Traditional Christian theology maintains that God created human beings, male and female, in the divine image. But the traditional pattern of characterizing God in scripture, authoritative confessions and creeds, and much of Christian theology throughout the centuries seems predicated on the assumption that God is masculine. God is traditionally referred to as ‘he’ and as ‘our heavenly Father’, for example; and even the figurative characterizations in scripture and elsewhere (e.g. as the faithful husband of wayward Israel) are predominantly masculine. By virtue of the relationship between masculinity and maleness, this pattern might seem to suggest that males bear the image of God to a greater degree than females, that they are more fit to represent God, and that whatever traits contribute to making someone masculine are somehow more divine than whatever traits contribute to making someone feminine. So many traditionalists have thought; so many feminists have objected.
There is a persuasive case to be made for the conclusion that the traditional pattern of characterizing God in predominantly masculine terms has been harmful to women.Footnote 1 To the extent that it encourages a false view of their own superiority, and enables various kinds of oppressive behaviour, there is also a case to be made for the conclusion that it has been harmful to men.Footnote 2 There is, furthermore, good reason to think that this harmful state of affairs is at least partly a result of the fact that women's voices have largely been excluded from the theological conversations that have shaped the tradition.Footnote 3 In the face of such considerations, it is an urgent question whether the traditional pattern of characterizing God in predominantly masculine terms is theologically mandatory. That is to say, it is an urgent question whether we are somehow morally or rationally (in the epistemic sense) required to uphold and persist in the traditional pattern, or whether doing so is somehow liturgically important. If the answer is ‘no’, then the traditional pattern is theologically optional.
Some, to be sure, have suggested that the harms done by the traditional pattern are overstated.Footnote 4 But even if that were true, it is undeniable that many people believe that women have been and continued to be harmed by the traditional pattern, and that Christendom's failure to adopt more gender-inclusive ways of speaking about God has presented such people with serious, ongoing obstacles to faith and to comfortable participation in their own religious communities. This fact alone constitutes reason to adopt more inclusive ways of characterizing God, unless there is some reason for thinking that the traditional pattern is theologically mandatory. But what reason could there be? Why endorse traditionalism, the view that the pattern of characterizing God in predominantly masculine terms is theologically mandatory, rather than inclusivism, the view that feminine characterizations of God are just as legitimate as masculine ones?
There is no explicit biblical affirmation of traditionalism. Nor can traditionalism be defended by appeal to facts about God's body; for, incarnational considerations aside, the Christian tradition speaks virtually unanimously against the idea that God has a male or female body. Consequently, defences of traditionalism tend to argue either for the singular theological fittingness of masculine characterizations of God, or for the conclusion that the biblical pattern of characterization indicates that God wants to be characterized in predominantly masculine terms. Both kinds of argument would be undercut, however, if it could be shown that masculine characterizations of God are no more metaphysically accurate than feminine ones. Given the perceived harms of characterizing God in predominantly masculine terms, why would a perfectly loving God insist on being characterized in this way if feminine characterizations are equally accurate? Why would it be theologically fitting to continue in a metaphysically misleading pattern of characterization that raises obstacles to faith and the religious life? It is hard to see good answers to these questions. Consequently, in defending the claim that masculine characterizations of God are no more (or less) accurate than feminine ones, I take myself to be arguing against the best available underpinning for traditionalism.
I am hardly the first to oppose the claim that masculine characterizations of God are more accurate than the alternatives. One of the main projects of so-called second-wave feminist theology in the 1980s and early 1990s was the idoloclastic attempt to show that the traditional Judaeo-Christian concept of God as masculine is a pernicious misrepresentation of God.Footnote 5 But, broadly speaking, the parties to the feminist side of the debate offered just two main strategies for defending the inclusivist position.Footnote 6 Some dismissed the traditional concept of God as an idolatrous invention, a mere projection onto the divine of male-centric values and fantasies.Footnote 7 Others appealed to divine transcendence. A suitably strong theology of transcendence will imply that relatively few (if any) characterizations of God are strictly metaphysically accurate, thus making it extremely difficult to support the view that masculine characterizations are metaphysically privileged.Footnote 8 On both ways of thinking, the traditional pattern of characterization is not due to its greater metaphysical accuracy but rather to the unjust suppression of feminine characterizations by sinful men in power speaking and writing under the influence of entrenched biases against women.
Feminists have largely moved on from these particular debates, in part because some of the underlying assumptions that helped determine the contours of the debates (e.g. that sex and gender are binary categories, or that gender supervenes on biological and psychological traits) are now more contested than they previously were. Even so, most inclusivists seem content with some variation on one of the two strategies I have just described. Most, that is, except for those working in the analytic tradition of philosophy. As is well known, analytic theologians tend to have very little sympathy either for projectionism or for strong theologies of divine transcendence. What then have they said about the feminist critique of the traditional concept of God? Remarkably little. Although frequently and in various ways targeted by inclusivists for uncritically embracing a masculine conception of God, those working in the analytic tradition have mostly opted simply to ignore the critique rather than to explore their own distinctive resources for addressing it. As a result, there is now a significant gap in the literature. What might one say if one is sympathetic towards feminist concerns about traditionalism but unsympathetic towards projectionism and strong theologies of transcendence? This is the question that animates the present article.
My point of departure is the following insight from Elizabeth Johnson's landmark book, She Who Is: ‘If women are created in the image of God, then God can be spoken of in female metaphors in as full and limited a way as God is imaged in male ones' (Johnson (Reference Johnson1993), 54). Out of context, the remark seems to suggest that, on Johnson's view, the primary rationale for inclusivism is the imago Dei doctrine. In fact, however, this is not the primary rationale that we find in the rest of She Who Is. Although the imago Dei doctrine certainly plays a role in her overall case, the metaphysical underpinning for her brand of inclusivism is, as I have already noted, the doctrine of divine transcendence.Footnote 9 My own goal in the present article is to provide metaphysical underpinning for inclusivism that dispenses with the latter doctrine and rests more theoretical weight on the imago Dei doctrine.
The main argument summarized
In the following sections, I argue from a particular understanding of the imago Dei doctrine (elaborated below) via the method of perfect being theology for the conclusion that masculine characterizations of God are no more or less accurate than feminine ones. To say that one way of characterizing God is more accurate than another is just to say that it comes closer to telling the straightforward, literal truth about what God is really like (intrinsically or extrinsically, essentially or contingently). Accuracy can, of course, come apart from legitimacy, permissibility, or propriety. It doesn't automatically follow from the fact that masculine and feminine characterizations come equally close to telling the truth about God that it is morally permissible or liturgically appropriate to characterize God in feminine terms. Accuracy is a matter of metaphysics; propriety is a matter of what is overall most fitting in light not only of what we know about the relevant metaphysics but also of what we know about God's preferences, likely harms and benefits to human beings and their relationships with God, and so on. Further premises, therefore, would be required to establish inclusivism, which asserts not the equal accuracy but rather the equal propriety of characterizing God in feminine terms. From the remainder of this section up until the beginning of the article's closing section, I shall be concerned with matters of metaphysical accuracy. In addressing these issues, I take myself indirectly to be addressing the question whether we are rationally required to characterize God in predominantly masculine terms. In so far as we can see that it is not most accurate to do so, we are not rationally required to do so. I then turn to consider questions about the moral and liturgical propriety of feminine characterizations of God in the final section of the article.
My main argument is simple. (i) God is most accurately characterized as masculine (say) only if God is masculine and God is not equally feminine. However, so I shall argue, (ii) God is masculine or feminine only if God is equally masculine and feminine. Therefore, (iii) God is not most accurately characterized as masculine. By the same reasoning, mutatis mutandis, God is not most accurately characterized as feminine (or as belonging to any other gender).
I take the first premise of this argument to be obviously true. Thus, my remarks in what follows shall focus on defending the second premise. Another way of stating that second premise is to say that gender is a divine attribute only if God belongs equally to both genders. I shall use the term genderism to refer to the negation of that premise, i.e. to the following thesis: gender is a divine attribute (i.e. God has gender) and God is either more masculine than feminine, or vice versa. My defence of premise (ii) shall consist in an argument for the conclusion that genderism is false. As I have indicated, my argument rests on what I take to be obvious implications of the thesis that women, like men, are created in the image of God.Footnote 10
Does anybody in fact endorse genderism? Initially, one might have doubts. It is commonly said that the general consensus in the Christian tradition is and always has been that God is beyond gender; it is also, admittedly, rather difficult to find theologians who clearly endorse genderism. However, careful attention to the texts commonly adduced as evidence of consensus that God is beyond gender reveals that those texts in fact only support the thesis that there has been consensus that God is beyond sex. (Cf. Cooper (Reference Cooper1998), 168–169.) Moreover, even in the absence of explicit theological affirmations of genderism, the history of iconography – in which God and each person of the trinity are overwhelmingly regularly depicted as male – bears strong testimony to the fact that masculine genderism has, on some level, been the dominant view throughout church history.Footnote 11 Finally, although it is indeed hard to find explicit affirmations of genderism as I have formulated it, it is easy to find affirmations of claims close enough in the neighbourhood. Robert Jenson (Reference Jenson and Kimel1992), for example, insists that there are neither linguistic nor metaphysical reasons for thinking that Jesus' address of God as ‘Father’ is non-literal, and that the Church decided during the fourth-century Arian controversies that this mode of address is both non-literal and absolutely to be preferred over feminine or neutral forms of address (ibid., 105).Footnote 12 Again, this is not an explicit affirmation of genderism; but it is a nearly explicit affirmation of the claim that God is literally our Father rather than our Mother or our beyond-gender Parent. It seems hardly a stretch, then, to say that Jenson affirms genderism.Footnote 13
The rest of the article unfolds as follows. In the next section, I lay out some of my assumptions about gender and the imago Dei doctrine that are needed to understand my argument against genderism. In the following section, I present that argument. Then, in the closing section, I consider whether my argument against genderism has negative implications for the moral and liturgical propriety of characterizing God exclusively or predominantly in masculine terms.
Gender and the imago Dei
In accord with conventions that are presently fairly standard, I use the terms ‘male’ and ‘female’ when talking about sex, and ‘masculine’, ‘feminine’, ‘man’, and ‘woman’ when talking about gender. Furthermore, I do not presume that gender necessarily correlates with sex. Accordingly, my terminology allows that there might be male women and female men. It also allows that someone might be a man or woman, masculine or feminine, without being either male or female. I shall generally try to remain neutral on the metaphysics of sex and gender; but for convenience, I will talk as if they are binary categories and as if sex is not socially constructed whereas gender might be. Nothing in my argument depends on these choices, both of which are now contentious.Footnote 14
There seem to be two different (but not mutually exclusive) things that one might mean by saying that a characterization of someone is gendered, i.e. masculine or feminine. First, one might mean that it involves the use of a pronoun or predicate that can be literally applied only to members of a particular gender. For example, it is a conceptual truth that the pronoun ‘he’ and predicates like ‘is a father’ or ‘is a king’ apply literally only to men. (Some will say that they apply literally only to males; but I imagine one would say this only if one thought that only males could be men.) I'll refer to characterizations that use pronouns and predicates like this as strongly gendered.
Second, one might mean that the characterization involves the application of a predicate that stereotypically applies to members of just one gender, even though it can be literally applied to members of both genders. Predicates like this I'll refer to as weakly gendered. In the literature, for example, we find claims to the effect that characterizations of God as powerful and a se are masculine, whereas characterizations of God as nurturing or as brooding over God's children are feminine. Those who make such claims do not generally mean to suggest (e.g.) that being powerful is a sufficient condition for being a man, or that being nurturing is a sufficient condition for being a woman. Rather, their view seems to be that these predicates are gendered by virtue of the stereotypes associated with each gender.Footnote 15
Both strongly and weakly gendered characterizations have been seen to be problematic. Strongly masculine characterizations of God obscure (or perhaps even tacitly deny) that traits definitive or stereotypical of womanhood reflect the image of God. The same is true of weakly masculine characterizations when their use greatly overshadows the use of weakly feminine ones. Mary Daly famously said that: ‘If God is male, then the male is God’ (Daly (Reference Daly1973), 19). By extension, we might also say that if God is (exclusively or predominantly) masculine, then the traits that are definitive or stereotypical of masculinity are more divine than those definitive or stereotypical of femininity. But, according to the second-wave feminist critiques, it is precisely this thought, together with the fact that gender typically (even if not uniformly) tracks sex, that has led to the subordination of females within the Christian tradition.
So much for preliminary terminological points. Let me now highlight two assumptions. First, I assume that sex is not a divine attribute: God is neither male nor female. (The incarnate Christ was male; but presumably his maleness is no more a divine attribute than his tallness, his hairiness, his musculature, etc.) Accordingly, I also assume that in offering a gendered characterization of God, one does not thereby commit oneself to the claim that God has, or is even imagined to have, male or female bodily features. The first assumption I take to be obviously true; the second is a natural correlate of the first.
Finally, I want to make some brief remarks on how I understand the version of the imago Dei thesis on which my arguments depend. There are various ways of interpreting the claim that women and men equally are created in the image of God, some of which put it in conflict with theories about gender that ought to be compatible with it. For example, one might read it as saying that some people have been created as women, others as men, and the intrinsic properties that make someone a woman are no less a part of the divine image than those that make someone a man. On that reading, one cannot affirm the equality thesis while also affirming (e.g.) that gender is a social construction, or that gender membership reflects something about where one stands in the structure of power and oppression within one's society.Footnote 16 But surely that is a bad result. The equality thesis ought to conflict with theories according to which some people by virtue of their gender bear the divine image to a greater degree than people of a different gender. It should not, however, necessarily conflict with theories according to which gender is socially constructed. Nor should it necessarily conflict with theories according to which gender is determined more by social position than by personality and behavioural dispositions – especially since the latter traits are clearly more relevant to the divine image, and were the sorts of traits primarily in view in the debates of the 1980s and early 1990s.
In light of all of this, I shall understand the equality thesis to be equivalent to the following claim: All human beings are created equally in the image of God; and if mental and behavioural characteristics contribute at all to gender membership, then the ones that contribute to making someone a woman are no more or less relevant to her bearing the image of God than those that contribute to making someone a man. I hope, too, that in showing how one might recast the equality thesis in such a way as to eradicate certain contested assumptions about gender, I will have gestured toward ways of reformulating any other claims in this article that might unintentionally presuppose contested theses about gender.
In the next section, I provide an argument against genderism. If genderism is false, the following disjunction is true: gender is not a divine attribute or God belongs equally to both genders. The first disjunct implies that strongly gendered terms do not literally apply to God – God transcends gender categories. Importantly, however, it does not follow from this that weakly gendered terms and predicates have no literal application to God. Thus, even if predicates like ‘is a se’ or ‘is nurturing’ are weakly gendered, one can still coherently affirm that gender is not a divine attribute. Likewise, one can affirm that gender is not a divine attribute while still maintaining that strongly gendered pronouns or predicates have an important place in our theology. For, after all, such terms might be applicable to God, and the relevant figures of speech might be theologically important. I affirm with scripture and the oldest creeds of Christendom that God is my heavenly Father. Likewise, I can agree with Julian of Norwich that ‘as truly as God is our Father, so truly is God our Mother.’Footnote 17 Both claims are theologically important; both claims express truths. And this is so even if both are metaphors.
The argument against genderism
Every divine attribute is such that God has it either essentially (i.e. of necessity) or contingently. Every divine attribute is furthermore either intrinsic to God or extrinsic. Thus, every divine attribute falls into one of the following mutually exclusive and jointly exhaustive categories: (i) essential intrinsic, (ii) essential extrinsic, (iii) contingent intrinsic, or (iv) contingent extrinsic.Footnote 18 Those who endorse the doctrine of divine simplicity, according to which God is identical to God's attributes, will insist that God has no attributes outside the first category. Those who reject divine simplicity and endorse an abundant theory of properties have room in their ontology for divine attributes that fall into some of the other categories.Footnote 19
In the argument that follows, I make three further assumptions. First, in order to give my argument widest scope, I assume that the doctrine of divine simplicity is false and that every non-paradoxical predicate that has a literal meaning corresponds to an attribute.Footnote 20 I take it that we can generally tell which predicates are meaningful; and a predicate has a literal meaning if, and only if, it is not irreducibly figurative. A predicate F is paradoxical iff, on the assumption that F is a meaningful predicate, there is something x such that Fx ≡ ¬Fx. (For example, on the assumption that is not a predicate that applies to itself is a meaningful predicate, it is paradoxical.) Note that I am not assuming that paradoxical predicates do not correspond to properties; I simply do not wish to presuppose that they do.
Second, I assume that the term ‘divine attribute’ refers only to attributes had by at least one of the three persons of the trinity independently of the incarnation. There is no question that Jesus of Nazareth was male; there can be little question that he was also masculine. But just as we would not want to say, on the grounds that Jesus was bearded (if indeed he was), that being bearded or being embodied is a divine attribute, so too we should not want to say simply on the grounds that Jesus was a man that being masculine is a divine attribute.Footnote 21
Third, I assume that strongly gendered terms (either pronouns or predicates) literally apply to God only if God has an attribute corresponding to a strongly gendered predicate.
With these assumptions in hand, I shall argue for each of the following three premises:
P1. Strongly gendered attributes are among God's essential attributes only if God belongs equally to both genders.
P2. The category of contingent intrinsic divine attributes is empty.
P3. Strongly gendered attributes are among God's contingent extrinsic attributes only if God belongs equally to both genders.
The three premises exhaust the four categories of divine attribute; thus, if all three are true, it follows that God has strongly gendered attributes only if God belongs equally to both genders. I take this conclusion to be equivalent to the thesis that genderism is false.
Let us begin with P1. The first step is to argue that no strongly gendered attribute is essential to all three persons of the trinity unless each person is equally masculine and feminine. This conclusion is delivered in two ways by the method of perfect being theology deployed under the assumption that it is neither better to be masculine than feminine nor better to be feminine than masculine. Here again the conviction that women and men equally are created in the image of God is crucial; for it is precisely that conviction that provides theological underpinning for the view that it is neither better to be masculine than feminine nor better to be feminine than masculine.
Here is the first way: it seems plausible that, for every attribute A that is essential to all three persons of the trinity, it is intrinsically better to be A than not-A.Footnote 22 The principle seems clearly to be true for attributes other than strongly gendered ones; thus it seems most sensible to say that it holds of strongly gendered attributes as well. But it is not intrinsically better to be masculine than to be feminine, or vice versa; thus, no strongly gendered attribute is essential to all three persons of the trinity unless each person is equally masculine and feminine.
One might object: every member of the trinity has essentially the attribute not being identical to the number 2. But it seems neither better to have that attribute nor better to lack it. (The number 3 has that attribute whereas the number 2 lacks it, but neither number is thereby better than the other.) Thus, there seems to be reason to reject my principle that for every attribute that is essential to all three persons, it is better to have the attribute than to lack it. I reply as follows: it is intrinsically better to be distinct from the number 2 than to be identical to the number 2; but the reason we don't think that the number 3 is thereby better than the number 2 is simply that it is also better to lack the property being identical to the number 3 than to have it. None of this implies that it is bad to be identical to the number 2 or to the number 3, just that something is intrinsically better if it is distinct from them (and from all other numbers, and, indeed, from anything else besides God).
Here is the second way to arrive at the conclusion that no strongly gendered attribute is essential to all three persons of the trinity unless each person is equally masculine and feminine: under the method of perfect being theology, the recipe for determining the essential divine attributes is (roughly) to sort the attributes we know of into two categories – great-making properties, and other properties. The divine attributes, then, comprise the best possible array of mutually compatible great-making properties. But, under the assumption that it is not better to be masculine than to be feminine, or vice versa (together with the assumption that there is no hybrid option), we can readily see that strongly masculine attributes belong in the category of great-making properties only if strongly feminine attributes do. A fortiori, strongly masculine attributes belong in the best possible array of such properties only if strongly feminine ones do.
It follows, then, that at least one person of the trinity is essentially masculine only if at least one person of the trinity is essentially feminine; and at least one is essentially feminine only if at least one is essentially masculine. For suppose every person of the trinity were masculine. If being masculine and being feminine are equally great-making properties, then God would have been no better or worse had every person of the trinity been feminine. But then God does not have the greatest possible array of compatible great-making properties, contrary to the hypothesis of perfect being theology. To avoid this consequence, one of the following must be true: (a) at least one person of the trinity is masculine whereas at least one other is feminine; (b) all persons of the trinity are both masculine and feminine; or (c) no person of the trinity is either masculine or feminine.
Suppose it turns out that two persons of the trinity (Father and Son, presumably) are masculine whereas only one person of the trinity is feminine,Footnote 23 and suppose that each divine person's gender is essential to him or her. In that case, does it follow that God is more masculine than feminine? In a word, no. This for two reasons.
First, and in general, a claim of the form ‘God is more F than G’ is true only if (a) every person of the Trinity is more F than G; (b) the divine nature is more F than G; or (c) the trinity as a whole is more F than G. Where masculine and feminine are substituted for F and G, (a) and (b) are ruled out by the imago Dei doctrine. If every person of the trinity were more masculine than feminine, or if the divine nature were more masculine than feminine, men and women would clearly not be equally created as such in the image of God – men as such would be greater image-bearers than women. And it seems that (c) is satisfied only when F and G are properties had by all three persons. For, after all, we don't say (e.g.) that God is more discarnate than incarnate, more non-indwelling believers than indwelling believers, less distinct from Father and Son than identical to Father or Son, etc. But we have already established that if all three persons of the trinity are masculine, all are equally feminine (and vice versa). Thus, on the supposition that two members of the trinity are essentially masculine and one is essentially feminine, it is still not the case that God is more masculine than feminine.
Second, and perhaps more importantly, to say that God is more masculine than feminine when two divine persons are masculine and one is feminine is equivalent to saying that the term ‘masculine’ more aptly describes the Godhead – more adequately gets at what God really is – than the term ‘feminine’. But to say this, in turn, is to subordinate the feminine person of the trinity – to say that she is somehow less real, or less really God, than the other persons. But, of course, this is just the heresy of subordinationism; it is not an option for Christian theology. If every person of the trinity were more masculine than feminine, men and women as such would not be equally created in the image of God – men would be greater image-bearers than women. Thus, I conclude that P1 is true.
P2 says that the category of contingent intrinsic divine attributes is empty. I reach this conclusion as follows. Suppose (for reductio) that God is intrinsically F but might have been intrinsically G, where being intrinsically G implies not being intrinsically F, and vice versa. Now, if God were G, then God would be better, worse, or neither better nor worse than God in fact is. If God would be better by virtue of being G, then God is not perfect; if God would be worse by virtue of being G, then God might not have been perfect. But God is necessarily perfect. Thus, being G is neither better nor worse than being F. But if that is true, then it seems to be a matter of sheer happenstance whether God is F or G. If intrinsic features of the divine nature depend on the divine will, then (since it is neither better nor worse to be G rather than F) there is no explanation in the divine will for why God is F rather than G; and divine aseity implies that there is no explanation outside God for God's intrinsic attributes. Thus, it is sheer happenstance that God is F rather than G; and, if so, then God's being F rather than G is wholly independent of God's will. But none of God's intrinsic features is wholly independent of God's will – this, too, is a consequence of divine aseity. Thus, it is not sheer happenstance that God is F rather than G. (Contradiction.)
If this argument is sound, it follows that God has no contingent intrinsic values or desires. If that is right, then one might think that either God's having the values and desires God has is a contingent extrinsic attribute of God or else God's creative options are severely constrained. The reason is that what God creates plausibly depends upon what God wants and what God values. Thus, if God might have created different things, it seems to follow that God might have wanted or valued different things, from which it seems to follow that God's having the values and desires God has is a contingent matter. So, if God has no contingent intrinsic values and desires, and if God might have created different things, then it must be a contingent extrinsic matter what God desires and values. The latter conclusion seems incompatible with divine aseity; for it seems to imply that divine desires and values – and hence perhaps also divine goodness – depends somehow on factors external to God. But denying that God might have created different things seems to raise problems for a robust understanding of divine freedom. So we might seem to face a terrible dilemma. But I reject the initial disjunction. Suppose God's values and desires are intrinsic. Nevertheless, they might (for all we know) be conditional on how certain matters of chance turn out; and God might (for all we know) necessarily value leaving certain matters – the truth values of counterfactuals of creaturely freedom for example, or the movements of fundamental particles – to chance. If that is right, then there is room to say that God might have created different things, and this despite the fact that divine values and desires are intrinsic.
Finally, I turn to P3: God has strongly gendered contingent extrinsic attributes only if God belongs equally to both genders. (Other examples of contingent extrinsic attributes might include being worshipped by Paul or being the creator of the Milky Way galaxy.) Many theories of gender maintain that gender is extrinsic by virtue of being at least partly socially constructed. Moreover, any credible theory of gender according to which gender is both extrinsic and potentially a divine attribute will maintain that a person's gender depends in part upon some combination of outwardly directed beliefs and desires, behavioural dispositions, or observed behaviour towards others. If gender depended entirely upon internal and internally directed mental states and behaviours, it is hard to see why it would not be intrinsic; and if it did not depend at all upon beliefs, desires, behavioural dispositions, or actual behaviours, then it is hard to see how it could possibly be an attribute of a disembodied personal God.
Given all this, it is easy to see why P3 is true. We have seen that God has no strongly gendered intrinsic attributes; likewise, God has no strongly gendered essential attributes. The question, then, is whether we have reason to think that there is anything about God's beliefs and desires or about the social significance of God's observable behaviour that would ground the literal application of strongly masculine terms to God but not strongly feminine ones, or vice versa. Scripture seems to characterize God as behaving in ways stereotypical of both genders, and scripture also seems to attribute mental states to God – maternal love and concern, for example – that are appropriately characterized by way of feminine imagery.Footnote 24 When theologians like Julian of Norwich apply strongly feminine terms to God, the characterizations usually resonate with rather than contradict what the Christian tradition seems to affirm about God's beliefs, desires, and behaviour towards creation; and likewise when strongly masculine terms are applied to God. Given all of this, there is good reason to think that either God has no strongly gendered extrinsic attributes at all or that God belongs to both genders.
To say that God belongs to both genders, however, is not yet to say that God belongs equally to both genders. Perhaps God belongs to both genders but nevertheless has a preponderance of masculine attributes and therefore counts as more masculine than feminine. As noted earlier, however, to say this is to fall foul of the equality thesis. Again: If every person of the trinity were more masculine than feminine, or if the divine nature were more masculine than feminine, or if the trinity as a whole were more masculine than feminine, then men as such would be greater image-bearers than women as such, contrary to the equality thesis. Thus, if gender is extrinsic, either God has no gender, or God belongs equally to both genders.
Implications for traditionalism
In the previous section, I defended the conclusion that God is masculine or feminine only if God is equally masculine and feminine. The only other premise in my main argument is the obvious truth that God is most accurately characterized as masculine only if God is masculine but not equally feminine. Thus, I conclude that it is not the case that God is most accurately characterized as masculine. By the same reasoning, mutatis mutandis, it follows that it is not the case that God is most accurately characterized as feminine. In this closing section I want to consider briefly the implications of this conclusion for the moral and liturgical propriety of characterizing God in exclusively or predominantly masculine terms.
The conclusion that God is not most accurately characterized as masculine poses a serious challenge to traditionalists who maintain that God ought to be characterized in predominantly masculine terms. The challenge is to explain why this should be so. The most natural defence would appeal to the idea that privileging masculine characterizations somehow gives us special insight into the divine nature that would be lost if feminine characterizations came to enjoy an equal place in our theology and liturgical practice. But it is hard to maintain this view while at the same time admitting that characterizing God in predominantly masculine terms is metaphysically misleading. The question for the traditionalist, then, comes to this: why think that masculine characterizations are to be preferred, given that preferring masculine characterizations is apt to mislead people about God's attributes?
Broadly speaking, I see only two responses to this question that have any hope of being viable. First, one might think that the risks (in terms of harm to human beings in general, or to the furtherance of what we reasonably take to be God's goals for creation) of de-privileging masculine characterizations of God outweigh the risks of continuing to privilege them. At best, I think, we have no reason to think that this claim is true; and feminists have, in effect, been arguing for quite some time now that the claim is resoundingly false. Second, one might think that, wholly independently of considerations about harm and (again) despite the fact that it would be metaphysically misleading, God simply wants to be characterized by human beings more in masculine terms than feminine ones. This seems to be the position of John Cooper (Reference Cooper1998), and perhaps there is some historical precedent for it as well.Footnote 25 In the next several paragraphs I will explain why I think his defence of that view is a failure.
Three chapters of Cooper's Our Father in Heaven are devoted to examining the ‘biblical pattern’ of representation of God and establishing the claim that the ‘only reasonable conclusion’ one might draw from that pattern is that ‘Scripture intends to portray God as a masculine person’ (ibid., 114). His argument for this conclusion rests mainly on two premises: (i) all proper names, basic titles, and common nouns that name God are masculine; and (ii) these forms of speech are to be privileged over mere ‘figures of speech’ in determining what the authors (human and divine) intend to convey about what sorts of gendered terms are appropriately applied to God. I won't contest the first premise; and I need not contest the part of the second premise that says that proper names, basic titles, and common nouns are to be privileged over mere figures of speech in determining what the human authors of scripture intend to convey about divine gender. The salient question is whether the absolute prevalence of masculine names, titles, and common nouns indicates anything about the divine will.
Cooper is explicit that it does indicate the divine will. After arguing (against some feminist authors) that the masculine portrayals of God in scripture were not necessary in order to accommodate the beliefs and values entrenched in ancient cultures, Cooper writes:
The fact that God could easily have chosen a gender-inclusive self-presentation … and did not do so is a strong reason for supposing that the exclusively masculine language for God in Scripture is not merely a temporary accommodation. In fact it is a good indication that God is not interested in promoting gender-inclusive language for himself. (ibid., 153, emphasis added)
At the same time, however, he affirms – emphatically, and multiple times – that ‘ontologically, God is genderless' (ibid., 169, n. 3). But he thinks that this fact is ultimately not relevant to the debate about how God is (morally and liturgically) appropriately characterized. Thus, after noting that the claim that God transcends gender is often taken to be a ‘key theological ground for the doctrine that God has no gender and thus for dealing with the Bible's masculine language for God’, he writes: ‘This move [is not consistent with Scripture and] in effect makes a theological-philosophical conclusion about God's nature more determinative of Christian language than the text and teaching of the Bible’ (ibid., 168).Footnote 26 He goes on later to say that the doctrine that God is beyond gender ‘is not stated in Scripture’ and that, although it ‘has a substantial biblical basis . . . its derivation involves some debatable cultural and philosophical assumptions about sex, gender, body, and spirit’ (ibid., 183). Thus, he finally concludes:
I strongly affirm the Christian doctrine that God himself is beyond gender. But I also conclude that because this doctrine is derived from Scripture and relies on extrabiblical assumptions, it ought not to be placed prior to or above Scripture in shaping the language of the Christian faith. (ibid., 184)
The upshot, then, is that scripture – the ‘biblical pattern’ – testifies against inclusivism; and the testimony of scripture trumps metaphysics as evidence about how God prefers to be characterized.Footnote 27
Many theologians will reject the claim that the testimony of scripture trumps metaphysics as evidence about how God prefers to be characterized. But that is not the route I wish to take. As a general rule, if my considered judgement were ever to be that scripture affirms p whereas my best theory in metaphysics affirms not-p, I would side with scripture and cast aside the metaphysical theory. On this much, I think that Cooper and I agree. But, as one might expect, I have yet to reach the verdict that scripture affirms a proposition that the overall best metaphysical theory denies. Nor have I ever reached the unfortunate conclusion – one that Cooper himself seems to have reached – that there is some proposition that our overall best metaphysical theory affirms but God would prefer for us to ignore in our characterizations of God. The reason, of course, is that my metaphysical theorizing and my interpretation of scripture proceed in dialogue with one another. Theory-building is a holistic process.
The problem with Cooper's argument is that he fails to appreciate the way in which philosophical reasoning is already involved at the level of figuring out what it means to take sides with scripture and to let it serve as evidence for a conclusion. Let us grant that ‘the biblical pattern’ of reference to God is one which portrays God overwhelmingly in masculine terms. Let us also grant (as seems plausible) that this pattern is prima facie evidence that God wants to be portrayed in overwhelmingly masculine terms. Still, just like the doctrine that God is beyond gender, the thesis that God wants to be portrayed in overwhelmingly masculine terms is not at all stated in scripture, but is instead merely derived from scripture with the help of additional, extrabiblical assumptions.
The question, then, is whether our total evidence – scripture, metaphysics, and whatever other extrabiblical assumptions we might appropriately appeal to in defending our views about how to characterize God – supports the thesis that God prefers to be characterized in predominantly masculine terms despite the fact that it is metaphysically misleading. Against Cooper's view, we might first note that the biblical pattern of reference to God is one that applies exclusively Aramaic, Hebrew, and Greek terms to God; but, of course, nobody would claim that this pattern is evidence that God wants to be talked about exclusively in Aramaic, Hebrew, and Greek. The reason, of course, is that we know independently – i.e. at the level of philosophical presupposition – that there is no reason why God should want this. There is nothing about God or about the words of Aramaic, Hebrew, and Greek that make them more apt than (say) English words for describing God; and there is no other reason that we can see why God should want to be described only by way of such words. So likewise, we might think, with gendered words. Moreover, we know from scripture that God cares about the truth; God is no deceiver. This, it seems, speaks strongly in favour of a view about how God prefers to be characterized that attributes to God an interest in characterizations that are not metaphysically misleading. (This is why I said earlier that the best route for a traditionalist to take is to oppose my arguments for the conclusion that masculine characterizations of God are not more accurate than feminine ones.)
Cooper maintains, in effect, that the moral and liturgical aptness of feminine characterizations of God depends on the proportion of feminine names, basic titles, and common nouns that are applied to God in scripture. He thinks that the fact that all of those are masculine is decisive evidence that God has no interest in gender-inclusivism (ibid., 114, 153). But, again, this claim is no less an extrabiblical philosophical thesis than its negation, and it is no less an extrabiblical philosophical thesis than those often invoked to support inclusivism (such as the claim that if feminine characterizations are no less accurate than masculine ones, then they are likewise no less morally or liturgically appropriate than masculine characterizations). Cooper's ‘biblical pattern’ is just that – a pattern, and one that is open to interpretation in light of our best overall evidence, philosophical presuppositions included.
I acknowledge that nothing I have said so far demonstrates that Cooper's conclusion is false. I have only shown that his argument for that conclusion fails. I also acknowledge that there is something quite natural in taking not just the overall biblical pattern of divine characterization but Jesus' own practice (as reported in scripture) as indicative of a preference on God's part to be characterized in predominantly masculine ways. It is tempting, indeed, to think that sticking with the traditional masculine modes of characterizing God is the safe option – safe because it is non-revisionary and is backed by the weight of centuries of practice. I grant that all of this constitutes some reason (partly pragmatic, partly epistemic) for affirming the thesis that God, for unknown reasons, simply prefers to be characterized in masculine terms.
Still, we must view these considerations in light of the very persuasive arguments for the conclusion that much harm has come from treating masculine characterizations of God as mandatory, and also in light of the very real possibility that implicit bias against women plays a non-trivial role in shaping our judgements about how the biblical pattern of characterization, including what we learn in scripture about Jesus' own practice, is naturally interpreted.Footnote 28 We must also recognize that there is nothing at all safe about insisting on a pattern of characterizing God that is ultimately metaphysically misleading (as I think we have very good reason to believe that the traditional pattern is) unless we also have very good reason for thinking that God simply prefers that pattern of characterization. To do that is to risk idolatry. Indeed, it is to risk other sins as well, given that the pattern in question has contributed to the oppression of half the human race, and given that, if we know anything at all from scripture about divine preferences, we know that God cares about justice. Much rides, then, on the claim that God prefers to be portrayed as masculine despite its being metaphysically misleading. As I have argued in this last section of the article, Cooper's case for that claim is a failure; and, for my part, I can see no other convincing reason to endorse it.
Typical defences of inclusivism assume either that the concept of God is a human invention or that divine transcendence precludes the literal application to God of just about any human term, thus making both masculine and feminine characterizations equally appropriate. My goal in this article has been to offer metaphysical underpinning for inclusivism that rests on neither of these assumptions, nor on controversial assumptions about the nature of gender. I have argued that if God is masculine or feminine at all, then God is equally masculine and feminine. I have also provided some reason to think that if this conclusion is true, then it is morally and liturgically appropriate to characterize God in feminine terms – the most salient reason just being that, judging from how God is portrayed in scripture, God seems to prefer not to be characterized in ways that are known to be misleading.Footnote 29
