Introduction

Attention: I am currently working on a new version of this post. This is now out-of-date, though it still largely represents my views accurately. Pending the release of that new post, I will keep this up - but be cautious in citing it as most sections are being significantly revised.

(Note: I am now working on a video version of this post! If you prefer that format, click here to check it out and subscribe to my channel. That being said, the more comprehensive and up to date version of this post will always be here.)

Writing essays explaining why one rejects theism or Christianity has become somewhat of a tradition for nontheists. Ever since Russell’s classic Why I’m Not a Christian, both professional philosophers and people in the online philosophy of religion space have written pieces summarizing their reasons for disbelief [1]. Partially inspired by these posts, and encouraged by a number of friends, I have decided to make a post in a similar vein. In this essay, I will summarize the various reasons that lead me to reject the neoclassical theistic view dominant in Western culture. This essay ended up being much more comprehensive than others I have read. These are complex issues, and I wanted to do my best to address what I thought were the most important considerations for and against theism. Still, a full realization of this project that addressed all relevant considerations would realistically require many book-length discussions, and so I have inevitably needed to leave out objections and arguments I would have liked to discuss. Nevertheless, I have included what are to my mind the most important factors in the current debate.

Two more caveats are worth noting before we proceed. First, the following constitutes my reasons for rejecting theism. I think justification is person-based, and while we can try our best to present arguments for our beliefs, how compelling various considerations seem is often personal. Therefore, while I will occasionally speak of what I think should be compelling to theists, this is only because I try my best to think in the most neutral perspective I can when considering evidence, and these arguments ultimately reflect my personal weighing of the evidence. Second, I am standing on the shoulders of giants. Most of the arguments I bring up in this essay are strongly influenced by philosophers I have read. Throughout this essay, I will cite them wherever relevant. However, one philosopher whose influence emanates across this entire project, and so is worth mentioning from the outset, is Paul Draper. Draper’s general methodology and work has been so impactful on my thinking that this project would have looked very different without his influence, and I would encourage those interested to look into the work he has written that has inspired this piece.

This essay will proceed as follows: In methodology, I will outline the general way I think about and evaluate the question of God’s existence. I will adopt a broadly abductive approach and suggest two primary factors relevant to the question: the evidential value of theism compared to alternatives and its inherent virtue as theory relative to alternatives. In evidence, by far the most significant section, I will outline the data that I find most compelling in favor of naturalism and in favor of theism and weigh the two. I will argue that the evidence in favor of naturalism comes out ahead, and that theistic evidence generally suffers from two inherent issues: underspecification and contentiousness. In theoretical virtue, I will outline factors that make theories inherently virtuous and argue that theism fares poorly. Finally, I will discuss the personal factors that have led me to nonbelief. Below is a table of contents summarizing these sections:

(Annoyingly, blogspot has no good option to create jump links within a post. If you'd prefer to read a version of the post that allows you to quickly jump to certain sections, here is a link to a version in google docs with that capability.)

Introduction

Table of contents

Methodology

Abduction and Bayes’ Theorem

What should we expect on theism?

Expansions to Theism

For Naturalism

Humans

Evolution

Dependence of Mind on the Physical

Cognitive Science of Religion

Success of Naturalistic Explanation

Suffering

Biological Role of Pain and Pleasure:

Flourishing and Languishing:

Triumph and Tragedy:

Geographical Distribution of Suffering:

Particularly horrendous evil:

Theodicies and Skeptical Theism:

Free will:

Soul-making:

Heaven & Defeat:

Skeptical Theism:

Ineffectiveness of Prayer

Religious Diversity

Flawed Religions

Nonresistant Nonbelief

Introduction to Theistic Evidence

For Theism

Necessary & Contingent Existence

Fine-Tuning

Consciousness

Moral Knowledge

Beauty

Religious Experience

Summary of the evidence

Prior Probability of Theism and Naturalism

Simplicity/Modesty

Precision/Informativeness

Fit with background knowledge

Personal Considerations

Summary of the Case

Works Cited

Footnotes

Log of Edits

Methodology

Abduction and Bayes’ Theorem

I follow many modern philosophers of religion in thinking the best way to assess theism is by comparing its merits as a worldview compared to alternative worldviews. In my view, the most clear way to make this comparison is abductively. The general idea behind this kind of reasoning is that we ought to prefer those theories that can account for the data we see and are likely antecedently. Put another way, we ought to look for the worldview that contains the best trade-off between inherent virtue and explanatory power.

What I intend to argue is that there is a worldview, naturalism, which consistently fares better than theism on both its inherent virtue as a theory and its explanatory power. Now, I do not intend to argue that naturalism is the best worldview on offer. Instead, I merely intend to argue that it comes out ahead of theism. So long as there exists a superior worldview to theism, that is a reason to reject it, even if there may be worldview that end up outpacing them both.

In order to analyze what we would expect on theism relative to naturalism, we need to give some account of each worldview. By theism, I mean the idea that there exists an omnipotent, omniscient, and wholly good non-physical person (and yes, I’m aware that some Classical Theists would not regard God as a person, but my critiques are still relevant nonetheless) who is the source of all of causal reality. By naturalism, I mean the idea that causal reality is fundamentally physical, and if any other order of causal reality exists, physical reality explains why it exists.

There are at least two ways to model abductive reasoning. One way, would be to use a kind of abductive schema such as the following (Dawes, 2015 110):

The surprising fact, E, is observed.
N would be a more satisfactory explanation of E than T.
It is reasonable to accept the best available explanation of any fact, provided that explanation is a satisfactory one.
Therefore, since accepting N entails rejecting T, it is reasonable to reject T.

The alternative approach would be to model arguments using Bayes theorem, a mathematical way to calculate conditional probabilities. In its simple form, Bayes theorem can be stated as follows, where ‘Pr’ stands for probability, ‘‘|’ can be read as ‘given’, ‘H’ stands for hypothesis, and ‘E’ stands for evidence:

Pr(H|E) = (Pr(E|H) *Pr(H))/(Pr(E))

The key takeaway from this theorem is that the way we update the probability of a theory based on new evidence takes into account both the evidential support the new data lends and the prior probability of that theory. For instance, imagine I have a coin I found on the street and I want to determine if it is a loaded coin (weighted towards heads) or a fair coin. Suppose I flip it once and it comes up heads. If my coin is loaded, this is a more likely observation to make than if it is heads. Yet, it seems wrong to conclude that my coin is loaded on the basis of this initial flip. And you would be wrong to form this conclusion: though the fact that that coin came up heads is more likely if your coin is loaded, prior to that observation it is far more likely that a random coin that you picked up on the street is fair than that it is loaded. This extremely low prior probability that you found a weighted coin might be overcome, though, if you flipped heads enough times in a row!

How can we apply Bayes theorem to the kind of considerations at play in my paper? My arguments could be summarized formally using Bayes theorem as follows, where ‘N’ is naturalism, ‘T’ is theism, ‘E’ is the evidence under consideration, and ‘K’ is background evidence [2]:

Therefore,

Pr(N|E&k) > Pr(T|E&k)

The above analysis, there are two key factors at play: the evidential strength of the data at hand and the prior probability of naturalism and theism. In the first part of this post, I am only going to be addressing the evidence for and against theism and naturalism. When making these judgements, however, we also need to take into account the background evidence. What is background evidence? Whenever we make a judgment about the probability of a particular piece of data on a theory we do it with some set of prior facts in mind. For instance, it would not make much sense to talk about the probability that we have a fair coin if we did not have in our background facts about the existence of coins and how they work. However, how you delineate the background turns out to be tricky in certain cases. This is because one has to be careful not to double-count data. For instance, imagine that you were trying to measure the evidential effect of the data of conscious suffering on theism and naturalism, but that you held the theory of evolution in the background, where the content of this hypothesis included not only that complex life evolved from simple life, but also that natural selection leads to predation, illness, decay, and death. Obviously, these facts entail the existence of conscious suffering, and so to evaluate the evidential effect of suffering on theism they cannot be included in the background (or they’d already be counted!). To see an instance where a popular apologetics channel made this mistake, see my video responding to CapturingChristianity. To avoid this problem, we need to use antecedent probability, which is the probability that a particular datum is true independent of the observations and testimonies that enable us to know the truth of that datum (Draper, 1989). In other words, we need to put aside the knowledge that enables us to know the truth of that datum and ask what we would expect on theism and naturalism prior to that knowledge.

Though it is clear what we should not include in the background, what exactly we should include in the background is up for dispute. In The Existence of God, Richard Swinburne suggests that we start with the more general facts and move towards specific facts, including the general facts in the background after evaluating them. In Theism and Explanation, Gregory Dawes suggests that we include everything in the background apart from the explicit facts we cannot include. My view is closest to something like the following: the background should include everything that is relevantly epistemically independent from the data in question, and that is not an unsubstantiated, controversial addition to one theory (see the section on theistic expansions for more on this latter point). Thus, when evaluating my arguments, it is crucial to remember that certain facts are always being taken for granted. This should clear up objections of the following form: you argue that datum x is more probable on naturalism than theism, but neglected to mention related datum y, which is more probable on theism than naturalism. The answer to these objections is that all the arguments here contain an implicit 'other evidence held equal' clause, and that we will get to the relevance of datum y later on. However, I do not intend to settle debates around what to include in the background or whether priors are subjective or objective (a controversy that will be relevant later on in the post) here. Though I fluctuate on this issue, at this stage I think Bayes theorem is a useful model for abductive reasoning, and we ought to treat it as such - but just thinking generally in terms of abduction and theory comparison is better than getting bogged down in the details of controversies over Bayesian epistemology.

What should we expect on theism?

Now that I have discussed the formal way in which my arguments might be formulated, we need to address an immediate problem in the following analysis: what exactly should we expect that God ought to desire and thus instantiate. This question is notoriously controversial. For some parts of my discussion, the likelihood comparison between theism and naturalism will not require a very fine-grained analysis of God’s desires, but in other cases it will be necessary to have a more clear understanding of what specifically to expect on theism. Thus, it is worth establishing the general guiding principles I will use here.

I think the most natural and immediate intuition about what God would do is that an omnipotent, maximally good being would instantiate the best state of affairs possible. Put another way, a perfect being should create the best of all possible worlds. We can state this principle as follows:

Strong Principle (SP): God will create the most intrinsically valuable world that God can create.

I am sympathetic towards SP, but I recognize that many reject it [3]. The most common argument against the principle is that since there are infinitely many states of affairs, there will always be a better world that God could have instantiated and so there is no best possible world among those that God can instantiate. Rather than trying to resolve this dispute, I will suggest a much weaker, alternative principle that we can use alongside SP to try to evaluate what to expect on theism (adapted from McKim, 2015):

Weak Principle (WP) [4]: God will avoid all states of affairs that frustrate flourishing and virtue such that their deepest and most important aspects are not realized. In short, God will avoid disaster.

I think this ‘disaster avoidance’ principle is weak enough to be amenable to naturalists and theists alike. If we are to expect anything from a perfect being, surely they should at least ensure that the deepest aspects of flourishing are realized [5].

Note that if we abandon any sort of limiting principle on God’s actions, theism will end up having no explanatory value. We need some way of separating out the states of affairs that are antecedently expected on theism from those that are not, or else theism will not be predictive at all. I suggest that my two principles are at least plausible attempts at this, and any theist who disagrees will need to suggest their own if they want to avoid theism failing in principle as an explanation.

When evaluating the data I do in the course of this piece, I will do so on the basis of both SP and WP wherever they are relevant. While I lean towards SP, I think my reasons to reject theism can be motivated even on WP and so even assuming the bare minimum I would expect on theism, those expectations remain consistently violated.

Expansions to Theism

One more preliminary matter worth addressing is the issue of how to evaluate mere theism vs. ‘expanded theism’. The idea here is that while the omnitheistic claim on its own may succeed or fail in explaining various data, most theists believe much more than the bare theistic claim: they want to add on other auxiliary hypotheses to theism that help build their worldview. For instance, Christians believe a whole set of historical, social, and theological claims not entailed by bare theism. Obviously, this post would be far too long and convoluted if I were to go through every possible expanded story theists might try to add to theism to account for the data in this post. Extended discussions on various theodicies, for example, are best left for another post. Therefore, I will do my best to address objections to various arguments as best as possible in this post but lest it become book-length I’ll refrain from indulging every possible theistic expansion defenders of theism might propose. There is one important point I want to make about theistic theory versioning, though, and it is as follows:

One may be tempted to think that the success of expansions to theism hinges solely on the extent to which they overcome the evidential problem faced by T, that is, that if Pr(E|T&A) ≧ P(E|SP) then ‘A’, the auxiliary hypothesis, has done its work. However, such an analysis would crucially ignore the probability of the expansion itself on T: Pr(A|T). Indeed, as Robert Adams and Paul Draper have pointed out, the correct way to evaluate the effect of an auxiliary hypothesis on T’s fit with the data is by using the weighted average principle, formulated as follows, where E is the data to be explained (Adams, 1985, Draper, 1989):

Put in regular language, WAP says that successful theistic expansions are those which account for the data well and are independently likely stories on theism antecedently. At this point, the theistic theory versioner may want to claim something like the following: ‘I am not interested in defending T, only my specific version of T (say evangelical Christianity). Therefore, WAP does not apply to me because I can simply evaluate the probability of the data on evangelical Christianity without ever considering T’. This response, however, would crucially ignore the fact that evangelical Christianity entails T, and so any reason to think that T is false is thereby a reason to think that evangelical Christianity, or any other specific version of T, is false (Draper, 2004). That is, unless evangelical Christianity can meet the crucial conditions just outlined. Therefore, the theory versioner cannot escape the burden of WAP.

Having now addressed the relevant preliminary considerations, I will list all the major facts that I think favor naturalism over theism, before addressing those facts that allegedly favor theism over naturalism.

For Naturalism

Humans

The existence of human beings is a particularly interesting piece of data because it is rarely discussed and, if anything, that life and consciousness exist generally is usually taken as evidence for theism. In fact, I think the fact that human beings exist specifically is one of the strongest pieces of evidence against theism [6]. Note then that here I am going to take it for granted that life exists (we will consider the relevance of this general fact later on), and ask what form of life we ought to expect on theism and naturalism.

As already mentioned, if a morally perfect God exists we should expect that they should value the exemplification of the best kinds of goods possible: the deepest values and states of being. At least, if we should not expect this, we should expect that they would avoid creating any state of affairs that would prevent the most important aspects and kinds of values and flourishing from being realized.

By way of introduction, consider the following passage in an excellent and underappreciated paper by Mark Walker:

“Imagine in the year 2100 that a global government issues an edict proclaiming that every child born to a human outside of the capital city will be of a new chimerical species. ‘Chuman’, as this species is known, is created from a human-chimpanzee cross. Chumans are mentally and physically challenged in comparison to humans, and have been genetically altered to have a strong propensity for violent outbursts (a propensity they wish they did not have). Clearly, such an edict would be morally wrong for any number of reasons. Indeed, implementing such a horrendous evil could well eclipse every other evil done by humanity—and this is certainly saying something. Picture the painful collapse of civil society as those born prior to the edict die off, and only chumans remain outside the capital city. For example, without the physical and mental capabilities necessary to master modern agricultural technology, we must imagine that most chumans quickly starve, and the world outside the capital falls into unimaginable barbarism, cruelty, and suffering. Let us suppose it is a mystery why the world government insisted on this action. One hypothesis is that they did so to ensure that the capital remained in perpetuity the center of power on the planet. Another hypothesis is that the motivation was to allow for a fuller expression of their benevolence and love: those humans remaining behind the capital walls could make the occasional sojourn into the more barbaric territory to assist and offer their love for chumans. In either case, surely we would be inclined to think that if the world government had any concern with doing what is best, they would not have insisted on the creation of chumans in the first instance. The world would have been much better if the world government had permitted those outside the capital to have human children. The parallel with God’s creation decision is perhaps apparent: when we think of God as the standard for a perfect being, then, comparatively speaking, it seems that our natures are at least as defective in comparison with His nature as chumans are in comparison with ours. Accordingly, it seems that God too is morally culpable for creating us with defective natures; defective, that is, in comparison with His nature. So, by analogous reasoning, God should create ontologically equivalent beings— other gods—rather than humans” (Walker, 2009 1-2)

Clearly, the hypothesis that the capital government that created chumans is a particularly loving, ethical one is a very bad one. And, the gap between God and us is far greater than the gap between humans and chumans. Note that Walker does not actually need to hold that God should create beings ontologically equivalent to Him, but rather just that God should make more impressive beings than humans. Human beings are certainly impressive relative to most other life known to exist. Our ability to engage in moral decision making, to reason, and to experience complex emotions is far above most other organisms on earth. However, we are far less impressive than the best kind of life an omnipotent being could create. As Paul Draper has noted, humans are “in terms of intelligence, a hair's breadth away from monkeys" (Draper, 2008). An omnipotent being with infinite resources could clearly create life that far transcends our own. A few specific examples of our limitations are worth noting:

Firstly, while we are certainly constituted to be selectively altruistic in some ways, our moral characters are highly flawed and often deficient. We are extremely prone to violence, bias, barbarism, jealousy, greed, and a litany of other vices. Secondly, we are highly intellectually ignorant beings that succumb to many cognitive limitations and are more likely to rely on cognitive shortcuts than think rationally. Thirdly, there are clearly many levels of flourishing we are simply incapable of experiencing quantitatively. For instance, imagine life whose architecture is such that their experiences of all positive emotions are hundreds of times more intense than our own. Fourthly, and most importantly, we are unable to realize the deepest level of and essential qualitative aspects of the most important values. Anyone who has been in altered states of consciousness knows that even within our limited biological architecture there are states of experience whose valuable aspects are far deeper than what can possibly be achieved in the course of normal human life [7]. Imagine the kind of experience that organisms with much more impressive capacities could engage in. I will provide three examples:

First, consider empathy. The current way we empathize is inherently limited and superficial on multiple levels. We know from psychological research on empathy that it is subject to many distortions and biases that cloud our ability to do it genuinely: for instance, we have trouble empathizing with those who are dissimilar to us. More importantly, we only empathize by generalizing from our own experiences, trying to imagine the perspectives and experiences of others. This means that our empathy is bound to be an artificial, biased extension of ourselves, rather than a true understanding of another person. However, empathy could have been deeper in multiple ways that overcome these limitations: firstly, imagine beings who were free of the many biases and distortions that cloud our empathetic thinking. These beings would still generalize based on their own experiences, and so still empathize very superficially, but could at least do it more genuinely than humans. Secondly, consider alternative life whose mental architecture would be not only free of these inherent biases when engaging in empathy, but could actually temporarily feel what others feel, experiencing the same emotions and sensations of other beings. While these beings would experience a much truer form of empathy than humans, they still have one limitation: though they might experience the immediate emotions of another, their interpretation of these feelings would still be clouded by their own perspective. Thus, thirdly, and most importantly, consider beings who could overcome this final limitation, placing themselves into the perspective of others temporarily, and so experience true, lucid empathy. These organisms would not just generalize based on their own experiences, but be able to temporarily have the complete experience of another, and so actually appreciate the nuance and breadth of their emotional/cognitive state, later incorporating that perspective into their own understanding.

Second, consider rationality. I already mentioned early that we are highly ignorant in many ways. However, our ignorance goes beyond our tendency to cognitive bias: it is likely that there are parts of the universe that we are fundamentally incapable of understanding. Life with much more impressive capacities than our own could certainly uncover the deepest aspects of the universe, engaging in the scientific enterprise in a far more impressive way than us. Third, consider love: while we are able to experience strong emotions of love towards other beings, we are rarely, if ever, able to experience true unconditional love towards others. Unconditional love is, in my view, the deepest kind of love possible because it is one free of attachment: it is one not dependent on one’s own desires, but love purely for another to be happy without condition. We are usually so clouded by delusions and selfishness that such a state is extremely difficult. Unlike in the case of empathy, it is arguable that the most dedicated people, for instance Buddhist monks, are able to approximate if not achieve this state. Though, whether we can really ever let go of all conditions on love is highly questionable. At the very least, for most of us this kind of state is impossible. Beings with a far more deep, impressive ability to love would be capable of reaching this kind of state, allowing them to appreciate an aspect of love that most of us can never achieve.

These are just a few examples of many ways that the way we are constituted limits our ability to realize the deepest kind of experience. It strikes me as extremely probable that there are also many virtues and experiences that other possible life could exemplify that we cannot even describe or imagine due to the limits of our own mental architecture. This is especially true when one considers that an omnipotent God could not only create any kind of biological life they please, but life that transcends biological limitations. An omnipotent being with infinite resources could clearly create life that is far more impressive than ours.

What ought we to expect on theism antecedently, then? If we buy into SP, then we should expect that God should create the most impressive kind of being they can create, and the existence of humans is plainly far out of line with this expectation. There are certainly possible degrees of flourishing far above what humans can experience and which could exist in other beings. Even on WP, the creation of humans relative to possible alternative life plainly constitutes disaster by my lights. Since WP does not demand that God create the highest quantitative level of flourishing or value that they can, we should not antecedently expect on WP that God maximizes, for example, the intensity of positive emotions. However, we should expect that God makes it such that whatever beings exist can appreciate the most important values and deepest aspects of those values. Even making the generous assumption that humans are constituted to experience the most important values that could exist, we clearly consistently experience the deepest aspects of these values. I already mentioned three examples of this in the cases of our empathy, rationality, and love. And, the complaint cannot be lodged against this argument that there is always a deeper level of these values that can be achieved. For instance, there is no deeper level of empathy than simply being able to truly and completely take on another person’s perspective. Indeed, this deeper kind of empathy just is what empathy truly exemplified would be. And this is just one example of the many ways in which we are constitutively limited in truly appreciating the deepest aspects of the most important values.

An alternative society with life far beyond our capacities could be engaged in an exercise of soul-making, virtue seeking, and raw experience unimaginably greater than our own. This kind of world is certainly far more in line with what we ought to expect from a perfect being than what we observe. Thus, the existence of humans is extremely improbable on theism. There are presumably an enormously large number of types of beings God could create, and amongst that selection, beings so unimpressive as humans are particularly unlikely.

What about the antecedent likelihood of human beings on naturalism? Since on naturalism all organisms are the process of gradual physical processes without supernatural guidance, and each level of complexity in those organisms requires that life surpass certain filters (ie. the emergence of sexual reproduction, survival for long enough to harness all resources on a home planet), the more impressive life is the less likely it looks on naturalism. Depending on how likely life should be to surpass each filter on naturalism, more impressive life will look relatively less likely. In my opinion, human life has already surpassed a great many filters and thus is quite unlikely. However, given enough physical space and opportunities that somewhat impressive life would emerge strikes me as likely. Irrespective, the probability of human life is clearly far more probable on naturalism than it is on theism, since human life is positively primitive compared to what is possible on theism. Importantly, on naturalism, life that transcends biological and technological limitations is impossible, whereas on theism it is not only possible but expected.

In sum, the existence of human beings is highly improbable on theism because human beings are enormously limited in experiencing the possible depths, nuances, and breadth of various good states and virtues that would have been possible in alternative forms of life. Human beings in virtue of being somewhat impressive are also fairly unlikely antecedently on naturalism, but we are the kind of flawed beings that underlying physical processes could produce given enough time and space [8].

There are two objections to my argument that I want to consider before moving on. The first is taken from a paper by Robert Adams (Adams, 1972). Adams makes two points that are relevant to the argument here: First, that it is implausible to think that it is wrong to bring into existence a being less impressive than one could have brought into existence. For instance, you would not have wronged anyone by failing to take a pill that would turn your child into a superhuman. Second, a grace-based argument that God’s love is one that is not based on the ‘merit of the person loved’. Because God loves not on the basis of the excellence of a particular creature, but for love’s sake, He has no obligation to create the most excellent creatures. God’s grace leads Him to love creatures irrespective of their excellence. Both Mark Walker and William Rowe have given excellent responses to Adams’ arguments, and I would direct readers to their pieces for a comprehensive response (Rowe, 2004, Walker, 2009). Here, I will just summarize a few important points:

Adams’ first argument does not constitute a serious objection to mine for two reasons: firstly, I did not rely on the principle that Adams finds implausible (namely, that it would be wrong to bring into existence a creature less excellent than one could have brought into existence). Instead, I relied on WP, which is far more plausible. Secondly, what we are concerned with in this argument is what is antecedently likely that God would do, not what it is possible for God to do. And, even if it is possible that God might make beings like us in the sense that it would not be absolutely wrong to do so, we clearly are not the kind of beings that are likely on theism. What a “reasonable person would expect beforehand” on the hypothesis of theism is not that a being so limited and unimpressive as humans would exist (Draper, 2008).

What about Adams’ second argument based on grace-based love? A serious issue here is that the connection between God’s grace and his lack of an obligation to create greater beings than humans is extremely tenuous. To use the analogy of ‘chumans’ from earlier, this ‘grace-based’ love would imply that we should not love chuman babies any less than human babies. But surely it would not imply that we should not prefer to create more humans than chumans. The reason why God would choose to create the world with more excellent creatures need not be because he loves them more than he would love the less excellent ones, but rather that the world with more excellent creatures would contain more of the deepest kinds of goods, be morally superior, allow for more soul-making, etc. Another way to put this is that God’s love towards beings is a neutral factor with respect to deciding on what world to create: God would love whatever beings He created equally across worlds. The world with beings far superior to ours, however, has additional reasons for being chosen. And, God obviously has other reasons than love for creating creatures. If all God cared about is whether he loved the creatures He created, a species of barbaric monsters that could not improve would be equally likely to be instantiated than any other being God might create. Clearly, this is implausible. God might love those beings as much as he loves other beings in virtue of His grace, but he has other reasons to not create them.

To be fair, Adams himself clearly thinks there is more to God’s decision to create the world than grace-based love, since he thinks the world also needs to satisfy certain conditions to be instantiated by God. However, I have argued that God’s creation of humans violates WP, and so does not even meet the minimum principles God should abide by. I have not argued that God is obligated to create the most impressive beings, just beings more impressive than humans.

Additionally, regardless of whether Adams’ suggestions can apply to my specific argument, I think they cannot be right. This is because his arguments rests on the idea that God has no obligation to create a certain kind of creature or world, in the sense that he has not harmed any being by failing to do so. This, however, overlooks the fact that a perfect being almost certainly has more to do than simply failing to violate any duties: a perfect being would also perform supererogatory actions.

A final problem with Adams’ argument is that, if it were correct, it would destroy theism’s predictive power with respect to the creation of humans in a different way. If the fact that God would love any being he creates equally leads us to fail to expect any particular degree of impressiveness for those beings on theism, then we have zero reason to expect any particular kind of being. Therefore, the probability that God would choose humans as opposed to any kind of being becomes extremely low in virtue of the specificity of that outcome.

The second possible objection I wish to address would suggest that God has in fact created the far more impressive life that I have alluded to. After all, the physical universe is very large: perhaps our (far) biological superiors are out there, or perhaps God has created beings that transcend biology and exist in some plane of existence beyond our own. My main issue with this objection lies in its reliance on speculation. The only real data we have is our own existence, and speculating that there might be more impressive life out there is of no help, since such a possibility cannot count as data. Sure enough, there could be more impressive life that exists which is relatively more expected on theism. There could also be much more unimpressive life out there which is even less expected on theism than our own life. Thus, the speculative possibilities cut in both directions and cancel out. A second problem with this objection is this: even if God has created more impressive life, the question remains as to why humans exist at all. Why would God not put that more impressive life in place of us? By hypothesis, these possible beings are far greater than us on every dimension of experience and virtue. By hypothesis, there is absolutely no reason to prefer humans over these alternative beings in any sense. Insisting that only humans could occupy our particular sector of the universe is no help, since it is God who sets up the design of the universe and even given that design there are clearly more impressive biological beings that could occupy earth. Thus, even if far more impressive beings do exist, the evidential problem still remains with respect to our existence.

Evolution

By ‘evolution’ I am referring to the fact that complex life evolved from more simple life on the basis of trans-generational genetic change (Draper, 2015). The truth of evolution, thus stated, is still denied amongst some fundamentalists. However, even those who favor the theory of intelligent design or theistic evolution likely accept this claim [9]. Indeed, I think the evidence for evolution in the face of modern science (including not just biology but applied biology in other fields ie. psychology) is so overwhelming that its denial is ludicrous.

Now, on naturalism, if complex life does come about, it must come about via some ultimately physical process such as biological evolution. Since causal reality bottoms out in physical mechanisms on naturalism, there is no way for complex life to emerge other than ultimately physical processes such as biological evolution. Additionally, there is the significant fact that we lack any plausible alternative candidate to evolution on naturalism (Draper, 2015). On theism on the other hand, while God could choose to bring life about by some physical process such as evolution, they need not. They could choose any number of non-natural means to bring about life, such as simply ‘popping’ life into existence miraculously, or using some gradual ‘special creation’ process similar to that described in the book of genesis. Therefore, antecedently, the probability of biological evolution is lower on theism than naturalism in virtue of the fact that on theism God could use any of the non-natural options for instituting life in addition to the naturalistic options.

While SP and WP are not even necessary in this case to establish the probability judgment, one could bring them in to try to show that God has reasons to use non-natural means to bring about life relative to biological evolution, and so increase the strength of this judgment. And, indeed, we have at least some reason to expect that God will prefer to intervene miraculously in the case of creating life but not in the case of other features of the universe, because life has tremendous significance on theism. Thus, God has reasons to mark-off life as special relative to other parts of the universe. God also has significant reasons not to use evolution, reasons we lack on naturalism, namely, that evolution is a highly tortuous, inefficient process that results in enormous amounts of suffering. However, this latter fact is something we will consider on its own in a later section.

One kind of objection to this argument would try to add some auxiliary hypothesis that God only acts through physical causes to theism. However, I think this argument is doomed to failure simply because there is no good way to argue that Pr(God only acts through physical causes|theism) is 1 antecedently. Furthermore, this kind of auxiliary hypothesis would require abandoning the idea of supernatural intervention in the world, a cost I presume most theists will not want to pay [10].

Dependence of Mind on the Physical

The next fact to consider is that our mental life is highly dependent on physical processes. Our energy, sleep, appetite, sexual drive, anxiety, mood, and essentially any other function one might think of can be directly tied to the functioning of particular physical brain states. However, the mind’s dependence on the body goes much deeper than this. Core parts of our identity such as our rationality, values, personality, memory are all also highly dependent on physical processes. Damage to certain parts of the brain can change your fundamental value system, and your reasoning about morality (Burns & Swerdlow, 2003, Koenings, 2007). The gradual decline of neural functioning later in life makes people gradually lose their memory and personality. Certain mental disorders alter brain functioning in ways that cause peoples’ entire belief structures to change (Lavigne et al., 2019). Here is one passage summarizing just a tiny fraction of the evidence for these changes, from Dove and Elpidorou’s Consciousness and Physicalism:

"The pull of consciousness’ uniqueness is strong, but one cannot be left unimpressed by a preponderance of evidence, both everyday and scientific, that points to the view that whatever consciousness is, it is intimately related to our biological makeup. One too many drinks and consciousness is affected; a blow to the head and consciousness might be gone. Bilateral damage to central thalamus and thalamic manipulations (such as reduction of thalamic metabolism and blood flow) can make consciousness disappear (Alkire & Miller, 2005; Bogen, 1995; Churchland, 2013; Laureys, 2005; Posner, Saper, Schiff, & Plum, 2007); fusiform lesions lead to prosopagnosia (Barton, Press, Keenan, & O’Connor, 2002 ; Kanwisher & Yovel, 2006); lesions in V1 (primary visual cortex) can cause blindness or severe loss of vision (Leopold, 2012); and middle temporal lesions can result in the loss of the visual experience of motion (Churchland, 2013; Ramachandran & Blakeslee, 1998). We know that consciousness is either lost or severely affected during generalized absence seizures (i.e., the abnormal and hypersynchronous discharge of neurons; Blumenfeld, 2011) and that the areas most affected by such seizures correspond to the ones that are altered in sleep and anesthesia (Tononi & Koch, 2008). Consciousness may have its own spectral allure, but the predictive and manipulative power of our brain sciences is a constant reminder that consciousness is not a free-floating, ungrounded phenomenon."

What all of this shows, at least minimally, is that our minds are highly dependent on the physical structure of our brains [11].

On naturalism, if minds are to exist, they must be ultimately dependent on or equivalent to physical mechanisms. The reason for this is that since on naturalism causal reality is fundamentally physical in nature, minds must ultimately depend on physical processes. On theism on the other hand, God could institute it such that minds are highly dependent on physical mechanisms. However, God need not: he could also create beings whose minds are not causally dependent on physical processes, or for whom only some parts of their mind are dependent on physical processes. The very fact that these latter options exist on theism but not on naturalism shows that Pr(Physical dependence of minds|Naturalism) > Pr(Physical dependence of minds|Theism). Furthermore, theism starts off from a radical metaphysical dualism wherein the fundamental mind is nonphysical, giving us inductive reason to infer that other minds will be likewise nonphysical.

Cognitive Science of Religion

The next piece of evidence I want to consider is the new wealth of evidence from the cognitive science literature regarding the evolutionary origins of religious belief. I have already written an extensive account of this argument for a research project during my undergraduate education, and I am trying to rewrite that draft into a version suitable for publication somewhere. For now, I am going to post a link to the initial draft of that paper. I also appeared on Emerson Green’s youtube channel to discuss the paper, so readers might look there for resources on this argument as well. If I do eventually get my draft of this paper published, I will edit this section and post a link to the article.

Success of Naturalistic Explanation

By the ‘success of naturalistic explanation’ I mean that the materialist research project of explaining phenomena in terms of physical causes and properties has been remarkably successful. Consistently, folk theories that appeal to spooky substances or supernatural agents have been swept by the wayside and replaced by physical theories. We now know the weather is not caused by agents but by differences in physical properties between places, that diseases are not caused by demons or curses but by microorganisms, and that life evolved not due to the sculpting of beings from dust but gradual changes to the genetic makeup of organisms.

In Theism and Explanation Gregory Dawes quotes the famed biologist Charles Lyell as follows: “It is not from enquiries into the physical world, present or past, that we gain an insight into the spiritual; we may arrive at conclusions unwelcome to our speculations.” Commenting on this quote from Lyell, Matthew Baratholomew says the following: “it is a revealing comment. Before 1859 natural philosophers in Britain had confidently believed precisely the opposite. They were certain that enquiries into the physical world were bound to elicit clear insights into “the spiritual,” insights so unambiguous that they could be used as a foundation for a defence of the Christian faith. In 1859 that foundation turned to dust, and Lyell’s brief statement can stand as an emblem of his recognition of that disaster.” (Dawes, 2015 132)

I do not think most theists would be too inclined to deny the success of the naturalistic program either. Indeed, many theists have sought to accommodate the findings of the sciences into their theological framework, rather than placing them in opposition. And I agree with this project in the sense that I do not think the natural sciences directly conflict with sophisticated theological models. However, I do think the success of the materialist research program nonetheless constitutes extremely strong evidence against theism. If naturalism is true, then since physical reality is fundamental and explains why any other order of causal reality exists, the materialist research program must be the successful one. If theism is true, there is no reason why the materialist research program had to be successful. Any number of supernatural or non-physical research programs could have been successful, either in place of the materialist program or alongside it. God did not need to work through physical causes or make it such that humans only successfully explained phenomena in terms of the physical.

Of course, some theists will be inclined to deny that the materialist research program is the only successful one. They may claim that in the realms of mentality, paranormal investigation, or complex biological systems there is still room for a supernatural research project. However, I simply deny that they would be correct in asserting that such programs are successful. The tides of history have and continue to point away from these programs and towards materialism. Even if these programs did have promise, I think any well-informed person would be hard pressed to deny that clear dominance of the materialist model.

Many theists would likely want to, at this point, appeal to the idea that God works through these secondary, natural causes, setting up the natural laws with the wisdom to know how the world will play out given his initial design. However, God does not (or at least usually does not) intervene to contradict this initial design, since it is the very functioning of these efficient causes that he planned out by setting the initial natural laws. I have already talked a little about this objection. The main problem with it is that while it is all well and good to posit this auxiliary hypothesis, we have no good antecedent reason to expect that God should work only through natural causes. That is, prior to our observation that causality tends to work through physical mechanisms, we have no good reason to expect that God will choose only to work through these physical mechanisms. Indeed, we have significant reason to think God will not [12].

A second objection I want to clear up would say that the probability that any research program would be successful on naturalism is incredibly small. This objection misunderstands the structure of the argument. As Paul Draper has noted in an article where he further motivates this kind of argument, this argument is “interested in the evidential significance of the success of science, given that there are explanations of one sort or another for natural phenomena. If the scarcity of brute facts in nature can somehow be shown to support theism over naturalism, so be it. But given that scarcity—given that natural phenomena typically do have explanations—the fact that so much in nature is known to have a naturalistic explanation (and no part of nature that could have a naturalistic explanation is known not to have one) strongly supports metaphysical naturalism over theism. After all, things could have turned out differently.” (Draper, 2008 299-300). As an alternative gloss, whether or not there will be a successful research program at all looks like it will be determined by whether rational agents exist in an intelligible universe. Sure enough, these facts are some of the most problematic for naturalists in that it looks plausible that theism explains it better than naturalism. And, indeed, we will consider these very facts later on when examining the evidence for theism. However, at the point that we are considering the evidential significance of the success of science, those facts will already be part of the background. And from the fact that they favor theism over naturalism, even very strongly, it would not follow that given those facts and so given that a research program exists at all, the success of natural science does not favor naturalism over theism.

Suffering

I thought for a while about how to write this section, given that this is a piece of evidence so commonly discussed that much of what I want to say has already been said. What I have decided to do is to summarize some of the most important points about the distribution and quality of suffering we see, closely falling the work of Paul Draper and a few others, and briefly explain why I think they count against theism. See (Draper, 1989) and (Draper, 2017) for the primary citations for this section, and for a more complete discussion.

Biological Role of Pain and Pleasure:

First, consider that pain & pleasure usually contribute systematically to biological goals such as survival and reproduction. We feel pain when it is evolutionary useful to, and pleasure when it is evolutionary useful to. If naturalism is true there is no reason to expect that pain and pleasure would not be systematically biological, in fact, we have reason to infer from the fact that all of our other functional states and feelings are deeply connected to biological goals that pain and pleasure will be no different. Since evolution designs us with such goals in mind, pain and pleasure will likely also align with these goals. There is one highly important difference, however, between pain and pleasure and other biological systems. Unlike most biological systems, pain and pleasure have moral significance: pain is intrinsically bad and pleasure is intrinsically good. On naturalism, this difference between pain and pleasure and other biological systems is irrelevant, since reality is fundamentally indifferent to the lives of sentient creatures. On the other hand, if theism is true, we have substantial reason to think that pain and pleasure will be treated differently than the rest of our biological functioning. Since pain and pleasure have moral significance, God has reason to instantiate them differently from the rest of our biological systems. For instance, God has reason to produce pleasure in the absence of biological utility, and God would have moral reasons for producing pain and pleasure, reasons which could conflict with pain and pleasure contributing to biological goals. If anything, we should expect that pain and pleasure, since they have moral significance, should contribute systematically to moral goals on theism.

What about pain and pleasure that is produced in the absence of biological utility? Often, pain and pleasure is biologically gratuitous in that it produces more suffering or pleasure than is necessary to achieve biological goals. First, note that on theism we have much greater reason to expect creatures to experience happiness rather than pain. Yet, biologically gratuitous pain is extremely common: think of the level of pain experienced by an animal set on fire. Second, biologically gratuitous pain and pleasure do not appear to have any connection to moral goals, something we would expect on theism and not naturalism. Pain does not impressively contribute to soul-building or virtue, nor does pleasure even necessarily contribute. Instead, these systems often work against the building of virtue: for instance, consider the key role pleasure plays in motivating the vice of addiction. Instead, biologically gratuitous pain and pleasure usually arise from the imperfect tuning of our biological mechanisms. For instance, we often experience biologically gratuitous pain due to the “failure of our organic systems to function properly,” such as pain caused by an organ failure (Draper, 1989). These pains are not perfectly biological useful, but they are appropriate in that it makes sense for us to feel pain in response to damage to our organic systems, and the excess pain is merely an accident resulting from a lack of fine-tuning—something we would expect if indifferent nature was at play here, but not expect from a God who could fine tune the development of biological systems as they wished.

Flourishing and Languishing:

Second, consider the fact that sentient organisms languish at extremely high rates and experience vast quantities of suffering of many kinds over the course of their lives.

The content of naturalism and theism alone already give us some reason to disfavor theism with regards to these facts. If theism is true, reality is fundamentally good and caring, and so we have reason to expect that organisms will flourish (not languish) at high rates. On the other hand, since reality is indifferent on naturalism, we do not have any positive reason to think languishing will not occur.

Now, suppose that we take into consideration evolution in the background. Recall that the probability in question is antecedent, so the explicit facts about the evolutionary process that involve the suffering of sentient organisms (ie. that widespread predation amongst sentient creatures occurs) will not be part of the background. What can exist in the background is the fact that we know that non-conscious agents such as plants have come about by a gradual, competitive process and have languished and gone extinct at extremely high rates. Now, minimally, we have no reason to think that sentient creatures will not also come about by this process and experience lots of languishing on naturalism. Indeed, we have inductive reason to infer that they will, given that non-sentient agents came about via this process. On the other hand, if theism is true, we have very strong reason to think that God would treat sentient agents differently from nonsentient agents, due to the possibility of intense suffering in the former case. Indeed, since the use of natural selective, competitive mechanisms would lead to the mass suffering of conscious creatures, God has extremely strong reasons to make complex conscious life come about in some alternative fashion. And, importantly, unlike on naturalism, God actually has many non-natural options for bringing about this life that would not involve suffering (such as a kind of special creation process). Even if God is to use evolutionary processes for whatever reason, we have reason to think they would guide this process such that sentient beings flourished far more than non-sentient beings.

It is important to emphasize the sheer amount of languishing being referenced in the above paragraphs. We are talking about hundreds of millions of years of a savage process of death, starvation, disease, illness, parasitism, and predation. It is hard to even comprehend the collective level of abject misery and pain that beings suffered during this process. God’s goodness and love gives us extremely strong reason to think that they would prevent this, a reason that simply does not exist on naturalism.

The picture gets worse for theism when we consider the difference between moral agents like humans and nonmoral agents like animals. Unlike animals, humans have the capacity for higher-order moral knowledge that grants them responsibility and significance that animals lack. On theism this difference is highly relevant. Even if God has a good reason for allowing humans to experience pain and languishing, there is strong reason to suspect that this reason will not extend to nonmoral agents. For instance, moral agents might need to experience some suffering for the sake of developing their moral characters. However, nonmoral sentient agents cannot develop such a character, and so need not suffer for it. On naturalism, in contrast, uncaring nature has no reason to treat nonmoral agents different than moral agents. Once again, the theist has reason to be surprised that the naturalist does not.

Triumph and Tragedy:

Third, consider the fact that people’s deepest wishes and desires of the hearts frequently remain unfulfilled. Even given the background that sentient agents experience much suffering, we still have reason to expect that the things people most want in life, their deepest wishes and desires, will be fulfilled on theism but not on naturalism. Yet, we consistently find no such fulfillment for human beings. People’s deepest goals are usually not accomplished, and those who deserve to have such goals fulfilled appear no more likely to have them fulfilled than those who do not. Indeed, it often seems that the morally worst people such as strongmen and dictators have been historically likely to succeed in their goals, while those who courageously campaign against them are often subject to the worst terrors, and fail just as often if not more than they succeed. If reality is fundamentally indifferent, this is exactly what we would expect.

Geographical Distribution of Suffering:

Fourth, consider the enormously inequitable distribution of suffering (Linford & Patterson 2016). What time period one was born into, where they are born, and what ethnic background and gender identity they happen to possess continue to have an enormous impact on their quality of life. The overwhelming ethical consensus is that these kinds of morally irrelevant differences between people ought to be irrelevant to the rights and opportunities they are afforded. Thus, we ought to expect that God, the exemplar of morality, will too be sensitive to these ethical concerns. Even if God has a good reason to allow all the suffering examined thus far, we have reason to suppose that this suffering will not come down disproportionately on certain groups of people on the basis of totally morally irrelevant facts about those groups. On naturalism, since uncaring nature is not sensitive to this moral consideration, we have reason to expect an inequitable distribution. Indeed, because there are far more ways for the distribution of suffering to be inequitable than equitable, the probability that it will be inequitable on naturalism is quite high.

Particularly horrendous evil:

Fifth, consider particular horrendous evil: evil that is utterly soul-crushing and life-changing for the victim, and that which involves moral abomination. Genocide, torture, and rape all fall into this category. I add this in as a special category above and beyond the others because a perfectly loving God would have particularly strong reason to prevent these evils in contrast to others. Since a morally perfect God would oppose such abominations to the fullest extent, they have every reason to prevent these evils. Additionally, many suppose that these kinds of evils are such that they cannot be justifiable under any grounds. Only the most radical utilitarian will suppose that horrendous torture is allowable if it enables some greater amount of flourishing later on. Thus, these evils are particularly resistant to theodical explanation and skeptical theism. Thus, we have significant reason to expect that, on theism, God will prevent these kinds of evils: a reason we lack on naturalism.

Theodicies and Skeptical Theism:

Because this would be an incredibly bloated post if I tried to address every possible objection and counter objection to my arguments (it is already very long) and my intention is just to communicate my personal reasons for atheism in general, I am avoiding delving too deep into objections to my arguments. However, I would be remiss if I did not say at least a little about objections to the problem of evil since it is such a common topic. Let me give my brief thoughts on the most popular theodicies and skeptical theism:

Free will:

The idea here is that free will is a great good, as it gives meaning to moral action and allows for meaningful relationships between people and God to be developed. Freely performed actions are more valuable than nonfreely performed actions. In order to allow for the existence of free will, however, God had to allow for the possibility of wrongdoing & horrific suffering. Thus, the existence of evil & suffering is justified because God cannot force humans to only take morally right actions without violating their free will, and a world with free will & suffering/immorality is better than a world without free will & without suffering. I have five major problems with these theodicies.

Firstly, I think the advocates of such theodicies seriously overestimate the value of free will. It is not at all obvious that the good of free will, in general, outweighs the disvalue of the horrors that it causes. In fact, it looks much more reasonable to think that a perpetrator’s freedom to cause suffering is a weightless concern in comparison with the disvalue caused by that freedom. Consider a world exactly like ours but in which people do not have the freedom to torture children. I think most people would agree that this world is superior to ours despite the loss of freedom to commit a certain action. Alternatively, suppose you had a button that allowed you to deprive a serial killer of their ability to kidnap and murder people. Any rationally thinking, moral person would press this button, because they would recognize that the horror caused by this individual’s freedom is not worth the suffering it entails. Indeed, the idea that free will is worth any horror it permits is fundamentally at odds with the basic way we have set up society: if someone commits a horrendous crime, we put them in jail, we do not preserve their freedom. The implicit value judgment in such laws is that the disvalue of the pain and suffering resulting from people's freedoms outweighs the freedom to inflict this suffering (Oppy, 2006 makes this point. Also see Ekstrom, 2020 for a much more elaborate and sophisticated defense of this general argument).

Secondly, everyone, including theists, must agree that freedom is subject to certain reasonable limits. This is because if God does exist, He has plainly limited our wills in many ways. There are plenty of evils I cannot possibly commit due to physical limitations. It could be that every time I lifted my finger a certain way I committed genocide, or that I could inflict pain 10x worse on any conscious creature I chose by yelling at it a certain way. If free will were an unlimited good, God would have granted us more freedom than they have. Thus, we need to consider what kind of limits we ought to expect on freedom if theism is true. Let us use the analogy of a parent and a child: When you have a young child you want them to exercise their will and freely explore their interests, yet not completely. You intentionally prevent the child from accessing particularly sharp objects, you don’t let them play with guns, and you don’t let them wander into the street without assistance. While this is in some sense limiting their will, doing so is completely reasonable because we know the horror that could result from not limiting them in any way. The underlying principle here is that a good parent limits a child’s responsibility until they are worthy of it or have developed it so as to prevent great tragedy & suffering. Similarly, God would limit our moral responsibility until we are worthy of it via freely developing the strength of character to handle it, and take it away when we abuse it too much. Yet, we are plainly not worthy of the moral responsibility we have, which is highly unexpected on theism (Draper, 1989).

Thirdly, free will clearly fails to account for much of the data about suffering that I cited. Free will on its own does not give an account of the existence of pain or its distribution, as there are many free actions that do not entail the existence of pain. Perhaps, however, we can sketch an account by which God uses pain in order to influence humans to perform morally right actions. Yet, pain often motivates morally wrong actions and the existence of pain has not impressively led people to take right actions, so we have yet more reason to not expect the facts about pain & pleasure we see on theism (Draper, 1989) Indeed, if the reason God allows suffering is for the sake of free will, that should give us more reason to expect that pain & pleasure would be tied to moral goals rather than biological goals, exactly the opposite of what we see. Finally, free will obviously does not account for the overwhelming majority of suffering throughout the history of earth: animal suffering for hundreds of millions as years has nothing to do with free will, and many instances of evil that humans are subjected to such as disease, decay, death, and natural disasters are often not the result of free will.

One response to the above point is to suggest that suffering in animals or suffering that is the result of natural processes is actually the result of the actions of free beings such as immaterial spirits or beings. This is a very poor response. Firstly, it is highly questionable that the freedom of malevolent beings to cause horrific suffering is worth the suffering it causes, as per the first two responses I provided here. Secondly, we can already perfectly adequately explain suffering amongst animals without reference to these immaterial beings, meaning that this response likely requires overdetermination‒that is, that there are multiple causes for every instance of natural suffering, which would be sufficient to explain the existence of that suffering on its own, yet jointly cause the suffering. Not only is overdetermination highly independently implausible, but it shows that that naturalist has a far more virtuous explanation for this suffering than theism: theist’s need to tack on an extra supposition that has no explanatory merit purely to explain the data. Thus, this response is tantamount to admitting that theism is worse at explaining the data than naturalism, and that the theist needs to conditionalize on evil (and so double-count the data) in order to generate this theodicy. Thirdly, this objection still fails to provide us with any good antecedent reason to expect that we would see a world rife with the kind of animal suffering we observe, because we would never predict ahead of time that, on theism, God would make a world filled with immaterial spirits who secretly overdetermine animal suffering of the kind we observe. Fourthly, the existence of immaterial beings requires a number of very difficult issues: How do these beings impact the physical world without us noticing? This would plausibly violate our best physical theories. How do persons exist without bodies? Not only does this fit very poorly with our background knowledge, but thinking it is antecedently likely that God would create persons without bodies undermines arguments in favor of theism, such as the fine-tuning argument. I have only scratched the surface with the challenges to this objection here, but I hope the reader can see how problematic this attempt to save the free will theodicy is.

Fourthly, free will theodicies pose a problem for belief in a paradisiacal afterlife (Nagasawa 2004, reprinted in Oppy 2006). Heaven is meant to be a place that overflows with goods and whose goods are incomparably greater than those present on earth. Yet, heaven is also meant to contain no moral or natural evil. If the existence of morally significant freedom requires that humans can engage in evil, and heaven is a place with essentially no evil, then humans do not have free will in heaven (unless they just never exercise it to commit evil, a possibility we will address). But, in that case, the greatest goods can be realized without significant freedom, which undermines the justification for the free will defense. Alternatively, if the existence of morally significant freedom is compatible with a place in which no evil is committed (heaven), then it is possible that God could bring about a situation in which we have morally significant freedom and yet there is no suffering or evil, which again undermines the free will defense. One response to this would be to suggest that though we are free to commit evil in heaven, we never do due to the presence of a perfect being or the nature of the heavenly environment. Yet, if a perfect being can help us along in heaven without contradicting our free will they could do the exact same for us on earth. Another response is to suggest that creating a heaven like place on its own with no free will is less valuable than creating a universe with significant moral responsibility and free will followed by a heaven without free will. This response looks like it is in significant tension with the orthodox understanding of heaven: firstly, since the goods in heaven are meant to be incommensurably greater than the goods of earth, free will must be an incomparably small good compared to those realized in heaven, and this makes it hard to see how realizing this insignificant good makes the suffering it entails worth it. Secondly, if our two options are first, a world with incommensurably great goods and no evil (heaven alone), and second, a world with the incommensurably lesser good of free will, comparable in magnitude to the badness of suffering, followed by a world with incommensurably greater goods, it looks eminently unreasonable to choose the second option. For, on the second option, evil is being introduced for the sake of the most infinitesimally small goods.

Fifthly, I think there is no incompatibility between the causal determination of our actions and our moral responsibility [13]. Thus, I find no problem with God determining the causal order such that we act in certain ways, yet us still maintaining moral responsibility. If God could do this, God could create a world in which we always engage in morally correct action and while preserving the moral value of free will.

Soul-making:

Soul-making theodicies are based around the idea that suffering allows for the cultivation of certain important virtues. The idea is that values such as love, courage, forgiveness, and sympathy (to name just a few) require suffering in order to supplant their genuine cultivation. For instance, we need to see people suffering in order to develop sympathy and need to face risk in order to develop bravery.

The major problem for this theodicy is that it ought to lead us to predict that evils are permitted insofar as they allow for the development of important virtues. Yet, this is manifestly not what we see: while at least some suffering does cultivate virtue, the majority of it clearly contributes to the degradation of people’s character and lives. Much of suffering, especially the most horrific suffering, is utterly soul-crushing. Additionally, as in the case of free will, this theodicy does not account for the evil we see that makes no contribution to soul-making. In this category sits the bulk of all evil, natural evil that afflicts animals, since animals can not meaningfully engage in soul-making. Against this, it might be said that animals can develop virtues. As example, animals might display perseverance and courage when rescuing an at-risk member of their species. This suggestion seriously stretches the limits of plausibility: firstly, the vast majority of suffering animals experience does not have any visible relation to cultivating virtues, secondly, when it does, it could be said only about animals of a high cognitive capacity such as lions or dolphins, but it appears impossible to say about animals with very rudimentary cognition who act in the absence of free will or any higher-order thought. Thirdly, I would go farther than this and suggest that animals do not possess propositional attitudes at all, which form the basis for understanding the soul-making process, thus making any application of soul-making to animals totally nonsensical (Stitch 1979, Churchland 2007). One possible way to extend the soul-making theodicy to account for the deficits I have adduced is to add in the defeat of evil and continuation of soul-making in the afterlife: this is a possibility I will address in its own section.

There are some other problems with soul-making theodicies, very well outlined by G. Stanley Kane in his response to John Hick, one of the primary modern proponents of the theodicy (Kane, 1975). I highly recommend his article for those interested in a more sophisticated critique of them than what I can provide here. A brief outline of two of his points: first, the advocates of soul-making theodicies suffer from a serious lack of imagination about how virtues might manifest in the absence of truly horrible evil. For example, virtues like perseverance and courage could arise in situations such as working hard on a doctoral dissertation or training for sporting events. Compassion, empathy, and sympathy can all arise when supporting those participating in these difficult endeavors. Virtues can also be developed in simulated situations: for instance, extremely realistic but nonetheless illusory dreams or visions could allow people to be confronted with scenarios that involve danger or risk, and could be made forceful enough to leave a lasting impression on people’s psychology. The response theists will likely give to this is that the virtues we can cultivate in the face of more horrific suffering and genuine danger are more significant than those we can cultivate in these scenarios. Even if this is true, however, this difference in value, which appears to be a matter of degree rather than kind, is going to need to outweigh the disvalue of all the suffering necessary for its cultivation, which is implausible.

A second point Kane makes is that there is an inconsistency between the soul-making theodicy and heaven, analogous to the one present with the free will theodicy. Heaven is meant to be a place that will be free of evil, suffering, and hardship in general. Thus, heaven will not be the kind of place where the virtues gained via soul-making can be exercised. If the cultivation and use of these virtues is so important, then it becomes unintelligible why they should not persist in heaven. Additionally, it makes their cultivation prior to heaven appear pointless. If we know that humans will ultimately enter into a place where they do not need to use them at all, their toiling to develop them becomes stripped of its meaning (Again, see Kane’s article for an elaborated version of this argument).

A final note about soul-making theodicies is that I think that, in general, people seriously overestimate the value of going through hardship as a means for developing good character. We often engage in the idealization of our own suffering and that of others to try to imbue it with a kind of meaning. In fact, however, the idea that suffering ‘makes you stronger’ is generally psychologically false: on the contrary, going through horrible circumstances puts you in a more vulnerable, mentally unsound, and unhappy place. This is especially true for the worst kind of traumas, which are exactly those the theist needs to justify for their theodicy.

Heaven & Defeat:

One common strategy to respond to the problem of evil is to suggest that prima facie unjustified evils will not remain unjustified: they will later be incorporated into some greater story for the sufferer in the afterlife. Two important ways this might happen are worth distinguishing (Chisholm, 1968). One is mere compensation, which is where the evil that occurs in this world is ‘balanced-off’ by some disproportionately greater good someone receives later on. For instance, suppose your phone is confiscated by the government but the next day the government awards you a new house for free. The good of getting a new house is (presumably) greater than the disvalue of losing your phone, such that you have been compensated for being robbed. However, in this case, there is no necessary connection between losing your phone and the good of getting a new house. You could have gotten the house without also having your phone stolen. Thus, this kind of relation between evils and later goods is plainly insufficient to help with the problem of evil. Instead, what most theists appeal to is defeat: where the evil that you experience has some necessary connection to a greater good: that is, the greater good could not be achieved without the evil. The evil that happens becomes incorporated into some state of affairs that is incommensurably better than it would have been without the evil. This ‘defeat’ of evil does help with the problem of evil because it justifies evil that occurs in that the whole state of affairs the evil allows for is better than it would have been without that evil. Since most evils are obviously not defeated in this life, theists almost always envision that this defeat will occur in the afterlife during something like the beatific vision.

The main problem I have with this strategy is that, antecedently, it does nothing to help us predict any particular state of affairs. We could, when confronted with any quantity or kind of evil, suggest that those evils will be defeated in the afterlife. We could do this if we lived in a completely hellish world filled with barbarism and torture. On its own, conjoining theism with the condition of defeat and the afterlife is compatible with any amount of suffering we see and so does nothing to raise theism’s predictive ability with respect to the data. So, we are obviously going to need some kind of reason to think that the evils we see are uniquely defeatable relative to other kinds of evils that might be allowed. However, as far as I can tell we have absolutely no reason to expect this: on the contrary, for most evils we see it looks completely befuddling how they could be defeated. Firstly, there are those evils rife in the animal kingdom that seem completely random and pointless and involve beings who lack the capacity to participate meaningfully in moral development. Secondly, there are those evils for which we have the strong intuition that nothing could justify these evils occurring. Think about the slow torture and murder, over weeks, of a child. I think many people will suppose that this suffering is so horrible and grotesque that it is simply impermissible for it to be allowed, and that there can be no whole combined with it such that it is justified. Notably, to think there is some such whole relies on a kind of radical consequentialism on which far-off goods in the afterlife necessarily connected to the original evils justify their allowance. There is also the worry that this entails moral skepticism: if horrible evils are necessarily connected to later states of affairs such that they are all-things-considered good for the sufferer, that seems to deprive us of our moral obligation to intervene and stop them (Maitzen 2009). Even worse, it makes it such that we have the least reason to intervene in the case of the most horrible evils, exactly contrary to our moral intuitions, since these must be the ones that picture into the greatest overall wholes such that they can be justified.

Skeptical Theism:

Skeptical theists, unlike theodicists, do not attempt to provide an explanation for why God allows suffering. Instead, they suggest that our epistemic situation is such that we cannot reliably make certain inferences about suffering. The most famous of these inferences that they doubt is the move from the appearance of much of the suffering we see being gratuitous to the conclusion that it is reasonable to believe that some of it is in fact gratuitous (Wykstra 1984). However, many evidential arguments from evil such as Draper’s make no such ‘noseeum inference’ and so are at face-value immune to skeptical theism. In order to try to challenge these arguments, skeptical theists have attempted to modify their approach in order to cast doubt on the probability judgment that the suffering we see is more probable on naturalism than theism. I do not think any such modification has been successful. I discuss my opinion on this issue at much greater length in my paper on the cognitive science of religion, cited earlier. I direct readers who want to hear my full thoughts on the matter there. Briefly, I have three main problems with skeptical theism:

The first problem is that, as I mentioned, I do not think skeptical theism cannot be modified to undermine evidential arguments from evil like Draper’s. The reason for this is threefold: First, the existence of unknown reasons God has for permitting evil is antecedently counterbalanced by unknown reasons God has for preventing evil, meaning the unknown reasons cancel out and leave us with the known reasons God has for preventing evil (Draper 1989, Draper 2014). Second, the data of suffering in question is extremely specific, making it antecedently highly unlikely on theism and naturalism in virtue of its specificity alone. However, the inferences made on the basis of the background knowledge raise its probability on naturalism but not theism, and skeptical theism does nothing to undermine this (Draper 2014). Thirdly, the very fact (if true) that it is unknown whether theism can predict the data of suffering, but it is known that naturalism can, can itself be modified into an argument from suffering to support naturalism (Tooley, 2019).

The second, more fundamental problem, is that if skeptical theism can be modified to undermine the kind of probability judgment I make here, this would be fatally damaging for theism in general. If the existence of unknown reasons God might have to allow some feature of the universe (or the unrepresentativeness of the sample) renders the probability of that feature on theism inscrutable, this applies not just to suffering but to all features of the universe we might cite in these kinds of arguments, including features one might take to favor theism. This would destroy all abductive arguments for or against theism, render theism with zero explanatory content, and force us to be rationally committed to agnosticism (at minimum) (Climenhaga, forthcoming makes a similar argument).

The third problem with skeptical theism is that its skepticism leaks into other areas and undermines our knowledge in other insidious ways. The most famous of such examples is that of moral paralysis, which I touched on in the section in defeat: if we think God exists and has an all-things-considered reason to allow suffering (because it will be a net benefit to the sufferer), that plausibly deprives us of our moral obligation to prevent horrific suffering (Maitzen 2009, Sehon 2010). I am still agnostic as to whether this argument works (Howard-Snyder 2009 provides a thought-provoking response to it). There are other concerns too: the very principles skeptical theists use seem to commit them to doubt other known facts such as the age of the earth or the truth of their own religious texts (Hasker 2010, Wielenberg 2010, Russell 2018,). This is because their principles are, in general, on very shaky epistemological ground (McBrayer 2009).

Ineffectiveness of Prayer

The most rigorous, extensive studies we have on the efficacy of intercessory prayer establish that it makes no relevant difference in the world (Masters & Spielmans, 2007), (Masters et al., 2006). For instance, people who are ill and prayed for are no more likely to recover than those who are ill and not prayed for. When I say ‘relevant’ difference what I mean is the kind of difference that would only be compatible with theism, but not naturalism. Naturalism is, of course, compatible with prayer making an indirect psychological difference to those praying or being prayed for, and thereby speeding the recovery of prayer recipients. But, on naturalism, there is no way that prayer could make a difference outside of some indirect natural route such as this, since there is no greater force that could actually intervene and cause prayer requests to be miraculously fulfilled. Thus, the probability that prayer would be ineffectual in the way examined by studies on intercessory prayer is essentially 1 on naturalism.

On theism, there is at minimum a mechanism by which intercessory prayer could be effectual; namely, there is a God who has the ability to intervene and realize prayer requests. This alone shows that the efficacy of prayer is more likely on theism than naturalism. Do we have any reason to expect that God would have the motivation to fulfill prayer requests? On the plausible assumption that a perfectly good God would be all-loving, I think we do. Part of love is a desire for the wishes and hopes of those loved to be fulfilled. If one loves another, one strives to fulfill the wishes and requests of that person so long as one can fulfill them and so long as they would be good for that person. Unlike us, God has no limits on their ability to fulfill wishes, and since they are all-loving, we have at minimum some reason to suppose that they would want to fulfill people’s wishes. Note too that this data is relevantly different from the data invoked by the problem of evil. We can envision people praying for particular things in a world without horrendous suffering or widespread languishing. Indeed, the kinds of things people pray for are often those things that would not make for a potent problem of evil. Thus, the data of unfulfilled prayer is relevantly different from that invoked in the problem of evil, and it constitutes some evidence for naturalism.

Religious Diversity

By ‘religious diversity’ I am referring to the fact that there is no clearly dominant religion, and the religion that is dominant has drastically changed over the course of human history. If theism is true then God is the ultimate exemplar of value. Forming a relationship with this exemplar and modeling one’s life in line with them would surely be an immensely important good, if not the most important good one could attain. Thus, if God exists, God has a reason to ensure that their revelation is present throughout history so all beings have access to their teachings and to a relationship with them. This obviously holds on SP, and the same expectation holds even on WP since following and knowing God is likely the most important aspect of a good life on theism, such that conscious moral agents without this option would be missing out on the most important parts of their flourishing.

If naturalism is true, on the other hand, then religion is simply a byproduct of biological and sociocultural evolution. Since humans are inherently disposed to latch onto a wide variety of religious concepts (see, again, the section on the cognitive science of religion), and since cultural evolution is tumultuous, we can expect that the chances of a single religious tradition dominating all of human history is extremely low. This inference is obviously undercut by the assumptions of theism for the reasons mentioned in the previous paragraph.

A second point worth making is that what religion one ends up falling into is determined primarily by family and communal ties, not an earnest seeking of the evidence. If theism is true, we have some reason to think that those who are most honestly exploring truth and a relationship with God will be the most likely to come to theistic belief. On the other hand, if naturalism is true, this is exactly what we would expect, for two reasons: First, since theism is in fact false, earnestly seeking the evidence will not reliably lead one to belief in theism; Second, since religion is primarily a byproduct described the cognitive science of religion model and general sociological trends, the primary means by which religion spreads is through credibility-enhancing displays from people within your own cultural unit, and from costly signaling from people within one’s culture (read my paper on the cognitive science of religion for more about this). This precisely leads us to expect that adherence to particular theisms will track what cultural circumstance one is born into. These expectations, however, are undercut on theism, since God has good reasons to reveal themselves to people for reasons other that they happened to be in a culture or family that subscribes to the correct religion.

One kind of response to the argument I have made is that the free will of human beings to choose to believe in God is a great good that requires the possibility that many humans will not come to theistic belief. However, nothing about ensuring that one theistic tradition would be historically dominant deprives humans of free will. Christianity is the dominant religion in the United States today, yet people can freely choose not to believe in God. What I am suggesting is just that Christianity ought to be a live option throughout history and across the world, not that God should force anyone to take this option. Indeed, if anything, this objection should lead us to expect that God would set up reality such that everyone has the free choice to believe in God. Yet, many persons have been born in situations where the existence of God was not even an idea one could contemplate. Thus, if anything, the idea that God would want people to freely choose whether to believe in theism should lead us to expect even more that the correct theistic tradition would be a live option across societies and times.

Another objection to this argument would suggest that propositional assent to theism is not the only way to have a relationship with God. Thus, a relationship with God is in some sense open to all people, even those not aware of the concept of theism. I address this objection in the second on nonresistant nonbelief, as it is also relevant to that argument.

Flawed Religions

In this category, I intend to include a variety of facts: that religions have been variously a force for good and bad in history, have both condemned and supported atrocities, that holy texts contain a mixture of interesting, morally commendable content, and banal, morally horrific content, and that religious believers do not impressively stand out as leading more impressive ethical lives than nonbelievers.

On the assumption that God will reveal himself to humans through a religion (an assumption most theists will accept), we have reason to expect that the religion that correctly portrays God should be extremely impressive. After all, this religion will represent the ultimate moral exemplar and the most perfect being there is, so surely this religion should stand out. We should expect the inspired texts and teachings of this religion to be far above those texts and teachings that are uninspired. Additionally, the representatives of this religion, in virtue of having access to these correct revealed teachings, should at least trend towards good moral behavior in comparison to those who do not follow this religion’s teachings. Yet, we do not find any such ‘stand-out’ religion. All religions have been a pretty big mixed bag: their holy texts often contain morally horrific content, their teachings do not appear clearly and obviously more profound than the teachings of other religions or secular teachings, and their adherents are, like the rest of us, prone to being very morally upstanding but also behaving awfully. Now, I am aware that this is a fairly contentious claim: perhaps the proponents of a particular religion will claim that their religion does in fact stand out as a moral exemplar amongst the crowd. For this reason, I doubt my argument will be convincing to such people: however, I think it is clear why this fact provides me with evidence against theism.

Nonresistant Nonbelief

In this category, I refer to those people who do not have a relationship with God and are not actively resisting this relationship in any way. Many nonbelievers fall into this category: there are many atheists and agnostics who would be perfectly happy to believe in God, and perhaps even desire to, and yet lack this belief. There are also many, perhaps even more, nonbelievers who fail to believe in God due to lacking proper access to the concept. Many millions of people throughout history have never even had the concept of theism available to them in their societies, and so not even had a chance to believe in God (Maitzen, 2006).

If theism is true, we should expect that God would always be open to a relationship with persons. The case for this has been made extensively by JL Schellenberg, and I summarize it more in detail in my cognitive science of religion paper (Schellenberg, 2012). In brief, an all-good God would be perfectly loving, and perfect love plausibly involves valuing a relational-personal love with creatures. Forming a relationship with the source of all value would arguably be the most important kind of good creatures could participate in. Thus, if God exists, we have reasons to expect that they will form relationships with creatures who are not actively resisting them in any way. On the other hand, if naturalism is true, then it is perfectly expected that many nonresistant nonbelievers will not believe in God, since cultural and biological accidents will often lead them to, and there is no overriding force compelling them to.

I think the most pressing issue for this argument lies in the question of what a relationship with God actually entails. There is a certain ambiguity in the conflation between assent to a proposition and relationship with a person: so far, we have been treating a relationship with God as if it requires having a certain belief. But, might it be possible to have a relationship with God without this belief? Perhaps just participating in the good in some way or having a connection to some positive force in your life is enough to be in a relationship with God, even if you do not think He exists. In this way, you might doubt that there are nonresistant nonbelievers.

It looks somewhat plausible to me that there may be other ways to have a relationship with God besides the relational-personal one on which you need to actually endorse theism. What does not look nearly so plausible is to deny that there is any significant, added good that one obtains by having this relational-personal kind of love with God. Put another way, there may be different versions of having a relationship with God: one is to simply live an ethical life, but another is to step into a relational-personal communion with God. Pursuing the latter in addition to the former looks like it would deepen one’s existing relationship with God, revealing new aspects to it and affording yet more opportunities to appreciate the divine. It looks very implausible to deny that having a personal relationship with God of the type that requires believing that He exists affords one no added benefit in addition to solely having a sort of non-personal, ‘pursuit of the good’ type relationship with God. Firstly, God is meant to be a person and relational-personal relationships are not just a kind of relationship one can have with persons, but the primary mode in which relationships with persons manifest. Secondly, many religious believers report having precisely this kind of relationship with God and benefitting from that relationship greatly. And, many nonbelievers (especially former believers) report the pressing lack of such a relationship as being a significant detriment to their lives. Thus, to endorse the strong form of this objection needed to fully undermine the hiddenness argument requires that you deny what religious believers and nonbelievers alike report.

A second issue with this objection is that it does not fully rebut all the motivations one could use to levy the hiddenness argument. So long as some important good is associated with correct belief in God, we have reasons to think God would care about correct belief and so prevent non resistant nonbelievers. Rejecting this will lead to us giving up the idea that correct belief is important to God, for reasons I elaborate in my paper on the cognitive science of religion (McKim, 2015). Most theists will not be happy with this: according to them, salvation requires propositional assent to theism, and proselytization is extremely important, to take just two examples. Thus, levying this objection to hiddenness has corrosive consequences for the standard beliefs of sectarian theists.

Introduction to Theistic Evidence

I diverge from many of my fellow atheists in accepting that certain facts about the world fit better with theism than naturalism. In order to adequately compare theism and naturalism, I cannot only focus on those facts which naturalism explains better. I must cover those facts I think theism most plausibly explains better than theism, and argue that the balance tilts in favor of naturalism. This is no easy task. I unexpectedly think the data favoring naturalism is more prevalent and stronger than the data favoring theism. However, this subjective impression is hard to argue for. Instead of trying to do such a weighing process, I will identify two extremely common problems with the data that favors (or is purported to favor) theism over naturalism. By showing how prevalent these problems are with data taken to favor theism, I hope my reader will gain an idea of why I think the evidential case for theism is inherently weak. Here are the two problem I have alluded to:

Disputed data: The existence of the data taken to favor theism tends to be much more controversial than the data that favors naturalism. In other words, whether the data theists appeal to as supporting theism actually exists is often uncertain. Obviously, if the existence of a piece of data is controversial, it cannot as strongly count towards our final tally than if the existence of a piece of data is known [14].
Understated evidence: Here, I follow philosopher Paul Draper who has identified the following common fallacy in theistic arguments: “They successfully identify some general fact about some topic that is more probable given theism than given naturalism, but all too conveniently ignore other more specific facts about that topic, facts that, given the general fact , are significantly more probable in naturalism than in theism.” (Draper, 2010 421-422). Put more simply, theism often explains general facts about the universe fairly well but fails when you examine the specifics/details of those facts.

This section will proceed as follows. I will identify the pieces of data I think most promisingly favor theism over naturalism. I will argue that they often fall prey to one or both of the above issues that I have identified. Finally, I will summarize my findings on the relative explanatory power of theism and naturalism, arguing that naturalism fares much better than theism.

For Theism

Necessary & Contingent Existence

[Note: This section is currently under substantive revision to delve deeper into some issues in the literature I only touch on briefly. Stay tuned!]

Arguments in favor of a necessary being that explains contingent reality take on many different forms. I obviously do not have space to consider them all, so I am going to be bold and try to outline the general form that these arguments take on. There are three general steps in these arguments. First, argue for some general kind of explanatory principle, very often the principle of sufficient reason or some variant of it. In its most simple form, this principle says that all contingent facts require an explanation. Second, argue that some amalgamation of things exists that needs explaining, for instance, the universe or the ‘Big Conjunctive Contingent fact.’ Third, argue that the explanation for this conjunction lies in a necessary being, and that this necessary being is God.

What reasons should we have for supposing that ‘God’ is the best candidate for a necessary being, over some naturalistic initial universe state, for instance? Many arguments have been suggested here, but one of the arguments most in vogue (and that has been around for a while) is an argument based on arbitrary limitations. According to this argument, a naturalistic initial state would contain a variety of arbitrarily set properties and values, while God is infinite in all their attributes, and this lack of limitations is either simple or more clearly a suitable candidate for necessity (or both).

I have a few objections to this argument from ‘arbitrary limits’. Firsty, the positive arguments for this contention are not very forceful, since they tend to rely on our intuitions that something that is unlimited is more theoretically virtuous than something limited. I only have a very tenuous intuition that this is true. Secondly, it seems that this problem is going to cut both ways, especially for sectarian theists. That God is triune in nature according to Christian seems very arbitrary—why are there three instantiations of God rather than two or four? Why is there one God instead of an infinite set of Gods? Wouldn’t a polytheistic foundation have less limits? Thirdly, there seems to be a sense in which having more causal power than is necessary to produce effects is theoretically unvirtuous. The theory in which the initial state has just enough causal power to bring about the universe seems more explanatorily parsimonious than the theory that tacks on all sorts of unnecessary causal power. Fourthly, it is not clear that theism is the only candidate theory for something that is unlimited. One can easily imagine some kind of initial state that lacks spatial or temporal boundaries, without having any moral properties whatsoever, for instance. Fifthly, given that the initial state is necessary, it is unclear whether asking why the initial state has value X rather than value Y is a valid question. After all, the initial state had to be that way since it was necessary.

Obviously, I do not have time to evaluate all the arguments that move from a necessary being to God here, but I direct others to resources that have done so. In sum, I think there are some intuitions that could lead one to concluding that theism is a better candidate for a necessary being, but I think they are far too weak to merit any deduction and at best serve as weak evidence for theism.

In addition to the problems with the inference from the necessary being to God, this argument clearly relies on disputed data. Many (if not most) philosophers are skeptical of the idea of a necessary foundation to reality, and I myself am not convinced that the first stage of contingency arguments that establishes this necessary foundation succeeds.

First, I’m not convinced of the PSR. I do not think any of the arguments for it are decisive. Perhaps the most common argument for it is that the PSR is very intuitively obvious and draws support from our everyday experiences: as far as we can tell, everything in our everyday life succumbs to explanation. I agree that our everyday experience is filled with reference to explanation, and that we have good reasons to assume inductively that new things we encounter will too have explanations. However, I am extremely skeptical of extrapolating that commonsense intuition to affairs so far removed from our experience as the existence of the universe, time/space/energy itself, etc. We know that our everyday, commonsense understanding of other matters such as time or motion completely breakdown when we break out of our local reference frame. For instance, our ‘folk’ theory of time is that it is absolute, non-relative, independent from space, and presentist in nature. Yet, general relativity teaches us that this folk theory turns out to be false. Why should we expect that our ordinary conception of causality and explanation will successfully extend to matters so removed from human experience as the existence of the universe itself?

Another argument for the PSR is that our scientific practice provides support for it and or relies on it. I agree that, in general, science assumes that phenomena have explanations and looks for these explanations. However, it is not at all clear that science supposes that we will always find explanations for events. It seems, rather, that science looks for explanations as far as it can, but, in fact, we often find that we bottom out in certain natural laws that are not amenable to further explanation. I am not uncomfortable admitting such brute facts into my worldview if need be. I do not think the theist will escape having to do this, anyways: when we get to the question of why God chose to make the world as He did they too will have trouble coming up with a fully sufficient explanation.

In addition to rejecting the argument for the PSR, I think there are positive reasons to reject the kind of PSR the theist needs for the LCA. One problem with applying the PSR to the universe itself is that at least one kind of naturalistic world-picture says the physical universe constitutes the totality of explanatory reality. Clearly, the totality of explanatory reality cannot itself have an explanation. Thus, we will need to bottom out in an infinite regress of explanations, a circle of explanations, or in an uncaused cause (Note that, again, the theist is going to have exactly the same options as the naturalist). It is perfectly coherent to suppose that the first link in the explanatory chain that constitutes the universe is relevantly unlike the other links in virtue of being the beginning of causal reality itself, and thus we have principled reason for holding that it would be unexplained unlike the rest of the events across the chain. The fact that such coherent naturalistic pictures exist suggests that explanatory principles that rule them out are false. After all, we can still garner all the benefits of such explanatory principles, including all the arguments for them, simply by limiting our explanatory principles to apply to all links in explanatory reality other than the one initial, unexplained link.

One argument against the principle that ‘all non-initial items have causes’ from Alexander Pruss and Robert Koons is that this principle succumbs to a skeptical challenge: since we cannot know whether the current state of the universe is the initial state, we have to be open to our current thoughts being themselves uncaused, which would undermine our rationality (Pruss & Koons, 2020). My response: just because it is possible that a skeptical scenario is true does not mean we succumb to skepticism. Knowledge does not require absolute certainty in the sense that we decisively rule out skeptical scenarios, just justification. Clearly, my thoughts prima facie appear uncaused, and there is ample evidence to think they are not. The mere possibility of a skeptical scenario does not undermine that. If it did, skepticism would leak into our beliefs with or without a more extended explanatory principle.

A second major problem with the PSR is that it is not clear that necessary facts can explain contingent facts. The basic problem here can be summarized as follows: consider the sum total of contingent facts, let us call the aggregate ‘the universe’. Now, if the universe has an explanation, its explanation is either contingent or necessary. Since we have stipulated that all the contingent facts are contained within the universe, it cannot have a contingent explanation (assuming that things cannot explain themselves. If they can, the LCA has other problems). Therefore, the explanation for the universe must be necessary. However, necessary facts must obtain. Therefore, anything that follows from necessary facts must also obtain. However, this would make the universe itself also necessary, thus absurdly implying that there are no contingent facts and destroying the basis for the LCA in the first place. This is the problem of modal collapse, and it has been outlined in many places and been subject to much discussion (Sobel, 2009, Van Inwagen, 2019). Obviously, I cannot do justice to the full breadth of this conversation here. Let me just highlight one common attempt to respond to modal collapse, and explain why I do not think it works:

One thing the proponent of the PSR might want to say is that explanations do not need to be entailing. If explanations are not entailing, then the universe does not automatically follow from its necessary explanation, and so it can remain contingent while still being explained by the necessary fact. First of all, it is not clear that necessary truths can be merely probabilistic reasons that incline a particular outcome. As Sobel points out:

“Necessary truths, since they are ‘true no matter what’, cannot matter to probabilities of other truths. Truths other than necessary truths are either logically entailed by or completely independent of them. It is, for example, necessarily true that it will either snow or not snow in this area later this year, which truth has absolutely no bearing on or relevance to whether it will snow this month. Nor is the problem the obviousness of this necessary truth. It is its ‘vacuity’, its ‘lack of contingent content’, that is the problem. Only truths that are not ‘true no matter what’ can be inductively or probabilistically relevant things that are not true no matter what. From this I conclude that necessary truths can be only deductively or necessitating relevant” (Sobel, 2009 224. See pages 223-228 for the further context that motivates this passage)

I think I buy that probabilistic explanations exist, but I certainly do not buy that they are sufficient explanations. Imagine two identical world states x and y that diverge at some particular time point t, and are no longer identical after this point. This is consistent with the idea of a probabilistic explanation, since x and y may contain the exact same probabilistic device prior to the divergence. There does not seem to be anything whatsoever in x and y that can be used to explain their divergence at t, since they are identical prior to t. The only thing you can appeal to is brute randomness. You can point to the probabilistic device that makes it such that there was some chance of each divergent outcome occurring, but why we have that particular probabilistic device and/or why the probabilities led to different outcome is just random. In sum, probabilistic explanations seem to admit some kind of brute contingency in the world, thus making them insufficient replacements for the strong PSR.

One final point is that weaker versions of the PSR end up entailing the strong, untenable version of the PSR again (Oppy, 2006 and see here). I do not have space to discuss these matters in depth, and I would rather use this opportunity to shout out the blog of a friend of mine who has given this matter (and other arguments related to the PSR) significant treatment.

Enough about the PSR, let us move on to another problem with cosmological arguments. I am highly skeptical of the general formation of a ‘big conjunctive contingent fact’ and other similar formulations. Firstly, there are serious set-theoretic concerns around the ability to form such a conjunct. One brief worry is that the BCCF would itself be a conjunct of the BCCF, since the BCCF contains all contingent propositions. Secondly, it is not clear to me that the BCCF needs any explanation above and beyond an explanation of its conjuncts, because the BCCF seems to be an arbitrary unification of facts put together by the human mind, not a substantive thing. Hume made this point well:

“In such a chain or series of items, each part is caused by the part that preceded it, and causes the one that follows. So where is the difficulty? But the whole needs a cause! you say. I answer that the uniting of these parts into a whole, like the uniting of several distinct counties into one kingdom, or several distinct members into one organic body, is performed merely by an arbitrary act of the mind and has no influence on the nature of things. If I showed you the particular causes of each individual in a collection of twenty particles of matter, I would think it very unreasonable if you then asked me what was the cause of the whole twenty. The cause of the whole is sufficiently explained by explaining the cause of the parts.” (Hume & Coleman, 2007)

To make this more concrete, imagine the little conjunctive dog fact. This is a conjunction of the fact that my dog is hungry and the fact that I desire to take my dog on a walk. What is the explanation for this conjunctive fact? It seems that all we need is an explanation for each conjunct: once you have explained both why my dog is hungry and why I want to take my dog on a walk, you are finished. There is no further explanation for the little conjunctive dog fact itself. The usual response to this kind of argument is to generate examples of conjunctions that do seem like they require a further explanation for the conjunction beyond an explanation of each conjunct. I have never found these examples compelling. I think the more important point is, perhaps, not that we explain the conjunction by appealing to its conjuncts, but that the very idea of the conjunction is an arbitrary uniting in the human mind, not a concrete thing. However, to be generous we might even cede to the theist that conjunctions are real, and that they sometimes need explanations. This is not enough to establish that all conjunctions require explanations, which is enough to cast doubt on this stage of the LCA.

In sum, I think arguments from necessary existence rely on highly dubious data that I have good reason to dispute. Now, is this data understated at all? There appears to be one sense in which it is. The general fact that the vast majority of objects we see are contingent, limited, and physical seems to fit better with naturalism than theism, at least following the presupposition many theists make that naturalism posits a brute, contingent origin to the universe while theism posits a necessary one. On theism, we have far more reason to suppose that the universe will be composed by nonphysical, necessary objects than on naturalism.

In summary, I think contingency arguments are interesting but I do not have strong reason to suppose that they conclusively establish the existence of a necessary being that explains contingent reality, nor that if they do, that this is better explained on theism than on naturalism.

Fine-Tuning

Fine tuning arguments appeal to the fact that the fundamental parameters of physics take on just the right values to allow for the existence of life. I have always been sympathetic to these kinds of arguments. On theism, since God is all-good, we have reason to expect that they would value conscious life, both because that life is intrinsically valuable and because this would permit for a relationship between these beings and God. On the other hand, on naturalism reality is fundamentally indifferent, and so there is no intrinsic part of the hypothesis that predicts finely tuned values for conscious life.

Despite my sympathy for this argument, I have increasingly lost faith in its effectiveness. Let’s start with the probability of fine tuning on naturalism. The reason the probability judgment looks so impressive here is generally because advocates for fine tuning assume a uniform probability distribution of the selection of the constants. In other words, of all the possible values the parameters of physics could have taken on, each one is given an equal epistemic probability of being selected. Is this what we should assume on naturalism? One of my friend’s over at knownanstrangerthings has argued that if we should assume anything about the way the constants would be selected on naturalism we should assume that there would be a kind of weighted distribution towards constants that allow for life. In brief, there are two reasons for this (but go read my friend’s post!): a posteriori, we know that natural processes tend to give rise to weighted probability distributions such as normal distributions. A priori, the central limit theorem appears to provide further reason to expect such distributions. And, if the constants were selected by a weighted distribution, we ought to assume that constants like ours resemble the mean.

A second reason to think the probability of the constants is not low on naturalism is because many specific versions of naturalism predict that we’d observe finely tuned parameters ie. a multiverse, a necessary initial state, modal realism. This response is not fully satisfying, however, since it requires specifying a specific version of naturalism and so lowering the prior probability of the naturalistic hypothesis. Although, since theism already starts out at a disadvantage when it comes to simplicity, accepting these expansions may not be too costly for the naturalist. Additionally, one could invoke the law of total probability here to show that the posterior probability on naturalism is the weighted average of all these possibilities. And, while I do not have any reason to regard these specific extensions to naturalism as very likely, I do not have strong reason to regard single-universe, 'random roll of the dice' naturalism as very likely either. Therefore, the initial naturalism prior looks much better than most theists suppose it to be.

The most significant problem for fine tuning arises when we examine the probability of the constants on theism. Recall that the reason God really cares about the parameters taking on certain values is not because he likes certain numbers, but that these values allow for something he does care about: conscious life. This judgment is only going to work, however, if we think that God needed to set the values the way they did in order to get life. Why should we assume this, though? As far as I can tell God’s omnipotence means they could create conscious life regardless of what the parameters of physics look like. Here are two ways they could do so: first, they could simply miraculously sustain life like ours under hostile conditions. Second, they could alter the laws that govern how consciousness can instantiate in physical matter. In other words, God could change the psychophysical laws such that physical structures radically unlike ours could be the subject of mental states. If this is the case, however, then since God can get what they really care about without fine tuning, fine tuning ceases to be evidence for theism. This latter argument has been presented quite forcefully by Neil Sinhababu, who also shows that theists are in no position to claim that such alternative psychophysical laws are impossible (Sinhababu, 2017). Indeed, if anything the range of options is much greater on theism than naturalism. For, on theism, God has access to not only all the physically possible ways the universe might be arranged but all the non-physical options: for instance, physical universes in which God is constantly miraculously intervening to alter matter and energy and entirely non-physical universes in which disembodied life exists with no physical substrate whatever (Manson, 2018).

The final problem with the fine tuning argument is that it understates the evidence (Draper, 2008). Even if conditions that permit conscious life are more probable on theism than naturalism, specific facts about the nature and condition of conscious life are more probable on naturalism than theism. I have already talked about some of these specific facts: that life is far less impressive than it could be, that life evolved via biological evolution, and that our minds are highly causally dependent on physical processes. A few I have not yet talked about are worth mentioning.

First, the universe is not teeming with life. If conscious life is inherently valuable, surely we would expect that God would populate the universe with lots of life. Indeed, the very reasons that lead us to expect conscious life on theism seem to equally lead us to expect to see lots of life. Second, the universe is enormously hostile to the existence of life. If naturalism is true, since reality is indifferent to conscious existence, this is highly expected, while on theism, if God has our existence in mind, we have some reason to suppose the universe would be far more accessible and explorable. Now, I have become somewhat skeptical of this second argument simply because there are many ways to generate counter-reasons why God might want to make the universe hostile. For instance, the universe being so hostile could help emphasize the epistemic gap between God and us or demonstrate the glory and beauty of their creation. One note about this counterargument is that it comes at a cost for the theist: if one of God’s goals is to put humans in awe and or fear of his creation, that is yet more reason for God to miraculously sustain life under hostile conditions, which undermines the fine-tuning argument. If we were miraculously sustained under such conditions, that would surely put us in tremendous awe of the universe, more so than now. This nicely highlights how the predictions theism makes are always nebulous—coming up with reasons why God might have done things a certain way is easy because there are so many possible desires God might have that can be appealed to. This is a flaw theism has though, not a virtue, and I discuss it further in the section on the prior probability of naturalism and theism.

Irrespective of the success of the above general argument, there are two senses in which the hostility of the universe strikes me as highly unexpected. First, the universe is much more set up for the facilitation of nonmoral agents, in that the simplest kind of life ie. bacteria or lower-order animals can come about much more easily than advanced life such as humans. This is much less expected on theism than naturalism, since God has reason to favor the creation of moral agents over nonmoral agents [15]. Second, for hundreds of millions of years after the big bang the universe was occupied purely by inanimate matter. It took ~400,000 years before the first elements formed and 100 million before the first stars formed. Presumably, a perfectly good God is creating a universe for the goods therein and there do not seem to be any goods at all in this very long state of lifelessness, especially if what God really cares about is conscious life pursuing morally valuable goals. Surely it is much more expected that the universe would be filled with animate activity from the beginning on theism than naturalism. Now, of course, perhaps God needed this inanimate process for some unknown reasons or there are unknown goods therein. However, we have no antecedent reason to regard this as particularly likely on theism and such unknown goods are counterbalanced by unknown reasons God could have had for preventing this inanimate state (in addition to the known reasons I have listed). Thus, I conclude that the fine tuning evidence is quite problematically understated.

Are there any good responses to the objections I have brought up? One kind of response to the primary objection I have made, that God can get conscious life without any particular set of parameters or constants, is to suggest alternative reasons God needs fine tuning. For instance, perhaps what God really wants is not conscious life but conscious life and predictable order and/or the discoverability of natural laws, and fine-tuning is needed for that. Of course, this objection will require new justification for why God cares about these other conditions. One possible reason God would desire these conditions is that genuine moral choice requires predictable consequences. I am not sure about this response because I do not know whether there would still be discoverable order if life was being miraculously sustained under hostile conditions. It certainly seems that there could be predictable order on the second possibility I mentioned, wherein God simply changes the rules by which consciousness can instantiate. Irrespective, these are still interesting responses worth exploring. There is certainly something intuitively appealing to me about the idea that theism predicts orderliness in a way that naturalism does not. On some days, I am tempted to weigh fine-tuning in favor of theism.

Consciousness

In the face of the problems with the fine tuning argument, I think turning to an argument directly from conscious life is a better path for theism. Since this argument directly targets the data God cares about, rather than targeting conditions that allow for the data that God cares about, concerns that God could have obtained that data without the requisite conditions dissipate. There are two ways in which theism predicts conscious life. First, since reality is fundamentally mental on theism, theism entails that some sort of mind exists. Secondly, since we have strong reasons to think God would value the existence of conscious life (for reasons I talked about in the section on fine tuning), we have reasons to think they would favor the creation of other conscious life. On naturalism, reality is fundamentally indifferent, so there is no particular expectation for conscious life to emerge. Now, given the scale of the universe and fine-tuning, one might argue that conscious life coming about is probable. Since we do not know at present what conditions allow for the emergence of life or how common those conditions are with a high degree of certainty, it is hard to evaluate this possibility. However, I am not sure this is necessary given that theism entails both that conscious life of some kind exists and arguably that other conscious minds will exist, while naturalism clearly does not [16]. And, at least epistemically, I see far greater reason to expect life just from the content of the theistic hypothesis than the naturalistic one.

I find this argument intuitively persuasive, and am compelled to count this as a chip in favor of theism. Unlike many other theistic arguments, there is no ‘disputed data’ here either. Even most illusionists about phenomenal experience will admit that conscious life exists in some fashion. Unfortunately, the evidence here is still understated, for the same reasons I outlined in the section on fine-tuning.

There may be a further way to argue that consciousness supports theism over naturalism if particular positions in the philosophy of mind turn out to be true. For instance, suppose that substance dualism is true. It is hard to see how purely physical processes could give rise to wholly non-physical substances, while theism already begins with a sort of radical dualism. Thus, there may be an argument that certain kinds of dualism fit better with theism than naturalism. Indeed, one could argue that the fact that our mental lives are so dependent on physical causes is problematically understated evidence for naturalism for precisely this reason. I think this argument has merit but I do not find it at all persuasive since it relies on the truth of dualism, a highly controversial stance in the philosophy of mind that I reject.

Moral Knowledge

Deductive moral arguments are some of the worst arguments for the existence of God. This is mainly because of the weakness of the premise that says that God is necessary for moral realism. I cannot help but think that people who think this premise is plausible are not acquainted with modern metaethics, a field replete with accounts of moral realism that are perfectly compatible with naturalism (Railton 1986, Boyd, 1988, Brink, 2008 & Foot 2010 are just a few of many). However, this does not mean moral arguments for theism are doomed. Abductive arguments that argue that something about morality is more likely on theism are more promising. Here, I am going to pick what I think is the best version of this argument possible: that the existence of moral knowledge is more likely antecedently on theism than naturalism

The argument would go like this: moral knowledge is inherently valuable. It allows for the development of moral virtue and the exercise of moral behavior, which is also valuable. On SP, since God would instantiate the best state of affairs possible, those affairs would clearly include something so valuable as moral knowledge. Even on WP, an essential part of flourishing is the appreciation of moral values and moral behavior, and so a world without moral knowledge would constitute a frustration of the most important values. Thus, that God would guide life such that life is availed of moral knowledge is highly expected. Now, consider naturalism. On naturalism, there is fundamental indifference, and so no source for value guiding the development of life such that it obtains moral knowledge. Life evolved based on what was most conducive to replication and genetic fitness, and that this would produce moral judgments that were true in addition to adaptive is not guaranteed. Indeed, that those judgments that are adaptive would also happen to be those that track moral truth seems to be an enormously unlikely coincidence. Thus, the likelihood of moral knowledge is low on naturalism, and certainly lower than it is on theism.

The kind of reasoning I have employed above is fairly similar to evolutionary debunking arguments against moral realism. I predict that the extent to which one finds these kinds of debunking arguments plausible will affect how promising the argument from moral knowledge appears. I myself am much more sympathetic to this argument than other naturalists. I accept that moral knowledge, at least in a very general form, is more likely on theism antecedently than on naturalism.

However, this argument succumbs to both of the general weaknesses I identified with theistic arguments. First, it relies on disputed data. This argument relies on the existence of moral knowledge, and I, unlike many of my fellow naturalists, lean towards moral anti-realism. Furthermore, I would guess most naturalists who accept into moral realism probably reject the kind of debunking arguments needed to motivate this argument, putting this argument in quite a difficult rhetorical spot. The kind of moral realist accounts I find most plausible (that is, the kind I think would be antecedently likely on naturalism) are those that are not so susceptible to evolutionary debunking. For instance there are naturalistic moral realist accounts on which it is not surprising that evolution would align us with moral truth, for instance an Aristotelian virtue-ethics based account (Hurthouse, 2010, Thomson, 2015). But, of course, if this is the kind of moral realism that is true then the data is no longer unexpected on naturalism. So, I am in a dilemma with respect to this data (and I suspect most naturalists will be too): either I accept that it is unlikely on naturalism, but I do not think it is data, or I think it exists but is likely on naturalism.

Of course, it is important to note that for those who accept moral knowledge and can provide no good antecedent reason to expect it on naturalism this will be relevant data. However, as in the case of a necessary being, even those who accept this data should acknowledge that it is at least more controversial than the data I cited in favor of naturalism, and so cannot count as strongly towards our final tally.

This data also strikes me as problematically understated. The type of moral knowledge that exists, presuming now that we do have such knowledge, is not the kind one would expect on theism. First, there are individuals who lack (at least partial) access to moral knowledge such as psychopaths and the severely disabled. On theism, we ought to expect that God would provide equal access to moral knowledge so as to not frustrate the flourishing of particular persons. Secondly, cultures throughout history have differed in how well they track moral truth. There appear to be some general principles that emanate through societies, but there are still deep and fundamental differences in moral values between societies in space and time. For certain cultures to be morally underdeveloped would frustrate the ability of those cultures to flourish, and so that God would not guide cultures such that they all have access to moral truths is unexpected on theism.

Beauty

The argument from beauty could go like this: The ability to appreciate aesthetic value is one important kind of virtue. Thus, we have reason to expect that God will set up life such that it can appreciate this value. On the other hand, on naturalism there is nothing necessarily adaptive about tracking beauty, and so we have no reason to expect that humans will track this value.

I do not think this argument works. Firstly, what the nature of our aesthetic faculties are, that is, whether our judgments of beauty track some objective fact or are subjective is controversial, and the former is required to meet the expectations given above. Secondly, there are reasons to think that we can give an account of why evolution would equip us with aesthetic faculties (Humphrey, 1973). However, I think the decisive blow to this argument lies in its understatement of the evidence. Life is not only replete with beauty but also lots of ugliness, and our universe does not contain a lot of beauty in sensory modalities outside vision. Even more problematic, our faculties are fundamentally limited such that we cannot experience lots of aesthetic beauty. Humans have just three types of cones whose firing blends together to produce the array of colors we see (many humans are born with only two types, or lose one over time). There are many animals with more than three color receptors, meaning they can perceive colors we cannot even imagine, blending together frequencies from the ultraviolet spectrum, something we can only detect through instruments but cannot perceive. The same kind of limitations apply to our other aesthetic senses. If we expected that God would guide human life to appreciate aesthetic value, we ought to expect that he would guide that life to appreciate its deepest aspects and relevant dimensions. Clearly, that is not the situation we are in: our sense of aesthetics looks much more tuned by imperfect evolutionary processes than a perfect creator. At the very least, I see no good way to argue that that is the situation we are in, and without that key premise this argument completely lacks rhetorical force.

Religious Experience

By ‘religious experience’ I am referring to the fact that many people report feeling in contact with some non-natural reality or supernatural agency. This fact is bolstered by the further evidence that these kinds of experiences are not isolated, they seem to be fairly universal phenomena across cultures and time periods. If theism is true, there is a strong reason to expect these experiences: A relationship with God is an extremely important good, and contact from God is an important vessel by which God can impart their teachings. On naturalism, there is no particular reason to expect such a prevalence of religious experience. Perhaps one could argue from our in-built evolutionary tendency towards supernatural thinking that religious experience is evolutionarily likely (see the section on the cognitive science of religion). I do think this is a good naturalistic explanation for religious belief, but even if this argument works, it still seems to me that theism essentially entails religious experience while naturalism does not.

While very general facts about religious experience strike me as evidence for theism, unfortunately for believers, I think that the specific facts are all more likely on naturalism and tilt the scale all the way back towards naturalism.

First, consider the diversity and indeed inconsistency of religious experiences: what culture one is in heavily determines the kind of experience one has, and the sort of revelation one receives is extremely variable. This is exactly what one would expect if these experiences are the byproduct of cognitive predispositions influenced by societal conditioning (the expectation of the cognitive science of religion model). On the other hand, if theism is true, we ought to expect far more consistency in these experiences given that people are contacting a single supernatural reality. Perhaps one could make a kind of perennialist argument that God forms relationships with different peoples in very different kinds of ways. I do not buy this argument, however, because the revelatory ideas and experiences of people are not just different but wholly contradictory. Antecedently, that religious experiences are so culturally dependent and variable is extremely probable on naturalism and the cognitive science of religion in a way theism cannot match.

Second, the specific nature of religious experiences fits what one would expect on naturalism. Extremely intense religious experiences of the revelatory kind are quite rare amongst normal, functioning people. They tend to be induced under conditions that prime one for psychotic/hallucinatory states: the use of psychedelic drugs, starvation/fasting, and extreme distress. People who are more prone to schizotypy are also more likely to report having these kinds of experiences (Crespi et al., 2019). On the other hand, religious experiences of the more mild variety such as feeling an intense sense of connection when in prayer or a religious community, are much more common under normal conditions. However, these experiences are far easier to explain naturalistically, and tend to be reported by those who already have the theoretical framework of the religion they follow ‘polluting’ their interpretation of the data so to speak. If theism were true, there is no reason why the intensity of these experiences would directly line up with these accompanying natural conditions. Indeed, if theism is true, we have some reason to expect that experiences will instead be associated with need; that is, experiences should appear most in the lives of those who need them. For instance, such experiences should be most common in the lives of those suffering from the most horrific evil, for instance. Yet many people experience divine silence rather than comfort during tragedies.

Thirdly, though religious experiences are common, they are still absent in many people. I have already elaborated on this particular fact in the section on hiddenness, so I will not expand on it here.

Summary of the evidence

I have argued that the evidence favoring naturalism is stronger than the evidence favoring theism. I make this judgment on the basis that, in part, the evidence for theism succumbs to two common issues not generally present with the evidence favoring naturalism. Now, let me briefly review the facts I considered that favor naturalism and see if they succumb to the same problems.

I take it that all the facts I cited in favor of naturalism are essentially indisputable except three: the cognitive science of religion, the ineffectiveness of intercessory prayer, and nonresistant nonbelief. Since the cognitive science of religion is a relatively new field of study, any theories coming out of that field are inevitably subject to revision and tentative in nature. They are subject to criticism as scientific theories, and only time will tell as to whether they survive scrutiny. While studies on intercessory prayer are rigorous, we are still working on limited, empirical evidence, making that data inherently less certain than more universally agreed upon facts. Yet, these two pieces of data are still less forcefully disputed than most of the facts favoring theism. Theists might deny that nonresistant nonbelief exists on the basis that people have hidden resistances—perhaps the noetic effects of sin lead them astray in ways that cannot be clearly seen, for instance. Aside from being antecedently implausible, this argument becomes very hard to make as we learn more about the actual sources of nonbelief: see my paper on the cognitive science of religion and (De Cruz, 2018) for more on this.

Are any of the pieces of evidence favoring naturalism understated? This is a harder question to answer, but as far as I can tell they are not, with one possible exception. Many theists could argue that the ‘dependence of the mind on the physical’ is an understated fact because more fully stated facts about consciousness demonstrate substance dualism. I agree that if substance dualism were true, this would be more expected on theism than naturalism, but I strongly disagree that the fully stated facts about consciousness favor dualism. On the contrary, I think the fully stated facts about consciousness even more strongly favor materialism, making them even more evidence in favor of naturalism.

On the other hand, I have argued that all of the facts cited in favor of theism except two— fine-tuning and consciousness—are highly disputable. Secondly, they all rely on understated evidence. In general, theism can do a fair job of accounting for highly general, unspecified facts—but the second we examine the specifics of those facts, we find that the details highly favor naturalism over theism. Thus, I conclude that the extant evidence available heavily favors naturalism over theism.

There will inevitably be facts that people think should have been included in this analysis. I cannot include every fact that has been mentioned in this discussion without making a post far longer than I have, such is the nature of this issue. What I have done is include the facts that I personally find the most strongly compelling for both naturalism and theism. In the future, I will make small followups this post where I address other facts that I feel I have neglected or that people suggest.

Prior Probability of Theism and Naturalism

As I mentioned earlier, there are two key factors we need to examine when evaluating the relative merit of theism and naturalism. The first is the explanatory power of the two theories, something we have now extensively outlined. The second factor is the prior probability of theism and naturalism. This is an inherently more difficult criteria to evaluate because there is more disagreement about exactly how to do it. Subjective bayesians would argue that there simply is no clear-cut, objective way to evaluate the prior probability of a hypothesis just from its content alone. Even thinking more broadly abductively without invoking Bayes theorem, while many philosophers and scientists agree that there can be certain parts of a hypothesis that make it more theoretically virtuous, they disagree on what virtues count and how to evaluate each virtue in question. In what follows, I am going to briefly identify the factors that seem most consistently applied and relevant, and argue that naturalism consistently fares better intrinsically as a theory than theism. This section will be very short relative to the first section both because I think it is less important and more contentious, but I might expand/elaborate on this section in another post:

Simplicity/Modesty

That we ought to prefer theories that posit less is a very widely accepted principle. The most straightforward reason for thinking this is that the more claims a theory makes, the more ways it has to be false, and thus the less probable it is prior to looking at any evidence.

Recall that naturalism is the view that causal reality is fundamentally physical and precedes any other order of reality. The parallel claim to this in terms of simplicity would be the reverse: that causal reality is fundamentally mental and precedes any other order of reality. Theism agrees with this claim, but adds on a number of controversial additions to it. Namely, that this reality is singular, omnipotent, omniscient, and omnibenevolent. Thus, theism is clearly less modest than naturalism than a theory (Draper, 2017).

An even simpler way to think about the difference between theism and naturalism is as follows: both theists and naturalists accept that the natural universe exists. However, naturalists only believe in this universe, while theists add on an additional, prior layer of mental reality to the universe.

In conclusion, naturalism is clearly a more simple theory than theism, giving us some reason to prefer it on this metric.

Precision/Informativeness

The precision or informativeness of a hypothesis refers to the degree with which we can precisely specify what a hypothesis does or does not predict (Dawes, 2015). The more precisely we can make predictions based on a theory, the better that theory is. A serious issue with theism is that it is wrapped in a certain layer of mysteriousness that makes it very hard to use to predict data. A divine, omniscient agent is a mind so unlike and far above our own that it likely has intentions and plans that we simply cannot comprehend or understand. God always has a near infinite set of choices that we can only understand some small subset of. Indeed, skeptical theists use this particular point to precisely attempt to undermine our ability to judge whether the suffering we see is really gratuitous, or whether we can really judge the likelihood of particular pieces of data on theism. I do not go so far as to embrace such extreme skepticism, but it remains true that all predictions we make about theism are inevitably going to be tentative due to our limited knowledge of this divine reality.

There is a sense in which naturalism, broadly stated, remains somewhat uninformative. Merely that reality is indifferent is a fairly vague hypothesis. However, at least with naturalism we are dealing with the physical universe: a limited system that operates based on laws we can study via science, and does not have infinitely many ‘choices’ to make. Thus, while there still remains some mystery about a fundamentally physical reality, it is far less mysterious than a divine agent.

Fit with background knowledge

Here, I am referring to the extent to which the posits of our hypotheses fit with our background understanding of the world (Dawes, 2015). The fundamental mechanisms that theism posits is unlike any mechanism we are familiar with. Usually, we understand things as being caused by prior physical states of the world that operate in lawlike ways. The actions of a supernatural being are a totally foreign, mysterious concept. We have no idea how such actions would work, not least because we understand actions as being done by creatures in physical ways that are effective only in virtue of the actions of physical bodies, not everywhere and anytime non physically, as in the case of theism. Even worse, it seems like our everyday understanding of mental states makes no sense when applied to the actions of a perfect, divine agent. Here, I draw on some of the literature on the problems with divine properties, summarized nicely by a friend of mine in this post. To give two examples they mention, our common understanding of mental states when applied to every person we know of is that they are temporal and that they exhibit intentionality in that they can represent objects outside of themselves. How to logically square these traits with the properties of God, who is timeless and omnipresent, is seriously problematic. But, even if they can be squared with theism, what they show is that theism fits very poorly with our background understanding of the way things work. On the other hand, naturalism appeals to mechanisms that we perfectly understand in everyday life: the mechanistic laws of nature and physical forces underlying the universe.

One note on this section: I am aware that it is a bit murky whether this would fall into the category of ‘prior probability’ or the data, considering that in this section I’m abstracting away from all the evidence. This section has some overlap with the ‘success of the naturalistic research programme’ evidence. I simply note this murkiness as a problem in general with such abductive comparisons and move on.

Personal Considerations

A highly neglected part of these posts are the personal factors and biographical facts that have led one to disbelief. As much as we want to believe that our beliefs are entirely based on dispassionate, intellectual consideration, the influence of our unique background undoubtedly plays a role. Even when our reasons are based on careful consideration of the arguments, the story underlying how we came to consider these arguments and the perspective we take on them is still, I think, essential. Therefore, I am going to talk about the personal background that has led me to the positions I hold today and about some of the reasons I am inclined towards atheism that have not come out in my post thus far.

I grew up in a nonreligious household. And by ‘nonreligious’ I do not mean ‘antireligious’ but something more akin to ‘irreligion’. My parents neither encouraged me to adopt particular views in favor of theism and other religions nor against them. This is not to say that religion was not a part of our discussions or activities—but rather that it was never presented in a prescriptive way. In this sense, I had quite a unique childhood compared to most people I know [17]. And, in my opinion, this was an extremely beneficial decision on my parents’ part. Being allowed to make my own decisions on this matter made me think and grapple with these questions in a much more complex, open way than I would have if I had a particular belief system foisted upon me. Indeed, I think part of the reason I became so interested in philosophy and pursued an academic career is because I was encouraged to think about these questions for myself at an early age. Furthermore, this lack of pressure from my parents freed me from the arational influence that such authority figures can have on our beliefs. Children face an immense, innate pressure to adopt the beliefs of their parents, and this often leads to undue credulity for their opinions or intentional rebellion against them. Even as one grows older I think these influences implicitly affect us more than we think they do.

I was, from an extremely young age, very skeptical of religion. The stories contained in religious texts always struck me as highly facile and obviously manmade accounts of the origins of life and the universe. Additionally, I was struck by the incredible credulity with which many people uncritically accepted these stories. The universe did not at all look to me like a place that divine intentions had touched. I do not think there was any time in my life where I believed in God. When I was very young, however, I was much more agnostic. I distinctly remember that I would sometimes try to talk or pray to God, and what I remember most from these attempts is that they made absolutely no appreciable difference to my life or psychology. Talking to God was no different than talking to imaginary friends or Santa Claus. So, I settled on the idea that it was made-up. As I grew older and actually learned more about the sophisticated arguments for and against God’s existence, my views became more informed, but the fundamental reason I rejected theism in many ways has stayed the same: the universe does not look like a place designed by God.

A second important factor in my personal development and view on the supernatural more generally came due to my foray into the psychedelic community. In college, I became fascinated by the psychedelic experience. For those unfamiliar with psychedelic drugs, they have an incredible ability to alter consciousness: for instance, making one feel as if all time has melted away or making one feel as if they are in contact with a fundamental consciousness. One interesting note about psychedelic drugs is that people who take them often experience long-lasting changes to their personality and belief structure. In fact, many people who take psychedelic drugs report becoming more spiritual due to their experiences. I can see why: these experiences often seem very significant and can seem as if they provide access into a special, spiritual realm of reality.

Despite the seemingly profound nature of the psychedelic experience, what psychedelics lead me to notice more is how susceptible we are to projecting our own agency and consciousness onto everything else. Much of the psychedelic community has a tendency to get wrapped up in woo ideas such as astrology, astral projection, ESP, telepathy, transcendental meditation, the occult, sacred geometry, alien abduction, spiritual channeling, and much, much more. These ideas were often fueled by big figures in the psychedelic movements like Terrence McKenna and more cult-like movements like the Brotherhood of Eternal Love. What underlies almost all these ideas is the hyper-projection of our own experience onto everything else. As I discuss in my paper on the cognitive science of religion, we are highly vulnerable to the hyper detection of agency, and that, combined with the cognitive biases we are all likely to succumb to, makes us very prone to buying into supernatural ideas. I have seen many people who use psychedelics get completely wrapped up in their own bullshit, for lack of a better word, guided by this in-built tendency. This bullshit almost always insidiously drifts towards the tendency to think that our consciousness creates reality, that our thoughts and mental lives are somehow important to the universe, and that fundamental reality must resemble in some way our internal experience. Unfortunately, this semi-delusional line of thinking has only created more confusion, and its failure as an idea in the psychedelic movement is strongly supported by the empirical evidence that these psychedelic movements totally failed to transform society as promised, and only lead invariably to disaster, and the individuals who tried to use them to glean insight into reality not only failed to gain any insight, but often ended up endangering their own lives. For those interested further in this subject, I highly recommend the work of James Kent, whose podcast I have linked here. His writing on the psychedelic movements is of the highest quality, and strongly inspired my views on the subject [18].

All of this might lead you to believe that I am against psychedelic usage or regard it negatively, but this is actually not true: I still think psychedelics are enormously interesting pathways into consciousness, and that they need to be legalized and studied rather than criminalized. What I did learn, however, from my foray into this rabbithole is that because we are so prone to go down this delusional line of inwards thinking and agency projection, we need to treat theories which try to imbue the universe with agency with extra caution and skepticism, and to resist the need to imbue everything that happens to us with meaning.

Summary of the Case

I can now finally summarize the overall case I have made for naturalism in this post. In general, the world is a highly messy, flawed, and seemingly random place that fits better with an indifferent hypothesis than one that posits fundamental perfection. Additionally, theism faces serious problems as an explanation in general: It is highly parasitic on very general, broad facts about the world but fails when it comes to the specifics. Additionally, naturalism also already has an intrinsic advantage compared to theism due to its simplicity, fit with the background data, and relative precision in comparison with theism. I also think we need to treat theories such as theism with extra caution, because we are so prone to projecting our agency into the universe and falling down the rabbithole of delusional thinking.

Obviously, there is so much to say on this issue that it would take many books to say it all. I hope I have at least given readers an idea of my basic reasons for disbelief, and this post is always subject to revision in the consideration of new arguments I have not had a chance to look at.

One final point is in order: I have devoted many years of my life to combing through the literature on philosophy of religion at this point, and read all the major authors arguing for and against theism, including at least hundreds of journal articles on the subject. There reaches a point that one feels there is not much left out there that could significantly change one’s mind. I feel myself starting to be in this position. In addition, toiling over a topic that is fundamentally about a reality I do not think exists has started to look more and more like a waste of time, to put it bluntly. There are more pressing questions I need to face. I think this is why many naturalists/atheists do not devote significant time to philosophy of religion. Theism is believed in widely enough and is prima facie viable enough to be taken seriously, but in many ways I feel I have now met my epistemic burden. Therefore, this post might, in many ways, be my goodbye to philosophy of religion (at least as an extremely serious pursuit) as I move on to new topics. I have much more to say in the way of final reflections on the topic, so stay tuned for that, but expect content on this blog to shift towards other areas of philosophy in the future. That is not to say I will never post about philosophy of religion, I am sure I will never give it up completely, but its status as my primary pursuit has now come to an end.

Footnotes

[1] See, for instance, Graham Oppy’s Why I’m not a Christian and this post by my friend Emerson Green.

[2] For another good source discussing these issues, read Jeffrey Lowder here: https://secularfrontier.infidels.org/2016/02/weighing-theistic-evidence-against-naturalistic-evidence/ and here: (Chandler & Harrison, 2012)

[3] See (Wielenberg, 2004, Steinberg, 2007) if you are interested in the kind of motivations that compel me to accept SP.

[4] I have adapted this weaker principle from Robert McKim’s excellent discussion of divine hiddenness (McKim, 2015). However, I have modified it slightly, since McKim’s original principle specifies flourishing for us (human beings) specifically, and in this piece I want to think much more broadly about what to expect on theism, and abstract away from even our existence.

[5] One worry is that a similar objection to SP can be levied against WP; namely, that there are infinite levels of flourishing and value that can be achieved. Firstly, states of affairs seem more plausibly infinite in extent than levels of flourishing or states of value. However, more importantly, note that this principle does not ask for the greatest quantitative level of flourishing to be instantiated, but just that the most important and deepest aspects of flourishing are realized. So, for instance, it is not a violation of WP that God fails to create the maximum amount of happy beings that they can create, but it is a violation of WP if the most important aspects of happiness are not realized or if happiness does not exist at all. As we will see later, there are many values for which it is unreasonable to suppose that there can always be ‘deeper aspects’ of those values.

[6] Since the existence of human beings is extremely specific, and unlikely on any hypothesis in virtue of that specificity, it may be preferable to say that my argument is from the fact that beings of our relative level of impressiveness exist. Because this makes no difference to the course of the discussion, I will simply note this and continue to talk about the human species.

[7] As an example, many people who take MDMA report experiencing a kind of love and empathy for every other living being that far surpasses anything that could possibly experience without the use of that substance, both in intensity and kind. Many people who take psychedelics report a kind of connection with the universe and loss of ‘ego/self’ that is normally unattainable.

[8] Remember, importantly, that in this discussion we are presuming that life exists on both theism and naturalism and asking what kind of life to expect. Thus, any objection that life itself is very unexpected on naturalism simply does not apply to this argument (we will get to that).

[9] I (unsurprisingly) regard both of these theories as obviously false.

[10] See my essay on the cognitive science of religion and Draper, 2015 for further responses to this kind of argument.

[11] Note, importantly, that I am not making the further claim that our minds are completely physical in nature. My point is compatible with dualisms that insist some parts of our experience must be explained in non-physical terms; even the most ardent dualists must still admit that clearly our mental lives are highly dependent on physical processes.

[12] Again, see my paper on the cognitive science of religion or (Draper, 2015) for a further discussion of this argument.

[13] For those curious why, I am sympathetic to Frankfurt-style compatibilism. I think there are two major challenges to compatibilism: first, the principle of alternative possibilities, and second, a kind of sourcehood challenge that says we are never the ultimate cause of our actions. I think Frankfurt adequately addressed the PAP, and that Susan Wolf showed that the sourcehood objection rests on an incoherent demand.

[14] Some would argue it cannot count at all. I will not go this far.

[15] Justin Schieber pointed this out to me.

[16] Perhaps the best argument against this is to claim that a perfect God would never create anything at all because God would never move the world from a state of perfection to imperfection. This is a plausible argument I am still thinking about.

[17] I also seem to be especially unique in the philosophy of religion field. Most nonbelievers in the field tend to come from religious backgrounds and then deconvert later on.

[18] One great example of his work, in addition to the podcast I link to, is his classic article on Terrence McKenna’s theory that DMT revealed ‘machine elves’: http://tripzine.com/listing.php?id=dmt_pickover

Works Cited

Adams, R. M. (1972). Must god create the best? The Philosophical Review, 81(3), 317. https://doi.org/10.2307/2184329

Adams, R. M. (1985). Plantinga on the Problem of Evil. In J. E. Tomberlin & P. Van Inwagen (Eds.), Alvin Plantinga (pp. 225–255). essay, D. Reidel Pub. Co.

Boyd, R. (1988). How to be a moral realist. In G. Sayre-MacCord (Ed.), Essays on moral realism (pp. 181–228). essay, Cornell Univ. Press.

Brink, D. O. (2008). Moral realism and the foundations of Ethics. Cambridge University Press.

Burns, J. M., & Swerdlow, R. H. (2003). Right orbitofrontal tumor with pedophilia symptom and Constructional apraxia sign. Archives of Neurology, 60(3), 437. https://doi.org/10.1001/archneur.60.3.437

Chandler, J., & Harrison, V. S. (2012). Probability in the philosophy of religion. Oxford Univ. Press.

Chisholm, R. M. (1968). The defeat of good and evil. Proceedings and Addresses of the American Philosophical Association, 42, 21. https://doi.org/10.2307/3130021

Churchland, P. (2007). The Evolving Fortunes of Eliminative Materialism. In B. P. McLaughlin & J. Cohen (Eds.), Contemporary debates in philosophy of mind (pp. 160–180). essay, Blackwell.

Crespi, B., Dinsdale, N., Read, S., & Hurd, P. (2019). Spirituality, dimensional autism, and schizotypal traits: The search for meaning. PLOS ONE, 14(3). https://doi.org/10.1371/journal.pone.0213456

Dawes, G. W. (2015). Theism and explanation. Routledge.

De Cruz, H. (2018). Divine hiddenness and the cognitive science of religion. In A. Green & E. Stump (Eds.), Hidden divinity and religious belief: New perspectives (pp. 53–68). essay, Cambridge University Press.

Draper, P. (1989). Pain and pleasure: An evidential problem for Theists. Noûs, 23(3), 331. https://doi.org/10.2307/2215486

Draper, P. (2004). More pain and pleasure: A reply to Otte. In V. P. Inwagen (Ed.), Christian faith and the problem of evil (pp. 41–54). essay, W.B. Eerdmans Pub.

Draper, P. (2008). Collins' Case for Cosmic Design. Internet Infidels. Retrieved January 20, 2022, from https://infidels.org/library/modern/paul-draper-no-design/

Draper, P. (2010). Cumulative Cases. In C. Taliaferro, P. Draper, & P. L. Quinn (Eds.), A companion to Philosophy of Religion (pp. 414–424). essay, Wiley-Blackwell.

Draper, P. (2013). The limitations of pure skeptical theism. Res Philosophica, 90(1), 97–111. https://doi.org/10.11612/resphil.2013.90.1.6

Draper, P. (2015). Evolution and the Problem of Evil. In L. P. Pojman & M. C. Rea (Eds.), Philosophy of religion: An anthology (pp. 271–283). essay, Cengage Learning.

Draper, P. (2017). God, Evil, and the Nature of Light. In The Cambridge companion to the problem of evil (pp. 65–84). essay, Cambridge University Press.

Draper, Paul, "Atheism and Agnosticism", The Stanford Encyclopedia of Philosophy (Summer 2022 Edition), Edward N. Zalta (ed.), URL = <https: plato.stanford.edu="" archives="" sum2022="" entries="" atheism-agnosticism="">. <div></div></https:>

Foot, P. (2010). Natural goodness. Clarendon Press.

Hasker, W. (2010). All too skeptical theism. International Journal for Philosophy of Religion, 68(1-3), 15–29. https://doi.org/10.1007/s11153-010-9252-7

Hume, D., & Coleman, D. (2007). Dialogues concerning natural religion and other writings. Cambridge University Press.

Humphrey, N. K. (1973). The illusion of beauty. Perception, 2(4), 429–439. https://doi.org/10.1068/p020429

Hursthouse, R. (2010). On virtue ethics. Oxford University Press.

Inwagen, V. P. (2019). Metaphysics. Routledge.

Kane, G. S. (1975). The failure of soul-making theodicy. International Journal for Philosophy of Religion, 6(1), 1–22. https://doi.org/10.1007/bf00136996

Koenigs, M., Young, L., Adolphs, R., Tranel, D., Cushman, F., Hauser, M., & Damasio, A. (2007). Damage to the prefrontal cortex increases utilitarian moral judgements. Nature, 446(7138), 908–911. https://doi.org/10.1038/nature05631

Koons, R. C., & Pruss, A. R. (2020). Skepticism and the principle of sufficient reason. Philosophical Studies, 178(4), 1079–1099. https://doi.org/10.1007/s11098-020-01482-3

Lavigne, K. M., Menon, M., & Woodward, T. S. (2019). Functional Brain Networks underlying evidence integration and delusions in schizophrenia. Schizophrenia Bulletin, 46(1), 175–183. https://doi.org/10.1093/schbul/sbz032

Maitzen, S. (2006). Divine hiddenness and the demographics of theism. Religious Studies, 42(2), 177–191. https://doi.org/10.1017/s0034412506008274

Manson, Neil. A. (2018). How not to be generous to fine-tuning sceptics. Religious Studies, 56(3), 303–317. https://doi.org/10.1017/s0034412518000586

Masters, K. S., & Spielmans, G. I. (2007). Prayer and health: Review, meta-analysis, and research agenda. Journal of Behavioral Medicine, 30(5), 447–447. https://doi.org/10.1007/s10865-007-9120-9

Masters, K. S., Spielmans, G. I., & Goodson, J. T. (2006). Are there demonstrable effects of distant intercessory prayer? A meta-analytic review. Annals of Behavioral Medicine, 32(1), 21–26. https://doi.org/10.1207/s15324796abm3201_3

Railton, P. (1986). Moral realism. The Philosophical Review, 95(2), 163. https://doi.org/10.2307/2185589

Rowe, W. L. (2010). Can god be free? Clarendon Press, Oxford University Press.

Russell, B. (2018). The problem of evil and replies to some important responses. European Journal for Philosophy of Religion, 10(3), 105–131. https://doi.org/10.24204/ejpr.v10i3.2590

Schellenberg, J. L. (2012). The wisdom to doubt: A justification of religious skepticism. Cornell University Press.

Sinhababu, N. (2017). Divine fine-tuning vs. Electrons in Love. American Philosophical Quarterly, 54(1), 89–98. https://doi.org/10.2307/44982126

Sobel, J. H. (2009). Logic and theism: Arguments for and against beliefs in god. Cambridge Univ. Press.

Steinberg, J. R. (2007). Leibniz, creation and the best of All possible worlds. International Journal for Philosophy of Religion, 62(3), 123–133. https://doi.org/10.1007/s11153-007-9136-7

Stich, S. P. (1979). Do animals have beliefs? Australasian Journal of Philosophy, 57(1), 15–28. https://doi.org/10.1080/00048407912341011

Swinburne, R. (2014). The existence of god. Clarendon Press.

Theriault, J., Waytz, A., Heiphetz, L., & Young, L. (2017). Examining overlap in behavioral and neural representations of morals, facts, and preferences. Journal of Experimental Psychology: General, 146(11), 1586–1605. https://doi.org/10.1037/xge0000350

Thomson, J. J. (2015). Normativity. Open Court.

Tooley, M. (2019). The problem of evil. Cambridge University Press.

Walker, M. (2009). The anthropic argument against the existence of god. Sophia, 48(4), 351–378. https://doi.org/10.1007/s11841-009-0137-0

Wielenberg, E. J. (2004). A morally unsurpassable god must create the best. Religious Studies, 40(1), 43–62. https://doi.org/10.1017/s0034412503006796

Wielenberg, E. J. (2010). Sceptical theism and divine lies. Religious Studies, 46(4), 509–523. https://doi.org/10.1017/s0034412510000247

Wykstra, S. J. (1984). The Humean obstacle to evidential arguments from suffering: On avoiding the evils of ?appearance? International Journal for Philosophy of Religion, 16(2), 73–93. https://doi.org/10.1007/bf00136567

Log of Edits

12/10/22: Fixed typos throughout, updated citations, and fixed a few incorrect hyperlinks (notably, I linked the to the wrong post on the PSR on my friend's blog). Additionally, added some minor argumentative points in the sections on pain & pleasure and fine-tuning in line with some good suggestions from friends of mine.

12/15/22: Added novel arguments to the sections on fine-tuning and religious diversity that cropped up in old notes of mine. Specifically, I point out that the fact that your likelihood of being a theist is primarily set by cultural/familial circumstance is much more likely antecedently on naturalism than theism, and I argue that theists are in a bind when responding to certain objections to the fine-tuning argument: one of the key reasons for denying that the hostility of the universe is evidence against theism prevents you from objecting to the claim that God would miraculously sustain life under hostile conditions.

12/23/22: Added an additional way in which the fine-tuning evidence is understated, based on something Justin Schieber pointed out to me: The universe is far more conducive to the production of nonmoral agents than moral agents.

2/5/2023: Updated the section on Robert Adams’ paper to make it a little more comprehensive and added a link to the youtube version of the post!

2/7/2023: Added a few more lines about how to choose background evidence, to clear up misconceptions about what any one argument in this post assumes.

2/10/2023: Added a response to an objection that has come up a few times on the success of naturalistic science argument: namely, that naturalism does not predict that a research program would be successful to begin with. Also, updated the sections on the ineffectiveness of prayer, free will, and humans with some minor argumentative/style changes. Added a response to the idea that free will in the natural world is caused by the free will immaterial/demonic spirits. Still left to do: update the sections on soul-making and defeat, and expand the section on compatibilism and free will theodicies.

Why I'm an Atheist

Introduction

Table of contents

Methodology

Abduction and Bayes’ Theorem

What should we expect on theism?

Expansions to Theism

For Naturalism

Humans

Evolution

Dependence of Mind on the Physical

Cognitive Science of Religion

Success of Naturalistic Explanation

Suffering

Biological Role of Pain and Pleasure:

Flourishing and Languishing:

Triumph and Tragedy:

Geographical Distribution of Suffering:

Particularly horrendous evil:

Theodicies and Skeptical Theism:

Free will:

Soul-making:

Heaven & Defeat:

Skeptical Theism:

Ineffectiveness of Prayer

Religious Diversity

Flawed Religions

Nonresistant Nonbelief

Introduction to Theistic Evidence

For Theism

Necessary & Contingent Existence

Fine-Tuning

Consciousness

Moral Knowledge

Beauty

Religious Experience

Summary of the evidence

Prior Probability of Theism and Naturalism

Simplicity/Modesty

Precision/Informativeness

Fit with background knowledge

Personal Considerations

Summary of the Case

Footnotes

Works Cited

Log of Edits

Comments

Post a Comment

Popular posts from this blog

A Comprehensive Critique of the Argument from Psychophysical Harmony

Intuition, Metaphilosophy, and Psychophysical Harmony: a Response to Bentham's Bulldog