Free will theodicies that rely on a good afterlife to explain how earthly evils are ultimately outweighed or defeated are seemingly inconsistent when they claim that freedom with respect to good and evil choices is a great good, while also maintaining that a heaven with no freedom or opportunities to make such choices is supremely valuable.¹ This implies that either, i)

Heaven is not such a great good after all, because it lacks the freedom to choose between good and evil, or

Simon Kittle terms the first explanatory difficulty concerning the nature of heavenly goods the lack of value problem (2018, p. 100). How could morally trivial choices between, for example, worshipping God by playing the harp or singing in a choir, be valuable enough to justify the cost of creating a world of evil and suffering? Kittle suggests that the main challenge for a choice-based account of heavenly freedom is to give some idea of how the types and number of choices that an agent might face in heaven add up to a level of freedom that it would be desirable to have (2020, p. 465). However, such an account is typically not prioritised when divine goodness is understood primarily in terms of a loving God’s desire for union with other personal beings, and heaven is taken to be the consummation of that desire. A heaven that fulfils the goal of divine-human union seems to make heavenly choices to advance this goal superfluous. Moreover, where conceptions of heavenly sanctification, glorification, or deification imply uses of human freedom, the idea of perfecting the will tends to be the focus, such as with Timothy Pawl and Kevin Timpe’s understanding of how persons in heaven might grow in virtue (2017) and Eleonore Stump’s account of how the divided will’s progressive integration around the good removes psychological barriers to deeper union with God (2018). Kittle suggests that these non-choice-based accounts of heavenly freedom typically locate its value in certain internal states of the will, such as being able to realise ones wants, having lower-order desires aligned with higher-order desires, being free from psychological obstacles or being able to act in accordance with what is good and true (2018, pp. 100-101).

I shall argue that a good-maximising view of divine benevolence and God’s creative purpose provides for a choice-based account of heavenly freedom that answers Kittle’s challenge. On this view, God creates not solely or primarily for the sake of uniting with other personal beings, but to maximise the goodness of a world created because a world where other goods exist apart from the being of God is better than one where only God exists. As part of this world such a God could be expected to bring about a realm of enhanced agency, a maximally good paradisal environment where self-aware persons with freedom and goodness of will have the greatest possible scope for embodied agency, so that they can fulfil two related purposes, contributing to maximising created goodness and increasing their proximity to God. Before I discuss what goods might be involved and how choices between them could be supremely valuable, I first outline my methodological approach to conceptualising heavenly life.

This paper assumes that theistic concepts of heaven are appropriately evaluated in inference to the best explanation comparisons of theism and its rival worldviews, where theism concerns belief in an omniperfect God and the successful worldview is the one assessed to have the greater explanatory power (Oppy, 2018).³ As such evaluations entitle the theist to suppose the truth of theism when advancing their hypothesis, they may construct any explanation of the world that is consistent with this supposition, to consider how it might fare in an inference to the best explanation comparison.⁴ Given the apparent gratuity of many earthly evils, theism will lack sufficient explanatory scope if it does not suppose that all evils would ultimately be outweighed and/or defeated by greater goods in a possible afterlife. This follows from the widely held assumption that a perfectly good, all-powerful God would not permit or cause any evil that is preventable and pointless with respect to God’s purposes,⁵ and the sufferer-centred requirement which regards the suffering of the innocent as permissible only if they benefit from the good secured by the means that causes their suffering.⁶ As the claim an omniperfect God exists cannot provide for a comprehensive explanation of the apparent gratuity of many evils if the claim a good afterlife exists is false, the existence of a good afterlife is more probable than not on the supposition that theism is true. Stated differently, the existence of an omniperfect God and no good afterlife is less likely than the existence of both given the apparent gratuity of earthly evils.

To conceptualise this afterlife, I combine what may be termed top-down and bottom-up inferences concerning its possible nature. Top-down inferences derive features of an afterlife from a conception of God; bottom-up inferences derive this content from goods known to our experience. Aquinas may be taken to have used a mix of top-down and bottom-up reasoning to derive his account of the highest human good from the notion of an essentially personal and loving God and our worldly experience of the goodness of personal relationships.⁷ The concept of an omniperfect God from which I derive some key features of a possible heaven prioritises divine benevolence rather than love, and a good-maximising view of God’s creative purpose. Here, I follow J. L. Schellenberg in distinguishing the benevolence of a God who selflessly desires the good from a divine goodness understood primarily in terms of an essentially relational notion of love where God is solely motivated by a desire for personal relations with others for the sake of such relationships as goods in themselves (1993, pp. 20-23).

A benevolent God may be supposed to have created the world because a world where other goods exist apart from the being of God is better than one where only God exists. It also seems reasonable to suppose that such a God, by virtue of their omniperfection, has the will, knowledge and power to maximise created goodness, the good that results from God’s creative activity, where to say that something is maximised is to say that no more of it could possibly exist or that it has been increased to the greatest possible extent. On this view, an omniperfect God who actualises a world that is less good overall than some other possible world would not be perfectly good, suggesting that such a God would choose to create the best possible world, or any one from a set of best possible worlds of equal value. The claim God is omniperfect does not entail the claim an omniperfect God maximises created goodness, but I suspect that the former strictly implies the latter. At the very least, a good-maximising purpose is implied or made probable by divine omniperfection.⁸

The central objection to this concept of God is the paradox of maximisation, that however much good an omniperfect God creates, there would always be other goods that could be added to those that already exist, making their maximisation impossible. While talk of maximising non-moral goods does seem paradoxical, I am not persuaded this is also the case for moral goods that entail the possibility of moral evils.⁹ On theism there seemingly must be a threshold above which further instantiations of evil or certain quantities of types of evil become gratuitous.¹⁰ This suggests that for possible worlds containing moral goods and logically necessary moral evils there is a size of world or a total number of moral agents beyond which the cost of moral goods becomes prohibitive because of the resulting amount of moral evil.¹¹ I shall argue below that a good afterlife cannot contain good and evil choices, but this kind of response to the maximisation problem is consistent with the notion that God created an earthly life with the freedom to make such choices as a necessary condition of bringing heaven into existence.

Granting the coherence of the idea of a good-maximising God, such a deity could be expected to maximise created goodness by means of an environment with the following general features: 1)

A maximally good environment would be paradisal, containing no bad states of affairs caused by natural evils.

We can surmise from 1) to 5) that the world of our experience, given the moral and natural evils it contains, is not this maximally good environment. The epistemic assumptions outlined at the start of this section also entitle the theist to suppose this ideal environment to exist in a good afterlife. The theist is justified in inferring something of the nature of this possible afterlife from their experience of earthly goods, because on theism our world must provide the necessary conditions for bringing a good afterlife about. This implies there is a fundamental relationship between at least some of the properties of this life and those of the next life, given that the coherence of means-end rationality depends on the means and end sharing some properties that are intrinsically valuable to both. Furthermore, if earthly existence and the good afterlife share important characteristics, and the latter is a much greater good than the former, then the afterlife must contain a fulfilment of goods that are found in the world of our experience, where the notion of fulfilment refers to the fullest or highest realisation of some good.

This means that inferences concerning the heavenly fulfilment of embodied agency, a key feature of the maximally good environment outlined above, may be derived from earthly experience of embodied being and free will. By employing such bottom-up inferences I am not, however, supposing that the good afterlife is a mere continuation of mortal life. I agree with Marylin McCord Adams that any package of heavenly goods which are merely commensurate with earthly ones would not be valuable enough to make heaven a supreme good (1988, p. 235). Rather, the post-mortem fulfilment of embodied agency is taken to involve unexperienced goods that are strongly analogous to experienced ones, but which are only somewhat commensurate with them because material conditions in an afterlife greatly amplify or enhance those goods in terms of their scope, quantity and quality. The next section will outline some of these possible conditions.

To illustrate part of a bottom-picture of the nature and value of embodied agency in heaven, I will draw on Robert Kane’s neuro-physical and systems theory account of free will (2011; 2014). Attention to the neuro-physics that might be involved is appropriate given the fundamental continuity supposed between this life and the next. It also follows when presupposing dualistic holism, which views human beings to be essentially personal-spiritual-physical unities without affirming that we are made from one metaphysical constituent or precluding a non-physical dimension surviving the dissolution of the whole at death (Cooper, 2018, p. 416).¹⁶ Any continuation of life after death seemingly must involve some form of dualism, but dualistic holism looks to provide the most plausible account of post-mortem survival, since I take it to mean that a person is complete not only when they are embodied, but when their non-physical dimension is united with the particular body of which it is an integral part. Dualistic holism therefore suggests that an embodied afterlife would resurrect persons with the specific neurophysiologies made by their character forming choices in mortal life, as well as the unique physical traits of their personal identities.¹⁷

In employing Kane’s theory I endorse an incompatibilist view of free will. Compatibilism may appear a better fit for a heaven created by a good-maximising God, since an agent who is causally determined to always choose the best option, for example, would be a better maximiser of created goodness than an agent free to choose the less than optimal alternative. However, a theistic explanation of moral evil would seemingly have to involve some variety of incompatibilism.¹⁸ I am also doubtful that the second problem of heavenly freedom outlined in the introduction can be addressed adequately by a theodicy that offers a libertarian explanation of earthly moral evil while taking a compatibilist view of free will in heaven, for example when the beatific vision is thought to necessitate moral impeccability and the love of God.¹⁹ The explanatory power of any free will theodicy therefore depends, in my view, on distinguishing freedom with respect to good and evil choices from other valuable aspects of libertarian free will, and providing an account of how a good afterlife is dependent on the former and continues the latter. Nevertheless, I am not committed to Kane’s event-causal explanation of undetermined choices. Rather, I am inclined towards hybrid explanations of agency and free will that combine event-causation with an agent-causal view, where the significance of certain choices periodically brings the agent qua substance into awareness of herself as an agent to exert causal power over the settling of decisions between competing options. This possibility is in line with a recent proposal for causal pluralism by Joseph Martinez (2023) and Kane’s own suggestion that agent and event-causal views are compatible (2011, p. 396).²⁰

Kane’s theory does, however, provide a set of ideas that can help to illustrate how one feature of libertarian freedom, its scope to develop the potentials of agents, could give it value in a heavenly environment. In his account of free will there are self-forming actions, concerning those intentional choices that can shape an agent’s character when their wills are not already set one way on doing something before they act. Such choices are undetermined when they bring into conflict competing neural networks or volitional streams, goal-directed cognitive activities aimed at setting a will in a certain way, requiring agents to make efforts of will to resist the temptation to do either one of the actions they strongly want to do. The resulting uncertainty stirs up a chaotic disequilibrium in the neural processes of the brain, which opens a window of opportunity that screens off complete determination by prior mental events. As the agent is an information-responsive complex dynamical system, the chaotic brain state stirred up by a volitional conflict can give rise to emergent capacities, as the self-forming action replaces existing constraints on the behaviour of parts of the system with novel constraints suited to a new relationship between the whole and its parts. For Kane, it is this reciprocal causal influence of wholes to parts and parts to wholes which characterises the interaction between conflicting volitional streams within an agent’s larger neuro-physical system, or the self-network as he terms it, and the setting of the will through self-forming actions (2011, pp. 384-387, 396).

This picture has two implications for our discussion of the lack of value problem: not only does free will require neuro-physical flexibility, but libertarian freedom appears to be essential for neuro-physical growth. For example, a God-made world in which determinism or compatibilism precludes morally evil actions would probably not produce the sort of chaotic brain states from which new capacities may emerge, because it implies a narrowing of the range of possible choices open to the agent that seems likely to keep their self-network operating within existing constraints. This could also be the case in a heavenly environment in which agents have libertarian freedom but where the range of goods is so limited that choices between them can only be trivial. There is no space to develop these intuitions here, but in the good afterlife conceptualised in the next section I shall suppose that free will requires cognitive flexibility and derives a significant part of its value from the neuro-physical change and growth enabled by undetermined choices involving the widest possible range of goods.

A good-maximising God could be expected to create a realm of enhanced agency (REA), a paradisal environment where self-aware embodied persons with freedom and goodness of will have the greatest possible scope for agency. The term enhanced is chosen to describe the relatively higher levels of agency in this realm compared to those experienced in our earthly lives, but the terms optimal or maximal or ultimate could also be used. The thought motivating this idea is that agency exists in degrees. Generally, as our competence, well-being and resources increase, the range, quality and impact of our actions expand. The good-maximising view of God implies that such a deity would want self-aware embodied persons with freedom and goodness of will to have as much agency as possible, since this would produce the most created goodness overall and optimise the proportions of higher and lesser goods in an environment containing the greatest quantity and diversity of created things and beings.

REA, however, is not assumed to contain the same material conditions we find in earthly life. For example, I am not supposing that agency is enhanced by making everyone incredibly wealthy and technologically advanced. How then can we form a picture of the kinds of goods that might exist in this supposed realm, one that builds on the six general features of a maximally good environment outlined in the previous section? The foregoing discussion suggests three pre-requisites. Firstly, these goods must be derivable from the first order goods of embodied being and freedom of will. Secondly, their fulfilment should plausibly constitute human flourishing and the highest of REA’s created goods. Thirdly, since its environment lacks moral and natural evils, REA cannot contain the sorts of second order moral goods found in a world like ours, such as actions to fairly distribute scarce resources, dispense justice for wrongdoing, alleviate suffering, and protect people from harm. I propose that three types of non-moral goods satisfy these prerequisites: i)

learning, the activity of acquiring knowledge of the true nature of reality, the self and others.

These non-moral goods can be attributed with positive moral value through the actions of good-willed agents. They are also plausibly inclusive of all human goods in REA, especially when presupposing a notion of the person as essentially embedded in relations of one sort or another with other persons and objects, since learning, creativity and relatedness activities may all be taken to be inherently relational in some way. On this view, even simple sensory pleasures, like tasting food, could be tokens of relatedness for example, perhaps with one’s self as we enjoy being at one with our bodies and emotions, or with God through feelings of deep gratitude toward the ultimate source of such goods.

In REA, agents’ participation in learning, creativity and relatedness goods would be empowered by any material condition that maximises their internal capacities for agency and minimises external constraints on their actions. These conditions might include: 1)

Everlastingness. To remove any environmental constraint on the production of created goods, a good-maximising God would sustain space and time in REA indefinitely. This implies that the physical systems that make up REA’s environment are somehow perpetually regenerative or non-entropic.

Before considering how learning, creativity and relatedness goods might be experienced in such conditions, it is worth noting some points about the conceivability of 2) and 5). How REA could contain agents with biologically optimised bodies and be an afterlife for human beings with a prior earthly existence is possible in the following way. The dualistic holism presupposed in the previous section is assumed to provide for a resurrected person’s psychological continuity with their pre-mortem selves, where the immaterial dimension that survives the dissolution of the whole carries with it that part of a person’s consciousness which is essential to their subjective experience of reality, including their memories and personality. Granting this assumption, we further suppose that this immaterial dimension not only retains a person’s unique psychological identity, but stores information concerning their genotype, so that their specific biological identity can be reconstituted with new material after death.

A mechanism for this possibility is suggested by quantum theories of consciousness that distinguish phenomenal consciousness from deeper levels that might underlie ordinary awareness (Baruss, 2008, pp. 258-260). On this view, a person’s field of consciousness conceivably extends to the molecular level where it could retain genetic information. The notion that our resurrected bodies are reconstituted from this genetic code also suggests how the re-embodied person might have the perfected biological capacities outlined above, since the reintegration of a person’s physical and non-physical dimensions at resurrection could provide an opportunity for the new body to be formed to the maximum potential of its given genotype. For example, perhaps the process of re-embodiment fully realises this potential by activating a latent capacity for protein coding lying dormant in so-called junk DNA, those non-coding genes that constitute a substantial majority of the human genome about which little is known concerning their role in our development and physiology (Palazzo and Gregory, 2014). Conceivably, then, the process of resurrecting our physical bodies could optimise the specific genotypes we had in mortal life to enhance our physical and mental attributes for existence in REA.

The idea of a multi-dimensional environment facilitating movements across parallel dimensions is the most speculative of the conditions outlined above, having been partly inspired by biblical and Quranic stories of prophets ascending to heaven and angels mysteriously appearing from nowhere. These stories suggest how a being entering our world from another dimension, or moving the other way, might seem to us like they were instantaneously popping into or out of our realm of existence. The conceivability of this idea perhaps depends on whether there is a theory of space-time that could make it possible, a problem which cannot be explored here. Nevertheless, it is worth clarifying why such an environment could be agency enhancing and therefore something a good-maximising God might be expected to create.

I am assuming that the ability to move instantaneously from one point in space and time to another would give enhanced agents greater scope to contribute to created goodness by expanding the range of goods available to them and easing access to these opportunities. However, an environment that facilitates agency in this way would seemingly have to mediate such movements through parallel dimensions, since if space and time within a single dimension were to have the fluidity to allow agents to move to any spatial-temporal point at will, then life in that realm would be too chaotic and unpredictable. No one could expect effects to follow from causes, to form reliable beliefs about reality, or to take responsibility for the consequences of actions - an unintelligible existence that would significantly constrain purposeful uses of embodied agency.

In other words, the act of traversing from one parallel dimension to another is not assumed to be the kind of occurrence that would disrupt the ordinary course of events within a discrete spatial-temporal order. This implies that events in a parallel dimension are generally invisible to agents in other dimensions, except to those agents moving between them, otherwise the experience of multiple co-existing realities would be utterly incoherent.²³ It also suggests that these movements would be motivated by goals that cohere with the overall purpose of a multi-dimensional REA, so that such actions are meaningful in this context, rather than arbitrary and chaotic. One possibility is that these dimensions are hierarchically ordered, with lower levels overlapping with higher ones, and higher dimensions existing in increasing degrees of proximity to God, providing agents with an arena for a personal adventure from lower to higher levels of REA, an idea I shall return to below.

Given the six material conditions outlined above, we can suppose that participation in REA’s learning, creativity and relatedness goods would be sustained indefinitely as self-exhausting and repeatable pleasures. John Martin Fischer defines the former as those

pleasurable activities the single performance of which ends any further need to do them again, and the latter as pleasures that are highly fulfilling and completely satisfying every time they are experienced (1994, pp. 262-264). Noting that Bernard Williams (1973) assumes all pleasures to be self-exhausting in his argument against the desirability of immortality, Fischer rightly suggests that an everlasting life could be sustained by repeatable pleasures. In a maximally good REA where physical bodies are biologically optimised there is seemingly endless scope for repeatable pleasures, such as those involving sensory experiences. For example, loving physical intimacy with a significant other is a relatedness good involving sensory pleasures that could be repeated indefinitely.

The same good may be experienced as both a repeatable and a self-exhausting pleasure. For those who never tire of the taste, the pleasure of enjoying vanilla ice cream is repeatable, but it is self-exhausting for anyone for whom the pleasure of taste involves the sensation of discovering new flavours. Similarly, the self-exhausting pleasure of experiencing new taste sensations is repeatable if, with respect to ice creams for example, there is an endless variety of flavours to try.²⁴ This would seemingly be the case for learning, creativity and relatedness in REA. The pleasure of learning is self-exhausting when a particular thing is learned and repeatable if there is always something novel to learn. Creativity is a self-exhausting pleasure when a creative project is completed, and a repeatable pleasure in conditions that continually stimulate the imagination with new ideas and supply unlimited opportunities to act on these inspirations. Likewise, relatedness is a self-exhausting pleasure whenever a new personal relationship is formed and a repeatable pleasure when there is endless time over which to reacquaint with others after long intervals apart.²⁵

Learning, creativity and relatedness goods may also comprise some of the same acts. For example, composing a piece of music is a self-exhausting creativity good which involves actions that become repeatable relatedness goods when the music is performed for audiences. Similarly, becoming proficient in a new language is a self-exhausting learning good that involves activities that are repeatable relatedness goods when cultivating friendships with speakers of that language. Moreover, if we suppose that agents’ biologically optimised bodies give them the ability to enter a flow state at will, then even the most challenging activities would be enjoyable. The improved temporal and spatial visual processing in these cognitive states could make the performance of complex tasks seem effortless and automatic, and so immersive that agents fully experience the present without feeling the passage of time (Sinnett et. al 2020).²⁶ As REA’s goods would exist interdependently in a matrix of causal relations where every good occasions opportunities for others, their dual aspect as repeatable and self-exhausting pleasures, their hybrid potential as learning, creativity and relatedness activities, and the flow states in which they can be experienced, would plausibly sustain agents’ participation in them indefinitely.

It could be objected that boredom will inevitably ensue in an environment that is maximally good and everlasting, because at some point all things will have been experienced. I have already suggested that divine participation in REA’s goods would mean that opportunities to learn about or relate with God, for example, would never be exhausted. However, even if in a maximally populated environment the repeatable relatedness good of befriending new people, for example, cannot not be enjoyed forever because REA’s population size is ultimately limited, Fischer’s point about repeatable pleasures still stands for other goods. Indeed, it could be that only a subset of the total goods available in REA would be enough to sustain an everlasting life, given the repeatability of many of its pleasures. In other words, while there is a theoretical limit to the amount of goods that anyone could enjoy, in practice there would always be more goods to experience. This is perhaps more plausibly so if earth were just one of many feeder worlds that contributed to the population size and diversity of REA.

The foregoing outline of the kinds of goods and experiences that might exist in REA raises two questions. How might choices in this environment be undetermined when the neuro-physical conditions for indeterminacy appear to be absent? On Kane’s view, self-forming actions are free when the will is not set one way or the other and the chaotic brain states produced by volitional conflicts prevent past influences from completely determining an action. Yet in REA agents are already set on being good-willed and there are no good and evil choices, seemingly precluding actions that are self-forming in the way that Kane describes. Secondly, how might acts of choice in this landscape of goods bring about some value, avoid a cost, or have meaning? As time is not a constraint in an everlasting life, it may not really matter if a repeatable relatedness good, for example, is chosen over a self-exhausting creativity good at a particular point in time, suggesting that such choices would be trivial. Like deciding whether to worship God by playing the harp or singing in a choir, Kittle’s lack of value problem remains.

On the first question, Kane’s theory does not imply that self-forming actions must involve incompatible purposes, as with good and evil choices. Rather, Kane allows for certain complementary goods to give rise to similar neuro-physical conditions, such as when deciding between powerful present desires and long-term goals, or when faced with difficult tasks for which we have aversions (2014, pp. 41-42). On this reading, good or better and now or later choices between present and future goods or easy and difficult tasks could bring volitional streams into conflict, even if both options are directed toward the same purpose. For example, giving up a known present good for an unknown future alternative is not always easy because of the way our emotions are bound up with the things we value in the moment. Such fond attachments can be significant barriers to change, involving lengthy periods of reflection before the

requisite willingness arises. Conceivably, then, choices between complementary goods in REA could motivate the kind of volitional conflicts needed for self-forming actions to be undetermined.

On the second question, choices in REA may have value if its various goods differ in their intrinsic properties, and agents have purposes by which they attribute values to different goods and which motivate choices between them. The discussion in previous sections implies that goods in REA vary in their essential qualities. There are three types of goods, learning, creativity and relatedness; their tokens involve the greatest possible quantity and diversity of created things and beings; and there is a hierarchy of value, with higher and lesser goods existing in optimal proportions and the highest of these coming about through agents’ beatific visions of God. A possible purpose for agents can be derived from a conceivable way the beatific vision in REA might differ from standard Christian accounts of this state as a continuous and permanent immersion in or union with the being of God. This notion of beatific vision arises from a soteriology in which a heavenly divine-human union is our redemption from a fallen existence that is essentially separated from God, but the good-maximising view of heaven does not entail this aspect of Christian theism. It therefore need not take the beatific vision to be the totality of all experience in REA or conceive of heavenly life as in some way coextensive with the being or essence of God.²⁷

Instead, I propose that beatific visions could be peak experiences within heavenly life and its greatest reward. As such, they would be temporary, episodic events, with each agent conceivably experiencing multiple forms and various levels of these visions at different points throughout their life in REA. Unlike concepts of divine-human union, this notion of heavenly life need not explain how the incomprehensible metaphysical gap between created beings and an infinite God is closed without imperilling the freedom and personal existence of the former. More importantly, it suggests a possible purpose that could motivate agents’ choices: the desire to become closer to God through beatific experiences. On this view, REA’s environment exists in a state of proximity to God, where agents can live in varying degrees of closeness to the divine. Learning, creativity, and relatedness activities periodically precipitate forms of beatific vision that inspire and intensify agents’ participation in these goods and increase their levels of proximity to God. These beatific experiences may disclose new knowledge about a particular aspect of God’s mind and being that brings us a little nearer to the divine, or perhaps they refine our capacity for creative agency, bringing us closer in likeness to a creative God.

This possibility need not imply that agents instrumentalise learning, creativity, and relatedness goods for the purpose of obtaining beatific experiences. Conceivably, the reasons why God would grant such experiences would be unknown to agents. Perhaps a certain act is rewarded because of its impact on chains of secondary causes and effects known only to God. It may be that participating in goods for their own sake without thought of their reward is most pleasing to God, or is the only kind of intention that gives a good action the sort of far-reaching consequences that merit God’s reward. Not knowing whether a particular act will precipitate a beatific experience, a good-willed agent would understand that using goods as means to this end would be inefficacious, an epistemic position that seems likely to encourage them to participate in learning, creativity and relatedness activities for their own sake, to experience surprise, awe and gratitude when actions are rewarded with visions of God.

This landscape of goods and motives provides for a choice-based account of heavenly freedom in which agents have the widest possible scope for what Kittle terms implementation control, the freedom to choose how to bring about already settled-upon goals (2020, p. 461). The ultimate aims of agents are set: contributing to maximising created goodness and increasing their proximity to God through beatific experiences. As there is epistemic uncertainty about how God’s gifting of beatific visions is contingent on their participation in learning, creativity, and relatedness goods, the intermediate goals by which agents might advance their purpose are completely open-ended, and can be formulated in any way they choose. In REA, agents would therefore be free to form volitions around many and various intermediate goals linked to innumerable goods. Moreover, since increasing an agent’s proximity to God involves movement towards their highest possible good, choices involving such goods could make a difference to how the best interest of every agent is realised. Conceivably, choices between good or better options, or taking an opportunity now or later, might affect the shape and timing of an agent’s progress, with decisions which accelerate that progress having greater value than slower paths. If such choices represent real forks in the road, then one good would have more value than another if choosing it were to bring an agent to a new phase in their personal development.

Kane’s concept of an agent as an information-responsive dynamical system is suggestive here. Personal growth in REA could be characterised as a process of neuro-physical complexification, where the chaotic brain states produced by profound choices give rise to a new relationship between the whole and the parts of an agent’s self-network and a step-change in the complexity of that system. This possibility does not primarily concern changes in the nature of the will, as all enhanced agents must already be good-willed. Rather, what develops is a person’s capacity for embodied agency, where goodness of will is one element in a person’s neuro-physical make-up, alongside intellect, imagination, knowledge, memory, emotion and the physical senses. These interconnected powers conceivably deepen their integration with one another with each step-change in neuro-physical complexity, enhancing a person’s capacity for embodied agency in the process.

In a multi-dimensional heaven of increasing levels of proximity to God, progression to higher dimensions could be linked to choices that develop this capacity for agency. This is because a heaven structured in this way need not be characterised by an exclusivism where lesser goods in lower dimensions are abandoned for better goods in higher dimensions, thereby narrowing an agent’s range of choices. Rather, the agent who can traverse heavenly dimensions would be able to experience the full plurality of goods in all dimensions available to them, with every good, whether higher or lower, contributing to maximising created goodness overall. This implies that agents with such choices would need the cognitive capacity and competence to participate in a much wider and more complex array of created goods, making the opening up of higher dimensions of heavenly existence dependent on self-forming actions bringing

an agent to new thresholds in their neuro-physical development. It also suggests why REA’s afterlife must be progressive, since a person’s pre-mortem life will not have prepared them for the higher levels of agency possible in this environment.²⁸

In contrast, then, to concepts of heavenly sanctification which are progressive in providing for a perfecting of wills for union with God, REA is progressive in the sense that increasing one’s proximity to God is agency enhancing. An agent’s progression might go something like this. Maximising created goodness consists of participation in learning, creativity and relatedness goods; these activities may occasion beatific visions that increase proximity to God; these beatific experiences introduce new ideas and motivations that can bring about volitional conflicts concerning present and future goods or easy and difficult choices; resulting self-forming actions develop neuro-physical complexity; this enhances agency and the capacity to contribute to maximising created goodness; participating in new goods occasions opportunities for further beatific experiences that increase proximity to God.

Objection 1) The good-maximising view instrumentalises heavenly goods.

A concern with the virtuous cycle just outlined is that an agent taking one of its steps as primary might be instrumentalising other steps for that purpose in a self-regarding way. Not only could this mean that a learning good, for example, may be employed to obtain a beatific experience, but that an agent might seek to use those visions to enhance their agency, making God a mere means to their end. I have already suggested how the enhanced agent’s epistemic position could work against this sort of instrumental mindset. It also seems plausible, given I have presupposed that persons and learning, creativity and relatedness are all essentially relational in some or other way, that agents would be predisposed to value these activities as intrinsic goods, with their relationship to ultimate purposes becoming primary only periodically when actions are self-forming.

A more comprehensive answer might employ resources used by virtue ethicists to explain how other-regarding desires and an ultimate concern for one’s own flourishing can exist in an agent’s motivational structure without making the former secondary and conditional.²⁹ Nevertheless, that REA’s progressive afterlife is agency enhancing need not imply that individuals take as primary the goal of fully realising their own potential for agency. Rather, this desire, by virtue of the enhanced agent’s goodness of will, is assumed to be secondary to the two purposes of maximising created goodness and being as close as possible to God. As such, these purposes constitute each other in the following sense: by desiring to maximise created goodness agents increase their proximity to God by sharing in the divine motivation behind the creation of REA. Stated differently, the flourishing of the kingdom is the common desire of both King and subjects. This suggests that everyone in REA would be interested in the flourishing of everyone else and the flourishing of the whole, which should preclude agents from having desires that are problematically self-regarding.

Objection 2) Choices will ultimately become trivial in a maximally good environment.

REA is a maximally good environment, which entails there is an upper limit to the amount of created goodness it contains. That there is a limit also follows from the idea that this heavenly life fulfils the good of embodied agency, which implies there is an essential constraint on the personal transformation possible in this environment. This means that agents’ progress must come to an end, both in terms of increasing their levels of agency and their proximity to God, at which point there will no longer be any self-forming actions or purposes motivating choices between goods. In other words, choices become trivial when agents achieve their maximum potentials.

Perhaps the objection is answered if progress in REA is understood in terms of the capacities for agency and for proximity to God. Fully realising these capacities need not imply an end to choice-making or that choices involving these capacities lose their value. Just as perfecting one’s eyesight does not constrain the good that can be done with the ability to see, fully realising one’s capacity to know God, for example, would not limit opportunities to learn about God’s nature within the bounds of that capacity. Once all agents reach their maximum potentials, and REA’s virtuous cycle is no longer progressive in the way outlined above, participation in REA’s goods would still be meaningful, in the same way that the goods of friendship can persist without the parameters of such relationships forever widening and deepening, and the friendship becoming ever more intimate.

The purposes of promoting created goodness and being as close as possible to God could, then, continue to give choices significant value. Conceivably, it would also be within the power of the fully realised agent, by virtue of their innate creativity and unique life experiences, to always be able to generate purposes that are personally meaningful and which motivate their choices. In this context, there would be no actions that are self-forming in a transformative sense, yet REA could remain a dynamic environment where certain choices require neuro-physical flexibility, since in Kane’s terms acts of choosing seemingly have the scope to alter the relationship between the parts and the whole of a person’s self-network without increasing its complexity.

Objection 3) The fate of children and the severely cognitively disabled.

As part of the process of resurrection discussed earlier, I suggested that individuals would be recreated with genetically perfected bodies at an optimal age to participate in REA’s created goods. If we assume that an adult body is optimal, what then happens to children whose earthly lives end prematurely, or to those with severe congenital cognitive disabilities? If individuals from either group were resurrected with perfected adult bodies in an environment that enhanced their agency, it seems likely they would lack the self-control and competence needed to act in a way that is not harmful to themselves and others. This implies that those who die young would be resurrected as children to mature in REA under the care and tutelage of adult guardians. Conceivably, individuals who are severely cognitively disabled might also undergo a similar period of guided maturation. If their mental capacities are taken to be equivalent to those of non-disabled persons at immature stages of physical development, then they could be recreated with bodies at the chronological ages they would ordinarily be with these levels of cognition, but without the congenital problems that resulted in their pre-mortem disabilities.

Objection 4) The temporariness of beatific experiences would cause extreme suffering. The concern here is that just as the heroin user goes from experiencing bliss to extreme suffering when deprived of the drug, the temporary and episodic nature of REA’s beatific visions would cause agents to suffer when not immersed in these experiences. The drug analogy can be extended to characterise enhanced agents as so utterly dependent on these temporary experiences that like the heroin addict their autonomy is threatened, undermining REA as a choice-based account of heavenly freedom involving libertarian free will.

There are perhaps two replies to this objection. Firstly, the heroin analogy is not helpful to anyone supposing the beatific vision to be continuous and permanent, since it invites the challenge that the body has no practical use in such an existence by suggesting how an agent transfixed by the vision of God might be like someone forever trapped in a physically inactive drug-induced state. Secondly, the analogy’s force against the idea of temporary beatific visions seemingly depends on how they are construed as experiences of God’s essence. The vision of God’s essence in its totality may be distinguished from experiencing a certain aspect or dimension of that essence. The former, whether for example that essence is taken to be infinite love or an immutable and timeless perfection, is conceivably so overwhelming for any created being that withdrawal from it could only elicit feelings of tremendous loss and deprivation, but why might the latter possibility suggest otherwise?

To extend the drug analogy, experiencing certain attributes of God’s essence may be less like an all-consuming intoxication and more like a mind-altering drug that changes the user’s perception of reality. For example, a beatific experience might involve a vivid apprehension of a particular expression of God’s generosity at work in a relatedness good shared by two agents that permanently changes their self-understandings and their interaction with one another. Visions of this sort could give an agent experiential knowledge of an aspect of God’s essence that functions more like propositional knowledge in that it cannot be taken away once acquired and can be built upon with each new experience. On this view, agents come out of their beatific visions understanding that they have gained something from these experiences, and feeling immensely grateful for having had them, rather than regretting no longer being in these states.

Their impact on the mind could also be such that their absence is less felt, not only because seeing an aspect of God’s essence is less overwhelming than seeing its totality, but because divine attributes are revealed indirectly through experiences of created goods, moderating the force of these visions. We may further suppose that the intensity of beatific visions is tailored to each agent’s capacity to absorb them without harmful effects, so that no one suffers negative feelings when such experiences end. Perhaps in REA’s multi-dimensional heaven this capacity is greater in agents with the neuro-physical complexity to exist in its higher dimensions. There is more to be said about how attributes of God’s essence might be temporarily experienced as rewards or gifts in a realm of enhanced agency, but I see little reason to doubt the coherence of this idea, which does more to save the phenomenon of free will in a choice-based account of heavenly freedom than a permanent beatific state that necessitates moral impeccability and the love of God.

A good-maximising God could be expected to create the realm of enhanced agency, a heaven for self-aware embodied persons with freedom and goodness of will that maximises their internal capacities for agency and minimises external constraints on their actions. In this dynamic environment where agents participate in learning, creativity and relatedness activities to contribute to created goodness and become closer to God, choices are potentially transformative and ultimate flourishing requires a capacity for neuro-physical flexibility and growth that comes with having freedom of will. A more complete account would show how embodied agents with everlasting lives can change neuro-physically while remaining durably good-willed. This dual capacity may be partially explained by the way goodness of will is expressed in a person’s neuro-physical make-up during a mortal life involving good and evil choices, and suggesting how this might be re-created when a person is resurrected. These possibilities are properly discussed when answering the second problem of heavenly freedom outlined in the introduction, as part of a theodicy that incorporates this concept of heaven in its explanation of earthly evils. I have argued in this paper that a good-maximising view of God can address Simon Kittle’s lack of value problem by providing for a choice-based account of heavenly freedom that describes, in outline, what goods might exist in heaven and how choices between them would be supremely valuable.