Introduction
In this article, I will explore the implications of engaging with non-ideal theorising for the philosophy of religion, specifically, the problem of evil. In 2005, Charles Mills criticised mainstream analytic philosophers for relying on overly idealised conditions when theorising. Far too often, when analytic philosophers construct their theories, they do so by ignoring relevant aspects of how the world actually is – most problematically, for Mills, by ignoring oppressive structures by which the social world operates – in favour of an idealised representation of the world. This approach to theory construction Mills – following John Rawls – calls ideal theory, writing that ‘[w]hat distinguishes ideal theory is the reliance on idealization to the exclusion, or at least marginalization, of the actual’ (Reference Mills2005, 168). By contrast, non-ideal theorists avoid abstracting away from the messy details of the social world, but would rather seek to include these ‘non-ideal realities’ in their theorising.
Mills’s highly influential, though controversial, critique has contributed to the rise of a debate in political philosophy between ideal theorists and non-ideal theorists, and in recent years the critique has influenced other philosophical sub-disciplines as well, prompting scholars to initiate projects in non-ideal philosophy of language (Beaver and Stanley Reference Beaver and Stanley2019, Reference Beaver and Stanley2023; Keiser Reference Keiser2023), non-ideal philosophy of mind and language (Engelhardt Reference Engelhardt2024), non-ideal feminist theory (Tessman Reference Tessman2009), non-ideal social ontology (Burman Reference Burman2023), and non-ideal epistemology (Lackey Reference Lackey2021; McKenna Reference McKenna2023). This rise of non-ideal theorising in the broader philosophical academic world – a sign of which we can see in the publication of The Routledge Handbook of Non-Ideal Theory (2024) – raises questions regarding the theoretical assumptions regularly employed in the philosophy of religion as well. While some have observed that philosophers of religion tend to focus and rely on abstract, idealised versions of their objects of analysis (see e.g. Chignell Reference Chignell, Eckel, Speight and DuJardin2021; Knepper Reference Knepper, Eckel, Speight and DuJardin2021; Panchuk and Rea Reference Panchuk and Rea2020), to my knowledge, only one philosopher of religion – Sameer Yadav (Reference Yadav, Hereth and Timple2020) – has in a publication considered (though he does not utilise or engage with) non-ideal theory as a possible theory-choice to be applied in the philosophy of religion.Footnote 1 Due to this lacuna, my aim in this article is to apply non-ideal theorising to a paradigmatic example of how the problem of evil has traditionally been approached, in order to discern the implications non-ideal theorising may have for (one area of) the philosophy of religion. Due to the limited space of an article, I will limit my discussion to the theodicy (and epistemology) developed by Richard Swinburne.
Background: the critique of ideal theory
The critique of ideal theory starts from the observation that the methodological approach traditionally favoured by (analytic) philosophers has tended to rely on optimal, or ideal, conditions in theory construction. When theorising, philosophers often rely on simplified versions of their object of analysis, by abstracting away (much) personal and contextual details, leaving them with a version of their object of analysis which is easier to work with, but which also differs significantly from what it actually looks like. Åsa Burman paints a picture of the theoretical world philosophers tend to operate with – a world which is harmonious, where people get along with each other and go on picnics together. It is an equal society where conflict is neither frequent nor severe, and where people are race-less and gender-less, because differences in race, gender, or sexuality are not considered relevant. Such a world has abstracted away from the messiness of everyday life, such as material disparities and differences between individuals, leaving an idealised version of the world and the people who inhabit it (Burman Reference Burman2023, 1–2). Critics of this approach – non-ideal theorists – not only argue that such idealised theories paint a flawed, unrealistic image of our world, but that philosophers who construct such idealised theories, or ideal theories, end up abstracting away factors relevant for theory construction. As such, idealisations are inadequate starting points for analysing social problems, as they ‘tend to be partial and distorting, obscuring the heterogeneity and complexity of actual experiences and practices’ (Medina Reference Medina2013, 11). For the non-ideal theorist, the problem, here, is not with abstractions and idealisations, per se, since non-ideal theorists will also rely on abstractions and idealisations, as these are necessary for theorising. Rather, the problem concerns the types of abstractions and idealisations typical of ideal theorising.
Ideal theory has been criticised (and defended) on many different fronts, with different critics focusing on different aspects and tendencies of ideal theorising.Footnote 2 For our purposes here, my approach to non-ideal theory and methodology follows a Millsian tradition, which specifically concerns itself with analysing social injustices and oppressive social structures and conditions. Thus, for Mills, and non-ideal theorists following him, the most problematic feature of ideal theory is the tendency to abstract away from oppressive social relations and their legacy, and their effect on people and society. According to Mills, ideal theory is characterised by six traits:
Idealised social ontology. Ideal theory ‘will typically assume the abstract and undifferentiated equal atomic individuals of classical liberalism. Thus it will abstract away from relations of structural domination, exploitation, coercion, and oppression, which in reality, of course, will profoundly shape the ontology of those same individuals, locating them in superior and inferior positions in social hierarchies of various kinds’ (Mills Reference Mills2005, 168. Emphasis in original).
Idealised capacities. Ideal theory portrays individuals with ‘completely unrealistic capacities attributed to them’ (Mills Reference Mills2005, 168). Most often, individuals tend to be portrayed as hyperrational, or as unaffected by structural prejudice and bias.
Silence on oppression. ‘Almost by definition … little or nothing will be said on actual historic oppression and its legacy in the present, or current ongoing oppression, though these may be gestured at in a vague or promissory way’ (Mills Reference Mills2005, 168).
Ideal social institutions. By consequence, the ways in which systemic oppression shapes social institutions, along with how these institutions serve to systemically disadvantage women, the poor, or racial minorities will not be covered in ideal theory.
Idealised cognitive sphere. An idealised cognitive sphere will be presupposed. ‘In other words, as a corollary of the general ignoring of oppression, the consequences of oppression for the social cognition of these agents, both the advantaged and the disadvantaged, will typically not be recognized, let alone theorized. A general social transparency will be presumed, with cognitive obstacles minimized as limited to biases of self-interest or the intrinsic difficulties of understanding the world, and little or no attention paid to the distinctive role of hegemonic ideologies and group-specific experience in distorting our perceptions and conceptions of the social order’ (Mills Reference Mills2005, 169).
Strict compliance. In ideal theories, ‘[e]veryone is presumed to act justly and to do his part in upholding just institutions’ (Mills Reference Mills2005, 169), ‘thus abstracting away from how people actually behave and how oppression influences our behaviors and interactions’ (Engelhardt Reference Engelhardt2024, 11).
The trait silence on oppression is especially important, since it will have consequences for how the other traits are theorised. Importantly, as seen above, forms of oppression, like racism or sexism, may merit mention in theoretical works, but ideal theorists will tend to abstract away from how racism and sexism manifest in individuals and institutions, and how they structure society. So, while hyperrational agents may be unrealistic, agents unaffected by structural prejudice (which affects how much credibility they grant different people) is not only unrealistic; such idealisations also obscure oppressive social relations and their effects.
According to Robin McKenna, one of the problems which arise when you engage in idealisations ‘is that you might end up ignoring phenomena that are of real interest because you work at a level of idealization from which they are rendered invisible’ (McKenna Reference McKenna2023, 3). For example, ‘[i]f you ignore the fact that there are power differentials between epistemic agents, you are hardly going to consider the epistemological consequences of social power differentials between epistemic agents (e.g. do we afford more credibility to agents with more social power)’ (McKenna Reference McKenna2023, 6). Later, I shall consult critical epistemologies to illustrate such phenomena and their epistemological relevance to the problem of evil.
Again, not all abstractions and idealisations are problematic for the non-ideal theorist. But ideal theorists tend to adopt idealisations which obscure features of the actual world (McKenna Reference McKenna2023, 26; Burman Reference Burman2023, 3), especially those concerning oppression, that is, ‘oppression-obscuring idealizations’ (Engelhardt Reference Engelhardt2024, 10). As such, ideal theories
are best viewed as tendencies. Ideal [theorists] tend to make certain idealizations while non-ideal [theorists] tend to avoid these idealizations. … Even the non-ideal [theorist] engages in some idealization and abstraction for the simple reason that idealization and abstraction is a necessary part of theorizing. The crucial question concerns which idealizations you engage in … The non-ideal [theorist’s] complaint is that the ideal [theorist] tends to make idealizations that they should not make given the questions or issues they are trying to address (McKenna Reference McKenna2023, 20. Emphasis in original).
One of Mills’s points, as McKenna highlights, is that
‘if we build the sorts of idealizations that are typical of ideal theory into our theorizing, we will lack the tools to understand certain things. If we ignore the reality of oppression, human limitations, and the flawed nature of our environment … then we will lack the tools to understand and explain these things’ (McKenna Reference McKenna2023, 25).
In other words (for our purposes here), proper explanations require a proper understanding, and a proper understanding cannot be achieved if we abstract away relevant, non-ideal features of reality from our theoretical accounts.
Non-ideal theory as methodology
‘But what does it actually and precisely mean to engage in non-ideal philosophy?’ asks Hilkje Hänel and Johanna Müller (Reference Hänel and Müller2022, 33). While there are some understandings or assumptions that non-ideal approaches concern a turn to the real world in some sense, or that non-ideal approaches are topic specific, Hänel and Müller argue that ‘non-ideal philosophy is distinctive in its use of a certain methodology and that this methodology is, in an important sense, prior to specific topics (such as injustice, oppression, etc.) insofar as it reveals these topics’ (Reference Hänel and Müller2022, 37). Another theorist, Lisa Tessman, distinguishes between two ways in which one can employ a non-ideal methodology – a weak sense, and a strong sense. ‘In a weak sense, simply employing a methodology of examining actual rather than counterfactual/hypothetical ideal(ized) worlds qualifies a work as an instance of non-ideal theory’ (Tessman Reference Tessman2009, xviii). By contrast, for a work to qualify as non-ideal theory in the strong sense, it cannot merely examine and theorise actual lives, but must also
focus on the lives of those who live under conditions that are particularly distant from the ideal (in the sense of perfect), a distance that is generated and sustained by systemic sources of injustice. While no one lives an ideal … life or under ideal … conditions, some people live worse or more difficult lives, and under worse or more difficult conditions, than others. … Developing theory that reflects the lives of women and others who face systemic injustice requires theory that is non-ideal(ized) in this stronger sense: it focuses on the actualities of people whose lives, through injustice, are kept distant from an ideal (Tessman Reference Tessman2009, xviii. Emphasis in original).
Non-ideal theory in Tessman’s weak sense would be extremely weak indeed. In my view, such projects are better labelled ‘applied philosophy’, or ‘experimental philosophy’. When doing non-ideal theory (in the Millsian tradition) one must remember that, for Mills, the trait silence on oppression is especially important, because it affects how the other traits are theorised, and because it has the effect of hiding or marginalising the experiences of oppressed people. Thus, in my view, any non-ideal philosophy must account for oppression, or other systemic injustices, and their effects to qualify as non-ideal theory. I therefore endorse an understanding of non-ideal theory which is closer to Tessman’s non-ideal theory in the strong sense. As such, my specific approach to non-ideal theory could be understood as a critical non-ideal theory. My understanding of non-ideal theory also endorses Hänel and Müller’s view that non-ideal theory requires a certain methodology, not merely a turn of focus to certain topics, like oppression. On my view, a (critical) non-ideal methodology will approach a topic from a perspective which takes into consideration, and seeks to illuminate, the existence of unjust social power relations, for the purpose of determining their relevance for the topic and inquiry in question.Footnote 3
This can be done in different ways, for example, by choosing to engage, or not engage, with existing theories relating to the phenomenon under consideration. For instance, Jeff Engelhardt has recently defended a de-idealisation project, which seeks to engage with already existing (ideal) theories in order to replace them. ‘To de-idealize a theory is to replace its model of society as free of oppression with a model that represents systemic oppression and its effects’ (Reference Engelhardt2024, 2). Another way in which one can apply non-ideal theorising to existing theories is by investigating if, how, and to what extent an existing theory relies on ideal conditions (this is closer to what I intend to do in this article). One could see such investigations as preceding de-idealisation projects à la Engelhardt. It is not my purpose to de-idealise the problem of evil, in Engelhardt’s sense, in this article. Rather, I will examine a sample of how mainstream Anglo-analytic philosophy of religion has tended to approach the problem of evil, from a perspective informed by Mills’s approach to non-ideal theory. I do not seek to replace Swinburne’s theodicy with a non-ideal theodicy, nor do I seek to do applied or experimental philosophy.
Theodicy-construction
The problem of evil asks why a loving, omniscient, omnipotent God would allow the existence of evil and suffering. Theodicies are theories which seek to provide answers to this question, answers which attempt to show how God can be just, loving, and all-powerful despite the existence of evil.Footnote 4 Typically, theodicies do this by appealing to the existence of greater goods; the evils in the world are unfortunate, but necessary consequences of certain goods in the world which would not (could not) exist without the possibility of evil and suffering (Ooi Reference Ooi2022, 2022; Speak Reference Speak, McBrayer and Howard-Snyder2013, 205). Construction of theodicies centre on identifying such greater goods and showing how they are indeed greater (better or more valuable) than the evils that exist in the world (Trakakis Reference Trakakis2008, 7, 8). The theodicy constructed and defended by Richard Swinburne relies on three such greater goods: free will, character-building (or soul-making), and the notion of being of use.
Starting with the first, most (if not all) theodicies rely on the existence of free will, the idea being that God has given humans the gift of free will so that they may freely choose how to act. But with the freedom to act however one chooses comes the possibility that humans may act in selfish and harmful ways, thus giving rise to so-called moral evils – that is to say, evils caused by intentional human action. But despite this possibility, the existence of free will is typically understood as a good worth having, because its value is greater than the evils it can give rise to (Trakakis Reference Trakakis2008, 9–10).
Though Swinburne is not committed to a purely Irenaean soul-making theodicy à la John Hick, Swinburne still stresses ‘the value of suffering for the human beings who suffer, in enabling them to form their souls for good. By showing courage and sympathy in the face of their suffering and that of others, people can become naturally good people’ (Swinburne Reference Swinburne1998, 42). As such, Swinburne endorses the view that suffering can form our characters for the better, and that it is good that we have this possibility available to us.
Thirdly, Swinburne is notable for defending the view that ‘being of use’ is a great good, writing that
[t]here is also a further benefit for the sufferer, which lessens the evil of their suffering, if their suffering provides the opportunity for others to do morally good actions additional to helping the sufferer. ‘Being of use to others’ by what we suffer involuntarily is often a significant benefit for the person who is of use (Swinburne Reference Swinburne2023, 14).
Therefore, for Swinburne, suffering provides opportunities for other people to exercise their free will by choosing whether or not to help alleviate such suffering. But suffering is also good for the one who is suffering, because it provides opportunities for them and other people to form their souls and build character (Reference Swinburne1998, 233). As such, even a life filled with great suffering is never meaningless, because it provides opportunities for choice and being of use (Reference Swinburne1998, 236).
This summary of Swinburne’s theodicy will suffice for the following analysis, though I will be engaging with specific passages from Swinburne’s work when needed.
Slavery, the Racial Contract, and structural ignorance
In this section, I will consider Swinburne’s account of the benefits that American chattel-slavery provided both enslavers and the enslaved, and argue that Swinburne’s treatment of the beneficial opportunities that arose from God allowing the slave-trade to exist presupposes ideal epistemological conditions.
Swinburne’s treatment of the American slave-trade is rather short and lacking in details. However, we can discern an epistemic dimension: Swinburne notes that, in kidnapping and selling people from the African continent into slavery, the white enslavers did so ‘only because over centuries they had been taught to despise black people as uneducated savages, and had been taught that there was nothing wrong in enriching yourself at the expense of the suffering of uneducated savages’ (Reference Swinburne1998, 245). He writes that innumerable people, like preachers and politicians, contributed to a culture that dehumanised black Africans and thus created a public attitude toward these people that made their enslavement morally unproblematic to Europeans.
But God allowing this to occur made possible innumerable opportunities for very large numbers of people to contribute or not to contribute to the development of this culture; for slavers to choose to enslave or not; for plantation-owners to choose to buy slaves or not and to treat them well or ill; for ordinary white people and politicians to campaign for its abolition or not to bother, and to campaign for compensation for the victims or not to bother; and so on. There is also the great good for those who themselves suffered as slaves that their lives were not useless, their vulnerability to suffering made possible many free choices, and thereby many steps towards the formation of good or bad character (Swinburne Reference Swinburne1998, 245).
Though Swinburne is right that individuals in the white population did have the free will to oppose the institution of slavery, and that enslavement made this possible, he is severely underestimating the cognitive effects of culture. For in order to make the choice of opposing slavery, the dominant population first needs to recognise the immorality of slavery, since recognition of a practice as immoral is the first step toward fighting against it.
Charles Mills, in The Racial Contract, develops a framework through social contract theory for understanding societal cognitive dysfunctions. According to Mills, there exists a contract of exploitation in society. While social contractarians have usually only acknowledged and distinguished between a political and a moral contract, Mills argues that the notion of a social contract also tacitly presupposes an epistemological contract, which prescribes norms for cognition. The general purpose of this Racial Contract (the political, moral, and epistemological contract) is to privilege one group (whites) over another (non-whites), and to enable the exploitation of the underprivileged. It follows that all members of the privileged group are beneficiaries of the Racial Contract, even though some may express their dissatisfaction with it and perhaps even actively fight against it. For Mills, this contract of exploitation – the contract of and for the dominant, privileged social group – requires its own moral and empirical epistemology, which determines through norms what is to count as moral and factual knowledge of the world. It is this epistemology that provides us with a moral and intellectual compass, that is to say, the ability to see the world through a standard which becomes ‘objective’. This gives rise to an agreement or contract about what is to count as the correct, objective interpretation of the world. But when it comes to such social contracts – whose purpose is the privileged and dominant social position of some, and the underprivileged, non-dominant social position of others – that which is officially sanctioned as ‘reality’ is divergent from actual reality (Mills Reference Mills1997, 9–18). As such, the Racial Contract requires an agreement to misinterpret the world, which means that people must learn to see the world erroneously, but
with the assurance that this set of mistaken perceptions will be validated by white epistemic authority, whether religious or secular. Thus in effect, on matters related to race, the Racial Contract prescribes for its signatories an inverted epistemology, an epistemology of ignorance, a particular pattern of localized and global cognitive dysfunctions (which are psychologically and socially functional), producing the ironic outcome that whites will in general be unable to understand the world they themselves have made (Mills Reference Mills1997, 18. Emphasis in original).
What Mills is arguing here, is that members of society – through this epistemology of ignorance – are living under a cognitive model which precludes them from accurately understanding social realities. As Mills clarifies, the ignorance stemming from the Racial Contract need not be based on bad faith or bigotry, but can operate in a socio-structural fashion. Moreover, the structural ignorance that the Racial Contract gives rise to ought to be understood as a cognitive tendency, not as an absolute, since some people of the dominant social group can and will overcome it (Mills Reference Mills, Sullivan and Tuana2007, 21–23). For Mills, the Racial Contract also gives rise to moral ignorance, by which he means ‘not merely ignorance of facts with moral implications, but moral non-knowings, incorrect judgements about the rights and wrongs of moral situations themselves’ (Mills Reference Mills, Sullivan and Tuana2007, 22. Emphasis in original). In short then, this social contract creates a structural, socio-epistemic and moral blindness.
For Mills, such structural cognitive dysfunctions would certainly play a central role in the context of slavery. Thus, considering the work of Mills, moral considerations have a socio-epistemological component, in that in cases of social oppression and domination, ideologies upholding the oppression and domination will seep into the culture of society, erecting epistemological barriers which serve to protect the dominant political order and hierarchy. However, Swinburne has not accounted for any such politico-cultural influences, nor the existence of structural ignorance which would directly affect the free choices of individuals of the dominant population. Rather, he seems to presuppose an idealised cognitive sphere, and ideal (supremely rational and moral) cognitive abilities. Or, adopting Lisa Tessman’s words, Swinburne, like ideal theorists, seems to rely on ‘idealized moral subjects acting under background conditions that are presumed not to have been impacted by a history of oppression or by the ongoing effects of oppression’ (Tessman Reference Tessman2009, xviii). But if Mills is right, then any analysis of ‘oppressive social systems requires oppressors characterised by what Tessman calls “the ordinary vices of domination,” such as cruelty, indifference, contempt, and arrogance, which make for “degraded,” “twisted” forms of moral character’ (Kidd Reference Kidd, Kidd, Battaly and Cassam2021, 69; Tessman Reference Tessman2005, 54).Footnote 5
Further, if we assume Mills’s framework, and the (southern) white US population was indeed living under a Racial Contract which created a culture which prescribed a moral ignorance with regards to slavery, which in turn prevented the white population from understanding the immorality of slavery – and recognition of the immorality of slavery is essential for opposing it (on moral grounds) – then Swinburne’s explanation for the reason why God allows slavery would be incomplete at best. By not accounting for the epistemic factors and consequences of the politico-cultural environment in which agents are embedded, Swinburne gives the impression that the failure of agents to fight against slavery is a failure of the will. In other words, Swinburne’s theodicy encourages focus on factors internal to the individual agent, while ignoring external factors and non-ideal social conditions and social relations that affect the agent, and influences their moral compass.
Considering that the Bible and Christianity were used to justify American slavery on moral grounds, how could a person raised in such a culture notice the error of their community? Living in such a community – in which abolitionist voices may be a rare occurrence – would be akin to living in an echo chamber, in which dissenting voices are excluded from serious consideration. Without an explanation of how Swinburne expects the individuals in his theodical world to break free, so to speak, of the echo chamber in which they are situated, Swinburne’s theodicy seems to presuppose a supremely rational homo economicus, able to easily see through any cognitive bias, collective prejudice, and ideological influence. Since we are decidedly not perfectly rational agents, this idealised presupposition demands too much of us, even if our environments were conducive to make optimal decisions. As David Beaver and Jason Stanley point out, we are limited creatures; we do not have perfect memory, we do not have infinite amount of time, and we are prone to make errors when reasoning, and not just when we are tired (Beaver and Stanley Reference Beaver and Stanley2023, 355).
Swinburne’s epistemology
Here, one might wonder whether such societal cognitive dysfunctions that Mills outlines in his epistemology of ignorance are in tension with Swinburne’s Principle of Honesty. As he states, Swinburne considers it a moral truth that
‘God has an obligation not to make a world in which agents are systematically deceived on important matters without their having the possibility of discovering their deception’ (Swinburne Reference Swinburne1998, 139. My emphasis).
The word ‘possibility’ is particularly important in this quote. On the one hand, given that the white population had the possibility to break free of their echo chamber, would suggest that God has not forsaken Their obligation according to the Principle of Honesty. But on the other hand, is mere possibility sufficient for honesty? If I allow an entire society to be deceived by propaganda, even though they have the possibility to discover their deception, am I being honest? Being omniscient, God would certainly know that propaganda has the power to influence and distort the cognition of millions of people for generations; God would also be aware of the harm the propagandistic deception causes. This raises the question of why God would create humans so susceptible to manipulation and bias which might impede their moral awareness, and by extension, impede their character development. Of course, one might insist that humans must have the capacity to deceive each other on massive scales in order to have free will. But, first, such appeals would seem to suggest that the freedom to do harm is more important than the freedom from harm. And, second, even if we assume that societal epistemic ignorance originally came into being due to deliberate deception (as opposed to unconscious self-deception), after a generation the echo chamber would be firmly established, and there may be no human left who would seek to deliberately deceive the population in order to keep the echo chamber in place. For example, if men established a patriarchy by deliberately deceiving and convincing people that women are inferior (as opposed to erroneously concluding that women are inferior due to male bias), after generations, it is not obvious that there would still be men who circulated a patriarchal narrative they knew was a lie. Rather, the patriarchal social contract or echo chamber might circulate the narrative of female inferiority on its own, without agents exercising their free will to deceive others. In such non-ideal cases (slavery included) it would seem like massive amounts of people, entire societies, are being systematically deceived, without any human agent deliberately upholding the deception. Here, I maintain that Swinburne’s epistemology – due to its silence on oppressive social relations and their epistemological effects – leaves unanswered questions in relation to his theodicy and especially in relation to non-ideal epistemological conditions like those of Mills’s epistemology of ignorance.
Swinburne’s Principle of Credulity likewise seems to run into problems when tested against non-ideal conditions. While Swinburne’s Principle of Credulity is prescriptive, the social ontology of Swinburne’s theodicy is meant to be descriptive. However, with any understanding of the psychology and cognitive landscape absent from Swinburne’s account of slavery and the opportunities it provided for emancipatory action, one can only seek such epistemological understanding by turning to Swinburne’s epistemology. As such, even if Swinburne himself would not want to apply his Principle to such a non-ideal environment – as his principle was developed to accommodate religious beliefs and religious experiences rather than oppression – if we find problems in applying his epistemology to his theodicy, then we would seem to have found inconsistencies in his philosophy.
Swinburne defines the Principle of Credulity in the following way: ‘other things being equal, it is probable and so rational to believe that things are as they seem to be … By “seem” (or “appear”) I mean “seem epistemically”; the way things seem epistemically is the way we are initially inclined to believe that they are’ (Reference Swinburne1998, 20–21). The application of this principle can be considered in, at least, two ways: personally, as a principle for assessing one’s own rationality, or intellectually, as a principle for assessing other people’s rationality. Considering the former in the non-ideal epistemic context of oppression as theorised by Mills in The Racial Contract, this principle would encourage people not to distrust the pro-slavery rhetoric of the media, politicians, and other outlets which may distort causes and conditions of oppression in their attempt to defend the institution of slavery. Swinburne continues:
If it seems to me that I am seeing a table or hearing my friend’s voice, I ought to believe this until evidence appears that I have been deceived. If you say the contrary – never trust appearances until it is proved that they are reliable – you will never have any beliefs at all (Reference Swinburne1998, 21).
While I’m sure Swinburne would agree that appearances can be more deceiving in some areas or dimensions of (social) life than others, and that we ought to be more critically minded in those areas, we can still see the problem with trying to universalise this principle. Allowing for more critical reflection in certain areas of social life, without accounting for these areas specifically, one ends up assigning them to the margins, thus implying (a) that they are not part of ‘normal’ circumstances, and (b) that they are less important to theorise; or, channelling Mills, by representing non-ideal cases as simple aberrations, one conveys the message that these cases are not worth theorising in their own right (Reference Mills2005, 168). Moreover, Swinburne seems to assume that people are blank slates – not affected by the culture in which they were raised, a culture which would groom or socialise them in favour of the slavocracy in which they were born – and/or not emotionally invested in the outcome, and that they are so rational as to be able to easily escape such cultural influences as soon as they are presented with counterevidence.
But considering the application of this principle intellectually – that is to say, as a principle for assessing other people’s rationality – also brings difficulties for his theodicy. If we are to assess the mentality of people living in a slavocracy who endorse slavery, and finding that they are, indeed, epistemically and morally affected by the ideology permeating society at the time, then a host of questions seem to arise which demand our attention. Are these people rational in endorsing slavery or not? After all, it may seem to them that the enslaved people are savages, and that slavery is for their own good – that the enslavers are doing the enslaved a kindness (cf. the notion of the white man’s burden). According to Swinburne’s Principle, they would seem to be rational in this belief. Moreover, how can one mature – through fighting against slavery – if one believes that endorsing slavery is morally good, while simultaneously believing that one does not have good reasons for believing that one is wrong? If epistemic corruption can affect entire societies, leading people to believe that endorsing slavery is morally good, then, as the theodicy suggests, is being of use as a slave still (to be counted as) a greater good? Furthermore, considering Swinburne’s prescriptive Principle against oppressive social conditions raises questions regarding the epistemic responsibility of agents in his theodical world. If José Medina is right that ‘social injustices breed both ignorance and irresponsibility’, in that ‘under conditions of oppression, subjects have a harder time maintaining their status as responsible ethical and political agents (at least in part) because they have a harder time being epistemically responsible’ (Medina Reference Medina2013, 131), then, applied to his theodicy, Swinburne’s (prescriptive) epistemology leaves unanswered questions regarding ethico-epistemic responsibility under oppressive conditions and cases of structural ignorance. What consequences do these and other questions have for the theodical model being theorised? If slavery is a justified evil because it provides people the chance to freely choose to fight against it, and because being of use as a slave provides the enslaved the opportunity to give other people the opportunity to fight against slavery, but societal epistemic corruption can epistemically hinder people from recognising slavery as an evil to be fought against, and recognition is a prerequisite for moral choices, then it would seem highly relevant to account for such non-ideal epistemic factors in one’s theorising. If I am right that non-ideal theorising reveals epistemic obstacles in the process of decision-making, then non-ideal theorising would seem to limit the plausibility of Swinburne’s theories.
But there are other signs that Swinburne’s epistemology is based on ideal conditions. He writes: ‘I suggest that it is a principle of rationality that (in the absence of special considerations), if it seems (epistemically) to a subject that x is present (and has some characteristic), then probably x is present (and has that characteristic); what one seems to perceive is probably so’ (Swinburne Reference Swinburne2004, 303. My emphasis). ‘It seems to me … intuitively right in most ordinary cases … to take the ways things seem to be as the way they are’ (Swinburne Reference Swinburne2004, 303–304. My emphasis). Here, Swinburne’s choice of words – ‘in the absence of special considerations’ and ‘in most ordinary cases’ – are illuminating, because they suggest that Swinburne believes his Principle works under normal circumstances, and that any ‘special considerations’ are aberrations. As such, he seems to suggest that under normal circumstances, things are probably as they seem to be. But if Mills’s epistemology of ignorance is correct, then this would seem to suggest that Swinburne is presupposing that normal circumstances are those which are not epistemically distorted by the effects of oppression and domination. If this reading is correct, then this would suggest that Swinburne is working under the assumption that either (i) oppression and domination do not have distorting epistemological effects, or (ii) oppression and domination constitute (historical) deviations from the norm, and so, are not part of ordinary circumstances. But ‘[a]s Judith Shklar points out, the history of philosophy leads us to think falsely of justice as the norm and injustice as the aberration’ (Fricker Reference Fricker2007, 39). Mills agrees, writing that ‘[s]exism and racism, patriarchy and white supremacy, have not been the exception but the norm’ (Mills Reference Mills, Sullivan and Tuana2007, 17. Emphasis in original. This point is repeated in Reference Mills2013, 38). If these statements are correct, as I believe they are – after all, patriarchal views of women as inferior, defective, irrational, and less intelligent than men have influenced the culture, politics, social structure, religions, and worldviews of countless societies throughout history – would it not be overly naïve to consider oppression, domination, and exploitation as anomalies in human civilisation and history (today included), to the point that their epistemological and social ontological features and effects do not merit consideration in one’s theorising?
Swinburne himself considers the limitations of the Principle of Credulity (Swinburne Reference Swinburne2004, 310ff), thus suggesting that it is not meant to be universal, as in applying in all circumstances – though he does maintain that it applies in ordinary, or under ‘normal’, circumstances. However, the special circumstances he considers that would limit his Principle do not account for oppressive social conditions, which may influence the cognition of epistemic agents, as theorised by Mills. And so, if Swinburne believes that the Principle of Credulity applies in ordinary (or under ‘normal’) circumstances, but oppressive social conditions are not aberrations in world history, then it would not simply be a matter of adding such social conditions to Swinburne’s list of exceptions to his Principle. That is, the Principle of Credulity cannot be saved by a mere addendum if Mills, Fricker, and Shklar are right that oppressive social conditions have not been the exception in history, but the norm (Mills Reference Mills, Sullivan and Tuana2007, 17).
Omissions in theodicy and their implications
While we have already, to some extent, seen that Swinburne seems to presuppose ideal conditions in decision-making, there is another aspect of decision-making that we have yet to explore – inter-agent communication. Theodicies that emphasise that suffering provides opportunities for people to help people who are suffering (like those that appeal to character-formation or soul-making) tend to presuppose ideal interactions between agents. Such theodicies claim that God would be justified in allowing a person A to suffer for the purpose of person B having the opportunity to freely choose to help A. In such an example, the notions of character-building (or soul-making) and being of use, central to Swinburne’s theodicy, together stipulate that suffering is good for the person who is suffering and for any people who may be in a position to help. For the latter, one person’s suffering provides an opportunity for them (B) to exercise their free will by making a decision to help the one who is suffering, and so develop their (B’s) (moral) character. For the former, being of use by providing this opportunity for another person is good for them (A) – that is to say, the one who is suffering. However, this example implies that B is fully aware that A is suffering. But in our actual, non-ideal society, for many B’s out there, B may not believe A’s testimony that A is suffering.
Here, we might consider Miranda Fricker’s framework of epistemic injustice, specifically her notion of testimonial injustice. As Fricker explains, testimonial injustice ‘occurs when prejudice causes a hearer to give a deflated level of credibility to a speaker’s word’ (Fricker Reference Fricker2007, 1). Since (gendered and racial) bias and prejudice may cause us to give certain people lower levels of credibility than they are due, Fricker is challenging the ideal nature of communicative exchanges presupposed in standard examples used the epistemology of testimony. Fricker not only points out that our epistemic practices are not always ideal in nature, but that the (non-ideal) environment in which we are situated, and people’s social identity, can and do influence our epistemic practices.
Returning to our example, if A is, say, a woman, and B is a man, then the phenomenon of testimonial injustice would illustrate a case where B may be in a position to help or alleviate A’s suffering, but due to a structural identity prejudice, B does not believe A’s testimony, and so does not choose to help. This example reveals another epistemological dimension of decision-making, one theodicists do not seem to have considered. If it is true, as Fricker’s framework suggest, that our non-ideal environment can hinder ideal epistemic practice, and B’s helping A is contingent on B believing A’s testimony, then it would seem highly relevant for the plausibility and function of Swinburne’s theory to account for such non-ideal circumstances and their epistemological effects.
Moreover, moving from individuals to groups, testimonies from Black Americans regarding the limitations and obstacles their oppressed situatedness provides them have historically fallen on mostly deaf ears, as white Americans have tended to deny the existence of structural limitations on Black Americans opportunities. The authors of one study found that ‘[m]any whites appear to believe that great progress toward removing barriers to blacks’ opportunity has occurred in recent years and that blacks now have opportunity equal to whites. Many whites also believe that blacks now have better than average opportunity due to unfair preferential treatment accorded them – i.e., there is a belief in widespread “reverse discrimination’’' (Kluegel and Smith Reference Kluegel and Smith1982, 518). Another study, from the Pew Research Center, found that
blacks are more likely than whites to say that black people are treated less fairly in the workplace (a difference of 42 percentage points), when applying for a loan or mortgage (41 points), in dealing with the police (34 points), in the courts (32 points), in stores or restaurants (28 points), and when voting in elections (23 points). By a margin of at least 20 percentage points, blacks are also more likely than whites to say racial discrimination (70% vs. 36%), lower quality schools (75% vs. 53%) and lack of jobs (66% vs. 45%) are major reasons that blacks may have a harder time getting ahead than whites (Pew Research Center Reference Research Center2016, 5–6).
Considering the enormous amount of research showing that Black Americans do suffer widespread discrimination and disadvantages in society,Footnote 6 if the findings above are correct, then there seems to be an epistemological barrier between reports from Black Americans of widespread discrimination, and white Americans believing such reports. If taking action to help people who are being unjustly treated is contingent on people first believing that other people are being mistreated, then our non-ideal realities would seem to illustrate conditions in which people who may be in a position to help, do not believe, and so, do not help either. This puts a certain strain on theodicies, like Swinburne’s, which do not account for non-ideal conditions, in that they do not seem to be universally applicable. As such, the theodical tendency to omit agential features removes from consideration how those features are, or can be, relevant for theodical theorising.
Tendencies in theodicy
From this discussion, we can discern some tendencies in Swinburne which, in turn, reflects how theodicists typically operate. As certain critics have observed, theodicies are typically individualist in nature (Tilley Reference Tilley1991, 236–238, 242, 244; Pinnock Reference Pinnock2002, 140), focusing on individual people from which nearly all personal and contextual details are abstracted. This may be a necessary feature of theodicies, as their aim – provide explanations for all forms of sufferingFootnote 7 – is universal in nature (Hick Reference Hick and Davis2001, 38; Swinton Reference Swinton2018, 21). Thus, the abstract (featureless) individuals may be meant to represent anyone and therefore everyone. But in abstracting away from personal features and contextual details, this theoretical representation also abstracts away people’s specific group identity and the hierarchical social relations through which people exist and interact with each other (see Kidd Reference Kidd, Kidd, Medina and Pohlhaus2017, 391). Since unjust and oppressive social hierarchies are group-based, the focus on individuals fails to capture such unjust hierarchies, and fails to adequately represent the social contexts in which individuals act and react to other people’s suffering.
Though theodicies often, if not always, rely on the notion of free will – and Swinburne himself believes it would be difficult to construct a satisfactory theodicy which did not rely on free will (Reference Swinburne1998, 241) – theodicists do not typically account (adequately) for psychological factors which influence people’s views, acts and decisions, like structural prejudice and implicit bias, through which stigmatised social groups suffer injustices and disadvantages. When bias is mentioned by theodicists, it is typically individual, not structural; like a character vice an individual can overcome through personal effort, rather than a structural psychological mechanism affecting the cognition of agents of entire societies. Instead, theodicists typically speak of desires and temptations (Davis Reference Davis and Davis2001; Hick Reference Hick1977, Reference Hick and Davis2001; Swinburne Reference Swinburne1998), which do not capture the socio-psychological mechanisms at work in structural prejudice and implicit bias.Footnote 8 As such, theodicists overlook how the non-ideal structure of a society can produce false beliefs that harm already vulnerable, disadvantaged, stigmatised populations. In other words, they overlook how social ontology affects epistemology in non-ideal settings. Theodicists therefore tend to severely underestimate the nature of bias, as they seem to assume that bias is an anomaly, a deviation from how social cognition ‘normally’ operates. Here we may remember that according to Mills, ideal theory is characterised by an idealised cognitive sphere, where cognitive obstacles are limited to individual biases, and little or no attention paid to the influences of ideologies.
In her account of tendencies in, what she calls, ‘ideal social ontology’, Åsa Burman writes that in social ontology there has been a tendency to use depoliticised examples, in the sense that the examples used are either not obviously political, or, when political examples are used, their political dimension remains underdeveloped (Burman Reference Burman2023, 80). I would argue that we can discern a similar tendency to use depoliticised examples by philosophers of religion when engaging in discussions on the problem of evil. We often see seemingly non-political examples, like parents, dentists, fawns, and natural disasters. One might find this tendency odd in a topic like the problem of evil, which ostensibly concerns itself with all the suffering that exists in the world. But even when examples with a political dimension are used – such as the European twentieth-century Holocaust, slavery, or September 11 – the focus is nigh exclusively on the harm they caused and the beneficial opportunities they provided, not on the underlying socio-political conditions and contextual details of these events. Swinburne’s treatment of slavery, as seen above, where the details of social life in a slavocracy are all but absent, demonstrates this tendency. As I have argued, an implication of this omission of contextual details is that Swinburne overlooks how such details may be relevant for his theodicy. Here, it is worth remembering that non-ideal theorists have remarked that a consequence of not fully, or adequately, developing the political dimensions of political examples is that these dimensions are hidden from view, and rendered invisible, which makes it all the more difficult to understand how these dimensions may be significant for one’s theorising.
The focus on oppression is particularly significant for the problem of evil, since theodicists, as some critics have observed (see Pinnock Reference Pinnock2002; Tilley Reference Tilley1991), have tended not to investigate or sufficiently account for oppression – its manifestations, implications, legacy, and its connection to other institutions and worldviews – though racism and sexism may merit mentions as examples of evil in the theodical literature. If this observation is correct, and theodicists have typically not attended to the details of oppression, then it seems increasingly likely that they have – like Swinburne – also overlooked how these details may be relevant, or have significant implications, for their theorising.
Since theodicies aim for universal answers, they have, by necessity, constructed broad theories capable of capturing suffering in diverse forms. The result has typically been one-size-fits-all theories. Moreover, theodicists have tended not to develop theories – specifically the environment in which humans live – with research from other (empirical) disciplines. Rather, theodical discourse practice has tended to prioritise abstract reflection (see, e.g., Geddes Reference Geddes2003; Surin Reference Surin1986; Trakakis Reference Trakakis2008) thus ending up being methodologically a prioristic. As a consequence, the social ontological models in theodicies have tended to leave out aspects of the social world which are significant for theorising beliefs and ideas that influence action.
Conclusion
In this article, I have found that Swinburne displays tendencies typical of ideal theorising. I have argued that in abstracting away contextual details and agential features in his idealised model, Swinburne overlooks how the details he excludes from consideration are significant for his theodicy. Moreover, given that the tendencies displayed by Swinburne are not limited to him, but can be found across the theodical literature, I propose – from this limited discussion of one paradigmatic sample of theodicy-construction – the hypothesis that theodicists are engaging in ideal theory.
Furthermore, I have aimed to show that accounting for non-ideal conditions in this traditional area of philosophy of religion reveals that traditional approaches to the problem of evil have left much un(der)theorised, especially in relation to the social world and human cognition and interaction under conditions of oppression. I find that non-ideal theory fundamentally changes the problem of evil. Adopting a non-ideal framework reveals, at a minimum, that the problem of evil is more complex than traditionally assumed by philosophers of religion. If I am right, then it seems to me that any future construction or defence of theodicy would require a substantial reconsideration of the entire theodical enterprise.
Competing interests
The author declares none.