AEJPR AGATHEOS - European Journal for Philosophy of Religion 2004-9331 Nordic Society for Philosophy of Religion in cooperation with the University of Gävle and Uppsala University aejpr.v2i1.51853 10.69574/aejpr.v2i1.51853 Research article Resisting Desire: Models of Christ’s Temptations HillJonathan Theology and Religion Department, University of Exeter, UK Correspondence email address: J.C.Hill@exeter.ac.uk 19062025 2025 2 1 4 25 2025 © 2025 Jonathan Hill This is an Open Access article distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), permitting all use, distribution, adaptation and reproduction in any medium or format, provided the original work is properly cited. Abstract

I propose a categorisation of models of Christ’s temptations. The main division is between models according to which he felt a desire to perform sub-optimal acts (“internalism”), and models according to which he did not (“externalism”). Most analytic theologians today are internalists. I examine different varieties of internalism and argue that they all have serious problems, which do not apply to externalism. I conclude that externalism emerges as the strongest model.

 Keywords incarnation temptation desire impeccability Introduction

The New Testament depicts Jesus as being “tested” or “tempted,”1 sometimes explicitly, sometimes implicitly, to perform sinful (or at least sub-optimal) actions.2 Much of the theological and philosophical discussion of these incidents has revolved around the question whether Jesus was capable of sin, and if not, how these could have been real temptations for him.3 My focus in this paper is rather different. Instead, I want to ask: what should Christians think these temptations were like for Jesus? In particular, did he want, in any meaningful way, to succumb to them?

I propose a category scheme of different possible answers to this question.4 I shall argue that the view that Jesus did not experience any desire to perform sub-optimal actions, although unpopular today, faces the least difficulties.

 What is temptation?

Let us start with a brief sketch of what temptation is. Most broadly, we can say that temptation occurs when a person considers, as a live option, the prospect of performing an action that they believe to be not the best one they could take. More narrowly, temptation involves entertaining an action which is not merely not the best, but positively wrong.

There could be different reasons for this, which we can broadly divide into internal and external factors. The most obvious internal factors are desires. If I am considering whether to eat a bar of chocolate or not, I find myself particularly focusing on the option of eating it because I feel a strong desire to do so. And the most obvious external factors are suggestions from other people. If somebody is urging me to cheat on my tax return, I find myself thinking about doing so, even if it would never have occurred to me to do so otherwise. In both cases I could be said to be undergoing temptation, but for very different reasons. And in some cases both factors may be present. Somebody may be tempting me to do something which I also feel a desire to do – perhaps a desire I already had, or one that is aroused by the suggestion.

To reflect these two kinds of cases, let us distinguish between two senses of “temptation.” First, we can define external temptation, or temptatione:

Temptatione: A person p is temptede to perform action t iff: (1) factors external to p are causing p to focus in particular on t rather than on other possible actions; and (2) p believes that t is not the best course of action available to them (or, more narrowly, that it is a wrongful course of action).

As this indicates, the person being tempted (the “temptee”) must believe that the proposed action is wrong (or, more broadly, contrary to their long-term goals). If I try to persuade somebody to do something that I know is wrong, but she does not, this is not temptation. Must the temptee also believe herself capable of performing the act in question? A.T. Nuyen (1997, pp. 92–93) suggests so. Certainly, anyone trying to tempt a person to do something she believes herself incapable of doing would have to try to persuade the temptee that she can do it. But I do not see why such an attempt could not form part of the temptation (as opposed to a necessary precursor to it). A good example of this is James Hogg’s Private Memoirs and Confessions of a Justified Sinner, in which a sinister stranger (probably the devil in disguise) persuades the Calvinist Robert Wringham to commit a series of murders. Wringham initially believes himself incapable of such acts since he is one of the elect and therefore predestined not to commit them, but the stranger convinces him that the doctrine of the perseverance of the saints means that he can commit such acts safely precisely because he is elect. This certainly seems like temptation, and indeed in the narrative the stranger makes this argument alongside his case for the desirability of committing the murders, not as a preamble to it.

Temptatione is, historically speaking, the older sense of “temptation.” The English word derives from the Latin tento, the root meaning of which is to try. The person who “tempts” is trying, or testing, an object (or person).5 But today, the word is increasingly used to refer not to an action by one person upon another but upon a kind of experience that a person may undergo. John Bigelow, Susan Dodds, and Robert Pargetter, for example, write:

In acting, or failing to act [in certain circumstances], you will be displaying strength or weakness of will; but this is because you are caught by a particular kind of conflict among your desires. We will call this kind of conflict temptation. When facing temptation, a person has a choice of at least two courses of action: one of which requires strength of will, and one of which is weak-willed. (1990, p. 40, emphasis original)

Let us call this internal temptation, or temptationi:

Temptationi: A person p is temptedi to perform action t iff: (1) p feels a desire to perform t, (2) p believes that a is a possible course of action for p; (3) p believes that performing t would conflict with their long-term goals; and (4) this results in a motivational conflict within p.6

It should be clear that this kind of analysis of temptation is fundamentally different from temptatione. Here there is no mention at all of an agent tempting the subject. Instead, the emphasis is on a conflict among the subject’s own motives. I use the word “motive” here because “desire” is ambiguous. Some authors (particularly philosophers) use the term quite broadly to mean any internal state or disposition that motives a person to behave in a certain way.7 However, others (particularly psychologists) use it more narrowly. Wilhelm Hofman and Lotte Van Dillen give a typical definition of “desire” as:

… an affectively charged motivation towards a certain object, person, or activity that is associated with pleasure or relief from displeasure… Desire is the feeling of wanting to have or do something, and thus motivates behavior. In contrast to general motives, desires are “about” certain objects or people. (2012, pp. 317–18, citing Kavanagh, Andrade, and May (2005))

There are two key elements here. First, desires are affective states – that is, they are particular feelings that occur at particular times. They are not dispositions. Second, desires have particular objects. They are not general attitudes. So, for example, my general wish to be a good person is not a desire on this definition. We can call that, instead, a long-term goal. But my specific feeling of wanting to do this good act right now (because it will make me feel better or avoid feeling guilty) is a desire.

Temptation, on this kind of analysis, occurs when desires and long-term goals pull in opposite directions. Wilhelm Hofmann, Malte Friese, and Fritz Strack, for example, write that “most instances of temptation can be described as a tug-of-war or conflict between impulses on one hand and self-control on the other” (2009: 162).8

Temptationi, unlike temptatione, requires the person to believe themselves capable of the action in question. I can certainly be said to feel a desire to perform an action I cannot do (I wish I could fly, for example), but if I do not think I can do it, I will not feel any impulse to attempt it.9 But the presence of such an impulse is essential to temptationi, because the whole point of temptationi is that the person is torn between two possible courses of action.10

 Externalism and internalism

Our two kinds of temptation yield two broad categories of models of Christ’s temptations, which I shall call externalism and internalism.11 They are easily defined:

Externalism: Jesus never experienced temptationi.

Internalism: On at least some occasions, Jesus experienced temptationi.

To put it more succinctly: when undergoing temptation, did Jesus experience any desires to perform the action in question? The internalist says that, on at least some occasions, he did. The externalist says that he never did.

Externalism has a long history. One of its clearest proponents is Augustine of Hippo, who writes:

He was indeed tempted, but he was not endangered, that he might teach you, when you are endangered by temptation, to answer the tempter, and not to go following after him, but to go away from the danger of temptation. However, as he said here, “My soul is troubled now,” so also where he says “My soul is sad, even to death” and “Father, if it is possible, let this cup pass me by,” he has taken up men’s weakness that he may teach him, saddened and troubled in the same way, to say what follows: “Nevertheless not what I will, but what you will, Father.” (Tractates on John 52.3.2 in Rettig, 1993, p. 282)

For Augustine, although Jesus might have acted and spoken as if he were really torn between doing God’s will and doing his own will, this was a sort of performance done for other people’s benefit. Jesus spoke as if he were reluctant to do the right thing, so that he could show us (who really are reluctant) how we should proceed. But he felt no such reluctance himself. If it was temptation at all, it was temptatione only, never temptationi. As Augustine writes elsewhere, “Let whoever believes that the flesh of Christ had desires opposed to the spirit be anathema” (Unfinished Work in Answer to Julian IV.47, in Teske 1999, p. 426).12

However, in the way I am defining externalism, one can be an externalist without going quite that far. Externalism as I define it denies only that Jesus felt any desire to perform a sub-optimal act. It is not the denial that he felt any desires, not even the denial that he ever felt the desire for a sub-optimal state of affairs to obtain. For example, the externalist can hold that, when tempted to turn stones into bread, Jesus could have felt hunger and the desire not to be hungry. (Indeed, to deny this would be tantamount to docetism.) What Jesus did not feel, according to externalism, was a desire to act to relieve his hunger.

This means that the externalist can do more justice to Gospel narratives such as Gethsemane than Augustine’s interpretation suggests. The accounts in Matthew 26 and Luke 22 explicitly give a picture of Jesus feeling significant distress at the prospect of his imminent arrest. But it is striking that Jesus is portrayed as praying only that the Father might do something to save him. Nowhere is there a suggestion that Jesus felt any desire to take any action himself which might save him. A desire not to die is not the same thing as a desire to take action to avoid dying.13

So one could easily interpret this scene as indeed showing Jesus in the grip of contrary desires – the desire not to die, and the desire that his Father’s will be done – without having to see it as a case of temptationi. On this view, his anguish was not a matter of his agonising over whether he should do his Father’s will or not. Rather, it was caused by the knowledge that, in doing his Father’s will, he was going to suffer. He did not want to suffer, and he hoped that his Father’s will might change, but there is no suggestion that he entertained the possibility of disobeying it if it did not.

Another passage that may seem to contradict externalism is Hebrews 4:15, according to which Christ “in every respect has been tested as we are, yet without sin (χωρὶς ἁμαρτίας).” If the only difference between Jesus’s temptations and our own is that he did not sin, it seems that at least some of his temptations should, like ours, have involved sub-optimal desires.

However, this assumes a narrow conception of “sin” as referring only to action, or possibly intention, as opposed to desires.14 But ἁμαρτία has a very wide range of meanings that go well beyond this. In Epictetus, for example, it can mean not only actions that are in some way wrong or mistaken but the state of living in ignorance or under the influence of passion (Roberts, 2014). If one interprets ἁμαρτία this broadly, then Hebrews 4:15 would support externalism, because sub-optimal desires would count as a form of ἁμαρτία.15

I think, then, that the New Testament is neutral between internalism and externalism. Nowhere does it state that Jesus did (or did not) feel any desire to act in a sub-optimal way. However, most analytic theologians today who have written on the subject of Christ’s temptations do seem to hold internalism. Rolfe King gives a representative statement:

[T]he capacity to be tempted implies (as it seems to) an actual potential desire to sin… It is helpful to reflect how remarkable it is that Christ actually had the capacity to be tempted, for this must mean he experienced at least an initial desire, or attraction at some level of his being, to sin. This is what the pull to sin entails. (King 2015, p. 74)16

We can immediately see a key commitment of internalism. As we have seen, temptationi requires that the person undergoing temptation believes themselves capable of performing the act under consideration. If Jesus underwent temptationi, then, he must either have been peccable or have not believed himself to be impeccable.17 Few analytic theologians want to say that Jesus was peccable,18 so that leaves the possibility that he was impeccable but did not know this. It follows, then, that anyone who endorses internalism (and rejects peccability) must hold a kenotic, or quasi-kenotic,19 view of the incarnation according to which Jesus – at least in his human nature – lacked certain knowledge.20

The idea that Jesus’s human knowledge was limited is very common among theologians today, including most analytic theologians. However, internalism does not merely entail a claim that Jesus shared common beliefs of his time and place, such as a belief that the sun revolves around the earth. It entails the claim that Jesus lacked key knowledge about his own moral and spiritual status. If Jesus did not know that he was impeccable, then either he did not know that he was God, or he did not know that God is impeccable (Pawl, 2019, pp. 140–41). But then his own teaching about himself, or about God’s attitude to sin, was not reliable. One verse sometimes used to support the notion of Jesus’s human ignorance is Mark 13:32, where Jesus says that the Father, but not the Son, knows the time of the eschaton. But this is not really a parallel case. Jesus states that nobody knows this information, presumably either because it is inherently unknowable or because the Father chooses not to share it. But Jesus’s divine identity, and the incompatibility of sin with the divine nature, are key elements of Christian doctrine, central to Christian preaching. They are not merely knowable, but actually known. Moreover, both are rooted in Jesus’s own mission and preaching. If, as Christians typically want to say, Jesus’s life and teaching is the most perfect manifestation of God’s nature available to us, and it demonstrates to all people God’s perfect love and goodness, it would be extraordinary to suppose that the one person least able to recognise this was Jesus himself.

I think most Christians would resist the suggestion that Jesus was mistaken about his own identity or the divine nature, since his reliability about these matters must be foundational to orthodox Christian belief.21 A Christian theology which makes claims about Jesus that are not held to derive from Jesus’s own teaching at all drives too great a wedge between Jesus himself and belief about him. Why, for example, suppose that any of these beliefs are true, if Jesus – the primary object of belief – did not teach them? In particular, I would raise a soteriological worry. In Mark 10:45, Jesus interprets his future death as “a ransom for many.” Christian theologians have usually held that Christ’s full divinity and humanity are both necessary for the atonement to work (either because they are constitutive elements of atonement, or because they are necessary preconditions). It follows that if Jesus did not know he was God, but did know that his mission was to save humanity, he did not know how salvation works. Perhaps he did not know, after all, that his mission was to save humanity. But not only would this require discounting the passages in the Gospels where he knows this (and indeed those where other people know it, such as John the Baptist (John 1:29)), it implies a reading of Christ’s passion where the Father instructs him to undergo suffering and death without even knowing why, and that seems very problematic to me.

 Bounded and open internalism

I suggest that internalism can be divided into a number of different varieties. The most important distinction is between what I shall call bounded internalism and open internalism.

Bounded internalism: Although Jesus experienced sub-optimal desires during temptationi, his psychological state was such that it prevented him from acting upon those desires.

Open internalism: On at least some occasions, Jesus’s psychological state would have permitted him to act upon his sub-optimal desires.

So according to bounded internalism, Jesus’s psychological makeup set strict boundaries on what he was able to do. This entails that Jesus was impeccable.22 The open internalist, however, holds that nothing about Jesus’s human nature prevented him from acting on his sub-optimal desires. This could mean that nothing at all prevented him from doing so. He never did anything wrong, but he could have. Let us call this unsafe open internalism. Clearly, this entails the view that Jesus was peccable.

An alternative version of open internalism which has proved popular in recent years is that Christ’s divine nature acted as a sort of safeguard that was never invoked. Thomas Morris set out probably the most influential version of this theory.23 For Morris, Christ’s divine nature stands ready to block him from ever succumbing to temptation. As a result, he is impeccable. However, it is never called upon to do so. Jesus always resists temptation purely through his human will. Jesus’s situation is therefore like a Frankfurt case: he cannot sin, but the mechanism that ensures that he cannot sin never activates.

Let us call this safe open internalism. Importantly, if Christ’s divine nature never plays a causal role in his resisting of temptation, then the causal story that explains his resisting of temptation is the same as it would have been if there had been no safeguard at all. So although safe open internalism can explain why Jesus was in no danger of succumbing to temptation, it cannot explain how he in fact did not succumb.

This, in my view, is a fundamental problem with open internalism (of either variety). The open internalist holds that, when tempted, Jesus felt precisely the same kinds of desires, and precisely the same psychological resources to resist them, that we all do. But we all, sometimes, succumb to temptation, and Jesus never did. Richard Swinburne, for example, writes:

That [Jesus] performed supererogatory good actions, while we ordinary human beings did wrong, although both he and we were equally well positioned to do good or ill is… crucial for theories of the Atonement. (Swinburne 1994, p. 207, emphasis added)

But if Jesus was no better positioned to do good than we are, why was the outcome in his case so different from ours? Surely it wasn’t just a happy chance! One might say, of course, that Jesus was perfectly virtuous, and we are not, but that does not explain the difference – it just redescribes it.24 One might appeal to the role of his divine nature, or perhaps of the Holy Spirit,25 but the open internalist thinks that his divine nature or the Holy Spirit never did act to influence his will (though the safe open internalist thinks that it was ready to do so). If this had happened, then his psychological state would have changed, and this is no longer open internalism. For example, if the divine nature were to direct Jesus’s human will to resist temptation, or make him desire only to do good, Jesus’s psychological state would no longer permit him to choose the wrong action – and that is bounded internalism. So open internalism either fails to explain Jesus’s perfect goodness or collapses into bounded internalism.

 Varieties of bounded internalism

I think, then, that despite its popularity, safe open internalism is not viable. Bounded internalism, by contrast, does offer some explanation of Jesus’s success in resisting temptation. We can distinguish between several different varieties, based on what kind of explanation they offer.

Desire-secured bounded internalism

The first possible explanation, which I shall call desire-secured (DS) bounded internalism, is that Jesus’s desires were unlike those of other people, and this is why, although he felt sub-optimal desires, there was never any chance that he would act upon them.

Just to complicate matters even further, we can distinguish between two kinds of DS bounded internalism. The simplest variety appeals to the nature of the sub-optimal desires themselves. Let us define a property that desires may have, which I shall call inherent motivational viability:

Inherent motivational viability: a desire d to perform an action a is inherently motivationally viable to the degree that, in the absence of reasons to do otherwise, the person experiencing d is likely to perform a.

Clearly, inherent motivational viability comes in degrees. A desire has maximal inherent motivational viability if the person experiencing it is certain to act upon it, and it has no inherent motivational viability if the person experiencing it is certain not to act upon it. Most desires – perhaps all of them, for non-divine persons – fall somewhere in between these two extremes.

One could hold that Jesus’s sub-optimal desires had no inherent motivational viability for him. He may have felt a sort of fleeting desire to turn stones into bread, for example, but there was never any chance of his actually doing it, because the desire was simply not strong enough to motivate him. Let us call this minimal desire (MD) bounded internalism.

Some ancient Christian theologians developed a theory similar to this on the basis of the Stoic concept of propatheiai. For the Stoics, emotions are fundamentally rational in nature, but experience shows that we do have emotional responses which are purely instinctive and involuntary, in the form of reactions to certain kinds of stimulus. These are not true passions, because they are felt wholly involuntarily, but if we assent to them, they bring about true passion. These are propatheiai or pro-passions.26 Origen appealed to this idea to explain those passages of the New Testament that attribute passion to Jesus. For Origen, Jesus could not have experienced true passion, as this would be a mark of imperfection. So, drawing attention to the fact that we are told that Jesus “began” to be distressed in Gethsemane (Matthew 26:37), he argued that this meant that Jesus was experiencing the propatheia of distress rather than distress itself (Commentary on Matthew 92). Didymus the Blind developed the idea further, arguing that Christ must have experienced propatheiai, for that is part of the natural condition of human souls. So where Origen’s emphasis is on the fact that Christ felt propatheiai as opposed to full passion, Didymus’ emphasis is on the fact that Christ felt propatheiai as opposed to nothing at all (Layton, 2000, pp. 276–81).

Could at least some propatheiai be desires? One could plausibly think so. For example, suppose I smell freshly-baked bread. I feel an immediate, instinctive desire for the bread. If I allow my attention to remain on the smell, I feel a full-blown desire for the bread. (That, at least, might be how a Stoic would analyse the situation.) Neither Origen nor Didymus explains precisely why Jesus felt only propatheiai, or how this relates to his sinlessness. If they had held (1) that his propatheiai were a kind of desire with no inherent motivational viability, (2) that he was incapable of feeling true passions (owing to his divinity), and (3) that these two reasons were why he never gave in to tempting desires, then they would have been proponents of MD bounded internalism.

MD bounded internalism offers a straightforward account of why Jesus never succumbed to temptation, despite experiencing desires to do so. But it does not seem to me to be an adequate account. This is because it requires the possibility of desires with no inherent motivational viability at all. If Jesus’s sub-optimal desires had any inherent motivational viability – even very low – then there was no guarantee that he would not act upon them. But such a guarantee is necessary if this is to be a version of bounded internalism. And the possibility of a desire that is so weak it has no inherent motivational viability whatsoever is not coherent. A desire, as we defined the term earlier, is a feeling of wanting, and is therefore inherently motivating. If I feel a desire then there is a certain state of affairs that I want to bring about, because I think it will bring me pleasure or relieve me from pain, and so I have a motive to act in such a way to bring about that state of affairs.27 Moreover, if I have a motive to perform an action, and that action is within my power, there is at least some probability that I will perform it.28 This probability may be very low. It is reasonable to think that the more out of character the action is, the lower the probability that I will perform it. But as long as I have a motive at all, that probability is not zero. People do act out of character sometimes, for better or for worse.

So to say that Jesus felt some desire to perform sub-optimal acts, and yet that this desire was so slight that there was zero probability of his actually performing them, is incoherent.29 One might say, perhaps, that the reason why sub-optimal desires held no inherent motivational viability for Jesus was that, being supremely virtuous, sub-optimal actions simply had no appeal to him. But then that is a denial that he felt any desire to perform them at all, and the position collapses into externalism.

An alternative version of DS bounded internalism is to suppose that Jesus’s sub-optimal desires were reasonably strong – as strong as our own, perhaps. But they had no power to motivate him to action because they were overridden by much stronger motives to perform only the best actions available to him.

To articulate this, let us define a new property that desires may have:

Actual motivational viability: a desire d to perform an action a is actually motivationally viable to the degree that, in the situation when d actually occurs, the person experiencing d is likely to perform a.

Actual motivational viability differs from inherent motivational viability in that it takes into account other motives to action that may also be in play. Suppose I see an unattended wallet and the thought crosses my mind that I might steal it. But I probably won’t actually do that. Why not? It is not (merely) because my desire to steal it has low inherent motivational viability. It is because I also have a long-term goal to be honest, which has higher inherent motivational viability, as well as a desire not to get into trouble, a long-standing habit of not stealing things, and so on. The inherent motivational viability of all of these, considered together, greatly outweighs the inherent motivational viability of my desire to enrich myself. So whatever the inherent motivational viability of that desire – which may, perhaps, be quite high – its actual motivational viability is very low.

Now we can describe what I will call outweighing motive (OM) bounded internalism. On this view, Jesus’s sub-optimal desires had non-zero inherent motivational viability (perhaps some of them had quite high inherent motivational viability), but their actual motivational viability was zero. This is because Jesus’s long-term goal to perform only the best possible action was so supremely important to him that any desire to the contrary had, in practice, no motivational power over him at all. We could, perhaps, say that this was because the strength of his long-term goal meant that on any given occasion he felt a very strong desire to do the right thing, in addition to any desire he might also have felt to do the wrong thing. Consequently, when Jesus experienced temptation, it was not simply a matter of a sub-optimal desire conflicting with a long-term goal. Rather, he experienced a sub-optimal desire conflicting with a (much stronger) optimal desire that was in line with his long-term goal.

I find OM bounded internalism more intuitively attractive than MD bounded internalism. It allows us to attribute normal human sub-optimal desires to Jesus in a way that MD bounded internalism does not. However, I think that OM bounded internalism is just as incoherent as MD bounded internalism, and for much the same reason. I can certainly imagine being in a state where two strong desires battle against each other, with the outcome uncertain. But I cannot imagine a state in which one desire is overpowered by another to such a degree that it has no chance of motivating action, and yet is still a felt desire. This is, once again, because any desire that is felt provides a motive for action.

If Jesus’s desire to act rightly in accordance with his Father’s will was so supremely strong that it was impossible for him not to act upon it, this can only be because that desire simply displaced any desire to do otherwise. He experienced what Adam Pelser calls “motivational silencing” of his sub-optimal desires: he might have perceived, intellectually, that there was a case to be made for turning stones to bread, but any appeal that doing so might have had for him was drowned out by the overwhelming appeal of obeying the Father. And, as Pelser goes on to observe, “motivational silencing is tantamount to the virtuous person’s inability to experience the inner psychological experience of temptation” (2019, p. 85). If this is so, then OM bounded internalism really collapses into MD bounded internalism, which itself collapses into externalism.

 Control-secured bounded internalism

Both MD bounded internalism and OM bounded internalism seek to keep Jesus immune to sub-optimal desires by appealing to unusual features of his desires. But a different approach to bounded internalism is possible, according to which Jesus’s desires were (perhaps) quite ordinary. What was remarkable about him was, rather, his ability to determine which of his desires he acted upon. On this view, he had perfect control over his motives and his actions. We can call this control-secured (CS) bounded internalism.

One simple way to express this is to appeal to the fact that Jesus was God. More precisely, the person of Christ was identical with the divine Son, the second Person of the Trinity. As such, all decisions that he made were made by the divine Son, who cannot be swayed by sub-optimal desires because of his perfect divine freedom and his impeccability. John McKinley characterises views of this kind as his M1, the “sinless by inherent impeccability” model of Christ’s sinlessness, which he attributes to a number of early Christian theologians, notably Tertullian, Origen, and Augustine (2009, pp. 97–102).30 For McKinley, this model is inadequate because it portrays Jesus’s experience during temptation as different in kind from everyone else’s (2009, pp. 100– 02, 299–301).31 In my view, though, it is inadequate for a more fundamental reason, which is that on this view, Christ’s humanity plays no role in his human decision-making. Orthodoxy states that Christ had a full human will as well as a divine one, as laid down at the third Council of Constantinople in 681. Precisely what this means is a matter of dispute, but what it must at least mean is that Christ had a human way of making decisions. That is, the mental and physiological machinery by which he acted was the same as ours, with the same fundamental way of operating, no matter how different the end results may have been. To deny this is to deny his true humanity. So any notion that the divine Son simply exercised his divine will and his human nature carried it out is untenable. We need instead to suppose that Christ exercised his self-control through the operation of his human mental machinery.

What, then, is the mental machinery by which we exercise self-control? Researchers typically characterise it as something that requires effort.32 When faced with conflicting desires, we experience difficulty in choosing the better course of action, and sometimes we are not equal to the challenge. But why is this? What is actually happening when we make this effort, and why do we so often fail?

Much psychological research into self-control has focused on the phenomenon of “ego depletion,” a sort of volitional exhaustion in which a person becomes less able to exercise self-control.33 There is substantial experimental evidence that people who have used self-control in one task are worse at using it in another task shortly afterwards, compared to people who come to the second task fresh. Strikingly, this is so even if the two tasks are quite different in nature. One of the most well-known experiments, for example, showed that people are more likely to give up trying to solve difficult puzzles if they have previously had to resist the temptation to eat chocolate (Baumeister et al., 1998).

Michael Inzlicht and Brandon Schmeichel have proposed what they call the “process model” of self-control to explain this phenomenon. On this view, there are two elements to the process. First, the use of self-control makes a person less motivated to continue exercising self-control, precisely because it is difficult. Recalling that self-control occurs when a person has a tension between an impulsive and situational desire and their long-term values, we can say that exercising self-control in one situation causes the impulsive desire to become stronger and the long-term one weaker in subsequent situations, if they occur soon after (Inzlicht and Schmeichel, 2012, p. 455). Second, and relatedly, the use of self-control makes a person less aware of the need for self-control in subsequent situations. There is evidence that, in such situations, the person is less likely to notice that the impulsive desires conflict with her long-term goals, making her less likely even to try to resist them (2012, pp. 455–57).

Suppose we accept this view as a general account of what happens in failures of self-control. If I give in to temptation, it is not simply, or at least not always, because my desire to perform the illicit act is too great. It is also because of a cognitive failure: I may not recognise (or refuse to acknowledge) that it is illicit.34 Such an account is similar to Aristotle’s analysis of akrasia, at least on the traditional reading of that account: I may know that eating the chocolate is against my own interests, but when faced with it, my desire for it temporarily clouds this knowledge.35 The process model refines the account by suggesting a more complex relationship between the strong desire and the cognitive failure: rather than the desire straightforwardly causing the failure of recognition, it could also be that the failure of recognition is what allows the desire free reign. It is because I forget how bad for me the chocolate is that I feel the desire for it so strongly – as well as my desire causing me to forget.

Now suppose that the following things were true of Jesus:

He had a core value of always doing the best thing in any given situation.

He always knew what the best thing to do in any given situation was.

Consequently, whenever he felt a sub-optimal desire, he was always aware of how it went against his long-term goals.

In any given situation, his sub-optimal desires were always weaker than his optimal desires.

The last of these looks like a restatement of OD bounded internalism. But OD bounded internalism seeks to secure Jesus’s moral safety only by appealing to his stronger desires to do good, and as we have seen, that is not enough. Here, we take the idea that his desires to do good outweighed his desires not to do so, and we combine it with the idea that he also had perfect understanding of what he ought to do. Even if he had to exercise self-control repeatedly, this did not cause in him the cognitive failures that it causes in other people. If this were so, then no matter how unpleasant or tiring it may have been for him, he would always successfully exercise self-control in every situation.

We could add, as part of this idea, that Jesus maintained clarity of purpose in what Kristian Myrseth and Ayelet Fishbach have called “epsilon-cost temptation” (2009, pp. 247–48). This refers to cases where giving in to the temptation on this occasion (e.g. eating this cake) would have negligible harmful effects, but doing so regularly (e.g. eating cakes every day) would be more harmful. For Myrseth and Fishbach, a key factor in self-control is the ability to recognise such situations as potentially habit-forming and therefore more in conflict with long-term goals than they might superficially appear. If (2) were true of Jesus, he would not be susceptible to this insidious kind of temptation, because he would perceive every situation under the “wide frame” that encompasses future possibilities (2009, p. 248).

There are two ways we could envisage CS bounded internalism working. One would be to say that, through his hypostatic union to the divine Son, Christ’s human mind simply had a perfect moral understanding. Niamh Middleton, for example, writes:

…the human will of Jesus, guided by his divine will, would have prevented propatheia from descending into pathos. The salient point is that this act of the will would not have resulted from obedience to the Father’s will alone, but from the perfect breadth and depth of a moral knowledge that would perhaps be better termed a moral vision illuminated by love. (2021, p. 114)36

We could call this moral vision (MV) bounded internalism. Maximus the Confessor held a view that was, in some respects, along these lines. He held that, in Gethsemane, Christ’s human will “seeks to avoid the passion,” though his divine will embraces it (Louth, 1996, p. 187, quoted in Heidgerken, 2021, p. 103).37 But Christ’s human will, although genuinely human, does not operate like ours. It is only a “thelemic” will, that is, a natural appetitive inclination towards something. We also have a “gnomic” will, that is, the disposition to deliberate about what to do. Christ, according to Maximus, did not have this. It was a part of his human perfection that he never deliberated, but was driven directly by his perfect, rational human desires, since he always understood perfectly what the best thing to do was.38

Alternatively, we could say that Jesus’s perfect moral understanding was itself part of a greater intellectual vision he always enjoyed, namely his constant awareness of the divine nature. This is the traditional view that Jesus permanently experienced the beatific vision.39 We can call this beatific vision (BV) bounded internalism. An advantage of this view is that it allows us to suppose that Jesus’s goals were wholly unified. One reason why self-control sometimes fails is when a person’s attention and energy is split between trying to resist the temptation and some other task (possibly not one involving self-control). Hiroki Kotabe and Wilhelm Hofmann give the example of a person in a work meeting trying to resist the temptation to eat a cookie. If she also has to focus on her contribution to the meeting, her attention is divided, and this makes it harder to resist the temptation (2015, p. 628). But if Jesus’s perfect moral vision was a function of his perfect awareness of the divine nature at all moments, he could never have experienced this kind of cognitive dissolution. He only ever had a single overarching goal, which was to glorify and obey the Father. All his other goals, such as his moral values, were functions of this single goal, on which his attention was wholly focused at all times.

I think that this is a coherent account. But one weakness of CS bounded internalism, which applies to both varieties, is that it assumes a highly intellectualist philosophy of action. If we say that the reason Jesus was unable to do wrong was that he had a perfect understanding of what is right, both general and episodic, then we are committed to the view that the intellect determines action. On this view, which is associated with Socrates but found in later Christian authors such as Bonaventure and Godfrey of Fontaines, it is impossible for a person to do something in the full awareness that it is not the best thing to do. It contrasts with a voluntarist philosophy of action, which emerged more clearly in the later Middle Ages with thinkers such as Duns Scotus and William of Ockham, and which holds that a person always has the ability to act against what the intellect recommends.40

This is not necessarily a problem in itself. Intellectualism could be true. But it does have ramifications that the bounded internalist may not want to accept. In particular, it is hard to reconcile with libertarian free will, since a key claim of libertarianism is that a free act is one that is not determined by any prior factors, including the intellect. Certainly not all Christians believe in libertarian free will, but many not only believe in it but hold it to be a key component of moral theory (and, perhaps, theodicy). For those Christians, CS bounded internalism is not viable. If Jesus had libertarian free will, then even his experiencing the beatific vision permanently would not have guaranteed that he would not misuse that free will.

There is one other problem with this view – indeed with all versions of bounded internalism. As we saw earlier, internalism (of any variety) requires that, if Jesus was unable to sin, he was unaware of this fact. However, if his inability to sin was based on, or manifested in, his psychological boundaries, it is hard to see how he could have been unaware of it in times of temptationi. If I wrongly believe myself to be psychologically capable of a certain unethical act, and I then find myself in a situation where I have a strong motivation to perform that act, but do not do so, it seems to me that the experience must show me that I am in fact not psychologically capable of performing it. My psychological inability would be the reason why I do not do it, and I would be aware of this in deciding not to do it, because I would realise that, really, I have no decision to make at all.

If this is so, then even the best version of bounded internalism – CS bounded internalism, whether that be MV bounded internalism or BV bounded internalism – looks incoherent. Open internalism does not suffer from this problem, because on that view Jesus is either not impeccable at all (unsafe open internalism) or his impeccability is secured without involving his psychological makeup (safe open internalism). But as we have seen, open internalism has serious problems of its own in its inability to explain how Jesus, uniquely, never actually sins.

 Conclusion

Categorising models of Jesus’s temptations on the basis of how they handle his desires, rather than focusing primarily on the question of impeccability, has helped to identify key issues that different models face. I have suggested that all versions of internalism face difficulties of one kind or another. Externalism, however, faces none of them. The externalist can hold that Jesus was impeccable, that he was aware of his own impeccability (and, perhaps, of his divinity), and that he nonetheless did feel strong desires for things that it would have been wrong for him to try to bring about. Externalism does not require any kind of kenotic or quasi-kenotic restriction on Jesus’s human knowledge, but it would be compatible with such a theory if one had other reasons to hold it. My conclusion, then, is that externalism is the strongest model of Jesus’s temptations.41

 References BathrellosD.2004The Byzantine Christ: Person, Nature, and Will in the Christology of Saint Maximus the ConfessorOxfordOxford University Press BaumeisterR.BratslavskyE.MuravenM.TiceD.1998Ego Depletion: Is the Active Self a Limited Resource? Journal of Personality and SocialPsychology745pp.125265https://psycnet.apa.org/doi/10.1037/0022-3514.74.5.1252 BaumeisterR.GailliotM.DeWallC.N.OatenM.2006Self-Regulation and Personality: How Interventions Increase Regulatory Success, and How Depletion Moderates the Effects of Traits on BehaviorJournal of Personality746pp.17331801https://doi.org/10.1111/j.1467-6494.2006.00428.x BlowersP.2012Maximus the Confessor and John of Damascus on Gnomic Will (γνώμη) in Christ: Clarity and AmbiguityUnion Seminary Quarterly Review633&4pp.4450 BrehmJ.SelfE.1989The Intensity of MotivationAnnual Review of Psychology401pp.10931https://doi.org/10.1146/annurev.ps.40.020189.000545 CarterJ. A.KallestrupJ.PalermosS. O.PritchardD.2014Varieties of ExternalismPhilosophical Issues241pp.63109https://doi.org/10.1111/phis.12026 CrispO.2007aDivinity and Humanity: The Incarnation ReconsideredCambridgeCambridge University Presshttps://doi.org/10.1017/CBO9780511805332 CrispO.2007bWas Christ Sinless or Impeccable?Irish Theological Quarterly722pp.16886https://doi.org/10.1177/0021140007082165 DavidsonI.2008Pondering the Sinlessness of Jesus Christ: Moral Christologies and the Witness of ScriptureInternational Journal of Systematic Theology104pp.37298https://doi.org/10.1111/j.1468-2400.2007.00331.x DestréeP.2007Aristotle on the Causes of AkrasiaInBononichC.DestréeP.Eds.Akrasia in Greek Philosophy: From Socrates to PlotinusLeidenBrillpp.13965https://doi.org/10.1163/ej.9789004156708.i-308.39 ErginelM.2016Akrasia and Conflict in the Nicomachean EthicsBritish Journal for the History of Philosophy244pp.57393https://doi.org/10.1080/09608788.2016.1176890 ForsythP.T.1909The Person and Place of Jesus ChristBostonPilgrim FrederickD.2010Unmotivated Intentional ActionPhilosophical Frontiers51pp.2130 GaineS. F.2015Did the Saviour See the Father?London; New YorkClark GalotJ.1986Le Christ Terrestre et la VisionGregorianum673pp.42950 GibsonJ.1995The Temptations of Jesus in Early ChristianitySheffieldSheffield Academic Press GrösslJ.2021Christ’s ImpeccabilityInArcadiJ.TurnerJ.Eds.T & T Clark Handbook of Analytic TheologyEdinburghClarkpp.21530 GrösslJ.von StoschK.Eds.2021Impeccability and Temptation: Understanding Christ’s Divine and Human WillLondonRoutledgehttps://doi.org/10.4324/9780367618933 GulF.PesendorferW.2001Temptation and Self-ControlEconometrica696pp.140335https://doi.org/10.1111/1468-0262.00252 HartT.1995Sinlessness and Moral Responsibility: A Problem in ChristologyScottish Journal of Theology481pp.3754https://doi.org/10.1017/S0036930600037285 HeidgerkenB.2021Salvation Through Temptation: Maximus the Confessor and Thomas Aquinas on Christ’s Victory Over the DevilWashington, DCCatholic University of America Press HoffmannT.2011Intellectualism and VoluntarismInPasnauR.Ed.The Cambridge History of Medieval Philosophy vol. VCambridgeCambridge University Presspp.41427https://doi.org/10.1017/CHOL9780521762168.032 HurstF.2009Forty Gospel Homilies by Gregory the GreatPiscataway, NJGorgias InwoodB.1985Ethics and Human Action in Early StoicismOxfordOxford University Press InzlichtM.SchmeichelB.2012What is Ego Depletion? Toward a Mechanistic Revision of the Resource Model of Self-ControlPerspectives on Psychological Science75pp.45063https://doi.org/10.1177/1745691612454134 JedwabJ.2011The Incarnation and Unity of ConsciousnessInMarmodoroA.HillJ.Eds.The Metaphysics of the IncarnationOxfordOxford University Presspp.16885https://doi.org/10.1093/acprof:oso/9780199583164.003.0009 KavanaghD.AndradeJ.MayJ.2005Imaginary Relish and Exquisite Torture: The Elaborated Intrusion Theory of DesirePsychological Review1122pp.44667https://doi.org/10.1037/0033-295x.112.2.446 KingR.2015Atonement and the Completed Perfection of Human NatureInternational Journal of Philosophy and Theology761pp.6984https://doi.org/10.1080/21692327.2015.1043931 KotabeH.HofmannW.2015On Integrating the Components of Self-ControlPerspectives on Psychological Science105pp.61838https://doi.org/10.1177/1745691615593382 LaytonR.2000Propatheia: Origen and Didymus on the Origin of the PassionsVigiliae Christianae543pp.26282https://doi.org/10.1163/157007200X00161 Le PoidevinR.2023And Was Made Man: Mind, Metaphysics, and IncarnationOxfordOxford University Presshttps://doi.org/10.1093/oso/9780199676576.001.0001 LeftowB.2014Tempting GodFaith and Philosophy311pp.323https://doi.org/10.5840/faithphil201312301 LewisR.ed.1997Middle English Dictionary: TAnn Arbor, MIUniversity of Michigan Presshttps://doi.org/10.3998/mpub.6785 LimJ.2022The Necessity of Beatific Knowledge in Christ’s Humanity: A Re-Reading of Summa Theologiae III, Q. 9The Thomist864pp.51542https://doi.org/10.1353/tho.2022.0034 LokeA.2009On the Coherence of the Incarnation: The Divine Preconscious ModelNeue Zeitschrift für systematische Theologie und Religionsphilosophie511pp.5063https://doi.org/10.1515/NZST.2009.004 LokeIA.2023The Impeccability of Christ, Divine Omnipotence, and Moral PerfectionFaith and Philosophy402pp.24361https://doi.org/doi:10.37977/faithphil.2023.40.2.5 LouthA.1996Maximus the ConfessorLondonRoutledge LyonsE.1994Jesus: Self-Portrait by GodDublinColumba McFarlandI.2007“Willing is Not Choosing”: Some Anthropological Implications of Dyothelite ChristologyInternational Journal of Systematic Theology91pp.323https://doi.org/10.1111/j.1468-2400.2006.00227.x McKinleyJ.2009Tempted For Us: Theological Models and the Practical Relevance of Christ’s Impeccability and TemptationMilton Keynes; Colorado Springs, COPaternoster MiddletonN.2021Jesus and WomenCambridgeLutterworth MorelandJ.P.CraigW.L.2017Philosophical Foundations for a Christian Worldviewsecond ed.Lisle, ILInterVarsity Press MorrisT.1986The Logic of God IncarnateIthaca, NYCornell University Press MunzerS.2023Temptation, Sinlessness, and ImpeccabilityInternational Philosophical Quarterly631pp.91108https://doi.org/10.2139/ssrn.4934118 MurrayM.ReaM.2008Introduction to the Philosophy of ReligionCambridgeCambridge University Press MyrsethK.FishbachA.2009Self-Control: A Function of Knowing When and How to Exercise RestraintCurrent Directions in Psychological Science184pp.24752https://psycnet.apa.org/doi/10.1111/j.1467-8721.2009.01645.x NuyenA.T.1997The Nature of TemptationThe Southern Journal of Philosophy351pp.91103https://doi.org/10.1111/j.2041-6962.1997.tb00827.x OddieG.2005Value, Reality, and DesireOxfordOxford University Press PawlT.2019In Defense of Extended Conciliar Christology: A Philosophical EssayOxfordOxford University Press PawlT.2021Conciliar Christology, Impeccability, and TemptationInGrösslJ.von StoschK.Eds.Impeccability and Temptation: Understanding Christ’s Divine and Human WillLondonRoutledgepp.94115 PelserA.2019Temptation, Virtue, and the Character of ChristFaith and Philosophy361pp.81101https://doi.org/10.5840/faithphil2019117119 PorterJ.2000Responsibility, Passion, and Sin: A Reassessment of Abelard’s EthicsThe Journal of Religious Ethics283pp.36794https://doi.org/10.1111/0384-9694.00054 RettigJ.1993Tractates on the Gospel of John 28–54Washington, DCCatholic University of America Press SwinburneR.1994The Christian GodOxfordClarendon SwinburneR.2011The Coherence of the Chalcedonian Definition of the IncarnationInMarmodoroA.HillJ.Eds.The Metaphysics of the IncarnationOxfordOxford University Presspp.15367https://doi.org/10.1093/acprof:oso/9780199583164.003.0008 TeskeR.1999Answer to the Pelagians III. Unfinished Work in Answer to JulianHyde Park, New YorkNew City Press TranT.2004The Beatific Vision of Christ the Wayfarer (Viator) and His FaithAngelicum811pp.85120 WareB.2010The Man Christ JesusJournal of the Evangelical Theological Society531pp.518 WeinandyT.2004Jesus’ Filial Vision of the FatherPro Ecclesia132pp.189201 WhiteT.J.2015The Incarnate LordWashington, DCCatholic University of America Press WrightR.2008Refining the Prediction of Effort: Brehm’s Distinction Between Potential Motivation and Motivation IntensitySocial and Personality Psychology Compass22pp.682701https://psycnet.apa.org/doi/10.1111/j.1751-9004.2008.00093.x How To Cite This Article HillJonathan.2025“Resisting Desire: Models of Christ’s Temptations”AGATHEOS: European Journal for Philosophy of ReligionVol2No1pp.425

The Greek term πειράζω is ambiguous. Its root meaning is to try, suggesting the translation “test.” Sometimes this “test” takes the form of a moral test upon a person, in which case “tempt” may be more appropriate. On this, see Gibson (1995, pp. 325–26).

We may distinguish between temptations to sin and temptations to sub-optimal behaviour. In the desert, at least some of the things the devil tries to persuade Jesus to do are clearly sinful, such as worshipping him. It is less clear whether this is the case with the agony in Gethsemane, assuming this is a case of temptation (on this, see Munzer 2023, p. 99). Would it really be sinful for Jesus to seek to avoid arrest? It would certainly be sub-optimal, since it would be better for him to die to save humanity than not to do so, but that is not the same thing as being morally wrong. In what follows I shall largely ignore this distinction and speak only of sub-optimal behaviour.

Discussions of Jesus’s impeccability in the analytic tradition include Morris (1986, pp. 137–62), Crisp (2007b), McKinley (2009), Ware (2010, pp. 9–17), Leftow (2014), Pawl (2019, pp. 132–64), Grössl (2021), Loke (2023), and Munzer (2023), as well as the papers in Grössl and von Stosch (2021). Discussions within the broader context of non-analytic theology include Hart (1995), Lyons (1994, pp. 62–65), Davidson (2008), and Gaine (2015, pp. 168–69). For a longer list of references, see McKinley (2009, p. 9 n13).

In his important work on Jesus’s temptation, John McKinley (2009) also examines a number of theological models. However, his categories focus on the question of Jesus’s impeccability, yielding quite a different categorisation scheme from mine, though I shall note points where they overlap.

Chaucer’s Marquis, for example, seeks “to tempte his wyf, hir sadnesse for to knowe” (The Canterbury Tales, The Clerk’s Tale, line 452) – meaning to test his wife’s fidelity. For more examples of Middle English usage of tempten, see Lewis (1997, pp. 196–97). The Wycliffe Bible of 1395 uses tempten in some cases to refer to God testing people, e.g. Deuteronomy 8:2, 33:8; 2 Chronicles 32:31; Hebrews 11:17. We do not use “tempt” this broadly today, though the related word “attempt” does retain a wide range of meanings.

This leaves open the possibility that someone might have long-term goals that are immoral, and who feels a desire to do something good. An example might be the Star Wars character Kylo Ren, who has a long-term commitment to evil which requires him to murder his father, and has to struggle with a desire not to do so. It feels a little odd to call this “temptation,” as Pelser (2019, p. 84) observes, but it would fall under our definition here. For the purposes of our discussion, however, we will ignore such cases and assume that the long-term goals involved in temptation are positive ones.

Graham Oddie defines desires slightly differently from this, but still more broadly than I am doing, as an experience that something is good, such that to desire something is the same thing as for that thing to seem good (2005, pp. 41–42, 47–81).

The word “most” here problematises the definition, as it implies that there might be other kinds of temptation, but the authors do not describe them.

Timothy Pawl (2021, pp. 99–100) argues against this on the grounds that, for temptation to occur, it is enough to have the desire to perform the action in question, even if one is certain that one cannot actually perform it. He gives the example of a person who has been told by God that she will be severely tempted but will not succumb, and who is entirely convinced that this is true: such a person would still experience true temptation. But I do not think this is a true counter-example. Such a person might well believe that she will not succumb to the temptation, but this is not the same as believing that she cannot succumb to the temptation. We can express this modally: the person in this example knows that she will not succumb in the actual world, but not that she will not succumb in every possible world. But if Jesus knew himself to be impeccable, he knew that he would not succumb in any possible world. It seems to me that I can, plausibly, feel a strong impulse to act in a way that I know I act in some possible worlds even if I somehow know that the actual world is not one of them, but not that I can feel such an impulse to act in a way in which I know I do not act in any possible world at all.

Whether temptatione is really temptation at all is controversial. For some authors (e.g. Lyons 1994, pp. 60–61, Crisp 2007b, p. 155, Murray and Rea 2008, pp. 82–83, McKinley 2009, p. 280, Pawl 2019, pp. 146– 47), it is an essential feature of temptation that the temptee feel some desire to perform the act in question. This, I think, reflects the degree to which temptationi has displaced temptatione as the primary meaning of the word today. However, for our purposes, it does not really matter whether we call temptatione “temptation.” This is because the primary reason for applying these terms to Jesus at all is because the New Testament says he was tempted – but, as previously noted, the term πειράζω is even more ambiguous than the English “tempt,” it and covers cases of “testing” much more broadly conceived. It would be consistent with the biblical claims to suppose that Jesus was never tempted (in a modern sense of the word) at all, only tested.

The terms “externalism” and “internalism” are commonly used in analytic philosophy to refer to rival theories in epistemology and philosophy of mind. On this, see Carter et al. (2014). My adoption of these terms is unrelated to this meaning.

For a similar view, see also Gregory the Great, Forty Gospel Homilies 14, in Hurst (2009, p. 102).

Brian Leftow (2014, pp. 11–14) suggests, though does not explicitly endorse, an account of this kind.

A well-known example of such a theory is that of Peter Abelard, who argued that sin consists solely in consenting to a wrong act, not in desiring it. On this, see Porter (2000).

Thomas Aquinas gives this interpretation of this verse: Summa Theologiae III q. 41 a. 3

Similar views are expressed by, inter alia, Hans Urs Von Balthasar (1983, pp. 24–25), Thomas Morris (1986, p. 137), John McKinley (2009), Andrew Loke (2009, p. 62), Richard Swinburne (2011, p. 161), Joseph Jedwab (2011, p. 178), Brian Leftow (2014, pp. 14–19), and Robin Le Poidevin (2023, p. 185).

This latter possibility can be divided into two: he might (1) have believed himself to be peccable, or (2) have been uncertain one way or the other. On this, see Munzer (2021, pp. 92–93). (1) seems more problematic since it would ascribe to Jesus a significant false belief. McKinley (2009, p. 241) rejects (1) but considers (2) to be possible.

Millard J. Erickson (1991, pp. 562–64) is probably the most notable theologian to defend Christ’s peccability, given his otherwise mostly conservative evangelical theology (though Charles Hodge also held this view). For more examples and discussion of this view, see McKinley (2009, pp. 205–17).

I use the terms “kenotic” and “quasi-kenotic” very broadly here, to mean any view according to which the incarnate Jesus lacked access to full divine knowledge. This includes Oliver Crisp’s categories of “ontological kenoticism,” “functional kenoticism,” and “krypsis” (2007a, pp. 122–53), which extends beyond authors who would describe their views as kenotic.

Thomas Morris (1986, pp. 146–49) argued, on these grounds, that Jesus must have lacked knowledge of his impeccability: temptation requires only “epistemic possibility” of succumbing, not ontological possibility. This view goes back at least to P.T. Forsyth (1909, pp. 300–02). Other authors to defend it include Richard Swinburne (1994, p. 204), Gerald O’Collins (1995, p. 271), and J.P. Moreland and William Lane Craig (2017, pp. 610–13). For more on this view, see McKinley (2009, pp. 219–43). McKinley splits it into two models (M8 and M9) depending on whether it is combined with ontological kenoticism, but for our purposes we can gloss over this distinction.

On this, see Pawl (2019, pp.140–43). Pawl argues that an important element of traditional Christology is that Jesus knew himself to be God, in which case he could not have believed himself capable of sinning.

At least, a sort of impeccability. There are different kinds of inability. For example, I am unable deliberately to mow down innocent pedestrians while driving, in the sense that it is a psychological inability for me – I could not bring myself to do it. This is different from the sense in which I am unable to lift a one-tonne weight or even from the sense in which I am unable – at least for now – to converse in Klingon. The doctrine of divine impeccability is usually driven by an intuition that God is metaphysically incapable of sinning – for God to sin would be a contradiction in terms. To say that Jesus was incapable of sinning in the same way that I am incapable of deliberately killing pedestrians clearly falls short of this. I think, then, that most bounded internalists would probably want to say that sinning was humanly impossible for Jesus and metaphysically impossible for him as well. In what follows, though, I will ignore these different kinds of impeccability and assume that bounded internalism simply entails impeccability.

Other versions of this theory have been articulated by Oliver Crisp (2007b, p. 177), John McKinley (2009, pp. 290–93), Bruce Ware (2010, pp. 15–17), and Brian Leftow (2014, pp. 22–23).

Bruce Ware, for example, answers that Jesus “utilized all of the resources given to him in his humanity” (2010, p. 16) such as prayer, Bible study, and receptivity to the Holy Spirit. But an answer of this kind just pushes the question back a stage. Why did Jesus do all of this, and we do not, if Jesus had only the same resources and abilities that we do? How, indeed, do we know that he did, and that he did not have to rely on his divine nature after all?

John McKinley (2019, pp. 299–313) argues for a Christological model according to which the Holy Spirit both veils Jesus’s impeccability from him and empowers him to resist temptation.

The major sources for this theory are Seneca’s On Anger and a fragment of Epictetus quoted by Aulus Gellius in Attic Nights 19.1.14–20, quoted in Inwood (1985, p. 177). Cicero also refers to it in Tusculan Disputations III 83. On this theory, as well as Origen’s and Didymus’ use of it, see Layton (2000).

This does not necessarily mean that all actions have motives. On the possibility of motiveless acts, see Frederick (2010).

By “probability” here, and throughout this discussion, I mean it in a classical sense. In this sense, to say (for example) that there is a 5% probability of a twenty-sided die rolling 1 is to say that, of all the possible cases of a twenty-sided die being rolled, 5% of them will yield a 1. Such a conception of probability is neutral about whether and how these outcomes are determined. If we could measure the motivational force of desires as easily as we could the probability of die rolls, we could say that one desire has a higher probability of resulting in action than another, but this would not commit us to any particular account of free will, determinism, or the like.

To clarify: the incoherence lies in holding that the reason why there was zero probability of Jesus’s performing the act in question is that his desire to do so had insufficient inherent motivational viability. One might hold that his desire was potentially motivating, but there was some other reason why there was no chance of his acting on it. That would be a different kind of bounded internalism, such as the ones discussed below.

As McKinley points out, this is not to say that these theologians defend this model consistently, and alternative views can also be found in their works.

Clearly this objection rests upon McKinley’s view that Jesus’s experience during temptation should

have been the same as everyone else’s, which is not my view.

This has been recognised as a major element of self-control since Brehm and Self (1989).

Ego depletion occurs when the tasks involving self-control follow in quick succession, not if they are more widely spaced out. Self-control works rather like physical exercise. If I spend an hour lifting weights, I will become tired and less able to lift them afterwards – but if I do this regularly my ability to lift weights will improve over time. Similarly, if I resist temptation, I will find it harder to resist another temptation soon afterwards – but if I resist temptation regularly, my ability to do so will improve over time. On this, see Baumeister et al. (2006).

Hiroki Kotabe and Wilhelm Hofmann express a similar idea within the context of a version of the resource model, according to which the resource that gets depleted when exercising self-control is simply “cognitive abilities that rely on directed attention” (2015, p. 626, emphasis in original)

Some recent interpreters have argued that this is not Aristotle’s view, and he thinks that at least sometimes the akratic agent acts even with full awareness of the wrongness of her act. On these differing interpretations, see Destrée (2007) and Erginel (2016).

Middleton uses the terminology of propatheiae here, but this is not MD bounded internalism, because in her conception, the fact that Jesus experiences only propatheiae is not the explanation for why he never acts wrongly. His perfect will provides that explanation, and his lack of full-blown sub-optimal desires is another effect of this will.

However, Maximus did not think of the interior impulse to sub-optimal action as Jesus’s own desires. Rather, it was a constant interior attack upon him by demonic forces attempting to subvert his will (Heidgerken, 2021, p. 107). One might, then, interpret Maximus as closer to externalism than to internalism.

On this, see Bathrellos (2004, pp. 99–174) and Blowers (2012). McFarland (2007) draws out the implication that, for Maximus, to will something is not the same thing as to choose it.

Traditionally, the doctrine of Christ’s beatific vision would accompany a version of externalism, not internalism. On Aquinas’s arguments for Christ’s beatific vision, see Lim (2022). Galot (1986) gives a much-discussed rejection of the doctrine, as does Weinandy (2004). For defences of the doctrine, see Tran (2004), Gaine (2015), and White (2015, pp. 236–74).

On intellectualism and voluntarism in the later Middle Ages, see Hoffmann (2011).

I presented earlier drafts of this paper at the Logos Institute, St Andrews, and the Joseph Butler Society, Oriel College, Oxford. I would like to thank both these institutions for their kind invitations and the audiences for the very fruitful discussions, which helped to guide my thinking on this topic.