God is often defined or described as being omniscient (all-knowing), omnipotent (all-powerful), and omnibenevolent (all-loving). Theists typically hold that only one entity fits this definition or description. Many deem this to be a self-evident truth. Indeed, most theists – the ones I know, at any rate – take it for granted that, once these triune attributes are in place, monotheism follows as a matter of course. Yet, despite (or perhaps because of) its prevalence, this monotheistic conviction is rarely argued for.

I do not want to argue against theism directly. I do want to argue, though, that the ‘mono’ portion of monotheism rests on nothing more than an aesthetic preference. In order to show this, I want to copy/paste the traditional idea of God so as to obtain two Gods.

Imagine a universe ruled by two deities who fit the standard description in all respects and who have agreed to a crisp division of labour. ‘Day shift God’ oversees all the events that occur while the sun is up, whereas ‘night shift God’ oversees all the events that occur while the sun is down (since the sun always shines somewhere on our planet, we can assign a fixed reference point, say, Mecca or Rome). This cyclical division results in an exhaustive coverage. If something happens, it happens either during the night shift or during the day shift. Time of day thus determines which of the two Gods will intervene or refrain from intervening. Apart from their chosen range, the day and night shift Gods act exactly the same as the monotheistic God.

From our humble human perspective, then, the two sources of Godly intervention would be indistinguishable. Because the empirical evidence for God's existence is underdetermined, any inference from what we observe to monotheism becomes inconclusive, leaving room for a ditheist possibility.

It might be thought that the obvious response to this scenario is that if a God only has power over the day and not the night, then such a God is not omnipotent, but perhaps ‘demipotent’. Polytheist models tend to assign different attributes to different gods (duBois Reference duBois2014), so a natural reflex is to carry this uneven distribution over to any polytheist account. However, this is not applicable here, insofar as the two Gods of my model retain the standard triune attributes (I am thus using a capital G, as a reminder of this). Both Gods know what happens during the shift of their colleague, otherwise they would fail to display the required omniscience. Likewise, both Gods are equally suited to act or refrain from acting, thereby preserving their omnipotence. Finally, given that both Gods are all-loving, both care about the world and its inhabitants twenty-four hours a day. So, whenever the God of the day shift performs a certain action, the God of the night shift nods in agreement, and vice versa.

I am introducing these day and night shifts, not because I believe in them, but to tease out a strong intuition. Once this intuition has been brought into the open, we shall be able to assess its merit and status. Let us therefore ask: is the ditheistic ‘shift work’ that I have just sketched impossible? Here are three potential responses:

1. (1) It is possible that the world is governed by two Gods ranging over day and night, respectively; so the idea of there being only one God does not go without saying.

2. (2) It is possible that the world is governed by two Gods ranging over day and night, respectively; but in the world we live in there is only one God.

3. (3) It is not possible for the world to be governed by two Gods ranging over day and night, respectively.

Response 1 can be considered a concession. Although polytheist stances have historically been the object of much bias, an open-minded inquirer can simply embrace my ditheist account as a genuine theological possibility.

Response 2 counters my scenario with a hefty positive claim, namely that the world is governed by only one God. I do not think anyone is in a position to defend this stance, since most of the traditional arguments for theism can be recycled to promote a polytheistic account. The argument from design, for example, can employ the same premises and entailments to support a conclusion about two designers. Likewise, biases aside, there is no reason why cosmological arguments could not accommodate some kind of causal overdetermination. One can of course fall back on a sacred text to justify monotheism, but I would consider that a sign of philosophical failure.

Response 3 is, to my mind, the most interesting. It flatly denies the possibility that I am trying to evince. What reasons, then, might one give to support response 3? Philosophers rarely mobilize their argumentative prowess to defend a cause that they do not consider a ‘live option’ (in William James's sense). As a result, one's pre-reflective inclinations can set the direction of one's subsequent argumentation. Yet, in spite of monotheism's home field advantage, I think that response 1 is a perfectly sensible one.

It may be helpful to review all the theistic commitments that my scenario keeps intact. Am I suggesting that we live in a Godless world? Not at all. Am I suggesting that we are loved by a God, come what may? Yes – in the same manner allowed by regular theodicies (in fact, we double down, since we are constantly loved by two Gods). Can these Gods do anything? Yes, each God can do whatever they want – and what they want is to cooperate perfectly in their shared governance of the world.

Regarding this last point, a proponent of response 3 might object that, despite the non-overlapping daytime and nighttime shifts, the very existence of another God poses a cap on each God's power, thereby contravening the demand for omnipotence. This is not quite right. The Gods of my model both have day and night power (in the strict philosophical potency/act sense), but each chooses actually to deploy their gifts in a 12-hour period. Hence, despite their freely chosen domains of action, both Gods are omnipotent – they only choose to let the other God do what they otherwise would do (in the same benign way that a monotheistic God can part seas but need not always do so).

It is tempting to picture power as the occupation of some expanse or territory. On this picture, if one King-like being reigns over all the surface area available, then no other being can gain a foothold. Yet, what is at stake is not a spatial expanse but a list of powers. If we list all that the day shift God can do and all that the night shift God can do, the two columns will be identical.

James Baillie and Jason Hagen nevertheless think it is obvious that there cannot be two omnipotent beings. They use an analogy with football to illustrate why:

This analogy begs the question. Indeed, the comparison with football clubs only works because the authors have chosen a zero-sum game. If, however, we switch to another sport, the supposed obviousness of the intuition breaks down. There is no reason, for example, why two field athletes could not each put the shot 25 metres exactly.

As tempting as it is to erect systems of beliefs on seemingly obvious truths, the strategy will boost one's confidence only if one disregards other truths. For instance, one can justify a monotheist picture by invoking the fact that only one sports team can win the playoffs. However, one can also justify a ditheist picture by invoking the fact that it takes two to tango. Or again, one can say that a circle has only one centre, but one can also say that a circle has no privileged point on its circumference. Who is to say that sport is more important than dance, or that a centre is more important than a circumference? I see no principled way to adjudicate such conflicting appeals. They move only those who already agree.

If supreme greatness is defined as a one-seat position from the outset, then the philosophically relevant commitments are smuggled – without argumentation – in the very choice of definition. Figuring what fits or fails to fit with a definition is not critical work. We should instead ask why one should accept a given definition. So, while Baillie and Hagen draw an analogy with a competitive activity, why not emphasize cooperation? It is customary to assume that,

Yet, we could just as easily claim that, were there two Gods and one of them resolved on a course of action, the second God would resolve on the same course of action, thereby demonstrating that he was also an all-powerful god.

There is something somewhat macho about forcing the theological situation into a contest of wills where only one party can be the victor. One telling symptom of this is the fact that most conversations foreground omnipotence at the expense of omniscience and omnibenevolence. Looking at omniscience, why can't two beings know all that there is to know? Knowledge is not a contest or a resource that fills a receptacle, so it is unclear why one supremely knowledgeable being should exclude another. Likewise, why can't two parents love their child equally? Few monotheists worry that God will act at variance with his former decisions, so I do not see why I should worry that the day and night shift Gods will act at variance with each other's decisions.

Clearly, ‘[t]he question concerning the possibility of the existence of two omnipotent beings is crucially dependent on the adopted characterization of omnipotence’ (De Florio and Frigerio Reference De Florio and Frigerio2015: 310). Omnipotence is usually thought to be ‘the power to do anything logically possible’ (it is also sometimes defined as ‘a power exceeding that of any other possible being’, but this is clearly question-begging). Is it possible for the day shift God to disagree with the night shift God? Yes – if we drop the other attributes. But, given that the definition of God that interests me also includes ‘all-loving’ and ‘all-knowing’, it is not possible that the two Gods might disagree. That is why taking on the three attributes matters. I would thus define omnipotence in the traditional way, as ‘the power to do anything logically possible’, but I would assess this logical possibility alongside other premises that are relevant. Good logical reasoning is holistic and does not exclude relevant information. Discussions that treat omnipotence in isolation are thus a non-starter and will not yield correct conclusions.

If being on the same page about a project is a metaphysical possibility, surely two omnibenevolent Gods can realize it, if anyone can. A ditheist-friendly source of inspiration would therefore be the happy marriage, where ‘what one wants is never contradicted by the other's will’ (De Florio and Frigerio Reference De Florio and Frigerio2015: 317). Since the most philosophically important decisions are made before arguments are ever crafted, focusing on the neglected attributes of omniscience and omnibenevolence might undermine monotheism's monopoly on the mainstream theological imagination.

So far as I can see, the Godly exclusivity at play in the monotheistic picture is either a matter of necessity, a matter of empirical fact, a result of parsimony considerations, or a result of aesthetic considerations. The exclusivity in monotheism cannot be a matter of necessity, since necessity admits of no exceptions and I have put forward a counter-example. It also cannot be a matter of empirical fact, since that would require devising some means of settling the question by observable testing and no such observation or test exists. The exclusivity might be a result of parsimony considerations. However, it is unclear why ‘more parsimonious’ should be equated with ‘more likely to be the case’. After all, it would be more parsimonious for the human species to have only one sex, but the fact is that we have two. I do not see why two omnipotent Gods should feel bound by parsimony. Even when dealing with natural phenomena, it is hard to find a non-circular argument establishing that simplicity tracks truth. So, in the supernatural realm, confidence that simplicity is an argumentation-settler seems to me doubly misplaced.

I submit that, in the final analysis, the real obstacle to my day and night shift proposal is not an argument, but an aesthetic preference: it simply offends a prior sensibility to countenance two Gods instead of one. However, there is nothing in this offence that is principled (unless, that is, aestheticians have somehow devised an algorithmic procedure akin to formal logic, which they have not). In fact, one could just as easily insist that a world-view involving two Gods is more pleasing, since it celebrates a model of cooperation that monotheist theological accounts do not provide.

My day shift and night shift proposal aims to dislodge, not just arguments, but pictures, analogies, and intuitions. The influence of such cultural schemes on philosophy of religion (and philosophy generally) cannot, to my mind, be overstated. In keeping with the Jesuit adage that core beliefs are all settled before the age of seven, a person raised on my ditheist model could insist that the throne atop the theological order must be a two-seater. This necessity would be taken as self-evident. In fact, since a tag-team of Gods would mean that humans are under the protection of twice as much wisdom and care, a convinced ditheist would find it unthinkable to prefer monotheism.

Culture naturally colours how we regard conflicting hypotheses, but we should distinguish a prior conviction from a priori conviction. As a belief, monotheism is not only widespread, but dearly held. This shared emotional investment can make monotheism look like an axiom. I have argued, however, that the familiar attributes of omnipotence, omniscience, and omnibenevolence can all be maintained while countenancing two Gods, not one. My argument does not mean that one must cease being a monotheist. It does mean, though, that in the absence of better reasons, commitment to a one-God picture is very likely a matter of personal taste.