Within analytic philosophy of religion, a lot of attention has been devoted to the nature of faith and the reasons why some people fail to have it. However, as proven by the existence of religious seekers, rebels, fence-sitters, and speculators of all stripes, full-blown faith and lack of faith are not the only general attitudes towards God one may have. In this article, I will discuss some consequences of this fact from the normative viewpoint.
Though it is intuitive to see the question ‘Ought one to have faith?’ as the practical counterpart of the theoretical problem of the existence of God, the parallel is incomplete. While God cannot exist only in part, if faith is gradable like belief, it is possible to have faith which is less than full-blown. The label ‘partial faith’ seems to be useful as an umbrella term for many different existential attitudes towards God which combine a positive or neutral element with a negative one. I will argue that, on some relatively non-controversial assumptions about God and faith, it is likely that partial faith (as opposed to full-blown faith or lack of faith) is the all-things-considered best option – in other words, that some people at some times ought to have faith only partially.
By outlining the considerations which speak in favour of partial faith, I hope to give expression to a view silently influential in the social and cultural sphere but rarely explicitly defended by philosophers. As I will try to show, the intuitions which underlie this view can also shed some light on the nature of faith, its varieties and connection to personal experience – as well as on the problems which ensue if this connection is severed.
Faith and reasons: preliminary assumptions
In this article, I am concerned with theistic faith (and not faith in general – e.g. faith we may have in a friend or a cause) of the kind normally associated with religion, but I am not focusing on theistic faith as religious. This means that I am leaving open the questions of what counts as religion, whether faith can occur entirely outside of religious traditions, and how conforming or not conforming to orthodoxy as defined in one's religious community impacts one's faith. I am also bracketing all theological issues relevant to my topic – for instance, the significance of revelation, the consequences of the Fall, and the role of merit versus grace.
For the purpose of the discussion, I propose the following definition of faith: faith consists of (a) a positive cognitive attitude towards the existence of God – for example, belief that God exists; (b) a positive conative attitude towards the existence of God – for example, desire for God to exist (cf. Howard-Snyder and McKaughan (Reference Howard-Snyder and McKaughan2021), 2–5). Notably, this allows for the possibility of faith constituted by one attitude (like hope) with both a cognitive and a conative dimension (cf. McKaughan (Reference McKaughan2013), 112–113). This, depending on how we define the attitudes involved, may include at least some varieties of non-doxastic faith, while other varieties – significantly for our topic – will fall under the description of partial faith.Footnote 1 Similarly, we can talk about a positive cognitive and conative attitude towards God's non-existence, which taken together constitute lack of faith. Since (some if not all) such attitudes are gradable, so is faith and lack of faith.
On such a definition, someone who has faith – barring the cases of akrasia – normally tends to act as if God existed and as if God's existence is a good thing, but such a tendency or such actions are not a constituent of faith. Neither is religious commitment, even though, insofar as one embraces a particular religious tradition which makes claims about God, one's acting on faith includes conforming to that tradition – which means that partial faith can, and often does, result in partial religious commitment.
Since God's existence and nature have impact on the nature and contents of the entire universe, having faith leads to having a certain overall perspective, within which everything is seen through the lens of God's positive presence. In other words, that God exists and that it is a good thing are, as John Bishop has put it, ‘framework principles’ (Bishop (Reference Bishop2007), 80) of the general perspective of the person of faith. I assume – with Bishop, and contrary to the authors like D. Z. Phillips (cf. e.g. Phillips (Reference Phillips1988), 38–53, 115–122) – that such framework principles can be understood realistically. I also do not rule out the possibility that one's perspective may have other framework principles besides these two. Depending, among other things, on the exact attitudes constitutive of one's faith and on one's concept of God, it may also include other attitudes towards God (and not just towards God's existence), such as admiration or trust. (If God does not exist, such attitudes are also possible, though, in a sense, misplaced (cf. e.g. Forbes (Reference Forbes2006), 36–51).)
It can be argued that belief and desire are ‘paradigm cases’ of the cognitive and conative attitudes which can constitute faith, since all such attitudes are ‘belief-like’ and ‘desire-like’ (which does not mean identical with belief and desire) in their capacity to shape one's worldview and drive one's motivation (cf. Howard-Snyder (Reference Howard-Snyder2018), 121). Nonetheless, the fact that faith can consist of the attitudes other than desire and belief matters a lot, being crucial for explaining why different people at different times can have very different kinds of faith, with different accompanying motivations and expectations – and different amount of voluntariness. (Though the question of which attitudes can be constituents of faith, or what conditions an attitude has to fulfil to be eligible, is important and interesting, I have no space or need to discuss it here.)
The factors leading to the discrepancies between particular perspectives resulting from faith are not limited to the nature of the attitudes which constitute faith. Due to individual differences in experience, personality, and cognitive make-up – but also due to different concepts of God – such perspectives can be extremely different for different people (and for the same person at different times) even within one religious tradition. To illustrate: the general outlook of Augustine of Hippo at the time when he was writing his Confessions must have been very far from the one Thérèse of Lisieux had when entering the monastery at the age of fifteen, even though, presumably, they both had full-blown faith.
The differences between various kinds of faith are something of more than technical importance and can go so far as to put various kinds of faith in tension with each other. Different varieties of faith have different advantages, which can be mutually exclusive (more on this later), and none seems to be always available to everyone. An important factor in this is that different people have different reasons to have faith, and have faith for different reasons – something which seems to be at the root of the fact, sometimes underestimated by philosophers, that faith can mean very different things for different people at different times (which might be one of the reasons for a broad disagreement about what it means for faith to be full-blown).
A common implicit assumption is that the core normative reason to have faith is that there is a God such that being in the best possible relationship to God is the greatest good (and thus ought to be desired above all else) and the quality of one's relationship to God is proportionate to the degree to which one has faith. Correspondingly, there is an implicit assumption that the core normative reason not to have faith is that there is no such God. Thus, assuming that there is or that there is not such a God prejudges the matter: if God exists, full-blown faith is the right attitude to have, and if there is no God, the right attitude is not to have faith. If the project of defending partial faith is to have any chance of getting off the ground, no such assumptions can be made.
Is it problematic to refrain from any assumptions about God's existence in a normative discussion of faith? It can be argued that it does not make sense to talk about faith in isolation from a relationship to God; some authors even consider such a relationship to be a constituent of faith. This, however, raises some problems. One difficulty is that if faith involves or implies being in the right relationship to God, then it seems that it is not possible to have faith if God does not exist – a somewhat unfortunate conclusion. Another problem is that defining faith as necessarily leading to or involving a positive relationship to God has an implication that faith automatically causes or entails such a relationship, which may lead to difficulties in the concepts of God ascribing to God the kind of agency normally possessed by persons (cf. Sider (Reference Sider2002), 67).
As a way of avoiding these issues, I propose to assume that the right (i.e. best possible) relationship to God is not a constituent or effect of faith but its aim, in the sense in which truth may be considered the aim of belief. If the aim of faith is a relationship to God, then, if God does not exist, faith does not fulfil its aim no matter how high the degree of the attitudes constitutive of it. At the same time, if God exists, faith contributes to a relationship to God – that is, if God exists, then, all other things being equal, one's relationship to God is better if one has faith than if one does not have it. My point, as will soon come to light, is that, for any two cases, all other things tend not to be equal.
Importantly, one does not have to assume God's existence or non-existence to talk about normative reasons to have or not have faith, including but not limited to the reasons to be a theist or an atheist. The set of reasons to have some conative attitude towards the existence of God includes facts such as ‘Some people report that they acquired help from God after praying for it’ (a reason to have a positive conative attitude) or ‘Some people exert violence on others to fulfil what they consider God's will’ (a reason not to have it). In addition to those, the set of reasons to have some cognitive attitude towards God's existence includes such facts as ‘Some arguments for God's existence are sound’ (a reason to have a positive cognitive attitude) or ‘The existence of evil is hard to reconcile with the existence of God’ (a reason not to have it). (As we can see, a reason not to have a positive attitude amounts to a reason to have a negative one.) The important thing is that we seem to live in a universe in which all those and many similar states of affairs are true and constitute reasons for opposite attitudes – in other words, a universe which is, as John Hick puts it, ‘religiously ambiguous’ (Hick (Reference Hick1989), 12).
Robert McKim, who focuses on evidence as the reason for belief in God's existence and other religious beliefs, notes that our universe exhibits not simple but extremely rich religious ambiguity (McKim (Reference McKim2012), 131–151): the reasons to have faith and not to have it, rather than balancing each other so as to prompt universal suspension of judgement (as they would if the ambiguity was simple), stand in different relations to different agents and cannot be compared on agent-neutral grounds. This is because at least some reasons to have faith and not to have it (e.g. events which may or may not be supernatural) can only be interpreted as reasons within a particular perspective, and different agents have access to, and can move to, different perspectives. Thus, the reasons to have faith and not to have it are numerous but not all weighings of them are available to every agent. I will assume that, as an upshot, it is likely that not all kinds of faith are available to everyone – but also that not everyone can have full-blown faith and not everyone can entirely lack faith.Footnote 2
If McKim is right – as I assume he is – different weighings of reasons to have faith and not to have it are optimal for different people at different times. This means that different people at different times ought to weigh the reasons to have faith and not to have it differently in the general normative sense of ought: all-things-considered ought rather than moral, prudential, or epistemic ought (with ‘all-things-considered ought’ may include moral, prudential, epistemic, and possibly other considerations). The way individual perspectives work, some reasons for and against faith are experiential – related to the personal experience (perception, memories, emotions, etc.) of a particular person in such a way that no one else can have the same or larger amount of access to them. In addition to that, not everyone (and perhaps no one) seems to have equal amount of access to the reasons which are non-experiential, that is, equally accessible to two or more people, even if – as I will assume for the sake of simplicity – the disparity in this respect is not significant. I will also assume, in line with the prevalent view on the reasons to be agnostic, that the set of non-experiential reasons for and against faith, taken in insolation, displays religious ambiguity of the simple kind.
Though I do not specify what exactly constitutes the reasons for cognitive and conative attitudes, I assume that the reasons for conative attitudes intertwine with the reasons for cognitive ones, both in the sense that conative attitudes correspond to cognitive attitudes towards states of affairs involving values (which means that the same thing can be a reason for a cognitive and a conative attitude) and in the sense that cognitive attitudes can be reasons for conative ones and the other way round. All of that is not without an impact on the dilemmas relevant to partial faith. (When I speak of ‘dilemma’, I do not imply that having faith or not having it is necessarily a result of a conscious decision, only that the stakes are high and it is far from obvious which option is optimal.)
When speaking in terms of ‘likelihood’ and ‘seeming’ throughout the article, I do not mean probability in any measurable sense. My topic touches upon two complex and mysterious areas – that of the transcendent reality and that of other minds. No one can know what exact reasons other people have to have faith and not to have it (though we can make sensible guesses), nor how exactly God responds to the particular ways in which such reasons are weighed (though we can speculate and reflect on it). To an extent, a topic like this necessitates relying on intuitive analogies with better-known phenomena. Given all that, I do not think a water-proof argument for the claim that in some cases partial faith is optimal can be given. Instead, much less ambitiously, I want to show that there are serious reasons to believe such cases exist and throw some light on what they are.
Doubts, worries, struggles and enquiries: partial faith and its varieties
On the above account of faith, the main difficulty with defining partial faith is establishing its limiting cases. Partial faith is not the same thing as non-maximal faith: someone who has maximal faith has the cognitive and conative attitude constitutive of faith to the maximal (absolute) degree – the heights to which faith does not need to rise in order to be full-blown. At the same time, it is far from obvious at which point on the faith spectrum faith stops being full-blown. In addition to that, it is not clear what constitutes minimal faith – faith which is as small as it can be without becoming lack of faith.
Though any precise distinctions are bound to be arbitrary, one can do worse than make use of the natural points of division between positive and negative attitudes concerning God's existence. A natural dividing point for the cognitive attitudes is agnosticism – suspending whatever cognitive attitude one might have about the existence of God.Footnote 3 Its equivalent and thus a dividing point for the conative attitudes is basic openness to God – a state neutral between the positive and negative conative attitude towards the existence of God, when one merely entertains the possibility that God's existence is a good thing.
I will assume that the cognitive constituent of one's faith is full-blown when it goes above agnosticism – that is, when one neither has a positive cognitive attitude towards God's non-existence nor suspends whatever cognitive attitude one might have about whether God exists – and the conative constituent of one's faith is full-blown when it exceeds basic openness to God, that is, when one neither has a positive conative attitude towards God's non-existence nor limits one's conative attitude towards the existence of God to basic openness. By the same token, one has a full-blown cognitive constituent of lack of faith when one neither suspends whatever cognitive attitude one might have about whether God exists nor has a positive cognitive attitude towards the existence of God, and one has a full-blown conative constituent of lack of faith when one neither limits one's conative attitude towards God's existence to basic openness nor has a positive conative attitude towards the existence of God. One has partial faith when one's cognitive or conative attitude towards God's existence is less than full-blown (which includes the possibility that they are both so) but one does not combine a full-blown positive cognitive attitude towards God's non-existence with a full-blown positive conative attitude towards it. (Of course, defined in this way, partial faith is also partial lack of faith.)
The division I propose is meant to grasp the intuitive difference between the people with an attitude normally associated with religion (full-blown faith) and two kinds of people which do not fit this description. One kind (partial faith) is constituted by those who have some receptivity to God, in the sense that someone who is receptive ‘finds the prospect of being the subject of some state or event attractive in some respect but is neither firmly convinced that he should be nor is irresistibly drawn to being the subject of that state and event’ (Cuneo (Reference Cuneo2017), 673), where the source of the attraction can be cognitive or conative, and the other kind (lack of faith) is constituted by those who do not have any such receptivity.Footnote 4
On such division, faith is full-blown even if it is shallow or weak – a natural diagnosis when the degree of one or both constituents of faith is low but still above the point of neutrality. In contrast, partial faith encompasses the cases in which the engagement with God is either neutral or conflicted. The paradigm example of the first kind of case (neutral engagement) is faith neutrality – agnosticism combined with basic openness to God, which constitutes the state of complete cognitive and conative neutrality towards God's existence; the second kind of case (conflicted engagement) is exemplified by the atheists who ardently regret that God does not exist or those who firmly believe that God exists but wholeheartedly resent God's existence.
One's faith is partial when one follows a reason not to have a positive cognitive attitude towards the existence of God (the kind of attitude I will henceforth call God-related doubt), which makes the cognitive constituent of one's faith less than full-blown, or when one follows a reason not to have a positive conative attitude towards God's existence (the kind of attitude I will call God-related worry), which does the same to its conative constituent. Hence, we can distinguish two basic types of partial faith, based on doubts and worries respectively – each with its accompanying stereotypes, partially reflected in their extant philosophical analyses.
One basic kind of partial faith occurs when the degree of the conative constituent of faith is much higher that the degree of its cognitive constituent – in other words, when reasonably high degree of a positive conative attitude towards the existence of God is combined with serious doubt related to it. It is natural (though not necessary) for such partial faith to be combined with the attitude of spiritual enquiry, which, for the purpose of the discussion, I will define as active openness to the possibility of acquiring new reasons for cognitive attitudes which could dispel one's God-related doubts but also strengthen them. While an enquiry may consist of research, it can also involve religious practice – such as prayer, sometimes intended to function as an experiment (an approach warned against in some religious traditions but nonetheless present in them).
In the other basic variety of partial faith, the degree of the cognitive constituent of one's faith is much higher than the degree of its conative constituent: one has a reasonably high degree of a positive cognitive attitude towards God's existence but also serious God-related worries. Spiritual struggle, the attitude it is natural to combine with this kind of partial faith, can be defined as active openness to acquiring new reasons for conative attitudes, conducive to achieving some kind of evaluative reconciliation with God – but also to widening the evaluative gap between God and oneself. (While the idea of evaluative reconciliation with God is thorny enough to be a topic of a separate article, it needs to be noted that even though its default mode recommended by religious traditions is striving to change one's evaluative judgement which leads to a worry, the same traditions also accommodate the possibility of a more negotiatory approach, exemplified for instance by some forms of petitionary prayer. Thus, changing one's conative attitudes is not necessarily the intention or hope behind a struggle – even if other possible outcomes may be difficult to describe.Footnote 5)
It has sometimes been argued that all or some kinds of struggle and enquiry are incompatible with full-blown faith – for example, because they cancel the risk associated with acting on faith (cf. e.g. Buchak (Reference Buchak, O'Connor and Callahan2014)). While there is no space or need here to discuss this issue in detail, I will follow Elizabeth Jackson in assuming that there are different kinds of spiritual enquiry, which may include some which are compatible or even required by faith but also some which do not make any difference to it (cf. Jackson (Reference Jackson, Rasmussen and Vallier2020), 81–83). Such an assumption, which can be extended to struggle, is consistent with the possibility that some varieties of struggle or enquiry (e.g. the ones involved in agnostic prayer) are only possible on partial faith.
Traditionally, the kind of partial faith combining a high degree of positive conative attitude towards the existence of God with a low degree of positive cognitive attitude is typically associated with the stereotype of a ‘seeker’ – someone who embarks on the enquiry concerning God's existence guided by the desire to get to the truth about the matter, but also by the desire for a relationship with God. Its opposite is the much more negative stereotype of a ‘struggler’, associated with the variety of partial faith which combines high degree of a positive cognitive attitude towards the existence of God with a low degree of a positive conative attitude. The figure of a ‘struggler’ partially overlaps with that of a sinner, which reflects the intuition that low or absent conative constituent of faith is accompanied by rebellion or fear – two sides of the spiritual equivalent of ‘fight or flight response’.
In spite of how widespread the stereotypes of seeker and struggler are, it is important to note that they are rarely an accurate portrayal of someone who engages in a spiritual enquiry and struggle respectively. If the reasons for cognitive attitudes are interconnected with the reasons for conative ones, the connection between enquiry and struggle is also closer than suggested by the connotations of the words ‘struggle’ and ‘enquiry’, emotional/volitional and intellectual respectively. Out of the consequences it has for partial faith, the most important one is that persons of partial faith (and those who lack faith) cannot be neatly divided into ‘pure’ seekers or enquirers, the equivalent of ‘nonresistant nonbelievers’ (Schellenberg (Reference Schellenberg2015), 74–88), and ‘tainted’ or ‘resistant’ strugglers. A person whose faith is partial – barring very unusual cases – combines some features of the stereotypical seeker with those of the stereotypical struggler.
Notably, it is possible to have partial faith in more than one God, in the sense of entertaining more than one concept of God at the same time. Someone like that may, for instance, have a relatively high degree of belief in, but only basic openness to, the existence of God1, while also being agnostic about God2 but strongly desiring God2 to exist. Though ‘faith in more than one (monotheistic) God’ may sound paradoxical, it can be argued that most people are in precisely such a position: different concepts of God coexist even within one religious tradition, so wavering between the positive cognitive or conative attitude in relation to, for instance, ‘God of justice’ and ‘God of mercy’ can be an essential part of one's spiritual life.
Even in the absence of major dilemmas in relation to God's nature, most people entertain different concepts of God in the sense in which one can entertain different concepts of one's friend when wondering what to expect from them in a particular situation. Admittedly, ‘different concepts’ as a way of describing such a state of mind may sound artificial, and one might ask how different the two pictures of God have to be to deserve the name of ‘concepts’. Still, for our purposes it is enough to agree that, if one has a positive cognitive attitude towards the existence of a God different from the God towards whose existence one has a positive conative attitude, the larger the difference between the two Gods, the more reason it provides not to call one's faith (in either of them) full-blown.
Introducing different concepts of God in the context of partial faith leads to a possibility which I will mention only briefly: partial faith in God1 constituted by full-blown faith in God2 only partially different from God1, that is, having some but not all features of God1 – having the nature of God1 only to some extent. This idea leads to a lot of questions concerning which it is easy to have conflicting intuitions. Would Spinoza have said that those who believe in Deus sive natura partially believe in the God of Judaism? Would the authorities who excommunicated him have shared his view on the matter? Is desiring the God of open theism partially desiring the God of classical theism? If J. L. Schellenberg's ultimism is a kind of faith, does an ultimist have partial faith in all the Gods that might possibly exist? Since paying these questions the attention they deserve is beyond the scope of this article, I will only note that the topic of partial faith becomes even more important if the latter is understood as encompassing such ‘unintentional ecumenicism’.
The kinds of partial faith described above, however, are just models or ideal types. Partial faith comes in numerous varieties, in which enquiry and struggle may play a central role or be entirely absent, and which give rise to very different combinations of attitudes, ranging from hopeful indecision through puzzled inkling to bitter, fearful confident belief. Sometimes, such attitudes constitute a stage of the ‘faith journey’ (or of the process of ‘losing faith’), but they can also be permanent. Whether permanent or temporary, partial faith of a particular kind can be more or less reasonable for a particular person at a particular time.
Tragedies, paradoxes, and unusual arrangements: why partial faith?
An obvious thing to say in support of partial faith is that it is a justified response to reasons pulling in opposite directions. To the degree that one has both the reasons to have a positive cognitive or conative attitude towards the existence of God and the reasons not to have it, one is justified in having a cognitive or conative attitude towards God's existence which does not amount to full-blown cognitive or conative constituent of faith or lack of faith. In a universe displaying extremely rich religious ambiguity, at least some people are likely to have such a set of reasons pulling towards and away from positive cognitive and conative attitudes towards God's existence that some kind of partial faith is justified as a response.
The above reasoning can be rejected on the grounds that, if God exists and the best possible relationship to God is the greatest good, all reasons not to have full-blown faith (or even maximal faith) ought to be entirely disregarded, since anything short of full-blown faith is in tension with such a relationship or hope for it. Thus, full-blown faith is the only faith worth having – even if one has faith only in the event God exists, as a form of Pascal's wager. This claim is based on the assumption that, even if the quality of one's relationship to God is not directly determined by the degree of one's faith, some kind of indirect determination is likely to occur – for instance, because such a relationship requires complete trust in God and trust, in turn, requires disregarding any reasons one may have for God-related doubts and worries (cf. e.g. Adams (Reference Adams1984), 6–12). I will argue that such indirect determination is partial at most, and that there are factors which in some cases are likely to make such faith better for such a relationship.
Faith made-to-measure: normative pluralism about faith
Intuitively, there is at least one situation in which partial faith is the best possible option as far as God is concerned – namely, when full-blown faith is not possible. But, if God exists and the best possible relationship to God is the greatest good, what can make partial faith the optimal option for someone whose options also include full-blown faith?
Given that there are multiple concepts of God, the relationship to God can be conceived in multiple ways. Whether God is personal or not seems to be especially significant for the nature of such a relationship. If God is personal, then one's relationship to God is a relationship with God, and insofar as God, being a person, is analogical to human persons, such a relationship is analogical to the relationships between humans – even if it is difficult to establish the limits of the analogy. It can be assumed that if one considers God to be personal and God's existence to be a good thing (i.e. has a full-blown conative constituent of faith in a personal God), one also has motivation (i.e. a positive conative attitude) to be in the best possible relationship with God as a person – where ‘best possible relationship’ can also be conceived in different ways, depending on what features God is considered to possess in addition to personality. From now on, by ‘relationship with God’ I will mean ‘personal relationship with personal God’.Footnote 6 I will also assume that one's motivation for being in the right relationship to God can contribute to such a relationship in a positive way even if one's concept of God and of such a relationship does not correspond to reality.
Each weighing of reasons for cognitive and conative attitudes brings with it advantages and disadvantages, including but not limited to advantages and disadvantages for the relationship to God. The weighing of reasons which is all-things-considered optimal if the relationship to God is not taken into account is likely to differ from the weighing optimal for such a relationship. This, however, does not mean that God-related advantages of particular weighings can be neatly divided from non-God-related ones – especially if God is personal. The nature and quality of interpersonal relationships significantly depend on the well-being, moral character, and other characteristics of the persons involved and are partially determined by the experiences and other factors which determine these characteristics. If God is a person, the same factors are likely to have an impact on the relationship one may have with God.
On the same set of experiential reasons to have faith and not to have it – that is, for the same person at the same time – different kinds of faith have different God-related advantages and disadvantages. This is because they lead to different relationships to God, with different benefits and drawbacks – something which is likely especially if God is a person. What we know about the relationships between people suggests that no interpersonal relationship can have all the benefits that may possibly occur in every relationship between two persons. This seems to be also true about human relationships with God, even if some benefits which are incompatible in the relationships between humans can coexist in the relationship between God and a human being (for instance, some mystics seem to have experienced God both as a parent and as a lover).
The kind of faith best for a particular person at a particular time is the one which leads to the relationship to God which at that time is best for that person, and this will mean different kinds of faith in different cases. It might well be the case that, on the same set of experiential reasons for cognitive and conative attitudes, two or more kinds of faith lead to different but equally good relationships to God. What matters for us, however, is that in some cases drawbacks are likely to outweigh benefits. If the drawbacks which stem from other factors than the shortage of faith can outweigh the benefits stemming from one's faith being full-blown (for instance, arguably, in a case when someone's concept of God is extremely infelicitous morally or otherwise), then, for the same person at the same time, some kinds of full-blown faith lead to a relationship to God which is not optimal. However, a kind of faith which leads to an all-things-considered flawed relationship to God on one set of experiences can lead to an optimal one on another – that is, for another person or for the same person at another time – so faith cannot be assessed without referring to experience.
If this kind of normative pluralism about faith contradicts some common religious intuitions concerning faith as an ideal uniting the faith community, it also supports and throws light on some other ones – especially that about the connection between faith and individual vocation. To use an example from the Hebrew Bible: Ruth's relationship with God was probably very different from Isaiah's, but they both seem to have had the best possible relationship with God, partially resulting from a very different kind of faith in each case. On the assumption that Isaiah's prophetic identity was essential for the relationship he had with God, if the cognitive constituent of his faith was exactly like Ruth's – consisting (for all we know, at least at the beginning) of acceptance based on the requirements of a relationship with a close family member rather than belief based on personal revelation – his relationship with God would probably have been flawed.
The exact way and extent to which one's perspective determines the quality of one's relationship to God is necessarily shrouded in mystery. This is so especially if God is personal: the relationship one may have with God, like every interpersonal relationship, is determined by the nature and attitudes of both participants, and, more than any other relationship between two persons, has aspects which are difficult to comprehend for at least one of them. With this reservation, I will now say more about the reasons to think that the amount of one's faith is not the only aspect of one's perspective which impacts the quality of one's relationship to God – a necessary condition for partial faith to sometimes be optimal as far as God is concerned.
Normative crossroads: partial faith and evaluative dilemmas
When it comes to assessing the impact of one's overall perspective on one's relationship to God, it is important to remember that such perspective involves more attitudes towards God than just the attitudes constitutive of faith. Depending on one's cognitive and conative attitudes towards other things, such attitudes will be more or less positive – for instance, if I think God is personal and omnibenevolent, I will trust God, and if I believe God punishes sins but wish it was not the case, I might fear God. On some sets of experiences or even universally, some positive attitudes towards God (e.g. longing or paying attention) are more consistent with neutral or even negative cognitive or conative attitude towards God's existence. In some cases, such attitudes may lead to a better relationship to God than the attitudes resulting from full-blown faith on the same set of experiences.
Admittedly, the situation in which the set of positive attitudes towards God which leads to the best possible relationship one may have to God is only possible on partial faith involves a paradox. It might even seem that the experiences which lead to them are possible only for people with very specific personality types – perhaps an agnostic philosopher preoccupied with figuring out whether God exists, who would lose an interest in God immediately after stumbling upon an argument for theism which he found conclusive (a natural enquirer), or a dauntless antihero in the vein of Melville's Ahab, for whom opposing God, seen as a cosmic sparring partner, is a compelling existential challenge, but who is incapable of worship (a natural struggler). Nonetheless, while these may be paradigmatic cases of the situation in which partial faith is better as far as God is concerned, taking a closer look at the factors at work in such cases will show that the situations in question are not necessarily that uncommon.
The examples described above show that it is possible to have a reason for a God-related doubt or worry such that taking it into account leads to partial faith and disregarding it so as to have full-blown faith contradicts some value relevant to such a relationship. When the harm to the relationship which comes with such contradiction exceeds the benefits resulting from giving up the worry or doubt, partial faith is the best option. As I will try to show, something like this can happen also to someone who is neither a natural enquirer nor a natural struggler.
What does it mean for a weighing of reasons to ‘contradict a value’? On many if not all concepts of God, some values (especially but not exclusively moral ones) are relevant to the right relationship to God in the sense that pursuing them (in the broad sense of acting so as to realize them) is beneficial for or even required by such a relationship – either directly (e.g. as an act of obedience) or indirectly (e.g. as causing one to develop a virtue which contributes to relating to God in the right way). In some cases, however, one may have a reason for a positive cognitive or conative attitude both towards the pursuit of some value being (in a particular case) ruled out by faith and towards its being required (in the same case) by the right relationship to God. For instance, one may have a reason to believe that faith requires disregarding a particular experience (and thus desire disregarding it) and a reason to believe that the pursuit of a value relevant for the relationship to God requires taking this experience into account (and thus desire to do it). This creates a dilemma whose one horn leads to faith which is less than full-blown; on the other horn, the relationship to God is indirectly harmed by not pursuing a value whose pursuit it requires.
It can be argued that such dilemmas never actually occur: if having a God-related worry always means following a reason to have a positive conative attitude towards something incompatible with God's existence or requirements (cf. e.g. Kahane (Reference Kahane2011), 681–685), any such reasons ought always to be disregarded in favour of full-blown faith. On this reasoning, any ascription of value which contradicts faith is false, so disregarding the reason for such ascription cannot harm one's relationship to God – the dilemma is illusive. But is this apparent illusiveness itself an illusion? Religions tend to provide precise guidelines on how to proceed in the situations which might look like dilemmas of this kind, so the possibility of such cases is difficult to defend within a religious tradition, but the seriousness of the threat is demonstrated by how often such guidelines are needed.
Since each of the two constituents of faith can consist of a voluntary attitude (like acceptance) or an attitude not under voluntary control (like desire and belief), a common assumption is that in cases of apparent dilemmas, the benefits for the relationship to God which stem from adopting a voluntary attitude leading to full-blown faith always override any drawbacks which might stem from not pursuing a value which contradicts such faith. Such a claim agrees with a common intuition that it is the voluntariness of faith – making a decision to take a cognitive or conative step beyond or even contrary to what our experience tells us – that makes faith virtuous, since the core of the relationship with God is something voluntary, like worship or obedience.
In answer to that, it can be noted that – as illustrated by the above comparison of Isaiah and Ruth – voluntariness and non-voluntariness as the features of the constituents of faith both have benefits and drawbacks, and it is not obvious that the benefits which stem from voluntary attitudes are always greater. After all, non-voluntary attitudes tend to be formed in response to experiential reasons, and, intuitively, one's relationship with God is better when one's personal experience confirms God's existence and its being a good thing – as is the case with any interpersonal relationship. (Notably, the experiential grounding of the relationship to God is less likely to matter if God is not personal. If God is not a person, the best possible relationship to God is not at the same time the best possible relationship with God, and it is more likely that the optimal weighing of reasons to have faith and not to have it does not put much weight on experience.)
Which values should be considered relevant for the relationship to God? Different concepts of God imply different answers to this question, pointing out different virtues and other existential goods. Notably, it can be argued that the nature of human relationship to God is such that pursuing any value is relevant to it – for instance, because God is the creator of all values, so pursuing any value brings us closer to fulfilling God's plan for the universe (or just makes the universe an overall better place). On a similar reasoning, epistemic virtues are likely to be relevant for the relationship to God because having them makes it easier for us to acquire the knowledge about both God and values.
Hidden reasons: partial faith and human flourishing
The connection between value and cognition might explain the possibility that following some kind of reasons not to have faith is obligatory even if God exists. This has been defended in the context of the cognitive rather than conative constituent of faith – for instance, as a condition of the reasonability of evidentialism given the existence of God. In such cases, the value contradicted by disregarding the reasons not to have faith is the pursuit of truth (or a virtue corresponding to the disposition to engage in such pursuit). It can be argued that some extant arguments for the claim that openness to evidence contributes to the excellence of one's faith can easily be reformulated as arguing that such openness, while potentially endangering faith, contributes to one's relationship to God.Footnote 7 If I am right about this, such reformulations – along with the related construals of non-doxastic faith – provide illustrations of cases in which partial faith is (actually or potentially) the optimal option as far as God is concerned.
That neutral or negative cognitive attitude towards God's existence can sometimes be optimal for the relationship to God is less controversial than that the same may be the case with the conative attitude – not only because of the intuition that intellect is less likely than will to fall into error, but also because conative attitudes seem to be more important for interpersonal relationships than cognitive ones. Still, if the universe is religiously ambiguous and the reasons for cognitive attitudes intertwine with the reasons for conative ones, the symmetry between the cognitive and conative constituents of faith in this respect is likely to be greater than such intuitions suggest – especially if each of the two constituents of faith can consist of a voluntary or non-voluntary attitude.
Notably, the harm to one's relationship to God which results from a God-related worry may not depend on whether the state of affairs involving values which the worry reflects actually occurs – or even on whether one has a positive cognitive attitude towards this state. There are many possible ways in which the weighing of reasons which results in giving up a worry might be harmful for such a relationship: the harm might occur because one has not pursued or realized the value (perhaps not developing some virtue) or because one believes that the value has not been realized; it might also occur because of some aspect of the experiences resulting from facing the dilemma, such as intense suffering. (To use an analogy: if Lady Macbeth demands that Macbeth kill Duncan for the sake of his devotion to her, Macbeth's devotion to his wife is likely to diminish independently of whether or not he does kill Duncan and even if – perhaps on utilitarian grounds, in a possible world where Duncan is a horrible tyrant – killing Duncan is the right thing to do; it might even diminish if Macbeth himself believes he ought to kill Duncan.)
Intuitively, the idea that the realization or pursuit of a value may be incompatible with faith brings to mind the situation of a tragedy, probably involving conflicting desires and obligations. In spite of that, such incompatibility can occur in a range of different situations, some of which do not involve any major internal conflict – especially if God is personal and absolutely free. As noted by Robert Audi: ‘It could well be that God prefers us to be intellectually virtuous and morally and aesthetically balanced, even if this means, for many of us, or for some periods in the lives of some of us, that we are less religiously committed than, on the basis of our psychological capacities, we could be’ (Audi (Reference Audi2011), 103). While this may seem an unusual arrangement, it may well be the case that for some people at some times full-blown faith is inconsistent with general flourishing, so God, being omnibenevolent, prefers them to have partial faith solely because of that – a possibility supported by the existence of analogical cases involving the relationships between human beings.
In a nutshell, then: if God exists and the best possible relationship to God is the greatest good, partial faith is the optimal option for someone whose options also include full-blown faith if the overall perspective resulting from some kind of partial faith is better for one's relationship to God than the overall perspective which results from full-blown faith of any kind. A situation like that occurs when one has such reason for God-related doubts and worries that, given God's nature and its implications, disregarding this reason leads to a worse relationship to God than taking it into account. While proving that there are such reasons would be difficult, what we know about the range of experiences different people can have at different times allows us to believe that they are likely to exist – especially if God is personal.
One possible issue with this line of reasoning is that it conflates the relationship to God in the abstract with the relationship with God as a person. As a result, it might make too much of the consequences of God's being personal, extending too many features of personal relationships between humans onto the highly mysterious relationship a human being may have with the personal God; it can be argued that any argument based on premises of this kind is tenuous at best, so partial faith cannot be defended on such grounds. It needs to be noted, however, that the less assumptions we are allowed to make about the inner workings of human relationships with God, the less reasons we have to accept any claim about what contributes to such relationships in a positive way; considered on purely apophatic grounds, partial faith cannot be said to be less likely to lead to the best possible relationship to God than any other option.
Does all that imply that for some people at some times the best possible relationship to God may require lack of faith? Taken together, the line of reasoning I described does show it is possible. At the same time, it shows that lack of faith is less likely than partial faith to lead to such a relationship – and not only because a case in which an entirely negative general attitude towards God would lead to relating to God in the overall best positive way is hard to imagine. The fact that partial faith is overall more justified is relevant to the question of which weighing of reasons is optimal for the relationship to God in virtue of the relevance that the pursuit of truth has for such a relationship. Overall, in the universe in which non-experiential reasons for and against faith display simple religious ambiguity, partial faith seems to be all-things-considered better than lack of faith, especially if God exists – though lack of faith may well be the optimal option for some people at some times.
Of course, in a religiously ambiguous universe like ours the question whether partial faith can be the optimal option if God does not exist deserves equal attention, and could be a topic of a separate article of equal length. Since there is no space to discuss it here in detail, I will only note that the answer to it largely depends on what role we ascribe to purely prudential reasons for cognitive and conative attitudes.
Why partial faith matters: some meta-level remarks
Views and perspectives akin to partial faith are currently a matter of increased philosophical interest – and continental philosophy was the first to focus its attention on the topic. Especially post-secularism, the worldview associated with those who ‘have been instructed in, and remained ambivalently committed to, both religious and secular modes of seeing and being’, has sometimes been understood as a set of ‘partial faiths’ (McClure (Reference McClure2007), 8–9), but a similar idea can be found at the core of such contemporary proposals as Gianni Vattimo's ‘weak thought’ or John Caputo's ‘religion without religion’ (cf. Lohrey (Reference Lohrey2019)), and, perhaps, Richard Kearney's anatheism.
This appreciative interest is a new thing: the traditional picture of partial faith tends to identify it with either tepid indifference or bitter internal conflict. An example of the first portrayal can be found in ‘the miserable condition of the sorry souls who lived without infamy and without praise . . . mingled with that base band of angels who were neither rebellious nor faithful to God, but stood apart’ (Alighieri (Reference Alighieri1971), 27) – the inhabitants of the antechamber of Hell as imagined by Dante. The second kind of depiction is exemplified by the characters as different as Augustine of Hippo's younger self, depicted in the Confessions as having ‘two wills, one new and the other old’ (Augustine of Hippo (1912), 425), and Satan from Milton's Paradise Lost, who experiences ‘the hateful siege of contraries’ (Milton (Reference Milton2005), 200). In both cases, the feature of partial faith which comes to the foreground is its moral or psychological imperfection – a normative rather than descriptive property. Another thing which unites both portrayals is that they have largely been painted against the canvas of a perspective resulting from full-blown faith.
The fact that partial faith has usually been discussed from either entirely religious or entirely secular viewpoint might be the reason why it has so often been interpreted in terms of a moral or intellectual vice: pride, negligence, laziness, instability, or succumbing to inordinate desires. (Though the latter may seem to be a concept confined to moral theology, a connection often made by atheistic thinkers between religion and wishful thinking suggests that there exists its secular equivalent.) It might also be why even the authors as restrained in their normative judgements as William James have historically tended to focus in their analyses on the cases in which ‘the religious spirit is unmistakable and extreme’ (James (Reference James2002), 36) and their opposites rather than on what cannot be categorized as belonging to any of the two sets. In short, there are reasons to think that the map of partial faith which full-blown faith has drawn contains gaps or even distortions.
If we are interested in how partial faith looks from its own viewpoint – conflicted and open to change but not necessarily flawed or incomplete – outlining the considerations in favour of it may be a useful first step. In my view, the core consideration of this kind is the difference some kinds of personal experience make for the relationship to God – something easiest to show on the account of reasons for and against faith which is neither entirely externalist nor strongly normative.
It can be argued that it does not make sense to discuss faith in normative terms without assuming that, if we ought to have it, it is because God exists – and if internalism about reasons leads to the conclusion that partial faith is acceptable, all the worse for internalism about reasons. One possible answer to this is to note that the picture emerging from externalism is one dominated by the sharp division between those who have faith and those who lack it, while on the internalist picture, which contrasts rather than merges the shades and degrees of different possible attitudes towards God, those whose faith is partial form a faction both large and diverse. Notably, it is the second framework that seems to reflect social reality more closely, with many people seeking, wavering, conflicted, undecided, confused, or unsure. This suggests that, by embracing more internalist intuitions about faith, we may be able to avoid a major blind spot.
While everything I said is compatible with different notions of what exactly constitutes a reason for an attitude, it becomes more intuitive if we assume that some reasons for cognitive and conative attitudes are purely prudential. In addition to that, it implicitly assumes that at least some benefits (perhaps overridden by other considerations in some or all cases) are attached to authenticity or personal integrity in the sense introduced by Bernard Williams in his discussion of utilitarianism – namely, that ‘one who displays integrity acts from those dispositions and motives which are most deeply his, and has also the virtues that enable him to do that’ (Williams (Reference Williams and Williams1981), 51). In the end, perhaps the best way to elucidate the main thought behind this article is to re-read a part of Williams' argument against utilitarianism, replacing ‘utilitarianism’ with ‘full-blown faith’, ‘utilitarian/non-utilitarian’ with ‘related/unrelated to full-blown faith’, and ‘human welfare’ with ‘relationship to God’:
There are many and various forms of dispositions, patterns of feeling and desire, which can motivate people to attitudes contrary to full-blown faith; some themselves virtues, some more particular projects, affections and commitments. . . . I think that it is wrong to try to reduce all questions of the assessment of such dispositions to considerations related to full-blown faith, and indeed that it is incoherent, since there is no coherent view of the relationship to God which is independent of such issues as what people care for, in the spirit unrelated to full-blown faith, with regard to such things as these dispositions. . . . The difficulty is that such dispositions are patterns of motivation, feeling and action, and one cannot have both the world containing these dispositions, and its actions regularly fulfilling the requirements of full-blown faith. (ibid.)
Does this mean that my defence of partial faith is radically incompatible with the common religious view that everyone ought to have full-blown faith? Not necessarily. That partial faith is the best option for some people at some times does not rule out the possibility that for everyone there will be a time when full-blown faith will be the best option – in other words, that everyone will ultimately reach the point when full-blown faith will be available and optimal for them. This would mean that everyone ought to have full-blown faith at some point – but for some people at some times it means (to invoke Augustine once more) not yet. While the existential weight of permanent partial faith might make it more philosophically interesting, everything I said can, if the reader so wishes, be read as referring to partial faith held only temporarily.
It can be hoped that if devoting more attention to partial faith helps in understanding the role played in faith by individual experience, it may also shed light on some issues related to faith in general. Daniel J. McKaughan writes:
The task of clarifying the precise role that attitudes such as hope, trust, and acceptance can play in connection with religious commitment, reflecting on their epistemological implications, and elucidating plausible constraints on the cognitive dimensions of these attitudes and combinations of attitudes is one of the most exciting projects in philosophy of religion today. (McKaughan (Reference McKaughan2013), 117)
Analysing the consequences of different cognitive and conative attitudes constitutive of different varieties of faith (or of faith as understood in different ways) becomes even more difficult – but also more interesting – if we include in our considerations different varieties of partial faith as an alternative way to relate to God in a potentially positive way. I will now point out two issues related to this which seem to me most in need of discussion.
One interesting problem is that of conflicting attitudes. If many different attitudes can play the role of a cognitive and conative constituent of faith, what happens when one has two or more cognitive or conative attitudes which lead in opposite directions? Do such cases exemplify a variety of partial faith I did not include in my account? In the words of the protagonist of James Wood's Book Against God, a novel which can be interpreted as a literary analysis of the situation in which one has both atheistic acceptance and theistic belief: ‘after all, belief and unbelief are not absolutes, and not absolute opposites. What if they are quite close to each other, I mean belief shadowed by unbelief and vice versa, so that one is not exactly sure when one begins and another ends?’ (Wood (Reference Wood2003), 49).
Another issue concerns the possibility of different frameworks when it comes to dividing full-blown from partial faith. On my account, at least some kinds of non-doxastic faith constitute partial faith (which, if partial faith can sometimes be the optimal option, does not make non-doxastic faith any worse), while, presumably, those who consider belief to be essential for faith would say non-doxastic faith is always partial. What would be the consequences of adopting a framework in which it is (e.g.) acceptance that is considered essential? Would it make sense to say that one kind or concept of faith constitutes partial faith in relation to another concept or kind? (And can some conflicts within religious communities be explained by the possibility that different members of the community consider different concepts of faith to be normative in this sense?) In short, can the concept of partial faith be helpful in explaining why different kinds of full-blown faith – even or maybe especially in the same religious tradition – seem sometimes to be in conflict?
I hope to have shown that seeing in a more favourable light the attitude of those who have faith only partially can be philosophically fruitful in more than one way. Apart from moving forward the debate on the nature of faith, paying more analytic attention to the problem of partial faith might help bridge the gap between analytic philosophy of religion, strongly focused on orthodox Christianity, on the one hand, and its continental counterpart, more concerned with liminal and ambiguous kinds of spirituality, on the other. Ultimately, however, perhaps the significance of the attempts to defend partial faith lies primarily in their challenge to the common view that it does not deserve the respect religious and secular worldviews grudgingly pay each other in both private and public life. If some people at some times ought to have faith only partially, taking partial faith's voice into account may be a good idea in the universe as religiously divided as it is religiously ambiguous.
Acknowledgements
I would like to thank Daniel Howard-Snyder, Timothy J. Mawson, Daniel McKaughan, two anonymous reviewers, and the participants of the grant seminar – Piotr Gutowski, Przemysław Gut, Marcin Iwanicki, Piotr Lipski, Piotr Szałek, and Joanna Teske – for their comments on the earlier versions of the article.
Financial support
The research for this article was funded by the John Templeton Foundation and the Ian Ramsey Centre for Science and Religion at the University of Oxford as a part of the grant project ‘Religious Faith and Practical Aspects of Life’ headed by Piotr Gutowski within the research scheme New Horizons for Science and Religion in Central and Eastern Europe.
