Thursday, June 9, 2011

A Divine Memorial to the Past? Memories in Presentist Truthmaking

An older, unpublished paper of mine I wrote in grad school, no longer quite up-to-date:

In A Treatise on Time and Space, J R Lucas suggests the theory that it is God's memories that ground purported truths about the past. More recently, Alan Rhoda has argued at length for this view, noting that it is common fair among process theists. There are some troubles for this theory, though, which show that such a theory needs to be radically restructured and restated if it is to be at all viable. In this paper, I explore these problems and, in doing so, the question of how exactly to formulate the divine memories theory in a plausible way. It turns out that producing a version of the theory that preserves all of its purported strengths and yet still avoids the problems of the other versions is much more difficult than it seems – and, indeed, we seem to have very good reason to be skeptical that such an ideal version will ever see the light of day.

Presentists take it that everything which exists is present and exists at the present. The past, in a very strong sense, is no longer and the future, correspondingly, is not yet. This creates a problem, however, with accounting for truths purportedly about the past. If only present things exist on the presentist view, in what are presentists supposed to ground past-tensed truths? This ‘Grounding Problem’, as it is sometimes called, has elicited numerous responses, almost all of them attempting to point out some present entities or facts that are supposed to be doing the grounding of past-tensed truths. Some appeal to primitive or brute past-tensed states of affairs or properties, others to arrangements of abstract maximal propositions, and still others to temporally distributional properties (among other things).[1]

Despite such varied responses, many of them have met a number of objections – that they are metaphysically ‘cheating’, that they do not really guarantee the truths they are meant to ground, that they have implausible logical consequences, and so on.[2] One view that might be put forward as an ideal solution to all of these problems is to suppose that all the grounding work for past-tensed truths is done by God’s memories – so that God’s remembering my past trip to Maui grounds the truth of the past-tensed statement that I went to Maui. J. R. Lucas (1973) has suggested such a view, as does Alan Rhoda (draft), who notes that this view is common fair among process theists such as Hartshorne (1984). Such a theory appears initially to have many advantages over the other theories. Among them, it seems to accord well with a general theistic perspective, particularly the general idea of a kind of ‘metaphysical supremacy’ for God. When in doubt about what grounds a certain fact or about whether to postulate additional entities to explain something, why not just turn to God for our full explanatory needs instead? And what more grand version of the dependence of the world on God than that where the very past itself exists only in and because of the divine mind?

At first blush, the statement of such a view appears rather straightforward:

GMem1: It was the case that p iff God remembers that it was the case that p.

But we quickly run into a problem here – is the content of God’s memories past-tensed or present-tensed? From ordinary memory statements like “God remembers that I went to Maui” or “God has a memory of me having gone to Maui” one might think that the contents of memories are in general past-tensed, or at least that they are so when dealing with wholly past objects or states of affairs. Indeed, Rhoda (draft) seems to write at times as if this was his view. If this is correct, we could put GMem1 more clearly as follows:

GMem2: It was the case that p iff God has a memory whose content is that it was the case that p.

I think, however, that we ought to reject GMem2 and instead assign the past-tense involved in ordinary memory-statements not to the content of the memory state itself but rather to the temporal perspective of the speaker on the content. So “Sam remembers that he hit the ball” tells us (at least) that (1) Sam has a memory whose content is normally expressed with the present-tensed “I am hitting the ball”; and (2) the content of that memory is ascribed to a time earlier than the memory. This situation is similar to that involving statements such as “At one time, Sam believed he was the tallest man in the Communist Party”, where the “was” does not indicate that Sam once believed some past-tensed statement about his comparative height in the Communist Party but rather indicates the speaker’s own current, shifted temporal perspective on the purported obtaining of that content. So the analysis of GMem1 should, perhaps, more exactly read as follows:

GMem3: It was the case that p iff God has a memory whose content is that p.

But why is GMem3 needed by the divine memory theorist as opposed to GMem2 in the first place? Well, consider what would happen if we regarded the content of a memory to be past-tensed as in GMem2. The right-hand side of the biconditional in GMem2 contains exactly what we needed to find grounding conditions for in the first place (that is, its having been the case that p). Because of this fact, GMem2 is simply not a successful statement of the grounding conditions for it having been the case that p – it is plainly circular, since whatever grounds the right side is a function of what does so for the left. The sentence ‘it was the case that p’, even though it is used all on its own on the left-hand side and, arguably, appears in an intensional context on the right, still appears on both sides in a manner objectionable enough to defeat the account. To put it in a different way – to give the right-hand side of the biconditional content requires that we are already independently able to give content to the left (since the content of the right incorporates – or at least is a function of – the content of the left). And doing this for the left-hand side will, among other things, require giving it grounding conditions. But this is just what we cannot do since it is precisely the right-hand side of the biconditional which is meant to do that job for the left-hand side in the first place. As a statement of grounding conditions, GMem2, then, simply fails. So the divine memories theorist should formulate their view as GMem3 has it, not as it is in GMem2. The content of God’s memory must be present-tensed (or even maybe tenseless), not past-tensed as GMem2 would have it.

Now that we have GMem3, do we have yet a perfect formulation of the divine memories view? Unfortunately not – we are instead faced with a brand new problem that needs solving. After all, what makes something a memory in the first place? What seems to make something a memory with the content that p – as opposed to some other attitude towards p – is, at least partly, that it is true that it was the case that p. Additionally, for episodic memory, we would also require both that one has a past (perhaps causal) acquaintance with its being the case that p and that this past acquaintance is the cause of the current memory. If Rhoda (draft) is right that God’s acquaintance with facts is direct and that his current memories are a result of these past acquaintances, this additional condition may be required on all of these divine memories which are meant to be doing the grounding of past truths.

But now there is trouble – as just mentioned, the fact that it was the case that p is one of the grounds for the fact that God has a memory with the content that p. And not just that, if we apply the conditions for episodic memory to God’s memories, then all sorts of past-tensed truths will be involved in grounding the fact that God has a memory with the content that p – including the fact that it was the case that p itself. But, on GMem3, the fact that God has a memory with the content that p is itself supposed to ground the fact that it was the case that p! We clearly have a vicious circle that we somehow must break out of. If we want to keep something like the divine memories view of presentism, I take it that the only option is to come up with some other way of picking out the appropriate mental states which are supposed to be doing the grounding work – that is, other than as memories – and in such a way that we do not already presuppose what we are supposed to be explaining – that is, the truth of things like its having been the case that p.

So, where we let ‘M’ designate some type of mental state of God’s which is supposed to meet these criteria just mentioned, the divine memories view should really be formulated something like as follows:

GM: It was the case that p iff God has a mental state of type M with the content that p.

But, having been forced into GM, the divine memories presentist is now faced with challenges they did not formerly seem to face. Many presentist accounts of the grounding of past truths, for instance, are susceptible to conceivability arguments against their proposed truthmakers. Consider a verificationist account, for instance, on which past truths are grounded in present evidence.[3] If this account were correct then, given the current evidence, it would necessarily follow that we have exactly the past truths we in fact have. But this does not seem right. It is certainly conceivable that our universe have the evidence it in fact has yet have a completely different past (say, because God decided to miraculously make it so at this particular point in time, with no taking into account anything that came before). So it seems false that evidence is what grounds past truths since the two seem to be only contingently related.

Now, one virtue of cashing out divine memories presentism in terms of memories (as it was done in GMem1-3) was that it logically guaranteed the truth that it was the case that p – no conceivability argument was possible against it.[4] But now that we cannot specify M in GM in terms of memories, it looks like the view is probably going to be susceptible to conceivability arguments after all – it seems likely that it will indeed be conceivable that God have a state M with the content that p and yet it fail to be true that it was the case that p. Indeed, it will be conceivable precisely because of this that M is not a memory at all, since (as was already mentioned) to be a memory is at least in part to have some content that p which was formerly the case.

But if one cannot already assume that M is a memory, it is not clear there is any other way of specifying M such that it will logically guarantee the truth that p was the case. M cannot be some kind of belief or knowledge since, unless p is still true, that would imply that God knows or believes something false, which is impossible given divine infallibility. Perhaps it is a kind of perceptual state; but if such a state is to account for cases where p is presently false, it cannot be of the sort that guarantees the veridicality of its content. We cannot appeal to causal facts either, since causal facts, on a presentist view, will be partly about the past and hence in need of the same grounding as the truth that it was the case that p. Rhoda (draft), in his argument for the divine memories view, puts it this way:

This dual reference—to a predecessor state and a successor state—naturally requires our analysis of “c caused e” to quantify over both c and e. The presentist, however, will insist on placing at least one of those quantifiers within the scope of a tense operator. Thus, if “c caused e” then either e exists and it was the case that c exists, or e existed and it was then the case that c had existed.

The only way to save such attempts at typing M seems to be to regard the content of M not as p but as its having been the case that p. But once we do this, we are again faced with the same problem as that which plagued GMem2 since we are explaining its having been the case that p in terms of God having some mental state with the content of its having been the case that p. GMem2 is, in fact, just one particular instance of this class of doomed views. But if, as seems obvious now, such a way is blocked, it does not seem that any way of typing M can get out of the conceivability problem without either running into the problems faced by GMem2 or those faced by GMem3.

So once we properly formulate the divine memories view, one of the main virtues it had over other presentist views seems to evaporate. The divine memory view, as formulated in GM, seems to lack any resources to block conceivability arguments against it. It appears to be possible that God have such states and yet the past be different than it in fact was. But the only alternatives to this version seem to be versions like GMem2 and GMem3 which seem to be plainly unviable. So GM seems to be the only remaining version of the divine memories view left on the table, problems and all. But if the presence of such problems is taken to be good evidence against a version of presentism, as well it should, that means we have good reason to look elsewhere for an appropriate theory of time and persistence – either to a non-presentist view or to some presentist theory which can in fact do better. Despite whatever initial appeal it might have had, it seems then that the divine memories constitute little more than a divine memorial to the past and are simply not the presentist truthmakers some presentist-leaning theists may wish them to be. God may be ultimate or exalted and much may depend on him for its existence, but just not in this particular way in this particular case.


Bigelow, John (1996) “Presentism and Properties” in Tomberlin, James (ed.), Philosophical Perspectives 10: Metaphysics. Oxford: Blackwell, 35–52.

Bourne, Craig (2006) A Future for Presentism. Oxford: Oxford University Press.

Cameron, Ross (draft) “Truthmaking for Presentists”

Craig, William Lance (2000a) The Tensed Theory of Time: A Critical Examination. Dordrecht: Kluwer Academic Publishers.

_____ (2000b) The Tenseless Theory of Time: A Critical Examination. Dordrecht: Kluwer Academic Publishers.

Crisp, Thomas (2007) “Presentism and the Grounding Objection” Noûs 41.1: 90-109.

Hartshorne, Charles (1984) Omnipotence and Other Theological Mistakes. Albany: SUNY Press.

Keller, Simon (2004) “Presentism and Truthmaking” in Zimmerman, Dean (ed.), Oxford Studies in Metaphysics Volume 1. Oxford: Oxford University Press, 83–104.

Lucas, J. R. (1973) A Treatise on Time and Space. London: Methuen & Co.

Ludlow, Peter (1999) Semantics, Tense, and Time. Cambridge, MA: MIT Press.

Parsons, Josh (2005) “Truthmakers, the Past, and the Future” in Beebee and Dodd (eds.), Truthmakers: The Contemporary Debate. Oxford: Oxford University Press,161-174.

Rhoda, Alan (draft) “Presentism, Truthmakers, and God”

Sider, Theodore (2003) Four Dimensionalism: An Ontology of Persistence and Time. Oxford: Oxford University Press.

[1] See Bigelow 1996, Bourne 2006, Cameron draft, Craig 2000a and 2000b, Crisp 2007, Keller 2004, and Ludlow 1999 for various presentist options.

[2] For various criticisms see, for instance, Cameron draft, Sider 2003, and Rhoda draft.

[3] See, for some brief discussion, Parsons 2005.

[4] Rhoda draft says something similar in favor of the divine memories view over and against many other presentist views.

No comments: