The Anselmian objection and omniscience as an ability

Let us begin by considering a more explicit version of the Anselmian objection:

If the kenotic theologian adopts an orthodox Christology, she must deny the conclusion, and so she must reject one of the premises. Since she accepts (3) as a part of kenoticism, the remaining options are to deny or modify (1) or (2).

Let us look at the first alternative. The kenoticist might consider modifying (1) by affirming that Christ is the ‘greatest possible being’ with respect to his divine nature but not his human nature. But if the kenotic theorist uses this strategy, she thereby loses the motivation for kenoticism in the first place. For in that case one could simply maintain that Christ lacked knowledge of the date of the parousia with respect to his human nature but not his divine nature.Footnote ⁸ But if one merely relativizes properties to different natures, then Christ does not ‘give up’ any of his divine properties at all: he continues to be omniscient. In other words, we can relativize the predicate ‘greatest possible being’ to the divine nature only if we are prepared to deny (3). But the kenoticist is already committed to (3), so she will not modify (1) in this way.

How about the second option, to deny (2)? Supposing that Christ is not omniscient, does it follow that, possibly, there is a being greater than Christ? The kenoticist might deny this by claiming that it is impossible for Christ to be omniscient given other properties that he has by virtue of being human. In other words, the kenoticist could suggest that it is metaphysically impossible that Christ instantiate both the property being omniscient and his other human properties. If that were the case, then premise (2) would be false.

Prima facie, however, there is no explicit contradiction between any human property Christ possesses and the property being omniscient. At a minimum, there is no obvious contradiction in affirming that Christ was human and that he knew the date of the parousia. But if there is no obvious contradiction, then kenoticism is in trouble because the Anselmian concept of divinity requires that Christ be omniscient if it is metaphysically possible. In sum, the Anselmian objection against kenoticism states that if Christ lacked omniscience, he would be less-than-the-greatest. But a divine being cannot be less-than-the-greatest. Since kenoticism affirms that the incarnate Christ lacked omniscience, kenoticism leads to the conclusion that Christ was not divine.

In light of the Anselmian objection, some philosophers have attempted to amend the kenotic position by showing how Christ could be both omniscient and non-omniscient in different senses. One of the most promising approaches is made by Stephen Davis (Reference Davis1983).Footnote ⁹ Davis offers the view that Christ was omniscient in the sense that he had access to all knowledge even though he willingly refrained from accessing it. Consider the following possible scenario. Let A be some subject, p some proposition, and x some action. Suppose that

Davis writes: ‘The sense in which A does not know about p is this: A cannot at present truthfully answer relevant questions about p. The sense in which A does know p is this: A can answer them at any time A wants to do so (by doing x)’ (Davis (Reference Davis1983), 126). Similarly, Christ was omniscient in two different senses:

Thus, Davis's proposal offers a response to the Anselmian objection in that it affirms that Christ was omniscient in some sense, namely, that he had the ability to gain knowledge merely by performing some other action. Call Davis's account the omniscience-as-ability view.

Though initially appealing, the omniscience-as-ability view is vague and stands in need of explication. One problem is that the view implies that omniscience is a power, ability, or capability, which seems strange.Footnote ¹⁰ What exactly does it mean to say that the possession of knowledge is an ability? An ability to do what? Knowing a certain proposition is typically understood as possessing a true belief that meets certain epistemic conditions (justification, warrant, etc.).

An even more pressing worry for Davis is that he does not describe exactly what sort of action x the agent must be able to perform such that by performing x, the agent thereby knows p. This is a tricky matter because not just any action will be adequate. Some types of actions will make the possession of knowledge too easy. Suppose my father knows all the facts about the English Premier League since the league's formation in 1992. And let us say that if I were to ask him who won the league in 1995, he would tell me the answer. But of course, it is certainly false that I know who won the 1995 title just because there is some action I can take (asking my father) such that by taking it, I would receive the answer. Thus, the sort of action that Davis affirms must be restricted more narrowly than simply any action.

Davis provides an example that might point us in the right direction.Footnote ¹¹ Suppose I ask you to tell me the sum of 5,793 and 63,201. You do not immediately know what the answer is, but you could perform some action, say a mental calculation, such that if you were to perform it you would know the answer. There might be a sense in which you ‘know’ what the sum is even before doing the mental calculation. Perhaps omniscience-as-ability could be construed in this sense of ‘knowing’. Unfortunately, this suggestion is also unclear. Do I know, for example, what the product of 5,793 and 63,201 is? Although I lack the ability to calculate the product in my head, I do have the ability to punch the numbers in my calculator. But in the latter case, it is intuitive that I do not know what the answer is without the use of my calculator.

One might attempt to avoid this counterexample by stipulating that the action which the agent performs must be internal to the agent (e.g. a mental calculation) instead of external to the agent (e.g. using a calculator). But this modification also runs into problems. Consider another example. According to the legend, Archimedes discovered his famous principle while sitting contemplatively in his bathtub.Footnote ¹² Upon his sudden discovery, he jumped up and ran naked through the city shouting ‘eureka’. Assuming that Archimedes used only internal (mental) resources while sitting in his bathtub, we would have to say that he actually knew his principle prior to his ‘eureka’ moment. After all, he had the internal ability to discover his principle beforehand. But, intuitively, Archimedes did not know his principle prior to his ‘eureka’ moment: he learned something new on that occasion. The upshot of all this is that the omniscience-as-ability view, while promising to respond to the Anselmian objection, is ultimately unclear and stands in need of modification.

Semi-kenoticism

There is a better variation of kenoticism in the same neighbourhood. My proposal is to retain the normal understanding of knowledge, namely, as the possession of a belief which meets certain epistemic conditions. Following a well-known distinction in the literature, we should then differentiate occurrent beliefs from dispositional beliefs.Footnote ¹³ Moreover, we should differentiate both of these from dispositions to believe. I will briefly elaborate on these before continuing.

Occurrent beliefs are those beliefs which are at the forefront of one's consciousness. As I type these words, I have an occurrent belief that I own a laptop. I occurrently believe that I have five fingers on each hand and that 5 + 5 = 10. Nevertheless, I have another set of beliefs that are not occurrent in this way. My dispositional beliefs are those beliefs that I possess but that are not immediately present to my mind. Such beliefs, however, can be made occurrent at some later time. Edward Wierenga explains the distinction as follows:

Examples of my dispositional beliefs might include, for example, that FC Barcelona were the victors of the 2011 UEFA Champions League, that the square root of 121 is 11, and that Alvin Plantinga is taller than Eleonore Stump. Dispositional beliefs are those beliefs the possession of which disposes their possessors to form a corresponding occurrent belief when certain conditions are met. Wierenga suggests that the condition that needs to be met for dispositional knowledge to be turned into occurrent knowledge is merely that the person consider the proposition in question.Footnote ¹⁴

A dispositional belief, however, is different from a disposition to believe. A disposition to believe is not a belief at all. Rather, it is the readiness to form a belief through some sort of process. Robert Audi (Reference Audi1994, 424) uses the analogy of a computer as an illustration.Footnote ¹⁵ Occurrent beliefs are like pictures on a computer screen, and dispositional beliefs are like pictures that are stored in the computer's memory but that can be retrieved and brought to the screen at any moment. A disposition to believe, by contrast, is not a picture stored in the computer. Rather, it is analogous to the ability a computer has to generate a brand-new picture on the screen. A designer can use the computer, for example, to draw a picture that will end up being displayed on the screen. The computer did not have the picture stored in its memory before the designer used it. Instead, the computer had the capacity or ability to show the picture on the screen as a result of the designer's input. The computer analogy is helpful for illustrating these distinctions. As I will indicate in the next section, however, the analogy falters when we consider the nature of beliefs in greater detail.

We can now see the reason why Davis's omniscience-as-ability model is inadequate: it conflates dispositional beliefs with dispositions to believe. That is why, for instance, he mistakenly maintains that one knows the results of mathematical calculations such as adding 5,793 and 63,201. In reality, the agent has a disposition to believe the answer because there is mental calculation that one can make such that one would then know the correct result. Nevertheless, she does not have a corresponding dispositional belief regarding that proposition. The fact that Davis does not distinguish between the two contributes to the vagueness in the omniscience-as-ability approach.

I suggest, along with Cullison (Reference Cullison2006), that one might think of the incarnate Christ's knowledge in terms of a mixture of occurrent and dispositional beliefs. According to the view I will call semi-kenoticism, Christ is omniscient in the full sense: he believes all true propositions, and all his beliefs meet the appropriate requirements for knowledge. On this view, however, the vast majority of the incarnate Christ's beliefs were dispositional rather than occurrent. Since Christ's knowledge involved a mixture of both occurrent and dispositional beliefs, we can say that Christ was mixed-omniscient.Footnote ¹⁶ Thus, the incarnate Christ very likely did not occurrently know about the nature of subatomic particles. He very likely did not occurrently know about the general theory of relativity, or what events were taking place on the other side of the planet, or the first 200 digits of pi. As the gospels record, he also did not occurrently know the date of the parousia. Yet Christ was divine, and he therefore knew such things dispositionally. Normal human beings convert their dispositional knowledge into occurrent knowledge on a regular basis, but they have dispositional access only to a limited range of propositions. As an omniscient being, however, Christ would have had dispositional access to all true propositions. Moreover, he would have had the ability to access knowledge that no normal human could have accessed. Thus, the gospels record instances in which Christ's knowledge is sometimes regarded as evidence of supernatural abilities.Footnote ¹⁷

At this point, one might wonder: in what sense is semi-kenoticism kenotic, since the view affirms that the incarnate Christ continues to be omniscient? The answer depends on how we think of the pre-incarnate Christ. Suppose one maintains that the pre-incarnate Christ had all of his beliefs simultaneously and occurrently. This is the way God is supposed to have knowledge on certain traditional views, such as that of Aquinas.Footnote ¹⁸ Semi-kenoticism would then be the view that Christ freely gave up one mode of omniscience – knowing all true propositions occurrently – in favour of another mode: knowing some true propositions dispositionally. So, upon becoming incarnate, Christ does not give up omniscience simpliciter but only a particular mode of omniscience.

So far I have described dispositional beliefs as beliefs that an agent has the ability to convert into occurrent beliefs. But what conditions need to be satisfied in order for this conversion to occur? For normal human beings, the typical condition that needs to be met is merely that the subject consider the proposition in question.Footnote ¹⁹ Let us say that a dispositionally believed proposition p is consideration-sensitive just in case a subject would inevitably form an occurrent belief p simply upon considering p. But this does not seem to hold in the case of Christ, for Christ considers the question ‘What is the date of the parousia?’ and still answers ‘The Son of Man does not know.’ We now have a worry: if dispositional knowledge is consideration-sensitive, Christ does not seem to have such knowledge.

In response, it is not clear that all dispositional beliefs must be consideration-sensitive. Notice that the definition of consideration-sensitivity requires that the subject would ‘inevitably’ form the occurrent belief upon considering the dispositional belief. But as long as ‘inevitably’ is not understood as ‘necessarily’, it is broadly logically possible that someone should both consider some dispositional belief yet fail to make it occurrent.

Consider the following scenario: I have memorized the directions from Chicago to Washington DC. I have dispositional beliefs about which highways to take, which exits to get off, and so forth. You ask me how to drive from Chicago to Washington DC. The following scenario seems possible: I recognize your question, and I recognize that I dispositionally know the answer. However, I then do something unexpected: I immediately inject myself with a mind-slowing drug that makes it impossible for me to answer your question. Or less dramatically: at the moment you ask, I suddenly force myself to think of something entirely different. Now, it is still true that I have dispositional beliefs about how to drive to Washington DC, but I merely prevent those beliefs from becoming occurrent. I would not be able to give you the directions. To be sure, the beliefs would have become occurrent under ordinary circumstances. I suggest that these ordinary circumstances are what the ‘inevitably’ locution captures.Footnote ²⁰

In the case of Christ, we do not need to imagine anything as drastic as mind-slowing drugs. Rather, the omnipotent Christ could simply will that his dispositional beliefs, when they are considered, should fail to yield their corresponding occurrent beliefs unless he should will otherwise. In other words, his dispositional beliefs will not be consideration-sensitive but rather consideration-and-allowance-sensitive. A dispositional belief p is consideration-and-allowance-sensitive just in case one would inevitably form an occurrent belief with respect to p upon (i) considering p and (ii) willing that p become occurrent.

In sum, the semi-kenotic model affirms that the incarnate Christ was omniscient: he had an exhaustive combination of occurrent and dispositional beliefs. Moreover, during his earthly life, some of Christ's dispositional beliefs were such that he did not will that they become occurrent. An omnipotent being would have no trouble accomplishing this feat.Footnote ²¹

Semi-kenoticism and Anselmian divinity

I argued earlier that kenoticism falls prey to the Anselmian objection. The objection stated that if Christ was divine, he was the greatest possible being. But if he lacked omniscience, then he was not the greatest possible being; therefore, kenoticism entails that Christ was not divine. Since I am contending that semi-kenoticism is a more plausible view, I must now consider whether semi-kenoticism can circumvent the Anselmian objection.Footnote ²² Proponents of the Anselmian objection might offer a parallel argument:

Since Anselmians agree with (10), the key premise which the semi-kenoticist needs to reject is (11). In fact, there are good reasons to think that (11) either lacks support or is false. Notice first that (11) seems to depend on the more general principle:

Initially, (G) might seem plausible in the context of finite creatures like ourselves. I cannot drive to Washington DC if my knowledge remains in a dispositional state: I am able to get there only if I actually recall the directions. But the same reasoning will not apply in the context of an omnipotent being. An omnipotent being's abilities will not be limited or impeded by dispositional knowledge: if he wishes to get to Washington DC, he will do so. Part of the underlying intuition behind (G), I think, stems from the fact that we often make mistakes because our dispositional knowledge fails to become occurrent when we need it to. I forget to pick up groceries on the way back from work because my wife's request remains in a dispositional state throughout the day. But again, this is not a problem for an omnipotent being. If some piece of Christ's knowledge remains dispositional, it is only because he has not willed that it become occurrent. So, our initial intuitions in support of (G) need to be revised when considering an omnipotent being. It seems that (G) is inconclusive as it stands, and we are therefore left without a reason to affirm (11).Footnote ²³

Even if (G) is correct, however, there is a good reason to think that (11) is false. The problem is that being occurrently omniscient might in fact be inconsistent with having human properties. To be occurrently omniscient is to have conscious knowledge of all true propositions simultaneously. But it is plausible that being an embodied human being requires having physical faculties that prevent the possibility of simultaneously entertaining an infinite number of occurrent beliefs. This should be evident from the mere fact that Christ could not have had occurrent knowledge of an infinite number of beliefs while he slept. When a person sleeps (dreams set aside), she has no occurrent beliefs at all: no proposition is at the forefront of her consciousness. If this is the case, however, (11) is false since it is impossible for Christ by virtue of being human to be occurrently omniscient. Human beings generally only entertain one or two occurrent beliefs at a time, and on the semi-kenotic model Christ's mental life would have followed this normal pattern.

It might be objected that possessing human properties is also inconsistent with being dispositionally omniscient and that therefore the incarnate Christ could not have been dispositionally omniscient either. But this objection is unclear, for how could one demonstrate that it is impossible for a human being to have an infinite number of dispositional beliefs? One initial strategy might be to assume that beliefs in general are physical entities – patterns of neural activity, perhaps – and that dispositional beliefs are ‘stored’ in parts of the brain. Just as a computer's hard drive cannot store an infinite amount of information, so Christ's brain would not be able to store an infinite number of dispositional beliefs.

In response, here is where the computer analogy that I employed earlier breaks down. Unless one subscribes to a reductionistic conception of beliefs, there is no reason to think that beliefs are ‘stored’ in the same way that pictures are stored on a hard drive. A more plausible way of thinking about beliefs in general is as a certain type of relation. Beliefs involve a subject's standing in a ‘belief-relation’ to a certain proposition p, where p is the content of the belief.Footnote ²⁴ As I look outside my window, I see a red car parked on the side of my street. I now stand in a new belief-relation to the proposition there is a red car parked outside. Moreover, such belief-relations can be categorized as either occurrent or dispositional. When my teacher first taught me astronomy in grade school, I came to stand in the occurrent-belief-relation to the proposition Jupiter is the largest planet in our solar system. But now (assuming I am not thinking about astronomy) I stand in the dispositional-belief-relation to the proposition Jupiter is the largest planet in the solar system. Upon considering the proposition again, I have the ability to stand once more in my original occurrent-belief-relation. If this view is correct, then it reveals an important disanalogy between encoded information on a hard drive and a dispositional belief. Given that dispositional beliefs are not like stored pictures on a computer, it remains unclear why it would be logically impossible for Christ to stand in the dispositional-belief-relation to all true propositions.Footnote ²⁵

I suspect that the underlying difficulty is that human beings normally acquire dispositional knowledge by first acquiring occurrent knowledge. In other words, normal human beings need to learn about the world before having dispositional beliefs about the world. And, of course, there is not enough time and resources for an ordinary human being to learn all that there is to know. But this marks a fundamental difference between Christ and normal human beings. On the semi-kenotic model, Christ was dispositionally omniscient not because he learned all there was to know but simply because he was divine. He would have possessed the ability to access any information whenever he so willed. Just as I have access to my dispositionally believed proposition Jupiter is the largest planet in the solar system, Christ would have had similar access to his dispositionally believed propositions. The semi-kenotic model, however, involves the affirmation that Christ freely refrained from accessing some of his infinite knowledge.

In sum, there does seem to be something problematic about a human being's having occurrent knowledge of all true propositions. By contrast, there does not seem to be any problem with having dispositional knowledge of all true propositions. What all this shows is that (11) is either inconclusive or false given the semi-kenotic model, and since the Anselmian objection needs (11) to succeed, I conclude that semi-kenoticism is safe from it.Footnote ²⁶