Defence of the premises

Some of our beliefs are motivationally peripheral. I believe that the globular cluster M13 primarily consists of very old stars. But this belief affects very few of my actions. If I came to believe that M13, unlike other globular clusters, consists of young stars, my life would not significantly change. Other beliefs, however, are more central. For instance, I believe that I have a wife and kids. If I ceased to believe that, my life would change in widespread and significant ways, and many of the things I am now motivated to do, I would no longer be motivated to do. My belief in the existence of my wife and kids, then, is one of the beliefs that are motivationally central to my life.

Centrality and peripherality is, of course, a matter of degree. Somewhere intermediate between my belief that I have a wife and kids and my belief that M13 is mainly composed of very old stars, there is my belief that brushing teeth is healthy. The centrality and peripherality of a belief differs from person to person, and some variation is appropriate. It is appropriate that the belief about M13 be more central to the life of an astronomer specializing in stellar evolution than it is to my life. Moreover, note that when I talk of a belief being motivationally central, or at the motivational centre of a life, I am not claiming the belief is the centremost belief – indeed, all claims of centrality in this paper are to be understood as capable of holding by degrees.

Beliefs can be false but central, and maybe such false beliefs can be central to a flourishing human life. One can imagine a doctor who leads a flourishing and intellectually sophisticated life bringing an expensive cancer treatment to the needy. A belief in the effectiveness of the treatment will be motivationally central to her life, but her life is not much less a flourishing human life should it turn out that all the studies that claimed the treatment to be effective were in fact wrong.

However, while a false belief might be found at the motivational centre of a flourishing and intellectually sophisticated life, it seems less likely that there be an impossible belief at the motivational centre of a flourishing and intellectually sophisticated life for a significant length of time.

One reason for this is that an impossible proposition entails all propositions. An intellectually sophisticated person or community reflects particularly on the entailments of beliefs that are motivationally central, and some of the motivational centrality is apt to transfer to the entailed claims. As a result, there is some likelihood that if a motivationally central proposition were in fact an impossible proposition, then the person or, especially, community would come up with an entailment q of the motivationally central proposition such that believing q would be damaging to flourishing. For instance, if they believed that circles are squares in a motivationally central way, they might draw the logical conclusion that pleasures are pains, and then they might torture people in order to give their victims pleasure. But to commit torture is significantly harmful to one's flourishing.

Second, an important aspect of human flourishing involves humanly, and hence morally and intellectually, excellent activity flowing from motivationally central beliefs. Maybe it is possible that humanly excellent activity would flow from beliefs that are so far wrong as to be impossible, but it does not seem very likely. A humanly excellent life would be an examined life, and a part of the point of the examination is to ensure the compatibility of one's beliefs.

Normally, in fact, we would expect humanly excellent activity to flow from motivationally central beliefs that are not just possible but true. But we can imagine everyday cases where a mistake about, say, a scientific matter underlies an instance of humanly excellent activity. We can, perhaps at a stretch, imagine a case where humanly excellent activity flows from some impossible and motivationally central belief, but those cases are even less common. If humanly excellent activity flows from some motivationally central belief, then, in the absence of any further information, we should say that the belief is more likely than not to be at least possible.

Therefore, we have reason to accept premise (4). We need to remember, however, that the ‘probably’ in the premise and the conclusion is a probability that does not take into account other evidence (perhaps other evidence of the very same sort, coming from lives centred on other beliefs – see below). It is not an all-things-considered probability.

Premise (5) is easily argued for by citing plausible examples: Abraham Heschel, al Ghazali, Alvin Plantinga, Augustine of Hippo, Clairvaux Abbey, Francis of Assisi, Jean Vanier, Teresa of Avila, the Pittsburgh Oratory and the student groups associated with it, the Taizé Community, Thérèse de Lisieux, Thomas Aquinas, and various personal friends of mine whose humility I will not embarrass by naming them. Some examples that make (5) true are famous, many others less so, and we can all add to the list.

The belief in a MGB typically enters motivationally into the lives of these persons and communities in multiple ways. It makes them see the natural world around them as created by the MGB, and it makes it possible for them to see their neighbour as made in the image of the MGB. It gives them hope in a providential government of the world. It confers a deep felt meaning on their lives and the lives of those around them, by entailing that somehow behind this physical reality there is that than which no greater can be thought. It gives fruit for meditation and grounds contemplation, which not only are constitutive parts of a person's flourishing, but profoundly help form distinctive character traits. Divine love provides an example for imitation. And so on.

Moreover, the belief in the existence of a MGB is entailed by more specific religious beliefs that are motivationally central in the lives of many of the persons and communities mentioned, such as by the Christian belief that the maximally great being became human and died that we might live. The argument, then, could be run with those more specific beliefs, to the conclusion that these more specific beliefs are possible, and hence so is the belief in a MGB (if p is possible, and p entails q, then q is possible).

By (4) and (5), we thus conclude that probably it is possible for a MGB to exist, and hence by S5, probably there exists a being that is perfectly good, omnipotent, omniscient, and creator of all contingent beings in all worlds. Assuming, of course, there is no further evidence to the contrary.

We can, I think, sharpen (4) by specifying that the probability of the possibility of the belief increases roughly in proportion to such factors as: how motivationally central the belief is, how flourishing the individual or community x is, how much of x's humanly excellent activity flows from that belief, how rational the motivational connection between the belief and the humanly excellent activity is, how intellectually sophisticated x is, how long the time span involved is, how large a community x is (assuming x is a community), and so on. The examples I gave in support of (5) differ to some extent in respect of some of these factors, and hence the support they offer to the existence of the MGB varies. However, in all of the cases, at least it seems to be true that the belief that there is a maximally great being, love for whom should consume one's entire life, is very close to, if not at the very centre of, the motivational life, and the flourishing life of virtue is largely motivated by that belief.

Moreover, the greater the number of individuals and communities that there are which satisfy the conditions in (4), the more probable it is that the belief in question is possibly true.

Kripkean cases and two-dimensional semantics

There are, however, a number of Kripkean cases where an impossible belief may be at the motivational centre of a life. The issue is that Kripkean views raise many empirical errors to the level of metaphysical impossibilities. Consider three examples, the first two perhaps fictitious, but if so then not by much, and the third entirely factual: (i) George believes that Dorothy is his biological daughter and this belief is near the motivational centre of his life, but let us suppose Dorothy is not his daughter, having been swapped at birth. (ii) Patricia believes that electrons are manifestations of a field, and her life-work as a physicist is centred on this belief, but in fact electrons are essentially particles. (iii) A fair amount of Dr Livingstone's activity was based on the assumption that the Lualaba River was the Nile; but, in fact, the Lualaba was the Upper Congo, and so it is metaphysically impossible that the Lualaba be the Nile.

It seems quite likely that in some such cases, the false belief could be close to the motivational centre of a flourishing life. One response would be simply to concede this claim, but hold that such cases are relatively rare – most beliefs that are motivationally central in most flourishing lives are in fact possible. This concession, however, while it preserves my argument for the probability of the possibility premise of the ontological argument, does allow that the existence of these cases lowers that probability.

A more daring response to the Kripkean cases would be to affirm that the motivational centrality of the beliefs in the three cases, and in most similar cases, detracts from the flourishingness of the lives and the human excellence of the activities. At the end of the last section, I suggested that the more flourishing the life and the more the humanly excellent activities flow from the belief, the better the case for the belief's possibility, and hence in these cases the argument for possibility is, as is to be expected, weaker. George would flourish more if he focused less on the biological aspects of paternity. Patricia would live a more intellectually open scientific life if she were more open to the possibility of field theories of electrons being false. And Dr Livingstone would perhaps have done more good to the science of geography were he not focused on a Quixotic quest for the sources of the Nile.

I leave it to the reader to judge the plausibility of responses of this sort. I think there is something to them. But instead, I want to press a two-dimensionalist response, which requires modifying (4).

A standard way to see what is at the heart of two-dimensionalism is to consider the sentence S, ‘Water is H₂O’, and two worlds: w ₀ being the actual world, and w ₁ being a world very much like this one, except that the predominant colourless, wet, tasteless, life-supporting liquid with all the same behaviour as H₂O is not H₂O but XYZ. We can then say the following two things. First, the sentence S expresses a proposition p, that water is H₂O, which proposition is true in both worlds, even though in w ₁ the word ‘water’ as used by the denizens of w ₁ picks out not H₂O but XYZ. Indeed, it is a standard Kripkean claim that p is necessarily true. It is usual to call the proposition that water is H₂O the ‘secondary intension’ of S. But there is a another way to consider S and the two worlds. We can talk of the ‘primary intension’ of S, which in w ₀ says of the colourless, wet, tasteless, life-supporting liquid predominant in w ₀ that that liquid is H₂O, and which in w ₁ says of the colourless, wet, tasteless, life-supporting liquid predominant in w ₁ that that liquid is H₂O. The primary intension of S then is true at w ₀ but false at w ₁. A rough and ready way to think of the primary intension of S is as what S would have meant had S been said in the world at which its truth-value is being checked.

We can then distinguish between a sentence token S being metaphysically necessary, namely its secondary intension being true at every world, and its being a priori, namely (simplifying slightly)Footnote ² its primary intension being true at every world. We can likewise say that S is metaphysically possible provided that S's secondary intension is true at some world, and S is conceivable (in this technical sense) provided that (simplifying again) S's secondary intension is true at some world. Now we can modify (4) to say:

1. (7*) If x is a human individual or community that leads a flourishing and intellectually sophisticated life of significant length while holding a motivationally central belief expressed by a sentence token S, then, probably, either S is conceivable or it expresses a proposition that is possible.

The examples (i)–(iii) are no longer a problem. It may be metaphysically impossible that Dorothy is George's biological daughter, but it is certainly conceivable: there are possible worlds where the primary intension of ‘Dorothy is George's biological daughter’ is true – i.e. worlds where George's biological daughter grows up under George's parentage, and is named ‘Dorothy’. There are, not implausibly, possible worlds where the primary intension of ‘Electrons are manifestations of a field’ is true – worlds where a field rather than particle theory holds. And there is a world where a river named ‘the Nile’ that ends in Egypt extends back to the river named the ‘Lualaba’ in the Congo, and that is a world where the primary intension of a sentence expressing Dr Livingstone's belief about the source of the Nile is true.

Of course, now we have a gap between what we can conclude from (4*) and (5), and what we want to conclude. What we learn from (4*) and (5) is that either possibly or conceivably a MGB exists. What we need for the possibility premise of the ontological argument is that possibly a MGB exists.

However, the conceivability of the existence of a MGB entails the possibility of a MGB existing. It is a standard claim in the two-dimensionalist literature that ‘There is a god’ has the same primary and secondary intensions (see, e.g., Chalmers (Reference Chalmers, Lepore and Smith2006) ), and it seems plausible that the same is true of ‘There is a maximally great being.’ After all, ‘There is a maximally great being’ does not contain indexicals, and does not appear to make reference to any individuals picked out de re like ‘Dorothy’, or to any ostensively referred-to qualities or kinds like ‘water’. One might, of course, think that greatness is picked out ostensively, as a quality that is paradigmatically had by sperm whales in respect of size, Albert Einstein in respect of intellect, Francis of Assisi in respect of peace, etc. But that seems mistaken. If greatness were picked ostensively, it seems we would have to say that it is conceivable that a mass murderer exhibits greatness in respect of peace, since there is some world where the best-known person named ‘Francis of Assisi’ is a mass murderer, and that just does not seem right.

Now, a claim whose primary and secondary intensions are the same is possible if and only if it is conceivable, and is necessary if and only if it is a priori (with ‘conceivable’ and ‘a priori’ understood in the technical two-dimensionalist sense). If this is right, then from (4*) and (5) we conclude that, probably, it is the case that possibly or conceivably a MGB exists. Since the conceivability of the existence of a MGB entails the possibility of a MGB existing, we conclude that, probably, possibly a MGB exists, and hence, by S5, a perfectly good, omnipotent, omniscient creator exists in all worlds.

It may be that not all cases can be handled by the two-dimensionalist move. For instance, it could well be that materialism about minds is a priori false, in which case a neuroscientist whose flourishing life is centred on materialistic assumptions about the mind will be a case of someone who leads a flourishing life motivationally centred on an assumption that is neither possible nor conceivable. I myself suspect that materialistic assumptions harm the flourishingness of one's life. But even they didn't, (4*) only claims that the belief is probably possible or conceivable, and it seems plausible that cases of beliefs that are neither possible nor conceivable but which occur at the motivational centre of a flourishing and intellectually sophisticated life will be quite rare.

In any case, the two-dimensionalist move of replacing (4) with (4*) decreased the number of counter-examples, and hence improved the probabilistic argument.Footnote ³ And, of course, we can make (4*) into a proportionality claim the way it was done with (4) at the end of the second section.

Non-theists and atheists

A major objection to the argument is cases of people or communities whose lives are flourishing and are centred on beliefs whose possibility and even conceivability is incompatible with belief in a perfectly good, omnipotent, and omniscient creator, beliefs such as those of atheists or of non-theistic religions. But if such beliefs are possibly or conceivably true, then we have an incompatibility with the conclusion of our ontological argument which says that a perfectly good, omnipotent, and omniscient creator exists in all worlds.

In response, note first that this line of criticism leaves (4) and (4*) intact. These claims embody probabilistic principles, so there is nothing wrong with noting that sometimes they will lead us astray. As applied to theistic cases, they raise the probability of theism, and in the absence of other evidence they would raise it above a half. But now the worry is whether the existence of non-theistic or atheistic cases of application of (4) and (4*) doesn't provide evidence against theism that neutralizes the evidence for theism that (4) and (4*) gave us.

I shall argue that non-theistic and atheistic cases are not as worrying as one might initially think. First, we shall consider non-theistic religious views. About each one, I have a choice of one of three basic moves: (A) admit that significant evidence is provided for the possibility of the truth of the belief, but argue that the belief is in fact compossible or co-conceivable (with ‘conceivable’ understood in the two-dimensionalist sense) with the existence of a MGB; or (B) argue that the belief in question either is not motivationally central to a flourishing life or less centrally or less rationally vel caetera (see the end of the second section for the various factors to be considered) contributes to that life than theistic belief does in the case of the kinds of flourishing theists I have mentioned; or (C) argue that even if (4) and (4*) give us evidence of the possibility of the truth of the non-theistic belief, nonetheless this evidence is undercut by the existence of a decisive argument against that belief.

I shall consider three sorts of non-theistic views, meant to be paradigmatic and particularly common, but with no pretence to exhaustiveness: polytheism, pantheism, and illusionism.

Typical polytheistic religions accept the existence of powerful supernatural agents, and this may well be motivationally central to the lives of the adherents. However, the existence of powerful supernatural agents by itself is compossible with the truth of theism. Indeed, the Western monotheistic religions traditionally accept the actuality of the existence of angels and demons. This is a type-A response in my typology above.

Now, there are ways of adding to the polytheistic claim that render it incompatible with theism. Within the scope of this paper, I cannot handle all possible combinations, so let me give one that is fairly common. One might believe one of these deities is both the greatest actual being and yet fully finite: this appears incompatible with the existence of a MGB, since a MGB is the greatest actual being in every world. However, it is not clear why the belief in the finitude of the greatest actual being should significantly contribute to flourishing (see also what I will say later about atheism and negative doctrines) – this is a type-B response.

Furthermore, it might even be that the belief in question is compatible with the existence of a MGB, since if a MGB could take on a human nature, as Christians believe, perhaps a MGB could take on the nature of a great angel as well, and in a world where that happened there would be a greatest actual being who is supernatural and has a finite nature, and in that sense is finite, while yet having an infinite divine nature.

Pantheism requires a different response. One kind of pantheism simply states that, as a matter of fact, all of reality is God. Applying (4) and (4*) to this kind of pantheism only yields the claim that it is conceivable or possible that all of reality is God. But the traditional theist agrees with this conceivability or possibility claim. For there is indeed a possible world where God does not create anything, and hence He is the whole of existence.Footnote ⁴ Thus, some pantheisms may be subject to a simple type-A response.

But what about the flourishing pantheist who thinks pantheism is necessarily true? It is not completely clear that the necessity claim is what is at the motivational centre of the flourishing pantheist's life. But suppose it is. It now seems likely that our pantheist will agree that God, or the whole of existence, is a maximally great being. We may now be able to distinguish what is at the motivational centre of the life of pantheists into two separate beliefs: (i) the belief that there is a maximally great being, and (ii) the belief that necessarily the whole of existence is the maximally great being. It could be that in some cases (i) is much more central than (ii), and much more productive of a flourishing life. Moreover, (ii) makes it difficult to take seriously the presence of evil, since a maximally great being does not have any evil as part, quality or privation, while the whole of existence plainly does have evil as part, quality or privation. And, of course, battling evil is an important aspect of human flourishing. Thus, while (ii) may promote certain kinds of flourishing, such as love of nature, it detracts from the struggle against evil that is an important part of our flourishing under our present circumstances. Thus, we can offer a type-B response in this case.

A different kind of paradigm case is provided by those of the Eastern religions that claim either that all existence is illusion, or at least that selfhood is an illusion. I shall call these ‘illusionist’ views, though some of the defenders of the views might prefer the term ‘non-realist’. In any case, since it is essential to the concept of God that He have selfhood and not be illusory, these claims are not compossible with theism. Here, there does not seem to be available a type-A response.

One kind of type-B response would be to note that in that in the lives of many ordinary adherents of these religions, the heady illusionist metaphysics and epistemology is not what is motivating them. Rather, they are motivated by the belief that not focusing on their own selves will, through a cycle of rebirth, yield happiness. But this belief is compossible, or at least co-conceivable, with theism. However, likewise some ordinary theists may not be motivated by the existence of a MGB, but by the existence of a very great being, and in any case this line of argument will not apply in all cases.

Instead, I want to push a different type-B response. The belief that all existence is illusion or that there are no selves, if taken seriously, detracts from human flourishing. A central part of human flourishing is love for and friendship with others. But to love someone is incompatible with (consistently) disbelieving in her existence or her selfhood. One can love the idea of someone while disbelieving in her existence, but that is a different state. Love calls on us to appreciate persons as independent selves for qualities that they in fact genuinely possess, and in loving someone, we do appreciate the person for qualities we take or at least hope her to possess. But if we disbelieve in the person's existence, we surely have to disbelieve in her qualities as well, and insofar as our disbelief is firm, we do not even hope that she has these qualities. And if we disbelieve that the person has selfhood, then our love cannot consistently respect her in the distinctive way in which we love calls us to respect beings endowed with selfhood.

Granted, a person who believes that all existence is illusion can exhibit love. But this love will be inconsistent with that belief. Now, given how central love and friendship is to human flourishing, this will not be a case where the central excellences in human activity flow from the belief in question, and hence this will be a case where the more precise version of the principle at the end of the second section (or its possibility/conceivability variant) does not assign all that much probability to the possibility of the belief that all existence is illusory.

But in the cases both of illusionism and pantheism, one can also make a type-C move. We have decisive arguments against the possibility of illusionism and the possibility of necessary pantheism (i.e. of the doctrine that pantheism holds necessarily). These arguments defeat the weight that (4) and (4*) give to these possibilities, and leave theism unaffected. The argument against the possibility of illusionism is that, plainly, if something is illusory, then there is a self to whom it appears as an illusion and so neither all of existence nor all selfhood is illusory. Granted, educated non-realists will not be convinced, but the argument is, nonetheless, self-evidently sound.

The argument against necessary pantheism, and hence by S5 against the possibility (or conceivability – for in this case, like that of theism, conceivability, and possibility come to the same thing) of necessary pantheism, is that God is perfectly good, and a perfectly good being cannot have evil as part, quality, or privation, while in fact there is evil in reality, so it cannot be that necessarily God is the same as all of reality. Of course there is more to be said about these arguments, but the basic strategy is clear: I can directly argue against the possibility of the doctrines, and thereby undercut the evidence against theism that (4) and (4*) yield.

Of course, my opponent might make the same move, by using the problem of evil. A disadvantage of a probabilistic argument is that it is always open to responses that attempt to decrease the probability of the conclusion through independent arguments. But of course this paper is not the place to give a response to the problem of evil – to do that, I would have to avail myself of all the available proposed theodicies, and maybe a dollop of sceptical theism.

The final, and most difficult, case I shall consider is that of the non-religious atheist – someone who disbelieves in the existence of God, and does not espouse any non-theistic religious view. The case of religious views that include atheism, such as Samkhya, can perhaps be handled more easily with a type-B response: what is motivationally central are beliefs about liberation and how one can achieve it, rather than the atheism itself.

With regard to non-religious atheism (which I will just call ‘atheism’), I will bite the bullet and say that there are very few if any cases of atheistic individuals or communities where the atheism, as such, motivationally contributes significantly to flourishing and humanly excellent activity. Atheism is a negative doctrine, and one does not expect belief in purely negative doctrines to make significant contributions to flourishing when they are motivationally central – our lives should be focused on how things are rather than on how they are not.Footnote ⁵ While atheism may make one somewhat more self-reliant, it is also a view that takes away from the apparent value and meaning of the universe and other people. For any intrinsic value in the universe or in other people that the atheist can acknowledge is one that the theist can acknowledge just as well – but the theist can also acknowledge a deeper value of the universe and people in virtue of divine creation and participation in divine goodness. This is a type-B response.

Moreover, to continue this type-B response, one might argue that while it seems likely that theism tends to be central to the motivational life of theists, to many atheists atheism is not motivationally central. The atheist probably does not kiss her husband because she thinks there is no God, and it is arguably unlikely that she helps the needy because she thinks there is no God (though see below for the ‘we-are-on-our-own’ view), and so on. But a theist might well kiss her husband because she believes that the spousal relationship is a reflection of God's love, and she might help the needy because the needy are children of God. In fact, it seems to me that there is something crabbed in a life motivationally centred on a negative doctrine like atheism.

Still, there are atheists who insist that atheism implies that we are on our own, and that this can be a central, background motivation in a flourishing life (Howard Sobel, conversation (2009); see also Russell (Reference Russell1917) ).

At least three moves can be attempted in response. First, we could quibble that atheism does not entail that we are on our own: atheism is compatible with conspiracy theories about aliens controlling human history. Nonetheless, atheism combined with commonly held and uncontroversial beliefs does imply that we are ‘on our own’, and to get motivational power in theism, we also need to combine it with commonly held and uncontroversial beliefs – thus, the belief in a maximally great creator of all contingent beings only gives reason to treat one's spouse well when the theism is combined with the claim that one's spouse is a contingent being. So this response probably fails.

Second, we can insist that the belief that we are on our own is less central than the corresponding theistic beliefs, thereby giving a type-B response. In the second section, it was noted that the greater the degree of centrality, the more evidence there is for the possibility claim. Hence, if there is less motivational centrality of atheism for the flourishing atheist than of theism for the flourishing theist, we should be able to say (assuming all else is equal) that the flourishing atheists motivated by their atheism provide less evidence for the possibility of atheism than the flourishing theists motivated by their theism provide for the possibility of theism. Hence, the balance of the evidence from principles (4) and (4*) would still in favour of the possibility of the existence of a MGB.

So, is ‘we are on our own’ less central than corresponding theistic beliefs? Well, one way that a ‘we-are-on-our-own’ belief motivates is by making it impossible to leave things to providence or other supernatural influences – if something is worth doing, it is we who need to do it. But if that is how the belief functions, then it functions as a way of removing a potential defeater: belief or hope in supernatural things, allegedly, provides a defeater for reasons to aid our neighbour, and the belief that ‘we are on our own’ removes that defeater. The rational role of the belief is, thus, negative – it does not provide a reason on its own, but removes a potential defeater. This appears to be motivationally less central.

In fact, if we see the most motivationally central beliefs as ones that actually rationally move us, then it will be relevant that ‘we are on our own’ lacks rational motive force on its own – it only counters the allegedly pernicious force of supernaturalism. And whether supernaturalism in fact has such force is unclear. After all, as we face a world of vast problems, it may be that only with a belief that we are not on our own can we rationally avoid despairing over whether our own actions have a chance of making a difference.

On the other hand, there is a another way that ‘we are on our own’ functions, and this is perhaps best seen in Russell (Reference Russell1917). We cower together in our smallness before the vast uncaring cosmos, and this moves us to cherish our fellowship with one another. We can, however, likewise be moved to fellowship by the common apprehension of the mysterium tremendum et fascinans of the MGB, and I want to suggest that this fellowship is likely to be the more flourishing one, since it is not centred on something negative, our smallness and the uncaringness of the universe, but on something positive.

Being united by a common enemy – the vast universe – may be phenomenologically experienced as a deep unity, but probably is not the best kind of fellowship. One observes, for instance, the way revolutionaries fight with apparently deep unity against a regime – and then turn on one another when the enemy is gone, thereby providing evidence that there was something lacking in their apparent unity. Furthermore, the kind of attitude of standing firm in the face of the vast cosmos just does not appear to be as flourishing a relationship to that cosmos as the warmth in St Francis's song about ‘our brother the sun’ and ‘our sister the moon’ (Arnold (Reference Arnold1962), 224).

Third, we can say that there is a sense of being ‘on our own’ that is compossible or co-conceivable with theism: while the theist does not believe that God retired from active duty in the way that deism holds He did, nonetheless the theist can hold that God could have chosen to do so. Thus, the theist can hold that deism is possible or conceivable, and hence that it is possible or conceivable that we are indeed ‘on our own’. This leads to the question whether we should take ‘we are on our own’ or atheism to be the motivationally central belief in the case of the atheists under consideration. If it is the ‘we-are-on-our-own’ belief, then a complete type-A response can be given.

Let me emphasize that I am not claiming that there are no atheists who lead flourishing lives, or even that there are few of them. Rather, I am claiming that it is unlikely that, insofar as their lives are flourishing, atheists are centrally and rationally motivated by atheism.

One might, of course, as a referee for this journal suggested, consider in place of atheism a different belief that entails atheism: the belief that all causation is natural causation. One might think that this belief fares better. After all, are not many great scientists rationally motivated in their flourishing search for an understanding of the physical world by this belief? However, first of all, the most central of beliefs are going to motivate not just one's scientific life, but also one's interpersonal life. And there the belief that all causation is natural causation does not seem helpful. It might even induce worries about free will and responsibility that are deleterious to flourishing. The theistic scientist, on the other hand, can be motivated both in her interpersonal life by her seeing others as images of the MGB as well as in her scientific life by a belief that a MGB exists and would likely give created agents epistemic powers at least somewhat commensurate with their thirst for knowledge.

Moreover, it is not clear how rational the move from a belief that all causation is natural causation to being motivated to search for an understanding of the physical world is. It might be thought to work something like this: if all causation is natural causation, then that gives one reason to think we can understand the world through empirical investigation, and that makes the empirical investigation worth trying. However, surely, this is enthymematic. That all causation is natural causation is prima facie compatible with the claim that the things we seek causes for in fact have no natural causes because they have no causes at all. It is only when we conjoin the naturalist claim with some version of the Principle of Sufficient Reason (PSR) that we get the claim that the things we seek causes for have natural causes. And it is not clear that an intellectually sophisticated atheist can reasonably believe the PSR, since the cosmological argument becomes available, or can find an appropriate weakening of the PSR, and hence it is not clear that the move from ‘all causation is natural causation’ to being motivated to engage in empirical investigation is a rationally justified one.

But the less rational the move from the belief to the motivation, the weaker the evidence for the possibility of the belief according to the remarks at the end of the second section. Furthermore, the naturalist scientist needs some reason to hope that the causes are ones that we can in fact find, and in fact it seems that it is the theist who is better placed to have that hope for the reasons in the last paragraph. It is indeed plausible that ‘science is an outgrowth of western Latin Christianity, connected with it in much the same way as Gothic architecture’ (van Inwagen (Reference van Inwagen and Morris1994), 53).

Let me end the discussion of atheism with this pertinent quotation:

Modal aspects

Plantinga's discussion of his ontological argument uses the notion of a being's being maximally excellent in a world w. Plantinga says that a being is maximally excellent in w only if it is omnipotent, omniscient, and perfectly good (to which one should add: and only if it is the creator of any and all contingent beings), and a being is maximally great only if it exists in all worlds and is maximally excellent in them all. The objectionFootnote ⁶ now before us is that while the belief that there actually is a maximally excellent being is central to the lives of flourishing theists, the belief that there is a maximally excellent being in all worlds is not central to the lives of flourishing theists.

I think this kind of objection may be predicated on a mistaken understanding of how we should read Plantinga-style ontological arguments. On one reading, we begin by stipulating that a maximally great being is one that exists in all worlds and is maximally excellent in them all. On that reading, when I claim that the belief that there is a maximally great being is central to the lives of many flourishing theists, I may be claiming the centrality of the belief that there is a being that is maximally excellent in every world.Footnote ⁷ And then the objection that what is motivationally central is only this-worldly maximal excellence is salient.

However, we should not read Plantinga-style arguments as proceeding by stipulation. After all, Plantinga never defines maximal greatness as maximal excellence in all worlds. He only claims that maximal excellence in all worlds is a necessary condition for maximal greatness. Rather, we should understand the arguments as starting with the property of maximal greatness, and then arguing that maximal greatness in fact entails maximal excellence in all worlds (cf. Plantinga (Reference Plantinga1978), section 10.7) – that a being that was only, as a contingent matter of fact, maximally excellent would not be maximally great, say because it would not be a being than which a greater cannot be conceived, or because a being that is excellent only contingently has a certain modal fragility that derogates from greatness. Thus, while I initially said that it is the possibility premise (2) that is the main controversial premise, nonetheless premise (1) is a genuinely substantive though plausible claim about what maximal greatness in fact entails, a premise that can also be disputed.

One way to see what is going on is to think of a maximally great being as one that has all perfections. Suppose that Plantinga is, in fact, right that it is a perfection to have maximal excellence in every world. It does not follow from this that the belief that there is a maximally great being is only motivationally central to x if the belief that there is a maximally excellent being in every world is motivationally central to x. One can believe that there is a maximally great being – say, a being that has all perfections – without believing that there is a being that has maximal excellence in every world, just as one can believe that Frank has inherited all of Susan's property without believing that Frank has inherited Susan's horse (say, because one does not know that Susan had a horse, or because one fails to draw the relevant inference).

That God exists and is maximally great is, I think, motivationally central to the lives of many flourishing people and communities. If my arguments are right, it follows that it is likely that possibly or conceivably, there is a maximally great being. But if being maximally great in fact entails being maximally excellent in every world, and if ‘there is a maximally great being’ has the same primary and secondary intension, then it follows from this that it is likely that in every world there is a maximally excellent being.

It is not relevant to my arguments whether, in fact, the belief that God is maximally excellent in every world is motivationally central to anybody. What I need is the centrality of the belief that there is a maximally great being (or, alternately or perhaps synonymously, a being than which a greater cannot be conceived or one that has all perfections – I can run the argument on any of the three formulations). And that is central to the lives of many flourishing theists. Nor do I need the claim that maximal greatness entails maximal excellence in all worlds to be central to anybody. I only need the claim to be true.

One might modify the objection, by saying that there is a small list of perfections, as omniscience, omnipotence, and perfect goodness, and all that is motivationally central to the lives of flourishing theists is the claim that there is a being that has these perfections. But I think this is mistaken – the fact that God has all perfections is motivationally powerful. For instance, a part of what may motivate a theist to imitate God's mercy is the idea that mercy is the sort of thing that is exhibited by a being that has all perfections. And a significant part of what makes the Christian belief in the Incarnation be so motivationally powerful in the lives of theists is the idea of the humility of a being that has all perfections becoming one of us.

Moreover, perhaps another belief could have played just as central a role in the motivational lives of theists as the belief that there is a maximally great being. Maybe the belief that there is a being that is maximally great except for not existing in any world containing unicorns could have been just as motivationally central. (I am not sure. Beliefs with exception clauses are less likely to be motivationally central.) But I am not saying that a belief that could be motivationally central to a flourishing life is likely possible or conceivable. I am saying that a belief that is motivationally central to a flourishing life is likely possible or conceivable.