At Investigations 304, p.102, Wittgenstein demurs from a suggestion that in a hypothesis entertained earlier – not Wittgenstein's own – the conclusion, when related to pain and pain-behaviour, had been that ‘the sensation itself is a nothing’: ‘The conclusion was only that a nothing would serve just as well as a something about which nothing could be said’. (The “beetle” in the possibly empty box, Inv 293, would seem to be the example in mind.) Simon Blackburn takes this out of its context, to place it suggestively in a new context, in the wake of a chapter headed ‘God’, and in juxtaposition to Hume's insinuated question that ‘If the whole of Natural Theology … resolves itself into one … undefin'd Proposition, … If this Proposition affords no Inference that affects human Life, or can be the Source of any Action or Forbearance … ’, why trifle with anything so otiose?Footnote ¹

The shifted context could invite us to consider “A nothing would serve just as well as a divine nature about which nothing could be said”. But if Hume was referring there to the position on ‘the Being of a God’ for which his character Demea ‘might cite all the Divines, almost, from the Foundation of Christianity’, and on which he did cite Malebranche, to refer to as ‘Being without Restriction … ’; that should put us on the alert. ‘Being without Restriction’ could well be something about which nothing can be said, but need not be a something about which nothing can be said.

In English, expressions of the form ‘being a ___’ can be ambiguous as between being of the ___ kind (being a cat, being feline, say), and the same, but in addition being one and one only individual of the ___ kind (being a cat, being a feline). Someone may enter a room, sniff, and say ‘Cat’, as though by way of explaining the smell in the room, or his own response; yet without necessarily knowing whether this cat or that, one cat or more, might have been in action.

Being without restriction, if there is any, is precisely not being a being of any kind. Indeed one way of asserting existence-without-restriction, if that should be called for, is by denying to it not just numerical unity, but any determinate quantity or cardinality, including nought; or indeed by denying to it any determinate existence, denying to it existence in any determinate mode of existence.Footnote ² Or it is, if you can coherently suppose an “it” to deny anything of.

I propose, in response to the shifted context, to leave undisputed the contention concerning ‘ … a something about which nothing could be said’, to consider rather “A nothing would serve just as well as something about which nothing could be said”. This allows me, as a restriction to ‘a something’ might not, to look for a counter-instance from the theological area which the shifted context is presumably inviting me to look: where ‘God’ may stand for “being without restriction” in the way recognised by Hume as having been used ‘from the Foundation of Christianity’ (Dialogues Pt 2), in the God of Demea's Divines.

Since the time of the Deists, and in the generality of academically discussed “theisms” today, we are fairly typically expected to consider God as a something, a determinate set of determinate somethings. Divines of earlier times, however, were often, as Hume makes Demea observe, more typically expected to consider God as something, but not necessarily as a something, or any determinate number of somethings. Wittgenstein may or may not have been aware of the historically uneasy fit between the contention that God exists, and the (post-Deistical?) expectation that this should mean not only that something but that a something will then exist, an entity with identity in some kind. But he recognised, at any rate in the manuscript just cited, that if God exists as older Divines expected, it is not as a determinate value of the bound variables we can use in our more than merely dialectical arguments. If the essence of God guarantees his existence, ‘that really means that here there is no question of existence’: no question, at any rate, of existence as tractable within our “scientifically” serious arguments. I do not need to dispute that, neither did the Divines alluded to by Hume.

The contention that something will ‘serve just as well as … ’, given the shifted context, invites a further preliminary, in the question ‘would serve … for what purpose?’. In the text from Hume which Blackburn juxtaposes to Wittgenstein's, the answer to be understood is: For the purpose of supposing it, in order to draw inferences of practical consequence in affairs of humans, simply as humans. The crucial conclusion in Dialogues Pt 12, which Hume needs to make stick against his primary (Deist) targets in that work, is: ‘If the whole of natural Theology, as some People seem to maintain, resolves itself into one simple, tho’ somewhat ambiguous, at least undefin'd Proposition, that the Cause or Causes of Order in the Universe probably bear some remote Analogy to human Intelligence: If this Proposition … affords no Inference that affects human Life, or can be the Source of any Action or Forbearance, … ’Footnote ³ why should he or anyone bother to dispute that? And if getting such practical inferences should indeed have been a purpose driving Demea's older Divines, as it arguably was driving the natural theology of the Deists and their successors, a first conclusion could very well be that, for that purpose, something about which nothing could be said, even where something in no way finite, in no way determinate, should be concerned, could then indeed be in no better case than a nothing.

This is where a difference in background perspective comes into play. To seek to draw such consequences from a divine nature, may have seemed desirable to Hume's targeted Deists, and may seem routine for the generality of today's academic theists (and their internal critics): it need not be the only purpose of importance to inquirers. It need not be a purpose of some quite serious inquirers at all: any more than finding a deity capable of telling the time or feeling the cold need not be. There can be other purposes, for which these might want to suppose something about which nothing could be said, while not wishing to “suppose nothing”, even if that could be done. Before giving an example, however, there is a concession to be allowed.

If ‘something about which nothing could be said’ is meant to cover ‘about which nothing could be denied’ as well as ‘about which nothing could be affirmed’, then indeed I would not wish to dispute that that something may be in no better case than a nothing, even for discursive purposes other than drawing inferences from it. If there were anything of which nothing could even be denied coherently, how could it be successfully referred to in intelligible speech of any level, even a conversational level (as in Locke's ‘merely civil’ use of language)?

A nothing cannot be supposed (i.e., supposed to be something) without absurdity, regardless of why you might wish to be able to suppose it. If it is no better than an unjustifiable reification, of merely rhetorical and no scientifical worth, as perhaps to be seen in Das Nichts nichtet, a nothing is not something susceptible of being supposed anyway, even for something to be denied of it.

But if something not finite, not determinate, in any way, can coherently be supposed (to be), then at least something can be denied of it; or else there is no “it” distinct from anything else, or even distinct from nothing, to be supposed (to be) by us. You might thus wish to deny of it that it is susceptible of being frustrated, of being interfered with; perhaps to deny of it that it can be lacking in anything that exists or can exist in extra-mental reality.

A crucial point about something in no way finite, in no way determinate, is that if there is any, it is has an ontological distinctness which owes nothing to distinctness or determinateness in any kind. It is distinct from nothing (by being something) and is distinct from anything that is finite, determinate, in any way (by being not finite, not determinate, in any way). By being itself, whatever it may be, no question of indistinctness (and hence of vagueness or indeterminacy of reference) can arise. It is not distinct from itself; and if it is not finite, if it is not determinate in any way, there cannot be more than one of it, and it cannot have distinct parts. It is thus absolutely distinctly an “it” of which something can be denied, no matter what expression we may use with the aim of referring to it, when seeking to deny something to it.

The specifics of the expression chosen as a marker are to be seen as of no importance to our understanding of the thing marked (in this case, the strictly infinite), if they can tell us nothing of the nature of the “it” in question. ‘Something not finite in any way’, or ‘God’ (as when used by Demea's older Divines on the understanding that ‘God’ is to stand for something in no way finite, if there is any, and for nothing else in extra-mental reality) may be more helpful – or unhelpful – heuristically or mnemonically than, say, ‘Jabberwocky’ or ‘Mickey Mouse’, if used to signal the same referential intention on the part of a speaker; but can tell us strictly nothing of the nature of the thing referred to, if that nature is being supposed to be something in no way finite, in no way determinate. Expressions, if and when used to stand for anything strictly infinite, are to be understood as Sinnlos to us, lacking in signification, save in a conversational level of speech unusable for even broadly “scientifical” purposes; if and precisely because they have succeeded in standing for the strictly infinite, if there is any. The expressions can ‘mark’ what they refer to, without ‘signifying’ anything of it to us. Even ordinary marker buoys can tell us that something is being marked, without telling us whether it is lobsters or contraband; though many of them are in fact designed so as also to signify something of what they mark. That is a bonus in a marker, beyond what is necessarily called for in marking as such. Expressions intended to mark something not finite in any way, if there is any, can tell us nothing at all of it, while yet marking it absolutely determinately, if it is there at all. It cannot be mistaken for nothing, if it is something. It cannot be mistaken for anything else, if everything else that is something is in some or other way determinate. Such intended markers for something not determinate in any way can fail of success, only in the case where nothing strictly infinite exists; where only a fool or a knave would want the attempt to succeed.

I have been considering the case of something you might wish to refer to, of which nothing could be affirmed. The more sweeping case of just any ___ you might seek to refer to, but about which nothing could even be denied, is quite different. Could you ever be faced, in the imaginary “ ___ of which nothing could be denied”, with anything susceptible of being referred to determinately, even as an object of someone's actual judging thought? I see no reason to think so, and would doubt that Wittgenstein would have been tempted to imagine so.

In the case of any strictly infinite existent, the case of anything existing but not in any determinate mode of being, coherent denials are not impossible. ‘The strictly infinite is not finite in any way’, ‘ … is not green’, ‘ … is not to be identified with the Forth Bridge’, ‘ … is not blindness’, ‘ … is not a chimaera’, and so on, can be perfectly coherent, and even accounted true, provided – and it is an almighty proviso – that ‘the strictly infinite’ itself does stand for something in extra-mental reality. Unlike a nothing, then, the strictly infinite, if there is any, can be supposed [to be something], for the purpose of making coherent denials of it.

For some denials, a further possibility may arise. There could be an interpretation where ‘ … of which nothing can be denied’ might be meant to extend only to those denials where some “thing” more properly speaking may be concerned; an interpretation in which chimaeras, say, or blindness, or the bloom on the cheek of youth … might not be counted as things proper, but in which other things (say, instantiated integral “forms”) might be. The way could then be opened up to further arguments. In particular, the way could be opened for a further possible purpose: unavailable in the case of a nothing, but available in the case of something supposed in no way finite, something about (the content of) which nothing can be known. In particular, the way could be opened to arguments to the effect that if any determinate existents – say, the wisdom to be found in Socrates – do not of their nature have to be limited, merely to exist at all, then they cannot but be found existing also, without the limitation of being-found-in Socrates, in the strictly infinite nature itself, provided there is one; and indistinctly from it.

If that may be true of the wisdom of Socrates, it may not be true of the red of his nose. Red, or for that matter moral goodness, might have to be found, if at all, only in limited forms, determinate in at least some way; whereas descriptive goodness or wisdom might not have to be restricted in this way. Red, for example, when found in the world of the categories, will be of particular surfaces, or particular wave-lengths. Moral goodness can logically be ascribed only to something taken as already circumscribed in some way. I do not pursue this possibility for further argument here, and do not offer it as a conclusion, but historically it was a matter of importance to some of Demea's Divines.Footnote ⁴

Two ways would then be possible, in which it need not be true that ‘a nothing would serve just as well as a something about which nothing could be said’. And if there is anything that is not finite in any way, there is no way in which it is not going to be “something about which nothing can be said”, if affirmations concerning its supposed nature or doings are what is in mind.

In a first possible way, it can in principle be supposed [to be]: for the purpose of denying something of it. The same cannot be said for a nothing, if a nothing cannot without absurdity be supposed [to be anything] in the first place, for anything to be denied of it. In the case of a nothing there is no “it”, no “something” to deny anything of. This is a first way in which a nothing would not necessarily serve just as well as a something about which nothing could be said (where ‘said’ has the sense of being affirmed, not necessarily of being denied).

A second possible way is likewise not excluded. Something strictly infinite, if there is any, is something about which nothing can be said, in the sense of being affirmed, yet that strictly infinite something can be supposed [to be something], for the purpose of referring to it obliquely in a relational affirmation: for example, by asserting of something other than the strictly infinite, this tree, perhapsFootnote ⁵, that it bears some non-necessary relation to the strictly infinite: e.g. that it is non-necessarily ontologically dependent on the strictly infinite, or is non-necessarily known as an existent by it.Footnote ⁶ (Only non-necessary relationships can be had by anything other than itself, to anything strictly infinite. If there is anything that is not finite in any way, it cannot logically be said to be able to suffer necessitation in any way.)

If there is in reality nothing strictly infinite, then ‘This tree is known to the strictly infinite’ cannot be true. If there is something strictly infinite, and if the tree genuinely is itself a determinate something – which may or may not be the case for real-world trees – then ‘This tree is known to the strictly infinite’ can be true. Whether the supposition or presupposition of something strictly infinite should be true rather than not, is thus crucial for the truth or otherwise of the relational assertion envisaged.

By contrast, you cannot logically suppose nothing [to be something] for the purpose of affirming some determinate entity to be non-necessarily related to it, or indeed for any other purpose. Of course if this tree, say, is a determinate entity, it is necessarily not nothing; but this follows from its being said to be an entity. It need not be imagined as the affirmation of a relationship between the tree and some (unjustifiable) reification, something called nothing.

As against the contention I considered, taking further the suggestion offered by Blackburn's transplanting Wittgenstein's formula into the context of Hume's Dialogues, I conclude that there is at least one unstrained interpretation of ‘something about which nothing could be said’, and there are at least two possibly coherent purposes, for which it may not be true that ‘a nothing would serve just as well as a something about which nothing could be said’. The interpretation is one which takes ‘something about which nothing could be said’ to stand for something strictly infinite, “being without restriction”, if there is any, and understands ‘about which nothing could be said’ as ‘about which nothing can be affirmed’. On this interpretation, denying things of that strictly infinite something, if there is any, is not excluded. Neither is asserting of something other than it, some non-necessary relation to it. Both of these are possibly coherent purposes, and historically were of importance to at least some of Demea's older Divines.

That is my conclusion here, but it could be worth emphasising that it is asserted to hold only for denials of things to anything strictly infinite, or for affirmations or denials (made in licensable unanalysed or incompletely analysed speech) of non-necessary relations borne by things other than it, to it. For everything else, including not just beetles but Wittgenstein's “beetles” in boxes, I have no particular reason to dispute that, at any rate for drawing inferences that humans can understand, something of which nothing could be said, can indeed be no better for such a purpose than a nothing. Someone accepting my conclusion need thus have no quarrel with the answer expected from Hume's rhetorical question cited at the outset of the present discussion, from the passage used by Blackburn to shift the context of Wittgenstein's cited conclusion to the effect that ‘a nothing would serve just as well as a something about which nothing could be said’.

My own conclusion, of course, against the contention suggested by Blackburn's fresh context for Wittgenstein's formula, depends crucially on the possibility of making the supposition of infinite existence without incoherence. That in turn depends on being able to provide at least one non-incoherent use of ‘exists’ so as to cover both the existence of determinate things, as tractable in some broadly Fregean or comparable way, and infinite existence. This is a large proviso, even if it may not be the lost cause from the outset that it can seem.Footnote ⁷ If it can be made out, however, my conclusion comes at the cost noted by Hume towards the end of his Dialogues: it allows no possibility of drawing inferences of practical consequence to humans from an essence about whose content it is supposed that nothing can be known. This sufficiently stymied the theology of Hume's Deist primary targets, but does not have to trouble others, whether engaged in theological explanation, or committed to worship of a God not finite in any way. The strictly infinite God of Demea's older Divines, and of my conclusion above, is precisely something of which nothing can be known, something of which one cannot but be silent; something, that is, of which nothing intelligible to us can be predicated properly, absolutely, and with truth.

My conclusion comes too, if it is made out, at a related cost, not troublesome to at least some of the older Divines, and at least noted by Wittgenstein in another place. Apropos, it would appear, of an ontological argument, he said: ‘That the essence of God guarantees his existence – that really means that here there is no question of existence’.Footnote ⁸ There is indeed in anything strictly infinite, no question of the existence we can recognise in the existence of determinate things, the existence recognised in the values of our bound variables, if entity is to be allowed from the outset only where there is identity in some kind. If there is a strictly infinite divine nature, it cannot be ‘a thing among the things’, and need not be expected to be. Moreover, if that nature's essence is nothing other than its existence, as was notoriously held by a theologian who had rejected the most famous of ontological arguments, it should likewise be allowed to hold that there is in play in a strictly infinite divine nature no question of an essence either, i.e., no question of an essence of the kind from which effects can be read off; or indeed of an essence of any determinate kind.

Not only Hume's Deists but the generality of academic theists nowadays could be expected find that cost unwelcome. They commonly take God to have attributes in the sense of the designata of significant predicates; take God to be or to have a strictly provable or strictly disprovable existence; take God to be morally good or wicked (and hence to be presupposed as already circumscribed in at least some way, so as to be susceptible of being evaluated in a strong evaluation) … . In short, are they not to be seen as precisely understanding God to be determinate, circumscribed, in at least some way; by being this God with these non-relational attributes, and not some other?

If they are, they may find more that is unwelcome to them. For if the God of their understanding is in any way circumscribed, determinate, finite, how can it then be excluded that that determinate God could on their assumptions have to come down to being understood either as a thing within a mere sum of things, or as the sum itself; in which case there is ultimately no difference between the world around us, as they will then have to see it, and the world around us, as viewed by the ancient atomists and other thoroughgoing and consistent atheists. True, either one of the things, or the sum of things itself, will have the label ‘God’ affixed to it by the “theists” in question. But their label on a toy of necssity leaves it a toy of necessity; their label on blind necessity itself leaves it blind necessity. A mere sum of things, and nothing besides – if that is what the world around us should come to – is of its nature necessary; and the things of that sum, including ourselves, our ideas, and the God of the academic theists (unless it is the sum itself) are toys of necessity. A mere sum of things could then be arguedFootnote ⁹ to amount to a “system” within which it could be found that everything that can occur within the system must occur, and must recur to any approximation you like, indefinitely many times. Within the assumptions of that kind of system, the God of their understanding – unlike that of Demea's Divines – may indeed escape a charge of being “something of which nothing can be said”. But for the purpose of providing a theology of any explanatory worth, or indeed for the purposes of worshippers when praying for things to happen, it has appeared above that that need not be an insuperable difficulty. How serious is the charge which can then be made out?

By contrast, relying on a God finite in any way – a God identified with the God of their understanding – is going to be otiose or pointless. It does not exclude a “mere sum of things” metaphysics, and can be accommodated within one whether as a thing within the sum, or as the sum itself. If a “mere sum of things” metaphysics should in fact be true, that may not be the only inconvenience anyway. Scientific explanation itself is going to be possible only if by some fluke we should happen to inhabit a region of the universe in which things occur and succeed other things in ways susceptible of explanation; not ways necessarily inexplicable; as they will have to be ultimately within the sum of things overall. As for worship of such a God, it is either going to be absurd, in the case where you would be going through the motions of illusory worship (pretending to do something voluntarily within a totally necessitated system towards) the (necessary) sum of things itself; or rather pointless, in the case where you would be going through the motions of illusory worshipping, or praying to, something that will ultimately have to be some fellow toy of necessity, though (in vain) called God.

So while the God of the Deists, and indeed the God of today's academic theism fairly generally may indeed be better than a nothing for purposes of drawing inferences that humans understand, so obviously better that no-one would be likely to think otherwise; how is it for more specifically theological purposes? Or those of worshippers? Is it for those purposes any better than a nothing? And it is supposedly, after all, in connexion with such purposes that it is typically considered within today's academic philosophical theology.

For purposes of serving a putatively explanatory theology, or purposes of worshippers, such a determinate God – unlike the God of Demea's Divines, as argued above – could well serve no better than a nothing. Could serve even worse than a nothing, I might add, if the nothing is less likely than the Gods of today's academic theists are, to raise false hopes, or to massage prejudices which have no business to be encouraged. I do not count that as a further conclusion here, because the gaps left in the argument for it are too large and too contentious to be made good in the space available here. But would it be all that difficult to fill them? Even without prompting from Prof. Blackburn's transplantation of Wittgenstein's remark to a Humean context, would academic philosophical theology not be advised to look afresh anyway, with less disdain than it has done in post-Deistical times, on the supposition of a referent for ‘God’ that is not finite in any way; the supposition of something rather importantly other than nothing, whether for the purposes of academic philosophical theology or of serious worshippers? Hume, Wittgenstein, and Demea's older Divines did not close off that possibility. Why not at least entertain it?