Hostname: page-component-74656f9884-8r8nv Total loading time: 0 Render date: 2026-11-08T16:36:14.734Z Has data issue: false hasContentIssue false
  • English
  • Français

‘His death belongs to them’: an Edwardsean participatory model of atonement

Published online by Cambridge University Press:  13 February 2017

JONATHAN HILL
JONATHAN HILL*
Affiliation:
Department of Theology and Religious Studies, University of Exeter, Amory Building, Rennes Drive, Exeter EX4 4RJ, UK
*
e-mail: j.c.hill@exeter.ac.uk
Rights & Permissions [Opens in a new window]

Abstract

The Participatory Model of Atonement (PMA) offers an alternative view of Christian salvation, drawing on Pauline theology. It conceives of sin as a contagion which can usually be escaped only by dying. By ‘participating’ in Christ's death, the believer can escape its effects without having to die. This notion of ‘participation’ is obscure. I consider a possible way of clarifying it using metaphysical ideas taken from Jonathan Edwards. ‘Participation’ might involve becoming similar to Christ through the action of the Holy Spirit, to such a degree that a person might be called identical (in some sense) with Christ.

Information

Type
Original Article
Information
Religious Studies , Volume 54 , Issue 2 , June 2018 , pp. 175 - 199
Check for updates
Copyright
Copyright © Cambridge University Press 2017 

It is commonplace, in discussions of the Christian concept of atonement, to draw attention to the fact that (unlike the Trinity and incarnation) Christian orthodoxy has never sought to define the doctrine of atonement beyond a general affirmation of its truth. The Symbol of Nicaea states that the Son became incarnate ‘for us humans and for our salvation’ (Tanner (Reference Tanner1990), I 5), but neither the bishops at Nicaea nor their successors at later councils tried to explain how Christ's life, death, or resurrection brought about salvation, or indeed what salvation itself consists of.

It is also commonplace to describe later Christian reflection on the atonement in terms of various competing ‘models’, offering different explanations of how Christ saves sinners.Footnote 1 It has become standard to categorize these models into two main types: ‘objective’ models (versions of which are attributed to Anselm of Canterbury and John Calvin) and ‘subjective’ ones (attributed to Peter Abelard).Footnote 2

‘Objective’ models, in brief, are those where Christ is regarded as having performed some action that affects God's attitude towards sinners: either he makes satisfaction on their behalf (as in Anselm and Aquinas) or suffers the penalty for their sins in their place (as in Calvin and later defenders of ‘penal substitution’, particularly in the evangelical tradition).Footnote 3 ‘Subjective’ models, by contrast, view Christ's action as directed towards sinners themselves, with salvation consisting in (or depending upon) a change in lifestyle inspired by the love demonstrated in Christ's life and death.Footnote 4

There are well-known advantages and disadvantages to both groups of models, leading many theologians either to try to hold them together at the same time or to find alternative models that combine the best features of both.Footnote 5 One such alternative approach is the ‘participatory’ model of atonement (PMA).

The participatory model

The PMA has a slightly unusual history. Arguably a central element of Paul's theology, it has been well known to New Testament scholars for decades, but has been conspicuously absent from theological surveys of atonement models. Only now are philosophical theologians starting to take it seriously. In particular, Tim Bayne and Greg Restall (Reference Bayne and Restall2009) have defended this model on the grounds that it avoids some of the pitfalls associated with its more commonly discussed rivals.

The central biblical text on which the model is based is Romans 6–7, where Paul writes about how, through Christ, believers may escape their old life and begin a new one. He states:

Do you not know that all of us who have been baptized into Christ Jesus were baptized into his death? Therefore we have been buried with him by baptism into death, so that, just as Christ was raised from the dead by the glory of the Father, so we too might walk in newness of life. For if we have been united with him in a death like his, we will certainly be united with him in a resurrection like his. We know that our old self was crucified with him so that the old body of sin might be destroyed, and we might no longer be enslaved to sin. For whoever has died is freed from sin. But if we have died with Christ, we believe that we will also live with him. (Romans 6:3–8; all biblical quotations are from the NRSV)

From the perspective of a philosophical theologian, this passage looks very like an attempt to explain how Christ's death and resurrection save believers. We may summarize a model inspired by this passage like this:Footnote 6

  1. (1) Sin is an ontological reality that binds human beings, corrupting their nature, destroying their relationships, and leading them to disobey God. This results in each person suffering from a sinful identity: they are not simply people afflicted by sin, they are sinners.Footnote 7

  2. (2) The only way for sinners to escape sin and cease being sinners is to die and lose this old, sinful identity.

  3. (3) Through participation in Christ via faith and the sacraments, human beings can share his death and ‘die’ to their old identities.

  4. (4) They can also share his resurrection and gain new identities.

  5. (5) As a result, human beings can be remade with new identities, free from the power of sin and the guilt that it brings.

At the heart of the PMA is the notion of participation, which is twofold. First, sinners are able to identify with Christ; in Pauline terms, they are ‘in Christ’. This means not simply sympathizing with him or imitating him, but in some sense becoming identical with him. Second, as a result of this, sinners acquire Christ's properties, or at least some of them. The idea rests on Leibniz's Law: if x is (numerically) identical with y, then y must have all the properties that x does, just as (say) Clark Kent has all the properties that Superman does, because they are one and the same. Christ died: the sinner who participates in Christ therefore dies too (in some sense). Christ rose again: the sinner will therefore rise too. Because he has died and risen again, Christ is free from the power of sin: the sinner is therefore free from it, too. This is what salvation means. We should therefore be careful to distinguish the PMA from theological discourse that speaks of participation in Christ but views it as distinct from the mechanism of atonement and salvation. Many theologians might be willing to use this language, but only in conjunction with other explanations of atonement. The PMA is distinct in that it makes this language primary.

The PMA combines elements of both the ‘objective’ and ‘subjective’ models of traditional analysis, but is distinct from both. Like ‘objective’ models, something really happens on the cross and in the garden tomb. Christ's death actually accomplishes something that would have been theoretically impossible otherwise. But what it accomplishes is not the paying of a debt or the satisfaction of a moral law, but the transformation of human beings. And like ‘subjective’ models, the focus is on sin as a personal reality and on the need for moral and spiritual change in sinners. But sinners are changed not by imitating Christ's example, but by sharing in his identity through participation.

Bayne and Restall note a number of positive elements to the PMA. First, it presupposes a largely non-deontic view of sin. Bayne and Restall criticize ‘objective’ models on the grounds that they characterize sin ‘as purely a failure on our part to fulfil our moral obligations’ (Bayne & Restall (Reference Bayne and Restall2009), 151). They argue that this is an inadequate notion of sin, and that while sin certainly does include deontic elements (a sinner is, among other things, someone who has failed in her moral duties), this is not its most important character (ibid., 152–153). Rather, sin is more fundamentally relational and ontological: the sinner is someone whose relationship with God is broken, and who is in some way metaphysically corrupted.

Where traditional ‘objective’ models see salvation as consisting in, or depending upon, a legal or quasi-legal resolution of transgressions committed by the sinner, the PMA sees it as involving the transformation of the sinner. The sinner is, above all, someone who suffers from being in a sinful condition, analogous to an incurable disease (or, as Bayne and Restall call it, ‘a sickness and a pollutant’ (ibid., 154). Precisely what this involves needs to be set out in more detail, but for now we may appeal to the traditional Christian doctrine of concupiscence, the notion that – whether as a result of biological inheritance or our social context, or both – we are subject to overwhelming desires that inevitably result in our committing sinful acts. Like an incurable disease, the sinful nature can be escaped only by either dying or undergoing a miraculous transformation.Footnote 8 This is why atonement has to involve death: death is not, on this model, a punishment or a payment; it is the only way of escaping sin.

This leads to the second positive element of the PMA, on Bayne and Restall's analysis. It presents the fundamental problem of sin, and the nature of salvation, as focused on identity.Footnote 9 Sinners escape sin by literally (on some level) becoming new people. Bayne and Restall write:

Paul's conception of sin is not, primarily, deontic. He does not see Christ's death and resurrection as the salve for a troubled conscience . . . The participatory strand in Paul's theology takes sin to be a problem of our identity – the atonement is not a matter of balancing the moral ledger but involves the inauguration of a new form of humanity. (ibid., 160)

By changing the identity of sinners, they escape sin conceived in ontological terms as a kind of corruption of human nature, because they are refashioned without it. Not only this, but the deontic element of sin is addressed too. Because sinners have changed their identity, they are no longer subject to the power of concupiscence that was bound up with their old identity. Moreover, they are no longer the same people who committed their past transgressions, and are therefore free from the guilt of those past transgressions.

Weaknesses of the model

But these strengths of the PMA are bought at the cost of some serious potential weaknesses. One, which Bayne and Restall do not address, is the possible incoherence of its understanding of the nature of sin. Sin, in this model, is portrayed as a corruption of human nature that must be purged through remaking the self. Does this mean sin is a physical corruption, literally a kind of illness of the body? If so, and if this corruption is removed by participation in Christ, then in theory one could distinguish between those who have been saved and those who have not by using a microscope – an absurd consequence. Perhaps instead it is a metaphysical corruption, of the soul rather than of the body. But in that case, this model of the atonement will require a belief in a soul, as a substance distinct from the body – something about which many theologians today, like most philosophers, are dubious.

Alternatively, perhaps sin is a purely psychological corruption, akin to non-chemical addiction or compulsive behaviour. But then the talk of ‘sickness’ and ‘pollutant’ must be metaphorical to some degree. And it is not clear why a change in identity of the kind envisaged by this model should eliminate (or reorder) these psychological traits and not other ones, such as those that constitute our personalities.

Moreover, if we conceive of sin in terms of concupiscence – an overwhelming tendency to commit sins – and we think of participation in Christ as remaking individuals’ natures such that they escape this power of sin, then those who have faith in Christ and participate in the sacraments should be free from the power of concupiscence and able to refrain from sinning. But this is inconsistent with experience.

These issues with the conception of sin relate closely to an even more serious problem with the PMA, which is the concept of ‘participation’ itself. What exactly does it mean to ‘participate in Christ’? How can the believer be considered ‘identical’ with Christ in any serious sense? One possible answer is this: to say that the Christian ‘dies with Christ’ (as Paul puts it) is to say that she identifies with Christ to such a degree that she replicates Christ's death in her own life, in a spiritual way. On this interpretation, ‘identity with Christ’ is more of an experiential reality than an ontological one: the Christian identifies with Christ rather as the sports fan identifies with a team. Just as the fan can say, ‘We lost today’, even though he did nothing himself, so too the Christian can say, ‘I died to sin’. By living as Jesus did, the Christian escapes the corrupting influence of sin.Footnote 10 ‘Identity’, in this version of the model, means something akin to national or cultural self-identity.

Such an interpretation may seem attractive, but it risks stripping the PMA of everything that distinguishes it from other models of the atonement. Bayne and Restall imagine the following objection:

The Christian doesn't literally die on the cross with Christ, and she isn't literally reborn with Christ in his resurrection. What sense can we really make of participating in the death and life of Christ? There is really no such thing as a participatory model of the atonement; instead, all we have are a motley and confusing assortment of metaphors. (ibid., 161)Footnote 11

This version of the PMA is effectively a version of the ‘subjective’, Abelardian model of the atonement, framed in Pauline language. Even so, it does still have some distinctive elements to it. The traditional exemplarist model focuses on Christ as a moral exemplar, primarily of love, and views atonement as consisting of imitating that loving example. What we might call the ‘subjective participatory model’ focuses more on identity, and on the sinner's need consciously to change her self-identification away from self and towards Christ.Footnote 12

But can there be an ‘objective participatory model’?Footnote 13 Such a model would take ‘identity’ not in the social sense but in an ontological sense. It would hold that the Christian is ‘identical’ with Christ in the sense of numerical identity, at least on some level.Footnote 14 On such a model, the Christian does die on the cross with Christ. Unlike the subjective version of the PMA, such a model is not reducible to one of the existing models of atonement. But it takes us back to the major problem of how the Christian can be considered identical with Christ, and what ‘participation’ consists in. If x is identical with y, this means that ‘x’ and ‘y’ are simply different names for the same thing, like ‘Clark Kent’ and ‘Superman’. If x is identical with y, anything that is true of x is true of y. But clearly the Christian is not identical with Christ in this sense, and does not share all of his properties. The Christian does not become a Galilean carpenter.Footnote 15 But then what kind of identity is involved, and how does the Christian acquire some of Christ's properties but not others?Footnote 16

Bayne and Restall offer two suggestions in response to this problem. One is to point out that we do commonly think of moral identity as being sometimes distinct from personal identity, and so it may be plausible to think of the atonement as involving a change in one without a change in the other. The other is to note that other central Christian doctrines, such as the incarnation and Trinity, if taken literally, also involve problematic identity statements (ibid., 162–163).

In themselves, these are no more than starting points, since they do not explain how we should conceive of the notion of identity that is employed in the PMA. But we can develop a clearer approach, and a more detailed model, using insights from the American Puritan theologian Jonathan Edwards (1703–1758).Footnote 17

Edwards on identity

Jonathan Edwards did not hold the model of the atonement described above.Footnote 18 But his treatment of other doctrines, notably original sin and incarnation, closely parallels the PMA. In his writings on these topics, we find Edwards deeply concerned with questions of identity: how one person can be identical with another in some sense while distinct in another, and how properties of one can be ascribed to another.

In what follows, I briefly examine Edwards's treatment of these doctrines, focusing on elements of particular relevance to our purpose. After this, I suggest ways in which these elements could be used to construct a version of the PMA which Edwards himself might not have endorsed but which builds on his thought.

I begin with an argument Edwards develops regarding the doctrine of original sin, in book IV, ch. 3 of one of his last work, The Great Christian Doctrine of Original Sin Defended. He argues that Adam's descendants are, in some sense, identical with Adam, and consequently share his guilt:

God, in each step of his proceeding with Adam, in relation to the covenant or constitution established with him, looked on his posterity as being one with him . . . From which it will follow, that both guilt, or exposedness to punishment, and also depravity of heart, came upon Adam's posterity just as they came upon him, as much as if he and they had all coexisted, like a tree with many branches . . . I think, this will naturally follow on the supposition of there being a constituted oneness or identity of Adam and his posterity in this affair. (WJE III 389–390; all italics in Edwards's quotations are original)

Edwards's framing of original sin as involving ‘identity’ between Adam and his descendants diverges from typical Reformed approaches to this topic at the time. It was more usual to argue that God justly imputes Adam's sin to his descendants because Adam acts as his descendants’ representative.Footnote 19 Here, then, Edwards unfolds a version of the original sin doctrine both original in its details and strikingly similar to the PMA. Just as, in the latter, sinners are somehow identical with Christ and share his properties, here they are somehow identical with Adam and share his properties, namely his guilt.Footnote 20

The major objection to this doctrine, which Edwards ascribes to his antagonist John Taylor, is correspondingly similar to the major objection to the PMA: ‘that Adam and his posterity are not one, but entirely distinct agents’ (WJE III 394). Edwards's reply to this objection is an elegant piece of theological metaphysics.Footnote 21 He argues that ‘created identity or oneness with past existence, in general, depends on the sovereign constitution and law of the supreme Author and Disposer of the universe’ (WJE III 397). That is, personal identity over time always depends upon God's decree. Edwards argues that the existence of any given object (e.g. the moon) at any given moment (e.g. t2) must be caused either by itself at an earlier time (t1) or directly by God. But the former is impossible, because that would violate the principle that an object has no causative power in times or places where it does not exist (WJE III 400). Therefore, the moon at t1 and the moon at t2 are both caused directly by God and have no causal relation to each other, any more than objects far apart in time and space do (WJE III 401). God's conservation of the universe ‘is altogether equivalent to an immediate production out of nothing, at each moment’ (WJE III 402).

We may call Edwards's view an extreme version of occasionalism:Footnote 22 God does not merely directly cause every event in the world, but recreates the entire world (with minimal changes) from moment to moment. So where we think a single individual persists over time, in fact there is a series of quite distinct individuals, each lasting for only a moment.Footnote 23 This is a clear rejection of endurantism (the view that the whole of an object is present at any given time and endures from one moment to the next); Edwards unequivocally holds four-dimensionalism.Footnote 24 It may also seem like a rejection of perdurantism (the variety of four-dimensionalism that holds that objects are genuinely temporally extended, so that its temporal parts compose a genuine unity). However, Edwards thinks that God chooses to regard the distinct temporal stages as a single individual persisting through time. This means that he can be seen as endorsing a version of perdurantism.Footnote 25 I-at-t1 am a temporal part of a whole that includes me-at-t2 and a host of others – but these different temporal parts form a single whole only because God regards them as doing so.Footnote 26 Edwards's main point is that If God can do this, then he can also regard each descendant of Adam as identical with Adam, just as he regards each successive me as identical with each other (WJE III 404). If it is just for God to hold me-at-t2 guilty for something I-at-t1 did, it can be just for God to hold me guilty for something Adam did.Footnote 27

God's decision to regard me-at-t2 as identical with me-at-t1 is not a whim.Footnote 28 Identity over time is connected to psychological similarity, and Edwards cautiously endorses Locke's consciousness-based theory of personal identity. On this theory, I am the same person who had cereal for breakfast this morning because my ‘consciousness’ extends back to that point, as shown in the fact that I remember it (WJE III 398).Footnote 29 Clearly, though, there is no psychological continuity between Adam and his descendants. So how can God regard them as identical? Edwards points out that ‘[t]here are various kinds of identity and oneness, found among created things, by which they become one in different manners, respects, and degrees, and to various purposes’ (WJE III 404–405). These different kinds of identity are founded on God's will, which itself is ‘regulated by divine wisdom’ (WJE III 406), and the divine wisdom may have different reasons for its decisions. Edwards notes two kinds of reasons: first, God may regard two things as identical if doing so has ‘analogy and harmony with other laws’ (which Edwards does not really explain);Footnote 30 and second, he may do so because it results in beneficial consequences (WJE III 406). He goes on to argue that both of these conditions apply to the identification of Adam with his descendants. Moreover, he insists elsewhere in the book, Adam's status as the first human makes him peculiarly appropriate to be regarded as identical with all of his descendants (WJE III 259–260).Footnote 31

We may, then, summarize Edwards's theory of identity, as expressed in On Original Sin, like this:

  1. (1) No object persists through time; there are only discrete objects existing momentarily at each instant. Call this the Continuous Creation Thesis.

  2. (2) Identity through time depends entirely upon the will of God, who is able to regard two or more discrete objects at different times as identical, thereby making them identical. Call this the Divine Voluntarism Thesis.Footnote 32

  3. (3) God does this when he has a good reason to do so, either because (a) it is consistent with his other decisions, or because (b) it has good consequences. Call this the Divine Wisdom Thesis.

  4. (4) In normal cases of personal identity over time, God regards the different temporal entities as identical because of psychological similarity between them, which itself holds only because he has chosen to create these different temporal entities with that psychological similarity. (Edwards does not say whether this is an example of 3a or 3b above, or an entirely different basis for identity.) Call this the Psychological Similarity Thesis.

We can see how some of these principles can be applied to other doctrines by considering Edwards's approach to one of the thorniest problems in Christian philosophical theology: that of incarnation.

Incarnation and identity

Traditionally, Christians believe that Jesus of Nazareth, the historical human being who lived in Galilee, was identical with God the Son, the second Person of the Trinity. But it seems to follow that the Son has properties that theologians are traditionally reluctant to ascribe to God. He is, for example, temporally and spatially located, he undergoes change, and he suffers and dies. Theologians have proposed various solutions to this problem, but the most popular throughout Christian history is what is now known as ‘compositionalism’. On this view, the Son became incarnate not by being transformed into a human being like a caterpillar being transformed into a butterfly, but by acquiring or ‘assuming’ a human body and soul. A relation of some kind exists between the Son on the one hand (the divine nature) and the human body and soul on the other (the human nature), and it is in virtue of this relation that the Son can be called genuinely human.Footnote 33 The human nature is located in time and space, suffers change, and so on, but the divine nature does not. Most of the historical Christological disputes have been between people who accepted this general story but disagreed about what relation holds between the two natures.Footnote 34 The model continues to be defended today.Footnote 35

A problem with this approach to incarnation is that if Christ's human nature is a particular human individual – the body, animated by a rational human soul, walking around Galilee – which is ‘assumed’ by the divine Son, then it is distinct from the divine Son. This, indeed, is demanded by orthodoxy: the Definition of Chalcedon specifies that there is ‘no confusion’ between Christ's two natures (Tanner (Reference Tanner1990), I 86).Footnote 36 But given this, it is hard to see how the Son himself is human at all. He seems, rather, to be associated with a human being. This sounds like Nestorianism, the heretical view that there are two persons in Christ, one human and one divine.

So a traditional metaphysics of incarnation has to find a way between two unacceptable extremes. Jesus of Nazareth has to be identical with the Son in the sense that they are one and the same person. But he cannot be identical with the Son in the sense that his human nature is the same thing as his divine nature.

One again, the problem parallels the one we have seen with the PMA. In each case, we have the claim that a human being can be considered identical (in some sense) with a divine figure, while at the same time being non-identical (in some other sense) with that figure. In the case of incarnation, Jesus is ambiguously identical with the Son. In the case of atonement, the sinner is ambiguously identical with Christ.

A common image used to portray the relation between Christ's two natures is that of the body and soul. Just as normal human beings (assuming some kind of substance dualismFootnote 37) are composed of two concrete particulars which nevertheless compose a single genuine unity, so too is Christ.Footnote 38 For some theologians, this means that just as the soul uses the body as a kind of instrument, so too God the Son uses his human nature as an instrument to act in the physical world.Footnote 39

There is biblical warrant for applying the same idea to the relation between Christ and his followers. Paul calls the church the ‘body of Christ’Footnote 40 and seems to mean by this – among other things – that each individual believer acts like an organ of Christ: Christ uses them instrumentally to achieve his aims. But there is nothing in these passages to suggest that being the ‘body of Christ’ has anything to do with the process of dying to sin found elsewhere in Paul. Moreover, it is questionable whether such an analogy is adequate to ground the transference of properties, let alone identity.Footnote 41

Edwards on incarnation

As with his handling of original sin, Jonathan Edwards recognizes that the doctrine of incarnation raises a problem of identity. The incarnation involves identity not over time but between a temporal object and a non-temporal one. But it is easy to extend the Divine Voluntarism Thesis to cover non-temporal objects as well as temporal ones. If God is capable of regarding objects at different times as identical, why should he not be capable of regarding temporal objects as identical to non-temporal ones too?

Given this, it may seem that Edwards has an easy answer to the problem of how Jesus can be identical with the Son. The Divine Wisdom Thesis allows him to hold that God can regard Jesus as identical with the Son if doing so has beneficial consequences. Clearly, it does have beneficial consequences, as it allows for both the incarnation and the atonement. Edwards could therefore construct an argument for the union of Christ's natures much like his argument for the union of Adam with his descendants.

However, Edwards does not take this easy route. Perhaps, when he wrote on the incarnation, he had not yet developed the Divine Wisdom Thesis (On Original Sin, where it appears, was written right at the end of his life). Even if he had developed it, though, he might have found it unsatisfying as an explanation of how the incarnation works.

Edwards evidently wanted to find such an explanation. Although he wrote little on the incarnation in his published works, the topic appears a number of times in his Miscellanies, a long series of short notes written throughout his life for his personal use.Footnote 42 In them we find Edwards experimenting with ideas that are often not fully developed. One of the most important is Miscellanies 487, entitled ‘Incarnation of the Son of God and union of the two natures of Christ’, written in late 1730 or early 1731 (WJE XIII 87–88; Withrow (Reference Withrow2011), 126–135). In this short text, Edwards tries to answer the question we have just been looking at: how can Jesus be considered identical with the divine Son? He writes:

’Tis not [just] any communion of understanding and will that makes the same person, but the communion of understanding is such that there is the same consciousness. Thus [1] the man Christ Jesus was conscious of the glory and blessedness the Logos had in the knowledge and enjoyment of the Father before the world was, as remembering of it (John 17:5). (2) He has the same spirit or disposition towards the Father; not as believers have a filial spirit, or spirit of children, but he has the Spirit of the only begotten of the Father. He is disposed towards the Father as being his own Father in the manner that he is the Father of the Logos, and is conscious of a respect in the Father to him as the only begotten Son . . . (WJE XIII 529)

Although this text was written over twenty-five years before On Original Sin, the reasoning parallels parts of the argument in the later book. Edwards appeals to the Psychological Similarity Thesis, but here applies it to Christ. Jesus is identical with the Logos because of the qualitative similarity – the ‘same consciousness’ – that he bears towards the Logos. The fact that Jesus and the Logos are distinct entities is irrelevant, just as the fact that I-at-t1 and I-at-t2 are distinct entities does not prevent God from decreeing (on the basis of our psychological continuity) that we are identical.

But there is more to Edwards's argument here. The properties in virtue of which Jesus resembles (and is thus identical with) the Logos are relational properties. Edwards lists them in the passage quoted above:

  1. (1) Consciousness of the eternal knowledge and enjoyment of the Father

  2. (2) Knowing the Father as his own Father

  3. (3) Knowing the Father's attitude towards him as the Son

In other words, what makes somebody identical to the Son is bearing a certain relation (or set of relations) to the Father. This idea goes back at least to the Cappadocian Fathers, who argued that the Persons of the Trinity are distinguished solely by their mutual relations.Footnote 43 Shortly before he wrote Miscellanies 487 on the incarnation, Edwards had begun work on his Discourse on the Trinity, in which he developed the argument that the Son is the Father's idea of himself.Footnote 44 Because this reflexive idea is perfect, it constitutes a divine person in its own right. As Edwards puts it:

God's idea of himself is absolutely perfect, and therefore is an express and perfect image of him, exactly like him in every respect. There is nothing in the pattern but what is in the representation – substance, life, power, nor anything else – and that in a most absolute perfection of similitude; otherwise it is not a perfect idea. But that which is the express perfect image of God, and in every respect like him, is God to all intents and purposes, because there is nothing wanting; there is nothing in the Deity that renders it the Deity but what has something exactly answering of it in this image, which will therefore also render that the Deity. (WJE XXI 114)Footnote 45

So for Edwards, what makes the Son the Son of the Father is that he is the Father's idea of himself. What makes him a divine Person is that, as idea, he represents the Father perfectly and therefore resembles him perfectly. In Miscellanies 487, he makes no reference to the notion of the Son as the Father's idea of himself, but this fundamental relation of ‘absolute perfection of similitude’ is present. Just as this relation constitutes the identity of the Son within the Trinity, so too it identifies Jesus as the incarnate Son.

We may, then, pick out the following claims from Edwards's arguments in Discourse on the Trinity and Miscellanies 487:

  1. (1) The Son is the perfect representation of the Father, meaning that he resembles him perfectly, and bears certain relations towards him as a result. Call this the Divine Resemblance Thesis.

  2. (2) Only the Son bears this resemblance, and these relations,Footnote 46 to the Father. Call this the Unique Sonship Thesis.

Despite the Unique Sonship Thesis, Edwards accepts that there is a thing that, in some sense, is not identical with the Son, yet which bears the relevant relations to the Father: Jesus, or Christ's human nature. But he seems to hold a hidden additional thesis:

  1. (3) Relations are (or can be) identity-constituting. Call this the Identity Constituting Thesis.

This means that, if Jesus perfectly resembles the Father and bears towards him the same relations that the Son does, he simply is the Son. This is a Christological parallel to the Psychological Continuity Thesis, but it has a Trinitarian basis.

There is another line of thought in Miscellanies 487 (and much of Edwards's other writing on the incarnation): Christ's human nature is united to the divine nature through the indwelling of the Holy Spirit.Footnote 47 Edwards points out that believers are united to God through the Holy Spirit, who dwells within them. Why, then, should Jesus not be united to God in a similar way?

As the union of believers with Christ be by the indwelling of the Spirit of Christ in them, so it may be worthy to be considered, whether or no the union of the divine with the human nature of Christ ben't by the Spirit of the Logos dwelling in him after a peculiar manner and without measure. Perhaps there is no other way of God's indwelling in a creature but by his Spirit. (WJE XIII 528)

Here we see the fundamental idea that believers’ participation in the divine via Christ is not a unique phenomenon but of the same kind as Jesus’ own participation in the divine.Footnote 48 In other words, sanctification is the same kind of thing as incarnation, though it is not identical. Edwards immediately goes on:

The Spirit of Christ's dwelling in men causes an union, so that in many respects [they may be] looked upon as one: perhaps the Spirit of the Logos may dwell in a creature [i.e. Christ's human nature] after such a manner, that that creature may become one person [with the Logos], and may be looked upon as such and accepted as such. (WJE XIII 528)

Here again the Trinitarian basis of Edwards's understanding of incarnation is clear. In the Discourse on the Trinity, Edwards repeats and develops the Augustinian idea that the Holy Spirit is identical with the bond of love between the Father and the Son (WJE XXI 121–122; Hastings (Reference Hastings2005), 283). If, then, Jesus’ identity with the Son is founded upon his bearing the same relation to the Father that the Son does, the Holy Spirit must be involved.

Edwards effectively identifies two kinds of relations that hold between the Father and the Son. One is ontological, involving the Holy Spirit. The other is psychological, involving knowledge of the Father that only the Son has. Jesus bears all of these relations to the Father, and is therefore identical with the Son. The two kinds of relations are closely related, though, because it is through the indwelling of the Holy Spirit that Jesus has the psychological state of the Son:

The man Jesus becomes one person by a communion of knowledge and will; but as in believers all divine knowledge is by the Spirit . . . so, I suppose, ’tis by the Spirit that divine knowledge and consciousness is given to the man Jesus. And so, as ’tis by the Spirit of God that a divine temper is given to men and angels, so I suppose ’tis by the Spirit of the Logos that the man Jesus hath the spirit and temper of the only begotten Son of God. (WJE XIII 530)

We may, then, supplement our previous summary of Edwards's reasoning with these additional points:

  1. (1) Jesus’ union with the Son is similar in kind to believers’ union with Jesus. Call this the Consistent Union Thesis.

  2. (2) Union of creatures to God is always achieved through the indwelling of the Holy Spirit. Call this the Unifying Spirit Thesis.

An Edwardsean model of the PMA

We are now in a position to construct a participatory model of the atonement using these materials from Jonathan Edwards. Leaving exposition of Edwards entirely aside now, our purpose is to see whether someone who accepts (some of) the Edwardsean claims we have just examined could use them to defend a version of the PMA (not necessarily one that Edwards would have accepted).

Let us begin by recapitulating the claims we have seen in Edwards:

  1. (1) Continuous Creation Thesis – a controversial metaphysical commitment that not all theists will be willing to accept.

  2. (2) Divine Voluntarism Thesis – again, a controversial claim, whether taken in its strong or weak versions.

  3. (3) Divine Wisdom Thesis – this seems much more plausible.

  4. (4) Psychological Continuity Thesis – a widely, but not universally, held thesis.

  5. (5) Divine Resemblance Thesis – a plausible Trinitarian claim.

  6. (6) Unique Sonship Thesis – a claim that most traditional Trinitarians would accept.

  7. (7) Identity Constituting Thesis – a claim that might be generally acceptable in a solely Trinitarian context, but which is much more tendentious as a general metaphysical principle.

  8. (8) Consistent Union Thesis – a plausible and unproblematic claim.

  9. (9) Unifying Spirit Thesis – a plausible claim, at least for most Christians.

Edwards appeals to (1)–(4) of these in his defence of the doctrine of original sin, and he could (but doesn't) appeal simply to (2) and (3) to defend the incarnation. The same applies to the PMA. A theologian who accepts (2) and (3) could argue that, if God were to regard repentant sinners as identical with Christ, the benefits would be very great; therefore divine wisdom leads him to do so. However, the Divine Voluntarism Thesis is problematic enough for this to be an unappealing solution.

It would be better to use the Trinitarian and incarnation-based claims. Just as Edwards uses (5), (6), and (7) to show that Jesus is the Son, so too a theologian who accepts these claims could use them to argue that Christians are identical with Christ. An obvious difference between the two cases is that, on Edwards's account (and any Christian could plausibly agree), Jesus was psychologically and relationally exactly similar to the Son;Footnote 49 but Christians clearly are not. But the defender of the PMA might appeal to (9): just as (according to Edwards) the Holy Spirit works within Jesus to render him exactly similar to, and therefore identical with, the Son, so too we might say that the Holy Spirit works within Christians to make them exactly similar to the Son as well.

But if the Holy Spirit is really at work within Christians in this way, it is a long process, one that takes place throughout the Christian life and beyond. But the effects of Christ's atonement are supposed to be available right from the start of the Christian life. In the passage from Romans with which we began, for example, Paul locates the Christian's ‘dying with Christ’ in the past. But if this dying with Christ is based upon identity with Christ, then the Christian must be identified with Christ at an early stage of her Christian faith, not as something to look forward as a result of the activity of the Holy Spirit.

One possible solution is to see identity as a matter of degree. Just as a person may become more and more like Christ over time, so she becomes more and more identical with him. As soon as the Holy Spirit begins to work within her, giving her even a little of what Edwards calls ‘the spirit and the temper’ of the Son, she is identical with the Son to some degree. This is a very problematic solution. On the classical view of identity, any x and y simply are or are not identical; identity does not come in degrees. While the notion that a person may be partially (qualitatively) similar to the Son is readily comprehensible, the notion that a person may be partially (numerically) identical to him is not.

This solution could be saved if we introduce the notion of different kinds of identity. Because the Christian bears the appropriate relations to the Father (and, perhaps, to the Spirit), the Christian is genuinely identical with the Son, but (crucially) this is a different sort of identity from the mundane variety. As Bayne and Restall point out, ‘moral identity’ is arguably distinct from ‘numerical identity’, and is ‘a matter of one's commitments, values and relationships’ (Bayne & Restall (Reference Bayne and Restall2009), 163). We could develop this idea of different kinds of identity further. We might distinguish between ordinary identity (identityo) and divine identity (identityd). Identityo may itself be subdivided, depending on what our theory of identity is; for our purposes, any theory will do. What matters is that identityd is something different. Two things may be identicald if (a) one of them is a divine Person and (b) they are both characterized by a certain set of exactly similar relations, such as those we have already considered.Footnote 50

The viability of this model rests, to a considerable degree, upon the plausibility of claim (7) above, the Identity Constituting Thesis. But this may be reasonably plausible, at least to a traditional Trinitarian Christian, if we restrict that thesis to relations involving divine Persons. It would mean that some of Christ's properties, such as his having died and his being resurrected, can be truly said to apply to the Christian, because the Christian is truly identicald with Christ even without being identicalo with him. This version of the PMA thus avoids the problem of collapsing into the Abelardian subjective model. However, it needs to account for the fact that only some of Christ's properties apply to the Christian: the Christian does not acquire the property of having been born in a stable, for example; so why should she acquire the property of having died? It must also explain why the properties that are transferred to the Christian apply to her in a distinctive way. The Christian has not died in the same sense as Christ has died. But if the Christian is genuinely identical with Christ, she should have.

We could fill these gaps in the model by holding that when the Christian becomes genuinely identicald with Christ, this does not mean that (some of) Christ's properties are automatically transferred to the Christian by Leibniz's Law. Rather, it could mean that God chooses to transfer some of them – or, which may come down to the same thing, that he chooses to regard the Christian as bearing those properties. It now makes sense that these properties apply to the Christian in a different way from how they apply to Christ, and it also makes sense that only some properties are transferred or imputed (and only from Christ to the Christian, not vice versa). Note that, although this suggestion is inspired by Edwards's Divine Voluntarism Thesis, it does not require holding that thesis (in either its strong or weak versions). We are not suggesting that God directly causes the Christian to be identicald to Christ; that identityd is already based on the Christian's bearing certain relations to the Father and to the Son.Footnote 51 But the consequences of being identicald to Christ may be determinable by God, guided by his wisdom.

The plausibility of the model as here sketched stands or falls with the plausibility of the distinction between identityo and identityd. The very notion of different kinds of identity is problematic in itself. And the claim that identityd comes in degrees is more problematic still, as we have seen. Moreover, if identityd does not result in the automatic transference of properties by Leibniz's Law, then how is it a kind of identity at all? The defender of this model might appeal to the fact that identityd has these counterintuitive features because it concerns the divine – and identityo has, on this account, been insulated from them, which means that our normal, classical intuitions about identity still normally apply. But this appeal to divine mystery still feels unsatisfying.

An alternative solution is to keep the same structure of the model but to remove the appeal to numerical identity at all. Instead of inventing ‘identityd’, a mysterious relation which has little in common with identity as we normally understand it, we can appeal instead just to the similarity that holds between the Christian and the Son as a result of the action of the Holy Spirit. We might say that, ordinarily, such similarity does not result in numerical identity. In the case of Jesus, there was perfect similarity, such that Jesus really was identical to the Son as a result of the Divine Resemblance Thesis, the Unique Sonship Thesis, and the Identity Constituting Thesis. The action of the Holy Spirit on Christians results in their becoming similar to the Son, and perhaps even bearing similar relations to the Father as the Son does, but not to such a degree that they are literally identical to the Son, as Jesus is.

This revised version of the model presents the participation of Christians in Christ solely as a matter of (partial but increasing) resemblance as a consequence of the action of the Holy Spirit, rather than as a mysterious kind of numerical identity. On this view, how do Christians acquire the benefits of Christ's death and resurrection? We could appeal, as we did before, to direct divine action. If the Christian is becoming more like Christ as a result of the action of the Holy Spirit, this must consist of the Holy Spirit causing the Christian to bear more of the properties that Christ does. Among these properties could be some of the effects of having undergone death and resurrection. As we saw earlier, Paul talks in Romans of being ‘buried with him by baptism into death’. By becoming more like Christ through the activity of the Holy Spirit – initiated by baptism – Christians may share in Christ's death in the sense that they acquire the properties that Christ acquired through dying: escaping the power or contagion of sin (however precisely that is understood), becoming remade in a ‘spiritual body’, and so on.Footnote 52

On this model, atonement is a lengthy process. There may be a particular moment – perhaps, as Paul suggests, the moment of baptism – when the believer ‘dies’ in the sense of acquiring key properties of Christ, but the process continues. As Paul writes to the Philippians:

I want to know Christ and the power of his resurrection and the sharing of his suffering by becoming like him in his death, if somehow I may attain the resurrection from the dead. Not that I have already obtained this or have already reached the goal; but I press on to make it my own, because Christ Jesus has made me his own. (Philippians 3:10–12)

As in the model we have sketched here, Paul considers himself to be ‘becoming like [Christ] in his death’, which we might see as the gaining of properties in common with Christ that Christ had in virtue of having died. Resurrection, however, still lies in the future, as a state to be pursued and made possible only ‘because Christ Jesus has made me his own’.

There is one more advantage to this model. We have seen Edwards's suggestion that Jesus’ identity as the Son rests in some way upon his sharing certain properties of the Son. He enjoyed perfect similarity to the Son, as a result of which he was the Son. If the Christian is undergoing a long process of gaining properties in common with Christ – i.e. becoming more and more like Christ – then there might come a point at which the Christian too comes to be perfectly like him, and therefore also identical with the Son. This is the traditional Christian doctrine of theōsis or divinization.Footnote 53 To discuss this fully is well beyond the scope of this article, but as is well known, Orthodox Christianity lays particular emphasis on this doctrine, according to which the ultimate goal of humanity is to become divine, usually understood as sharing in God's properties though not his nature.Footnote 54 Here we may simply note that this version of the PMA unites this doctrine not only with incarnation but with atonement as well.Footnote 55

Participation beyond the PMA

The ‘objective’ versions of the PMA described above aim to answer the most serious objection to the model: how can the Christian be said to ‘participate in’ Christ, and why should this have salvific value? There are still gaps in the model. In particular, the objections raised earlier regarding the conception of sin as a corruption have not been addressed. These objections to the model may prove harder to answer than those we have focused on here.

But there is another way of looking at participatory atonement, which is – once again – suggested by Jonathan Edwards. As noted above, Edwards did not hold the PMA as outlined in this article. But he did sometimes use language very reminiscent of it. In Miscellanies 281 he writes, for example:

The threatening, ‘Thou shalt surely die’ [Genesis 2:17] is properly fulfilled in the death of Christ . . . [By] this threatening, according to the proper and true sense of it, is to be understood that mankind, or the human nature, should die; so that death should belong to all the persons. Now if any way can be found out . . . whereby an actual death should as much belong to all the persons as if every person died, so that it may properly be called their death and yet they saved alive, the threatening is fulfilled. But this is done by the union of their persons with Christ's person: they are made so much one that his death belongs to them; they own it, it is their death, in as proper a sense as Adam's sin is their sin – they sin in Adam and they die in Christ. (WJE XIII 379–380)Footnote 56

In this passage Edwards gives no indication that he thinks that the atonement consists in this ‘owning’ of Christ's death.Footnote 57 But in Miscellanies 398 he presents it as an essential prerequisite for atonement:

Query. What are those qualifications and circumstances of the person of Christ that renders it a just, proper and suitable thing, that he should be looked upon by God as representing those he undertook for in suffering, so that his suffering shall be looked upon as equivalent to theirs, and answering to it, or so that his suffering should be really equivalent, and properly answering to their suffering?

Ans. 1. One thing is, that he was so united to them, so much one with them, that he may be justly looked upon as the same. (WJE XIII 463)Footnote 58

In this latter passage, Edwards goes on to develop a rather different account of the union between Christ and believers, based not (explicitly) on the Holy Spirit but on Christ's self-giving love for humans. But what is striking is that, here, Edwards offers an interpretation of the Reformed penal substitution approach to atonement that incorporates a strongly participatory element.Footnote 59 One of the common objections to penal substitution is that it would be unjust for God to punish one person in place of another.Footnote 60 Edwards answers that, in a sense, Christ is the same person as the sinner, and so it is just for God to punish him.Footnote 61

This is a very different use of the notion of participation compared to the model we have been developing. In that model, punishment plays no role: Christ's death is not the transference of the punishment due the sinner but the means to allow the sinner to escape her old, sinful identity and acquire a new one. Edwards shows that the concept of participation may be a useful element in a number of different possible soteriological models, not just the version of the PMA that we have been considering in this article, and it may help those models overcome common criticisms. If this is so, then the task of showing participation to be a meaningful and plausible notion is of lasting value to a wide range of theological thought about atonement.Footnote 62

Footnotes

1. The question of how these ‘models’ should be understood is a difficult one. A traditional view is that they are ‘theories’, offering competing explanations. Lewis (Reference Lewis1952), 53–55, for example, likens the doctrine of atonement to our knowledge that eating nourishes us, and the different models to rival theories of nutrition. Since the mid-twentieth century, there has been growing dissatisfaction with this understanding of the models as rival theories, with a tendency to regard them instead as complementary metaphors. For a brief overview of this trend, see Sykes (Reference Sykes1997), 6–10, and for an influential argument for the ‘metaphor’ approach, see Gunton (Reference Gunton1988), 27–52. I do not want to address this issue here, except to say that I use the terminology of ‘models’ for its neutrality on this question, and that I assume that any reasonable model can at least be interpreted as an explanation of how atonement happens, whether that explanation is regarded as literal, exhaustive, limited, or purely metaphorical.

2. The distinction between these two basic kinds of model, and the attribution of their originals to Anselm and Abelard respectively, goes back at least to Ritschl (Reference Ritschl1872), who devoted the first chapter of his examination of the doctrine to these two figures; see especially ibid., 22–24. Aulén (Reference Aulén1931), 17–20, claimed that this had become the ‘traditional account’, with all atonement theology being assimilated to one of these two models, which he called the ‘objective’ and ‘subjective’. In fact it is hard to find any such simplistic accounts in authors before Aulén, but Aulén's own influence led to a widespread assumption that this dual classification, and the terminology he used to describe it, were indeed standard. The classification continues to be found in more recent authors; see, for example, Fiddes (Reference Fiddes1989), 26–31 and Graham (Reference Graham2010), 133–135. The usefulness of the classification is controversial, though, given that virtually every model has both objective and subjective elements to some degree; however, this dispute does not greatly affect the argument in the present article.

3. For modern defences of broadly Anselmian models, see Stump (Reference Stump1988), Aspenson (Reference Aspenson1990), Brümmer (Reference Brümmer1992), and Brown (Reference Brown2006). For penal substitution, see Stott (Reference Stott1986), esp. 156–163, and Holmes (Reference Holmes2005). See also Tidball, Hilborn, and Thacker (Reference Tidball, Hilborn and Thacker2008), which contains both defences and criticisms of penal substitution from an evangelical viewpoint.

4. We can distinguish between two kinds of subjective model. ‘Exemplarist’ models hold that Christ's soteriological significance is solely as an example to believers to make changes to their lives in imitation of him, much as Martin Luther King may serve as an example to people inspired by his actions. What we may call ‘infusion’ models, by contrast, see the change in believers’ lives as deriving not (merely) from their own efforts but from God working directly on their hearts in some way. Peter Abelard, usually taken as the first proponent of a subjective model, has traditionally been interpreted as an exemplarist (e.g. by his contemporary Bernard of Clairvaux), but some writers have argued that he was actually an infusionist (Quinn Reference Quinn2009 [1993]). Fiddes (Reference Fiddes1989), 140–168 is a modern infusionist. Rashdall (Reference Rashdall1919), 435–464 gives a classic modern statement of exemplarism; Hick (Reference Hick1993), 127–133 and Weaver (Reference Weaver2001) can also be seen as exemplarists

5. Gunton (Reference Gunton1988) and Fiddes (Reference Fiddes1989) are good examples of the former approach. Aulén (Reference Aulén1931), Swinburne (Reference Swinburne1988), White (Reference White1991), and Mulcahy (Reference Mulcahy2007) can all be seen as examples of the latter. Aulén, in particular, was so influential that his ‘classic’ model of the atonement, often known as ‘Christus Victor’ after the title of his book, is sometimes regarded as a third major kind of model.

6. I do not intend what follows as an exposition of Paul's thought. Rather, I aim to describe and assess a model of atonement that draws on the ideas in Paul's texts. Whether Paul himself would have endorsed the model as presented here is a question that this article cannot answer. For discussions of ‘participation’ in Paul, see Sanders (Reference Sanders1977), 463–472; Dunn (Reference Dunn1998), 390–412; Hooker (Reference Hooker2000), 522. Bayne & Restall (Reference Bayne and Restall2009), 159 focus in particular on Hooker's exposition of Paul.

7. This does not necessarily mean that being sinful is essential to humans or part of human nature. Rather, it means that sin should be thought of as a state in which a person exists (perhaps contingently), rather than as a set of behaviours or an external force.

8. In this respect, the participatory model is similar to the ‘Christus Victor’ model – though whereas ‘Christus Victor’ models typically envisage sin as an external force that is defeated, here it is primarily an internal corruption that is escaped.

9. ‘Identity’ has different meanings. In one sense, which I shall call ‘self-identity’, ‘identity’ means the kind of person you are or how you regard yourself, including what groups you consider yourself to belong to. Cultural or ethnic ‘identity’ is of this kind, and when we speak here of a ‘sinful identity’ and a need for sinners to change their identity, this is the kind of identity that is meant. Numerical identity is something distinct and means simply being one and the same thing. The ambiguity is confusing because, on at least some versions of the PMA, the purpose of the atonement is to change sinners’ self-identity, but the mechanism by which it does so involves changing their numerical identity. Bayne and Restall rightly focus on the role of identity as central to the PMA but do not always clarify this ambiguity.

10. This could be as a result of the Christian's simply being inspired by Jesus’ example, in which case this model will parallel the ‘exemplarist’ subjective model; or it could be as a result of the Holy Spirit's action upon her enabling her to do so, in which case it will be like the ‘infusionist’ subjective model. See n. 4 above.

11. Note that Bayne and Restall present this as an objection to the PMA in general, rather than to the ‘subjective’ version of it discussed here; they do not explicitly consider such a version.

12. This version of the model is also similar to some modern restatements of the Anselmian ‘satisfaction’ model, in which atonement involves the sinner ‘making use of’ Christ's death as an offering to God. See, for example, Swinburne (Reference Swinburne1988), 27–29 and Stump (Reference Stump1988), 64–72. In the version of the PMA suggested here, the sinner similarly takes a conscious attitude towards Christ's death, but it is one that involves appropriating that death as an end to sin and change of identity rather than as an offering to God.

13. While the ‘subjective’ PMA sketched above is arguably of interest in itself, in the rest of this article I am concerned only with the possibility of an ‘objective’ PMA, and all subsequent references to the participatory model should be taken accordingly.

14. I say ‘on some level’ because there may be different kinds of numerical identity. Ordinarily the law of excluded middle applies to numerical identity – either x and y are identical or they are not – and one might think this should hold for Christ and the Christian too. The problem facing the proponent of an objective participatory model is how to find a way plausibly to deny this, and one possible solution is to argue that there are different kinds of numerical identity, and Christ and the Christian are identical in only some of them. On this, see later, under ‘An Edwardsean model of the PMA’.

15. Moreover, identity is symmetrical – so if the Christian is identical with Christ, Christ ought to share all of the Christian's properties, obviously an absurd situation. I am grateful to Mark Wynn for this point.

16. One possible answer to this question could rest on a strongly realist theory of universals. If (a) there is such a thing as human nature, and (b) every individual human instantiates human nature in its fullness, and (c) this instantiation is in some way an identity relation, then (assuming the transitivity of this identity relation) it follows that all human beings are in some way numerically identical with all other human beings, including Christ. Whether anyone has actually held such a theory of universals is a matter of dispute (arguably the most likely candidate is William of Champeaux, whose extremely realist theory of universals is known only from Abelard's Historia Calamitatum, ch. 2). A theory along these lines has sometimes been attributed to Gregory of Nyssa, although others interpret him as a collectivist (someone who thinks that a universal is identical with the sum of all its particulars, in which case he would reject (b) and probably (c) above). On Gregory and other patristic authors, see Cross (Reference Cross2002) and Zachhuber (Reference Zachhuber, Chiaradonna and Galluzzo2013). Whether or not they or other ancient Christian theologians explicitly held such a theory of universals, a theory of this kind seems to be implied by the model of atonement sometimes known as the ‘physical theory’ which has been attributed to the early church, according to which atonement is achieved by God transforming human nature itself through the incarnation. On this, see Kelly (Reference Kelly1978), 172–173, 185, 377–378, 381; Hart (Reference Hart, Hart and Thimell1989); Turner (Reference Turner2004), 67–74. I am grateful to Morwenna Ludlow for pointing out the relevance of this theory to the current topic. An obvious problem with such a model is that if the Christian is identical to Christ in this way, she is equally identical to Judas. For this reason, as well as the intrinsic implausibility of such a realist view of universals, I will not consider this possible answer further.

17. References to Edwards's works are to volume and page number of the Yale edition of The Works of Jonathan Edwards (WJE). In what follows, my primary aim is not the detailed exposition of Edwards but the identification of particular ideas in his works which may be useful in developing a version of the PMA – even if Edwards himself would not have accepted this model.

18. The details of Edwards's soteriology lie outside the scope of this article. Edwards typically uses language associated with a Calvinistic penal substitution model (e.g. WJE XXI 312; XIII 391, 401 – see Hamilton (Reference Hamilton2013), 396 n. 8 for a long list of citations) but also arguably gives hints towards the ‘governmental’ model, a variant on penal substitution, which was developed further by later theologians influenced by him, particularly his son, Jonathan Edwards Jr (e.g. WJE XIII 402). See McClymond and McDermott (Reference McClymond and McDermott2012), 249–251, and for an older but still useful survey, Rudisill (Reference Rudisill1971). Mark Hamilton (Reference Hamilton2013) discusses these different interpretations of Edwards, arguing that Edwards held penal substitution, and rejecting (Hamilton (Reference Hamilton2015), 56–67) the claim that held a governmental model. For a modern defence of the governmental model itself, see Crisp (Reference Crisp2008).

19. See Crisp (Reference Crisp2005, 96) and Holifield (Reference Holifield2006, 150). For more on this doctrine, sometimes known as ‘federalism’ (because Adam is the ‘federal head’ of humanity), see Rea (Reference Rea2007), 326–327. Rea argues, in my view rightly, that this doctrine is inconsistent with the common-sense intuition that people cannot be held morally responsible for situations over which they have had no control.

20. Edwards's view is thus a version of what is sometimes called ‘Augustinian realism’, i.e. the view that all of humanity is somehow present in, and directly accountable for, Adam's act of sinning. See Hamilton (Reference Hamilton2013), 399–402.

21. Edwards did not invent this metaphysics as a response to this problem; he had consistently asserted it throughout much of his career. See Crisp (Reference Crisp2012), 25.

22. For the connection between Edwards's view of time and his occasionalism, see Crisp (Reference Crisp2011b), 49–56.

23. Edwards talks about ‘moments’ without defining them, and in particular without stating whether they are temporally extended or not.

24. For a clear statement of four-dimensionalism, including its different varieties, see Sider (Reference Sider2001), 53–71.

25. On Edwards's view as a version of perdurantism, see Crisp (Reference Crisp2005), 104–110. Note in particular that Edwards's emphasis on the fact that each moment is a new creation arguably means that a four-dimensional individual is not really an individual at all, but an aggregate of temporally distinct individuals – which could be seen as a denial of perdurantism. See Crisp (Reference Crisp2005), 109–110. Rea (Reference Rea2007), 332–345 argues that Edwards may be interpreted in a way like this, as a sort of stage theorist rather than as a perdurantist. On stage theory in general, see Sider (Reference Sider2001), 188–208.

26. Note that God does not merely regard or treat me-at-t1 and me-at-t2 as identical, in a sort of legal fiction. On Edwards's view, God's decision means that I-at-t1 and I-at-t2 really are identical. See Crisp (Reference Crisp2005), 106–107.

27. For more on this position, see Helm (Reference Helm2003), 50–55 and Crisp (Reference Crisp2012) 25–26. Michael Rea has proposed a variant on this view known as ‘fission theory’, which builds on stage theory to suggest that one or more temporal stages could be shared by more than one four-dimensional individual. Thus, I might consist of all the stages between my birth and my death (as on conventional stage theory), plus those stages of Adam in which he sinned. One consequence of this is that if everyone has those stages of Adam in common, then for part of Adam's life he was identical with billions of individuals, before splitting into them. See Rea (Reference Rea2007), 341–345, and, for a critical discussion, Hudson (Reference Hudson2014), 55–77.

28. Roderick Chisholm argues that, on Edwards's view, identity over time is arbitrary, and there is nothing to stop God from regarding any set of objects as identical. See Chisholm (Reference Chisholm1976), 138–139, and, on Edwards's answer to this objection, Helm (Reference Helm1997), 169–171 and Crisp (Reference Crisp2005), 96–101.

29. On Edwards's use of Locke's views on personal identity, see Helm (Reference Helm2003). Edwards is also at pains to point out that if I-at-t2 resemble me-at-t1, this is only because God has seen fit to create me in that way (WJE III 398–399). Moreover, Edwards holds that the fact that psychological resemblance over time is a fit basis for identity over time is itself down to God's will (WJE III 399; Crisp (Reference Crisp2005), 99). An anonymous reader for Religious Studies points out that this makes Edwards's appeal to psychological resemblance self-defeating: if it all comes down to God's will, the Lockean theory of personal identity does very little work.

30. Elsewhere, notably in Miscellany 1263, Edwards argues that all laws depend on God's will, which is ‘tied to no other rules and laws but the directions of his own infinite understanding’ (WJE XXIII 202–203). Edwards himself articulates the Lockean theory only in the context of stressing the role of the ‘arbitrary constitution of the Creator’ (WJE III, 399). This suggests that the ‘analogy and harmony with other laws’ that Edwards refers to really refers only to consistency in God's decrees: God may regard two things as identical not by reference to some external standard of identity, but if doing so is consistent with other decisions he makes.

31. As Crisp (Reference Crisp2005), 101–104, points out, this is effectively a federalist strand to Edwards's thinking that accompanies his Augustinian realism.

32. Edwards derives the Divine Voluntarism Thesis from the Continuous Creation Thesis: because there cannot be any basis for identity over time in objects themselves, it must lie in God. But the Continuous Creation Thesis is a hard pill to swallow (it arguably undermines the doctrine of original sin itself; see Crisp (Reference Crisp2005), 130–133). Fortunately, it may be possible to hold at least a version of the Divine Voluntarism Thesis without the Continuous Creation Thesis. We can see this if we distinguish between two possible versions of the Divine Voluntarism Thesis. The Weak Divine Voluntarism Thesis holds that God can regard any two objects as identical, but this is not the only ground there may be for identity over time. The Strong Divine Voluntarism Thesis (which Edwards endorses) holds that the divine will is the only ground there can be for identity over time. The Strong Divine Voluntarism Thesis requires the Continuous Creation Thesis. One could not consistently hold both that objects endure (or perdure) from moment to moment without being recreated and that all identity over time depends directly on God's will. This is because if objects endure or perdure, that alone seems to be the basis for identity over time – further action by God is not required. But the Weak Divine Voluntarism Thesis does not require the Continuous Creation Thesis. A person could think that most objects do endure or perdure through time in some way that does not depend directly on God's will, but add to this the claim that in some cases there is identity over time as a direct result of God's will. (Edwards hints at the possibility of such a view himself, when he suggests that those who are not convinced by his metaphysics may still recognize God's right to regard Adam and his descendants as morally identical (WJE III, 409).) Such a view would require holding that there are different kinds of identity over time, some of which depend on God's direct decrees and some of which do not. This is problematic. If identity is not normally a matter of God's will, how can God sometimes ‘make’ things identical that ordinarily would not be? The Divine Voluntarism Thesis, then, requires metaphysical commitments that may be unpalatable. In its strong form it requires the Continuous Creation Thesis; in its weak form it requires the view that there are different kinds of identity over time.

33. Note that on this view, ‘divine nature’ and ‘human nature’ refer not to sets of properties but to concrete particulars. For this reason, this approach to incarnation is sometimes called ‘concretism’, following Plantinga (Reference Plantinga1999). Note also that I speak of Christ's human soul in addition to his body, because most defenders of compositionalism have spoken in these terms, but compositionalism does not require belief in a soul distinct from the body. A non-dualist compositionalism would be quite possible, in which the Son assumes a living, thinking, human body.

34. On the different ways in which the two natures could be related to each other, see Marmodoro & Hill (Reference Marmodoro and Hill2010).

35. For defences of models of this kind, see inter alia Leftow (Reference Leftow2002), (Reference Leftow2011); Stump (Reference Stump2002); Crisp (Reference Crisp2011a).

36. Even anti-Chalcedonian theologians, such as the Monophysite Severus of Antioch, would have agreed that Christ's human body and soul are not identical with his divine nature. Severus merely denied that his human body and soul should be called a ‘nature’. On Severus’ Christology, see Torrance (Reference Torrance1993).

37. For this analogy to be useful, we do not necessarily have to accept dualism as true, only regard it as conceptually possible.

38. The image goes back to the fourth-century theologian Nemesius of Emesa (Daley (Reference Daley2002), 177).

39. This view is found in John of Damascus; see De Fide Orthodoxa III.15 (Marmodoro & Hill (Reference Marmodoro and Hill2010), 481–482). Thomas Aquinas elaborates on it, arguing that Christ's human nature is like an organ, an internal instrument rather than an external one; see Summa Contra Gentiles IV.41.

40. Romans 12:4–5; 1 Corinthians 12:12–27.

41. If Christ uses me to accomplish his aims, I might legitimately be called his agent. Whether an agent can be said to be identical with the one who sends her, and if so in what way, is a fascinating question that is beyond the scope of this article. We may note, though, that some scholars, notably Peder Borgen, have argued that John's Gospel presents Jesus as God's agent, and that his identity as divine is grounded on precisely this relation. See Borgen (Reference Borgen and Ashton1986) as well as Loader (Reference Loader1992), 171 and Thompson (Reference Thompson2001), 230–238.

42. On the status of the Miscellanies and other unpublished works, compared to Edwards's writings intended for publication, see Crisp (Reference Crisp2012), 3–5.

43. For a philosophical overview of this, see Rea (Reference Rea2009). Note that Edwards's version of this doctrine departs from the traditional one in key respects. The usual view is that the individuating characteristics of the Persons are simply their dependence relations towards each other, e.g. the Son is begotten by the Father. Edwards goes beyond this, identifying other characteristics of the individual Persons that distinguish them from each other, though he conceives of these properties as playing a role within the life of the whole Trinity. For example, the Son is the divine Reason and the Spirit is the divine Love: the Father reasons with the Son and loves with the Spirit (WJE XXI, 132–34; Strobel (Reference Strobel2013), 51–58). Edwards seems inconsistent about whether these individuating properties count as relations. He states that they do at WJE XXI, 146, but he seems to distinguish them from ‘mere modes or relations’ at WJE XXI, 131. On this, see Crisp (Reference Crisp2012), 127–135 and Pauw (Reference Pauw2002), 73. Here, I assume for simplicity's sake that the individuating properties are all relations. This distances my account from Edwards's, but someone who holds an Edwardsean doctrine of the Trinity could easily adapt it to include other individuating properties.

44. Edwards did not complete the Discourse on the Trinity in a form ready for publication, or even give it a title. For its date, see WJE XXI, 109. A full examination of Edwards's Trinitarian theology, the nature of which has long been a matter of dispute, is well outside the scope of this article. For a discussion of the points considered here, see Crisp (Reference Crisp2012), 120–127.

45. Edwards's argument here presupposes a strong resemblance theory of representation.

46. The relations other than perfect similarity are vital here, since the Holy Spirit also perfectly resembles the Father, but bears other relations to him.

47. The role of the Holy Spirit in Edwards's Christology has been widely noted (Hastings (Reference Hastings2005); Withrow (Reference Withrow2011), 126–135). Withrow, following Stephen Holmes and Carl Trueman, notes (127) that the idea goes back to Didymus the Blind, via John Owen.

48. This similarity between the union of human and divine in Christ and the union of believers to the divine – and the union of the divine Persons in the Trinity – is a central theme in Edwards's theology (Hastings (Reference Hastings2005), 282–284; Hamilton (Reference Hamilton2013), 404–412). Edwards himself draws attention to the similarity, in his view, between Trinity, incarnation, and sanctification at WJE XIII, 330; see also WJE XIII, 448.

49. By ‘psychologically exactly similar’ I am thinking in terms of dispositions rather than of experiences. That is, we might say that Jesus’ behaviour showed exactly the same personality as the Logos, rather than that Jesus and the Logos underwent precisely the same experiences. For example, if (per impossibile?) Satan had tempted the Logos, he would have responded as Jesus did, but this is not to say that the Logos straightforwardly did experience temptation despite his perfect union to Jesus. I am grateful to Mark Wynn for raising this issue.

50. This may seem similar to the concept of ‘relative identity’, according to which identity is sortal: some a may be the same F but not the same G as some b. On this view, there can be a sense in which a and b are identical and another sense in which they are not. But the distinction between identityo and identityd is different: rather than all identity being sortal, we are suggesting that there could be different kinds of identity altogether, because of the fundamental ontological difference between God and creatures. On relative identity in general, see Geach (Reference Geach1967–1968), 3–11 and Griffin (Reference Griffin1977). Peter van Inwagen has used this concept to argue for the coherence of the Trinity (van Inwagen (Reference van Inwagen1988) ) and the incarnation (van Inwagen (Reference van Inwagen1994) ).

51. God would still be the ultimate cause of the Christian's being identicald to the Son, because he would be directly or indirectly responsible for the fact that the Christian bears the relations in virtue of which she is identicald to the Son (I am grateful to Oliver Crisp for raising this point). But this would not be a matter of God directly making the Christian identicald to the Son in a more strongly Edwardsean manner.

52. On this model, the logic of salvation is the reverse of that found in most subjective models. Whether examplarist or infusionist, these models normally see a more moral lifestyle (however conceived) as the goal of the process of salvation. On the model suggested here, a more moral lifestyle is certainly part of the process, but the goal is to acquire not Christ's moral properties but those that permit escaping sin.

53. There is a wide literature on this topic, but good starting points are Russell (Reference Russell2004) and Finlan & Kharmalov (Reference Finlan and Kharmalov2010).

54. It is perhaps no coincidence that Jonathan Edwards is widely recognized as an unusually prominent western proponent of the doctrine of theōsis (McClymond (Reference McClymond2003); Strobel (Reference Strobel2013), 192–207).

55. The connection of the doctrine of theōsis to that of the incarnation is well established, and has been at least since Athanasius’ famous claim that ‘God became man so that man might become God’, meaning that the doctrine of incarnation must be true because the doctrine of theōsis is. On Athanasius’ argument, see Finch (Reference Finch2010). But while incarnation is closely linked to theōsis in traditional Orthodox thought, there is not always a clear explanation of how that link works: why does the Son's becoming human allow human beings to become divine? The PMA, in the form sketched here, gives a possible answer.

56. This text was written some time in 1727 (WJE XIII, 19–20). See also XIII, 429; XVIII, 52–53; XX, 466–467, the last of which was written in the mid- to late-1740s (WJE XX, 38). Edwards himself cross-referenced these entries together, indicating that these ideas were important to him over a number of years.

57. Elsewhere, he considers other benefits – besides salvation – that he thinks derive from union with Christ. See e.g. WJE XIII, 304.

58. This was written in 1729 (WJE XIII, 27–28), not long before the text on the incarnation discussed above.

59. Edwards argues in a sermon of 1729 that the elect are one with Christ in the same way as sinners are one with Adam (WJE XIV, 404–405). Mark Hamilton has shown that, consequently, participation is central to Edwards's doctrine of atonement (Hamilton (Reference Hamilton2013), 402–404).

60. See, e.g., Stump (Reference Stump1988), 62–63, and Graham (Reference Graham2010), 131–132. Holmes (Reference Holmes2005), 119–122 seeks to answer the objection, while Lewis (Reference Lewis2009 [1997]) argues for the coherence of the transference of punishments in general.

61. This is the version of penal substitution that Mark Hamilton ascribes to Edwards, though he notes that the evidence for it in Edwards is ‘only suggestive and not definitive’ (Hamilton (Reference Hamilton2013), 412). Drawing on ideas from Karl Barth, Oliver Crisp has also developed a version of penal substitution much like this, where

Christ and the elect together compose one metaphysical entity that persists through time’ and ‘the punishment due for the sin of the derivatively elect is transferred to Christ . . . Although he is not the one who has sinned . . . because he is a member of this larger entity, he may pay the consequences of the sin of other members of the same entity. (Crisp (Reference Crisp2009b), 440–441)

Crisp's version is slightly different from what Edwards suggests. Edwards seems to envisage that Christ and the elect are ‘the same’, which I take to mean identical in some sense. Crisp suggests that Christ and the elect are distinct parts of a greater whole. This is more readily comprehensible but raises questions: in virtue of what do they compose a greater whole that is more than just a mereological sum? How can the punishment for sin committed by one part be justly inflicted on another part, even if they are both parts of the same greater whole? On problems of this kind, see Rea (Reference Rea2007), 329–332. Crisp considers responses to them (Crisp (Reference Crisp2009b), 443–447), but they do not arise at all with the Edwardsean version sketched here.

62. I am grateful to Oliver Crisp, Morwenna Ludlow, and Mark Wynn for their helpful comments on earlier versions of this article, as well as to members of the Centre for Philosophy of Religion at Leeds University, where a version was presented in October 2015.

References

Aspenson, S. (1990) ‘In defense of Anselm’, History of Philosophy Quarterly, 7, 33–45.Google Scholar
Aulén, G. (1931) Christus Victor (London: SPCK).Google Scholar
Bayne, T. & Restall, G. (2009) ‘A participatory model of the atonement’, in Nagasawa & Wielenberg (2009), 150–166.Google Scholar
Borgen, P. (1986) ‘God's agent in the Fourth Gospel’, in Ashton, J. (ed.) The Interpretation of John (Philadelphia PA: Fortress), 67–78.Google Scholar
Brown, D. (2006) ‘Anselm on atonement’, in Davies & Leftow (2006), 279–302.Google Scholar
Brümmer, V. (1992) ‘Atonement and reconciliation’, Religious Studies, 28, 435–452.CrossRefGoogle Scholar
Chisholm, R. (1976) Person and Object (London: Allen & Unwin).Google Scholar
Crisp, O. (2005) Jonathan Edwards and the Metaphysics of Sin (Aldershot: Ashgate).Google Scholar
Crisp, O. (2008) ‘Penal non-substitution’, Journal of Theological Studies, 59, 140–168.CrossRefGoogle Scholar
Crisp, O. (2009a) A Reader in Contemporary Philosophical Theology (London & New York: Clark).Google Scholar
Crisp, O. (2009b) ‘Original sin and atonement’, in Flint & Rea (2009), 430–451.CrossRefGoogle Scholar
Crisp, O. (2011a) ‘Compositional christology without Nestorianism’, in Marmodoro & Hill (2011), 45–66.CrossRefGoogle Scholar
Crisp, O. (2011b) Revisioning Christology: Theology in the Reformed Tradition (Aldershot: Ashgate).Google Scholar
Crisp, O. (2012) Jonathan Edwards on God and Creation (Oxford: Oxford University Press).CrossRefGoogle Scholar
Cross, R. (2002) ‘Gregory of Nyssa on universals’, Vigiliae Christianae, 56, 372–410.CrossRefGoogle Scholar
Daley, B. (2002) ‘Nature and the “mode of union”: late patristic models for the personal unity of Christ’, in Davis, Kendall, & O'Collins (2002), 164–196.Google Scholar
Davies, B. & Leftow, B. (eds) (2006) The Cambridge Companion to Anselm (Cambridge: Cambridge University Press).Google Scholar
Davis, S., Kendall, S., & O'Collins, G. (eds) (2002) The Incarnation: An Interdisciplinary Symposium on the Incarnation of the Son of God (Oxford: Oxford University Press).CrossRefGoogle Scholar
Dunn, J. (1998) The Theology of Paul the Apostle (Edinburgh: Clark).Google Scholar
Edwards, J. (1957–2008) The Works of Jonathan Edwards, 73 vols, various eds (New Haven CT: Yale University Press).Google Scholar
Fiddes, P. (1989) Past Event and Present Salvation (London: Darton, Longman & Todd).Google Scholar
Finch, J. (2010) ‘Athanasius on the deifying work of the Redeemer’, in Finlan and Kharmalov (2010), 104–121.CrossRefGoogle Scholar
Finlan, S. & Kharmalov, V. (eds) (2010) Theosis: Deification in Christian Theology (Cambridge: Clarke).CrossRefGoogle Scholar
Flint, T. & Rea, M. (eds) (2009) The Oxford Handbook of Philosophical Theology (Oxford: Oxford University Press).CrossRefGoogle Scholar
Geach, P. (1967–1968) ‘Identity’, Review of Metaphysics, 21, 3–12.Google Scholar
Graham, G. (2010) ‘Atonement’, in Taliaferro & Meister (2010), 124–135.Google Scholar
Griffin, N. (1977) Relative Identity (Oxford: Clarendon).Google Scholar
Gunton, C. (1988) The Actuality of Atonement (Edinburgh: Clark).CrossRefGoogle Scholar
Hamilton, S. M. (2013) ‘Jonathan Edwards on the atonement’, International Journal of Systematic Theology, 15, 394–415.CrossRefGoogle Scholar
Hamilton, S. M. (2015) ‘Jonathan Edwards, Anselmic satisfaction and God's moral government’, International Journal of Systematic Theology, 17, 46–67.CrossRefGoogle Scholar
Hart, T. (1989) ‘Irenaeus, recapitulation and physical redemption’, in Hart, T. & Thimell, D. (eds) Christ in our Place: The Humanity of God in Christ for the Reconciliation of the World (Exeter: Paternoster), 152–181.Google Scholar
Hastings, A., Mason, A., & Pyper, H. (eds) (2000) The Oxford Companion to Christian Thought (Oxford: Oxford University Press).Google Scholar
Hastings, W. R. (2005) ‘“Honouring the Spirit”: analysis and evaluation of Jonathan Edwards’ pneumatological doctrine of the incarnation’, International Journal of Systematic Theology, 7, 279–299.CrossRefGoogle Scholar
Helm, P. (1997) Faith and Understanding (Edinburgh: Edinburgh University Press).Google Scholar
Helm, P. (2003) ‘A forensic dilemma: John Locke and Jonathan Edwards on personal identity’, in Helm & Crisp (2003), 45–59.Google Scholar
Helm, P. & Crisp, O. (eds) (2003) Jonathan Edwards: Philosophical Theologian (Aldershot: Ashgate).Google Scholar
Hick, J. (1993) The Metaphor of God Incarnate (London: SCM).Google Scholar
Holifield, E. B. (2006) ‘Edwards as theologian’, in Stein (2006), 144–161.CrossRefGoogle Scholar
Holmes, S. (2005) ‘Can punishment bring peace? Penal substitution revisited’, Scottish Journal of Theology, 58, 104–123.CrossRefGoogle Scholar
Hooker, M. (2000) ‘Paul’, in Hastings, Mason, & Pyper (2000), 521–524.Google Scholar
Hudson, H. (2014) The Fall and Hypertime (Oxford: Oxford University Press).CrossRefGoogle Scholar
Kelly, J. N. D. (1978) Early Christian Doctrines, rev. ed. (New York: Harper & Row).Google Scholar
Leftow, B. (2002) ‘A timeless God incarnate’, in Davis, Kendall, & O'Collins (2002), 273–299.Google Scholar
Leftow, B. (2011) ‘The humanity of God’ in Marmodoro & Hill (2011), 20–44.CrossRefGoogle Scholar
Lewis, C. S. (1952) Mere Christianity (London: Bles).Google Scholar
Lewis, D. (2009 [1997]) ‘Do we believe in penal substitution?’, in Crisp (2009a), 328–334.Google Scholar
Loader, W. (1992) The Christology of the Fourth Gospel: Structure and Issues (Frankfurt am Main; Bern; New York; Paris: Lang).Google Scholar
Marmodoro, A. & Hill, J. (2010) ‘Composition models of the incarnation: unity and unifying relations’, Religious Studies, 46, 469–488.CrossRefGoogle Scholar
Marmodoro, A. & Hill, J. (eds) (2011) The Metaphysics of the Incarnation (Oxford: Oxford University Press).CrossRefGoogle Scholar
McClymond, M. (2003) ‘Salvation as divinization: Jonathan Edwards, Gregory Palamas and the theological uses of Neoplatonism’, in Helm & Crisp (2003), 139–160.Google Scholar
McClymond, M. & McDermott, G. (2012) The Theology of Jonathan Edwards (Oxford: Oxford University Press).Google Scholar
Morris, T. (ed.) (1988) Philosophy and the Christian Faith (Notre Dame IN: University of Notre Dame Press).Google Scholar
Mulcahy, E. (2007) The Cause of Our Salvation (Rome: Gregorian University).Google Scholar
Nagasawa, Y. & Wielenberg, E. (eds) (2009) New Waves in Philosophy of Religion (Basingstoke: Palgrave Macmillan).Google Scholar
Padgett, A. (ed.) (1994) Reason and the Christian Religion: Essays in Honour of Richard Swinburne (Oxford: Clarendon).Google Scholar
Pauw, A. (2002) The Supreme Harmony of All: the Trinitarian Theology of Jonathan Edwards (Grand Rapids MI: Eerdmans).Google Scholar
Plantinga, A. (1999) ‘On heresy, mind and truth’, Faith and Philosophy, 16, 182–193.Google Scholar
Quinn, P. (2009 [1993]) ‘Abelard on atonement: “Nothing unintelligible, arbitrary, illogical, or immoral about it” ’, in Crisp (2009a), 335–353.Google Scholar
Rashdall, H. (1919) The Idea of Atonement in Christian Theology (London: Macmillan).Google Scholar
Rea, M. (2007) ‘The metaphysics of original sin’, in van Inwagen & Zimmerman (2007), 319–356.Google Scholar
Rea, M. (2009) ‘The Trinity’, in Flint & Rea (2009), 689–735.Google Scholar
Ritschl, A. (1872) A Critical History of the Christian Doctrine of Justification and Reconciliation (London: Hamilton & Adams).Google Scholar
Rudisill, D. (1971) The Doctrine of the Atonement in Jonathan Edwards and his Successors (New York: Poseidon).Google Scholar
Russell, N. (2004) The Doctrine of Deification in the Greek Patristic Tradition (Oxford: Oxford University Press).Google Scholar
Sanders, E. P. (1977) Paul and Palestinian Judaism (London: SCM).Google Scholar
Sider, T. (2001) Four-Dimensionalism: An Ontology of Persistence and Time (Oxford: Oxford University Press).CrossRefGoogle Scholar
Stein, S. (ed.) (2006) The Cambridge Companion to Jonathan Edwards (Cambridge: Cambridge University Press).CrossRefGoogle Scholar
Stott, J. (1986) The Cross of Christ (Downers Grove IL: IVP).Google Scholar
Strobel, K. (2013) Jonathan Edwards's Theology: A Reinterpretation (London: Clark).Google Scholar
Stump, E. (1988) ‘Atonement according to Aquinas’, in Morris (1988), 61–91.Google Scholar
Stump, E. (2002) ‘Aquinas’ metaphysics of the incarnation’, in Davis, Kendall, & O'Collins (2002), 197–218.Google Scholar
Swinburne, R. (1988) ‘The Christian scheme of salvation’, in Morris (1988), 15–30.Google Scholar
Sykes, S. (1997) The Story of Atonement (London: Darton, Longman, & Todd).Google Scholar
Taliaferro, C. & Meister, C. (eds) The Cambridge Companion to Christian Philosophical Theology (Cambridge: Cambridge University Press).CrossRefGoogle Scholar
Tanner, N. (ed.) (1990) Decrees of the Ecumenical Councils (London: Sheed & Ward; Washington DC: Georgetown University Press).Google Scholar
Thompson, M. (2001) The God of the Gospel of John (Grand Rapids MI: Eerdmans).Google Scholar
Tidball, D., Hilborn, D., & Thacker, J. (eds) (2008) The Atonement Debate (Grand Rapids MI: Zondervan).Google Scholar
Torrance, I. (1993) Christology after Chalcedon: Severus of Antioch and Sergius the Monophysite (Norwich: Canterbury).Google Scholar
Turner, H. (2004) The Patristic Doctrine of Redemption (London: Wipf & Stock).Google Scholar
van Inwagen, P. (1988) ‘And yet they are not three Gods but one God’, in Morris (1988), 241–278.Google Scholar
van Inwagen, P. (1994) ‘Not by confusion of substance, but by unity of person’, in Padgett (1994), 201–226.CrossRefGoogle Scholar
van Inwagen, P. & Zimmerman, D. (eds) (2007) Persons: Human and Divine (Oxford: Oxford University Press).CrossRefGoogle Scholar
Weaver, J. D. (2001) The Nonviolent Atonement (Grand Rapids MI: Eerdmans).Google Scholar
White, V. (1991) Atonement and Incarnation (Cambridge: Cambridge University Press).CrossRefGoogle Scholar
Withrow, B. (2011) Becoming Divine: Jonathan Edwards's Incarnational Spirituality within the Christian Tradition (Eugene OR: Wipf & Stock).Google Scholar
Zachhuber, J. (2013) ‘Universals in the Greek church fathers’, in Chiaradonna, R. & Galluzzo, G. (eds) Universals in Ancient Philosophy (Pisa: Edizione della Normale), 425–470.Google Scholar