for progress.”
I’d put it differently:
Error is the price we pay for exploration under incomplete knowledge. Progress comes from calibration, not error itself.
Error isn’t creative. Calibration is.
Likewise, you argue that metaphysics requires eternal objects because we can think about unrealized possibilities. I’m not convinced the ability to negate requires an ontological realm of unreal entities. It requires only a system capable of representing states that are compatible or incompatible with reality. That seems a much weaker commitment.
My biggest point of departure is here:
“Without God, possibilities…would remain relatively nonexistent.”
I wonder whether this conflates possibility with value.
God need not be understood as maintaining an inventory of possibilities. God can instead be understood as the ground of the reality within which possibilities become meaningful. Reality itself constrains what is possible.
That’s why I wonder whether Whitehead, Lewis, and much of modal metaphysics begin one step too late. Rather than asking,
What is the ontology of possibility?
I’d ask:
What makes any possibility physically realizable while excluding infinitely many others?
To me, that’s a more fundamental question.
On that view, novelty isn’t reality retrieving something from a warehouse of eternal forms. It’s a genuinely new stable organization becoming possible because the evolving world has reached a configuration where that organization no longer violates the underlying constraints.
In other words:
Novelty is discovered, not stored. It is licensed by reality, not retrieved from a catalogue.
That’s why I wonder whether eternal objects are actually doing explanatory work, or simply relocating the mystery. If coherence, recurrence, and constraint are sufficient to explain why stable forms emerge and persist, what explanatory role remains that requires an ontology of eternal objects?
Let me try to explain myself better.
I am not arguing that possibility requires its own ontology. The ultimate category here is Creativity, a formless productivity or Ungrund (Böhme, Schelling), sheer indeterminacy rather than any stock of possible forms. Determinate actualities are not drawn from it as if thawed from a freezer but cut out of it by their own becoming. Whitehead's name for that self-cutting or decision by which an actuality is achieved is concrescence, of which God’s primordial valuation is the non-temporal instance, conditioning every temporal cut without preceding or determining it in advance.
As I wrote in the essay you’re replying to, eternal objects carry no coordinates, no metric, none of the pre-given structure that would make them a state space. Any coordinate/state spaces that come to be are built and rebuilt through actualizations, not imposed from above by pre-determined possibilities.
On my view, coherence and recurrence already presuppose the form of definiteness. To be inherited, recognized, contrasted, or tracked, something must be recurring. Definite potentials that which recurs. The point is not that a determinate form was stored somewhere before its emergence, but that emergent determinacy cannot be made intelligible as such unless there is some non-actual pole of definiteness involved. Note that the actual and potential are poles not separate worlds. There is one world, but the historic becoming of that world involves the finite actualization of infinite possibilities. Those possibilities are not “somewhere else.” They are in the infinite actuality, God; and precisely as you say, God is the ground within which possibilities become meaningful.
Regarding error as the price of progress, Whitehead is saying that non-conformal or false propositions, precisely because they are active throughout the living world and not merely the logical judgments of we speaking animals, can lure organismic feelings beyond conformity to the already actualized. “Error” is the risk internal to adventurous novel comparisons/contrasts. As far as I can tell, your idea of calibration works only if we are talking about an already constituted epistemic agent correcting its world- and/or self-model. I am after a more basic ontological question.