I’ve argued that the problem of no best world in fact generates a moral dilemma for God. But some interesting and important moral dilemmas for perfect beings assume that there is a morally best possible world. We can show that there are cases in which God actualizes the best possible world and nonetheless violates a moral obligation! Suppose that the earlier Moses receives his divine message the better. But suppose that the set of possible times at which Moses can receive his message begins after 7am and ends exactly at 12 noon. Moses can receive his message at any one of the infinitely many possible times after 7am and up to (and including) 12pm. An essentially perfectly good agent is required to pass the divine message on to Moses at some time after 7am, but for every time t after 7am, there is some other time t’ (t > t’ > 7am). So there is no earliest time after 7am at which an essentially perfectly good being can pass the divine message on to Moses. Now consider the principle in (1.0).
1.0 It is morally necessary that A if and only if some time t at which A is true is better than any time t’ at which ~A is true.
OA ≡ (Et)(Vt')((A is true at t) & ((~A is true at t') & (t' < t)))
Let At symbolize the proposition that the divine message is passed on to Moses at time t. Prior to 7 am, an essentially perfectly good agent is morally required not to pass the divine message on to Moses at any time after 7am.
1.1 O~A12pm & O~Aj & … & O~Ak
For any time t after 7am at which he passes the divine message to Moses there is a better time t’ (t’ < t) to pass the message on to Moses.
1.2 O(A12pm v Aj v … v Ak)
But it is better that Moses receives the message at some time during the interval [7am, 12pm) than at any time outside the interval.
(1.1) and (1.2) are consistent. But in Best World Moral Dilemmas it is necessarily true that at each time t after 7am, t is the best possible time for an essentially perfectly good being to pass the divine message on to Moses. We know that, necessarily, some interval of time or other elapses before the perfect being passes on the message to Moses. Since not even omnipotent beings can change the past we know that at each time t after 7 am, t is the best possible time to pass on the message on to Moses. For arbitrarily selected interval [7am - t) suppose that exactly [7am - t) has elapsed before the perfect being passes on the message to Moses. At time t, t is the best possible time to pass the divine message on to Moses and at every time t’ earlier than t, t’ was the best possible time to pass the divine message on to Moses. A perfect being that passes the divine message on to Moses will actualize the best world that he can, but he will nonetheless fail to fulfill some moral requirement.
So how long will God wait after 7 am? Which t will he pick?
Does he randomly select it?
Hi Mathis,
That's a good question. I don't know any in principle way to select a time. The problem of course is that it does no matter which time he selects, he will both actualize the best possible world and do something wrong.
Are you familiar with what van Inwagen has argued in his Gifford lecture series on the problem of evil?
If I recall correctly, he argued that, since there is no minimal amount of evil God must allow, the amount of evil he allows must be arbitrarily selected.
Perhaps this is exactly what would happen in this scenario as well.
Hi Mathias,
Yes, I'm familiar with PvI's argument, and I think it's badly mistaken ('On Vague Eschatology' F&P, '08). But you're right that there is some analogy with the present problem. But there is this odd difference. No matter what time t God chooses to relay the information, he does actualizes the best world he can and manages to do something wrong anyway. That's weird, I think.
As the matter stands, you have 1) given a conception of time such that for every t there is another time t' between t and 7am and 2) given a conception of moral correctness such that it is wrong to tell Moses Q at any time t' after 7am when there is another time t (where t is before t') when you could have told him Q.
You have defined moral wrongness and time in such a way that God cannot (in some cases) be morally correct. Surely this is a problem (even for agnostics who want a reasonable conception of morality). But it is not a difficult problem.
Either there are not an infinite number of times between any two times or else it is not wrong for God to not tell Moses Q at the earliest possible time. The theory of morality given requires the impossible: it requires God tell Moses Q at the first time after 7am. But there is no such time according to the theory of time supposed. So, either the theory of time is wrong or the theory of morality is wrong.
No reasonable theory of morality can require you to do something at location/time/moment specified that does not exist - for example, a theory of morality could not require that you to love your neighbor on the day between Monday and Tuesday. There is not such a day, and so your inability to follow the commandments on such a day is not a moral failing. Similarly, (if) there is no earliest time t after 7am, (then) your inability to tell Moses something at time t isn't doing something wrong.
Personally I think both the theory of time and morality given are highly suspicious, but you could keep either as long as you reject the other. Certainly any reason to keep one is a strong argument against keeping the other.
As the matter stands, you have 1) given a conception of time such that for every t there is another time t' between t and 7am and 2) given a conception of moral correctness such that it is wrong to tell Moses Q at any time t' after 7am when there is another time t (where t is before t') when you could have told him Q.
I don't think I offer a "theory of time". I assume only that in the half-open interval between (7am and 12noon], there is no instant of time that is the earliest instant after 7am. Do you think there is an earliest instant after 7am?
Second, I don't think I offer a "conception of morality". I make the perhaps most common assumption about a morally perfect being that it must perform some action A such that there is no A' that is better. In some cases this yields the result that there is no action satisfying the description. That's one way to generate a moral dilemma. In fact, Michael Slote discusses rational dilemmas that have a similar form. There are other dilemmas of roughly similar form for utilitarians. These occur for all maximizing views of rationality and morality. I see nothing particularly untoward in the description, since it is perfectly consistent. There are no contradictions forthcoming, but only a dilemma.
Sure, you give neither a theory of time nor a theory of morality. But you make claims about both. You claim that between any two times there is another and that it is morally incorrect to do something at time t when it would be better to do it at an earlier time.
I think that it just cannot be the case that you can generate a genuine moral dilemma here. You can certainly try, and have. I just don't understand why.
If there is an earliest instant, you can require the agent to tell Moses Q then. If there is no earliest instant, you can require the agent to tell Moses Q as soon after 7am as possible. Surely time can be divided up into usable moments which are discrete and numerable. We actually use such a way of demarcating time. Do this and require that Moses be told at any time between 7am and noon, inclusive of 7am.
It just doesn't seem like a compelling moral dilemma to suppose that a morally perfect being would be required to do something that is impossible due to a rather controversial claim about the nature of time.
Moral dilemmas are generally caused by the conflict of two seemingly genuine moral requirements. This is not such a case. This is a conflict between a proposed moral requirement and claim about the nature of time with which it is incompatible. But a similar (and equally plausible) moral requirement can be formulated to work with this theory of time.
It just doesn't seem like a compelling moral dilemma to suppose that a morally perfect being would be required to do something that is impossible due to a rather controversial claim about the nature of time.
Here's the claim I make. In the half open interval (7am, 12noon) there is no earliest instant after 7am. I guess I'll ask again whether you think there is such an earliest instant. If your answer is yes, I'd say that you're the one holding the controvesial position. I don't know anyone who thinks that there is an earliest instant under these assumptions. But maybe I ought ot be better informed about that.
I'm pretty familiar with the literature on moral dilemmas and I'm pretty sure (again, I could be wrong) that such dilemmas are not in general ". . .caused by the conflict of two seemingly genuine moral requirements." "Seeming requirements", as far as I can see, don't yield any all in obligations. But that is all aside.
Perhaps, you don't find it compelling. That's all good. But it oughtn't be because there are controversial assumptions. I assume that a perfect being cannot perform an action for which there is available a better (do you deny that?), and there is no instant that is the earliest one after 7am (do you deny that?). Pretty modest assumptions, I say.
I don't see any reason to think a scenario like this, where these are the only relevant considerations, is possible. In any realistic situation, I suspect there will be a multitude of incommensurable values, and there won't be the sort of domination that this case requires. For instance, maybe, the longer Moses waits for the message, the longer he gets to exercise the value of preparing himself for God's message. So, it is not better simpliciter that the message be given earlier, since then the value of Moses' preparing himself is curtailed.
I don't see any reason to think a scenario like this, where these are the only relevant considerations, is possible.
Really? But you think, on the other hand, that Grim Reaper cases require time and effort to refute. GR cases are prima facie possible and this, relatively simple case, is obviously not prima facie possible. Interesting.
Fair enough. :-)
One issue is that values of worlds must be compared holistically. There are almost always holistic factors involved. For instance, some times for delivery of the message instantiate such objectively valuable properties as "being a time t such that, when t is expressed in binary and in units of Planck time counting from the Big Bang, it represents a 48khz 16bit wave file containing a beautiful and infinite symphony." Other times harmonize better with times at which other significant events happened, and hence contribute to the lawlike unity of the world. I do not know how to describe the world you are giving without taking into account such values, all of which an omniscient and omnibenevolent being will take into account.
The cases I give do not directly involve value. And anyway they are aimed at atheists who typically don't think there is any deep correlation between value and possibility.
Hi Mike,
The argument seems to require the assumption that the best world contains a scenario like the one you describe. But why think that?
Also, you say: A perfect being that passes the divine message on to Moses will actualize the best world that he can, but he will nonetheless fail to fulfill some moral requirement.
Does it need to be the case that a perfect being fails to fulfill an obligation in order for the argument to work? On your premises, it seems like even if telling Moses were purely supererogatory, a morally perfect being who told him at 7:01 would still perform a worse act than one who told Moses at 7:02. That might still be a real problem for moral unsurpassability, even if no obligations are in play.
One more thought: I take it the reason God performs the better act the earlier he tells Moses is because telling him earlier has a better outcome than telling him later. Right?
If so, I think there will probably be a few possible ways to deny the inference from better outcome to better act. For one, you could think that goodness of acts is a function of goodness of motives; and that God's motives in telling Moses remain the same throughout the interval.
The argument seems to require the assumption that the best world contains a scenario like the one you describe. But why think that?
I think it requires that some world has such a scenario. I'm just pointing up a conceptually interesting fact that performing the best available action is consistent with doing something wrong.
Does it need to be the case that a perfect being fails to fulfill an obligation in order for the argument to work? On your premises, it seems like even if telling Moses were purely supererogatory, a morally perfect being who told him at 7:01 would still perform a worse act than one who told Moses at 7:02
That's right, I think, if God can perform a supererogatory action.
If so, I think there will probably be a few possible ways to deny the inference from better outcome to better act. For one, you could think that goodness of acts is a function of goodness of motives; and that God's motives in telling Moses remain the same throughout the interval.
Right, sure. But that is to consider a different case, right?
We can show that there are cases in which God actualizes the best possible world and nonetheless violates a moral obligation
I guess I'm thinking of an argument from Bruce Langtry to the effect that, necessarily, maximal worlds contain no bad actions. I think Langtry thinks this is supposed to be an a priori truth. I'm not sure how plausible the claim is, but it seems like someone who was really convinced of it would deny that the best world contains your scenario.
(Also, I don't see what's to stop us from concluding that the world you describe isn't a real possibility, since it contains God violating a moral duty.)
But that is to consider a different case, right?
I was thinking that your argument requires an implicit premise linking the goodness of the outcome of God telling Moses earlier to the goodness of God's act of telling Moses. What makes it the case that telling Moses earlier is better than telling him later?
I guess I'm thinking of an argument from Bruce Langtry to the effect that, necessarily, maximal worlds contain no bad actions. I think Langtry thinks this is supposed to be an a priori truth.
Purported a priori truths are corrigible. I'm offering a correction for anyone who might believe as Langtry does.
Also, I don't see what's to stop us from concluding that the world you describe isn't a real possibility, since it contains God violating a moral duty
That's one conclusion, but it seems hasty and not especially informative. It doesn't teach us anything. It would be good to know why such cases are not possible, since the failure to fulfill an obligation in a moral dilemma is not like the failure to do so in non-dilemmatic situations. We'd need to know why God could not find himself in a moral dilemma. On the other hand I've described a way in which he might find himself in one, and the assumptions I employ don't seem very complicated or implausible. One might urge (I would) that such examples give us opportunities to revise or precisify inherited conceptions of God that were never given the right sorts of testing in the first place. Conceptions which no one bothered to challenge with an inventive thought-experiment or two.