The Modal Possibilities of Salvation

Since I've started working on universalism, my reading list has shifted. Which has compelled me to read Oliver Crisp's chapter on Barth vis-à-vis universalism in "All Shall Be Well". And it's just as well that I had already decided not to go the route of comparing Barth to a predetermined standard of universalism. But from the analytic perspective, Crisp does something I find questionable in stating his definition, and I'd like to play with it more.

You see, Crisp begins with the statement that there is a difference that must be respected between "all human beings will be saved" and "all human beings must be saved." Now, on that bare point I can agree, albeit for different reasons—I mean it in deontic terms, and he means it in alethic terms. But he bases this distinction on possible-worlds semantics, such that it appears the difference is between "all actual human beings will be saved" and "all possible human beings will be saved." That is, Crisp makes the question to be answered whether universalism is true, or necessarily true.

Now, I'm not sure this is really the modal question that must be answered with regards to salvation. But the game is then to figure out what the good questions are, and to ask them.

(Warning: logical operators ahead. Set browsers to auto-detect Unicode!)

Now, the trick to Crisp's dilemma is that in Kripkean terms it might not exist. That is, if all actual human beings will be saved, and there are no actual human beings who will not be saved, then in the actual world we can represent "all human beings will be saved" as ◻S, which is the same in alethic terms (that is, "as a truth claim") as Crisp's "all human beings must be saved." (Deontic terms, in which we speak of obligation and permission, are another matter.) This claim, in either form, therefore implies that all human beings are in fact saved, in spite of the fact that the alternative is false but not impossible. It could hypothetically be otherwise, in other words, but it is not actually otherwise.

But Crisp isn't reading out of Naming and Necessity. He relies on there being a difference between this strong claim ◻S, "it is necessary that 'all human beings are saved' is true," and the weaker claim ⋄¬S, "it is possible that 'all human beings are saved' is not true." As Crisp himself puts it,
If a particular doctrine commits its defenders only to the conclusion that all human beings will be saved, then there are possible worlds in which some human beings are not saved. That is, had things worked out differently, it would have been true that not all human beings are saved (i.e., had some other state of affairs obtained, not all human beings would be saved).
Now, this gives us a window into what system of modal logic Crisp is using. Let's lay the implications out more clearly than he does in the text. First: salvation is a possibility in world w (⋄S). However, there exists a world v accessible to world w in which this possibility is not conserved. So far, we know we're not in S5, because this is not Euclidean. That is, there exist possible worlds u and v, both accessible from w, in which two different things are true. In one, universal salvation remains possible (⋄S, or even u ⊨ S if we reach the eschaton without it collapsing), and in the other it is no longer true and therefore impossible (¬⋄S, because v ⊭ S). Therefore in world w, we cannot say that universal salvation is necessarily true (¬◻S). We can only say that it is contingent, meaning that both S and ¬S coexist as possibilities of w, and so the truth value of S is indeterminate (or simply not yet determined) as a future condition.

There are two possibilities in Crisp's phrasing. Either this contingency is one in which (a) we are actually in world u, and the eschaton has been determined to involve universal salvation in this world, but we posit that there are in fact worlds like v in which it has been determined otherwise; or (b) we are actually in world w, and the eschaton—and so also human fate as to salvation—is yet to be determined by actions in the present or the future.

Now, the problem in distinguishing between these is Crisp's choice of verb tenses. He repeatedly talks about salvation in the future tense, and his negations thus seem to be in the futur antérieur. This pushes towards the second option, a conditional based on present or future actions. This is a contingency in which salvation is only ultimately determined at the eschaton. While he doesn't hold firmly to such a view, he also doesn't drop the assumptions behind it as he moves forward with his argument:
Suppose all human beings will as a matter of fact be finally saved. If it is true that nevertheless God could have brought about a world where not all human beings are saved, this has important implications for the nature and purposes of God. For if it is a matter of contingent fact that all human beings are saved, then it is not the case that God had to save all human beings. He was free to refrain from saving all human beings. That is, God could have decreed to save some fraction of humanity less than the whole, even though, as a matter of fact, he has not chosen to do so.
Here Crisp has shifted into deontology, which is a problem with common-language philosophy when one is trying to make fine modal distinctions. It's hard to stick firmly in one system without being unreadable except to logicians. (And I am obviously willing to run that risk, or I wouldn't be talking in symbols!) But let's try to be clear. In deontic terms, ◻S means that God is obligated to save all human beings, which is the normal implication of "must." The claim ⋄S then means that God is permitted to save all human beings, whether or not God actually does so. And just as in the alethic, ⋄S implies ⋄¬S, such that it is also permissible for God not to save all human beings.

The reason this shift works in Crisp's chapter is that we have a strong aversion to claiming that God is subject to necessity. Therefore, in deontic terms, with respect to God as an agent and salvation as an action, we will forcefully deny ◻S as a claim that "God must save all human beings." But in alethic terms, with respect to salvation as a matter of objective fact, Crisp has added exactly nothing to his earlier point. He has not given us a reason to reject ◻S where it means "it is necessarily true that all human beings are saved." In point of fact, as theologians, we have every reason to accept the possible coexistence of (a) both ⋄S and ⋄¬S with respect to the permissibility of God's action, and (b) ◻S with respect to the fact of our salvation.

This is because we are not in the present/future contingency that Crisp has set up as his basic dilemma. In fact, we are downstream of a very different contingency! I can certainly grant that, had things been otherwise, we might be in a world in which not all are saved. But—and especially when one is discussing Barth—the point of contingent decision here is well past: it is the incarnation, life, death, resurrection, and ascension of Jesus Christ. And that point of contingency in time is based on an eternal and non-contingent being-in-act of God. This God, for whom all possible actions are permissible and none are either obligatory or prohibited, has in fact acted to save. And whatever the scope of God's saving action, its results must be said to be necessary in all possible worlds accessible from the point of this reality of Christ.

In fact, if we suppose with the tradition after John that Christ is the embodiment of the divine logos present in the act of creation, then there are in fact no possible worlds, and no actual worlds, in which this action of God is not in fact true. There is no God who has created all things, both things that are and things that are not, and who has also decided otherwise than to act in Christ. Whatever the scope of the action of God in Christ, its reality is therefore necessarily true—true in the actual world and also true in every possible world.

Again, there is nothing inconsistent about this statement as regards the fact of the result of God's action. It is true that God is free to have acted in any such way as God should choose to have acted—and being and acting in Jesus Christ is the freely chosen expression of God's own free will. The action is contingent on God's willing it, but we have warrant to state unequivocally that God both wills it and has in fact done it. And therefore, whatever is true of salvation in Christ, it is necessarily true in this and all possible worlds.

This is why the deontic detour that Crisp makes is so misleading. Its rhetorical effect masks its lack of relevance to the discussion about what we may say is in fact true about God's action in Christ and its scope for humanity. And it wrecks his logic! It leads him to make the following four points, only the first two of which may be upheld:
  • that God is free;
  • that God is metaphysically independent of his creation, i.e., exists a se;
  • that God must act in a way consistent with his nature;
  • and that God is essentially such that, for any possible theatre of creation that includes the creation and fall of human creatures, he must bring about the salvation of all humanity in that theatre of creation.
By shifting from alethic discussion of the facts of the effects of an action to deontic discussion of the possibility of that action in the first place, Crisp has wedged two "must" statements into Barth in places where "does in fact" is the correct operator. And this will only partially redeem those last two points, because they also involve the subjection of God's being-in-act, respectively, to an objective nature that governs what God may do, and to an obligation subordinate to that nature by which God must (as though by reflex!) save things God has made.

Neither of these statements are Barth. Without diving for citations, I can tell you that I know from my reading in the CD that Barth explicitly rejects both of them. God's nature is God's self-determination, which is why we now say "being-in-act," because God's being and God's action are nothing but the uninterruptible realization of God's will. And we say of God's will, and consequently of God's being and acting, that they are eternal even as they intersect changing time, and this is the nature of God's faithfulness. God is not capricious. God is never other than who God has been and what God has done, even though God is not bound by those events in time, and can and will do new things. The consistency the events partake of is in the will of God, not in the history in which they appear. And so God may save us because we are God's own creatures, but there is nothing about our creaturely being that compels God's saving act. Nor is there anything compulsory about God that forces God to save creatures. God saves because God wills to save; God does not destroy because God wills not to destroy.

Where does this leave us? God's saving action in Christ is necessarily true, in this and all possible worlds. God has in fact acted in this way, and we are obligated to recognize that fact. What remains in question is what that fact implies for the whole of humanity in all its parts. The question of universalism is not whether God has willed and acted to save, but whether, in God's willing and acting to save, all human beings are in fact saved; or whether there is a further contingency involved in saving individuals; or whether there is no contingency involved, but the fact of the matter is that only some human beings are saved. These are the canonical options: universal salvation, contingent salvation, and partial salvation.

The problem with representing these three modally is that two of them are not future conditionals. They are attempts at describing the actual world, and only one of them can in fact be true. The question for the outside pair is whether we live in a world, in the wake of Christ, in which God has acted to save all, and therefore all are saved—or in which God has only acted to save some, and therefore only some are saved.

But this pair of mutually exclusive possibilities is itself exclusive of the other option! The middle term, the contingent option, suggests instead that God has made salvation possible for all in Christ, but not actual. And this, rather than the other two, is the universe in which Crisp's dilemma exists—in which salvation is possible but not necessary, and there is an open question as to whether or not God will in fact save. There is a possible world, in the contingent scenario, in which all are in fact saved, but it is only one choice situated among other possibilities that cannot all be realized.


Post a Comment