Triadic Differences and Theological Coherence: Oliver O’Donovan's Reflections on Friendship as a Locus for Comparing Resurrection and Moral Order and Ethics as Theology

This article leverages the theme of friendship in Oliver O’Donovan's Entering into Rest as a locus of comparison between his earlier Resurrection and Moral Order and the Ethics as Theology trilogy. It does so by using demonstrable methodological differences between the two moral-theological projects to illumine a fundamental theological coherence. The article pursues this task in five sections. The first expounds O’Donovan's reflection on friendship in Entering into Rest. The second articulates the triadic approach adopted in these reflections. The third examines a similar triad articulated by O’Donovan in Resurrection and Moral Order. The fourth uses this triadic connection to demonstrate theological coherence between Resurrection and Moral Order and the Ethics as Theology trilogy. The fifth concludes by raising a critical query about the arbitrariness of the triadic methodology employed by O’Donovan. The article’s conclusion considers more generally—in light of O’Donovan's pursuit of ethics as theology—the inherent contingency and incompleteness of theological ethics.


Introduction
Oliver O'Donovan's Ethics as Theology (ET) trilogy-and especially the final volume Entering into Rest (EIR)-has to date received limited attention compared to his influential work Resurrection and Moral Order (RMO) (with notable exceptions). 1 This article makes a small contribution to this underdeveloped discourse by firstly pursuing a prima facie parochial task: expounding O'Donovan's reflections on friendship in the chapter 'The Communication of Friendship' from EIR. 2 It is via this initially narrowed focus that I intend to make some more general observations about O'Donovan's moral theology. For the moral minutiae of EIR unveils some crucial continuities and discontinuities between RMO and the ET project. It illumines deliberate methodological differences between RMO and EIR-especially concerning the moral appraisal of specific themes-that in turn demonstrate profound theological coherence across O'Donovan's moral-theological corpus. This article pursues this task in five sections. The first section expounds O'Donovan on friendship in EIR. The second section articulates the triadic methodology adopted in these reflections. The third section then examines a similar triad in RMO. The fourth section leverages this triadic methodological connection between RMO and EIR to demonstrate the fundamental theological coherence of RMO and ET. The fifth and final section will then conclude by raising a critical query concerning the arbitrariness of his triadic methodologies. By way of conclusion the insights that can be gleaned from O'Donovan about the pursuit of theological ethics more generally-in particular the inescapable contingency and incompleteness of faith seeking moral understandingwill be considered.

Friendship
In this first section I intend to expound O'Donovan's treatment of friendship in EIR. The chapter 'The Communication of Friendship' begins with a reflection on coexistence in general. Human nature is 'created social', he argues. 3 But harmonious coexistence is not itself friendship. For there are several social encounters in which we expect civility 1 but not 'continuing connection' 4 -we envisage friendliness but not friendship. Here O'Donovan draws on the work of Jean-Yves Lacoste. For Lacoste, friendly coexistence has moral valence: 'in even the ordinary passing encounter of those who co-exist (unvirtuously) in peace there is a kind of "promise"'. 5 This 'promise'-to conduct interactions civilly-is to both reveal and live in accordance with our social nature. The neighbour is the 'fulfillment of the self … God offers us our self, God offers us our neighbour, and the gifts may be distinguishable, but not separable'. 6 'Friendship', though, 'is distinguished from all these polite or friendly encounters precisely by its ambitions to endure as a particular relation'. 7 This feature exposes the inherent limitedness of friendship; we cannot sustain enduring friendships with everyone. This same observation has been made in influential writings like Cicero's Laelius and Aristotle's Nicomachean Ethics-our capacity for friendship is restricted, therefore we must choose our friends wisely. 8 This comes with 'a sense of tragedy [that] hangs over the classical discourse of friendship', a tension between the ideas of 'perfection and universality'. 9 Next, O'Donovan searches for particularly Christian insights on these issues. Augustine and Ambrose both indicated they were 'impressed enough' with the reflections of Cicero and Aristotle. 10 But most crucial for O'Donovan is Aelred of Rievaulx's review of Cicero's De amicitia. 11 For Aelred successfully holds together universality and perfection. He provides a way forward by reconciling the expanding social tendencies of humans with the exclusivity of good friendship. He does so by: drawing friendship into a Christian understanding of pilgrimage, growth, and final attainment, allowing its beginning, development, and perfection to contribute to moral and spiritual growth, and especially associating friendship with wisdom. 12 Humans sojourning towards the eschaton are free to find as many friends as possible along the way. People can welcome new friendships whilst deepening old ones. This kind of move situates friendship in love (caritas). 13 It is a means of moral formation, the perfecting of devotion, through the acts of giving and forgiving. O'Donovan enters this discourse with friendliness: To be friendly is to conduct an encounter with no prospect of familiarity beyond its conclusion, but to conduct it in a manner that might be appropriate if it did have such a prospect … in being friendly to the [stranger] we say, in effect, 'I am not your friend, nor you mine, nor in this life are we ever likely to become friends; yet in God's eternity, and even in this life if it should so transpire, a friendship between us will be no bad thing'. 14 This openness to relational endurance shapes our friendly posture to our barista or the tradesperson visiting our home.
But there is also a dark-sided complexity to friendship. Friendship can be tarnished in at least two ways: by denying the possibility of 'forming' new friendships or failure in 'sustaining' existing friendships. 15 These two forms of denial, namely exclusion and betrayal, are the 'two complaints that echo through the Psalter'. 16 O'Donovan first examines exclusion. He argues 'a certain tension between old and new friends is natural enough', and that loyalty in friendship is virtuous. 17 But an excess exclusivity is destructive. Friendship 'cannot bear the weight of too much identity-conferral', and so we must be open to new friendships. 18 Loyal friendship must not descend into destructive exclusivity and moral tunnel vision.
O'Donovan secondly reads betrayal 'as an extension of the logic of competition into a general condition of society'. 19 Hostility becomes embedded in our modes of relating: 'being-against is more fundamental than being-with, and can always trump friendship and set it aside'. 20 But 'friendship in a world of rivalry … can be no more than a provisional truce'. 21 Perhaps an openness to new friendships remains, but 'beginnings have no way of growing to become serious or lasting'. 22 This, given our inherent sociality, is ultimately an 'abandonment of ourselves'-betrayal is self-betrayal. 23 O'Donovan begins to draw the chapter to its conclusion with a question: 'Is there to be found, then, a universal that can embrace the perfecting influence of friendship?' 24 Christians have typically turned to the friendship Christ offers to his disciples to answer this question. 25 This, to be clear, is more than mere friendship with God, a concept not unfamiliar in pagan antiquity. Christians had in mind something more concrete: 'friendship with Christ, however, is a moment of fulfilment, a concrete, material encounter'. 26  Friendship with Christ is not a major feature of the New Testament. But, O'Donovan argues, this theme emerges at an important structural moment in the Gospel of John (15:1-25). Jesus suspends his teaching on the Spirit to reflect 'upon the permanent relation of the disciples to his own person'. 28 His absence in exaltation will not be a total absence. There will remain a sustained and vital friendship-something like a vine, its branches, and viticulturalist-'constituted by joyful adherence to Jesus' teaching, and supremely to the command of mutual love'. 29 How, though, 'may we know the friendship of Jesus' given his exaltation? 30 It is by attending to the historic resurrection, the community it created, and the future parousia that we encounter in Christ. 'The present friendship, mediated by the Holy Spirit, a gift rooted in the past of Pentecost and in the future coming of which it is a pledge, is a real presence, but in absence'. 31 The Spirit speaks to us of Christ in history and the church-'where two or three are gathered together in my name, there am I in the midst' (Matt. 18:20). 32 In the congregation, then, friendship finds fulfilment. 'To be friends with the risen Jesus is to have the same friends as he'-philadelphia. 33 This is 'the basis from which purer and more intense affection may be developed'. 34 Moreover, this 'brotherly' group is characteristically hospitable, serving and welcoming the outsider. O'Donovan is seeking to resolve the issues of loyalty and exclusion raised earlier.
This outward oriented vocation can only be discharged amongst a people who 'make the congregation visible and local'. 35 O'Donovan is claiming that hospitable philadelphia necessitates visibility in the world. This leads to a discussion regarding the visibility of the church in history. If the Spirit does indeed speak to us of Christ 'in the church', 36 then where in history shall we look for the Bride of the Lamb? O'Donovan thus proceeds with a reflection on the value of the church as a visible institution: Looking for the community of Christ's friends is not simply a matter of looking there, at the institutional structure, but of looking through the institutional structure to discern what God has given it the power to reveal … because what it shows, the reality of the friendship of Christ, is not natural to its structure and conduct, but looks eschatologically beyond it. 37 O'Donovan acknowledges that this kind of institutional discernment is often difficult. Institutional forms of the church have at times hidden the Bride of Christ. 38  'problem is not a matter of the church's existence, but of its visibility'. 39 Therefore, 'our response can only be to look for it, which means, to let our own love for Christ and for his followers lead us to where we find him and them present'. 40 This search will appropriately begin with historical institutions, but will also ultimately go beyond them. The 'body of disciples is found wherever it is revealed to us by its confession of its Lord, by its adherence to his teaching and obedience to his universal mission'. 41 This is where the community of Christ's friends are to be found.

The Triadic Approach to Ethics in Entering into Rest
In this second section I intend to examine the approach or method employed by O'Donovan in this ethical survey of friendship. This will ultimately serve as a helpful vantage point to examine some broader matters concerning his moral theology. The triadic approach used here is explicitly articulated at an important juncture in EIR where the text pivots from broader theological considerations towards the examination of specific spheres of human life, the objects of 'moral thinking' (e.g., friendship, work). 42 O'Donovan articulates his approach: 'beginning in each case from the created human existence that gives rise to them, and then proceeding, by way of the wrongs which they expose to us, to the work of God sanctifying them, for service, assembly, and witness'. 43 He begins with creation, moves through sin, and ends in an ecclesiologically oriented sanctification.
This triadic pattern emerges in the chapter 'The Communication of Friendship' in EIR. O'Donovan begins with 'created human existence' by providing what Sarah Coakley calls a 'rich phenomenology of moral existence' and sociality. 44 He proceeds through friendship 'by way of the wrongs which they expose to us' 45 or-as Philip Lorish and Charles Matthews put it-'attuned to the complexities and paradoxes of a sin-riddled creature' like exclusion and betrayal. 46 He finally concludes with 'the work of God sanctifying them, for service, assembly, and witness'. 47 This twofold task consists of examining, firstly, the perfection of friendship in Christ and, secondly, its ecclesiological implications.
Even a cursory survey demonstrates that this same method is employed in other chapters such as ' creaturely contours of work: 'work as material labor, work as social contribution, and work as personal self-realization'. 48 He then examines the impact of sin on work: it can damage our 'creative relation to the material world', 49 our 'social relations to others', 50 and any 'meaningful justification for the efforts it exacts for us'. 51 Finally, O'Donovan examines its sanctified end in the justifying and sanctifying work of Christ and its subsequent call to service. 52 When examining specific ethical themes O'Donovan begins with creation as experienced; moves through the sin the specific subject illuminates; and ends in sanctification, emphasising perfection in Christ and its ecclesiological implications.

The Triadic Approach of Resurrection and Moral Order
The triadic approach of EIR serves as an illuminating point of comparison with O'Donovan's foundational RMO. The most helpful place to begin in RMO is the prologue to the second edition where a similar triad is found. 53 Here one can observe some methodological differences that in the end shed some light on the theological congruences (and differences) of RMO and ET.
In this prologue O'Donovan argues that the primary concern of RMO was to overcome any dichotomy between creation and kingdom ethics. The central thesis of the text is 'in the resurrection of Christ, where creation is restored and fulfillment promised, ethics had a foundation which embraced the partial truths of both these points of view'. 54 He thenmost interestingly for our purposes-proceeds to outline his approach to or method of reading specific subjects in ethics whilst gesturing towards some underlying influences.
He first cites Karl Barth's Ethics. 55 Barth's governing principle was that 'ethics must conform to the shape of salvation-history, and so has a threefold pattern corresponding to creation, reconciliation and redemption: an ethics of life, an ethics of law, and an ethics of promise'. 56 O'Donovan also cites James McClendon's Ethics. 57 McClendon proposed a similar approach: 'ethics covers "the sphere of the organic", "the sphere of the communal" and "the sphere of the anastatic"-which means we are "part of the natural order", "part of a social world" and "part of an eschatological realm"'. 58 O'Donovan notes there were 'some instructive difficulties' in the application of Barth  headings. But the allocation of subjects can be problematic. The subject of war is allocated under the third, eschatological heading for McClendon, but under the first, creational heading for Barth. 60 Friendship fell under creation for Barth, 61 but the anastatic for McClendon. 62 O'Donovan proposes a corrective solution to this arbitrariness: [T]he organization of ethics into creation, reconciliation, and eschaton cannot provide a selfevident principle for arranging the specific subject areas that ethics interests itself in … Each area has to be given, as it were, a salvation-history of its own. 63 The threefold governing principle ought not subsume subjects but be brought to bear on them all.
There are some prima facie similarities between this approach in RMO and EIR. The formulas employed seem to be purposefully triadic. Their application is consistent-both avoid the errors of Barth and McClendon and are brought to bear on a subject in their entirety. And the triads themselves have some familiar categories. For example, they both begin with an emphasis on creation; and there is overlap between eschatology and sanctification in the third pillar of the triad.
But further inspection also reveals amongst these similarities significant differences. Eschatology in the final feature of the triad in RMO is replaced in EIR by the language of sanctification and ecclesiological categories of 'service, assembly, and witness'. 64 Creation in RMO pertains to primal, objective realities gifted to humanity from the beginning. Marriage and work are a 'gift of creation', for example. 65 But in EIR creation concerns 'human existence' as experienced. 66 But the most obvious difference is the second part of the triad. In RMO it is called 'reconciliation' but in EIR the focus of the second pillar is firmly on sin and 'the wrongs' illuminated by a subject. 67 How does one account for these triadic differences? It could be postulated that these differences are merely a by-product of theological development: they demonstrate a developed inconsistency and incongruence between the earlier RMO and the later ET. This interpretation of development-as we will see further below-is similar to what Samuel Tranter suggests. 68 And O'Donovan himself says of ET: 'I am conscious that my later thinking about Theological Ethics has progressed since I laid the foundations for it in Resurrection and Moral Order'. 69 But the full picture of this development is more subtle than a wholesale change in theological outlook or methodological approach.
O'Donovan argues that ET is a 'necessary complement' to RMO. 70 The preface of Self, World, and Time articulates the essential goal of the project: to further account for the Spirit in his theological ethics. 71 He aims to examine implications for the Spirit on the 'forceful moral objectivism' of RMO. 72 It is an 'Ethics after Pentecost'. 73 This means the 'angle of vision [is] turned principally towards the subjective renewal of agency'. 74 Hence O'Donovan's theological ethics 'takes a step closer to the point of view of the deliberating agent … reflecting from the inside'. 75 It concerns 'Pentecost and Moral Agency'. 76 Andrew Errington has argued this shift in perspective also corresponds to greater attention to practical reason in the project. 77 He observes that the categories of self, world, and time are the 'three key elements of natural moral reasoning' 78 -in Coakley's terms, the 'three core phenomenological categories' central to the project. 79 These categories correspond to the virtues of faith, love, and hope, the faculties of practical reason sanctified that are-in Jennifer Herdt's language-'taken up in the graced redemption of the moral life'. 80 The Theological Congruence of RMO and ET It is this shift in emphasis-a change in the 'angle of vision' 81 -that best accounts for the triadic methodological differences examined in this article. In this fourth section I intend to push this argument further. I will bring each pillar of the triad in EIR-creation, sin, and sanctification-into contact with the RMO. In doing so I intend to demonstrate that each pillar of this triad-rather than being inconsistent with RMO-profoundly coheres and indeed finds substantive theological grounding in RMO. The triadic differences traced in this article somewhat counterintuitively illuminate what Errington has called the 'impressively unified picture' of RMO and ET. 82

Triadic Pillar I: Creation
The created order and the moral field have a central role in RMO. O'Donovan argues 'the order of things that God has made is there. It is objective, and mankind has a place within in it'. 83 Humanity is called to live in accordance with this objective reality-'the way the universe is, determines how man ought to behave himself in it'. 84 In other words, one must live congruently within the world's 'complex network of teleological and generic relations'. 85 Moreover, the central claim of RMO is that the resurrection of Jesus Christ has vindicated this same created order. The resurrection of the material human body of Jesus Christ was a reaffirmation of created reality, an elimination of the possibility that in Adam, 'creation was a lost cause'. 86 This vindication of creation has important implications for moral agency, for the possibility of deliberating the 'moral field' as Errington states it. 87 For in a redeemed cosmos human agents encounter the world as 'intelligible'. 88 This is further elaborated with the concept of wisdom. 89 O'Donovan argues: Wisdom liberates us from the persistent fear of that unutterable and unknowable uniqueness by enabling us to interpret each particular thing, in all its newness to us, generically, and so measure its difference from other things and respond to it appropriately according to its kind. 90 These comments concern the ability to morally deliberate when confronted with novel subjects. This does not pertain to subjects like friendship and work, as explored in EIR. Nevertheless, there is insight provided here into how O'Donovan generally conceives of moral deliberation. He adds, reflecting on casuistry: The moral agent approaches every new situation, then, equipped with the moral law … As he holds the moral law together in thought with a particular situation, it illuminates and interprets it, enabling him to reach a moral judgement about it. 91 One can arrive at conclusive moral knowledge through deliberation and reflection on particular cases. This is undergirded by earlier claims of an intelligible, ordered creation, vindicated in the resurrection of Jesus Christ.

This comment refers to Resurrection and Moral Order and
These theological emphases are congruent with the approach taken to specific subjects in EIR. But it is first worth taking a step of abstraction and considering the theological congruence between RMO and the series ET as a whole. The assertion that the creation is intelligible to moral agents in RMO corresponds well to the emphasis on moral agency and practical reason. Consider this passage from EIR: Creation is like a mountain with different profiles; it does not actually change shape, but it seems to do so as we move round it and change our angle of vision. To describe the mountain as it really is, not seen from north, south, east, or west, nor flattened out as seen from the air, is to describe what no human eye can see … Yet a mountain of created goodness and truth there is, just one mountain, from however many angles we may view it. Discourse about it is possible and necessary among different people who stand in different relations to it. 92 Here objectivity and subjectivity are held together; the latter is predicated upon the former. There is an intelligible created reality that agents can discern. There is ontology-an actual mountain-that vindicates subjectivity.
There is a similar point made in Self, World, and Time. O'Donovan asserts that ET has 'its angle of vision turned principally towards the subjective renewal of agency'. 93 He then immediately qualifies these comments: 'much must be said here about the objective order of created goods'-such distinctions are to speak only of the 'general directions' of the two texts. 94 This new emphasis on subjectivity is not to abandon the importance of an underlying, objective order that is perceivednamely, the mountain. ET is not replacing the 'forceful moral objectivism' of RMO but complementing it. 95 There is plenty of created order in ET and so, too, moral agency in RMO. O'Donovan himself notes that RMO attends to 'the renewal of human agency [more] than was apparent'. 96 There is, then, theological congruence between ET and RMO. The phenomenological approach taken in EIR-the engagement with practical reason in 'The Communication of Friendship'-is ultimately predicated on this kind of objective, creational ontology. To sagely meander through 'human experience' of sociality confident of discovering something useful assumes a vindicated created order 97 -there is indeed a mountain that our subjective perspective grasps, even if incompletely. The approach taken to specific subjects in EIR fits well with RMO.

Triadic Pillar II: Sin
The same can be said regarding sin. In RMO sin is the rejection of the created order and 'an inescapable confusion in [humanity's] perceptions of it'. 98 Humanity's ordained vocation in the cosmos as sapiential beings has been refused, and knowledge 'has been inescapably compromised by the problem of fallenness, the defacement of the image of God, and by the fallen creature's incapacity to set himself right with good will and determination'. 99 Hence sin is fundamentally epistemological: 'chiefly a matter of confusion and intransigence', as Errington suggests. 100 This helpfully explains the greater emphasis on sin in EIR's triad. Because if sin is an epistemic problem it is an agential problem. It is not on this account primarily the creation that is damaged, but creatures-the mountain does not 'actually change shape' due to sin but sin distorts 'our angle of vision' on the mountain. 101 No wonder a series-ET-that seeks to attend more closely to moral agency and subjectivity produces a triad that is more -according to Herdt-'finely differentiated attention to sin'. 102 For O'Donovan, moral agency is the primary sphere of sin's reign.
The methodological difference of the triads-a change from reconciliation to sin in the second pillar-thus points to a fundamental theological coherence between RMO and ET. For O'Donovan's earlier assertions about the epistemic impacts of sin in RMO accord with the ethical approach of EIR, namely, the proceeding 'by way of the wrongs which they expose to us'. 103 It is thus unsurprising to find in 'The Communication of Friendship' that the examination of exclusion includes a novella about soldiers in bewilderment about the opacity of their moral dilemma; 104 the reflection on betrayal concerns the false beliefs about rivalry in contrast to neighbourliness. 105 The wrongs that friendship exposes us to are cast as deeply epistemic and agential problems. The theological axioms of RMO fit well here.

Triadic Pillar III: Sanctification
The final pillar of O'Donovan's triad in EIR concerns sanctification. I noted above that this single task has a twofold nature: it examines both the perfection of the created reality in Christ and its ecclesiological implications. These two strands are similarly inextricable in RMO. Regarding the perfection of creation itself-namely, ultimate eschatological restoration-this is grounded in the resurrection in RMO. The resurrection of Christ-namely, as 'Jesus is given back to his circle of friends' 106 -is the sign that 'this order, with mankind in its proper place within it, is to be totally restored at last'. 107 This too fits well with EIR; the same ideas emerge in 'The Communication of Friendship'. O'Donovan argues that the perfection of friendship only makes sense 'in the context of the resurrection, the supreme justifying and sanctifying act of God set in the middle of history, which gives Christ back to his friends'. 108 This is not only semantically congruent with RMO-'Jesus is given back to his circle of friends' 109 -but theologically congruent too. The perfection of creation is again predicated upon the resurrection. It is true for all of creation, as is the focus in RMO. It is equally true for all the particularities of creation, such as friendship in EIR.
The perfecting reality of the resurrection has ecclesiological implications in RMO. O'Donovan argues that 'in Jesus God has reached out to us from heaven in love'. 110 The appropriate response is to 'love him in return by believing'. 111 This takes the form of neighbour-love too. For 'the church anticipates restored humanity'. 112 This ought to produce a love for the church, for 'all humanity lies implicitly within the church'; 113 the love believers have for one another takes the central place in the New Testament as the binding feature of the ecclesia. 114 But this same love also compels us to 'venture out beyond its present borders, to claim those who lie outside [the church]'. 115 The resonances with EIR are obvious; claiming that awaiting the restoration of humanity implies love for the ecclesia and mankind could equally belong in 'The Communication of Friendship'. O'Donovan makes the very same moves in this chapter as he does in RMO: the love of Christ leads us to love of the church-philadelphia-and that ought to lead to hospitality, a community that 'receives guests' and 'goes out' to those absent. 116 The same train of theological thought unfolds in both texts.
It is also worth recalling that a central feature of the ecclesiological implications in 'The Communication of Friendship' concerned the identification of the real, 'eschatological' church. 117 O'Donovan's suggestion was to begin with the visible institutions. Most illuminating is a comparison of this discourse with a discussion of conversion and baptism in RMO. 118 The question here is how one might identify the legitimate, eschatologically grounded conversions of others. The answer O'Donovan provides is to begin with baptism-it is a visible sign of conversion. 119 There is a similar suggestion made in EIR regarding the visible church-to identify eschatological realities one should begin with ecclesial signs. This is not simply to highlight an idiosyncratic similarity. Rather, it demonstrates that a shared theologic is operating in the background-eschatology and ecclesiology are bound together and to discern the former one must start with the latter. This is the fruit of theological congruence between the two texts.
Yet with this all considered a question remains: why does the third pillar of the triad change from the language of eschatology in RMO to sanctification in EIR? It is again due not to a wholesale change of theological substance but perspective. For O'Donovan, to speak of sanctification is to still speak of 'God's saving work in Christ'. 120 But it is now to do so 'from one particular angle': 'the transformation of our experience as agents'. 121 He is speaking of the same objective realities outlined in RMO-regarding eschatology and ecclesiology-from the subjective angle. This is true of the entire triad employed in EIR. The same theology of RMO is being refracted through a different lens-the lens of subjective moral agency-in EIR and ET.
Here it must be noted that voices like those of Tranter suggest that the differences between RMO and ET-and subsequently something like the triads examined hereare more disjointed than I have presented them. To be clear: Tranter does concur with much of what I have claimed here. He reads ET as 'a return to moral-theological first things … [that] revisits the themes of RMO, but it modulates them into a new key'. 122 He similar characterises ET as 'reimagining' or 'remapping' the same moral theological foundations established in RMO. 123 Hence Tranter does recognise a significant coherence between the two texts: 'there are certainly plenty of points of continuity'. 124 Yet Tranter also suggests that framing the changes in ET as a mere 'complement' to RMO is perhaps to 'smooth over' the comprehensiveness RMO originally claimed for itself. He suggests: 'surely something more like a revision than a supplementary perspective is being entertained … there are … points at which decisive departures are made from the earlier work'. 125 He therefore concludes the differences between RMO and ET are more akin to 'developments' or 'advances'-even 'tensions' as per the title of his book-rather than merely a change in perspective. 126 How the differences of these projects are most precisely characterised in the final analysis-whether as development as per Tranter or complement as per O'Donovan-is a dispute the central thesis of this article does not depend on. I think this is especially the case given that this disagreement is not that of a complete impasse: on the one hand Tranter still accepts that there is significant 'continuity' between RMO and ET and on the other hand O'Donovan is still 'conscious' his moral theology has 'progressed' between the two projects. 127 For the purposes of this article it is enough to note that the triadic differences between RMO and ET are not-on either reading-the result of wholesale theological change or complete contradiction. These real differences-or even perhaps 'developments' or 'tensions'-are primarily a result of an essentially familiar theology of RMO-the same 'salvation-history' 128 -being viewed from the subjective perspective of the moral agent in ET. Here I think even Tranter would be in agreement.
This change of perspective not only accounts for the triadic differences already traced in this article; it also explains the prima facie similarities of the triads. It is not mere coincidence that there are shared ideas like creation-even if this means different things in each text. It is unsurprising that the entire triad is brought to bear on a single subjectas learned from the errors of Barth. There is a deep theological coherence between RMO and EIR that parsimoniously accounts for both the notable methodological differences and similarities of their two triadic approaches to specific moral subjects.

Triadic Arbitrariness?
This article amounts to but a cursory comparison of RMO and ET. I have attended very particularly to the approach to specific subjects in EIR so as to provide a helpful entry point into a large and complex corpus. But this same focus ensures this article is inherently uncomprehensive and neglects vast subtleties in O'Donovan's theological ethics. It is therefore-mindful of these limitations-that, with some trepidation, I raise at this concluding juncture a necessary critical query: the methodological arbitrariness of O'Donovan's triadic approach.
It is O'Donovan himself who argues in RMO that 'it is pleasing to be able to speak of salvation-history triadically [but] it can only be convention'. 129 His concern is that otherwise we risk a 'preposterously truncated' gospel. 130 The ethical task ought to consider 'all the moments of the Christ-event', including an 'advent moment' that accounts for his earthly ministry. 131 Yet one wonders whether O'Donovan risks failing to heed his own warnings in the later EIR. The triad in EIR is not doing the same thing as in RMO. Further deliberation would be required to determine how to integrate an 'advent moment' into a theological ethics that aims to be especially attentive to agency. 132 But this broader concern about the conventionality of triads remains and ought to be brought to bear on EIR.
It is worth considering how the triad in EIR develops. It is the climactic thought of the section 'Body, Soul, and Spirit' in 'The Communication of Work'. 133 It is difficult to trace the development of his thought here. O'Donovan draws on 1 Thess. 5:23 to form a tripartite anthropological formula of body, soul, and spirit. 134 It seems this triad is the grounds for choosing the specific subjects O'Donovan examines in EIR. The three parts of the formula correspond to a topic-body and work, soul and friendship, spirit and meaning: 'we are bodies in a material world, centers of emotion and relation in a social world; centers of intelligence and purpose in a world of meanings'. 135 It is then, following this train of thought, that O'Donovan outlines yet another triad: the approach to specific subjects that has been examined closely throughout this article. 136 But why this triad? O'Donovan does not clearly indicate whether it in some way corresponds also to the triad of body, soul, and spirit just articulated. It seems in the end to be partly chosen simply because it is a useful triad. One recalls the comments from RMO here: 'it is pleasing to be able to speak of salvation-history triadically [but] it can only be convention'. 137 Perhaps then this triad in EIR is merely the fruit of an undue attraction to tripartite formulae. The project ET-notably a trilogy-is itself an exploration of triads: faith, hope, and love correspond to the sanctification of self, world, and time. 138 But in this case, as with the tripartite anthropological formula of body, soul, and spirit, the triads are at least clearly argued for. The same cannot be said for the triad employed to examine specific subjects in EIR.
The concern is not simply that this triad is unsubstantiated-though this is potentially problematic-but that it leads to more troubling questions. Is this a truncated account of moral agency? Are there other features that could have just as well been included? Is this formula in the end an arbitrary convention? These queries loom over the latter chapters of EIR. The allure of the triad risks becoming distracting and undermining an otherwise rich exploration of specific subjects.
But this critical query must be somewhat tempered. For this conventionality points to a more fundamental and deliberate arbitrariness undergirding ET as a whole. And this in turn illumines some of the broader implications that O'Donovan's understanding of his project might have for the pursuit of theological ethics more generally. For O'Donovan is not unaware of this issue of arbitrariness. He himself notes 'that there could be other possible starting points' for ET, pointing to the Sermon on the Mount as an example. 139 He adds, 'I entertain the fantasy that, were I to begin this enterprise again, I might take my starting point from the teaching of Jesus'. 140 Coakley rightly observes that this 'would have been to shape the project somewhat differently'. 141 David Elliot pushes the critique further: O'Donovan indicates no rationale for which approach to prefer among equally plausible alternatives. This leaves it unclear how necessary any given approach is to the kind of work he is doing. This is not a crippling problem, but it does create the impression that his work is at root 'occasional', perhaps even mostly idiosyncratic. 142 vista of salvation-history articulated in RMO. And he does so-in the final analysis-with a critical self-awareness of an arbitrariness and contingency that points beyond the horizon of our moral-theological systems to the true and living God that cannot be hemmed within them.