The following is a long-form project I undertook for my theological studies in 2021-22 (the main reason for the extremely light posting during that time). As the title suggests, it critically assesses the model of the Trinity put forward by the philosopher and theologian, William Lane Craig. A warning: it’s not for the faint of heart. At around 17,000 words — and dealing with a host of somewhat obscure or arcane issues — it requires a certain degree of patience. But for those willing to persist, I trust you will be rewarded!
Introduction
The doctrine of the Trinity, so central to Christian belief, has generated a near-ceaseless stream of controversy since its formal articulation in the fourth century. A key source of this protracted debate, exciting theologians and philosophers alike, is the so-called “three-in-one” problem: that is, how the one true God can at the same time be three distinct persons, all of whom are equally divine. Anyone trying to provide a compelling account of the dogma is immediately confronted with the task of avoiding contradiction, while endorsing the following seven propositions:
- The Father is God
- The Son is God
- The Holy Spirit is God
- The Father is not the Son
- The Son is not the Holy Spirit
- The Holy Spirit is not the Father
- There is exactly one God
This heptad forms a narrow strait that a person must navigate if she is to remain within the bounds of mainstream Trinitarianism. On one side lies the Scylla of tritheism, while on the other lurks the Charybdis of modalism. But the demand for rational consistency must also be met: how, for example, would the apparent identity relations of (1)-(3) be reconciled with (4)-(6), which imply personal distinctions?[1] With regard to the Trinity, fidelity to both the canons of logic and orthodoxy have long occupied the doctors of the church.
Several basic families of Trinitarianism have emerged as possible ways of resolving these dilemmas. One of them, Social Trinitarianism (ST), flourished in the second half of the twentieth century as part of a wider renaissance in Trinitarian thought. While a variety of social models exist, they are united by the conviction that the Trinity can be conceived of as a loving family or community of three psychologically robust selves. Among the champions of ST is philosopher and theologian, William Lane Craig, who has developed a model of the triune God he dubs Trinity Monotheism (TM). He argues that although Father, Son, and Spirit are fully divine, only the Trinity as a whole can properly be called “God”. Depicting the Deity as a soul-like substance possessing three sets of rational faculties – each of which is sufficient for personhood – Craig contends that the Trinity instantiates divinity, while the persons are divine in virtue of being proper “parts” of God.
Craig’s proposal has generated a paltry response from the broader theological and philosophical communities. The following essay aims to remedy this relative lack by offering an extended assessment of its viability. My contention is essentially negative in cast: although TM exhibits certain strengths, it is weakened by a series of conceptual, logical, and theological flaws – afflicting both Craig’s formal theory and his model – which together prove fatal. Whatever virtues it possesses, Craig’s proposal ultimately fails. Regrettably, anyone wanting to settle the various conundrums thrown up by the doctrine of the Trinity will have to search elsewhere.
The essay will unfold as follows. First, I will describe TM in detail, surveying its main contours. Second, I will briefly examine one of the key strengths of the scheme, relative to other social models of the Trinity: namely, a fairly robust principle of divine unity. These sections will constitute a prelude for my critical assessment of Craig’s proposal, the essay’s most important phases. I will divide this segment of my analysis between TM’s formal theory (the persons are divine because they are proper parts of the Trinity) and the model proper (God as a soul-like substance with three sets of cognitive, affective, and volitional faculties). Turning my attention to what I consider some of the proposal’s key deficiencies, I shall scrutinise it in conversation with previous criticisms. It is here that I will substantiate my basic position, arguing that TM does not provide a coherent philosophical – and indeed, theological – account of God’s tri-personal life.
Among other things, I will argue that the mereological foundation Craig uses to allege the persons’ full divinity, supplemented by his cat analogy, is too fragile a basis for the claims he wishes to make. More than that, I shall contend that the proposal has a deflationary effect on the persons’ status, implying different types or grades of divinity. I will explore Craig’s proffered Cerberus analogy, suggesting that neither it nor other alleged parallels throw sufficient light on the question of how three divine persons may constitute one God. I will also examine the implications of TM for the nature of the divine persons in relation to the entire Trinity – contending that the model Craig offers is both unstable and ambiguous, lacking sufficient clarity concerning the members’ ontological status. This will go along with critical comments concerning the lack of proper coherence between Craig’s model and other aspects of his thought. Finally, I will elucidate what is perhaps the most serious (and surprising) charge against TM: that whatever else he might say, Craig’s proposal ultimately entails God’s non-personhood.
Some notes on structure and methodology
Before moving on, several important methodological decisions ought to be clarified. First, the essay will be a largely philosophical (or philosophical-theological) examination of TM. My focus will be on the conceptual and logical issues thrown up by Craig’s proposal, and as such, I will concentrate less on specifically biblical or exegetical issues. That said, because the Trinity is an avowedly religious doctrine, such concerns – especially the comportment of TM with biblical texts and later creeds – cannot be entirely excluded. I will therefore make targeted reference to representative passages and formulae in cases of apparent conflict. Second, I will simply assume that God’s nature or being is somehow tri-personal (however this is to be understood), and that such a view is entailed by a reading of the relevant New Testament texts; arguing for this point will form no part of the essay.
Third, my approach will be largely critical, rather than positive and constructive. After supplying the necessary context, I will be occupied with assessing Craig’s scheme; I will not make proposals of my own (except, perhaps, en passant), and I will not be advocating other models. Finally, I will refrain from critiquing TM using tools and concepts that are themselves controversial. For example, Craig’s compositional approach to the relations between the persons and the entire Trinity can easily be criticised on the grounds that it violates the doctrine of divine simplicity (DDS). DDS, of course, hinges on the claim that in God there are no parts of any kind (whether material or metaphysical). Moreover, it has a long and venerable history, and continues to enjoy warm endorsement within the guild of philosophical theology. But while the doctrine remains compelling to some, it attracts substantial criticism from those who, like Craig, reject it as incoherent or unintelligible. To try and defend (in this case) DDS would require far more space than is available here, rendering the current enterprise unworkable.
Trinity Monotheism: A Summary
I turn now to Trinity Monotheism, which Craig developed during a particularly fertile period for analytic Trinitarian theorizing. TM is a species of Social Trinitarianism, or what has otherwise been dubbed “three-self” theories of the Trinity.[2] Proponents of ST conceive of God, not as a single individual or self, but as a trio of subjects, bound together in a loving, intimate, and harmonious way. Despite their diversity, ST theorists are united in their commitment to a group-notion of the triune God, in which the three members constitute a kind of divine society.[3] Underlying this conception is the crucial claim that Father, Son, and Spirit are persons in a modern, psychologically rich sense: they are not merely subsistent individualisations of the divine essence, nor “concrete particular non-properties”,[4] but fully-fledged personal agents akin to human subjects.
Craig’s basic thesis can be summed up as follows: while the persons of the Godhead are fully divine, it is the Trinity as a whole that may legitimately be called “God”. As he writes, “…the Trinity alone is God…the Father, Son, and Holy Spirit, while divine, are not Gods”.[5] This conviction underpins Craig’s entire proposal, which can be conveniently divided into two segments: his formal theory, which concerns the divine members as “parts” of God; and the model proper, where he conceives of God as a soul-like substance with three sets of cognitive faculties, each sufficient for personhood.
Craig’s account of the Trinity is partly motivated by philosopher Brian Leftow’s criticism of similar models. Leftow threw down the following challenge to champions of TM (part of a broader critique of ST thought):
Either the Trinity is a fourth case of the divine nature, in addition to the Persons, or it is not. If it is, we have too many cases of deity for orthodoxy. If it is not, and yet is divine, there are two ways to be divine – by being a case of deity, and by being a Trinity of such cases. If there is more than one way to be divine, Trinity Monotheism becomes Plantingian Arianism. But if there is in fact only one way to be divine, then there are two alternatives. One is that only the Trinity is God, and God is composed of non-divine persons. The other is that the sum of all divine persons is somehow not divine. To accept this last claim would be to give up Trinity Monotheism altogether.[6]
Craig disputes Leftow’s framing of the dilemma. He rejects the idea that the Trinity is a fourth instance of deity and denies that the Godhead is composed of non-divine persons.[7] Moreover, he repudiates what he deems an unjustified assumption on Leftow’s part – namely, that if there are two ways of being divine, the persons of the Trinity will invariably be hobbled by a kind of “diminished” divinity. Craig argues that there is, in fact, more than one way of bearing this property: by instantiating the divine nature; and by being a part of that exemplifying entity.[8] Deploying the notion of part-whole relations, he explains that the members of the Trinity are fully and unequivocally divine because they are proper “parts” of the Deity. The entire Trinity is God, for unlike the persons it instantiates the divine nature. But Craig insists that while divinity may be possessed as a result of different factors, this in no way undermines the status of the persons.
Craig’s motivation is clear: not only is he anxious to avoid a Quaternity; he also wants to ensure his scheme does not court tritheism by positing a series of individuals that independently exemplify divinity. On his account, there are not four gods since there is only one exemplification of the divine essence; Father, Son, and Spirit should therefore not be seen as identical with God in the same way that the whole Trinity is.[9] Craig recognises the potential strangeness of using mereological language to conceive of intra-Trinitarian relations (as God is not a material object extended in space). But he also notes that a certain part-whole relation seems to obtain in this instance, given, for example, that “…the Father…is not the whole Godhead”.[10]
Craig attempts to illuminate his theory with the analogy of a cat. He deploys his feline example to explain how the persons can be divine without exemplifying that nature, and why this does not entail an unfavourable disjunction between them and the entire Trinity. Although a cat’s DNA and skeleton do not instantiate felinity or “cat-ness”, they remain fully feline because they are proper parts of a cat (which does instantiate that essence).[11] Similarly, because the members of the Trinity are proper parts of God, their own status is safely underwritten. Craig writes that “far from downgrading the divinity of the persons, such an account can be very illuminating of their contribution to the divine nature”. He goes on to say that the entire Trinity possesses some properties because the members do (e.g., omniscience, omnipotence, perfect goodness), while the members enjoy other attributes – necessity, aseity, and eternity – “because God as a whole has them”.[12]
Craig realizes that this still leaves unexplained the reason why the persons should be viewed as metaphysically integral parts of the same entity, rather than three otherwise discrete beings (who happen to be united). He therefore cites a second, and somewhat infamous, analogy: Cerberus, the mythical three-headed dog of the underworld.[13] “Cerberus has three brains”, Craig writes, “and therefore three distinct states of consciousness of whatever it is like to be a dog”. Despite this mental plurality, Craig avers that “Cerberus is clearly one dog” and “a single biological organism” – points that he thinks can be used profitably to illuminate the notion of a single entity enjoying distinct, multiple, and co-occurring states of personhood.[14]
To be sure, the analogy requires elaboration, lest it fall short of showing how an immaterial being may “support” three persons while still being one. Craig therefore invites us to reflect upon the nature of the soul and how it can be applied to the Trinity. He argues that:
God is very much like an unembodied soul; indeed, as a mental substance God just seems to be a soul…endowed with three complete sets of rational…faculties, each sufficient for personhood. Then God, though one soul, would not be one person but three, for God would have three centres of self-consciousness, intentionality and volition, as Social Trinitarians maintain…God would therefore be one being that supports three persons.[15]
For Craig, God can be described as a soul-like substance, supporting or underlying three sets of cognitive, volitional, and affective faculties. God-as-soul is likened to its human analogue: while a human soul sustains a node of consciousness, the Deity is so richly-endowed that he bears three bundles of such “equipment”, each constituting a distinct individual. These entities are fully personal, possessed of agency and intentionality.[16] Craig not only contends that TM captures the spirit of orthodox creedal lore; he also claims that it pushes beyond mere recital of fourth-century language and tropes to provide an intelligible and philosophically cogent account of the Trinity.[17]
Assessing Trinity Monotheism: Relative Strengths
TM and the principle of divine unity
Although much of my analysis of TM will be critical in nature, I want to start by making a few comments in favour of Craig’s proposal. Conceiving of God as a kind of soul-like substance, of which Father, Son, and Spirit are proper “parts”, carries at least one key virtue lacking in many of the main “pro-social” alternatives to TM: a comparatively strong account of divine unity. Many forms of ST on the market arguably lack a sufficiently robust principle of divine oneness, which means that such models remain vulnerable to tritheism. In his broad-based study of ST, Daniel Spencer has perceptively suggested that most of the alleged solutions to the lurking dangers of polytheism proffered by Social Trinitarians fail in the final analysis.[18] The reason is that they rely on an “as if” reading of the members’ unity for their success: while the Trinity is conceived of as a divine society, the persons are treated as if they were one, based on such features as interpenetration, absolute mutual devotion, or an immediate, non-inferential awareness of each other’s internal states.[19] However, these appear to be mere verbal solutions, which seek to paper over the metaphysical gap between multiple selves existing, say, in a state of perfect harmony, and a being that is genuinely unified in nature.
Consider Richard Swinburne, the widely regarded Christian philosopher (and avowed Social Trinitarian). On his functional monotheistic view, three divine beings sharing one common, abstract nature exist together of logical necessity, thereby constituting an indivisible collective that is the source of all things.[20] Swinburne suggests that what he calls G1 actively (and timelessly) causes G2 to exist, thereafter permitting the second being’s ongoing existence; together, they then co-operate in bringing about G3.[21] But as certain philosophers have observed, this is logically compatible with three discrete individuals enjoying the same essence and a complete concord of wills.[22] There may not even be anything inherently impossible about three separate immaterial beings experiencing a deep kind of co-inherence or psychological interpenetration. Other models, like that of Keith Yandell, appear to be afflicted by similar defects. He argues that Father, Son, and Spirit each possess a set of “G” properties, sufficient for divinity. In order to individuate the persons, Yandell contends that all such properties – including those that appears to be identical – are bearer-specific. But he also insists that the persons cannot logically exist apart from each other, enjoy complete volitional unity, and possess non-inferential awareness of each other.[23] However, if critics are right, then he has also failed to provide justification for why the three should necessarily exist in a relationship of metaphysical dependence, or why their co-existence should entail a genuine, ontological unity. They remain three ultimate principles, whose status as discrete beings is logically consistent with what Yandell has advanced.[24] Ultimately, Spencer is correct: advocates of ST relying on such strategies to burnish their monotheistic credentials are not dissimilar to a person who fashions a megagon and calls it a circle.[25] Since this kind of polygon is not truly circular (regardless of appearance), it remains a mere approximation; the qualitative fissure that exists between the two shapes cannot be bridged by pretence or linguistic veneer.
Craig’s proposal, by contrast, seems to come closer to something resembling authentic metaphysical unity. His contention that the members of the Trinity function analogously to “parts” in one larger entity underwrites their essential oneness far more readily than the main rivals to TM. This is buttressed by his model: Father, Son, and Spirit are personal “elements” of the one divine soul, thus enjoying a kind of organic solidarity that does not rely on mere comity or volitional harmony. Craig’s position is reinforced by the apparent fact of his seeing the divine nature in concrete, rather than universal or abstract, terms. On a universal view, it is possible for a certain nature or cluster of properties to be exemplified in, or borne by, discrete tokens.[26] Humanity, for instance, is a universal that is instantiated by any number of separate individuals. This, too, would leave one’s Trinity theory vulnerable to charges of tritheism, where each person exemplifies the divine nature as an individual item. Craig’s model operates with a single trope of divinity, which seems to avoid the problem of multiple exemplification to which universal properties inevitably lead.[27] This would allow him at least to blunt the “three Gods” threat, thereby gaining an advantage over certain other iterations of ST.
I fear, however, that Craig may have won a Pyrrhic victory: securing the essential unity of the divine persons by way of his proposal comes at a cost. In what follows, I will attempt to show just how exacting that cost is.
Assessing Trinity Monotheism: The Formal Theory
We arrive now at the heart of my essay: a critical assessment of Trinity Monotheism. Does Craig’s scheme offer a satisfying version of Trinitarianism? Is it logically sound and internally consistent? And does it successfully observe some of the basic constraints of Trinitarian orthodoxy? Unfortunately, my fundamental answer to these questions is “No”. For ease of analysis, I will divide this section into two main segments: Craig’s formal theory; and the model proper.
Contending with creeds, sparring with Scripture
I start with Craig’s theory, which seeks to explain the divinity of the persons by way of part-whole relations. I suggest that it is riddled with a number of serious flaws; for the Christian, the degree to which it fails to properly cohere with orthodoxy is of utmost concern. Indeed, one must contend with the fact that Craig’s view of the persons’ divinity conflicts sharply with both early church councils and biblical texts. This is most obvious when it comes to his contention that the persons, while divine, are not properly God. The second council of Constantinople (AD 553), for instance, declares that there is “one God, even the Father of whom are all things…”. Judging by this particular statement, the authors of the creed were seemingly convinced that the Father, far from only being God predicatively, was to be identified with the ultimate principle of all creation. The Nicene Creed is also quite clear, enjoining Christians to “believe in one God, the Father almighty…”.[28] This is quite similar to the Apostles’ Creed, which declares belief “in God, the Father…”. Their basic coordination of the two figures in question appears to be unambiguous: the Father is God.[29] It is difficult to see how these clauses, occurring in creeds that have won near-universal acceptance, can be reconciled with Craig’s claim that the Father is “merely” divine by predication.[30]
Craig suggests that ecumenical councils, while important, should not set the bar for what counts as licit theological statements; that norming role, of course, belongs to Scripture. But while this may be true, his view encounters further problems, with the divergence between TM and the biblical narrative reflected in several key New Testament texts. Let two passages stand for many. First Corinthians 8:6a declares that there is “one God, the Father (cf. John 17:3a)”. This, too, seems to be saying that God is the Father. But if Craig is right, then Paul’s claim is incorrect, and the Apostle was wrong to apply the label “God” to this particular member of the Trinity. On Craigian Trinitarianism, we may recall, one would be forced to say that the Father is not, strictly speaking, God. One possible alternative is to interpret the clause predicatively: where “God” is used, Paul would simply mean something like “divine”. But this will not work, for the position of “God” in that clause suggests it is being used in nominal, not predicative, fashion. And why does the Apostle also say “one God…”? This would imply the existence of numerically one being possessing divinity – the Father himself. What, then, of the Son and Spirit? According to Craig, there are three divine beings, who help compose the triune God. However, if his claims concerning predication are accepted, then it seems that 1 Corinthians 8:6 must be read as referring to only onesuch individual. Anyone wanting to uphold orthodox Trinitarianism would be right to resist such bizarre interpretations.
Craig might respond that such an argument can be turned against a prima facie reading of the passage: if the one God is the Father, how do we account for the divinity of the other two members? Traditional conceptions of the Trinity can appeal to the idea that the Father is the principle of deity; the Son, meanwhile, proceeds from the Father as the eternally begotten Logos.[31] Whether or not this is ultimately satisfying, such a course represents one way of reconciling a surface reading of 1 Corinthians 8:6 with dogmatic claims concerning the Son (and Spirit). Craig, however, cannot make use of it, given his rejection of relations of origin within the Godhead.[32]
Second, consider Colossians 2:9, where Paul affirms that “in Christ all the fullness of the Deity lives in bodily form”. This clearly seems to set forth the claim that the divine essence resides completely in the man Jesus – i.e., that he is God in physical form or “solid reality”.[33] If certain commentators are correct, and this represents an expansion of Colossians 1:19, then the point is simply underlined: the plenitude of God, the origin and guarantor of all things, has filled Jesus Christ to the uttermost.[34] At the very least, this clashes with Craig’s insistence that the Son-Logos (who he would agree became enfleshed in Christ) is not the whole God, but “merely” one part or element of that larger entity. Could this really be the case if, as Paul declaims, the earthly Son bore fully the nature of the Creator himself?[35] That would introduce a strange contrast, since Jesus would have possessed the divine nature in greater measure during his worldly sojourn than did the Son-Logos in his pre-incarnate state. And of course, “God” in this passage cannot be read as the entire Trinity, for this would lead inevitably to heterodox positions concerning the various operations ascribed to the divine members. Patripassianism, or the view that the Father himself suffered in Jesus’ passion, is only the most obvious entailment of such a view. It seems that a text such as this should be interpreted as saying that the Second Person of the Trinity bears fully the divine nature. Craig’s theory, then, is questionable on biblical-textual grounds.
Mereological Objections to TM
Quite apart from exegetical problems, TM raises a clutch of philosophical concerns, traceable largely to the unusual manner in which Craig distinguishes the Godhead from Father, Son, and Spirit. I shall begin with faults relating to mereology (or the study of part-whole relations), which afflict his formal theory in a variety of ways.
Undermining God’s aseity
We might firstly reflect on the implications of Craig’s claim that some kind of part-whole relationship obtains between the persons and the Trinity. A mereological approach suggests that God is a composite of (at least) three “elements”: Father, Son, and Spirit. One key worry is related to the idea of ontological dependence, since God’s existence would seem to hinge on those elements. According to a major stream of classical mereology, composed objects are metaphysically posterior to their parts, which maintain them in being. Borrowing the language of contemporary metaphysics, any such item is grounded in its constituents.[36] Grounding denotes a state of non-causal dependence between multiple items; in the present context, it may be closely related to the idea of fundamentality, which here refers to God’s status as the ontological bedrock underlying all other realities and states of affairs.[37] For something to be fundamental just means that it is grounded in nothing else.[38] The trouble with saying that God is grounded in his constituents, however, is that it conflicts with his allegedly fundamental being. In particular, the claim compromises his aseity, or the conviction that he relies on nothing that is not himself for existence.[39] According to theologians holding to versions of divine aseity, God is self-sufficient and independent, which comports neatly with his bearing fundamentality. It is arguable, however, that a mereological God cannot bear this property, thereby failing to possess the kind of ultimacy fitting for the Deity.[40] It might be possible to argue that God depending on his parts does not violate the notion that he exists a se; because Father, Son, and Spirit are internal to the Godhead, reliance on them does not imply ontological dependence in the way that a creature relying on various external conditions would. One problem with this response is that it does not take account of the peculiar approach that Craig adopts, wherein he makes a clear metaphysical distinction between the persons and the entire Godhead.[41] He argues that although the Trinity is God the persons are “merely” divine by predication, while his division of attributes – some possessed by the whole Godhead, others “lodged” in the persons – suggests a similar difference. It seems, then, that God relies on components that are, strictly speaking, distinguishable from himself.[42]
But perhaps my objection is gratuitous: why assume that God is dependent on his parts because he is (as Craig’s theory suggests) complex? The unstated premise here is that wholes are always reliant on the elements that compose them.[43] However, it is arguably not always the case that a whole is posterior to its parts. Grounding relations may travel in the other direction: an entity could be ontologically prior to its constituents, whose existence is guaranteed by the totality. Instead of the entire Trinity relying on the persons, they would be grounded in the divine whole. And indeed, many of Craig’s explicit statements can be read as implying this very claim: the persons are divine because they are parts of the entity called “God”, finding their identity within it; they possess certain incommunicable attributes in virtue of the entire Godhead having them; and, to anticipate my evaluation of Craig’s model, the persons are likened to centres of consciousness, supported by the divine soul.[44] Even so, these points are belied by elements of Craig’s mereological approach to God, which strongly imply that he must, in some sense, rely existentially on his parts. Since for Craig the persons bear certain great-making attributes primitively (e.g., omnipotence, omniscience), it is only on their account that God himself does. Moreover, triunity – which Craig argues is integral to God’s instantiation of the divine nature – only appears possible because of the conjunction of the three persons. It means that key aspects of God’s essence remain grounded in his personal constituents – a state of affairs that, somewhat incongruously, ends up undermining it.[45]
Could recourse be made to some notion of metaphysical interdependence? On this view, a totality and its parts are bound together in a profoundly mutual way, such that there exists a relationship of reciprocal metaphysical reliance. The whole depends on its parts for existence, but the parts also rely on integration within a broader system for theiridentity and persistence in being.[46] By analogy, God and the persons could be said to exist in a relationship of ontological interdependence, where both “parties” rely on each other. Unfortunately, Craig will find no refuge here. He explicitly states that substances are ontologically prior to their constituents. While this conflicts with some of the already-noted entailments of his formal theory, it nevertheless rules out any notion of metaphysical interdependence from the outset. But even if such a position were permissible, problems would remain. For example, contemporary defences of mutual dependence theory also tend to deny notions of well-foundedness (e.g., the claim that any item, X, is either grounded in a fundamental entity, or itself bears fundamentality).[47] What this means is that neither the whole nor its parts can be ontologically basic. Such a move may be tolerable for contingent entities embedded in wider networks of existence and being; it is far less attractive when applied to an allegedly ultimate (i.e., fundamental) entity.
Failing to shore up the persons’ divinity
So much for God’s ultimacy. A more immediate and obvious objection to Craig’s formal theory is that it fails to provide sufficient logical warrant for the persons’ (full) divinity.[48] Recall that for him, the persons are divine because they are parts of the Trinity, just as (e.g.) a cat’s DNA is feline because it is part of the overall animal. But this does not seem to furnish a robust foundation for Craig’s claims. To see why, we might explore first the notion of transitivity. If Craig is right, and the persons are fully divine by virtue of parthood, then the nature of transitive relations implies that their parts (e.g., the Father’s cognitive or affective faculties) are as well.[49] This seems deeply implausible, however: those faculties do not exemplify all the great-making attributes one normally ascribes to divinity, such as aseity, omnipotence, omniscience, or unsurpassed moral goodness. Nor do they have eternity per se, even though they are constituents of purportedly eternal beings. It is therefore not at all clear why mere parthood should be deemed adequate to uphold the divinity of the three persons.
Craig is liable to raise two counter-objections at this point. First, he has insisted that some mereological analyses limit proper parthood to a certain “level” within a composite entity: a door may be part of a house, for instance, while the door’s handle is merely part of the door (not the dwelling as a whole).[50] Conceiving of the Trinity as an individual composed of individuals – where “individual” stands for something that is neither disguisedly plural nor mass – Craig argues that “parthood is not transitive across types of composition”.[51] He contends that since the Father himself is an individual while his faculties are not, parthood cannot devolve further to include the latter; although they may be a part of him, they are not, strictly speaking, a part of God. Only three “elements” – Father, Son, and Spirit – could be legitimate candidates for “unambiguous” divinity, for they are the only genuine components of the Trinity.[52]
Second, one might suggest that to count as a part in the relevant sense, something must make a “direct, functional contribution to the whole”.[53] Another term of art is “determinate function” – i.e., the defined, circumscribed purpose something may have, which limits the scope of mereological relations. A handle may make an immediate contribution to the proper functioning of a door, but not to the house in which the door is located. Similarly, while a set of cognitive and volitional faculties might contribute to the centre of consciousness Craig calls the Father, its role within the entire Trinity is indirect – mediated by the First Person, just as a door mediates the role a handle plays within a particular house. By another (though complementary) route, therefore, God’s authentic parts would be limited to Father, Son, and Spirit.
How should one respond to these rejoinders? It appears, firstly, that in some cases, there is nothing inherently invalid in saying that parthood is transitive across levels of composition.[54] So long as care is taken to specify the functional domain of a certain part (where that domain can be variable in scope), presuming the transitivity of the part-whole relation may be permissible.[55] While “the house has a handle” seeks to draw an untenable inference from the knowledge that a particular door has a handle, there appears to be nothing intrinsically wrong with saying “the house has a doorhandle”.[56] Formulating this kind of relationship via statements in the form of “X is a part of Y” would be even less problematic, since they do not carry the implication that Y is the functional domain of X; saying that a doorhandle is a part of the house, while a little unusual, remains technically correct.[57] And in the case of a metaphysically integrated entity like the Trinity, where the relevant constituents are held together necessarily, it is difficult to see why someone should object to such a construction. Contrary to Craig, extending the part-whole relation to an item’s sub-parts is not always illegitimate.
Of course, Craig is likely to still insist that because the Father is an individual but his faculties are not, transitivity cannot proceed, for different types of composition allegedly renders this an illicit move.[58] That appeal, however, would trigger pressing questions regarding consistency: after all, Craig has contended that both an entire cat and its distinctive parts are “fully and unambiguously” feline. This looks suspiciously like an attempt to traverse different types of composition: a cat is a good example of an individual (as used by Craig), even though its DNA and skeleton are not. Compositional distinctions, however, have not prevented him from saying that the latter bear the property of felinity just as much as the former. Granted, the parallel he initially sought to draw was between a cat and the entire Trinity, not the individual persons. But that is irrelevant to the matter at hand, which concerns the legitimacy of positing an extension of mereological principles – and any attendant properties – across various types of constitution. There appears to be no good reason to restrict this to a particular level of parthood, especially if sameness of composition does not impose a limiting principle. Craig is therefore faced with a dilemma. On the one hand, he can maintain the cogency of his cat analogy. Such a move would, however, compel him to concede that the divine persons are in principle susceptible to a similar procedure, even though they and their faculties are not of the same compositional type. That, of course, would lead him into the absurd position of saying that the persons’ attributes are just as divine as God is. On the other hand, he can continue to claim that “assuming…transitivity across different types of”[59] constitution is an illicit move – thereby rendering his cat analogy wholly untenable.
As for the second rejoinder – that direct functional contribution circumscribes the notion of proper parthood – I have already indicated that in some cases, the domain of a specific component within a larger whole is not necessarily a barrier to transitivity. Handles can still be said to be parts of houses, even if their role is determinate. Moreover, it is arguable that on TM, at least some divine properties do seem to play an immediate role in the proper “functioning” of the entire Trinity. Craig himself suggests that the whole Godhead, not the persons, bears necessity, aseity, and eternity in a primitive sense; such attributes help form the Deity’s nature in a direct, non-mediated way. While it might seem odd to think of such attributes in terms of parthood, a mereological model of God – one that seems to implicitly permit the use of mereological language across types of composition – leaves the door open to this concept. In any case, for all their importance as “components” of the Deity, necessity, aseity, and eternity are hardlydivine in the full-orbed sense. On the part-whole theory of TM, however, they would be just as divine as the persons or the Trinity.
One might object that these properties do not play a functional role within the Godhead; unlike an “active” attribute like omnipotence, they are static or inert in nature. However, this severely limits the concept of “functional” beyond its common meaning. A particular part may indeed operate by changing or actively maintaining the state of some object (e.g., a piston firing in a car to generate locomotion; a pair of lungs that permits regular respiration). But it may also lay down the conditions for other parts to function as required. It is arguable that a divine individual could not be omnipotent if it did not have eternity; the power possessed by such an entity would be contingent, subject to temporal constraints. Indeed, its occupancy within the bounds of space-time would see it subject to the vicissitudes of the historical process. Only an eternal, self-subsistent being can coherently be said to exercise unlimited power. In a broad sense, then, something like eternity plays a crucial functional role in upholding God’s omnipotence, and does so in the direct sense Craig deems important. But can it really be divine in the same way the persons are? Our initial contention therefore seems sound: mere parthood is insufficient to guarantee the persons’ divinity.
Diminishing the persons’ divinity
Craig’s mereological approach lacks the logical force necessary to guarantee the persons’ divinity, while also generating a series of conceptual confusions. And yet, his formal theory also has an actively deflationary effect on the nature of the persons – something that has been dubbed the problem of “diminished divinity”.[60] However exalted they may be, Father, Son, and Spirit bear this status in a downgraded sense when compared with the entire Trinity. To some extent this mirrors my earlier criticism concerning God’s ontological reliance on the persons, an implication Craig has seemingly failed to recognize. And yet, his theory also seems explicitly to require the subordination of Father, Son, and Spirit to the Godhead – a move that can only issue in a depreciated divinity, despite his insistence to the contrary.
Why might this be the case? For starters, it is entirely unclear how something may be fully divine (or human, ursine, planetoid, etc.) without instantiating that nature.[61] Craig’s contention that there are two ways of being divine is arguably a mere verbal solution to Leftow’s charge. That is, he takes two different states of being – instantiation and parthood, respectively – and insists that they can underwrite the same nature. It may be an ingenious linguistic strategy, but it also seems to gloss over a disjunctive definition of what it means to be (in this case) divine. Although the same term has been applied, the fact remains that it has been deployed to cover what appear to be very different conceptual zones. Craig, of course, thinks that his cat analogy provides a satisfying answer, illuminating what it means for something to be of a certain kind, whether through instantiation or parthood. To be sure, there is a genuine sense in which the characteristic features of a cat are feline (they do not belong to dogs or horses, after all). But mere predication cannot but have a depreciating effect; the part in question is what it is only “adjectivally”, depending on membership within the entire composition. Instantiation, by contrast, suggests a kind of metaphysical abundance that a thing’s components lack.[62] Indeed, within the logic of Craig’s scheme, an object exemplifying a certain essence and being a member within that same object do not seem to have the same ontological status. How, then, can they offer access to the same nature, and to the same degree?
It is worth examining Craig’s analogy in more detail. I have my doubts about its usefulness, although some objections are weaker than others. For example, while it is true that the atoms of a cat are not feline in the same way that its DNA or skeleton are,[63] one can simply restrict the theory to a thing’s proper or distinctive parts.[64] This would instantly rule out parts like atoms, since they are not unique to any one organism, having the capacity to “migrate” between material objects. But consider yet another part of a cat: its ears. They seem uniquely feline, both morphologically and physiologically, properly belonging to members of the domestic cat family. Indeed, they are perfectly fashioned to enable the animal to hear even the softest of sounds – important for the life of a (would-be) predator. However, it still seems hard to credibly maintain they are feline in the same way, and to the same degree, as the entire cat. For all the adaptive brilliance they exhibit, a cat’s ears cannot pounce, climb, catch rodents, or see well at night. These are properties and powers characteristic of felines considered as entire biological systems. That in itself appears to suggest a qualitative difference between the whole entity and a particular element composing it. It also suggests that the felinity of a cat’s ears, skeleton, or DNA are circumscribed, just as the divinity borne by Father, Son, and Spirit would be circumscribed if viewed through the lens of parthood. The suspicion that Craig’s analogy is finally unilluminating – and that his mereological approach actually diminishes the divinity of the persons – continues to linger.
Of essences and properties
This brings me to a related way of elucidating the problem. Consider the connected ideas of essences and properties – for example, the nature or essence of a human being. According to some conceptions of humanity, we are rational animals.[65] From that basic nature flow powers like language, a sense of humour, a capacity for narrative construction, abstract reasoning, and so on. Now, if Craig’s cat analogy is valid – and behind that, his part-whole approach to the persons’ divinity – then a human’s skeleton and DNA are, mutatis mutandis, “fully and unambiguously” human. Moreover, they are just as human as the entire human organism, albeit in a different way. But here is the rub. A human’s skeleton and DNA do not reason; nor do they use language or engage in personal I-Thou relationships. Only the entire human being bears and engages those properties. Even if an advocate of TM argued that such parts make a direct functional contribution to a properly functioning human, it is still the case that their powers and capacities are restricted in a way that the entire organism is not.
The crucial point is that properties are an important element in distinguishing what an object is – that is, identifying its nature or essence.[66] If something does not exhibit what are held to be the requisite attributes for a particular item,[67] then we have reason to question whether it is truly and wholly what it is claimed to be. In the case of TM, Craig singles out triunity as that property which distinguishes the entire Godhead from its members;[68] the implication of his view is that only the whole Trinity bears divinity in its fullest sense, for it alone has the attribute of tri-personality.[69] While it is arguable that God’s triune nature depends on the conjunction of the persons (which, as I have argued, undermines his ultimacy), the particulars of Craig’s theory introduce a qualitative distinction between him and the divine members. This means that despite Craig’s protestations, the persons, for all their splendour, are saddled with a kind of shrunken divinity – excluded from the fullness of that nature because they lack a key distinguishing attribute of Godhood. Indeed, if they do not have triunity, we are compelled to ask whether, on Craig’s account, they can be unambiguously and unqualifiedly divine.
But isn’t triunity essential to the divine nature?
At this point, Craig might insist that his proposal naturally proceeds, at least in part, from the prima facie plausible claim that triunity is an essential part of the divine nature; a distinction must therefore be made between the persons, who allegedly do not possess it, and the entire Trinity. To be sure, this fits well with basic Christian intuitions about God’s identity. Surprising though it may seem, however, we have reason to resist the notion. This requires some elaboration. First, while an attribute is held necessarily, it does not follow automatically that said property plays a constitutive role in a thing’s essence. As some critics have alleged, arguing otherwise is logically fallacious,[70] which would mean that triunity is not an essential part of the divine nature. Second, arguments purporting to demonstrate that tri-personality is qualitatively similar to features traditionally held to establish divinity are unpersuasive. Justin Mooney, for example, suggests that it underlies God’s maximal greatness, just as the properties of (e.g.) omnipotence and aseity do.[71] But while tri-personality may be superior to uni-personality, his construal presumes a “pro-social” form of the attribute (i.e., “loving, co-operative communities”), begging the question in favour of ST.[72] Furthermore, if divinity means maximal greatness – that is, the very summit of magnificence – then why would the existence of three persons within the Godhead be adequate? If a loving society per se is better than a solitary monad, would not, say, six, 13, or 745 persons be superior to a trio? A unified triad might get you some way towards a maximally great God, but the logic behind this argument suggests that there is nothing uniquely superlative about the number three. Such observations indicate that drawing a contrast between the Godhead and the persons on the grounds of triunity are somewhat shaky.
But let’s concede that there is some merit to Craig’s contention regarding tri-personality and the divine nature. Unfortunately, this would merely change the shape of the problem for him: saying triunity is constitutive of divine nature jars with other intuitions that exert an equally potent claim on the Christian’s acceptance. I have already suggested that embracing such a position invariably commits a person to different grades of divinity within the Godhead – intolerable for anyone wishing to affirm the full deity of the persons.[73] It also appears to conflict with creedal and biblical statements (e.g., Colossians 2:9) that are seemingly untroubled by the notion that the persons can instantiate the divine nature – quite apart from questions of triunity. The Gordian knot could perhaps be cut, but in ways that are barred to the Trinity Monotheist. Michael Willenborg, for instance, argues that the persons themselves are triune, possessing a common nature that includes existence within a relationship of triadic metaphysical dependence.[74] The Son would be triune in that he has a nature, possessed alike by Father and Spirit, and only exists if the other two persons do.[75] If successful, this would allow someone to embrace the essential character of triunity and the full divinity of Father, Son, and Spirit. Craig, however, cannot resort to Willenborg’s proffered solution, for the distinction between the divine members and the entire Godhead – under which lies the attribute of triunity – is essential to his theory. Remove it, and the entire philosophical structure collapses.[76] But nor can he credibly maintain the distinction, for the reasons already adduced.
A two-tiered Deity
Michael Rea has made a perceptive observation that dovetails neatly with the preceding lines of argument. He agrees that Craig’s theory effectively posits not one, but two types of divinity: the primitive nature, instantiated by the Trinity as a whole; and the divinity of the persons, which is derived or mediated.[77] This is certainly hard to reconcile with Craig’s subsequent claim that divinity is a single property that something can possess “as a consequence of various factors”. If indeed we have multiple, yet unequal, forms of the divine nature, then the insistence that only one such type of divinity exists cannot easily be maintained. But more importantly, a derived property or status is incompatible with full and “unambiguous” divinity, precisely because it depends on something else – in this case, the entire Trinity – for its actualisation. The problem of mereology and divine ultimacy has returned, only in reverse: Father, Son, and Spirit bear several of the great-making attributes essential to deity, but in derivative fashion. They do not seem to be fundamental, for what they are is grounded in the Godhead; if divinity is associated with being metaphysically basic (something that does not appear possible on Craig’s model, in any case), then the persons do not qualify. Yet again, this appears to be a kind of circumscribed or diminished deity, at odds with what is normally taken to be essential to theistic conceptions of God.
It might be possible to blunt the objection by arguing that the Trinity just is the three persons; on this reading, no substratum exists, acting as an ontological guarantor for their divine status. But as I have already observed, the strong impression one acquires from reading Craig’s reflections on the Trinity is that such a foundation does exist (e.g., God as a soul-like substance supporting Father, Son, and Spirit).[78] To posit that nothing divine exists over and above – or “beneath” – the persons would bring the theory into conflict with the model proper. But then if my contention is valid, it means that that the persons are metaphysically rooted in that underlying substance for their ongoing existence. Any lingering trace of aseity would dissolve entirely. It is therefore difficult to avoid the conclusion that this segment of TM entails two, unequal modes of divinity: one pertaining to the entire Trinity; and a lesser form, borne by the persons.
Assessing Trinity Monotheism: The Model
I turn now to the second half of Craig’s proposal, which concerns the model proper that he has developed for the Trinity: God as a soul-like substance supporting three sets of cognitive faculties, each sufficient for personhood. We have seen that Craig’s formal theory suffers from several defects. How does the model itself fare? I will outline a variety of criticisms, which I submit are overwhelming for its credibility.
Is Craig’s model metaphysically possible? Assessing the Analogies
Lurking beneath Craig’s model is the pressing concern that it may not be metaphysically viable – or at least not in any way that we can grasp conceptually. This will become clearer as we make our way through a series of proffered analogies.
William Hasker, who argues for a view similar to TM, has rendered the model in the form of a theoretical proposition:
Trinitarian Possibility Postulate: it is possible for a single concrete divine nature – a single trope of deity [i.e., God as a soul-like substance] – to support simultaneously three distinct lives, the lives belonging to the Father…the Son, and…the Holy Spirit (qualification in parentheses mine).[79]
This is where Craig’s now-infamous Cerberus analogy becomes relevant, which he thinks can show us how his model might be true. Recall that for him, Cerberus is a single, concrete trope of “dog-ness”, which exemplifies canine nature. Rather than supporting one canine consciousness, however, he is in possession of three states of dog-like cognition, volition, and affectivity.
The fundamental issue with Craig’s preferred analogy is that it thoroughly fails to illuminate his model, or show how it might be possible for one entity to support, sustain, or uphold three centres of cognition, will, etc. Consider the related – and real-world – example of dicephalic parapagus twins, who share one body below the neck. If Cerberus is really one instantiation of “dog-hood” with three centres of canine consciousness, then such twins would have to be considered similarly. But Brittany and Abby Hensel, perhaps the most well-known example of dicephalic parapagus twins alive today,[80] appear to be two distinct tropes of humanity, albeit sharing largely the same physical structure. They have their own personalities, aspirations, mental states, and ways of relating to other people. In short, the Hensels seems to exhibit all the trappings of what we would normally associate with distinct human individuals. If intuition is a reliable guide, it suggests that we are dealing, not with one case of humanity supporting two human-like sets of consciousness, but two human beings.[81] And because there is no salient difference between the Hensel twins and Cerberus, we should therefore regard the mythological guard-dog of Hades as three instantiations of “dog-hood”. Deployed consistently, Craig’s analogy yields the unwelcome conclusion that there are three gods, or three exemplifications of divinity, within the Trinity. This would not be triunity, but tritheism.[82]
Craig might respond that while Brittany and Abby are distinct persons, there exists only one human being in their vicinity. Multiple sets of consciousness, mental states, or personalities are consistent with this view, which means that more than one instance of humanity is not logically demanded. Mutatis mutandis, the same could be said of Cerberus: while there are three canine “persons”, there is only one instantiation of a dog. That might suffice to show how a divine soul could “support” three divine persons, without the latter devolving into three separate exemplifications of divinity. Such a move remains unconvincing, however. Embryological studies suggest that parapagus twins and dicephalic (“two-headed”) animals[83] are the result of incomplete twinning or embryonic fission: that is, where an embryo fails to properly divide. But behind the phenomenon lies a plurality of organisms, even if the process of bifurcation was stymied by certain factors in utero. Such organisms tend to have more organs than would be required for just one body, which may also suggest that if the pre-natal developmental process had continued unimpeded, two morphologically distinct exemplifications of (e.g.) humanity would exist. As such, these unusual occurrences do not seem to be genuine parallels to Craig’s Cerberus analogy, which remains unilluminating regarding his model.[84]
Grappling with some bizarre consequences
Pressing these analogies also leads to certain worrying metaphysical implications. Imagine, for example, that one of Cerberus’ canine members, Fido, is surgically separated. He still exists, even though he is not identical to the Fido that was part of Cerberus. The reason he is not identical is that on Craig’s reading, he was not a dog in the first place – merely a canine “person” that helped constitute one complete instantiation of “dog-hood”. This seems very strange, for it commits us to saying that the present animal is not the same entity as his pre-operative predecessor. Meanwhile, his present metaphysical status remains entirely unclear: is he a newly-formed substance, or merely a discarded “component” of the larger Cerberus organism? The only way, it seems, of avoiding such a dilemma is to say that Fido was and is a dog, before and after surgery.
Craig claims to have unearthed absurdities flowing from the aforementioned thought experiment, thereby nullifying it as an objection to his chosen analogy.[85] Suppose, for example, that Cerberus is a three-headed wolf. On the view propounded here, he is actually a set of three conjoined wolves. Now suppose once more that Fido is surgically separated from the rest of the organism, but is reconstructed using parts from a border-collie. Is he, then, a kind of wolf/border-collie hybrid? Or is he still a wolf made to look like a border-collie? And what was he prior to surgery? Although Craig thinks he has offered a sound rejoinder with these questions, there is reason to doubt his conclusion. For consider Fido, not as a member of Cerberus, but as a single, “stand-alone” canine (e.g., a wolf). Now imagine that due to injury or disease, Fido loses several body parts. Somehow, he has survived, with veterinary surgeons reconstructing him using parts from a border-collie. What breed of dog is the newly-minted Fido? Is he some kind of hybrid? And if so, what was he prior to surgery? I cannot detect any relevant difference between this scenario and the one Craig has offered – and yet, no one would dispute the claim that before the procedure, Fido was a wolf. More to the point, I do not think that anyone would argue with the contention that he was and is a single instantiation of “dog-hood”, despite the paradoxes yielded by this set of circumstances. Craig’s attempted rebuttal, I submit, is unsuccessful.
In any case, an implied Craigian interpretation of real-life cases generates similar absurdities. Consider again the Hensel sisters. If dicephalic parapagus twins truly are one exemplification of humanness (as the multi-headed Cerberus is allegedly one instantiation of “dog-ness”), then given the absolute nature of identity relations,[86] neither Abby nor Brittany is identical with the composite organism. But then it seems that we have two metaphysical classes of people populating the world: conjoined twins, neither of whom can be identified with a particular human being; and “ordinary” individuals, who can be so identified. There is, in other words, a specific kind of distinction between person and human being in the former case that does not obtain in the latter case. But how plausible is this sort of division? Moreover, how would we account for the fact that members of one alleged metaphysical class (i.e., the Hensel twins’ parents) somehow generated members from another class entirely? By contrast, simply thinking of such twins as two (conjoined) human beings – something that the logic of Craig’s Cerberus analogy prevents him from doing – would see this conundrum dissipate.
Finally, consider the troubling consequences Craig’s construal would have for the very nature of the Hensel “organism”. Because he is logically committed to the notion that only the total entity would instantiate humanity, his view falls prey to some acute inconsistencies. We might recall what I said earlier about human nature, and how it has often been conceived – i.e., as rational animality. Whether one accepts this precise formulation, it seems clear that what marks out humanity as a species, as well as a metaphysical “item”, is its capacity for rational thought. And as embodied individuals, we rely in some sense on our brains to exercise that capacity. But in the case of the Hensels, those brains – and the minds with which they are associated – are the possession of Brittany and Abby, not the organic whole. Moreover, the Hensel “organism” cannot be identical with either of its personal constituents, given classical identity relations.[87] Strictly speaking, it fails to bear the faculties necessary for rational thought, which means that it cannot (fully) instantiate human nature[88] – a conclusion that is sharply at odds with what Craig needs to claim. In the final analysis, both Cerberus and conjoined twins remain unyieldingly opaque as analogies, giving us little independent reason for thinking that the model underlying TM is sound.
Split brains and multiple personalities: some augmented analogies?
Philosophers and theologians sympathetic to Craig’s scheme have proposed different analogies, drawn from the worlds of neuroscience and psychiatry, in an attempt to demonstrate its intelligibility. Hasker has explored the possibility that medical phenomena like commissurotomies and multiple personality disorder (MPD) offer bona fide examples of several, distinct selves sharing a common substance (i.e., a human body).[89] In the case of commissurotomy (split-brain) patients, the brain’s two hemispheres have been severed, often in an effort to control seizures. Controlled experiments have yielded some startling results. In one example, the words “key” and “ring” were projected onto a screen, such that “key” appeared in the subject’s left visual field (and therefore transmitted to his right hemisphere) and “ring” in the right visual field (with corresponding transmission to the left hemisphere). Asked what he observed, the subject answered “ring” without knowing what kind of ring it was. Directed to point with his left hand (controlled by the right hemisphere) to what he saw, he pointed to a key, not a ring.[90] Some have interpreted such data as evidence for the emergence of two sets of (proto)-consciousness in one individual. For advocates and allies of TM, this may provide warrant for thinking that multiple and co-existing conscious agents can be grounded in the one exemplification of a certain nature.
For all its apparent plausibility, however, the analogy fails to throw sufficient light on Craig’s model. While researchers have gathered some fascinating data from such studies, philosophers and neuropsychologists have recently argued that the so-called “social ordinariness” of split-brain patients was the most common experience post-surgery.[91] That is to say, such individuals acted in a socially common way, with little or no indication of distinct streams of consciousness or global psychological bifurcation. Even clinical experiments themselves have generated ambiguous results: some patients evinced a marked division between the hemispheres during controlled activities, while others displayed a clear capacity for inter-hemispheric communication.[92] The “availability of content” afforded to split-brain patients tends to be chaotic and fragmented, with individuals frequently appearing to reach for phenomenal data in an uncoordinated manner across the two hemispheres.[93] This is not the kind of neat division between clearly demarcated mental “workspaces” one would expect if commissurotomies issued in the generation of distinct centres of consciousness – precisely because the force of the analogy depends (at the very least) on such “siloed”, independent nodes.
Even if the contents of one’s phenomenal field appear to be partitioned or incohesive, this does not necessarily entail a division into multiple conscious subjects or perceivers.[94] Indeed, experimental conditions still suggest that the same self is engaged in the task at hand, despite apparent conflict implying nodal fracturing within his or her consciousness.[95] This reflects the difference between what some philosophers have termed experience-based consciousness and agency-based consciousness: while an advocate might argue that the former can be divided, it still seems to be the case that a single, unified agent underlies those experiential fragments. Again, the social ordinariness of most split-brain patients tends towards an interpretation of behavioural unity, which persists despite possible fluctuations in perceptual integration.[96] It would therefore be very difficult to conclude that two discrete conscious subjects had arisen within the one individual human being.[97] As Hasker himself has conceded, “there is little to recommend” the argument that distinct persons develop as a result of this procedure – the very phenomenon that would be required to make commissurotomy patients true analogies to Craig’s “pro-social” model.[98]
As for those afflicted by MPD, it is claimed that two or more personalities inhabit the same body; the original personality has very often become divided, leading to parts that are “separate aggregates, each with its own memories, [forming] the nucleus for new, independently functioning constellations”.[99] Now, advocates have marshalled impressive evidence, supplemented by suggestive data gleaned from changes in localised brain activity,[100] that could imply the presence of co-occurring conscious agents in an individual suffering from MPD. This includes people testifying to being “intra-conscious” of their alters – experiencing intimate awareness of their thoughts and actions – and even a kind of inter-personal interaction with these identities.[101]
MPD could therefore bring us closer to illumination of Craig’s model than any other analogy on offer. But crucial questions loom: does the presence of certain “alters” really signal the emergence of multiple persons within the one human being (necessary for the phenomenon to act as a credible analogy to TM)? Or is it the case that individuals afflicted with this condition are instead suffering “merely” from the effects of one fragmented psyche? In fact, the phenomenon of multiple personalities faces potent problems. Even if a person’s alters are considered authentic psychic entities, the notion that we are dealing with a single fractured personality – where the various states are regarded as “parts” or “shards” of the basic ego – may ultimately be more plausible.[102] One key reason is that mainstream therapy with MPD patients continues to aim at re-integration of the various personalities that have been hived off the primary identity; that this remains the goal of clinical and psychological treatment implies the professional judgment, not that distinct personal agents inhabit one body, but that the individual in question suffers from a disintegrated personality, manifested in the appearance and evolution of semi-autonomous identities. The fact that a number of people afflicted with MPD have been able to achieve a healthy degree of psychic re-unification lends some weight to this position.[103]
Even on stronger interpretations of the available evidence, there are several crucial dis-analogies between the condition and Craig’s model. First, most advocates admit that the alters of an MPD patient are not fully-fledged persons in the way that two separate human beings are.[104] If they were, then the desired outcome of therapy with sufferers would literally end the lives of other people – in other words, homicide.[105] It is exceedingly difficult to see how anyone might accept this as an interpretation of therapeutic re-integration, regardless of how vivid said personalities are.[106] Furthermore, if such a goal is in any way possible, then it already suggests that we are not dealing with distinct subjects; after all, would it be possible for separate agents to undergo a process of therapeutic “fusion”? A psychological chimera seems implausible in the extreme (and would not be open to Craig, given his commitment to the divine members’ irreducibly first-person perspectives).[107] The problem, of course, repeats the one encountered in our exploration of split-brain patients: by failing to rise to the level of full personhood, the alters in an MPD patient do not constitute a suitable analogy for TM.[108]
Second, there is some evidence that what we might call subsidiary personalities are actually parasitic on the “host”, who remains the subject throughout one’s experience of MPD.[109] The alters, for all their vitality and uniqueness, seem to be at times ephemeral and intermittent, relying on the primitive personality as a kind of ontological anchor. But the members of the Trinity, at least on TM, are not related to each other in this way at all. It is not as if the Son depends on the Father as host – drawing his life from him as an epiphenomenon – nor the Spirit on either of them. And it will not do to say that the divine members are somehow parasitic on God as a whole. Aside from the bizarre spectacle of insisting on such a relationship between the Godhead and the persons, this would turn the Trinity into an unorthodox Quaternity: God, the primary or host identity; and Father, Son, and Spirit as subsidiary personalities, who exist in a co-occurring (though utterly dependent) relationship with him.
To be sure, a transcendent God, unbound by time and space, can never be fully captured by our mundane categories. But despite Craig’s insistence that TM provides an intelligible account of the Trinity, it seems that the analogies featured here ultimately fail to illuminate, thus leaving interested readers in a place of unrelieved agnosticism. Meanwhile, the metaphysical interior of the model remains utterly mysterious. Craig, for his part, insists that any analogy is simply a jumping-off point for reflection on his model, and has lambasted critics for apparently “fuss[ing] terribly” over the examples he offers.[110] But if neither Cerberus, nor the Hensel twins, nor split-brain patients, nor even sufferers of MPD, function as cogent parallels for TM, how can they be profitably deployed? What do they illuminate precisely? What are they launching pads for if they repeatedly thrust the advocate into a thicket of theological or philosophical problems? And if there are no other alleged analogies in the extant literature, what should we conclude about the purchase Craig’s model has on concrete reality?
Doubting the Coherency of Craig’s Model
Following in the wake of the above discovery is the disturbing realization that TM may not be metaphysically conceivable. Still, Craig and others might wish to endorse his model on the grounds that it allegedly satisfies orthodoxy and is not manifestly incoherent (even if one is not quite certain how it could possibly be true). Such a course would represent a retreat from the bolder claims he has made for its intelligibility. Worse still, close examination of the model calls into question any decision to embrace it. I turn now to concerns regarding Father, Son, and Spirit, exploring problems internal to Craig’s construction, as well as the tensions it generates with other aspects of his thought.
Are the divine persons substances?
An immediate worry concerning Craig’s model is that his position on the metaphysical nature of persons conflicts with his implied account of Father, Son, and Spirit, all while leaving their status entirely opaque. In describing the divine members as individual centres of will, cognition, and consciousness, Craig openly likens them to persons, akin to human beings. One would expect nothing less, given his avowed Social Trinitarianism. Moreover, Craig argues that persons in general are paradigmatic examples of individual substances (other typical cases being trees, dogs, or planets).[111] He even elucidates a set of features something must manifest for it to qualify as such:[112]
- The bearer of properties, even as nothing can bear it as a property.
- Said properties and capacities form a tight unity. The parts are what they are by virtue of the role they play in the whole, apart from which they would not exist.
- Persistence or sameness through change.
- Growth and development in a regular, law-like fashion, in accordance with what it is. There is an in-built tendency to realize any potentialities contained therein.
- Membership within a class of natural types, where each member has the same essence or is of the same kind.
- Bearing a principle of individuation, which distinguishes the entity from other entities that share its nature – i.e., something in virtue of which it is this X and not that X.
What happens when we apply these criteria to, say, the Father?[113] He seems to satisfy (1)-(3), as well as (6). Some might baulk at saying that the Father satisfies the sixth criterion, on the grounds that he cannot exist apart from membership within the Trinity. Craig himself suggests that the persons do not have the “stand-alone” quality of genuine substances.[114] Perhaps. But whatever principle of individuation one employs – and Craig seems to think the members can be distinguished according to the trappings of individual personality – it is clear that the Father is neither the Son nor the Spirit. Indeed, while they exist together, their complete individuality – not as mere relations, but as fully-fledged personal agents – is crucial to Craig’s model. As for feature (5), it is hard to deny that the Father is a member of a certain class of things, united by a common “whatness” – namely, divine personhood, and all the properties that flow from it. Craig denies that the persons of the Trinity instantiate an essence, but unless one is prepared to say that there is nothing in particular distinguishing Father, Son, and Spirit from other “items”, then it appears that they are constituents of a natural cohort. Consequently, the only possible feature the Father might fail to satisfy is (4), for it could be said that he is not subject to growth or perfection. But if this is true of the divine members, then it must also be true of God as a whole: he has no need of growth or development, for he is already complete. That God cannot meet all six criteria, however, does not seem to trouble Craig, since his account hinges on the notion that the Deity is a (primary) substance. As such, (4) may not be necessary when identifying individual substances, which means that Father, Son, and Spirit would qualify after all.
This is all well and good. However, according to Craig’s model, only God as a whole is a substance. He denies that the divine members are eligible, for on his account (whether as “parts” or as nodes of consciousness grounded in a soul-like substance), they do not enjoy a mode of being as fundamental as that of the entire Godhead. Craig cannot admit that the persons bear the metaphysical status of substance, since that would leave him with a surplus of entities: God as a whole, as well as Father, Son, and Spirit. Such a Quaternity is manifestly unorthodox, although it is difficult to see how this can coherently be avoided, given Craig’s prior metaphysical commitments. He therefore faces yet another intractable dilemma: accept the logical implications of his own view of substances and concede that the divine members satisfy the relevant criteria; or maintain his model. The former would see Craig move beyond Christian orthodoxy, while the latter would mean being at odds with himself.[115]
Again, Craig makes much of the claim that a true substance must have a “stand alone” quality, and likens Father, Son, and Spirit to hands: while a hand has a high degree of integrity, and could be said to have a kind of nature (from which certain properties flow), it is not a genuine substance in its own right.[116] There are at least a few problems with this rejoinder. First, if this kind of free-standing characteristic is essential to being a substance, then not even God himself would qualify on TM. For as I have already argued, Craig’s mereological theory (inadvertently) entails God’s dependence on his parts for both existence and exemplification. He would therefore fail to be a “stand-alone” entity, contrary to what Craig might insist. Second, it seems illegitimate to compare the persons with hands, for the latter are clearly mere instruments. Hands are used by the persons who have them; they do not possess independent agential power but are employed by those who do. The divine members, however, seem to bear such capacities; after all, it is not as if they are driven or directed by the divine substance. Granted, none of the members exercises power independently, if by that one means “unilaterally”. The conviction that Father, Son, and Spirit always operate in complete unison has long been part of the structure of orthodox Christian belief.[117] However, such agency remains qualitatively different from a hand: the latter has no self-determining or self-generating causal power, and only “acts” in so far as its owner does. One might continue to insist that neither a hand nor the persons can exist apart from the entities of which they are parts. This may be true, but it brings us back to the first objection – namely, how Craig’s compositional approach to God undermines ontological autonomy as a criterion for substantial existence.
Third, Craig’s reference to Father, Son, and Spirit having “sufficient integrity” to possess “natures” seems somewhat vague.[118] We still want to know what they are in a basic metaphysical sense. What are the persons if, on Craig’s reading, they are not individual substances? While they do not seem to be mere instruments (like hands), Craig is reluctant to commit himself to the logical entailment of his view. He gestures towards the idea that the divine members have natures/essences, but this is hardly a clarifying statement. On his account, only God as a whole has a nature; the persons may be divine, but only as elements within the Trinity. How, then, can they have an essence as well? And even if this were possible what, according to TM, might it be? The sense of confusion is only deepened when Craig, trying to elucidate his position, suggests that Father, Son, and Spirit are divine in virtue of being “God’s persons”.[119] I have already highlighted the problems associated with hinging divinity on something more fundamental than those who allegedly bear it. Beyond that, to claim that Father, Son, and Spirit are “God’s persons” (where one interprets that clause in a possessive sense) is to cast the Deity as primary subject. The members’ status, meanwhile, is shrouded in a metaphysical fog, which thickens when Craig simply calls them God’s “cognitive faculties”.[120] Persons just seem to be subjects: individuals possessing a first-person point of view, and capable of entering into I-Thou relationships. Craig would likely want to agree, given his declared commitment to a version of ST. But as I shall note in more detail below, mere “cognitive faculties” do not seem to rise to the level of personhood.
Ambiguities such as these have an obfuscating effect on Craig’s model. Does it mean that TM entails four substances? If so, then we begin to drift back towards a quartet of divine entities. On the other hand, if the persons are not themselves subjects, but merely modes or facets of God (who alone is true subject), then we have abandoned TM altogether, trading it for a radically one-self theory of the Trinity. Inherent doctrinal problems aside, such a move would conflict with the rudiments of Craig’s social model.
Persons, souls, and the Incarnation
There is another incongruity pertaining to Craig’s account of the divine persons, this time dovetailing with his construal of the Incarnation. Christian orthodoxy holds that around 2,000 years ago, the Second Person of the Trinity became incarnate, being joined with a human nature to form the man, Jesus of Nazareth. Craig adopts what he calls a neo-Apollinarian reading of the doctrine, in which the soul of the Son-Logos completes the human nature of the individual in question, thus realizing the emergence of a genuine person.[121] The keen-eyed reader is likely to be left wondering how these claims cohere: after all, on TM it is God as whole, not the persons, who is the soul-like substance. The Son-Logos is, by contrast, only one node of consciousness within the Godhead. How, then, could he furnish the man Jesus – himself an individual substance – with a rational soul?[122]
To be sure, this may not be a problem for other advocates of TM, so long as they are not committed to Craig’s Christology. How might Craig himself deal with the inconsistency? First, he could equivocate on the concept of “soul”. In the case of God, the soul is the mental substance capable of sustaining a coherent set of cognitive, affective, and volitional faculties. But how could the term be defined differently without suggesting a complete disjunction in referent? Indeed, in what way could the notion be applied to both the Godhead and the individual persons, if not to denote the same type of object and its capacities? One could argue that what helped compose the person of Jesus was a kind of sub-soul, which itself is one of God’s parts. But this would just see Craig collide with a more general problem, one that I will canvass below: how the Son can be a fully-fledged person without also being a fully-fledged soul. In any case, if the Son-Logos were not a rational soul in a sense similar to Craig’s usage, it is unclear how he might undergird the qualities necessary to guarantee the genuine personhood – including all the extant powers of reason and cognition – of the man Jesus.
Second, Craig could try and argue that the entire Trinity was united with a human nature to form Jesus, such that he would bear a rational soul. It should be manifestly clear why this is untenable. The idea that all three persons became incarnate in Christ leads, of course, to a variety of thorny theological problems. Not only does it invite the heresy of, say, Patripassianism; it would also make a nonsense of certain biblical texts (who, for example, was speaking from the heavens at Jesus’ baptism if all three persons were embodied in him [e.g., Mark 1:11]?). That leaves only the third alternative: admit that the Son-Logos is, in fact,a soul. But this, too, appears to be proscribed. Embracing it would destroy the coherency of Craig’s model, for the only entity that can be a soul is the Trinity itself. Only by revising his Trinitarianism or his Christology can he hope to remain (theo)logically consistent.
Of souls and support: some logical and conceptual problems
The problems of souls and divine personhood bleed into issues concerning the various concepts Craig uses to illuminate relations between the entire Trinity and the persons. To be sure, certain critics have misfired in their objections to this phase of his proposal. The term “support”, for example, has been censured for its apparent unintelligibility.[123] Such an objection is unnecessarily obtuse, however: Craig’s meaning seems perfectly clear, in that he intends something like “maintain in being”. The primary idea is that God – this soul-like entity – acts as the metaphysical basis underlying the three divine persons. His insistence that the word means the whole Godhead is “explanatorily prior” to Father, Son, and Spirit seems to bear this out.[124]
But while this complaint may have been successfully rebutted, it hardly renders Craig’s model sound. That is because some of his language and analogizing also appear to imply that where the Trinity is concerned, the divine being or substance can “be” three psychologically distinct agents.[125] For Craig, God is a mental substance – in other words, a richly endowed rational soul. Rational souls, meanwhile, are equated with persons, which Craig thinks is confirmed by our long acquaintance with other human beings.[126] He then compares the Trinity with what obtains between a human person and her soul: the one divine being supports three persons, “just as” an individual human soul supports one person.[127] But as Howard-Snyder asks, what does it mean, say, for my individual soul to “support” my person? On Craig’s view, it appears that I am absolutely identical with my person, although I am also absolutely identical with my being/soul. By the transitivity of identity, then, the particular soul-like substance in question (i.e., me) is absolutely identical with the person in question (also me).[128] This has some very concerning implications for Craig’s model: if we are to understand the model in the same way we understand relations between a human person and her soul, “then God, this single composite item, is absolutely identical with the three [members]”.[129] However, this seems logically false, for identity relations are usually seen as one-to-one – i.e., symmetrical – not one-to-many, in nature.[130] For X to be identical with Y, both items need to bear all properties in common; but this does not appear to be true of the entire Trinity and the persons, given the numerical differences involved.[131] Of course, Craig implies that God and the persons are distinct: notions of “support” suggest an asymmetric relationship between the divine members and the soul-like substrate. Yet somewhat confusingly, his framing of the matter also amounts to the claim that one item is indistinguishable from three items – a position which seems flatly incoherent, as well as incongruent with other elements of his model.
Now, Craig cannot say that souls and persons are distinct, for the metaphysical picture he has painted renders this an impermissible move. Could he appeal, then, to a notion like Composition as Identity (CAI) for clarification? Advanced by some scholars sympathetic to Trinitarianism, CAI holds that a conjunction of parts is, in certain contexts, identical to the whole thus composed.[132] They have then argued that because Father, Son, and Spirit compose the Trinity, they are constitutionally identical to it. However, several problems with this alleged solution spring to mind. First, I have been at pains to point out that Craig’s model creates a metaphysical distinction between the persons and the divine soul. While this conflicts with his broader claims concerning such concepts, it means that merely conjoining Father, Son, and Spirit (as per CAI) would be insufficient for achieving the desired degree of identity. Second, Trinitarian proponents of CAI neglect the enduring differences between the persons and the entire Godhead, even if its application to TM were legitimate. For example, there are certain properties the Trinity has which Father, Son, and Spirit lack.[133] The Trinity is triune; according to Craig’s theory, the divine members are not. Similarly, only the Trinity instantiates divinity, while Father, Son, and Spirit are merely divine by predication. Certain aggregates of beliefs as they apply to the persons and God, respectively, also imply a real distinction in characteristics. In a theoretical possibility left open by Craig’s construal of the “siloed” nature of the persons’ mental states,[134] the Father may affirm A and not B, the Son B and not C, and the Spirit C and not A. Consequently, the Trinity would “affirm” A, B, and C, in contradistinction to its three constituents. Contrast this with Craig’s human analogy: it seems undeniable that if my person affirms A that I myself affirm A, thus preserving identity. Instances such as these indicate that CAI probably cannot relieve the tensions resident in Craig’s portrayal of the Deity. And while he avers that “God, though one soul, would…be three [persons]”, it is doubtful whether he can coherently hold this position.[135]
These concerns reflect a basic instability in Craig’s model. Recall that for him, rational souls are substances with mental properties, while human souls are deemed to be identical with persons. Such statements – and the positions lying beneath them – cannot be easily reconciled with the way he conceptualises God or the divine members. For as we have also seen, Craig does not think of the divine persons as substances. Nor can they themselves be souls, since on his view the underlying divine substance is the only “soulish” element in the vicinity of the Trinity. And yet, they are the tri-fold loci of the Godhead’s mental properties. How can these various positions be reconciled, given Craig’s metaphysical priors? Indeed, how can Father, Son, and Spirit be persons without the quality of “soul-hood”, especially when Craig thinks of souls as entities with intellect and rationality? If a human soul is identical with a human person, then it seems reasonable to think of divine souls as identical with divine persons. Rational souls just are persons on Craig’s account. Yet worryingly, it is not at all clear that this is possible for the members, such are the constraints of TM: there is only one soul within the Trinity, which is the divine substance.
The minority reading of TM could see God modelled as one being enjoying a tri-fold mental life.[136] Again, this is implied (perhaps unwittingly) by some of Craig’s statements, casting the persons as God’s cognitive faculties.[137] He would then be a single intellectual substance, refracted through three dimensions of the one psychological state, with the identity of souls, personhood, and mentality preserved. That would be enough to remove some of the inconsistencies I have just noted. But as we have seen, interpretating Craig’s model in this way is highly problematic. For then Father, Son, and Spirit would no longer be persons in the true sense of the term, but merely mental aspects within the divine self/soul we designate “God”. The latter would be the personal substratum underlying the three members, who may be seen as his powers. However, while I myself may be endowed with the integrated capacities of rational thought, intentionality, and affectivity, such faculties do not rise to the level of personhood; only the total entity (i.e., me) can be thought of as a person. Consequently, the robustly “pro-social” commitments of TM would be violated, for interpersonal interaction, so essential to Craig’s model, simply cannot occur within oneself; it requires distinct and fully-orbed personal subjects capable of dialogic intercourse. And it can only be repeated that anyone appealing to this reading of TM would sail dangerously close to a one-self or modalistic conception of God, with the persons reduced to aspects, functions, or “life streams” of the entire entity.[138] It seems, then, that Craig’s model is both structurally unsound and irreconcilable with other aspects of his thought.
TM and God’s non-personhood
Cutting across much of the preceding discussion is perhaps the most remarkable entailment of TM: namely, that on the majority (and more plausible) reading of Craig’s model, God – this soul-like substance – is not a person.[139] Whatever other elements within the model imply, logical analysis of the relations between the members and the Godhead suggests the latter does not bear true personhood.
How could this be? It is important to note that Craig, like all good Trinitarians, is keen to avoid a Quaternity, and must do so to remain safely within the bounds of orthodoxy. God as a whole cannot therefore be a person in addition to the divine members, since this would lead to four such beings within the vicinity of the Trinity. We might also recall that as a champion of ST, Craig is committed to the full personhood of Father, Son, and Spirit. His model therefore casts them as three selves within the Godhead, bounded personal agents possessed of consciousness, self-reflection, rationality, and will. But where does that leave God? Once more, he is not identical with the persons: as I have already argued, positing this would be logically incoherent,[140] for one personal agent cannot be equivalent to, or indistinguishable from, three such individuals.[141] Arguing that God is identical with one of the divine members is also unsatisfactory, since this would mean that (e.g.) the Father is God, while the Son and Spirit are, say, ancillary nodes of divine “consciousness”. Aside from generating yet more absurdities (God having a primary consciousness as well as supplementary sets of such faculties), this move irretrievably diminishes the divinity of the Second and Third Persons – another sure route towards heterodoxy.
The only discernible alternative, then, is to say that on Craig’s model, the Trinity is not identical with any of the persons – which is another way of saying that God as a whole is not a person.[142] He may be a society or warm collective of personal beings – even beings joined together as an ontological unity – but he himself cannot bear the same status.[143] While Father, Son, and Spirit might enjoy a high degree of cohesion on Craig’s model, this comes at the cost of depersonalizing the total entity. To be sure, I am not referring to any particular philosophical account of personhood (e.g., a kind of Cartesian res cogitans), nor to a crude, anthropomorphised version of the concept, but to any individual that can (e.g.) credibly use first-personal pronouns, bears something analogous to rationality, and participates in an I-Thou relationship.[144] Such appears to be the portrayal of God in Scripture and tradition. But if the present line of criticism is correct, then it has catastrophic consequences for Craig and other advocates of TM.[145]
For one thing, a non-person does not have the capacity for rationality or intentional action; at a stroke, omniscience and omnipotence disappear as properties the entire Trinity can properly exercise. It also suggests that strictly speaking, God cannot do anything normally ascribed to divine individuals. Take creation: on TM, God as a whole could not be said to fashion the world. Creation is an act of rational, intentional power, requiring foresight and strategy. But a non-personal God cannot act, nor plan, nor intend; the persons may do this, but the numerical distinction between them and the entire Trinity on which Craig insists presents an insuperable barrier. Similarly, God in his entirety could not set about redeeming his people if he did not have the trappings of personhood. Redemption, like creation, is planned, organised, and deliberate. It also has as its end the re-forged relationship between God and the community of faith – something that seems to be possible only for true persons.[146]
One might also point to the vast array of statements in the Old Testament, especially in the prophetic corpus, wherein the divine voice is used. Whether in judgment or in mercy, the divine “I” is frequently employed by Israel’s God, as he reveals himself to his wayward people (e.g., Isa 45:5-7; 46:4; Jer 2:1c; 30:22; 31:1-2, etc).[147] But if God is not a person (or personal in a substantive sense), then the reams of material one finds in the body of written prophecy constitute grievous misrepresentations of the divine nature.[148] Indeed, since it appears that on TM he is (logically) deprived of personhood, how could he participate in an interpersonal relationship with his people? To rebut this objection, some advocates have appealed to the idea of corporate personality as a way of trying to show how the ancient Israelites could think of an assembly of individuals as a unified agent.[149] On this view, the God of the OT may be construed as a corporate person, capable of participating in I-Thou relationships akin to a single subject. But as a plausible model for early Jewish and Old Testament thinking about God, this remains deeply controversial, having been severely criticised on numerous historical and methodological grounds.[150]
As if all this were not calamitous enough, we confront a yet more fundamental issue: anything that is not a person has no credible claim to godhood. This is deeply dissatisfying, to say the least. On the one hand, Craig wants us to say that only the entire Godhead exemplifies the divine nature. But on the other hand, his model entails that he, God, does not possess a key feature of the divine essence. How, then, could the whole Trinity instantiate the nature that is claimed for it?[151]
Preserving the personhood of Craig’s God: some failed strategies
There are several possible rejoinders to this criticism, although I would argue that they all fail to assuage one’s doubts. Craig protests that the view just outlined has been cashed out by his critics in a highly “tendentious manner”, apparently relying on the belief that God must be a unitarian Deity.[152] But as Howard-Snyder and other critics observe, this is not a numerical claim about God, but pertains to the proper locus of personhood: if God as a whole cannot be identified with any one of the three centres of consciousness, then how can he genuinely be called a person?[153] They might bear all the trappings of personal agency, but the logical demands of identity relations means that he does not.[154] Craig has also complained that while the tradition has always maintained that God is personal, it has not held that he is a person.[155] However, this is a rather narrow, unnuanced reading of Christian doctrinal and intellectual history.[156] Modern Christian philosophers, for example, have leaned on the notion of God’s being a person.[157] And thinkers from earlier eras – even those like Aquinas, who stressed the deep, ontological differences between God and his creatures – were apparently comfortable with judicious application of the word “person” to the Deity.[158]
A respondent might argue that Craig’s Trinity, while not a person in the strict sense of the term, is so by analogy. In the case of God, we are compelled to use such language, given the metaphysical gulf lying between him and the creation.[159] Hasker thinks that God as a whole partially corresponds to what we normally think of as persons, given he is “composed” of three personal agents. He goes on to suggest that analogical notions are satisfied by treating God as if he was a person.[160] Lamentably, he has misapplied the lessons of analogy: the term “as if” is not analogical language, but that of mere appearance. Analogy is only applicable when the property (or properties) in question can truly be predicated of both parties to the alleged parallel. A good wine and a good man bear a certain quality of goodness, for example. Both items meet certain standards of excellence and approbation, even if what they satisfy is not exactly the same. The key point, however, is that any correspondence between a man and a bottle of wine hinges on genuine possession of goodness.[161] But if the present criticism of TM is valid, then God as a whole does not actually bear personhood; the divine members enjoy that property, but as I have noted, they are bounded conscious subjects. And since God as a whole is metaphysically distinct from Father, Son, and Spirit – and therefore not identical with any of them – he technically remains impersonal. In fact, by using terms like “as if”, Hasker implicitly concedes the point at issue, attempting to gloss over the non-personhood of TM’s God via linguistic pretence.[162]
It could also be tempting to suggest that even if the entire Trinity is not strictly a person, it could “borrow” properties constitutive of personhood from the divine members, in a manner similar to other composite forms.[163] Thus, God redeemed the multitudes because Father, Son, and Spirit were so determined. But while such an entity may bear certain qualities as a result of its constituents doing so, this is only the case when it has an antecedent capacity to do so.[164] Unless God possesses the prior ability to (e.g.) act deliberately and rationally – i.e., in a way one normally ascribes to a person – then it appears that he cannot perform those activities, even derivatively. That lack is complemented by the aforementioned observation concerning the circumscribed personhood of the divine members, which means that transitivity of the relevant properties is blocked. This simply underscores the gravity of the problem: the Trinity not only fails to fully instantiate the divine nature (not being a person); it is incapable even of drawing those powers and qualities from its members.[165]
God the group agent?
At this point, Craig could lean on an intriguing proposal developed by Chad McIntosh: the Trinity as a functional person.[166] An advocate of ST, McIntosh has conceded that many social theories of the Trinity are vulnerable to the charge I have been laying out. He therefore argues that the triune God may be seen as a person – functional yet genuine – distinct from the divine members. To do this, he appeals to the notion of group agency, which may be seen as a contemporary philosophical analogue to the allegedly ancient Hebrew idea of corporate personality. McIntosh contends that groups or collectives can exemplify and manifest the features one normally attributes to authentic individual agents (holding representational states, being committed to certain goals and positions, etc.), distinct from the members that compose them.[167] McIntosh deploys this as the basis for his version of group personhood, arguing that while the divine members are persons intrinsically (what they are by nature), the entire Trinity bears this property functionally – i.e., by what it does and how it acts.[168] And if certain ordinary groups can be considered persons, then the triune God – whose internal relations are much more tightly interwoven – would undoubtedly qualify.[169]
McIntosh’s account is certainly an ingenious one. But whatever merit the general concept of group personhood has, I do not think it can rescue Craig’s model. Several reasons spring to mind. The idea of group personhood as applied to TM seems to gloss over several substantive distinctions between the Trinity and its members. McIntosh himself acknowledges that his proposal means that they are qualitatively different. For social models like TM, Father, Son, and Spirit are three genuine selves, who can credibly use the word “I” and enjoy an integrated, first-person point of view (characterised, for example, by psychological interiority and irreducible subjectivity).[170] The overall entity, for all its claimed agential powers, lacks that same quality of selfhood; it has no internal mental life, for the divine members bear that capacity in a substantive, circumscribed (i.e., non-shareable) fashion. One might well ask where, if a first-person outlook is essential to true personhood, is that property “lodged” (so to speak) in God as a whole? There appears to be an unbridgeable metaphysical gap between God and the divine members, casting doubt on whether McIntosh’s account can be used to underwrite the former’s personhood – at least in the psychologically robust way that both modern laypeople and advocates of ST deploy the term.
But let’s say that God can legitimately be construed as a group person. It is then arguable that McIntosh has tried to have things both ways. He helps himself to claims of authentic personhood when necessary, before retreating to alleged differences between the Trinity and its members when his argument confronts unwelcome implications. Positing God as a (functional) person alongside the members seems to threaten a Quaternity of divine individuals. Alive to this concern, McIntosh insists that no such threat exists, because God, unlike the members, is not a hypostasis.[171] That may be, but if functional personhood is genuine personhood, then I do not see why hypostatization should be relevant; there would still remain four personal beings within the vicinity of the Trinity, which is one too many for Christian orthodoxy. Indeed, McIntosh argues for possible differences in beliefs between the members and the entire Godhead. While we may again wonder how the latter’s mental states are grounded, such a position would be intelligible only if we were dealing with a distinct individual.[172] And if being a hypostasis is germane to the discussion, then this arguably draws us back to the first objection. My point is that one cannot consistently argue that God satisfies the relevant conditions of personhood, while trying to resist heterodox conclusions by implying that this is qualitatively different from the kind borne by the divine members.
Furthermore, the idea of God as a corporate person, at least as McIntosh conceives of it, faces the sorts of mereological problems I have already raised. He suggests that God supervenes on the three divine members; in other words, he is existentially underlined by them. But as I observed earlier, God depending on his parts – a relationship that would also include his personhood – subverts his aseity and undermines the principle of divine fundamentality. Even if McIntosh’s account of the Trinity’s personhood was cogent in itself, Craig cannot coherently apply it to his own model, given the former’s insistence that the persons underwrite the Godhead.[173] The internal structure of Craig’s model is explicitly ordered in the opposite way, since the divine substance is “explanatorily prior” to Father, Son, and Spirit. And because the sets of relations propounded by McIntosh and Craig both appear to be uni-directional – albeit running along contrary routes – they cannot be made to properly gel. Ultimately, McIntosh’s proposal is of no help to Craig, who must make do with an entirely ersatz conception of God’s personhood.
Conclusion
William Lane Craig’s study of the Trinity, though less extensive than his treatments of other subjects, bears all the hallmarks that characterise his work: historical depth, theological sensitivity, philosophical erudition, and a steadfast refusal to rest easily on the authority of received tradition. Trinity Monotheism is a highly original attempt to resolve the many conundrums raised by the notion of a triune God, seeking to move beyond creedal formulae to provide an intelligible account of the Deity’s inner life. Moreover, TM seems to succeed where other forms of ST fail, underpinned by a stronger principle of divine unity than many of its theological cousins and forebears. Where many social models of the Trinity are bedevilled by the spectre of tritheism, Craig’s proposal can at least appeal to a genuine, metaphysically robust conception of God’s oneness.
But TM is not immune to criticism, as my essay has sought to show. If I am right, then it is vulnerable to a clutch of serious objections, touching on almost every element of Craig’s proposal. The mereological theory he deploys to explain the divinity of Father, Son, and Spirit is insufficient for the task. It cannot provide logical warrant for the persons’ divine status, with consistent application yielding the patently implausible conclusion that (e.g.) various attributes pertaining to God share the same property. This is additional to the theory’s deflationary effect on the divine members; despite Craig’s protests, one cannot avoid the overriding suspicion that it tacitly entails two tiers divinity. On this, critics like Daniel Howard-Snyder and Michael Rea are correct: the persons only enjoy a diminished divine status. Craig, of course, has insisted that his preferred cat analogy shows how something can be of a certain kind without exemplifying the relevant nature. But analysis of the alleged parallel simply underscores the problems at hand; the unfortunate conclusion is that Father, Son, and Spirit are divine only in attenuated fashion.
This is a reminder of just how mysterious Craig’s proposal is – a curious feature for something claiming the mantle of conceptual lucidity. For as we have also seen, all the available analogies in the extant literature fail for one reason or another to illuminate the model proper. Underlying those failures is the basic problem (as yet unresolved) of how three psychologically distinct selves can nevertheless constitute, or be supported by, one trope of the divine nature. And while it may be initially conceded that an advocate could still cling to the notion that TM is neither manifestly incoherent nor obviously heterodox, further scrutiny casts doubt on that position. From the metaphysical status of the divine persons to the consequences flowing from Craig’s use of soul-language to describe the Godhead, his model faces numerous – and in my view, insuperable – objections. The question of how the divine members can be persons without being substances lingers unanswered, as does the issue of how God can be a soul-like substance (and a mental substance at that) without also being a self. All this is to say that the various elements of Craig’s model are ill-fitting, which suggests that it is ineluctably incoherent. That is reflected in what is possibly the coup de grace for TM: the claim that the model entails God’s non-personhood. Indeed, while it may be surprising to some, logical analysis leads inevitably to the conclusion that on Craig’s view, God is not a person. In isolation, this criticism would be sufficient to render TM a dubious proposition; combined with the other issues identified in this essay, it simply completes the critical examination to which Craig’s views have been subjected.
Debate concerning the Trinity will no doubt persist unabated. But if my study of TM has contributed anything worthwhile, then it has exposed certain problems concerning one line of enquiry. Meagre though this might be, such an offering may help in channelling the wider discourse.
[1] See Daniel Howard-Snyder, “Trinity Monotheism”, in Thomas H. McCall and Michael C. Rea (eds.), Philosophical and Theological Essays on the Trinity (Oxford: OUP, 2009), 100-125.
[2] Dale Tuggy, “Trinity”, S.2, SEP (2009; rev. 2020), plato.stanford.edu/entries/trinity/.
[3] In addition to the “pro-social” Trinity theorists discussed below, see Plantinga’s classic statement. Cornelius Plantinga, Jr., “Social Trinity and Tritheism”, in Cornelius Plantinga, Jr., and Ronald J. Feenstra (eds.), Trinity, Incarnation, and Atonement: Philosophical and Theological Essays (South Bend: Notre Dame University Press, 1989), 21-39. For a critical summary of ST, see Brian Leftow, “Anti-Social Trinitarianism”, in Stephen T. Davis, Daniel Kendall, SJ, and Gerald O’Collins, SJ (eds.), The Trinity (Oxford: OUP, 1999), 203-204.
[4] Rea, “The Trinity”, in Thomas P. Flint and Michael C. Rea (eds.), The Oxford Handbook of Philosophical Theology (Oxford: OUP, 2009), 420.
[5] J.P. Moreland and William Lane Craig, Philosophical Foundations for a Christian Worldview (Revised; InterVarsity Press: Downers Grove, 2017), 589. Hereafter, I shall refer to Craig only.
[6] Leftow, “Anti-Social Trinitarianism”, 221; Craig, “Towards a Tenable Social Trinitarianism”, in McCall and Rea, Philosophical and Theological Essays, 95.
[7] Craig, PFCW,588.
[8] Craig, PFCW, 589-590.
[9] Craig, PFCW, 589.
[10] Craig, PFCW, 590.
[11] Craig, PFCW, 590.
[12] Craig, PFCW, 590.
[13] Craig, PFCW, 592.
[14] Craig, PFCW, 592.
[15] Craig, PFCW, 592-3.
[16] There is another, minority interpretation of Craig’s model, suggesting that the entire Godhead is the primary self, with the members as his cognitive faculties. This stems from some of Craig’s language, which can be ambiguous. However, his overriding commitment to ST strongly implies that we ought to see the persons as psychologically integrated individuals within the one Godhead. In any case, I submit that both readings are problematic.
[17] Craig, PFCW, 593.
[18] Daniel Spencer, “Social Trinitarianism and the Tripartite God”, RS 55 (2019): 194-195. Plantinga (“Social Trinity and Tritheism”, 39) offers an example of this phenomenon: God is one, since there is only one divine family or community. How is this meaningfully different from a family of distinct individuals?
[19] Spencer, “Social Trinitarianism”, 194-195.
[20] Richard Swinburne, The Christian God (Oxford: OUP, 1994), 180-181, 185.
[21] Swinburne, The Christian God, 172-177.
[22] Edward Feser, “Swinburne’s Tritheism”, IJPR 42 (1997): 178ff.
[23] Keith E. Yandell, “The Most Brutal and Inexcusable Error in Counting? Trinity and Consistency”, RS 30 (1994): 201-217.
[24] See William Hasker’s criticisms of Yandell: Metaphysics and the Tri-Personal God (Oxford: OUP, 2012), 158-161; “The One Divine Nature”, Theologica 3 (2019): 60-61.
[25] Spencer, “Social Trinitarianism”, 194. A megagon is a polygon with a million sides. To the naked eye, it can appear as a circle.
[26] This is analogous to humanity as an abstract universal being instantiated in three separate individuals. The danger of polytheism is not hypothetical: Prestige documents the drift towards an abstract view of the divine ousia in the sixth century, which led to, in his words, an “outbreak of tritheism”. See G.L. Prestige, God in Patristic Thought (London: SPCK, 1952), 272-273.
[27] That does not augur the end of the issue, since it remains an open question whether a concrete or particular property (such as the one trope of deity) is shareable among multiple and distinct things in the first place (e.g., the persons of the Trinity on ST). I remain neutral on that issue for present purposes, but see Tuggy, “Hasker’s Quests for a Viable Social Theory”, Faith Philos. 30 (2013): 173.
[28] J.N.D. Kelly, Early Christian Creeds (London: Bloomsbury, 2006), 369.
[29] Craig argues that many post-Nicene creeds bear traces of certain philosophical accretions (e.g., a drift towards DDS) that should make us question the identification of God with, say, the Father. It seems, however, that the phrase “God the Father” was commonly used in the post-apostolic church, well before DDS could exercise influence over creedal development. For numerous examples of “God the Father” appearing in early Christian writings (with slight variations), see Marcel Sarot, “Believing in God the Father”: Interpreting a Phrase from the Apostle’s Creed”, Hervormde Teologiese Studies 72 (2016): 1-4. Kelly, Early Christian, 133: “In the literature of the second and following centuries, ‘God the Father’ is so regular a description of the Deity that quotations illustrating it are superfluous”.
[30] We might also consider debates around the Athanasian Creed. Wierenga, a ST advocate, argues that with the Latin word deus (= “God”),the framers of that creed intended to declare that Father, Son, and Spirit are divine (i.e., bearing the property of divinity), not “God”. See Edward Wierenga, “Trinity and Polytheism”, Faith Philos. 21 (2004): 281-294. But Wierenga’s reading is strained: if the creed’s authors had wanted to say this, they would likely have opted for divinus instead. See Jeffrey E. Brower, “The Problem with Social Trinitarianism: A Reply to Wierenga”, Faith Philos. 21 (2004): 297-298.
[31] Louis Berkhof, Systematic Theology (London: Banner of Truth, 1948), 98.
[32] E.g., PFCW, 585, 593.
[33] N.T. Wright, Colossians and Philemon (TNTC; Leicester: IVP, 1986), 103.
[34] Wright, Colossians and Philemon, 103.
[35] Should Colossians 2:9 be read ontologically or functionally? Blackwell argues convincingly in favour of the former interpretation, concluding that Christ bore the divine essence in himself. This cannot be reconciled with Craig’s claim that the Son does not instantiate divinity. See Ben C. Blackwell, “You are Filled in Him: Theosis and Colossians 2-3”, JTI 8 (2014): 105-110.
[36] Joshua R. Sijuwade, “Building the Monarchy of the Father”, RS [nv] (2021): 4-5, 7.
[37] Sijuwade, “Building the Monarchy”, 5.
[38] Sijuwade, “Building the Monarchy”, 9.
[39] Tuggy, “On Counting Gods”, Theologica, 1 (2017): 195-196; Matthew Baddorf, “Divine Simplicity, Aseity, and Sovereignty”, Sophia 56 (2017): 409-410.
[40] This conflicts with Craig’s contention that the whole Trinity possesses attributes like aseity in a primitive sense: if God is reliant on his constituents, then he seems only to have it derivatively.
[41] See Tuggy, “The Unfinished Business of Trinitarian Theorizing” RS 39 (2003): 168, who argues that all ST models make a numerical distinction between God and the persons.
[42] Although I use personal pronouns in relation to God throughout this essay, this is largely conventional. As I shall argue later, it is not clear whether, on TM, God can even be considered a person.
[43] A.J. Cotnoir, “Mutual Indwelling” Faith Philos. 34 (2017):138.
[44] Craig (PFCW, 201) argues that in the case of metaphysical substances, the whole is ontologically prior to its parts, which would imply that God – described as a soul-like substance – is prior to his “parts” (i.e., the persons). It is hard to reconcile this with some of the already-noted implications of his compositional approach to the Trinity.
[45] Baddorf (“Divine Simplicity”, 403-418) argues that a complex God can retain aseity, based on the notion of the ontological priority of the whole over its parts. However, this can only really apply to God’s properties, where any concerns that grounded-ness might undercut divinity are irrelevant (since divine properties, unlike the persons, are not divine per se to begin with).
[46] Naomi Thompson, “Metaphysical Interdependence”, in Mark Jago (ed.), Reality Making (Oxford: OUP, 2016), 38-57.
[47] Thompson, “Metaphysical Interdependence”, 48.
[48] Howard-Snyder, “Trinity Monotheism”, 114.
[49] Howard-Snyder, “Trinity Monotheism”, 114.
[50] Craig, “Trinity Monotheism Once More: A Response to Daniel Howard-Snyder”, Philos. Christi 8 (2006): 108-109.
[51] Craig, “Trinity Monotheism”, 108-109.
[52] Craig, “Trinity Monotheism”, 110.
[53] Craig, “Trinity Monotheism”, 110.
[54] D.A. Cruse, “On the Transitivity of the Part-Whole Relation”, JoL, 15 (1979): 32.
[55] Achille C. Varzi, “A Note on the Transitivity of Parthood”, Applied Ontology 1 (2006): 142.
[56] Cruse, “On the Transitivity”, 32.
[57] Cruse, “On the Transitivity”, 32; Varzi, “A Note”, 142.
[58] Grimm argues that mereology adopts a liberal definition of individual, corresponding to any number of entity types. This seemingly undermines Craig’s “individualist” argument against transitivity. See Scott Grimm, “Degrees of Countability: A Mereotopological Approach to the Mass/Count Distinction”, SLTP 22 (2012): 588.
[59] Craig, “Trinity Monotheism”, 109.
[60] Howard-Snyder, “Trinity Monotheism”, 104ff.
[61] Howard-Snyder, “Trinity Monotheism”, 103.
[62] Leftow, “Anti-Social Trinitarianism”, 210-211.
[63] Howard-Snyder, “Trinity Monotheism”, 109-110.
[64] Craig, “Trinity Monotheism”, 106; Hasker, Metaphysics, 142-143.
[65] Many philosophers dispute the claim that things have essences or natures. However, given Craig helps himself to the idea (hence, his use of felinity as a stable metaphysical “core”, identifying certain cat-like objects), such appeals are legitimate. See Craig, PFCW, 199.
[66] For more on this, see Abner Shimony, “The Status and Nature of Essences”, Rev. Metaphysics 1 (1948): 38. Shimony uses a very broad conception of essence to include “any character” of a certain item, which is ambiguous. For a philosophically precise rendition of essences, see David S. Oderberg, “Essence and Properties”, Erkenntnis 75 (2011): 87; Oderberg, Real Essentialism (RSCP; London: Routledge, 2007), 44-47; Feser, Aristotle’s Revenge: The Metaphysical Foundations of Physical and Biological Science (Neunkirchen-Seelscheid: Editiones Scholasticae, 2019), 57, 60, 403-405.
[67] I am referring to intrinsic lack, not (e.g.) genetic defect or injury.
[68] Craig, PFCW, 589.
[69] Cf. Rea, “The Trinity”, 415.
[70] Howard-Snyder (“Trinity Monotheism”, 103, n.6) argues that Craig’s position fails, since God can have a property necessarily without it being constitutive of the divine nature. As another example, I might be left-handed in all possible worlds (i.e., necessarily), although that does not mean that left-handedness is part of my nature. Cf. Justin Mooney, “A New Logical Problem for the Doctrine of the Trinity”, RS 54 (2018): 6. That said, it could be argued that God’s triunity is far more fundamental to who he is than my left-handedness is to me.
[71] Mooney (“A New Logical Problem”, 6-7) also argues that triunity is, like other divine attributes, worship-worthy. But as a bare idea, this seems insufficient. For a property to induce worship of its bearer, it ought to be subject to clear conceptualisation; only then can it be assessed as something that should prompt veneration. But triunity simpliciter lacks the necessary clarity that would permit such an assessment: it remains ambiguous, requiring theoretical elaboration to have any force (to avoid heresy, for one thing). The advocate of TM would therefore be compelled to propose her favoured Trinitarian theory. But this brings us back to the problem at hand – i.e., whether triunity is an essential attribute of the divine nature. One would therefore have to presume the very notion at issue.
Mooney further suggests that triunity, like other great-making properties, plays an important role in identifying God (and presumably setting him apart from other, non-divine beings). But this claim is undermined by the realization that pre-Christian Jews seemingly had no trouble identifying Yahweh, despite having no clear conception of the Trinity. Indeed, one will be hard pressed to find any unambiguous reference to God’s tri-personality – even in latent or embryonic form – in the Old Testament, even though God’s people were capable of distinguishing him from false gods and could appropriately respond to his self-disclosures. The implied conclusion is that triunity is not required to uncover God’s identity.
[72] Mooney, “A New Logical Problem”, 7.
[73] Ironically, Craig elsewhere rejects the doctrine of eternal generation on the grounds that it creates different tiers of divinity within the Godhead (Craig, “Is God the Son Begotten in his Divine Nature?”, Theologica 3 [2018]: 22-32).
[74] Michael Willenborg, “The Persons of the Trinity are Triune Themselves: A Reply to Justin Mooney”, RS (2021): 1-8.
[75] Willenborg, “The Persons of the Trinity”, 3.
[76] Craig might respond that if there is no distinction between the persons and the entire Trinity, one runs into problems associated with the transitivity of identity relations (i.e., if the persons are identical with God, then that suggests identity with each other). Relations of origin (paternity, generation, spiration) might resolve the dilemma – although as we have seen, Craig rejects this (cf. n.72).
[77] Rea, “The Trinity”, 416ff.
[78] Many Trinitarian theorists have been careful not to privilege the divine substance over the hypostases, because it would suggest the latter has a metaphysically non-primitive status. Wilks observes this of the Cappadocians in the course of examining Zizioulas’ Trinitarianism. See J.G.F. Wilks, “The Trinitarian Ontology of John Zizioulas”, Vox Evangelica 25 (1995): 73; Stephen R. Holmes, The Holy Trinity: Understanding God’s Inner Life (Milton Keynes: Paternoster, 2012), 146. Cf. Basil of Caesarea, Ep. 52: “…in the case of God the Father and God the Son there is no question of substance anterior or even underlying them both”.
[79] Hasker, Metaphysics, 228.
[80] Lucy Wallis, “Living a Conjoined Life”, BBC News (25 April, 2013), https://www.bbc.com/news/ magazine-22181528.
[81] Eric Olson, “The Metaphysical Implications of Conjoined Twinning”, SJP 52 (2014): 27.
[82] See Howard-Snyder, “Trinity Monotheism”, 118-119, for a fictional parallel to the Hensel Twins (called “Twinsy”). I shall use the Hensel twins, as a real-life example seems to make the relevant points more vividly.
[83] Craig (“Trinity Monotheism”, 103) raises the (rare) phenomenon of two-headed animals as a real-life parallel to Cerberus.
[84] Alexandra Boyle, “Conjoined Twinning and Biological Individuation”, PS 177 (2020): 2395-2415. Boyle suggests that because we can conceive of the idea of conjoined twins being cloned from an original, embryonic aetiology is immaterial. However, because this hypothetical clone just is a copy of twins who experienced interrupted bifurcation, aetiology remains indirectly relevant (for the nature of the facsimile is constrained by the pattern set by the original). In any case, Boyle’s conclusion that such twins constitute one, psychologically divided human would not help Craig, for his model relies on there being three distinct, psychologically integrated individuals.
[85] See Craig, “Trinity Monotheism”, 103.
[86] Craig seems to endorse the absolute character of identity relations, judging by his critique of relative identity in his work on the Trinity. See Craig, PFCW, 590. For a defence of the principle of absolute identity, see H.E. Baber, “Almost Indiscernible Twins”, Philosophy and Phenomenological Research 52 (1992): 365-366. Cf. a summary of the issues in E.D. Bohn, “The Logic of the Trinity”, Sophia, 50 (2011), https://link.springer.com/article/10.1007/s11841-011-0265-1#Sec1.
[87] Again, see Baber, “Almost Indiscernible Twins”.
[88] To be sure, it may be possible to argue that the Hensel “organism” is capable of rational thought in an analogous sense, just because its personal constituents are. I deal with similar rejoinders below when discussing the claim that Craig’s model entails God’s non-personhood.
[89] Hasker, Metaphysics, 231-236. See also Trenton Merricks, “Split Brains and the Godhead”, in Thomas Crisp, David Vander Laan, and Matthew Davidson (eds.), Knowledge and Reality: Essays in Honor of Alvin Plantinga (Amsterdam: Kluwer Academic Publishers, 2006), 299-326. Merricks concedes that even if distinct centres of consciousness emerge within a split-brain patient, the idea itself is ambiguous, failing to rise to the level of genuine personhood.
[90] Hasker, “Persons and the Unity of Consciousness”, in Robert C. Koons and George Bealer (eds.), The Waning of Materialism (Oxford: OUP, 2010), 190.
[91] Wilkes stating that early researchers into the phenomenon were surprised by the manifest absence of behavioural disunity. See Kathleen V. Wilkes, “Multiple Personality Disorder and Personal Identity”, BJPS 32 (1981): 341-342. See also Edward H.F. de Haan, et. al., “Split-brain: What we Know Now and Why this is Important for Understanding Consciousness”, NR 30 (2020): 224-233, link.springer.com/article/10.1007/s11065-020-09439-3.
[92] Tim Bayne, “The Unity of Consciousness and the Split-Brain Syndrome”, TJP 55 (2008): 287.
[93] Bayne, “The Unity”, 288. But see Hasker, “Persons and the Unity”, 186-190 for a countervailing view.
[94] Yair Pinto, et. al., “Split-brain: Divided Perception by Undivided Consciousness”, Brain 140 (2017): 1231-1237.
[95] de Haan, “Split-brain: What We Know”.
[96] Bayne, “The Unity”, 286-289.
[97] Even Elizabeth Schechter, who argues the experience of split-brain patients offers evidence for two minds emerging within the one human being, does not think they rise to the level of persons capable of engaging in genuine interpersonal interaction. See Schechter, Self-Consciousness and ‘Split’ Brains – The Minds’ I (Oxford: OUP, 2018), 17.
[98] Hasker, “Persons and the Unity”, 179.
[99] Hasker, “Persons and the Unity”, 180.
[100] A.A.T.S. Reinders, et. al., “One Brain, Two Selves”, Neuroimage, 20 (2003): 2119-2125. This could over-interpret the data: localised brain activity may signal no more than a de-integration of consciousness, perception, memory, etc., consistent with psychic fragmentation.
[101] See Hasker, “Persons and the Unity”, 178-179 and the remarkable examples of MPD patients revealed therein.
[102] Using a variety of criteria (rationality, moral agency, etc.), Saks argues that the alters exhibited by a person with MPD are, in fact, separate persons. See Elyn R. Saks (with Stephen H. Behnke), Jekyll on Trial: Multiple Personality Disorder and Criminal Law (New York: NYU Press, 1997), 39-66. See Walter-Sinnott Armstrong and Stephen Behnke, “Criminal Law and Multiple Personality Disorder: The Vexing Problems of Personhood and Responsibility”, Nous 34 (2000): 304-305ff, for cogent replies to this claim.
[103] Philip M. Coons, “Treatment Progress in 20 Patients with Multiple Personality Disorder”, JNMD 174 (1986): 715-721.
[104] Armstrong and Behnke, “Criminal Law”, 304-305ff.
[105] Hasker, “Persons and the Unity”, 180.
[106] Some philosophers, defining personhood in narrative terms, argue that the alters of an individual with MPD may actually flow from a failure to hold together increasingly divergent and conflicting accounts of the self. The self is characterised as fragmented, but without implying the emergence of distinct and independent centres of consciousness. See Valerie Gay Hardcastle and Owen Flanagan, “Multiplex vs. Multiple Selves: Distinguishing Dissociative Disorders”, The Monist, 82 (1999): 649-652.
[107] Craig, “Does the Problem of Material Constitution Illuminate the Doctrine of the Trinity?”, Faith Philos. 22 (2005): 83.
[108] Applying MPD to the Trinity, Hasker (Metaphysics, 243)equivocates, saying the members are three “divine life streams”, analogous to certain alters. But is a “life stream” (no matter how rich) the same as a person?
[109] Armstrong and Behnke, “Criminal Law”, 302.
[110] Craig, “Trinity Monotheism”, 101.
[111] Traditionally, philosophers have distinguished between two types of substance: primary substances that correspond to individual entities; and secondary substances (i.e., a thing’s nature or essence). I am here talking about substances in the first sense.
[112] Craig, PFCW, 196-200.
[113] Howard-Snyder (“Trinity Monotheism”, 116-118) defends the claim that on Craig’s own terms, the divine persons are individual substances. The succeeding paragraph is indebted to his observations.
[114] Craig, “Trinity Monotheism”, 110-111.
[115] Howard-Snyder, “Trinity Monotheism”, 117.
[116] Craig, “Trinity Monotheism”, 110-111.
[117] Kyle Claunch, “What God Hath Done Together: Defending the Historic Doctrine of the Inseparable Operations of the Trinity”, JETS 56 (2013): 781-800.
[118] Craig, “Trinity Monotheism”, 111.
[119] Craig, “Trinity Monotheism”, 108.
[120] Craig, “Trinity Monotheism”, 108.
[121] Craig, PFCW, 610.For an extended explication of this critique, see James R. Gordon, “‘Twinsy’ and Trinity: An Assessment of Trinity Monotheism of William Lane Craig”, unpublished M.Div thesis, Trinity Evangelical Divinity School (2008), 9-10.
[122] Howard-Snyder, “Trinity Monotheism”, 109.
[123] Howard-Snyder, “Trinity Monotheism”, 120.
[124] Craig, “Trinity Monotheism,” 104.
[125] Craig, “Towards a Tenable”, 99.
[126] Craig, “Towards a Tenable”, 99.
[127] Craig, “Towards a Tenable”, 99; Craig, PFCW, 253.
[128] Howard-Snyder, “Trinity Monotheism”, 120: “…the one I am absolutely identical with supports my individual being”.
[129] Howard-Snyder, “Trinity Monotheism”, 120.
[130] Tuggy, “Hasker’s Quests”, 180.
[131] See Howard-Snyder, “Trinity Monotheism”, 107-108, n.18.
[132] Andrew Kirschner, “Will-Independent Mereological Trinity Monotheism: A Defence of the Logical Coherence of, A Priori Motivation for, and a Particular Model Concerning the Doctrine of the Trinity”, unpublished PhD thesis, University of Arkansas, (2019), 147ff; Bohn, “The Logic”, 367.
[133] Cotnoir acknowledges this kind of problem undermines stronger versions of CAI. See A.J. Cotnoir, “Composition as Identity”, in Cotnoir and Donald L.M. Baxter (eds.), Composition as Identity (Oxford: OUP, 2014), 13. Weaker versions may not be so susceptible, although they eschew the kind of numerical identity that is seemingly required here.
[134] Cf. Craig, “Does the Problem”, 83.
[135] Craig, “Towards a Tenable”, 99.
[136] Tuggy, “Hasker’s Quests”, 180.
[137] Craig, “Trinity Monotheism”, 108.
[138] Tuggy, “Hasker’s Quests”, 180.
[139] Howard-Snyder, “Trinity Monotheism”, 107ff.
[140] This is why Thomas McCall’s objection to critics of TM fails. He argues that they have not “demonstrat[ed] that God must be only one person rather than three to be truly personal” (italics mine). But as I have suggested, one self cannot “be” identical to three selves, for identity is symmetrical and absolute. See McCall, Which Trinity? Whose Monotheism? Philosophical and Systematic Theologians on the Metaphysics of Trinitarian Theology (Grand Rapids: Eerdmans, 2010), 31.
[141] Tuggy, “The Unfinished Business”, 168; Howard-Snyder, “Trinity Monotheism”, 107.
[142] Of course, this assumes that not being a person reflects a deficiency in God – something other Christian thinkers would dispute. For a philosophically rigorous response to major arguments against God being a person, see Ben Page, “Wherein Lies the Debate? Concerning Whether God is a Person”, IJPR 85 (2019): 297-317, https://link.springer.com/article/10.1007/s11153-018-9694-x#Sec6.
[143] Tuggy, “The Unfinished Business”, 168-169.
[144] Layman, advocating a model similar to TM, concedes that on such views, the Trinity cannot legitimately use first-person pronouns. See C. Stephen Layman, Philosophical Approaches to the Atonement, Incarnation, and the Trinity (New York: Palgrave Macmillan, 2016), 160. Cf. Brian Davies, Thomas Aquinas’s Summa Theologiae: A Guide & Commentary (Oxford: OUP, 2014), 348-352, whose criticisms of reductive accounts of God are well-taken.
[145] This compounds the incongruities in Craig’s proposal I have already noted. To wit: (a) God is an individual (mental) substance; (b) God is not identical with any of the persons; but (c) persons are, on Craig’s view, paradigmatic examples of substances, who bear rationality and intellect. See Howard-Snyder, “Trinity”, The Routledge Encyclopedia of Philosophy, https://www.rep.routledge.com/articles/thematic/trinity/v-2, 2015.
[146] Gary Legenhausen (“Is God a Person?”, RS 22 [1986]: 315-316) suggests that one’s supplicatory relationship with God need not imply personhood: God is perhaps far more than a person, since to be a person (allegedly) represents a limitation on his nature. But this proves too much: if saying God is a person threatens to limit his nature, affirming that he is a (one) God seems to have a similar constraining effect, for it picks him out as a distinguishable “item”. But then how do Christians affirm monotheism?
[147] Legenhausen (“Is God a Person?”, 314-315) canvasses an argument purporting to show that revelation does not require personhood: something can still be dubbed “revelation” merely if it issues in the acquisition of a message. A message, though, is a form of communication with intelligible content. How, then, could it be produced by something without a mind – i.e., a non-person?
[148] Thatcher comments that the Bible never applies the word “person” to God. This is hardly dispositive: if explicit use of certain terms was the yardstick for acceptable statements about the Deity, then we could not speak of him as a Trinity. And even if the word “person” is never used, the God of the Bible is arguably presented as bearing all the qualities normally associated with such beings. See Adrian Thatcher, “The Personal God and the God who is a Person, RS 21 (1985): 71-72.
[149] E.g., Chad McIntosh, “The God of the Groups”, RS 52 (2015): 177-180.
[150] For a trenchant critique of the “myth” of corporate personality, applied to ancient Hebrew thought, see Stanley E. Porter, “Two Myths: Corporate Personality and Language/Mentality Distinction”, SJT 43 (1990): 289-299.
[151] Howard-Snyder (“Trinity Monotheism”, 122-123) observes that God’s non-personhood undoes biblical anthropology, subverting what is the lynch-pin of the doctrine – Genesis 1:27 and the imago dei.
[152] Craig, “Trinity Monotheism”, 104.
[153] Howard-Snyder, “Review of Metaphysics and the Tri-personal God”, Faith Philos. 32 (2015): 112; Hasker, Metaphysics, 198, who misstates Howard-Snyder’s position as tending towards Unitarianism.
[154] This could also be cashed out in mereological terms: being a composite, God is not identical with any one of his parts. But since those “parts” just are bounded personal agents, he cannot be identical with any of them. See Tuggy, “The Unfinished Business”, 168.
[155] Craig, “Trinity Monotheism”, 104-105.
[156] Tuggy, “Hasker’s Quests”, 181.
[157] E.g., Swinburne, The Christian God, 125; Alvin Plantinga and Mark Tooley, Knowledge of God (Oxford: Blackwell, 2008), 1-2. I do not necessarily endorse the way Swinburne or Plantinga conceive of God. I merely note their views for the record.
[158] See Aquinas, ST, 1a.29.3. Aquinas also argues that “person” captures the various descriptions of God found in Scripture, writing that “what the word signifies is found to be affirmed of God in many places of Scripture; as that He is the supreme self-subsisting being, and the most perfectly intelligent being”. True, Aquinas discusses the use of “person” in the context of the Trinitarian members. But as the above quote suggests, he could also apply the term to God as a whole. Moreover, since for Thomas the divine fullness resides in the persons, to speak of their personhood is, ipso facto, to talk about the entire Deity’s.
[159] Hasker, “Has a Trinitarian God Deceived Us?”, in McCall and Rea, Philosophical and Theological, 44.
[160] Hasker, Metaphysics, 197. Hasker’s error partly lies in his conflation of analogical and non-literal language, which he thinks permits the claim that God is analogous to a person, even if he is not literally so. But this is incorrect: analogical language is not non-literal. See Davies, The Thought of Thomas Aquinas (Oxford: Clarendon Press, 1992), 70-71, on this point.
[161] Looking at the issue from another point of view, imagine if I treated a guilty person as if they were innocent, and decided thereby not to punish them (say, because I wanted to exercise mercy, or acknowledge my role in whatever act of wrongdoing they had committed). This would not mean that they were guiltless in fact, and my treating them as such would not imply otherwise. To use a term like “as if” is simply a facon de parler.
[162] For more on the nuances of the doctrine of analogy, especially in its medieval/Scholastic form, see Davies, The Thought, 70-74.
[163] See Howard-Snyder, “Trinity Monotheism”, 122, who raises this possibility only to reject it.
[164] Howard-Snyder, “Trinity Monotheism”, 122.
[165] I might also add that this exacerbates problems first encountered in Craig’s formal theory: no longer are certain great-making attributes open to God in a merely derivative fashion; if the present criticism is correct, then they are not open to him at all.
[166] See McIntosh, “The God of the Groups”, 171, for a description of the differences between functional and intrinsicist persons. Cf. Hasker, “Objections to Social Trinitarianism”, RS 46 (2010): 423.
[167] McIntosh, “The God of the Groups”, 168-170.
[168] McIntosh, “The God of the Groups”, 170-171.
[169] McIntosh, “The God of the Groups”, 172.
[170] Craig, “Does the Problem”, 83.
[171] McIntosh, “The God of the Groups”, 173-175.
[172] McIntosh, “The God of the Groups”, 172.
[173] McIntosh, “The God of the Groups”, 175