Hostname: page-component-74656f9884-c52kr Total loading time: 0 Render date: 2026-11-08T11:09:01.018Z Has data issue: false hasContentIssue false
  • English
  • Français

Social Trinitarianism and polytheism

Published online by Cambridge University Press:  09 February 2011

BRANDON CAREY
BRANDON CAREY*
Affiliation:
Department of Philosophy, University of RochesterBox 270078, Rochester, NY 14627-0078
*
e-mail: careybp@gmail.com
Rights & Permissions [Opens in a new window]

Abstract

Social Trinitarians attempt to solve the logical problem of the Trinity by claiming that there are three numerically distinct divine persons. A common objection to this view is that it is seemingly committed to the existence of multiple Gods and is therefore polytheistic. I consider Edward Wierenga's response to this objection, as well as two other possible responses, and show that each faces serious philosophical problems. I conclude that, in the absence of a better method of distinguishing the property of being divine from that of being a God, Social Trinitarians are committed to the existence of more than one God.

Information

Type
Articles
Information
Religious Studies , Volume 47 , Issue 1 , March 2011 , pp. 97 - 107
Copyright
Copyright © Cambridge University Press 2011

Introduction

The doctrine of the Trinity seems to commit its adherents to at least the following three claims:

  • FG The Father is God.

  • SG The Son is God.

  • FNS The Father is not the Son.

On a straightforward reading, where FG and SG are identity claims, and FNS is the negation of an identity claim, these theses constitute an inconsistent triad. On pain of inconsistency, then, the first project for philosophers who endorse the doctrine of the Trinity is to provide a consistent reading of FG, SG, and FNS.

One way of accomplishing this is by denying that FG and SG should be read as identity claims, and instead understanding them as predicating something of the Father and the Son, while maintaining that FNS denies the numerical identity of the Father and the Son. This idea is the foundation of Social Trinitarianism, according to which the Father, the Son, and the Spirit are numerically distinct persons, but have some important property or properties in common.Footnote 1 In this paper I address a common complaint made against Social Trinitarianism: that it entails polytheism. Ultimately, I will argue that, on any plausible understanding of what it is to be a God, Social Trinitarians are committed to the existence of more than one God, and thus committed to polytheism.

Social Trinitarianism

Taken at face value, FG, SG, and FNS are claims about identity.Footnote 2 Read this way, they entail the following:

  • FG′ The Father is (numerically) identical with God.

  • SG′ The Son is (numerically) identical with God.

  • FNS′ It is not the case that the Father is (numerically) identical with the Son.

But, of course, the Father and the Son are each numerically identical with God only if they are identical with each other as well, so all three of these claims cannot be true. This is the so-called logical problem of the Trinity: the conjunction of any two of the above claims entails the negation of the third. Any philosophically respectable account of the Trinity will have to avoid the resulting inconsistency by rejecting at least one of the primed claims and denying that FG, SG, and FNS should all be taken at face value.Footnote 3

The Social Trinitarian chooses to take only FNS at face value, endorsing FNS′, but denying that FG and SG express identity claims. Cornelius Plantinga (Reference Plantinga, Ronald and Cornelius1989) states the view as follows: ‘[T]he Holy Trinity is a divine, transcendent society or community of three fully personal and fully divine entities: the Father, the Son, and the Holy Spirit’, each of whom is ‘a distinct person’ (27, emphasis added). So, according to the Social Trinitarian, there are three non-identical persons, the Father, the Son, and the Holy Spirit, each of whom is fully divine. Since the Father and the Son are distinct, FG and SG cannot be identity claims, but must instead be predication claims.

The most obvious way to reformulate FG and SG as predication claims is as follows:

  • FG′′ The Father is a God.

  • SG′′ The Son is a God.Footnote 4

Since the conjunction of these claims is compatible with FNS′, the Social Trinitarian can avoid the charge of inconsistency by accepting that FG″ and SG″ are the correct readings of FG and SG, respectively. Call this first formulation of Social Trinitarianism, the conjunction of FG″, SG″, and FNS′, ‘ST1’.Footnote 5

Polytheism

An obvious objection to ST1 is that it entails the existence of the wrong number of Gods. Since, according to ST1, there are at least two distinct things that have the property of being a God, ST1 is committed to the existence of at least two Gods, and is thus a polytheistic view. Since at least some statements of the doctrine of the Trinity explicitly endorse that there is exactly one God, this seems like a devastating problem for ST1.

We need to be careful, however, in exactly how we formulate this charge. There is a perfectly reasonable sense of ‘polytheism’, for example, in which belief in the Greek gods is polytheistic, but on the assumption that a God (with a capital ‘G’) must at least be omnipotent, omniscient, morally perfect, etc., the existence of the Greek pantheon does not entail the existence of any Gods. My interest here is not in this sense of ‘polytheism’, but rather in the philosopher's sense of ‘polytheism’, which asserts that there are multiple beings that are omnipotent, omniscient, morally perfect, etc. This suggests the following understandings of the basic theistic positions, where being a God requires at least omnipotence, omniscience, and moral perfection:

  • Theism There is at least one God.

  • Monotheism There is exactly one God.

  • Polytheism There are at least two Gods.

The objection, then, is that ST1 is incompatible with monotheism and committed to polytheism, and this is clearly true. However, a different version of Social Trinitarianism can be formulated so as to avoid commitment to polytheism while still endorsing a plausible reading of FG and SG.

Edward Wierenga (Reference Wierenga2004) proposes a Social Trinitarian view that maintains consistency with monotheism by rejecting FG″ and SG″ in favour of two different predicative claims. Let us say that the property of being divine is just the property of having all of the divine attributes, such as omnipotence, omniscience, and moral perfection.Footnote 6 Wierenga then proposes to read FG and SG as:

  • FG‴ The Father is divine.

  • SG‴ The Son is divine.

Call the conjunction of these two claims with FNS′ ‘ST2’. ST2 is at least not obviously committed to polytheism – what is predicated of the Father and the Son is neither identity with God nor the property of being a God, but rather the property of having all of the divine attributes. It is thus open to proponents of ST2 to say that, although there are at least two divine beings, there is only one God, and thus polytheism is false.

However, Wierenga notes that ST2 faces a problem raised by Richard Cartwright based on the relationship between being divine and being a God. Cartwright's argument is as follows: ‘[E]very Divine person is a God; there are at least three Divine Persons; therefore, there are at least three Gods’ (1987, 196). Reformulating Cartwright slightly, being divine is sufficient for being a God, and so, in virtue of its entailing the existence of more than one divine being, ST2 entails the existence of more than one God.Footnote 7 The success of Cartwright's argument depends, of course, on the acceptability of the first premise, which Cartwright defends only by describing it as a ‘trivial truth’ (196).

The tradition of philosophical theism supports this premise, however, as the question of whether there is a God is typically addressed by philosophers by attempting to answer the question of whether there is a being with all of the divine attributes. But even a proof of the existence of such a being would not constitute a proof that there is a God unless being divine were sufficient for being a God.Footnote 8 The proponent of ST2 must deny this premise, however, and offer an account according to which the Father and the Son are both divine, but there is nevertheless only one God. That is, he must present some plausible way of distinguishing the property of being divine from the property of being a God. In the following sections, I will consider three attempts to construct such a position.

Wierenga's response

Wierenga responds by denying Cartwright's allegedly trivial truth and instead endorsing the following account of what it is to be a God:

  • G1x is a God iff (∃y)(y=God & x=y)

Though it is still presumably necessary for being a God that a thing be divine, it is not sufficient according to Wierenga.Footnote 9 So, although there are (at least) three divine persons according to ST2, they are not all Gods, because they are not all identical with each other (and thus not all identical with God), and so ST2 is not committed to polytheism.

Though this response does allow ST2 to avoid the charge of polytheism, there are good reasons to deny G1. First, Jeffrey Brower (Reference Brower2004) has noted an unfortunate consequence of accepting G1: G1 entails that polytheism is logically impossible. Brower writes, ‘[A]s Wierenga understands it, polytheism is logically impossible and hence trivially false, since it is logically impossible for more than one thing to be identical with God’ (299). G1 clearly does have this consequence, as anything that is a God will have to be identical with God, and so there cannot be more than one God. A closely related problem is that, if G1 is true, then monotheism is a logical consequence of theism. Since monotheism entails theism, this will have the (to my mind, at least) counterintuitive result that theism and monotheism are logically equivalent.Footnote 10

These consequences of G1 require us to say somewhat strange things about anyone who either accepts polytheism or accepts theism without committing to monotheism. If G1 is true, then the only logically consistent theistic position is monotheism. Those who endorse polytheism, as well as those who endorse theism but are undecided about monotheism, must be accused of failing to understand either the logic of identity or seriously misunderstanding what it is to be a God (in that they presumably think it is a matter of having some set of impressive properties, when in fact to be a God is just to be identical with a particular being). And while if monotheism is true, there must be something wrong with polytheism, one might have expected it to be a mere error in counting rather than a serious logical mistake. While this does not constitute a refutation of G1, it is a surprising result.Footnote 11

A more serious problem for G1 is that it gives counterintuitive verdicts on the truth of theism in certain thought-experiments. Suppose we discovered that ‘God’ does not refer, because although there is a being who is omnipotent, omniscient, morally perfect, and so on, this being has never communicated with humans or otherwise causally interacted with them in a way that would secure the reference of the name ‘God’.Footnote 12 Would we then conclude that theism is false because there is no x such that x is identical with God? My intuition is that we would not.Footnote 13 However, if G1 is true, then, since there is nothing which is identical with God in this case, there are no Gods, and so theism must be false.

Of course, the proponent of G1 may respond that the reference of ‘God’ is fixed by description. Perhaps, for example, the reference of ‘God’ was at some time fixed with the definite description ‘the omnipotent, omniscient, morally perfect creator of the universe’. On this view, the case above is a case in which theism is true, as the reference of ‘God’ would have been fixed to the being answering that description. However, a modified example can generate a similar problem for this view: suppose we discovered that there were two beings, each possessing all of the properties typically attributed to God.Footnote 14 Would we conclude that theism is false? Again, I think we would not. According to G1, however, this would be a case in which theism was false, since the definite description ‘the omnipotent, omniscient, morally perfect’ would not refer (because there would be two beings with these properties).Footnote 15

Both kinds of problems likely have their root in the same intuition: that sentences of the form α is a God attribute to their subjects membership in a kind. As such, it ought to be at least logically possible for the predicate ‘is a God’ to be true of more than one subject.Footnote 16 Since G1 is incompatible with this intuition, there is good reason to reject it. So, let us turn our attention to two strategies for claiming that, although polytheism is logically possible, the Social Trinitarian is not committed to it.

The greatest-being response

One way of avoiding the charge of commitment to polytheism, while still allowing for the logical possibility of more than one God, has its roots in the notion that a God is the greatest conceivable being. Using this idea, the proponent of ST2 might offer the following alternative to G1:

  • G2x is a God iff nothing greater than x is conceivable.

Add to this the claim that a society of three divine persons is greater than a single divine person, and the Social Trinitarian seemingly has an easy way of avoiding polytheism: the Father and the Son are both divine, but there is something greater than each of them, namely the Trinity of divine persons, and so neither of them is a God.Footnote 17

This strategy avoids the problems of G1, as, if there is a tie for greatness, there could be more than one God. Indeed, even if it is metaphysically impossible for there to be a tie, it would not follow as a matter of logic that polytheism was false; it would follow from the metaphysical truth that there can be no ties in greatness, in conjunction with the fact that a society of three divine persons is maximally great. It seems more reasonable to attribute the fault of those who endorse polytheism to a failure in judgment about such a principle than to a logical error or radical misunderstanding of what it is to be a God. Furthermore, since G2 uses no proper names or definite descriptions, the reference failure that generated the problem cases for G1 could only occur if there were an infinite series of increasingly greater things that could be conceived. Assuming that there is not such a series, G2 can avoid these problems as well.

There are, however, other reasons to doubt G2. First, it is not clear that the Trinity of divine persons is greater than any of its members in the relevant sense. For one thing, perfect-being theology has as its subject the perfection of beings, and it is difficult to see how the society of Father, Son and Spirit is itself a being at all.Footnote 18 More importantly, there is reason to doubt that the Trinity of divine persons is greater than any of its members, as, since each member of the Trinity is omnipotent, omniscient, and morally perfect, etc., the society of all three seemingly has no power or virtue that each member does not possess.

A further problem for this sort of response is that it seems quite reasonable to say that the divine attributes are just those attributes that would be had by the greatest conceivable being. However, the proponent of both ST2 and G2 must deny this. If the Trinity of divine persons is greater than its members, this is because it possesses some great-making property that they lack, such as the property of being triune. If this is the case, then, in virtue of being a great-making property, that property seemingly ought to be counted as one of the divine properties, in which case FG‴ and SG‴ are false; the Father and the Son are not divine, because they are not triune.

The proponent of G2 could certainly reject the intuitive claim that the divine attributes are just those that would be had by the greatest conceivable thing, and stipulate that being divine and being a God are distinguished by just the great-making properties that the Trinity possesses and the persons do not, but this rejoinder seems ad hoc – what reason is there for thinking that these properties are necessary for being a God, but not necessary for being divine, except that it allows ST2 to avoid the charge of polytheism? Furthermore, this view fails to match up with intuitions about the truth-conditions of theism as it predicts that, were we to discover that there are only two omnipotent, omniscient, morally perfect creators of the universe, we would think that theism was false. Since it seems much more plausible that we would think that theism and polytheism were both true in that scenario, G2 does not seem to provide the Social Trinitarian with an adequate defence against the charge of polytheism.

The independent-existence response

A final strategy for endorsing ST2 while avoiding commitment to polytheism draws on the idea of a God as an independently existing being. The idea is that, though each is divine, the Father and the Son each fall short of being a God, because each is dependent on the other in some way. This suggests the following necessary condition on being a God:

  • G3x is a God only if x is completely self-existent.

On at least some Social Trinitarian views, neither the Father nor the Son (nor the Spirit) is completely self-existent. For example, on Richard Swinburne's (1994) view, the three persons of the Trinity all causally sustain each other's existence,Footnote 19 in which case, assuming that not being caused by some other being is at least necessary for complete self-existence, none of them is completely self-existent. Thus, on this view, neither the Father nor the Son would be a God, even though both are divine, and so, according to G3, ST2 can avoid the charge of polytheism.

The most obvious concern with this response is that it is not at all clear why being completely self-existent should not also be a necessary condition for being divine. Indeed, such independence is typically included in more complete lists of the divine attributes, and so denying this property of the Father and the Son while affirming their full divinity seems contradictory.

A more serious concern, however, is that it is not clear whether the Social Trinitarian picture described here even endorses theism. According to G3, theism is true only if there is at least one thing which is completely self-existent, and neither the Father nor the Son nor the Spirit has this property. The only remaining (plausible) candidate for this property would seem to be the society containing all three, but is the Trinity of divine persons completely self-existent? This will depend, of course, on what is meant by ‘completely self-existent’.

The Trinity is not caused to exist by its members, but it does depend on them in a different problematic way: if the members of the Trinity failed to exist, then the Trinity would fail to exist.Footnote 20 Lacking this kind of counterfactual dependence on any other being seems like another plausible necessary condition for complete self-existence, in which case this sort of view will have the consequence that there are no completely self-existent beings.Footnote 21 As such, the proponent of both ST2 and G3 must deny that there are any Gods, requiring him to reject theism. While this trivially avoids the charge of polytheism, it is also obviously inconsistent with monotheism, leaving the Social Trinitarian in at least as much trouble as he was in to begin with.

Each of G1–G3 provides a way for the Social Trinitarian to endorse ST2 while rejecting polytheism, but each also entails at least one questionable claim. G1 is itself subject to serious doubt, and while G2 and G3 seem more plausible, when combined with ST2 and the denial of polytheism, they commit their adherents to other implausible theses. It may be that these accounts can be revised so as to avoid the concerns above, or that some other method of endorsing ST2 while denying polytheism will prove more promising. In the absence of a successful strategy, however, I conclude that Cartwright is correct in claiming that anything that is divine is a God, and thus that ST2 does not provide a way for the Social Trinitarian to avoid polytheism.

A different predicate?

It is of course open to the Social Trinitarian to deny ST2 and offer some other reading of FG and SG that does not entail that the Father and the Son are divine (and therefore Gods). For example, William Hasker (Reference Hasker2010) has recently proposed reading FG and SG as predicating neither divinity nor identity with God, but rather the property of being God. This property is to be understood as distinct from, and not entailing, either the property of being identical with God, the property of being a God, or the property of being divine. If it did entail any of these, then Hasker's proposal would either be inconsistent, in virtue of asserting the existence of three numerically distinct persons who are nevertheless all identical with God, or it would entail either ST1 or ST2, in which case it would also be vulnerable to the charge of polytheism.

However, I am in sympathy with Brian Leftow (Reference Leftow, Stephen, Davis and Gerald2010) in that I fail to understand what the property of being God could be if it does not entail either identity with God, divinity, or being a God. In fact, there seem to be good intuitive reasons to reject that there is any such property. If there is such a property, then ‘x is God’ is consistent with ‘x is not identical with God’, ‘x is not divine’, and ‘x is not a God’. But unless ‘x is God’ entails that some being other than x has one of these properties, it will then also be consistent with ‘God does not exist’, ‘there are no divine beings’, and ‘there are no Gods’. So, allowing for Hasker's property of being God seems to allow the consistency of the conjunctions ‘x is God, and God does not exist’, ‘x is God, and there are no divine beings’, and ‘x is God, and there are no Gods’, each of which intuitively seems inconsistent. And while the intuitive inconsistency of any two of these conjunctions might be explained away by appeal to denial of the third (e.g. ‘when I say that x is God despite not being either identical with God or divine, I mean that x is a God’), Hasker cannot avail himself of this strategy, as he is committed to the consistency of the conjunction of all three of these conjunctions.

Without some other way to explain the intuitive inconsistency of these claims, then, there is good reason to doubt that there is an independent property of being God. Without some such property, however, that FG and SG can plausibly be read as attributing to the Father and the Son, or some other understanding of the property of being a God, the Social Trinitarian seems committed to polytheism.Footnote 22

Footnotes

1. In recent papers in this journal, William Hasker (Reference Hasker2010)and Brian Leftow (Reference Leftow, Stephen, Davis and Gerald2010)give slightly different accounts of Social Trinitarianism. Hasker includes in Social Trinitarianism an additional claim about how the divine persons are related to one another, while Leftow views Social Trinitarianism as an explanatory project which ‘takes the three Persons as in some way basic and explains how they constitute or give rise to one God’ (2010, 441). Since the charge of polytheism depends only on Social Trinitarianism's commitment to three numerically distinct divine persons, I will stick with my somewhat more anaemic formulation.

2. I say that an identity reading of these claims is the ‘face value’ reading only because ‘the Father’, ‘the Son’, and ‘God’ are treated as proper names in most discussions (though the former two are disguised as definite descriptions), and sentences of the form α is β, where ‘α’ and ‘β’ are proper names, typically express identity claims (and sentences of the form α is not β typically express negations of identity claims). I do not mean by this to suggest that this is the correct reading of these claims, but only that it is the most natural first reading.

3. There are many ways to accomplish this besides Social Trinitarianism. For example, Brian Leftow (Reference Leftow, Stephen, Davis and Gerald1999, Reference Leftow2004) denies FNS, and reads FNS as denying that ‘the Father’ and ‘the Son’ refer to the same span in God's personal history. Alternatively, relative-identity theorists, such as van Inwagen (Reference van Inwagen1995)deny all three primed claims, and read FG, SG, and FNS as relative-identity claims such as ‘the Father is the same being as God’ and ‘the Father is not the same person as the Son’. While these views avoid the logical problem of the Trinity without endorsing Social Trinitarianism, they face other philosophical problems which I will not address here.

4. Some will bristle at the use of ‘is a God’. As I am using it, it could just as well be replaced by ‘is a deity’ or ‘is a thing of the same kind as God’.

5. Obviously, a complete Social Trinitarian theory will have to endorse parallel claims about the Spirit: that the Spirit is a God, but is numerically identical with neither the Father nor the Son.

6. Wierenga has explicitly endorsed this understanding of ‘divine’ in private communication, and Plantinga seems to be committed to it in his claim that the persons of the Trinity are ‘fully’ divine. It may be, however, that a view can still be appropriately called ‘Social Trinitarianism’ even if it attributes something less to the Father, Son, and Spirit – either that they have some, but not all of the divine attributes, or that they possess all of these attributes, but to a lesser degree than the Trinity as a whole. I take it that proponents of this strategy will have significant problems with orthodoxy, as they will have to affirm sentences like ‘the Father is not omnipotent’, or ‘the Son is morally perfect, but less morally perfect than it is possible to be’. However, such views may be able to avoid many of the problems below, and so the reader is free to think of these as problems for only a certain subset of Social Trinitarian views.

7. We have been focusing on just the Father and the Son, but the Social Trinitarian will also be committed to the claim that the Holy Spirit is divine (and distinct from both the Father and the Son), thus yielding Cartwright's conclusion of three Gods.

8. Strictly, the existence of a divine being would have to be sufficient for the existence of a God (who may be non-identical with the divine being). Social Trinitarians who accept that the existence of any one member of the Trinity metaphysically entails the existence of the entire Trinity can thus explain the philosophical tradition as an attempt to prove the existence of God (where ‘God’ refers to the whole Trinity) by proving the existence of a divine being (where that being is one member of the Trinity). Any such proof would then have to go on, however, to prove that the existence of a divine being entails the existence of God.

9. There is a complication here, as some, e.g. Brower (Reference Brower2004), have argued that God is not divine on Wierenga's view. I think that there is a perfectly good sense in which Wierenga can still claim that God is necessarily divine, and so I ignore this issue.

10. It is important to note that both criticisms here depend on Wierenga intending G1 as claiming that identity with God is more than just a metaphysically necessary and sufficient condition for being a God. It may be metaphysically necessary that God is the only being that satisfies the conditions for being a God (whatever they are), and yet nevertheless be logically possible that there be more than one God (in the sense that a contradiction cannot be derived from polytheism in first-order predicate logic with identity). It is plausible that Wierenga does intend more than this, as he offers G1 as a response to the question ‘What is it to be a God?’ (291), and thus presumably intends it as an analysis of ‘x is a God’.

11. It is at least surprising in the following respect: antecedently, one would have expected that the only interesting logical relations between these three theses would be that monotheism and polytheism are inconsistent with each other, and each entails theism.

12. Of course, in being the creator, this being would have some causal connection with speakers of English, but suppose there is no causal connection between the use of the name ‘God’ and this being.

13. As has been noted elsewhere (Kripke (Reference Kripke1980)), ‘God’ may be an unusual sort of proper name, which may cloud intuitions in this case somewhat. Thus, it may be that the proponent of G1 can find some other way to explain this intuition. In the absence of such an explanation, I take it that the case poses a problem for G1.

14. I construct the example with two divine beings, but a parallel case can be constructed with two societies of three divine beings. That is, if the correct description used to fix reference to God, as the Social Trinitarian may well maintain, is ‘the society of three omnipotent, omniscient, morally perfect creators of the universe’, a case can also be constructed in which ‘God’ fails to refer, but in which theism seems true.

15. The proponent of G1 may object that the cases I present here are metaphysically impossible. Since monotheism is a necessary truth, there are no possible worlds in which ‘God’ does not refer (indeed, its actual truth is enough to secure the reference of ‘God’). Assuming the truth of G1 and monotheism, this is certainly correct. However, I take it that, despite this fact, our intuitions about whether theism would be true in these situations can still be a useful guide as to the correct understanding of what it is to be a God.

16. Although there are surely specifiable predicates that are only true of one subject at any logically possible world, ‘is a God’ does not seem to me to be one of them. And while it may be a metaphysical necessity that God is the only member of God's kind, I can see no reason to think this is a conceptual or logical necessity.

17. There is an alternative way of using G2 to defend ST2, by claiming that one member of the Trinity is greater than the other two. In addition to facing the philosophical problems below, I take it that this strategy is theologically unacceptable to at least most Trinitarians.

18. For a discussion of what sort of entity this society might be, see Leftow (Reference Leftow, Stephen, Davis and Gerald1999).

19. It should be noted that the use of ‘causally’ here is non-standard. The Son, for example, ‘permissively’ causes the Father's existence by not doing anything to stop it. Nevertheless, both the Son and the Spirit are dependent on the Father's ‘actively’ (that is, in the normal way) causing their existence. If what is necessary to be self-existent is to not be actively caused by anything else, then the following response can be modified so that the only being which is not actively caused, the Father, is God, while the other members of the Trinity are merely divine. This response would avoid one of the objections below, but is, I suspect, unacceptable to many Trinitarians in virtue of making the Father metaphysically superior to the Son and the Spirit.

20. Given that the Trinitarian regards the existence of the Father, Son, and Spirit as necessary, if understood with the usual semantics, this counterfactual will be trivially true. For independent reasons, I think we should deny these semantics. The point I am trying to make can be thought of this way: though there are no possible worlds in which the Father, Son, and Spirit do not exist, at the nearest impossible worlds where they do not exist, the Trinity does not exist either.

21. This is on the plausible assumption that the Father, the Son, the Spirit, and the society of all three are the only candidates for this property.

22. An earlier version of this paper was presented at the 2009 Western Conference of the Society of Christian Philosophers. I thank the conference participants, and in particular Landon McBrayer, for useful comments and discussions. I am especially grateful to Ed Wierenga both for introducing me to the logical problem of the Trinity and for many helpful comments on earlier drafts of this paper.

References

Brower, J. E. (2004) ‘The problem with Social Trinitarianism: a reply to Wierenga’, Faith and Philosophy, 21, 295–303.CrossRefGoogle Scholar
Cartwright, R. (1987) ‘On the logical problem of the Trinity’, in idem Philosophical Essays (Cambridge, MA: MIT Press), 187–200.Google Scholar
Hasker, W. (2010) ‘Objections to Social Trinitarianism’, Religious Studies, 46, 421–439.CrossRefGoogle Scholar
Kripke, S. (1980) Naming and Necessity (Cambridge, MA: Harvard University Press).Google Scholar
Leftow, B. (1999) ‘Anti-Social Trinitarianism’, in Stephen, T.Davis, Daniel Kendall, & Gerald, O'Collins (eds) Trinity: An Interdisciplinary Symposium on the Trinity (New York: Oxford University Press), 203–250.Google Scholar
Leftow, B. (2004) ‘A Latin Trinity’, Faith and Philosophy, 21, 304–333.CrossRefGoogle Scholar
Leftow, B. (2010) ‘Two Trinities: reply to Hasker’, Religious Studies, 46, 441–447.CrossRefGoogle Scholar
Plantinga, C. Jr, (1989) ‘Social Trinitarianism and tritheism’, in Ronald, Feenstra & Cornelius, Plantinga Jr., (eds) Trinity, Incarnation, and Atonement (Notre Dame IN: University of Notre Dame Press), 21–47.Google Scholar
Swinburne, R. (1994) The Christian God (New York: Oxford University Press).CrossRefGoogle Scholar
van Inwagen, P. (1995) ‘And yet they are not three Gods but one God’, in idem God, Knowledge & Mystery: Essays in Philosophical Theology (Ithaca NY: Cornell University Press), 222–259.CrossRefGoogle Scholar
Wierenga, Edward (2004) ‘Trinity and polytheism’, Faith and Philosophy, 21, 281–294.CrossRefGoogle Scholar