In contemporary debates about ‘idolatry’, some thinkers accuse their co-religionists of idolatry. This fact is not new,Footnote 1 though it is no less surprising for its vintage, since it seems a reasonable assumption about members of a common religion that they worship a common deity. However, Halbertal and Margalit’s (Reference Halbertal, Margalit and In Goldblum1992) hallmark treatment of the topic shows that ‘idolatry’ admits as many meanings as there are critics of idolatry. In particular, they explain how idolatry can be defined against bad worshippers of God, both ‘outsiders’ and even ‘insiders’ to a religion. Some contemporary thinkers, distinguishing between ‘worshipping the wrong god’ and ‘worshipping God wrongly’, reserve ‘idolatry’ in its strictest sense for the first category. A variation of the second error, making factual mistakes or having wrong ideas about God, is often called ‘conceptual idolatry’.
This trend is a mistake. The distinction between ‘idolatry’ and ‘conceptual idolatry’ misunderstands the arguments from phenomenological and grammatical traditions that see conceptual interpretations of God-talk as idolatrous. The basic problem is not one of referring to ‘the right’ God; rather, phenomenologists and grammatical thinkers ask the more fundamental question of what it means for us to be able to think and speak about God in the first place. Conceptual God-talk constitutes idolatry because it undermines our ability to recognize God as God at all. This new claim might reasonably be called idolatry simpliciter, though this framing of idolatry comes with its own problems. I first survey phenomenological and grammatical accounts of the relation between concepts and idolatry, then examine analytic misunderstandings of the problem. I conclude by discussing how all parties, especially analytic-sympathetic thinkers, can move forward with renewed grasp on the question of idolatry.
Idolatry in the phenomenological tradition
Martin Heidegger (Reference Heidegger, Heidegger and Stambaugh1969, 56) posed the question: ‘how does the deity enter into philosophy, not just modern philosophy, but philosophy as such?’ He argues that such entry results from a basic confusion: being should be understood as the ground of entities, but metaphysicians seek a further ground for being, positing god as a self-causing entity on which being depends, obscuring the proper relation between being and entities. This is what Heidegger calls the ‘onto-theo-logical constitution of metaphysics’. The ontotheological god is one to whom ‘man [sic] can neither pray nor sacrifice. Man cannot fall to his knees in awe before the causa sui, nor can make music or dance before this God’ (Reference Heidegger, Heidegger and Stambaugh1969, 72). Such a god is introduced as a solution to a philosophical problem and remains limited to the role it plays in such a system: a metaphysical deus ex machina.
Heidegger inaugurated a tradition of anti-theological phenomenologists who argue that, in the words of Dominique Janicaud (Reference Janicaud and Prusak2000, 99), ‘phenomenology and theology make two’: the transcendent God cannot appear in the immanentist phenomenological field. Janicaud’s case relies, in part, on the argument that introducing theology into phenomenology constitutes returning to ‘metaphysics’, discussing something ‘behind’ or ‘beyond’ the sheer appearing of phenomena, some vague and viciously mysterious ‘thing in itself’. For present purposes, it suffices to say that criticism of metaphysics – a characteristic of phenomenology at large, but especially poignant in theological contexts – motivates the connection between concepts and idolatry. Arguments or theories that strive to get ‘behind’ phenomena or leave the mode of phenomenal appearing aside are seen as metaphysical. This condition portends to secure phenomenologists against abandoning everyday experience, attempting to systematize and organize it into some ‘more fundamental’ knowledge. Janicaud’s opponents respond that not only is he wrong, but that his methodological atheism begs the question of atheism in substance. Steven DeLay (Reference DeLay2022, 160) observes:
At bottom, [phenomenology that divides itself methodologically from theology] presumes to be able to know that, even were God to exist, God nevertheless would be unable to reveal himself as and for himself – God’s appearing is in effect ruled out a priori. To argue that any thought purporting to determine God is thereby illegitimate is just to say God cannot reveal himself determinately. But what evidence is there for that?
For his critics, Janicaud engages in metaphysics, since he imposes external criteria on appearances by arguing, a priori, that God cannot appear. Both parties reject metaphysics, but their difference about what metaphysics entails means that each accuses the other of illicit metaphysical speculation.
Those who claim that God can appear must explain this appearance’s possibility. This is a core project of Jean-Luc Marion, who in The Idol and Distance (Reference Marion and Carlson2001, 6) distinguishes the idol, which ‘reflects back to us, in the face of a god, our own experience of the divine’, from the icon, the ‘[v]isibility of the invisible, a visibility where the invisible gives itself to be seen as such’ (7). As he explains in God without Being:
The concept consigns to a sign what at first the mind grasps with it (concipere, capere); but such a grasp is measured not so much by the amplitude of the divine as by the scope of a capacitas, which can fix the divine in a specific concept only at the moment when a conception of the divine fills it, hence appeases, stops, and freezes it (Marion Reference Marion and Carlson2012, 16).
The idol is reducible to the concepts we bring to it, mirroring them back to us, deceiving us in making us think that we are discovering something new when we are simply projecting our preconceptions. The icon queries our concepts in its very appearing, not only by showing itself irreducible to them, but also by judging us and our concepts, opening us to a perspective other than our own.
Marion’s quest to think God without ontotheology, among other considerations, motivates him to develop the concept of ‘saturated phenomena’: certain phenomena – the event, the idol, the flesh, and the face/icon – thoroughly exceed our concepts and are ‘saturated’ with givenness. Marion classes revelation, the fifth type of saturated phenomenon, as the ‘paradox to the second degree’ and ‘saturation of saturation’. To avoid the illusion that language about God or any creaturely concept could encompass God’s self-revelation, Marion (Reference Marion, Caputo and Scanlon1999) strives to find a third way between cataphatic and apophatic God-talk: not naming God’s essence nor refusing to name God at all, but naming God as source of a given, as principle of a revelation that challenges the suitability of creaturely concepts. Marion insists that his determination of revelation as a saturated phenomenon is phenomenological, not theological (see Westphal Reference Westphal2006). The upshot is that ‘if God manifests (or manifested) himself, he will make use of a paradox in the second degree’ (Reference Marion and Kosky2002, 367n90). Marion, as a phenomenologist, ‘make[s] decisions only about the type of phenomenality which would render this phenomenon [i.e., God] thinkable’ (Reference Marion, Caputo and Scanlon1999, 39). Whether God is in fact given remains for theologians to answer. But if God is given at all, to anyone, it must be under the conditions of saturated phenomenality.
Marion’s account contains two key claims that other phenomenologists share. First, God appears as irreducible to conceptual analysis. Richard Kearney, for example, claims that God’s Kingdom appears only through certain ways of seeing and relating to ordinary persons – through love and justice enacted for and with other persons,Footnote 2 through our everyday acts of creating and making,Footnote 3 and other means. Since these activities are how God appears, we must make and work for the Kingdom: ‘[w]hen it comes to divinity, poiesis, not theoria, has the last word’ (Kearney Reference Kearney, Kearney and Clemente2018, 27). A necessary consequence of Kearney’s Kingdom-based phenomenology is that a God who does not appear in the world is not at all, since God cannot ‘be’ except as God appears in the being of the world (Kearney Reference Kearney and Manoussakis2006, 45). The faithful believe not in a God who is, but a God who promises that God will come.Footnote 4 Thus, Kearney ultimately advocates for an ‘anatheism’ – an after-theism (see Kearney Reference Kearney2010) – of a God who may be (see Kearney Reference Kearney2001), since there is no guarantee that the Kingdom will appear. John Caputo, like Marion, overcomes ontotheology through a phenomenological turn to the concept of ‘experience’. Just as experience involves ‘risk[ing] going where we cannot go, … hav[ing] the nerve to step where angels fear to tread, precisely where taking another step farther is impossible’ (Caputo Reference Caputo and Wrathall2002, 126), so ‘this sort of limit-experience … gives the name of God meaning, what we might call its phenomenological content, which is in the truest sense of the word experiential’ (Caputo Reference Caputo and Wrathall2002, 133), since God is that by which the impossible becomes possible.
Second, any claim about or attitude towards God that impedes our ability to experience God as beyond our concepts impedes our ability to experience, think about, or speak of God at all, thus constituting idolatry. Kearney (Reference Kearney2010, 175) is concerned to avoid the ontotheological ‘inherent temptation to violently impose its own version of the Absolute on others’, rejecting as idolatrous any ‘injustice committed in the name of God [which] is the greatest perversion of all’ (Kearney and Zimemrman Reference Kearney and Zimmerman2016, 39). He goes so far as to claim that
where Catholicism offends love and justice, I prefer to call myself a Judeo-Christian theist; and where this tradition so offends, I prefer to call myself religious in the sense of seeking God in a way that neither excludes other religions nor purports to possess the final truth. And where the religious so offends, I would call myself a seeker of love and justice tout court (Kearney Reference Kearney2001, 5–6).
Conceptual language that claims to describe a God who ‘is’ covers over the fact that God ‘is’ not, obscuring our need to make God along with God’s need to make us. This is especially the case when language about God ‘offends love and justice’, blocking our ability to engage in the acts through which God appears. Caputo agrees with the charge of idolatry against non-phenomenological approaches to God, concluding that
Orthodoxy is idolatry if it means holding the ‘correct opinions about God’ … but not if it means holding faith in the right way, that is, not holding it at all but being held by God, in love and service. Theology is idolatry if it means what we say about God instead of letting ourselves be addressed by what God has to say to us. Faith is idolatrous if it is rigidly self-certain but not if it is softened in the waters of doubt (Caputo Reference Caputo2007, 131).
Since God appears only in border-crossing ‘experience’ wherein the impossible is made possible, idolatry arises when believers take God to provide answers rather than questions, certainty rather than doubt, security rather than the ultimate invitation to go beyond: all of which are symptoms of conceptual approaches to God.
Phenomenologists give such narrow criteria for what constitutes encountering God because anything other than God’s appearing as beyond our concepts is indistinguishable from encountering the ordinary world. This is why Marion uses ‘love’ and ‘given’, implying a logic of appearance from beyond our ken, to describe how God arises to our thinking and speaking. Caputo turns to ‘experience’ and ‘event’, since both contain the logic of crossing boundaries and creating possibilities which did not exist before. These words are consciously self-undermining, guaranteeing our openness to further appearance, resisting our tendency to allow pre-conceived notions to determine thinking or talking about God. Kearney, in privileging poiesis over theoria, is more interested in highlighting practices (ministry to the poor, artistic creation, and others) which constitute both our making God and God’s making us. For phenomenological defenders of theology, conceptual analysis undermines our capacity to recognize a God irreducible to concepts, leading these figures to call such analysis ‘idolatrous’.
Idolatry in the grammatical tradition
Another group of thinkers critical of idolatry are the ‘grammatical Thomists’, readers of Thomas Aquinas inflected by ordinary language philosophy. For Brian Davies, as for other grammatical Thomists, idolatry is ‘“the failure to recognise the difference between God and creatures” … as if (in the language of Matthew Arnold), he lived on the next street’ (Davies Reference Davies, McCosker and Turner2016, 101). Thus, any language that makes God out to be creaturely in any respect undermines our capacity to speak of God at all and is fitly named ‘idolatry’. Herbert McCabe, for example, accuses several figures or movements as ‘idolatrous’, from process theology (McCabe Reference McCabe1987, 1) to free-will theodicians (McCabe Reference McCabe1987, 11; see also Davies Reference Davies2006, c. 5) to anyone who makes God ‘a member of the universe, subject to change and even disappointment and suffering’ (McCabe Reference McCabe1987, 18). Even anyone who should ‘suggest that a man ought to have the kind of certainty about his wife’s honesty, or about various press reports, that he has about the creed’ is not only ‘absurd’, but guilty of a ‘kind of idolatry’ (McCabe Reference McCabe and Davies2007, 13).
What these idolators share is their view of God as ‘a Top Person in the universe who issues arbitrary decrees for the rest of the persons and enforces them because he is the most powerful being around’ (McCabe Reference McCabe1987, 7), akin to a magnified creature. Certain faulty uses of the word ‘god’ cannot even refer to the true God at all. As McCabe insists:
We can use the word ‘God’ correctly or incorrectly, but the criterion for correct and incorrect use is not something we know about the nature of God. It is something that is thought to be true of our world. In other words, God’s being creator of the world is what gives us our meaning for the word ‘God’ (McCabe Reference McCabe and Davies2013, 386).Footnote 5
For McCabe, the meaning of ‘God’ is determined by God’s status as creator, as the answer to the question, ‘Why is there something rather than nothing?’ (McCabe Reference McCabe1987, 6; see Davies Reference Davies2003). Following McCabe, grammatical Thomists explicitly answer the question of how ‘God’ is apt to thought by positing ‘God’ as the end of inquiry to the question of why something exists rather than nothing (see McCabe Reference McCabe1987, 3–6); Davies Reference Davies2003), or to describe theology as the process of explicating the grammar of calling God beginning and end of all things (see Burrell Reference Burrell1979, c. 2; Mulhall Reference Mulhall2015, 47–54).
In this tradition, David Burrell links idolatry explicitly with the charge of ontotheology. Citing Kathryn Tanner (Reference Tanner1988) and Sara Grant (Reference Grant2002), Burrell argues that Christian doctrine requires a ‘non-competitive’ or ‘non-dual’ relationship between divine and creaturely action. A Christian doctrine of creation – and a Jewish or Muslim one too (see Burrell Reference Burrell1986) – demands that God be unlike any creature, since God creates the universe ex nihilo (‘from nothing’). God, therefore, cannot be subject to creaturely categories. The ontotheologian, on the other hand, treats God like an entity (see Burrell Reference Burrell2008). McCabe, Burrell, and Davies all agree that ‘idolators’ misunderstand how God shows up for us at all. For God to be an answer to the question of why there is something rather than nothing, God cannot have any creaturely ontological status: God cannot be a ‘something’. Accordingly, the ‘god’ of ontotheology is not the true God, and worship of or prayer to such a god is idolatrous.
If God is not a thing among things, then language about God cannot be absolutely continuous with language about creatures, since language ceases to be at home when moving from creature to Creator. We must acknowledge that our application of concepts to God is analogical, and in some cases ‘nonsensical’: conceptual God-talk can scarcely be conceptual at all, at least not in the usual sense. McCabe (Reference McCabe, Aquinas and McCabe2006), Burrell (Reference Burrell1973, Reference Burrell1979), Davies (Reference Davies1993), Nicholas Lash (Reference Lash, Hebblethwaite and Sutherland1982), Stephen Mulhall (Reference Mulhall2015), and other ‘grammatical Thomists’ offer unsystematic readings of Thomas Aquinas’s writings on analogy, arguing that the term picks out where language operates beyond its usual context rather than offering a systematic formula to relate our talk of creatures to our talk of God. We must be aware when the words we use have a radically different meaning outside their standard context. Thus, we often use words whose meaning we cannot fully articulate or understand. We know that God is good and wise and just, but not how. Theology carefully attends to the ways in which God-talk differs from other forms of discourse: it ‘is not concerned with trying to say what God is but in trying to stop us talking nonsense, trying to stop people making God in their own images, to stop us from mistaking our concepts and images and words for the mystery towards which they point’ (McCabe Reference McCabe and Davies2002, 216). By grammatical lights, analysis of God-talk is successful only when it self-consciously fails to give us handholds into God’s nature.
We also see why the distinction between ‘idolatry’ simpliciter and ‘conceptual idolatry’ cannot hold: any attempt to ‘identify’ God at all (except in the person of Jesus Christ) sees God as an object and ‘misrepresents’ God’s otherness from creation, obscuring God’s status as creator, and thus constituting idolatry. We cannot distinguish between worshipping the ‘wrong’ God and over-anthropomorphizing God or mis-predicating about God, since any predication or anthropomorphism about God whatsoever undercuts the very condition under which God is apt to thought and speech at all – God’s radical difference from creatures.
The grammatical Thomist case rests on philosophical views about the relation between grammar and metaphysics; this is what allows them to conclude that any inference from creaturely concepts to God’s immanent life constitutes idolatry. Mulhall (Reference Mulhall2015, 2–3), for example, follows Ludwig Wittgenstein ((Reference Wittgenstein, Anscombe, Hacker and Schulte2009), §371) in insisting that ‘Essence is expressed in grammar’ and that ‘[g]rammar tells what kind of object anything is (Theology as grammar)’. Against Francesca Aran Murphy (Reference Murphy2007, 89), who has criticized grammatical Thomism for ‘translating metaphysical concerns into concerns about the logic of religious language’, Mulhall responds that there is no ‘translating’ going on, since there is nothing for metaphysical concerns about God’s being to mean other than statements about the logic of religious language (Mulhall Reference Mulhall2015, 3–4). We can call grammatical Thomists critical of metaphysics in the sense that they are critical of any metaphysics that looks beyond the analysis of language.
Difference in the extent to which certain figures take this principle constitutes a divergence in grammatical Thomism (see Kerr Reference Kerr2016). Lash (Reference Lash1996, 134), for example, states that ‘[i]dolatry is a matter of getting the reference wrong: of taking that to be God which is not God’, and McCabe (Reference McCabe and Davies2007, 2) agrees that religious truth is a matter of propositional correspondence, albeit to facts about the world. Analogy remains fundamentally apophatic, securing theology against the temptation of seeing anything lower than God, even our concepts of God, as ultimate – in short, a series of ‘protocols against idolatry’ (Lash Reference Lash1988, 257). Religious language, though, is still fundamentally referential. But for Burrell and Mulhall, referential understandings of God-talk see metaphysics as something beyond analyzing grammar, seeking actively to understand the relation of God to the world. This view of religious language has some extreme entailments. If religious language is not referential, then no entity merits ultimate value, nor could we worship any ‘god’ but strange ones. Thus, we must attribute no ultimate value to anything whatsoever. This is how Mulhall describes religious belief and practice in both post-Kantian ‘continental’ and Wittgensteinian contexts (see Mulhall Reference Mulhall, Rosen and Leiter2007, Reference Mulhall, Kuusela and McGinn2011; Wolfe Reference Wolfe2017). Because we cannot relate to God as an object, we must understand religious forms of life as describing our relation to finite things. This restriction on ascribing ultimate value to finite things must include any conception or predication we claim to have or make of the divine.
Although phenomenological and grammatical thinkers agree that ontotheological conceptions of God are idolatrous, their conceptions of what it means rightly to identify God are very different. For phenomenologists, God must appear as God, and argumentative or rational probing into God’s existence constrains God to our creaturely ideas. For grammatical thinkers, God cannot appear at all at risk of being only a maximized creature; the only way to understand God is as a matter of radically questioning the origin or foundation of things as we find them, resulting in a transformed account of language and a new form of life in the face of divine mystery. Thus, while phenomenologists focus on God’s phenomenal appearing and resist conceptual analysis and argument, the grammatical approach sees conceptual analysis (albeit, only in an apophatic register) as the heart of theology. But they agree that certain conceptual approaches to God-talk undermine the conditions under which the word ‘God’ becomes meaningful at all, and thus can be called idolatrous.
Analytic (mis)interpretations of idolatry
Analytic thinkers recognize that the charge of ‘idolatry’ arises from schools of thought with different presuppositions than their own, so their first step tends to be explaining idolatry in their own words, leading to the ‘idolatry’–‘conceptual idolatry’ distinction. This section evaluates several analytic formulations of idolatry and arguments against conceptual language constituting idolatry. Because each reformulation of the problem of idolatry may be held by various thinkers, expressed in different words, I focus not on particular figures, but on characterizations of idolatry and responses to them. I conclude that no attempt – with the exception of Mark Johnston’s ‘foundational religious experience’ account – encapsulates the logic of ‘idolatry’ in phenomenology or grammatical thought.
Idolatry as mistake about god
A common interpretation of the charge of idolatry is that it picks out an error in someone’s notion of God. For example, the reason (on this view) that Burrell charges ontotheologians with idolatry is because God is, in fact, beyond being. Richard Cross (Reference Cross2008) lays out a problem facing those who accuse analytic-sympathetic thinkers of idolatry on the basis of ontological mischaracterization: false predication does not entail misidentification. If an ontotheologian were to claim that ‘God is not “beyond being”’, this might reasonably be interpreted as a de re claim: that which the ‘idolator’ and the critic of idolatry commonly call ‘God’ is not, in fact, beyond being. However, for the ‘idolator’ to be guilty of idolatry, they must mean the statement de dicto: whatever ‘God’ happens to pick out as a referent is not beyond being. On the latter interpretation, they would foreclose the possibility of ‘God’ picking out a referent that is beyond being, and they would fail to recognize such a referent as genuinely God. The critic must prove that the ‘idolator’, in making the former claim, also means the latter: a challenge, since the de dicto statement in no way entails the de re one.
Cross’s conclusion, that false predication cannot alone constitute misidentification, is correct. However, the previous section shows why the phenomenologist and grammaticist still have a case to make. The problem with ‘idolatry’ is not merely false predication, but the kind of false predication that makes it impossible for the ‘idolator’ to have recognized God in the first place. It is a severe category error, as if someone were to claim that ‘the number two is hungry’ – a claim that might license us to think that our interlocutor did not understand at least one of the concepts of ‘number’, ‘two’, and ‘hungry’, or perhaps even ‘is’. To claim that ‘God is not beyond being’ is to claim that the concept ‘being’ can adequately characterize God, making it impossible for God to have appeared as beyond all concepts, as totally unlike any creature. The phenomenologist and grammaticist offer criteria by which we might recognize the true God, and an ontotheological (for example) conception of God runs counter to all of them. This charge might reasonably be called ‘idolatry’, since no God-talk or worship of the ‘object’ of God-talk could be directed to the true God.
Conceptual idolatry as mistake about god
A related objection is that conceptual idolatry, not idolatry as such, constitutes ‘a baroque formulation of … the problem of being wrong about God’ (Wood Reference Wood2021, 117). Here, ‘conceptual idolatry’ – the problem of being wrong about God – is distinguished from idolatry proper – the problem of worshipping or praying to the wrong God: as Wood (Reference Wood2021, 118) clarifies, conceptual idolatry ‘is merely a cognitive error, the error of being mistaken about God and about our own limitations. In that case, it would not really be a form of idolatry at all.’ Similarly, David Decosimo (Reference Decosimo2022, 755) argues that analytic theologians ‘may do well to hear charges of “idolatry” as charges of heresy – of damaging and serious false teaching’. This is a weaker standard than Cross’s, since the critic of idolatry need not show that a conceptual idolator is an idolator tout court, since they are distinct accusations.
However, in light of the above discussion, we can conclude that the phenomenological and grammatical traditions are justified in seeing ‘conceptual idolatry’ as idolatry proper by their own lights. The charge of idolatry is not reducible to accusations of error or anthropomorphism, even if every tradition agrees that both are problematic. Conceptual error constitutes idolatry not necessarily because it is error, but because it is conceptual – at least, conceptual in the wrong way. Phenomenologists can only acknowledge God based on God’s appearing as slipping past our conceptual nets. Descriptions of God that claim human knowledge over God constitute decisive evidence that it was not God that appeared. Grammaticists worry that seeing God through creaturely categories undermines our recognition that we speak of ‘God’ only where sense-making ends; thus, our conceptual God-talk must deconstruct, rather than take as given, our concepts as applied to God.
Conceptual idolatry and modelling god
Sometimes, analytic theologians understand conceptual idolatry as a tendency to ‘substitute human conceptual frameworks for reality’ (Adams Reference Adams2014, 9), to ‘direct one’s thought towards a simulacrum of the phenomenon about which one aims to think or theorise rather than towards the phenomenon itself’ (Rea Reference Rea2020, 39), or to ‘worship[] one’s model of God rather than the true God’ (Crisp et al. Reference Crisp, Arcadi and Wessling2019, 17). Oliver Crisp, James Arcadi, and Jordan Wessling (Crisp et al. Reference Crisp, Arcadi and Wessling2019, 18) worry that ‘fear of idolatry can drive the theologian to avoid thinking about and describing God altogether, lest that description fail to map onto divine reality with full accuracy. And this may well be a form of idolatry itself, if God has revealed himself in order to be known and communed with.’ And analytic philosophers such as Jonathan Jacobs (Reference Jacobs and Kvanvig2015) offer models of God which allow and even entail apophatic restrictions on theological speech. It is not only analytic thinkers who use ‘idolatry’ in this sense: Karen Kilby (Reference Kilby2010, 66), for example, characterizes her worry about idolatry in the context of Trinitarian theology as ‘the possibility of being so robust, so confident that we know what we are talking about when we talk about the Trinity, that we are in fact projecting our most pleasing ideas onto God and making those the object of our worship’.
There are a few responses the phenomenologist or grammaticist can make. To some extent, this interpretation constitutes a variation of a generally epistemological interpretation of conceptual idolatry; thus, their responses in the above section should carry over here. The concern about modelling is different from error, since a model could be used, as Wood (Reference Wood2016, 57) notes, ‘to describe that which God is not, in order to see that God transcends all creaturely limitations’. However, not all models violate phenomenological and grammatical rules of God-talk: only those that purport to be of God. Wood points out that some models of maximal creaturely perfection can be used to show the difference between God and even the greatest creature; here, the critic of idolatry can agree. However, nobody takes such models to be of God. It is only models of God that could constitute idols.
As for Crisp et al., very few critics of idolatry seek to do away with God-talk completely, only to condition it on God’s terms. Marion and McCabe are two of the most sophisticated writers on theological language in the twentieth century precisely because they refuse to ‘avoid thinking about and describing God altogether’; they seek to speak in a way that respects how God reveals Godself ‘in order to be known and communed with’. Grammatical and phenomenological understandings of what conditions God sets on God’s self-revelation are responsive to a strand of theology exemplified by Karl Barth and Eberhard Jüngel. Against various forms of apophaticism, natural theology, and accounts of ‘religion’, Barth argues that since we cannot access God by our own power, God can be no object for our activities, and all human activities are fundamentally idolatrous: ‘[t]he power to be in the world and a man [sic], as man’s own power, is identical with the power to devise and form gods’ (Barth Reference Barth, Bromiley, Torrance, Thomson and Knight1956, 324). It is God who must overcome this gap. The only way to avoid idolatry is submission to God’s overcoming of our inability to worship God rightly, by having a theology and ethic totally centred on the person of Jesus Christ. Jüngel (Reference Jüngel and Gruder1983, 260) develops this Barthian theme in his claim that
whereas [the apophatic tradition] understood the revelation of God solely as the enabling of proper talk about the still greater hiddenness and mysterious superiority of God over the world, now [in Jüngel’s own thinking] God is to be grasped as a mystery which is communicable in and of itself in language.
Jüngel, like Barth, sees this communicability Christologically – Christ’s coming does not only declare God’s presence with creatures, but enacts it, such that God is perfectly communicable to creatures in, and only in, the person of Jesus Christ.
Responding to this strand of critique, Marion and McCabe do not appeal solely to God’s appearance in abstract terms alone: both adopt a Christological focus in their accounts of ‘revelation’. For McCabe (Reference McCabe1987, 177), it is precisely because ‘[t]he theologian uses a word by stretching it to breaking point’, because of apophatic constraints on God-talk, that ‘[o]ur language does not encompass but simply strains towards the mystery that we encounter in Christ’. God communicates to creatures completely in Christ not despite, but because of, McCabe’s strident apophaticism. Marion too takes his philosophical background to strengthen, rather than diminish, the Christological focus of his theology. After describing a phenomenon as that which ‘shows itself in itself and through itself only as much as it gives itself in and through itself’ (Marion Reference Marion and Lewis2016, 76), he offers his phenomenological interpretation of Christ’s incarnation:
what phenomenon has ever, without remainder and without reserve, respected the phenomenological program? What phenomenon has ever accomplished it ‘to the end, eis telos’? The demand of Christian theology here takes on the entirety of its immense claim: only the one who ‘loved his own until the end, eis telos’ (Jn. 13:1), to the point of saying in truth, ‘it is finished, tetelestai’ (Jn. 19:30), manifested, uncovered the phenomenon in itself and from itself. This phenomenon shows itself absolutely because he, and he alone, gives himself absolutely. Christ not only offers himself to be seen as a phenomenon among others, fulfilling the program of phenomenology; he fulfills it for the first and only time in actuality, in his actions, and becomes the phenomenon of all phenomena. He, the total and saturated agent of the putting into evidence of the absolute unseen, of the mystērion of God hidden since the origin of the ages, he who ‘was manifested in the flesh, ephanerōthē en sarki’ (1 Tim. 3:16), has at the same time spread the light everywhere around him. (Marion Reference Marion and Lewis2016, 76–77).
For Marion, Christ’s revealing the fullness of God only makes sense against a background understanding of the phenomenon as that which gives itself fully.
Thus, neither McCabe nor Marion sees their apophaticism as undermining the Christian’s ability to relate to God, or God’s ability to reveal Godself in God’s own terms. These terms, though, are not primarily linguistic, but take as their medium the life and mission of Christ, as well as this-worldly practices (such as receiving the Eucharist) instituted by Christ. Such a Christocentric turn only works within a broader skepticism of language’s ability to carve God at the joints, within Marion and McCabe’s insistence that we only recognize God as God when God is found to transgress all our concepts. These moves should help assuage Crisp et al.’s concerns about barring God’s self-revelation. All parties agree that they must interpret God-talk on the terms of God’s own revelation, while disagreeing about what those terms are.
Idolatry and Kant
Wood (Reference Wood2021, 110) argues that the disagreement about idolatry ‘only makes sense within the context of contemporary phenomenological [and, for our purposes, grammatical]Footnote 6 thought and its Kantian inheritance’. Since, for Immanuel Kant, ‘concepts are like filters that stand between our conscious experience and the sensory manifold that constitutes the phenomenal world’, and God is a noumenal object that ‘cannot pass through our conceptual filters’, God ‘cannot be [an] object[] of experience or knowledge’ (Wood Reference Wood2021, 119). Wood (Reference Wood2021, 119) claims that Marion goes further, arguing that by ‘trying to think about God using concepts that cannot apply to God, we are trying to turn God into the kind of thing we can think about, which is an attempt to constrain and confine God’. Wood disagrees with Kant and Marion on the noumena–phenomena divide and the conceptual constitution of objects: this rejection of Kantian epistemology allows Wood to see God-talk as referential and descriptive of Godself in a way Marion cannot.
Wood’s interpretation of Marion, however, is mistaken, since Marion challenges Kant on phenomenological grounds.Footnote 7 When it comes to God, Marion’s life’s work is to show that God can be given in experience, if we do not constrain God with creaturely concepts. Conceptual idolatry is idolatry because God can only be given by being experienced as beyond creaturely concepts. Wood’s (Reference Wood2021, 121) formulation of the phenomenological question – ‘How can we represent God to ourselves, when God is unrepresentable by human concepts?’ – considers the subject in an active and constructive role. The phenomenologist Emmanuel Falque (Reference Falque and Hackett2015, 7), a student of Marion, asks a different question: ‘How and in what way is God given to be seen today?’ Falque describes God’s appearing, without mentioning the subject or self-representation at all. Stricter Kantians, like Janicaud, cannot accept Falque’s framing of the question precisely for the reasons Wood offers. Should Falque frame the question as one of a ‘subject’s’ constitution of an ‘object’, he would beg the question against certain phenomenological thinkers, including Marion and himself! Between Wood’s and Falque’s questions stands the phenomenologist’s rejection of Kant’s strict divide between noumena and phenomena and Kant’s account of the conceptual construction of objects. Wood’s suggestion that he can avoid phenomenological critiques of idolatry by rejecting Kant’s epistemology cannot stand, since these very phenomenologists reject the same Kantian theses as Wood.
Idolatry and foundational religious experience: the case of Mark Johnston
The closest anyone in the analytic tradition comes to seeing idolatry along similar lines as phenomenological and grammatical accounts is Mark Johnston. Johnston (Reference Johnston2009, 35) even calls his method ‘phenomenological’, since ‘[w]e take the foundational religious experiences as they present to their subjects. We then look at the implied character of the putative spiritual beings who ostensibly appear in these experiences.’ He begins (10) by asking whether all uses of the word ‘god’ refer to the same thing: indeed, he goes so far to say that ‘[t]he best thing a believer can say in response to the question “Do you believe in God?” is “I can only hope that I do. I can only hope that I actually stand in a tradition in which God has genuinely revealed himself.”’ He then argues that all ‘supernaturalist’ religions are idolatrous, since they undermine religion’s grounding in experience. Like the phenomenologists, Johnston explicitly affirms (Reference Johnston2009, 68) that ‘the beneficiary of a revelation must see or hear the events as the Highest One manifesting himself. That content must be internal to the experience and not just to the subsequent beliefs that it prompts.’ Johnston offers a single criterion of recognizing God and argues that certain forms of religion are idolatrous, since they undermine our ability to identify God by that criterion.
Johnston, though, is no mere subjectivist, even if he rejects the Christocentrism of Marion and McCabe. He turns to process thought to argue that ‘The Highest One = the outpouring of Existence Itself by way of its exemplification in ordinary existents for the purpose of the self-disclosure of Existence Itself’ (Reference Johnston2009, 116). Thus, a worshipful attitude towards creation itself does not constitute idolatry, since creation is The Highest One’s outpouring of Itself through history and nature. It is supernaturalism, believing in a pantheon of gods to be placated or confessing life beyond this world, that is idolatrous. The only other world, for Johnston, ‘is this world properly received’ (Reference Johnston2009, 187). Johnston’s case turns on controversial readings of (mostly Jewish and Christian) Scripture, questionable interpretations of divine simplicity and analogy, and various other moves that should be subject to critical scrutiny. The scalpel is not my instrument here, though: my point is only to note Johnston’s sympathy with phenomenological and grammatical accounts of idolatry, and his uniqueness in this respect among analytic thinkers.
And ‘analytic’ Johnston remains. His critique of idolatry uses philosophical concepts and tools more palatable to analytic thought than ‘ontotheology’ or ‘grammar’. He is not post-metaphysical; he takes God-talk to be referential. His sources are almost entirely analytic. Nor does he see himself in continuity with standard ‘phenomenological’ or grammatical approaches – his only reference to Marion (Johnston Reference Johnston2009, 112) is a critical one, and he rejects Wittgensteinian suggestions of religion as a language game or religions as bearing family resemblance to each other (p. 37n1). Despite Johnston’s invocation of Heidegger on being’s self-unveiling, phenomenologists would worry about his attempt to ‘look at the implied character of the spiritual beings’ appearing in religious experiences, since this approach constitutes a transcendental (and thus, ‘metaphysical’) attempt to get behind the experiences as such. Most saliently, Johnston does not take ‘concepts’ of God to be idolatrous. He argues that we must have an antecedent concept of God to have an experience of God at all (Johnston Reference Johnston2009, 68–69), a conclusion anathema to traditional phenomenology. And although grammatical Thomists might agree to some interpretation of Johnston’s claim, the latter’s treatment of Thomas on analogy and simplicity makes clear that he runs counter to their approach. Nevertheless, Johnston explains how idolatry occurs by undermining the criteria by which God arises to thought and speech in the first place, a model isomorphic to phenomenological and grammatical accounts of idolatry.
Idolatry and god’s aptness to thought and speech
To do justice to this disagreement about idolatry, we must reframe how we think of idolatry in general. I have argued that ‘conceptual idolatry’ has been misunderstood by its critics as a charge distinct from idolatry, rather than as an argument that certain conceptual approaches to God constitute idolatry simpliciter. What phenomenological and grammatical thinkers show is that one can commit idolatry by undermining the very conditions under which God is apt to thought at all. Analytic thinkers fail to engage the charge of ‘idolatry’ as formulated by phenomenologists and grammaticists on precisely this point. Our understanding of ‘idolatry’, then, must expand to examine the means by which God becomes apt to our thinking and speaking, and whether ideas of God which undermine these ways constitute idolatry in its proper sense.
I do not think this impasse between the analytic tradition and its critics must be so. One area where analytic thinkers can engage fruitfully with their phenomenological and grammatical critics is by pointing to problems internal to their own accounts and not imposed through alternative terms. One problem that both phenomenologists and grammaticists might face, for example, is that if God only arises to thought and speech through the specific routes they describe, this must be the case not only for philosophers and theologians, but also for ordinary believers. Thus, phenomenologists and grammaticists might undermine the faith and practice of most religious believers globally: as Wood (Reference Wood2021, 115) puts it, ‘if we cannot worship God under false definite descriptions, then many – perhaps most – acts of worship would misfire’. Johnston embraces Wood’s argument to claim that many practitioners of organized religion are in fact idolatrous. Those who emphasize practices and form of life may want to avoid that route. Though phenomenological and grammatical thinkers could have the resources to combat this charge, it constitutes an objection to which they must respond.
Analytic thinkers could also consider, and indeed have considered, the ways God becomes apt to our thinking and speaking. Analytic thinkers have disputed, for example, whether we should adopt a descriptivist or a causal theory of naming for ‘God’ (see Alston Reference Alston1988; Harris Reference Harris1991; White Reference White1994). A decisive answer would help explain how uses of ‘God’ can refer felicitously or fail so to refer. So too have questions about determining the meaning of the word ‘God’ arisen in analytic circles, such as Jonathan Kvanvig’s (Reference Kvanvig2021) approach to ‘metatheology’. On this question, the phenomenologists give another metatheological route to determining the meaning of deity: God is the giver of certain phenomenal appearances uncircumscribable by our categories of thought. Kvanvig (Reference Kvanvig2021, especially cc. 4, 7) already discusses something approaching the grammatical Thomist determination of deity in his analysis of ‘Creator theology’, even if he does not specify the Creator–creature distinction the way they do. It is not that phenomenological and grammatical thinkers raise questions alien to analytic thought; they raise them in different vocabularies.
A significant sticking issue remains the relationship between theology and metaphysics. Contemporary analytic thought is more sanguine about interpreting religious language as straightforwardly descriptive of a God ‘out there’, as referring to an extra-mental object, than phenomenological and grammatical thinking, or even its own forefathers A. J. Ayer (Reference Ayer1952) and Wittgenstein (Reference Wittgenstein and Barrett2007). But, to give a sympathetic moment to the anti-realist, it seems that if there is any area in philosophy (and in life) to be suspicious of the way in which language tracks descriptive meaning directly, God-talk is an excellent candidate. There are philosophical and theological reasons to think that God may ‘transcend’ our speech in a way that nothing (else) does. And at any rate, as Johnston shows, there are ways to give ground to the phenomenologist – perhaps the grammaticist, too – without throwing referential interpretations of religious language out altogether. However the analytic thinker concludes their inquiry, it must be conducted.
Another matter of dispute is the relationship between the poles of the dyad ‘concept’–‘experience’. Phenomenological thinkers tend to exalt the latter over the former as a check against religion becoming a matter of theoretical assent rather than a form of life; grammaticists aim to accomplish the same goal by subjecting both concept and experience to strong apophatic riders. One might think, on the other hand, that analytic theology and philosophy of religion must be uncomplicatedly committed to ‘concepts’; if this entails a rejection of ‘experience’, then so be it. Such a conclusion, though, is hasty. After all, the attempt to navigate a nuanced relationship between ‘concept’ and ‘experience’ is a project with venerable provenance in Anglophone philosophy, especially as it arises in the interpretation and critique of Kant (e.g., McDowell Reference McDowell1996; Moore Reference Moore1997; Sellars Reference Sellars1968; Strawson Reference Strawson1966). This means that, to answer their critics, the analytic discussion of idolatry must draw upon work in philosophy of mind and philosophy of language in precisely the way analytic thinkers see the question as sharing a porous boundary with metaphysics and, to some extent, ethics.
By my lights, these matters of controversy in the dispute around idolatry are entirely philosophical. This formulation of the dispute about idolatry is entirely a dispute about the conditions under which the word ‘God’ is apt to thought and speech. If someone departs from a given thinker’s criteria for God’s aptness to thought and speech, then that thinker can call them idolatrous. For the phenomenologist, God must appear as God, as transcending our categories. For the grammaticist, God must be the reason for why there is something rather than nothing. In both cases, the very idea of God is tied up with God’s being unlike anything else. ‘God’ cannot ‘mean’ the One who is faithful to Israel, the One who is revealed in Jesus Christ, or any other such definition. For the phenomenologist and grammaticist to accuse their interlocutors of idolatry successfully, they must be absolutely confident that they have the only valid criteria for God’s aptness to thought and speech.
This observation raises a serious problem with the entire status quo of the ‘idolatry’ discussion at large: it turns out not to be a theological dispute at all, but a philosophical one on the grounds of rival semantic theories and rival accounts of religious experience. Co-religionists accuse each other of idolatry on the basis of disagreement over complex claims about metaphysics, the relation between concepts and experience, and the means by which God becomes apt to our thinking and speaking. If thinkers call their opponents ‘idolators’ based on philosophical criteria that their opponents do not accept, then idolatry, and all theological discourse generally, becomes completely determined by the philosophical, psychological, and linguistic conditions under which God can show up for us at all; theology thus becomes reducible to philosophical, psychological, and linguistic analysis. This objection constitutes claiming that Marion and McCabe fail to meet the challenge offered by Barth and Jüngel, since both thinkers construct their Christologies within the broader context of their philosophical criteria for God’s aptness to thought and speech.
In short, I agree that disputes around idolatry must involve the criteria by which God is apt to thought and speech. However, I am not confident in anyone’s ability to identify the exclusive determinations for the meaning of ‘God’. The ways in which ‘God’ becomes apt to thought and speech are many and messy, not isolable to a single condition. I would find it more plausible to define idolatry on dogmatic terms (as Barth and Jüngel do) – not because theology has precedence over metaphysics generally, but because the term under scrutiny is idolatry. There is, though, a more attractive alternative to both philosophical and Christological determinations: ‘idolatry’ must be understood in the context of the shifting and mutually informing exchange between philosophical and theological considerations, rather than subjecting one totally to the other. God-talk must be, as phenomenologists and grammaticists alike claim, grounded in particular patterns of speech and forms of life, rather than (as their practice occasionally evinces) the imposition of one’s own philosophical doctrines upon others with the cudgel of ‘idolatry’. Any other approach would constitute defining idolatry solely ‘according to one’s own subjective conceptual limitations’ (Jones Reference Jones2011, 9), thus ‘direct[ing] one’s thought towards a simulacrum of the phenomenon about which one aims to think or theorise rather than towards the phenomenon itself’ (Rea Reference Rea2020, 39).
Open access
