The knowledge and freedom of Christ

Consider the scope of Christ's knowledge, both what he knows via his human intellect and what he knows via his divine intellect. Call the knowledge which the person of Christ had in virtue of his divine intellect his ‘divine knowledge’, and call the knowledge which the person of Christ had in virtue of his human intellect his ‘human knowledge’.

The scope of Christ's divine knowledge just is the scope of the knowledge of God. And since God is omniscient, it follows that Christ, too, via his divine knowledge, is omniscient. For anything there is to be known, Christ knows it via his divine knowledge.Footnote ⁵

The scope of the human knowledge of Christ is a more contentious topic. Nearly everyone believes that there is something or other of which Christ's human intellect was ignorant.Footnote ⁶ For instance, Aquinas thought that Christ did not know, via his human intellect, the full scope of the divine power.Footnote ⁷ That said, there is also a consensus, at least according to Aquinas, that Christ's human knowledge includes knowledge of all future occurrences.

He writes:

On this view, which Aquinas says is universally affirmed,Footnote ⁹ Christ knew,Footnote ¹⁰ via both his divine and his human knowledge, all that would happen in the future.Footnote ¹¹ St Gregory the Great, too, affirmed this in a letter to Eulogius, Patriarch of Alexandria, in the year 600.Footnote ¹² Even today, professors at Catholic seminaries are forbidden from teaching that Christ didn't know all future states by his human intellect.Footnote ¹³

Furthermore, in addition to having exhaustive knowledge about the future states of the world, Christ's beliefs about the future were both certain and infallible. For, one present truth Christ would know via his human intellect when thinking about the future truths is that the source of these truths is God's direct revelation of them to him. He would know the source to be a truth-teller with the impossibility of being mistaken, and such knowledge would breed psychological certainty for Christ. Call the claim that Christ, in virtue of his created human intellect, had certain, infallible, exhaustive foreknowledge the Foreknowledge Thesis.

In this article, I will not defend the Foreknowledge Thesis. Rather, I will assume it for the sake of argument (along with the initial assumption of Conciliar Christology). Then, I will ask whether, if the Foreknowledge Thesis is true, Christ could be free with respect to his human will. In this article I consider an argument for the following thesis, which I will call the Conditional:

If the argument is successful in proving the Conditional, then any proponent of both Conciliar Christology and the Foreknowledge Thesis would, by modus ponens, be rationally committed to affirming that Christ's created human will was not free.Footnote ¹⁴ But, in so far as many – perhaps all – of the people who affirm the Foreknowledge Thesis are also people who wish to affirm the freedom of Christ's human will (again, Aquinas is in this camp), such a conclusion would be most unwelcome.Footnote ¹⁵

Indeed, while it is not itself a teaching of Conciliar Christology, the view that Christ's human will was free is certainly a common claim in church history, and important enough to have been expressed in other (non-ecumenical) conciliar statements. For instance, a Lateran Council in 649 includes:

Were the Foreknowledge thesis and Conciliar Christology to entail that Christ's human will was not free, this would be a serious problem for a large swath of traditional Christianity.

In what follows, I will be discussing the will and intellect that Christ had via his human nature unless I explicitly state otherwise.

The problem of explanatory priority

Consider the relation of explanatory priority. A is explanatorily prior to B when A is at least a partial explanation for B. For instance, I leave work at 4:30pm, at least in part because I believe that my wife has an appointment at 5pm. This belief is explanatorily prior to my leaving work at 4:30pm. Explanatory priority is transitive. If this belief is at least partially explained by my wife's telling me that she has an appointment today at 5pm, then my leaving work at 4:30pm is at least partially explained by my wife's telling me that she has an appointment today at 5pm. Note, though, that nothing is explanatorily prior to itself. It is incoherent for me to say that I leave work at 4:30pm, and this is explained, at least in part, by my leaving work at 4:30pm.

Alexander Pruss and Michael Rota both discuss the following difficult case of prophecy, which is useful for seeing the difficulty that arises for Christ's knowledge.Footnote ¹⁷ Jesus tells Peter that Peter will deny him three times before the cock crows.Footnote ¹⁸ Suppose that Peter denies Jesus, at least in part, because Jesus foretold it. Say, for instance, Jesus' foretelling puts the thought in his mind, and he thinks, ‘well, my Lord did tell me it would happen like this’. And suppose, as seems right, that Jesus knows that Peter will deny him three times before the cock crows, at least in part because Peter does in fact act thusly. In this case, Christ foretells the denial at least in part because Peter denies, and Peter denies at least in part because Christ foretells. Thus, by the transitivity of explanation, Christ foretells at least in part because Christ foretells. But this entails that Christ's foretelling is explanatorily prior to itself, which is impossible. What has gone wrong here?

Rota and Pruss both point to the same solution. They both reject that Peter denies Jesus, at least in part because Jesus foretells it. And they both point to scriptural reasons for thinking that this is the right reading of the event.Footnote ¹⁹ So, one way to get out of the difficulty is to deny that the earlier foretelling has any sort of explanatory role in the later action. If Peter's denial explains, at least in part, Christ's believing that Peter will deny, and Christ's believing in turn explains, at least in part, Christ's foretelling of Peter's denial, then Christ's foretelling cannot explain, even in part, Peter's denial.Footnote ²⁰

This response is adequate in the case of Peter's denial. But what if we change the example from an action of Peter's to an action of Christ's? Consider this case: Christ freely wills, with his human will, to drive the moneychangers out of the temple at t1.Footnote ²¹ His freely willing to run out the moneychangers explains, at least in part, his earlier knowing that he will run them out at t1. It had better not be that his earlier knowing that he will run them out explains, in any way, his later freely choosing to run them out. For, if his earlier knowing were to explain, in some way, his later choosing to run them out, then we would have an explanatory circle. His later choosing to run them out would explain, at least in part, his earlier knowing that he will run them out, and his earlier knowing would explain, at least in part, his later choosing to run them out. So by transitivity of explanation, his choosing to run them out would explain, at least in part, his choosing to run them out. And that's incoherent.

One might argue this way (where ‘→’ symbolizes material implication):

1. 1. If the Foreknowledge Thesis is true and Christ freely chooses what he does, then it is possible that Christ's foreknowing that he will perform action B is explanatorily prior to his B-ing [(FT&F)→◊EP(f,b)].

2. 2. It is necessary that Christ's B-ing is explanatorily prior to his foreknowing that he will perform action B [□EP(b,f)].

3. 3. It is not possible that (Christ's B-ing is explanatorily prior to his foreknowing that he will perform action B and Christ's foreknowing that he will perform action B is explanatorily prior to his performing action B) [∼◊(EP(f,b)&EP(b,f) )].

4. 4. Thus, it is not possible that Christ's foreknowing that he will perform action B is explanatorily prior to his B-ing [∼◊(EP(f,b) ) from 2, 3].

5. 5. Thus, it is false that the Foreknowledge Thesis is true and Christ freely chooses what he does [∼(FT&F), from 1, 4 MT].

6. 6. Thus, if the Foreknowledge Thesis is true, then it is false that Christ freely chooses what he does [FT→∼F, from 5].

Premise 6 is a reiteration of the Conditional. The argument is deductively valid. Thus, if the premises are true, we will have derived the truth of the Conditional. Are the premises true?

Premise 1 states that if Christ's human will was free, and if he had exhaustive and certain foreknowledge of all future states, then it is possible that that knowledge explain, at least in part, why he chooses something. But why think that?

It seems possible that Christ, prior to deciding what to do in a particular circumstance, could turn his attention to his knowledge of what he in fact will do. For instance, the Gospel of John (18:4) says that Christ, while in the garden, knew all that would befall him. One of those things he knew would befall him was his later discussion with Pilate, where he said (John 18:37), ‘Every one that is of the truth, heareth my voice’.Footnote ²² It seems possible, furthermore, that Christ, when later talking with Pilate, could recall his knowledge in the garden, and could, at least in part, say what he does because that was what he remembered he would say. I see no reason to doubt either of these possible seemings, or to doubt the possibility of their conjunction. If Christ knows the future, and he is free to turn his mind's eye hither and thither, what precludes the mere possibility of his turning his mind's eye to his future actions, and later recalling these in a moment of choice concerning those actions?Footnote ²³

Note that the above support for Premise 1 does not require that Christ, at any time in his actual life, was motivated by his foreknowledge in deciding how to act. The premise does not state that Christ acted in this way; it merely states that it is possible that Christ acted in this way. An objection to this premise fails, then, if it merely shows that Christ did not act in this way; it likewise fails if it merely shows that Christ would not act in this way. For an argument against this premise to succeed it must show that Christ could not act in this way.Footnote ²⁴

But if one supposes that Christ has fully accessible, certain foreknowledge of what he will do at every moment in his life, and that he is free in his choices of what he considers prior to deciding something, why couldn't he consider his knowledge of his future actions? And if he could consider them, why couldn't he be motivated by that consideration?

Consider an analogous case: nothing bars Christ from considering someone else's future actions when deciding what to do at a time. He considers Peter's future denying before mentioning that denying to Peter. Furthermore, it would be bizarre if that consideration played no explanatory role in Christ's saying what he said to Peter. But now if Christ can consider someone else's future actions and be motivated to action, at least in part, because of those actions, what bars him from considering and being motivated by his own future actions likewise? For my own part, I see no such impediment. And so, for my part, I concede the truth of the premise.

Premise 2 gains support from the basic truth that ‘truth depends on reality’. It is because things are the way they are that the truths are the way they are. And even if someone does not affirm truthmaker theory, one should still think that if it is true that S does A, this will be true, at least in part, because S does A. Furthermore, it is because the truths are the way they are that Christ knows what he knows. For his human intellect, and for every other human intellect, knowing that p is explained, at least in part, by its being true that p. And so, given the transitivity of explanatory priority, Christ's B-ing is explanatorily prior to his knowing that he will B. Moreover, if it could not but be that if it is true that S does A, this will be true, at least in part, because S does A, then, as Premise 2 says, it could not but be true that Christ's B-ing is explanatorily prior to his foreknowing that he perform action B.

Premise 3 tells us that it is impossible that both Christ's doing something is explanatorily prior to his knowing that he'll do it, and his knowing that he'll do it is also explanatorily prior to his doing it. It cannot be the case that both A explains B and B explains A. Nothing can be prior to itself in the same sense of ‘prior’. And, in particular, Christ's acting as he acts cannot be both prior and posterior to Christ's foreknowing that he acts in that way.

Steps 4–6 are derived validly from Premises 1–3. If it is true that it is impossible that both X and Y be true (as 3 says) and it is also true that Y is true in all possible situations (as 2 says), then X can be true in no possible situations (as 4 says). For, if X were true in even one lone possible world, since Y is necessary, both X and Y would be true in that world. But then, contrary to 3, it is not impossible that X and Y both be true. Step 5 follows straightforwardly from a modus tollens, and 6 is equivalent to 5. But 6 is the Conditional. Thus, if Christ had exhaustive, certain foreknowledge in his human intellect, then his human will was not free. How ought a proponent of Conciliar Christology and the Foreknowledge Thesis to respond?

Responses to the problem

Consider, as a first response, the move that Rota and Pruss use in the Peter case. Can we claim that Christ's knowledge of what he will do does not play an explanatory role in what he chooses to do, because Christ forgets what it is that he will do prior to making his free choices? It seems to me that we cannot. Peter can forget, and hence not know that Christ has foretold the denial. But Christ cannot forget or be ignorant of anything, given this exhaustive foreknowledge of his own actions. So we cannot break the circle in explanation by appeal either to Christ's ignorance or to his forgetfulness.

As a second response, we might tweak the first. Perhaps we can claim that Christ suffers neither forgetfulness nor ignorance, and yet still does not consider what he will do prior to deciding what to do. For, we might say that Christ did not have occurrent knowledge at all times, in his human intellect, of all future events. On this response, he didn't, for instance, when flipping a table at t1, have occurrent knowledge of every instant of your life, including this one. Or, at least, we need not say that he had to have such occurrent knowledge. But then, if he did not need to have such occurrent knowledge at all times, perhaps he did not have the relevant occurrent knowledge at the times when he was choosing what to do. Then we could say that his decision at t0 of what he was to do at t1 was not explained, even partially, by his knowledge. And so Premise 1 is false.

In reply to the second response, grant that occurrent foreknowledge is what is required for explanatory priority, and grant that Christ needn't have occurrent foreknowledge of all future times at every moment of his existence. I deny that this entails that he would lack occurrent knowledge (and hence certainty) when deciding what to do. It seems to me that dispositional knowledge becomes occurrent knowledge the moment one turns one's mind toward the dispositional knowledge. And so the moment Christ turns his mind to his future activity at t1, he would have occurrent knowledge of what he in fact will do at t1 (given the Foreknowledge Thesis). And so even if he doesn't always have occurrent knowledge of all future events, he has it at all the moments at which he needs not to have it in order to decide what to do.

This reply to the second response is stronger than needed. For all the proponent of the argument from explanatory circularity needs is the mere possibility that Christ's dispositional knowledge of what he will do becomes occurrent knowledge. The second response to the argument from explanatory priority works only if it gives reason to think that Christ could not have occurrent knowledge of what he will do before he does it. But the second response gives no reason to think that it is impossible for Christ to have such knowledge. And so the second response fails.

As a third potential response to the argument from explanatory priority, one might deny Premise 2 because one might think that Christ's future action is not explanatorily prior to his past foreknowledge. For instance, suppose one thought that the divine intellect's knowledge is explanatorily prior to all the happenings in the world (some Thomists hold this view).Footnote ²⁵ And suppose that God infuses the foreknowledge into Christ's created intellect in such a way that the total explanation for why Christ believes what he does about the future is that it was infused into him by the divine intellect. In that case, the divine knowledge is explanatorily prior, at least to some extent, to how the world is (in particular, to how Christ acts). Furthermore, the divine knowledge is also explanatorily prior to Christ's foreknowledge. And, the objector claims, the divine knowledge which is explanatorily prior to Christ's foreknowledge does not involve or employ Christ's future action as a means to explaining his foreknowledge. In that case, there would be no worry of explanatory circularity, since at least one half of the circle does not obtain – the future action does not explain, even in part, the past foreknowledge.

Suppose this story is right. God's knowledge is the common cause both of how the world is, and of Christ's foreknowledge, which God infuses into Christ's human intellect. We might still ask: why does God infuse these beliefs into Christ's human intellect, rather than a set of beliefs that contains all falsehoods? And here, I think the explanation would be: God infuses these beliefs because these are the ones that get reality right; they are the ones that are true. Christ's foreknowing that he will B is explained, at least in part, by God's infusing such knowledge into his created human intellect. And God's infusing such knowledge into his created human intellect is explained, at least in part, by the world's being the way it is, and, in particular, by Christ's future B-ing. Thus, by the transitivity of explanation, Christ's future B-ing explains, at least in part, Christ's foreknowing that he will B. And so even if, as the objector maintains, Christ's future action is not explanatorily prior to his foreknowledge in any direct manner, I still think there is reason to deny that it is not explanatorily prior in some way or other. For, there is reason to think that even if the divine knowledge is ultimately explanatorily prior to all that occurs in the world, it is still, contrary to the claims of the objector, because the world is a certain way that God infuses these exact beliefs into the created human intellect of Christ. I don't think that this response, even if true, would show that Premise 2 is false. There's still an explanatory dependence – even if causal independence – between how the world is and what Christ believes.

In a fourth response to this argument, the response I favour, one can distinguish Premise 3. I concede that it is impossible for X to be explanatorily prior to Y and Y to be explanatorily prior to X, when explanatory priority is understood in the same sense in both cases. But I deny that it is impossible for X to be explanatorily prior to Y and Y to be explanatorily prior to X, when explanatory priority is understood in different senses. Consider an Aristotelian analogy. While it is impossible for X to be causally prior to Y and for Y to be causally prior to X in the same sense (e.g., efficient causation), it is not impossible for X to be causally prior to Y and for Y to be causally prior to X in different senses. Aquinas gives the example of walking: walking is the efficient cause of health, and health the final cause of walking.Footnote ²⁶ If it makes sense to say that there are two notions of explanatory priority in the argument presented in the previous section, then we can distinguish explanatory priority into those senses and conclude that the argument is invalid due to equivocation.

I believe that there are two senses of priority in play here. We can call them motivational priority and unique truthmaker priority, and define them as follows:

Some examples of what I mean by ‘the propositional content of Y’ may be of help here. If I know that you are coming over for dinner, then the propositional content of that knowledge state, as I mean the term ‘propositional content of that knowledge state’, is the truthbearer that you are coming over for dinner.Footnote ²⁸

Consider some examples. As an example of motivational priority, one event is my coming home early, and another is my wife's telling me that she has an appointment at 5pm, and at least part of the reason I acted to bring it about that I come home early is because my wife told me what she did. As an example of unique truthmaker priority, Bob's existing is prior in the unique truthmaker sense to my belief that Bob exists since that Bob exists is a truthbearer, and nothing that doesn't include Bob could be a truthmaker for that proposition. Bob is also a truthmaker for that something exists and that a human exists, though he is not a unique truthmaker for either. He is not a unique truthmaker in so far as there are things that are not, or do not include Bob (e.g. you or me) that could be truthmakers for that something exists or that a human exists.

By my lights, Premise 1 employs motivational priority but not unique truthmaker priority, and Premise 2 employs unique truthmaker priority but not motivational priority. In what follows, I will argue for this claim.

One plausible way to understand Premise 1 is to take it to claim that if both the Foreknowledge thesis and the claim that Christ is free are true, then it is possible that part of the reason why Christ chooses to do what he does in at least one circumstance is because he knows he will do it. Focusing on the consequent, we see that it is an assertion about his motivations, that is, the reasons why someone does something. The argument I marshalled in favour of Premise 1 in the previous section assumed it to be a form of motivational priority that was in question in the consequent. So there is some reason to understand Premise 1 as employing motivational priority.

Premise 2, though, does not employ motivational priority. A necessary condition for an explanatory claim to be a claim of motivational priority is that it express why someone intends something. Premise 2, however, does not express why Christ intends to foreknow his B-ing. It expresses why it is that something is foreknown, but it is silent about what motivates Christ to foreknow. (One might think that the act of foreknowing is not the sort of thing that Christ is motivated to do.) And so Premise 2 does not employ motivational priority.

Premise 2, instead, is plausibly understood as expressing the priority of being to knowing. It gets its intuitive pull from the claim that for Christ to know that he does exactly this, it must be true that he does exactly this, and for it to be true that he does exactly this, then reality must be exactly that way. Nothing that isn't reality being that way could be a truthmaker for that truth. Thus, the sort of explanatory priority at work in Premise 2 fulfils the conditions for being unique truthmaker priority.

The consequent of Premise 1, on the other hand, does not satisfy the conditions for unique truthmaker priority. A necessary condition for the truth of the right hand side of the biconditional defining unique truthmaker priority is that the thing substituted for the ‘Y’ variable, the second relatum of the explanatory priority relation in question, be a belief or knowledge state whose content is made true by the first relatum of the relation. The second relatum of the explanatory priority relation in the consequent of Premise 1, however, is Christ's performing some action, such as driving moneychangers from the temple. And driving moneychangers from a temple is not something that is a belief or knowledge state. Thus, a necessary condition for the truth of the right-hand side of the biconditional defining unique truthmaker priority is unsatisfied; the right side is false. So the left side is false as well. Christ's foreknowing that he will perform action B is not prior to his B-ing in the unique truthmaker sense.

What if we consider cases where Christ's foreknowledge is of future belief states? In that case the Y relatum would be a belief or knowledge state. Even so, the necessary condition for the right side of the biconditional would not be fulfilled, because the first relatum must be the truthmaker for the propositional content of the second relatum to be a case of unique truthmaker priority. But we can see through an example that foreknowledge of future belief or knowledge states would not be a necessary component of any truthmaker for the propositional content of the second relatum.

If we make it explicit that the explanatory relation in the consequent of 1 is unique truthmaker priority, it reads:

Now let the action that he foreknows that he will perform in the first relatum be that of coming to know that P, and let the B-ing of the second relatum be his coming to know P. His coming to know P, unlike his belief that P, is not the sort of thing that has propositional content. Coming to know something is not the same as knowing it. And so the second relatum is not the right sort of thing to be substituted for Y in unique truthmaker priority.

But even if we allow that the second relatum is the right sort of thing to be substituted for Y in unique truthmaker priority, the necessary condition is still not satisfied. For if an action of coming to know something is to have propositional content, it will either be the content of the thing known – that is, P, in this case – or it will be the activity itself – that is, that Christ comes to know P. The truthmaker for neither P nor that Christ comes to know P is the earlier foreknowledge that Christ had. We can see this by example.

Let Christ foreknow at some time, t1, that he will, at some later time, t2, know that Timothy exists. Is Christ's foreknowledge that he will later know that Timothy exists prior to his later knowledge that Timothy exists in the unique truthmaker sense? It is only if the propositional content of the second relatum, his later knowledge that Timothy exists, must have, at least as a part, the first relatum as its truthmaker. And remember that the propositional content of that act is either that Timothy exists or that Christ comes to know that Timothy exists. But neither of these is made true by Christ's prior foreknowledge that he will come to know that Timothy exists.

Christ's foreknowledge that Timothy will exist is not a truthmaker for the propositional content that Timothy exists. Timothy is the truthmaker for such content. And Christ's foreknowledge that he will come to know that Timothy exists is not a truthmaker for that Christ comes to know that Timothy exists. The actual coming to know is the truthmaker for that proposition. And so, even in cases where the foreknowledge is foreknowledge of future knowledge states, such foreknowledge is not prior to the future knowledge states in the unique truthmaker sense.

Thus, the relata in the consequent of Premise 1 do not fulfil the conditions for standing in a relation of unique truthmaking priority, but they do fulfil the conditions for standing in a relation of motivational priority. And the relata in Premise 2 do fulfil the conditions for standing in a relation of unique truthmaking priority, but they do not fulfil the conditions for standing in a relation of motivational priority. And so there are different senses of priority being employed in Premises 1 and 2. Or, perhaps better to claim, we have, as of yet, no reason to believe that there is a single notion of priority that applies to both Premises 1 and 2.

Premise 3 states that it is impossible for Christ's acting to be explanatorily prior to his foreknowing and also for his foreknowing to be explanatorily prior to his acting. If there are two notions of priority in play in the reasoning, then this statement is not obviously true. Rather, an argument would be required to show that the two sorts of priority in question cannot form a circle. We might represent motivational and unique truthmaker priority as EP_M(x,y), and EP_T(x,y), respectively. Then, with this representative schema, the first three premises, on this understanding of the argument in question, would need to be as follows:

1. 1_M. (FT&F) → ◊EP_M(f,b)

2. 2_T. □EP_T(b,f)

3. 3_M,T. ∼◊(EP_M(f,b) & EP_T(b,f) )

Why think that Premise 3_M,T is true, though? Until an argument is given for it, the proponent of Christ's human foreknowledge and freedom need not accept 3_M,T, though she can happily accept 3.

But perhaps we are not out of the woods just yet. Perhaps there is another notion of priority that is used univocally across premises 1–3. In such a case, the charge of equivocation would not stick. The argument from explanatory circularity would still have a rendering that is sound.

For instance, one might focus on the source of Christ's foreknowledge.Footnote ²⁹ Suppose one has a causal theory of knowledge.Footnote ³⁰ On the causal theory of knowledge, ‘S knows that p if and only if the fact p is causally connected in an “appropriate” way with S's believing p’.Footnote ³¹ We might take the fact that p to be Christ's B-ing, and S's believing p to be Christ's foreknowing that he will B. In such a case, there is a causal relation that holds between what I've called ‘b’ and ‘f’. And that relation gives us at least a partial explanation of why Christ foreknows; he foreknows because he fulfils the proper analysis of knowledge (assuming, as we are, the causal theory of knowledge here). And that analysis isn't contingent; it is of the nature of what knowledge is that it requires an appropriate causal connection between the fact and the belief, or, in this case, the act and the foreknowledge. We might represent the conclusion of this reasoning as follows:

1. 2_C. □EP_C(b,f)

Christ's B-ing is explanatorily prior, in a causal sense, to his foreknowing that he will B.

One might worry that a causal theory of knowledge, conjoined with the Foreknowledge thesis, requires backwards causation, since Christ's B-ing occurs after the foreknowing, yet the B-ing is causally related to the knowing. Goldman, the originator of the causal theory of knowledge, considers a similar worry. His response is that, while there can be knowledge of future things, all that is required is that some causally appropriate relation hold between the fact and the belief (or, again, in this case, the action and the foreknowledge). One such appropriate relation is that of having a common cause, he says.Footnote ³² While this relation might work for the cases of mundane knowledge of the future that Goldman considers, the common cause relation cannot be the relation the proponent of the argument in the previous section employs.

If the EP relation employed in Premises 1–3 is a common cause relation between Christ's B-ing and his foreknowing that he will B, then Premise 3 is false. If EP_C is understood as a common cause relation, then it seems entirely reasonable to think that, necessarily, EP_C(f,b) is true if and only if EP_C(b,f) is true; that is, necessarily, if f shares a common cause with b, then b shares a common cause with f (and vice versa). So understanding the causal relation in the way Goldman understands appropriate causal relations with future actions or facts will not safeguard the soundness of the argument. Some other EP relation is needed to secure univocity across the premises, and it is incumbent upon the proponent of the argument from the previous section to provide it.

Furthermore, for the argument to be sound a modified Premise 1 would have to be true:

1. 1_C. (FT&F) → ◊EP_C(f,b)

The truth of the consequent will require the possibility that foreknowledge be a cause of later action. And that causal relation will have to be whatever sort of causal relation one uses for explaining how future actions cause earlier mental states. One wonders here, though, what reason we have to think that that causal relation holds between a person's motivations and his actions. It is odd enough to think of motivating as, literally, a causal relation of the same type as the causal theory of knowledge employs in its analysis of knowledge. It is even more peculiar to note that it isn't the standard causal relation, but some backwards facing non-common-cause causal relation. At the very least, the proponent of 1_C owes an account of why one should affirm it.

If there were a single EP relation that holds both ways between Christ's B-ing and his foreknowing that he will B, then that would be a problem for the proponent of the antecedent of 1. I conclude that if the appropriate causal connection posited by the causal theory of knowledge is one such relation that holds both ways, then the argument that it is is still wanting.