The Buddhist analysis of causality

I asserted above that the Buddhist takes issue with theism insofar as the latter requires that God is in no way dependent on anything else; this seems to conflict with a central metaphysical commitment of Buddhism that all entities entering into causal relations are themselves dependent upon other things, that is, the doctrine of ‘dependent origination’. By way of illustration, the Buddhist Dharmakirti employs this doctrine consciously to undercut theism, arguing that a creator God would undergo a change in virtue of causing the universe to exist, and so that the theistic notion of an unchanging God who is the causal origin of the universe involves a contradiction (Jackson (Reference Jackson1999, Reference Jackson1986), Hayes (Reference Hayes1988)). Dharmakirti is employing the doctrine to deny, then, that there can exist an entity like God that enters into causal relations – like creation – and nevertheless is not dependent on what He causes.

From a classical theist perspective, the objection involves a fundamentally misguided metaphysical assumption, such that Dharmakirti presumes that God can be causally active only in virtue of an intrinsic change in God. Since God is metaphysically simple – having no composition at all, including composition between His substance and any properties that might ‘inhere’ in Him – God's causality does not involve such a change. We find Aquinas, for example, therefore arguing that God's causal relation to the universe is of a special sort, distinct from the kinds of causal relations we find in material objects. Whereas material objects causing motion in each other involve reciprocal accidents of ‘action’ in the agent and ‘passion’ in the patient, God's causality involves no such accident inhering in God. Creation depends on God as its efficient cause, but God depends on or changes in no way because of this causal relation, and thus merely has a ‘logical’ relation to creation.Footnote ¹ Aquinas is here safeguarding the claim that God is necessarily not dependent on anything else and cannot undergo any intrinsic change.

The special kind of causal relation that God has to creation also involves, for Aquinas and many other classical theists, that God created the universe ex nihilo, bringing creation into existence from nothing in the sense of not acting upon (e.g. ‘re-shaping’) anything already existing. For my purposes, following Aquinas (2020), I understand this latter claim to be a claim about the kind of causal dependence that the universe has on God, that everything in the universe depends on God for its existence, and not a claim about whether the universe began to exist at some moment in time. It would be compatible with the doctrine of ex nihilo creation that the universe has existed for eternity and nevertheless was eternally dependent upon God for every aspect of its existence (Reference Rooney, Stump and WhiteRooney (forthcoming)).

Understood in this way, however, the doctrine of creation ex nihilo is not a negotiable part of classical theism, but a central commitment. Further, as I've noted, it seems to me that Buddhist metaphysics requires denial of these kinds of ex nihilo dependence relations, because the Buddhist wants to hold that every entity entering into causal relations is ipso facto causally dependent on something else. Contrast this with McNabb and Baldwin who try to bridge this gap between Buddhism and theism by arguing that the Buddhist commitment could be qualified by denying that God is a ‘thing’:

What they propose, in short, is to claim that God is not relevantly an entity that falls under the doctrine of dependent origination. Since God is not a ‘thing’, then God needn't be causally dependent, as He wouldn't satisfy the condition that any thing entering into causal relations is also itself a causally dependent thing. Now, for my part, it seems to me that God is a thing or entity in the relevant sense – He exists, He can enter into causal relations with things, and we can attribute properties to Him such as ‘being outside time’. It is also hard for me to see that the Buddhist should straightforwardly admit that there is anything, and especially something that can enter into causal relations, exempt from the doctrine of dependent origination. If they did, that would allow them to draw many more exemptions, such as for an immortal human soul, which would undermine the soteriological aim of the traditional Buddhist system. It would thus seem to be theoretically ad hoc to make an exception for God, because that exception would not otherwise involve any modifications to the underlying Buddhist doctrine of causality. Even if Buddhist scholarship were to phrase the doctrine of dependent origination as a claim that all ‘things’ are necessarily subject to causal dependence, it looks like the point of the doctrine is that (really) anything, even if it is like God, that enters into causal relations is mutually dependent on what it causes. If the Buddhist is to be interpreted charitably, then it looks like dependent origination would rule out God.

Most importantly, though, I do not see that Baldwin and McNabb's move helps resolve the problem: even if we grant that God is not an ‘entity’, that would not seem to resolve the problem posed by traditional Buddhist anti-theological arguments unless God can enter into causal relations without being dependent in any way on what He causes. Baldwin and McNabb manage to carve out a space for God to be unique such that dependent origination does not apply to Him, yet do so in a way that bypasses the questions about the nature of causation entirely. But, as long as the account of causality remains a standard Buddhist account, there is something suspicious or question-begging about the exception for God. This contrast just illustrates, I think, that the central point at issue is whether whatever enters into causal relations is thereby dependent on something else. If God were essentially such a kind of non-thing that His kind of causality is different from all other entities – which is what I take Baldwin and McNabb's distinction to accomplish – this is an indirect route to the same conclusion. On my reading, however, no such essential distinction between kinds of entities is necessary. So, for example, if it was shown that there are kinds of causation that do not entail that whatever's doing the causing depends on what it causes in any relevant way, then even ordinary entities could be exempt in some contexts from the doctrine of dependent origination. The distinction between entities and non-entities is not what accounts for the move from Buddhism to theism; a distinction between kinds of causal relations is what accomplishes that task. All that needs to be argued, then, to move from Buddhism to theism, is not that God is a kind of (non-)thing exempt from dependent origination, but rather that the doctrine of causality cannot be as the Buddhist ordinarily takes it to be.

Huayan and inter-dependence

Interestingly, there are a range of formulations of the Buddhist doctrine of dependent origination, where some schools of Buddhism might come close to admitting different kinds of dependence relations which would be friendly to theism if modified in the right way. Proving that God exists or has such a relation to the world is beyond the scope of the article, but what I hope to propose is that the Huayan 华严 Buddhist tradition might have the resources to accommodate kinds of causal relations where something acts as a cause but only has a logical relationship to (and is not affected by) what it causes. My proposal requires a distinction between efficient and formal causality, two different ways in which things can ‘depend’ upon others for their existence. The Aristotelian terms are helpful to distinguish the typical use of the term ‘causation’ in English, where an effect is brought about by the power of the causal agent (efficient causation), and the way in which we appeal to the nature of some object to explain its behavior (formal causation). I will argue that Huayan, while denying that there are causal sequences where one entity depends on another as an efficient cause, admits there are formal causal sequences. The appeal to formal causes, in Huayan, appears to implicitly admit that the nature of a thing, at least in some cases, does not depend on that of which it is the nature. What I will propose is that one could appeal to the Huayan conception of formal dependence to prompt acceptance of the claim that something could enter into an efficient causal relation without depending in any way on what it causes.

The Huayan school is an idealist school of Chinese Buddhism that develops insights from an earlier idealist school of Indian Buddhism, Yogācāra. The Huayan school's perspective on reality is one where all reality is interconnected and interdependent, such that there are no individual entities or natures or substances, but only a grand web of relations. In the central text of the school, the Huayan Jing 华严经, the web of relations among all things is compared to the ‘Net of Indra’, a vast mythical net of jewels, where each jewel is reflected in every other in the net (Van Norden and Jones (Reference Van Norden, Jones and Zalta2019)).

The Huayan school denied that any causal agent exercises causal power by itself, let alone being able to act or exist without any other cause. That is, anything involved in a causal sequence depends on causal conditions in order to bring about its effect, and every cause also depends on its effect in terms of what both are. An important patriarch of the school, Fazang 法藏, comments on the Huayan Jing, saying:

Fazang assumes that a particle of dust undergoes a change when it comes to be an object of my thought, but also that my mind undergoes a change when it comes to have the dust particle as object of thought. The point of the example is that the nature of both of these entities (the dust and my mind) depend on each other in order to be individuated as cause and effect. In fact, Fazang holds that their coming to be so related depends on the whole net of every causal sequence that has ever, presently, or will ever occur. Consequently, all causation never really involves existential dependence where one entity brings another into existence. Instead, every individual thing depends on everything else for its existence and character (Garfield (Reference Garfield2015), Dushun (Reference Edelglass and Garfield2009)). This could be interpreted as a denial of efficient causal sequences, given that no effect really depends on its cause for its existence. Fazang seems to interpret the salvific import of Huayan teaching in this way, so that, by recognizing ‘that after all nothing arises and there is nothing to be found, perturbed thoughts naturally cease and erroneous discriminations are annihilated’.Footnote ²

Yet, following their Yogācāra predecessors, one finds Fazang and others of the Huayan school refer to a ‘One Mind’ that underlines the varied and fantastical illusion of the phenomenal world that appears to us (Fox (Reference Fox and Emmanuel2013)). Fazang thinks of reality as a realm where each phenomenal event is nothing more than a manifestation of the underlying ultimately mental nature of the whole. The underlying nature of all things is like a mind, which Fazang identifies in the commentary on the Awakening of Faith in Mahayana as identical with the ‘One Mind’ of the Yogācāra school and the ‘buddha nature’ (tathagatagarbha). This ‘One Mind’ that underlies all things, according to an interpretation proposed by Bryan Van Norden,

Graham Priest, to the contrary, interprets Huayan as holding a view where all that exists is identical to everything else, or even as perhaps nothing beyond the web of relations (so that there might not be any concrete particulars at all), and as denying that there is any underlying nature that might explain the goings-on in the phenomenal realm. Consequently, the whole universe is ‘non-well-founded’, where nothing that exists depends on the whole in any meaningful sense; Priest's vision is one in which individual entities depend on each other, ad infinitum, but the whole is in no sense dependent on any cause outside itself, whether a nature or an efficient cause (Priest (Reference Priest2014)). On Priest's reading of Huayan, there is no ultimate ontological ground for the nature of what exists – the whole Net of Indra is ‘ontologically groundless’.Footnote ⁴ My point is not to argue that such an interpretation is incorrect, but only that it does not appear necessary to read the Huayan texts in this way. To the contrary, Huayan texts frequently appeal to natures of objects as somehow explanatorily prior to the particulars and their behaviour.Footnote ⁵ Fazang, for example, appeals to the principle (理 li) underlying the phenomena （事 shi）as what accounts for the underlying structure (体 ti) of the function and behaviour (用 yong) of material particulars.Footnote ⁶ This is to say that, even if efficient causes are illusory, the Huayan texts make appeal to formal causal explanations for why there are (apparently) efficient causal sequences.

In particular, the appeal to mental nature, that is, to attribute all phenomena in the universe to a ‘One Mind’, would be unnecessary if the Huayan Buddhist thought phenomena were merely dependent on each other ad infinitum. The mental nature of the world explains, for example, why we apparently encounter sequences of efficient causation, such as matches lighting balls of cotton on fire. The cotton being on fire does not really existentially depend on the match; rather, the whole sequence depends on the underlying mental nature of reality. Consequently, Huayan does not seem to require that the whole of reality need be ontologically groundless, even if reality would not existentially depend on a prime efficient cause. Even though Huayan is sceptical about the nature of efficient causation, the school might then hold that there is an ultimate ontological ground in the formal causal order. In keeping with this interpretation, Fazang describes the True Mind as ‘pure’ or ‘undefiled’ or ‘unchanging’.Footnote ⁷ And, similarly, the Huayan sutra lists many respects in which ‘true thusness’ (identical with the One Mind in Fazang's commentary) has properties that are such that, for example, it cannot be lost and destroyed, pervades all things, is the true or essential nature of all things, and is unchangingly pure (Cleary (Reference Cleary1993)). The upshot seems to be that the relationship is that the material universe and all phenomena are merely phenomenal, not being identical with the true nature of things, and so that the nature of things is not affected by what is generated by it in the phenomenal realm. One finds confirmation in claims such as: ‘The real essence of things is permanent, suchness, beyond thought’,Footnote ⁸ or,

On this reading, then, the nature underlying the whole web of interdependent entities is the ultimate or true nature of reality, and Huayan is teaching a kind of radical monistic doctrine.

If my reading were correct, nevertheless, the Huayan Buddhist would be part-way to a possible ex nihilo order of dependence that brings them closer to theism than perhaps other Buddhist traditions. I say ‘part-way’ because the view does not clearly or distinctly distinguish the ultimate nature, the reason why things are such, from the universe of which it is the explanation. Instead, the Huayan tradition attempts to identify the two by saying that the One Mind does not exist in separation from the phenomenal realm that it sustains. Yet Fazang's position, like Yogācāra, is a kind of metaphysical idealism – the Mind explains the phenomenal world, and the phenomena depend on the Mind in some way for what they are. But then the question is what relationship exists between the One Mind and the phenomenal things that relate to it; that is, what it is involved in the ‘dependence’ envisioned?

Since we are operating within the ambit of formal causal dependence claims, it seems we could have only three possible ways something could depend on another for its nature. Some things can share the same nature because they are identical. On this first possibility, [1] the phenomena would be strictly identical with (i.e. distinguishable in no way from) the Mind. But I think there are reasons that Huayan might not want to embrace that the One Mind is strictly identical with all phenomena. Since all distinctions between individual things appear to be mind-dependent, it makes sense that Huayan would hold that the mental nature of the whole is not identical with the (apparent) nature of the individual phenomenal things that appear to us, nor with any individual phenomenal appearance. Fazang, for example, seems to hold that all that exists in the phenomenal realm (e.g. the independent individual entities that apparently stand as members in causal sequences) depends on this Mind for existence, but that the mind does not depend on those things for its nature. That is, there is some asymmetric relationship between the individual phenomena and the Mind. And the view claims that our identification of the nature of the phenomena is merely apparent, which would seem to lead to a contradiction if the Mind were strictly identical to the phenomena. Similarly, the One Mind is the underlying reality or true nature of things that explains the occurrences in the phenomenal order (e.g. apparent causal activity). If so, the One Mind would not seem to be ‘pure’ in the right way if the illusory phenomena were strictly identical with it – there would be no ground for any distinction at all between One Mind and phenomena if they were identical in that way.

If not identical, we need to introduce some other way that one or more things can depend upon another for their nature, but which allows distinction between the relata. There seem to be two options: either a relation, such as composition, where the identity of the parts depends on but is not identical with the whole they compose, and where the whole would not exist without those parts; or, a relation like instantiation, where the particular depends on the universal for what they instantiate but where the nature or existence of the universal does not depend on the particulars. We can provisionally grant that relations like efficient causation explain the nature of things in terms of the agents that cause them; yet, at least at this point, efficient causal relations are not the right kind of formal dependence that the Huayan Buddhist is willing to concede. Consequently, either [2] the phenomena bear a relationship to the Mind such that the phenomena depend upon the Mind, as if instantiating the nature of the Mind, but phenomena are not constitutive of Mind in terms of its identity, or [3] the phenomena are related to the Mind as if as parts to whole. While I think that, in fact, Huayan is somewhat diffident between these two possible readings of its vision of reality, it seems to me that either reading would make possible an argument for the existence of a theistic God.

On [2], the phenomena have a relation with the One Mind akin to instantiation, and we should consider the One Mind not as a particular but as a universal like ‘mental nature’ itself. The phenomena that instantiate the One Mind, then, do not enter into the identity of the One Mind (i.e. the mental nature of everything). The notion of the universal in question can nevertheless be interpreted in either Aristotelian or Platonic fashion (Armstrong (Reference Armstrong1978)). On an ‘Aristotelian’ reading, universals exist only insofar as they are instantiated by particulars, in a way similar to a multiply located common part of all the particulars that instantiate them. But, if this was the case for mental nature, that nature need not be instantiated at all and so need not exist. And yet the One Mind is supposed to characterize absolutely everything. On one hand, the Huayan Buddhist would not want to appeal to the particulars, the phenomena, as essentially or necessarily instantiating mental nature, because this would undermine the claim that no phenomenal things have essences. On the other, admitting that the One Mind, as it characterizes all the particulars, is contingent means that we would need an explanation, aside from One Mind, for why phenomena instantiate it since they do not do so necessarily. Usually, we appeal to efficient causality to explain why some particular instantiates a universal if it does so contingently, and, if we wanted to appeal to something else in the order of formal causality to explain the contingent instantiation of One Mind, we would have to come to some other universal which everything instantiates necessarily. This could be either by some other Platonic universal, distinctly subsisting from the instances and whose nature is such that it accounts for the instantiation of the particulars, or by reconceiving the nature of the One Mind as a particularized universal or property, for example, a nature that is a metaphysical constituent of a composite whole material object.

If the Buddhist opts for a Platonic reading, where universals subsist non-spatiotemporally and are distinct from their instances, the One Mind would be a subsistent universal, of a different ontological kind from all the phenomena, but not identical with them. For good reason did Neo-Platonists develop a model of ‘emanation’ to describe that dependence, because it looks like the relation between this one universal and its instances is akin to efficient causality, since the Platonic universal is related to, but does not enter into anything like composition with, the particulars (Wallis (Reference Wallis1995)). The relation that the Mind has to the particular phenomena would be what accounts for those phenomena having the nature they do, but the Mind does not depend on the particulars for its nature or its existence. This looks like a straightforward version of classical theism for the Huayan Buddhist. So the Huayan Buddhist should think of the instantiation relation instead in the Aristotelian way, rather than the Platonic way, if they want to avoid theistic implications.

This should then push the Huayan Buddhist to option [3], where the One Mind is composed of the phenomenal as parts, such that the Mind is upon them for its existence and where those parts are constitutive of its identity. We might initially imagine that the One Mind need not be like a particularized nature, on [3], but rather only like a material composite whole. A material object's nature is usually existentially dependent on its parts. There is just no more chair or Tibbles once we cut away enough matter. Even if we take the phenomenal parts as non-material components, such as properties or modes, the One Mind can similarly depend for its existence on its component phenomena – if they did not exist, neither would the Mind. Yet, Huayan needs to draw a distinction between the Mind and the parts, such that there is something to being the whole that is not identical with being the parts, even if the whole is entirely decomposable into the parts. We already saw the tradition claim that this ultimate nature of everything is in some way unaffected by the goings-on in the material things or phenomena, because that Mind is described at least in some respects as ‘pure’ and unchanging. This seems to imply that some aspect of the whole One Mind can be distinguished from the parts, as characterizing every part equally. Namely, the relevant common feature of all the parts is that they are mental, as this fact is supposed to explain the way in which our distinctions between phenomenal objects (e.g. cats and dogs) are mind-dependent and perhaps even trivial.

Appeal to the mental nature of reality appears to be doing explanatory work on the Huayan's idealist account of reality, and so we might think instead that the relevantly explanatory aspect of the One Mind is not that it is a composite whole but that its nature characterizes both the whole and the parts together. For this reason, one should not too quickly identify the One Mind with the composite whole. Now, what I mean by ‘nature’ here should be taken broadly, given the Huayan insistence that nothing has a nature or that all things are empty of nature. What I mean is that, if it were merely a brute fact that the phenomena relate to the One Mind in the way they do, rather than something following from the nature of either phenomena or Mind (in a broad sense), it would only be a contingent fact that they relate in this way. But then the positing of the One Mind would seem unexplanatory, since the claim that the phenomena compose the Mind as a whole, having a composition relation, seems like Huayan intends to explain certain facts about the phenomenal things (such as their being mind-dependent) in virtue of their composing or being characterized by Mind. So, even if we were to deny that there is any substantive essence of the phenomena or essence of Mind, Huayan would need to explain the way in which their explanation about the mental character of the phenomena and the One Mind holds in virtue of the nature of their composition relation. That is to say, there would still need to be something in virtue of which the parts compose the Mind such that it would be true that, necessarily, if the parts compose a whole, then they compose that whole they do. The latter claim still requires something like a claim that it is in the nature of a part–whole relation that undergirds the truths about composition – as long as and insofar as x composes y, then, necessarily, the proposition that x composes y is true.

This is, of course, very similar to how contemporary metaphysicians might argue that ‘truth-makers’ have a necessitation relation to those propositions they make true (MacBride (Reference MacBride and Zalta2021)). What I am proposing is that Huayan appears to accept that there is a truth-maker for the fact that the phenomena compose the One Mind, and that this truth-maker involves certain aspects of the particular mental nature of the composite whole. However, if they accept an explanatory distinction between the particular mental nature characterizing the One Mind, the One Mind composed of the phenomena, and the phenomena, we can ask questions about the mental nature that account for the composition of the Mind and phenomena. Is it merely contingent or is it necessary that the One Mind is composed of these phenomena – that the One Mind is actually composed of its parts in this way? If the Huayan Buddhist admits the One Mind is contingently as it is, they would be inconsistent if they left this contingency as brute fact. The Huayan have already appealed to one order of formal dependence such that the phenomena compose the One Mind, where they take that dependence to be explanatory in some way, so to fail to explain what it is in virtue of which One Mind is as it is (despite conceding it is contingent) by appeal to that Mind's nature or something else would seem arbitrary. Admitting that the One Mind were contingent, however, would lead the Huayan Buddhist right down the path toward a contingency argument for the existence of God, because they would admit the need to posit a further thing that is essentially such that it accounts for the existence and nature of everything else (since the One Mind is everything else). That looks like a necessary being, aka the God of classical theism.

I think the best way to read the Huayan tradition is that they implicitly presume that the One Mind is necessarily the way it is, composed of these phenomenal parts. In this respect, Huayan puts tension on its claim that nothing is characterized by an essence since it would be appealing to the essence of the One Mind to explain why the phenomena compose it. And the view resembles a version of pantheism, because the One Mind (of which everything else that exists is a part) is a necessary being. It does not seem to me implausible that Huayan, rightly interpreted, is indistinguishable from pantheism, as Buddhist thinkers might have more generally embraced doctrines that are ultimately indistinguishable from pantheism (Duckworth (Reference Duckworth and Powers2015)). The pantheist view, however, can be criticized on the basis that it seems problematic for a necessary being to have parts in any respect. Thomas Aquinas represents the classical theistic tradition when he rules out God having any sort of ‘metaphysical’ composition, such as having inherent properties or qualities, as following from the fact that God is essentially a necessary being (Lamont (Reference Lamont1997), Vallicella (Reference Vallicella and Zalta2019)). At the core of such arguments is the idea that what is doing the explanatory work on the pantheist account, what makes the whole a necessary being, is only one distinct metaphysical constituent characterizing all of the other parts; for example, the nature of the composite One Mind. The pantheist is inconsistently supposing that something like a metaphysical constituent or part can do this kind of work. If the whole is exhaustively decomposed into the constituents, meaning that the One Mind is nothing more than the phenomena suitably existing or arranged, etc., and the whole is a contingent thing (that need not exist), then the constituents of that whole, including its particular nature, are also contingent. If we did not appeal to a Platonic universal, we'd need another particular that relates to this whole by an extrinsic relation – unlike a metaphysical constituent – and accounts for that whole and its parts being characterized by their particular mental nature. That particular would look a lot like God, as noted earlier, since it would be something that essentially accounts for the composition of the pantheistic whole, but is not dependent upon it.

If the particular nature of the pantheistic whole is such that it does not depend upon the parts of that whole, where it could persist after the decomposition of the parts, the pantheistic particular nature would exist in the way in which a human soul is envisioned by Aquinas to be able to persist after the death of a human being. But this means that this particular nature of the whole would not be a necessary being, and only a contingent thing, since it would depend essentially upon its parts insofar as it is essentially the nature of one particular composite, just like my soul is essentially mine and not (for example) that of Anaximander. But, since I am a contingently existing composite, my soul is contingent too. The pantheistic Huayan Buddhist should not accept that this is possible for the One Mind, and instead should hold that the mental nature only relates contingently to the phenomena – in itself, it is necessary and eternal, despite whatever phenomena that nature characterizes at any one time. But this implies precisely that the mental nature is subsistent and not essentially dependent on the phenomena to exist at all.

The relation between that nature and what it characterizes is not then plausibly a constituent or a part, since parts are intrinsic to the wholes they compose, whereas the mental nature (on this option) is not essentially intrinsic to the whole it characterizes, in terms of having an identity distinct from any of the parts individually and all the parts as a whole, since the nature exists regardless of whether the parts do. There is no good reason, then, to think of the relation between this kind of particular nature and what it characterizes along the lines of ‘parthood’ at all – the mental nature merely has an accidental relation to all the things it characterizes, as even the phenomena or the One Mind (the composite) depends on the mental nature for what they are. But now one has embraced theism: the mental nature is a subsistent thing, a necessary being which everything else depends upon for its existence, but which only has an extrinsic relation to what depend upon it, such that it depends in no way upon the composite or the phenomena. That looks, naturally, a lot like some kind of efficient causality, which is why the theistic tradition has often understood God as being first efficient cause, rather than a formal cause, of the universe, since God necessarily only has an extrinsic relation to the universe He accounts for. It seems merely stipulative to refer to these distinct kinds of relations as ‘non-causal’. But it does not ultimately affect the conclusion that, if these relations were possible, it is possible for one thing to depend on another for its essential character or properties or existence, but where that relation holds extrinsically or contingently between the relata, and therefore that, if the Buddhist can accept formal causal dependence, the Buddhist can accept theism.

A concluding meta-metaphysical postscript

What I have illustrated is one way in which admitting the legitimacy of at least one kind of formal causal explanation by Huayan Buddhism entails that there are kinds of dependence such that it is not necessary for a formal causal explanans to depend on its explanandum. That is, if there are legitimate explanations that function like formal causal dependences, where one appeals meaningfully to the nature of something in order to explain its properties, including its composition, or its existence, then one could utilize considerations drawn from the order of formal causal dependence to arrive at the God of classical theism. This appeal to formal causality, importantly, makes the kind of dependence associated with theism – ex nihilo dependence – follow as a principled distinction, as opposed to a question-begging one, from the Huayan tradition's own assumptions. Since the Huayan tradition already apparently accepted formal causality as a legitimate kind of explanation, nothing new needed to be presumed at the outset about God being a distinct kind of entity or even that there are distinct kinds of efficient causal relation. Instead, one can move from a general notion of formal causal dependence to explicating, in terms derived from formal causal dependence, the claim that there are contingent beings which depend on others for their nature, and from that distinction between contingent and necessary beings arrive at a necessary being that is identical to the God of classical theism.

Indirectly, this illustrates that allowing explanation of phenomena by appeal to their underlying nature permits proofs for the existence of God which are compatible with strong versions of scepticism about efficient causality (or, as the Buddhist tradition might hold, that efficient causality among material objects involves only accidentally related causal series). Pretend David Hume was right about causality: there is no ‘dependence’ of effects on causes, but instead the material universe is merely a series of a regular succession of events. Maybe there aren't even objects ‘out there’ at all. It would turn out in the Humean world that there is no causal dependence between events, as what looks like causal dependence is actually only law-like regularity. It would be invalid, in this universe, to infer that there needs to be a first efficient cause on the basis of the succession of one event after another.

Nevertheless, one might think that the succession of events and its law-like regularities is not merely a brute fact but requires formal causal explanation. For example, the laws surrounding gravitation might be explained in terms of the structure of space-time, or the way that light propagates through a medium might be explained by the character of that medium alongside the character of light, or the behaviour of electrons in certain circumstances might be explained by field interactions. The explanans in each of these cases is not a causal agent, but a ‘structural’ or ‘formal’ feature of the explanandum. And here too we have dependence, because the event is explained by and would not occur without the relevant structure. But if we acknowledge the need for formal explanation in the particular case, we can extrapolate this kind of reasoning to the material universe as a whole. If these dependence relations in the formal order are meaningful, to show that: if the material universe (even a Humean one) depends on something else in a formal causal order, then that universe is a contingent entity. But the introduction of contingent-necessary distinctions such as these leads naturally to an explanation on which some being therefore accounts for the existence or nature of the contingent beings because it is intrinsically or essentially a necessary being. This kind of argument is not unusual, as Richard Swinburne (Reference Swinburne2004) has developed arguments that are based on similar considerations. In such arguments, God functions as a formal causal explanation of the universe, that principle in virtue of which the regularities and natural laws characterize the universe, without becoming something like a metaphysical constituent or structure of it (like a world-soul). No appeal to the nature of efficient causality is strictly necessary, although (as in my discussion) it could be introduced to explicate the unique kind of relation God has to the world.

This does not end the discussion with the Buddhist, nevertheless, as I think the Buddhist should then make some move to reject the implication that the God of classical theism is possible by doubling down on another area of their metaphysics. Since the Buddhist ceases to be a Buddhist by accepting theism, on my view, they should deny the assumption that got my whole argument started: that is, they should deny that there are orders of formal causal dependence. Denying formal causal explanation is a serious cost, as this requires that the Buddhist call into question global assumptions about metaphysics to avoid the theistic implication. There are, I think two possible ways to go. Both employ an appeal to the underlying nominalist thrust of the Buddhist world-view. What is at issue, in the end, is the overall aim of metaphysics as a project.

One direction is to claim that all reasoning of any kind, including reasoning about natures and laws, is deeply misguided, such that all reasoning is invalid (not truth-preserving) and does not result in true claims. But the cost of this option is that, as no reasoning is valid, the reasoning they might employ against non-Buddhist views is also invalid. Similarly, the idealist arguments that everything is ultimately mental and illusory would be invalid, as well as the reasoning that would lead one to attribute all phenomena to an underlying ultimate nature. If one goes in this direction, Huayan's apparently metaphysical claims about the mental character of the universe should rather be interpreted as ‘skilful means’ intended to occasion a more appropriately enlightened perspective, supportive of moral transformation, as having a practical purpose, but not as being reasoning that demonstrates the truth of Huayan positions.

The Buddhist might think that Buddhism is nevertheless literally true, even if reasoning cannot be employed to demonstrate it. Maybe the truth of Buddhism is only evident via direct experience, under the relevant circumstances. For this reason, I think it is not accidental that the Huayan school eventually gets absorbed by the later Chan school, as the Chan view that anything other than a direct experience of the world is falsifying would be a natural (and probably more theoretically stable) extension of the Huayan perspective (Hershock (Reference Hershock and Zalta2019)). This kind of move has its own difficulties, as it looks as if one still needs to differentiate enlightenment-preserving from non-enlightenment-preserving practices or experiences or circumstances or transmission lines by which one arrives at the truth by this kind of direct gnosis inherited from the appropriately enlightened Master. The Chan school thus accentuates conditions of dharma transmission – public lists of enlightened Masters who inherited the dharma without a break in transmission from the Buddha himself, double-checked, historically accurate, etc. But this either ends back in truth-preservation of some kind (they're just truth-preserving lines of transmission!) or in a question-begging appeal to authority.

A better way for the Buddhist to double down and reject formal causality is to reject anything like a ‘truth-maker’ principle for reasoning about the world. The Huayan sutra seems to accept that propositions about the existence of objects, let alone whether one thing has a property or composes another as a proper part, are similar to illusions.Footnote ¹⁰ I take it that these claims should be interpreted along the lines of thinking that propositions about the existence or composition of objects can be true propositions, yet only trivially so because their truth involves nothing like a truth-maker, for instance, a mind-independent way that the world is, that would necessitate their truth. Instead, all propositions are true in virtue of the contingent states of our minds or conceptual scheme. The Huayan then does not seem to be necessarily involved in any contradiction when they extend this perspective to their own global perspective and hold that the claim that the truth of ‘all things are mind-dependent and trivial’ requires no truth-makers; for example, there is nothing that it is to be a mind or have a mental nature apart from our conceptual schemes, so mental nature is not something which necessitates the truth of propositions.

Clearly, on this approach, what counts as a sufficient explanation of some phenomenon, in a given context, would not be like a formal cause or nature of anything. Instead, their explanations will be far closer to epistemology or logic than to natural scientific reasoning. The Buddhist will inquire about the structure of our mental states or conceptual schemes that give rise to the phenomenal world of objects we experience. Nevertheless, a difficulty is that this way of proceeding cannot posit anything in the mental structure or logical structure that necessitates the truth of the propositions they are explaining. All of the truths they consider, all the way down, should be contingently true without anything like truth-makers of any kind. But then it looks very much like we cannot be engaged in the reasoning or explanation business, because none of our modes of inference or explanation could be such that they necessarily are of such a logical form to preserve truth.

The Buddhist will still owe us a consistent and plausible account of how their view does not collapse into the rejection of any reasoning as truth-preserving and, if it does not, how their view is compatible with what we ordinarily think we know about the world. It seems to me that the most promising route (perhaps exemplified in Tibetan Buddhist traditions) is to make a move akin to other nominalists that embraces an austere reductionism about truth. I take it that the Buddhist should embrace a view like that of Wilfrid Sellars, for example, who developed a meta-linguistic nominalist theory. Sellars held, roughly, that the nominalist can abstract singular terms that otherwise look like universals, replacing them with metalinguistic abbreviations that indicate the linguistic rules which govern the statements incorporating such abstract terms. At no point would we need to appeal to something like a universal (Loux (Reference Loux2006)). The Huayan Buddhist can then explain their ontological claims about the One Mind, in this light, as a kind of ‘ontology made easy’ because of their underlying nominalist approach to metaphysics.

The dialectic here illustrates why it is insufficient to attack one aspect of the doctrine of dependent origination and try to carve out an exception from the doctrine (for example) for special classes of non-entities. Dependent origination is intimately connected with the meta-metaphysical background of the Buddhist nominalist system of metaphysics. To change the Buddhist doctrine of dependent origination in admitting different kinds of formal causal dependence is necessary and sufficient for moving from Buddhism to theism because, in making such a change, their system's nominalism is moderated in some small way. But then modification of the doctrine in the direction I propose – towards admitting the kinds of dependence that would permit the existence of God – would ipso facto require the Buddhist to embrace a different approach to metaphysics, moving from austere nominalism to (at least) a moderate realism in some context.

One might therefore think, on my view, that showing Buddhism is implausible really comes down to showing that (in at least one context) the Buddhist version of austere nominalism doesn't work or is implausible. For example, if formal causal explanations are good ones, predictive and informative, as scientific practice indicates, then there is very good reason to think that they are true – appeals to natures are informative because things really have natures. The success of formal explanations, then, would seem a good reason not to be an austerely nominalist Buddhist. But, as we know, the dialectic of realism and nominalism is not easily exhausted by one apparent counter-example; the Buddhist can always have a further nominalist explanation.

In the end, I would suggest that reasons to reject Buddhist austere nominalism are likely to be connected to what you think metaphysics is about. For the realist, metaphysical explanation is tightly connected to natural scientific explanation. Distinct orders of dependence, such as formal, final, or efficient causality, make far greater sense if one is engaged in an explanatory project that mirrors that of natural science, where, for example, appealing to a law of gravitation is a different kind of explanation from the way we appeal to the internal molecular structure of a compound. Probably for the same reason, the plausibility of theism is tightly connected with intuitions about explanation and theories of causality that are connected with methodology in natural science and with realism in metaphysics. By contrast, the Buddhist austere nominalist approaches metaphysics as if it were about something else – language, mental concepts, processes of thinking. And the Buddhist has the burden of proof, apparently needing to show that their austere nominalism is somehow necessarily, globally true, given how wide-ranging that nominalism needs to be to rule out any possible route for theism. Resolving that meta-metaphysical dispute, and showing that there is a problem in the reasons that motivate accepting either realism or nominalism, goes beyond the scope of this article. Nevertheless, if my suggestion here is correct, the move from Buddhism to theism is really a paradigm shift, not a tweak.