The future evolution of religion has long been an important topic in John L. Schellenberg’s philosophy. He has often invited us to practice “evolutionary religion”,1 a present-day religious stance aimed at accommodating what the religion of tomorrow might bring (Schellenberg 2019a). In recent writings, Schellenberg not only encourages us to anticipate future religious change, he also suggests that we could actively guide the evolution of religion (Schellenberg 2013, 75; 2019d, 20–21). This radical proposal raises several questions. Setting aside issues regarding our human capability to guide our own evolution as well as the question of whether this is an endeavor we should actually attempt, I concentrate on the matter of discerning religious progress.

The need to discern religious progress becomes crucial when moving from promoting progress in general and trying to guide the evolution of religion in a certain direction. Progress can always be promoted simply by investing resources and hoping for the best. Guiding or controlling a process, however, requires a way of detecting whether progress is being made. Without the ability to discern progress, anyone attempting to guide religious evolution will have no idea whether their attempts are successful.

Schellenberg understands religious progress as a process towards learning the truth about transcendent reality (more on this in the next section), which makes discerning progress anything but easy. Any transcendent reality lies outside the scope of ordinary experience, and there is currently no straightforward, independent way of checking whether changes in understanding of the transcendent represent progress.2 We cannot even be sure whether fundamental historical changes, as the move from animism and polytheism to monotheism, represent progress or regression (more on this later too). But perhaps changes might occur in the future that will make it possible to discern progress in a non-question-begging manner?

As will become apparent in subsequent sections, whether and how religious progress can be discerned depends on several distinct factors: how religion evolves, how intelligent life evolves, and the existence and nature of the transcendent reality itself. The number of possible scenarios in the deep future is seemingly endless, many lying far beyond the current human ability to imagine. It is beyond the scope of any treatment to explore all relevant scenarios, let alone all possible scenarios. I therefore concentrate on a few significant possibilities where the changes taking place are not too alien. These possibilities are mostly what I call best-case scenarios, where beneficial circumstances intuitively should make discerning progress easier than it currently is. If discerning religious progress is ever practically feasible, it should be under such favorable conditions.

In the next section, I give a brief introduction to the relevant aspects of Schellenberg’s philosophy, namely, evolutionary religion and the religion project. In the third section, I assess different ways of discerning progress if a religious cognitive faculty evolves. In the fourth section, I consider the possibility of measuring progress in terms of human flourishing if no relevant biological evolution occurs, and in the fifth section, I discuss how progress could be discerned if naturalism is true. The article ends with some concluding remarks.

Evolutionary Religion and the Religion Project

Since the interest of this article lies in discerning religious progress within the framework of Schellenberg’s evolutionary religion, it must begin by taking a closer look at the details of his position. In particular, clarity on his view of religious progress must be established.

Archaeological evidence suggests that roughly 50,000 years have passed since the dawn of religion. In comparison, humans might have as much as 750,000 years ahead of us as a species, and there might be as much as one billion years left for intelligent life to continue evolving on Earth (Schellenberg 2013, 15–16; 2019d, 16). Pointing to our potentially early position in the history of intelligent life, Schellenberg suggests that religion as we know it might only represent a primitive stage. It is a possibility that most important religious evolution still lies ahead of us, in the deep future (Schellenberg 2013, 2019c). In his view, an awareness of the deep future as a realm of religious possibilities should guide our religious life in the present (Schellenberg 2009, 2013, 2019a, 2019c). As mentioned, “evolutionary religion” denotes a present-day religious stance aimed at accommodating any changes the future might bring (Schellenberg 2019a).

According to Schellenberg, a religious stance needs to meet some basic conditions to qualify as a proper instance of evolutionary religion:

Evolutionary religion, so I suggest, will be

(1) temporally qualified, identifying what is appropriate religiously with what is appropriate to our immature stage of development (Religious Stage Relativism);

(2) diachronic instead of synchronic, situating our present stage in the context of processes and changes unfolding over vast periods of time taking us from the deep past into the far future (Religious Diachronism);

(3) cognitively modest, finding, for the expression of religious intellectual commitment, attitudes and propositional contents reflecting due appreciation for the audacity of a young species engaging the most profound and controversial matters (Religious Intellectual Humility); and

(4) attentive to the evolutionary benefits of redesigned religion, thinking about how religion itself might help us evolve toward ever greater maturity in all areas of human life, functioning as part of the solution to our many immaturities instead of as part of the problem (Religious Pragmatism). (Schellenberg 2019d, 21)3

Condition one and two are straightforward, but condition three, which speaks of cognitive modesty, requires clarification. Schellenberg (2007, 2013, 2019c) argues at length that the proper epistemic attitude towards religious propositions at religion’s early evolutionary stage is agnosticism and that neither religious nor naturalistic belief should be considered justified. However, that does not exclude the possibility of a religious life. Schellenberg advocates non-doxasticism, a religious stance compatible with agnosticism. In non-doxasticism, belief is substituted with some weaker cognitive attitude, like hope (Muyskens 1979), acceptance (Alston 1996), or voluntary assent (Schellenberg 2005).

Condition four is central to present purposes. The idea of guided evolution becomes explicit in Schellenberg’s explanation: “Evolutionary religion can be a vehicle, both for religious and for a more widely human evolution—a positive evolution that we, more and more, are able intentionally to guide.” (Schellenberg 2019d, 21, emphasis added) The idea is that humans should control the evolution of religion to achieve the positive developments we desire, both in terms of religion and human evolution at large. Even though it surely an intriguing idea to guide human evolution through religious means, I focus exclusively on guiding the evolution of religion.

Evolution itself is not a teleological process with preset goals. Attempting to take control of the evolutionary process means attempting to change that fact by imposing a direction of development towards some chosen goals. It is only from this perspective that it makes sense to talk about evolution as progress. Robert McKim (2019) distinguishes between two kinds of religious progress: practical progress, which concerns religion as a means for human flourishing, and progress of understanding, which concerns progress towards finding religious truths. There is a strong focus on progress of understanding in Schellenberg’s writings. Agnosticism might be the proper epistemic response to religious claims in the present, but that will change as humans learn more about the transcendent.

Schellenberg has written extensively on religious investigation and progress of understanding. I concentrate here on the sense specified in Religion after Science (2019), his latest and most extensive treatment of the subject.4 Schellenberg invites readers to consider “the religion project,” a great human project comparable with the scientific project.5 It encompasses all of humanity’s religious thought and activity, and its ultimate goal is to uncover the truth about transcendent reality6 by answering the question “Is there any such thing as a triply [i.e., metaphysically, axiologically, and soteriologically] transcendent reality, and, if so, how should we understand its nature? Are we even able to learn anything about such profound matters?” (Schellenberg 2019c, 10). The goal of the religion project can be met in three ways: humanity can learn that there is a transcendent reality and what properties it has; humanity can learn that there is no transcendent reality; or humanity can learn that the truth about transcendent reality will forever be beyond our cognitive grasp (Schellenberg 2019c, 61).

However, as I pointed out in a previous publication (2022, 117), the idea that humans might discover that knowledge about transcendent reality lies forever beyond our cognitive limits seems questionable. How could a species ever learn that no further cognitive evolution is possible? Is it even possible for a species to master the science of evolution so fully that it can be determined beforehand which future evolutionary paths are possible and which are not? Progress towards the third possible endpoint of the religion project seems highly improbable. Since Schellenberg has yet to address this issue, I set aside this alleged possible outcome. This leaves two ways of reaching the goal: to learn that a transcendent reality exists and what properties it has or to learn that no transcendent reality exists.

While religious evolution itself is a kind of cultural evolution, it can be based on both the biological evolution of human cognitive capacities (Schellenberg 2013,67; 2019a, 580; 2019c, 31; 2019d, 17) and the evolution of cultural factors (Schellenberg 2013, 26; 2019c, 133; 2019d, 21). In many instances, it is not clear which type of evolution Schellenberg has in mind, which is unsurprising since cultural and biological factors tend to be intertwined when it comes to human evolution (Ferretti and Adornetti 2014). In what follows, I discuss religious evolution of both types. The next section focuses on biological evolution and the development of human cognitive faculties. It is followed by a section considering cultural evolution only.

Best-Case Scenario: Beneficial Cognitive Evolution

The most straightforward way the human species could evolve in a direction beneficial for the religion project would be if our cognitive contact with the transcendent improved. Such a scenario does not seem far-fetched from an evolutionary point of view, since, if there is a transcendent reality, this kind of evolution would be pragmatically beneficial for the individual (Van Eyghen and Bennett 2022, 120–23). In this section, I explore the idea of how religious progress can be discerned if a religious cognitive faculty (RCF) develops. For present purposes, religious cognitive faculty (RCF) is defined as sensory functions that enable cognitive contact with the transcendent.

There are at least two main possibilities to consider regarding RCF. The first is that humans’ ordinary senses might improve in a way that enables the formation of reliable beliefs about the transcendent. Since the transcendent is beyond the physical world of our senses, such evolution will not allow us to “perceive” the transcendent itself, only its impact on the world. Successful cognitive evolution along these lines would allow humans to experience the transcendent in the same fashion as we can now be said to experience the minds of other agents.

Both philosophers of religion and cognitive scientists of religion have argued that humans already possess a capacity of this kind. Alvin Plantinga (2000) famously claims that humans have a sensus divinitatis, a “sense of God,” that works with and through the ordinary senses to produce warranted true religious beliefs about God. Hypersensitive agency detection device (HADD) theory (Barrett 2004) from cognitive science of religion claims that humans possess a “hypersensitive agency detection device” that enables us to detect agency not only in persons but also (supernatural) in nature. However, while Plantinga takes the sensus to be reliable, most proponents of HADD theory regard the device as overly active and unreliable, detecting agency where there is none in the supernatural case (Barrett 2004; Braddock 2016; Szocik 2017).7 I want to stress that I am interested not in the current reliability of a HADD but in the possibility that a reliable RCF (which might or might not be a development of a HADD) could evolve in the future.

The second possibility is that humans might develop a new kind of sense that allows for direct cognitive contact and lets us “perceive” transcendent reality in much the same way our ordinary senses let us experience the world. Successful cognitive evolution along these lines would empower us with a new sense faculty that could be analogous to an existing sense, like a “second sight,” but could also be something else entirely.

Of course, we cannot be certain that such direct contact with the divine is metaphysically possible. In a worst-case scenario, the border between what humans can experience and transcendent reality is absolute, in which case, no experience of the transcendent will ever be possible, even if the transcendent exists. The evolution of an RCF of the second, direct type requires a best-case scenario wherein experience of the transcendent is metaphysically possible and the reason humans currently cannot experience it boils down to evolutionary bad luck (like we presently happen to lack the ability to see ultraviolet, which most birds can).

I conceive of RCF broadly enough to include both direct and indirect experiential contact with the transcendent. RCF is not meant to include nonexperiential capacities (like improved reasoning powers), since such capacities would contribute to other forms of religious progress (like better philosophy of religion).

In what follows, I consider the question of how it could be discerned whether RCF evolves towards greater reliability. An indirect RCF might well evolve without there being any transcendent reality, like several proponents of HADD-theory suggest (Braddock 2016), but it seems unlikely that a direct-style RCF would develop in such circumstances. Just as fins do not evolve in the desert, it seems extremely unlikely that a new sense would evolve to let humans “perceive” that there is “nothing there.” However, this matters little for present purposes, since the problems connected with discerning whether RCF evolves towards greater reliability will be the same for both indirect and direct RCF. Quite naturally, a scenario where RCF evolves towards greater reliability is also a scenario where the transcendent exists.8 The possibility that an indirect RCF might evolve if there is no transcendent reality can be safely set aside.

Discerning Progress in Terms of Consistency

How could it be discerned whether RCF evolves towards greater reliability? The reliability of an information-gathering process is sometimes measured in terms of sensitivity, i.e., its ability to reject false negatives, and specificity, i.e., its ability to reject false positives (Joyce 2003). However, such criteria are inapplicable in this case, since there is currently no way of independently identifying false negatives or positives and no way of estimating these numbers. Other criteria are needed to measure the reliability of RCF.

Religious diversity is a primary obstacle to regarding RCF as reliable. Even if RCF were to provide clear perception of the divine, there would still be a strong reason to regard it as unreliable if it remained the case that different subjects had different and incompatible experiences. Reliable cognitive mechanisms should not be expected to produce several incompatible systems of belief (Palmqvist 2022, 108–9).9 It therefore seems intuitive to suggest that increased consistency between different religious outlooks would indicate that RCF has evolved in a more reliable direction. This is based on the assumption that an increase in agreement would be due to better cognitive contact with transcendent reality. Since increased convergence between religious views is a clearly discernible phenomenon, consistency seems a natural candidate for measuring religious progress. Put a bit more formally:

Premise one: If RCF evolves to become more reliable, religious subjects will be in a better position to make true claims about transcendent reality.

Premise two: If religious subjects were in a better position to make true claims about transcendent reality, there would be a discernible increase in consistency between religious outlooks.

Conclusion: A discernible increase in consistency between religious outlooks is evidence that RCF evolves to become more reliable.

The argument, however, is faulty since it rests on the false, hidden premise that all possible increases in consistency are due to improvements of RCF. Religious consistency can increase for other reasons, including deterioration of RCF, and it can even be the case that improvements in RCF lead to decreased consistency. Establishing a connection between consistency and reliability is surprisingly hard.

Here, I demonstrate the difficulties of connecting consistency with progress of understanding using Tiddy Smith’s (2020) argument for animism. A recent study on the religious beliefs of thirty-three isolated hunter-gatherer societies discovered that all thirty-three hold worldviews that include animism as a vital aspect (Peoples, Duda, and Marlowe 2016). Since the thirty-three societies are isolated, they can be treated as thirty-three cases of independent evidence. Monotheistic religion, by contrast, goes back to the same original sources: Zoroaster and a few other prophets in the ancient Middle East.10 Smith argues that this is indeed compelling evidence for preferring animism over monotheism.

Smith’s description of humanity’s current religious situation represents an excellent point of departure for discussing the relation between consensus and truth on an evolutionary picture (if, like Hendricks (2022), you disagree with Smith, consider it a thought-experiment). An obvious way of reading Smith is that RCF has already evolved to a degree where religious consensus is possible in animism. However, the spread of the axial world religions (monotheism in particular) through cultural diffusion has distorted the natural consensus.

Let us assume that monotheism too is a product of RCF, and that Zoroaster and the other prophets had a special kind of monotheistic RCF that functioned differently from everyone else’s. Instead of experiencing a plurality of agents in nature, the prophets became aware of one singular agent behind everything. What does this say about the evolutionary status of the prophets’ RCF? There are two main possibilities to consider:

Improved RCF: Monotheistic RCF constitutes an evolutionary improvement over animistic RCF, allowing for better cognitive contact with transcendent reality.

Deteriorated RCF: Monotheistic RCF functions poorly. It represents a retardation as compared to animistic RCF.

The problem is that one cannot tell which of these two alternatives is correct (at least not without already knowing the truth about transcendent reality). There is no way of telling whether the monotheistic disturbance of animistic consensus represents progress or a degeneration of understanding.

It might be objected that if thirty-three out of thirty-four independent witnesses agree, it is obvious that the thirty-three animists should be trusted rather than the lone monotheist. However, when beliefs are judged only by their coherence with other beliefs, it becomes hard to explain how radically novel ideas can ever be justified. This problem is especially significant in an evolutionary context, where advancement depends on new cognitive capacities that very few people will have initially. It is therefore far less obvious than in ordinary witness situations whether to trust the majority.

Given Schellenberg’s deep future perspective, would it not be premature to draw any conclusions regarding the prophets’ RCF? Should we not wait a hundred thousand years and see whether monotheistic RCF becomes dominant in the population? However, that a feature becomes dominant through evolution in no way implies that it represents a qualitative improvement. There are many examples of how physical features can deteriorate in an evolutionary process. Think of the tiny, useless arms of a T-Rex, or think of flightless birds, or think of seals’ poor ability to walk on land. Even if monotheistic RCF becomes dominant in a hundred thousand years, it could still represent a functional degeneration.

Religious progress of understanding cannot be discerned by measuring consistency, since we have no way of knowing whether an inconsistent exception represents progress or deterioration, and since it is possible for a deterioration to spread to the entire population, in which case consistency would represent universal regress. But if convergence between religious views is not enough to conclude that RCF is becoming more reliable, what is?

Discerning Progress in Terms of Self-Support

One philosopher who has explored the question of when religious extra-sensory “perception” can be regarded as reliable is William Alston (Alston is preoccupied with contemporary Christian mystical perception, but his ideas are equally applicable to the evolution of RCF). Alston’s account is extensive, and I will only take up one of his core ideas for consideration, namely, the requirement of significant self-support. To count as reliable, a belief-forming practice must offer significant self-support. For example, ordinary sense-perception produces a coherent picture of the world that allows for interaction (Alston 1991, 173–75). Alston (1991, 276) claims that Christian mystical perception offers self-support in terms of spiritual growth: “This significant self-support amounts to ways in which the promises God is represented by the practice as making are fulfilled when the stipulated conditions are met, fulfilled in growth of sanctity, in serenity, peace, joy, fortitude, love and other ‘fruits of the spirit.’” Hans Van Eyghen (2022, 120) has re-applied Alston’s argument in an evolutionary context, suggesting that “mystical experience provide guidance to interact successfully with God” and spiritual growth is a sign that this interaction is indeed successful.

This kind of self-support is not very persuasive. As pointed out by McKim (2019, 33), there is no straightforward connection between practical religious progress and religious progress of understanding. All major religions contribute to human flourishing in some way or another, but they cannot all be literally true, since they propose incompatible views about the transcendent. It seems obvious that religion contributes to human flourishing regardless of its truth. It is therefore hard to see how factors such as spiritual growth could be used as primary indicators of reliability in a non-question-begging way. This remains true even if what is “learned” through RCF explicitly forecasts that spiritual growth will take place (as in Alston’s example), because it is easy to see how followers of a false religion might still correctly predict a development that takes place for purely natural reasons.

However, there seems to be something intuitively right about Alston’s general idea that reliable sensory experience should produce significant self-support. If the connection between spiritual growth and reliability is too contentious, perhaps significant self-support can come in other forms? Might a more highly evolved RCF produce self-support of the same kind as sense perception? It is surely a live possibility that an evolved RCF might enable humans to interact with, control, manipulate and perhaps also make predictions about the transcendent in ways presently undreamed of.11

As pointed out by Alston (1991, 209–25), control of or stable predictions about transcendent reality should not be expected if transcendent reality turns out to be the God of perfect being theism. Attempts to manipulate a personal transcendent reality always get ambiguous results. For example, if one prays for the health of one’s spouse and they get a chronic disease, does this mean that God heard the prayer and intervened to prevent her death, that God ignored the prayer, that God does not exist, or that God decided to punish the person praying?

On the other hand, if the transcendent is impersonal or has impersonal aspects, manipulability is not ruled out as on the theistic picture. Whether the transcendent can be manipulated ultimately depends on its nature. If the transcendent is manipulable and humans eventually learn to manipulate it based on RCF experiences, it seems reasonable to regard such manipulation as a way of discerning that RCF is becoming reliable and that progress of understanding is being made.12

A worry is that manipulability will enter the picture too late to be of any use as a means of discernment in the near evolutionary future. Humans surely do not seem able to manipulate the transcendent in the present, and perhaps manipulability should be taken as proof that RCF has reached a stage where it is fairly reliable rather than as an indication that it is becoming reliable. Consider an analogy. Imagine that you are blind but partaking in a treatment that gradually restores your sight. In the first weeks, you see poorly, and you cannot be said to have reliable sight. During this time, attempts at interaction based on what you think you see fail. After the first week, however, your sight is reliable enough to make interaction with your surroundings possible. The point is that when manipulation becomes possible, your sight is already fairly reliable. It seems reasonable to suppose that the same holds for manipulation of the transcendent based on RCF. This suggests that even if humans have a developing RCF, we are presumably too early in religious evolution to expect manipulability, and therefore it is hard to see it function as a means of discerning progress in the near future.

Cultural Evolution and Human Flourishing

I now leave all RCF-scenarios behind in order to consider the possibility, perhaps much more likely than not, that no further religiously relevant biological evolution will take place and that humans will never get in better cognitive contact with the divine. Religion would of course continue to develop, and developments resulting in substantial progress of humans’ religious understanding might still take place. How could we hope to discern progress if all religious evolution is based on cultural factors only?

Schellenberg has written extensively on how to promote religious evolution by cultural means. In Schellenberg’s view, humans have thus far underperformed on the religion project, with widespread intellectual and religious shortcomings hindering progress. Schellenberg mentions examples such as self-importance, intellectual greed, dogmatism, premature belief, and misguided loyalty. He suggests that humans can promote change by freeing ourselves from these “immaturities” (Schellenberg 2019c, 38–50).

Would it be possible to discern progress by measuring how successful humans are in eradicating these immaturities? The straight answer is no. Getting rid of the immaturities would be a way of promoting religious progress, but it would not in itself constitute progress. Trying to discern progress this way would be like trying to measure the growth of flowers by measuring how much weed is removed from the flowerbed. A more direct approach is needed.

Psychological and sociological studies confirm that there is a strong connection between religious participation and human flourishing (VanderWeele 2017). Perhaps progress could be discerned by relying on this connection? The assumption at work is that if religion A is closer to the truth than religion B, religion A should lead to more human flourishing than B. This especially will be true if the transcendent has what Schellenberg (2019, 8) calls “transcendent benefit,” in which case a greater good will be available to those properly aligned.

In the previous section, I rejected the idea of viewing human flourishing as self-support for RCF. My main objection holds in this context as well: if false religions can contribute to human flourishing13 as much as true religions, human flourishing cannot be taken as an indication on religious progress. However, other possibilities remain. In this section, I first assess whether lack of flourishing could be understood as indicating a lack of progress, after which I discuss a best-case future scenario in which a new religion leads to unparallelled increases in human flourishing.

If it is assumed that religious truth always leads to human flourishing, and that false religions might or might not lead to human flourishing, it might seem reasonable to think that progress of understanding always must be accompanied by human flourishing. If so, human flourishing could function as a necessary requirement, and views that do not lead to increased human flourishing could be set aside as not viable candidates for progress.

Unfortunately, even on the assumption that true views lead to human flourishing, progress towards that truth does not necessarily lead to an increase in human flourishing. There can even be progress alongside a decrease. Consider a scenario where Christian theism is true and animism false. Christian theism, the true view in this thought experiment, instantiates the idea of a personal triple transcendency. Imagine an animistic cult experiencing considerable human flourishing. Their religious view is far from true. Not only do they worship nature spirits that do not exist, they also lack any concept of soteriology, and the spirits they worship are not tied to morality in any significant way. This means that their view, while metaphysically transcendent, lacks both axiological and soteriological transcendency. Clearly, if the cult where to change their religious view to one entailing the existence of a personal triple transcendency, this would constitute a substantial leap in progress, and it would do so even if the cult’s new view were to get the specific details wrong.

Imagine if, due to some strange turn of events, the cult’s leaders decide to introduce the worship of Xelarr, god of extreme renunciation. Like Christian theism, the idea that Xelarr exists entails the existence of a personal triple transcendence. Unlike Christian religion, Xelarr worship is built on extreme self-negligence with a focus on self-mutilation, and self-loathing is considered the highest virtue and spiritual goal. Xelarr offers salvation in the form of complete obliteration, but only to his most devout followers. After this religious development, the level of human flourishing in the cult drops significantly. While Xelarr does not literally exist, the cult now rightly recognises that a personal triple transcendency exists, and so they have made substantial progress of understanding. Therefore, a view cannot be dismissed as moving in the wrong direction just because it leads to a decrease in human flourishing.

Up until this point, I have assumed that false religion and true religion contribute equally to human flourishing. What if that assumption is wrong? Perhaps all current major religions, none of which can uncontroversially be said to lead to much more flourishing than the others, are all false, and the true religion, once it has arrived, will contribute to human flourishing on a scale previously undreamed of. Consider the possibility that, in the deep future, a “super-religion” emerges that leads to human flourishing on a completely new level. Would it not be reasonable to consider this an indication that the super-religion has finally brought humanity close enough to the truth for the soteriological greater good to be available?

Of course, this is on the ceteris paribus assumption that there are no other detectable differences between the super-religion and other religions. If all members of the super-religion also use a new super-drug or take part in a new gene therapy or mental training programme, such possible reasons for the rise in human flourishing would have to be eliminated first. Given that there are no such differences, it seems reasonable to suggest that an otherwise inexplicable rise in human flourishing would indicate religious progress, since human flourishing is what should be expected from alignment with a soteriological transcendence.

It might be objected that even if there are no other detectable differences between the super-religion and other religions, the belief system itself of the super-religion will be different. Perhaps humans could flourish simply by holding these beliefs, in which case it would be the beliefs themselves rather than their truth that lead to human flourishing.

I have two responses to this objection. First, while false beliefs regularly lead to human flourishing in all the false religions (whichever they are), it is a substantial assumption that some set of false beliefs no one has yet considered could do this exceptionally more efficiently. Second, and most importantly, if such beliefs are possible, we should expect them to increase human flourishing in roughly the same way as other false beliefs, only much better. Therefore, the rise in human flourishing would not be inexplicable. There would be natural detectable by psychologists specialised in the connection between belief and wellbeing.

The argument assumes that human flourishing would occur as a direct result of alignment with the divine. Many religions, especially in the Abrahamitic monotheistic tradition, contain the contrasting idea that the transcendent (in this case, God) brings about human flourishing indirectly, through natural means. However, my point is simply that if a super-religion emerges, it will be reasonable to explain the rise in human flourishing by assuming that it represents religious progress. If the connection between the divine and human flourishing is indirect, there will be no inexplicable rise in human flourishing.

Another assumption of the argument is that human flourishing can be detected independently of human connection with the transcendent. This would be denied by some religious traditions, especially those where ascetism and renunciation play a crucial role. While I think there is much to be said for a notion of a transcendent reality whose axiological and soteriological dimensions are consistent with an everyday notion of human flourishing, this assumption is nonessential to the argument. Human flourishing can be thought of as relative to what the religion in question says about a good human life. If the super-religion says that misery is the true form of human flourishing and its spread leads to a massive and otherwise inexplicable rise in misery, that should also be taken to indicate religious progress (though such progress might make humans reluctant to continue with the religion project).

There are many possible methods one could employ to discern progress if no relevant biological evolution takes place. Like for human flourishing, these methods would have to depend on measuring some empirical phenomenon on the assumption that it correlates with religious progress. These methods would therefore be similar to using human flourishing in two important regards. First, they would be ambiguous in the present, both with respect to competing religions doing equally well and with respect to competing naturalistic explanations. Second, they would only be truly useful in a future best-case scenario where a new religion is accompanied by uniquely high levels of the relevant phenomenon for which no natural explanation is easily found.

What If There Is No Transcendent Reality?

Schellenberg envisages that the religion project could end with a negative outcome by establishing beyond reasonable doubt that no transcendent reality exists. How could humans make progress towards learning such a truth, and how could such progress be discerned?

As previously mentioned, no beneficial biological evolution should be expected if there is no transcendent reality. Since creatures evolve to be better adapted to their actual environment, it seems highly unlikely that humans would evolve an RCF just to see that “nothing is there.” And there are no best-case scenarios analogous to the rise of a super-religion, which would make establishing the “negative” outcome relatively easy. If there is no transcendent reality, the only obvious path to religious progress is philosophical reasoning.

Some philosophers seem to think that defending the view that there is no transcendent reality is easy as long there are no strong arguments in favor of such a reality. In the debate over the existence of God, atheists sometimes suggest that the burden of proof lies on the theist, since they are the one making transcendent claims. I am skeptical of the idea that there can be default positions in philosophy in need of no defense. In my view, such claims are often fallacious, resting on either an implicit argumentum ad populum (the burden of proof falls on the theist because they have the minority view among philosophers) or an implicit argumentum ad ignoratiam (where the theist’s inability to prove God’s existence is taken to imply the truth of atheism). It should likewise be impossible to reach the “negative” outcome simply by complaining that proponents of transcendent views lack convincing arguments (even if this happened to be the case).

So how could the “negative” outcome be reached? The fastest and most straightforward way would be to find a knock-down argument for naturalism, establishing beyond doubt that the world, as described by science, is all there is. However, as pointed out by Schellenberg (2019c, 94–97), naturalism is not supported by knock-down arguments but by “economic” considerations in metaphysics, ideals of ontological simplicity, and, of course, the success of science. To reach the “negative” outcome, i.e., to settle the matter once and for all, something more definite is needed.

There is no easy route to the “negative” outcome. To reach it, all possible views of a triply transcendent reality need to be ruled out. As I see it, the only way to do so is by finding convincing arguments against one view, or group of views, at a time. Let us call this the Popperian method, after Karl Popper, who famously suggested that science progresses by falsifying incorrect theories (Popper [1959] 2005).14 On this picture, making and discerning progress would be uncomplicated. The more views about the transcendent that can be conclusively rejected, the more progress can be made. However, there are at least two major obstacles with this approach, the first having to do with the total evidence and the second with the range of theories about the transcendent that need to be falsified.

According to Schellenberg, humanity might currently not be in the evidential position required by a Popperian approach. The process of falsification will presumably be evidential, meaning that transcendental views will be rejected based on available evidence. But how can we know if we possess all relevant evidence? In The Wisdom to Doubt, Schellenberg (2007, 15–49) presses this point at length. At our early stage in history, humans should be aware of the possibility of evidence regarding the transcendent that is currently undiscovered, neglected, unavailable, or even undiscoverable, thereby distorting our total evidence. This is a fundamental reason for Schellenberg’s view that evolutionary religion needs to be cognitively modest. If taken seriously, it seems that a Popperian falsification requires a best-case scenario in which there is both enough of the total evidence to make correct falsifications and knowledge that we are in such a favorable epistemic position.

The next worry concerns the range of alternatives. In Popper’s original account, science progresses in terms of falsification, but since the number of possible theories to falsify is infinite, the process will never be completed. There will be no future state where all false theories are falsified and only truth remains (Popper [1959] 2005, 6). The same seems to hold for the religion project, since the number of logically possible theories about the transcendent is also infinite. Even if future species where to prioritize the religion project above everything else, working on it until the end of the universe with the help of supercomputers using all energy in the galaxy,15 they would never be able to complete it. The deep future might be vast, but it is not infinite.

Schellenberg does not agree with this conclusion. His position is that humanity is currently ignorant concerning the range of relevant possibilities and whether the range is infinite: “The truth is that we are not in a position to fix the range of relevant religious options: our ignorance is such that any traditional claim must be seen as possibly belonging to an indefinitely large set of actual and possible competing religious claims, which is at present, and may for some time remain, both nondefinable and nonadjudicable.” (Schellenberg 2009, 22). It seems another “best-case scenario” is required to reach the negative outcome: a scenario in which the range of relevant transcendent alternatives is discovered to be finite. Schellenberg seems to think that such a scenario is possible, but I am doubtful. If there is no transcendent reality, I cannot see how it could ever be discovered that the range of relevant possibilities is any less than the full range of logical possibilities, which is infinite. Historically, discoveries of nonexistence have been limited to alleged empirical phenomena with fixed properties, like unicorns, or scientific entities with crucial functions embedded in theories, like phlogiston or ether.16 The set of possibly existing things that might correspond to the concepts of “unicorn” or “phlogiston” is therefore very limited. But the set of possible views of a triply transcendent reality has no such restraints, and what other restraints could there possibly be?

It might be suggested that the investigation could be restricted to views that are at least minimally plausible, and that the project could be judged as effectively completed once all such views have been falsified. However, minimal plausibility depends on cultural context, and this will surely change by future cultural evolution.17 It would therefore be premature to deem the project “effectively complete” just because all views presently found minimally plausible have been falsified.

A Popperian process of falsification seems like an appropriate way forward in a scenario where no transcendent reality exists. In a best-case scenario where it is known that the evidence available is adequate, there will be discernible progress. However, the (negative) goal of the religion project will remain out of reach.

Concluding Remarks

To be feasible, Schellenberg’s grand idea of guiding religious evolution towards the goal of discovering the truth about transcendent reality requires a way of discerning religious progress of understanding. I have considered the possibility of discerning such progress in some best-case scenarios in the relative near future. Conclusions about the future are speculative by necessity, and given how hard it is to foresee the future, I am the first to acknowledge that my conclusions are tentative at best. With that being said, the prospects for discerning religious progress in the evolutionary near future do not seem particularly bright to me.

In the best-case scenario where an RCF evolves, the intuitive idea that progress could be measured by convergence turned out to be a dead end. If the transcendent is impersonal and manipulable, an ability to manipulate it would certainly be a discernible kind of progress, but presumably manipulability enters the picture too late to be of much use at our early stage.

In a scenario where no RCF develops and there is only cultural religious evolution, it will be hard to discern progress—unless we humans find ourselves in a best-case scenario where a new super-religion emerges that leads to otherwise unexplainable empirical effects, like an unprecedented rise in human flourishing (which would require that the transcendent affects the world directly rather than indirectly).

The best-case scenarios in which it is possible to discern progress do not only depend on the right evolutionary developments taking place, but also on the nature of transcendent reality. For example, if the transcendent is an unmanipulable, personal God who works through natural means (i.e., the view of the transcendent that is by a large margin the most popular today), there will be no progress on any of the considered scenarios.

If no transcendent reality exists, the most straightforward way to proceed is a process of Popperian falsification. To get off the ground, this approach would require a best-case scenario where the relevant evidence is known and where we humans know that we are in these lucky circumstances. While this method would never reach the goal of the religion project, it would at least allow for discernible progress.

At present, we are ignorant about all relevant factors. We do not know if the transcendent exists or what properties is has, and we have no idea whether the future will bring any best-case scenario (or if we are already in one!). This means that we have no idea which, if any, method we should use to discern any religious progress. It therefore seems that we are currently too ignorant for Schellenberg’s idea of guiding religious evolution towards the goal of the religion project to be practically feasible.

Of course, everything might change in the deep future. In X million years, intelligent life might find itself in a position where it is apparent that an RCF is evolving, or perhaps future intellectual discoveries will limit the set of possible transcendent realities. Or maybe something even more beneficial will happen that lies far beyond what we are presently able to imagine. In such case, guiding religious evolution towards the goal of learning the truth about transcendent reality might become a feasible project worth considering. But that future is far off, and it might never come.


  1. Schellenberg’s evolutionary religion has only begun to receive proper philosophical attention. The discussion has mostly been limited to which religious views and attitudes best fulfill its requirements (Elliott 2017; Palmqvist 2019, 2022; Rottschaefer 2016; Schellenberg 2009, 2013, 2019a) and how religious progress can be promoted (McKim 2019; Palmqvist 2022; Schellenberg 2019c, 2019d). The idea of guided religious evolution on which I focus in this article has not been previously assessed. [^]
  2. Of course, religious adherents usually regard their own view as progress compared to other views. Schellenberg is only interested in progress on the philosophical meta-level, which does not presuppose the truth of any specific religious view (remember that agnosticism is the proper religious stance today in his view). [^]
  3. This recent version of the conditions comes from “The Future of Religion” (Schellenberg 2019d, 21). It represents a revision of the original conditions in Evolutionary Religion (Schellenberg 2013, 75). [^]
  4. I do not consider here the idea of religious investigation within the framework of ultimism (Schellenberg 2009, 2013). [^]
  5. It might seem strange to view religion as a project. However, within the broader context of evolutionary religion, the idea makes considerable sense. (Schellenberg himself explicitly combines the ideas of evolutionary religion and the religion project in “The Future of Religion” (Schellenberg 2019d, 13, 21)). If humans are to guide religious evolution to increase religious progress, religion undeniably turns into a project. [^]
  6. Schellenberg is committed to the view that religious reality must be transcendent. If it turns out that religious naturalism is true, and that religiously significant structures or entities exist within the world described by science, discovering the truth about such religious reality falls outside the scope of the religion project (in such a scenario, the discovery of religious truths will instead fall within the scientific project). [^]
  7. For the contrasting opinion that HADD is reliable, see Van Eyghen 2019. [^]
  8. Of course, in a scenario where no transcendent reality exists, a development where humans lose our HADD and stop detecting agents in nature would in some sense constitute a cognitive improvement, but it would not be a scenario with improved cognitive contact with a transcendent reality (because there is none). [^]
  9. An objector might claim that it is perhaps the interpretation of its output rather than the RCF itself that produces the incompatible systems. To sidestep such objections, I assume that RCF contains both an element of observation and one of interpretation. One way to spell out this position is to say that religious experiences are theory-laden. [^]
  10. This is true at least regarding Abrahamitic monotheism. We also find monotheism in Hindu vedantism, sun worship and Akhenaton’s religious revolution in ancient Egypt. The point is that compared to the number of societies with animism and/or polytheism, these outbreaks of monotheism are few and far between. [^]
  11. It might be objected that a false picture of transcendent reality might still allow for manipulation. One need only think of Don Quixote and his fight against windmills to see that one can manipulate reality while being completely deluded. However, the objection can be met by an inference to the best explanation. Even though quixotical delusion is logically possible, it is a much better explanation of a subject’s apparent success in manipulating reality that she is in fact right in her assumptions. Similar arguments are often given for realism in the philosophy of science (see, for example, Miller 2016). [^]
  12. An objector might raise the concern that in attempting to manipulate the transcendent, humans have crossed the border from religion and magic. While not uncommon, this distinction between religion and magic presupposes a very theistic understanding of religion. If the transcendent is not personal, manipulation (in a broad sense) will be the default mode of interaction. [^]
  13. Note that false religions do not necessarily have to contribute to human flourishing for natural reasons. According to philosophers like John Hick (1989), all religions are literally false but still provide some kind of access to the divine (or “Real-an-Sich,” as Hick prefers to call transcendent reality). [^]
  14. To be fair, Popper also thought that one theory can be more “truthlike” or have a greater verisimilitude than its competitors. However, as I am interested in progress in a scenario where there is no positive truth about the transcendent to be found, the idea that different theories can be more or less close to the true picture of the world has little relevance for present purposes. [^]
  15. According to the famous Kardashev-scale, a civilization able to harness all energy in its galaxy (a type III civilisation) is the most advanced form of civilization theoretically possible. [^]
  16. This should not be conflated with when belief in certain entities has waned over time because they no longer seem relevant. It has not really been discovered that there are no demons or witches, and yet such beliefs have plummeted in the last centuries (at least in developed countries). Belief in demons or witches should therefore not be regarded as disproved in the strong sense relevant for Schellenberg’s project. [^]
  17. See, for example, Schellenberg 2019b, where the point is argued that traditional theism has become less likely as cultural developments have changed our conception of what it means to be good. [^]


Alston, William P. 1991. Perceiving God. Ithaca, NY: Cornell University Press.

Alston, William P. 1996. “Belief, Acceptance, and Religious Faith.” In Faith, Freedom, and Rationality, edited by Jeff Jordan and Daniel Howard-Snyder, 3–27. Lanham, MD: Rowman and Littlefield Publishers Inc.

Barrett, Justin L. 2004. Why Would Anyone Believe in God? Lanham, MD: Altamira Press.

Braddock, Matthew. 2016. “Debunking Arguments and the Cognitive Science of Religion.” Theology and Science 14 (3): 268–87. DOI:

Elliott, James. 2017. “The Power of Humility in Sceptical Religion: Why Ietsism Is Preferable to J. L. Schellenberg’s Ultimism.” Religious Studies 53 (1): 97–116. DOI:

Ferretti, Francesco, and Ines Adornetti. 2014. “Biology, Culture and Coevolution: Religion and Language as Case Studies.” Journal of Cognition and Culture 14 (3–4): 305–30. DOI:

Hendricks, Perry. 2022. “How to Debunk Animism.” Philosophia 50 (2): 543–50. DOI:

Hick, John. 1989. An Interpretation of Religion: Human Responses to the Transcendent. Basingstoke, UK: Palgrave.

Joyce, James. 2003. “Baye’s Theorem.” In Stanford Encyclopedia of Philosophy, edited by Edward N. Zalta.

McKim, Robert. 2019. “On Making Religious Progress.” In Current Controversies in Philosophy of Religion, edited by Paul Draper, 27–42. New York: Routledge. DOI:

Miller, Boaz. 2016. “What Is Hacking’s Argument for Entity Realism?” Synthese 193 (3): 991–1006. DOI:

Muyskens, James L. 1979. The Sufficiency of Hope. Philadelphia: Temple University Press.

Palmqvist, Carl-Johan. 2019. “The Proper Object of Non-Doxastic Religion: Why Traditional Religion Should Be Preferred over Schellenberg’s Ultimism.” Religious Studies 55 (4): 559–74. DOI:

Palmqvist, Carl-Johan. 2022. “A Faith for the Future: Why Non-Doxastic Traditional Religion Is the Preferable Form of Evolutionary Religion.” European Journal for Philosophy of Religion 14 (1): 95–121. DOI:

Peoples, Harvey C., Pavel Duda, and Frank W. Marlowe 2016. “Hunter-Gatherers and the Origins of Religion.” Human Nature 27 (3): 261–82. DOI:

Plantinga, Alvin. 2000. Warranted Christian Belief. Oxford: Oxford University Press. DOI:

Popper, Karl. (1959) 2005. The Logic of Scientific Discovery. New York: Routledge. DOI:

Rottschaefer, William A. 2016. “Schellenberg’s Evolutionary Religion: How Evolutionary and How Religious?” Religious Studies 52 (4): 475–96. DOI:

Schellenberg, John L. 2005. Prolegomena to a Philosophy of Religion. Ithaca, NY: Cornell University Press.

Schellenberg, John L. 2007. The Wisdom to Doubt. Ithaca, NY: Cornell University Press.

Schellenberg, John L. 2009. The Will to Imagine. Ithaca, NY: Cornell University Press.

Schellenberg, John L. 2013. Evolutionary Religion. Oxford: Oxford University Press. DOI:

Schellenberg, John L. 2019a. “On Evolutionary Religion (In Interaction with Rottschaefer, Elliott, Dumsday, and Palmqvist).” Religious Studies 55 (4): 575–91. DOI:

Schellenberg, John L. 2019b. Progressive Atheism. New York: Bloomsbury Academic.

Schellenberg, John L. 2019c. Religion after Science. Cambridge: Cambridge University Press.

Schellenberg, John L. 2019d. “The Future of Religion.” In Current Controversies in Philosophy of Religion, edited by Paul Draper, 11–26. New York: Routledge. DOI:

Smith, Tiddy. 2020. “The Common Consent Argument for the Existence of Nature Spirits.” Australasian Journal of Philosophy 98 (2): 334–48. DOI:

Szocik, Konrad. 2017. “Religion and Religious Belief as Evolutionary Adaptions.” Zygon: Journal of Religion and Science 52 (1): 24–52. DOI:

VanderWeele, Tyler J. 2017. “Religious Communities and Human Flourishing.” Current Directions in Psychological Science 26 (5): 476–81. DOI:

Van Eyghen, Hans. 2019 “Is Supernatural Belief Reliably Formed?” International Journal for Philosophy of Religion 85 (2): 125–48. DOI:

Van Eyghen, Hans, and Christopher T. Bennett. 2022. “Did Natural Selection Select for True Religious Belief?” Religious Studies 58 (1): 113–37. DOI: