Philosophy of Religion 35: 1-20. 1994. © 1994 KluwerAcademic Publishers. Printed in the Netherlands.
The God of classical theism and the doctrine of the Incarnation CHARLES J. KELLY Department of Philosophy, Le Moyne College, Syracuse, NY, USA
The tradition of classical theism (CT) from Augustine to Aquinas developed a philosophical theology which posited for God a knowledge of the world that Aristotle's first unmoved mover arguably lacked. Yet this timelessly eternal God was still viewed as essentially simple and immutable. In addition, it was considered to be without either contingent attributes or any intrinsic relations to something other than itself. It cannot suffer and eternally knows that it cannot. Is it not then logically impossible for such an unchangeable and impassible deity to achieve an identity with a creaturely condition so that it could be said that God is some one that suffered, died, and sufferingly knew the evils of our world? Must a classical theist not acknowledge that acceptance of the reality of such an ontological empathy requires a leap of faith contra rationem? Must CT not adopt the stance of one of its staunchest proponents, Kierkegaard's Johannes Climacus, and concede that for it the Christian doctrine of the Incarnation constitutes an 'Absolute Paradox' genuinely believed only if its intrinsic absurdity and rational offensiveness are humbly acknowledged? 1 Negative responses to these questions are best developed in a logical framework historically associated with CT and one intrinsically capable of explicating the amphibolous character of much theological discourse. Some aspects of this logic, an Aristotelian syllogistic not relying on a twoname theory of the proposition, will be sketched in Section 1.2 The task of clarifying, modifying, and defending some of the basic theses of CT by means of this logic can then be undertaken in Section 2. Finally, in Section 3, it can be argued that one widely accepted way of stating the Christian doctrine of the Incarnation is amphibolous. In this respect the 'Absolute Paradox' bears similarities to some of the basic claims of CT itself and to Russell's antinomy. It is true that some of Aquinas's own formulations of the doctrine of the Incarnation conflict with CT's conclusions about the
2 divine nature. Other formulations, however, can be shown to be at least not patently incompatible with CT once two senses of identity are differentiated and an important dimension of the logic of negative names is clarified.
1. Some basics of a revised syllogistic
(a) The syllogism and amphibolous sentential strings. As the power of a logic is best revealed in its capacity to solve puzzles, it is fitting to begin by understanding how syllogistic can handle one recorded by Petrach. In the midst of a letter to Tommaso da Messina, the great humanist embellished his attack on the absurdity of the Afistotelian logic of his day by reminding his Sicilian friend of that story about Diogenes who was attacked by an abusive dialectician who said: 'What I am you are not'. When Diogenes agreed, the dialectician affirmed, 'I am a man'. When Diogenes did not deny that, the clown added, 'therefore you are not a man'. Then Diogenes answered, 'your conclusion is in fact false, and if you wish to make it true you must begin your syllogism with me'.3 Though Petrach himself would no doubt object, the dialectician's syllogism is worth examining. It might be laid out as: [A]
What I am you are not I a m a man You are not a man.
To capture the validity of Argument [A], Aristotelian logic must appropriate in its own fashion the Russellian insight that the grammatical subject and predicate of a sentential string need not be its logical subject and predicate. Accordingly, this logic can parse each sentential string by surrounding its logical subject with brackets and its logical predicate with parentheses. In italics it can render the main syntactical device which, as it quantifies the subject, shows whether the predicate is being attached to or detached from the subject. Its affirming and negating role assumed by the quantifier, the copula is no longer operative. Pace Petrarch and his interlocutors, Argument [A] can then be cast as a sound syllogism in Darii: [A*]
What [I am] (you are not) (I am) a [man] (You are noO a [man].
The universal affirmative major premiss reads: Whatever is such that I am identical to it - is such that you are not identical to it. The minor is the particular affirmative: some man - is such that I am identical to it. The true conclusion is the particular proposition: some man - is such that you are not identical to it. In presuming the sentential string 'You are not a man' .to be false, Diogenes has implicitly construed it as the singular '[You] (are not a man)'. Here, it should be noted, the main syntactical device is not expressed: where the subject is a proper name or singular personal pronoun the quantifier is indifferently universal or particular and is fittingly suppressed. 4 It is clear that the sophism arising from Argument [A] trades on the amphibolous structure of 'You are not a man'. The illegitimate inference from the true particular '(You are not) a [man]' to the false singular '[You] (are not a man)' constitutes what might be called a fallacy of amphiboly. The puzzle is dissolved once it is recognized that two distinct parsings of the same sentential string can be different in meaning and truth value. (b) The parsing of complex subjects and predicates. In calling ' [You] (are not a man)' false, we must be more precise. It is actually a more finegrained parsing of this proposition that is false. To understand this, it is first necessary to recall that in the traditional Aristotelian Square of Opposition there are four propositional forms: 'Every [S](P)' (universal affirmative); 'Some [S](P)' (particular affirmative); 'Not Some [S](P)' or 'No [S](P)' or 'Every [S] not (P)" (universal negative); and 'Not Every [S](P)' or 'Some [S] not (P)' (particular negative). Now, it is possible to parse any proposition not only in terms of its main subject, main predicate, and main syntactical device or quantifier. Where the subjects and predicates themselves are complex terms, it is also possible to parse them in the same triadic fashion. Thus the singular '[You] (are not a man)' with its suppressed main quantifier could be refined as '[You] ((are) not a [man])'. In this rendering bold brackets surround the main subject and bold parentheses the main predicate. 5 The complex predicate itself is herein parsed as a universal negative 'proposition' rendered in normal form as 'not a [man] (are). The total claim is false and is contradictorily opposed to the true '[You] ((are) a [man])'. But, '[You] (are not a man)' is obviously true when refined as '[You] ((are not) a [man])'. For the latter, with its predicate functioning in effect as a particular negative 'proposition', is contradictorily opposed to the false '[You] ((are) every [man])'. (c) 'That' as a lexically ambiguous syntactical device. The notion that
4
subject and predicate terms themselves can be parsed has significant import both logically and theologically. Theological ramifications will be seen in Sections 2 and 3. The logical implications concern the nature of the Aristotelian Square vis-~t-vis the square developed in predicate logic. The key idea here is the recognition of 'that' as a lexically ambiguous syntactical device capable of functioning either in a conditional sense ('thatc') or in a conjunctive sense ('thatk'). 6 To capture this ambiguity one might consider the fact that predicate logic often gives 'Every thing that is a unicorn is white' a conditional interpretation: 'Every thing - is such that if it is a unicorn then it is white'. This is a paraphrase of the basic string which Aristotelian logic can parse as 'Every [thing] (that c [is a unicorn] (is white))', wherein 'that e' as a syntactical device shows the bracketed 'is a unicorn' serving as the antecedent subject of the predicate and the parenthetical 'is white' serving as the consequent predicate of the predicate. The regimented contradictory opposite, 'Some [thing] not (that c [is a unicorn] (is white))', is equivalent by obversion to 'Some [thing] ([is a unicorn] and (is not white))', which itself is also naturally expressed with 'that' assuming the syntactical role of conjoining the terms in the predicate: 'Some [thing] (that k [is a unicorn] (is not white))'. The result is that as in predicate logic the universal affirmative 'Every [thing] (that c [is a unicorn] (is white))' entails the particular affirmative 'Some [thing] (that c [is a unicorn] (is white))' but not the particular affirmative 'Some [thing] (that k [is a unicorn] (is white))'.
2. Three theses of classical theism
The scheme of parsing just illustrated can be utilized in explicating three basic claims of classical theism. (a) God is absolutely simple. This first thesis insists that all essentially affirmative propositions with 'God' as their logical subject are logically equivalent. 7 As God necessarily and eternally possesses every attribute and activity which can be regarded as complete, pure, and perfect, these must all be necessarily co-extensive. Thus, where '~b" stands for any verb or verb phrase (for example, 'knows' or 'is happy' or 'is identical to its existence') expressing any such self-contained activity or perfection (1) Whatever [eternally exists] (eternally and perfectly t~s) is a scheme for necessary truths both de dicto, and de re. And as its direct converse (2) Whatever [etemally and perfectly ~bs] (eternally exists) is likewise necessary, CT can echo Aquinas in claiming that in God to
exist, to know, and to love are identical.8 The thesis that God is absolutely simple is thus compendious for CT's commitment to the logical equivalence, that is, the mutual enthymematic entailment, of such de re and de dicto necessary propositions as '[God] (exists)', '[God] (knows)', and '[God] (loves)', wherein the predicates are said not with time but with eternity. Two points deserve mention. First of all, this admittedly difficult thesis also endorses the equivalence of such propositions as '[God] (is Perfect Knowledge)', '[God] (is Eternal Love)]', and '[God] (is the Existence of God)'. 9 The 'is' in these predications expresses the relational 'is identical with' and not the non-relational '=' which permits the intersubstitutivity of the terms which flank it. l0 CT need not commit itself to claiming 'God = Perfect Knowledge', 'Perfect Knowledge = Eternal Love', or 'God = the Existence of God'.ll Secondly, CT need not conclude that such necessarily co-extensive predicates as 'eternally exists', 'eternally knows', 'eternally loves', and 'is the Existence of God' are synonymous phrases. This is easily seen by considering such truth functionally equivalent propositions as '[A] (is trilateral)' and '[A] (is triangular)' or '[A] (has size)' and '[A] (has shape)'. Clearly their necessarily co-extensive predicates express concepts quite different in meaning. 12 (b) God cannot possess any contingent attributes. This thesis claims that in God there cannot be any contingent characteristics or accidents. 13 Such attributes would compromise God's pure and immutable activity. What is truly said of God is said only essentially and necessarily. A seeming difficulty with this thesis is lucidly expressed by Thomas Morris. 14 He claims that there are many contingent states of affairs. If Tom's now wearing a striped shirt happens to be one of them, then God certainly knows this fact. And, knowing this fact, God is in a specific state of knowledge, the state of contingently knowing that Tom is wearing a striped shirt. Morris thus concludes that there are both necessity and contingency with respect to God, that is, God possesses not only necessary properties but also contingent ones. Accordingly, he insists that the thesis at hand is false and whatever claims that lead CT to arffirm it are also false. Morris is implicitly suggesting that if Tom is now wearing a striped shirt, CT must accept (3)
[God] (eternally knows that Tom is now wearing a striped shirt)
as a contingent truth de re and de dicto. However, it is clear that CT would summarily dismiss (3) as necessarily false whether the parenthetical
predicate is taken to express a contingent or a necessary attribute. God's absolute immutability and independence require that 'in God there is no relation to creatures'. 15 This means that any reference to anything effected by God must not be contained in what is being predicated of God. (3) clearly violates this requirement. Fortunately, however, the underlying sentential string of which (3) is a particular rendition is receptive to other parsings which CT can acknowledge as true. One particularly appropriate rendering is (3.1)
(God eternally knows) that [Tom is now wearing a striped shirt],
wherein 'that' functions as the main syntactical device of the proposition. As the 'that' in (3.1) is lexically ambiguous, (3.1) can be refined either as the conditional (3.11) ([God] (eternally knows)) that e [Tom is now wearing a striped shirt] or as the conjunction (3.12) ([God] (eternally knows)) that k [Tom is now wearing a striped shirt]. Examination of both (3.11) and (3.12) indicates that what each predicates of God belongs eternally and necessarily to it and that each avoids expressing a relation in God to what is other. (3.11) is not contrived; it reflects CT's central intuition that any event has a divine essential and eternal activity, in this case God's epistemic activity, as a sine qua non. In fact, whatever occurs in time occurs in the presence of God's eternal knowledge; nothing can occur without it. Thus, Aquinas could call God's knowledge a cause of states of affairs. 16 Itself a necessary dictum, (3.11) contains a contingent bracketed antecedent subject and a de re necessary parenthetical consequent predicate. Assertion of the antecedent would even result by modus ponens in the affn'mation of God's eternal existence as '[God] (eternally knows)' and '[God] (eternally exists)' are equivalent propositions. (3.12) expresses the conjunction of the de re and de dicto necessary '[God] (eternally knows)' with the de re and de dicto contingent 'Tom is now wearing a striped shirt'. Again, there is no suggestion of God possessing a contingent property. In contrast to (3.11), (3.12) has the advantage of instantiating the intuition that propositions of the form 'A knows that p' entail 'p' .17 (c) God eternally knows evil. CT asserts God's self-awareness and can endorse 'God eternally knows God' whether this be parsed as
7 (4.1)
[God] ((eternally knows) [God])
or as
(4.2)
([God] (eternally knows)) [God].
In accordance with the previous two theses of CT, the predicates of these necessarily true propositions about God express nothing beyond the divine essence. This point should be kept in mind when analyzing a classical theist's assertion that God knows evils. 18 The coherence of the position requires that this claim be expressed as (5.1)
([God] (eternally knows)) every [evil]
rather than as (5.2)
[God] ((eternally knows) every [evil]).
The latter predicates an unacceptable relation in God to what is other than divine. With (4.2) and (5.1) serving as models of theologically correct parsings, CT can suggest that GOd knows evil and whatever else is other than GOd with the same eternal knowing activity by which it knows itself. It has been alleged that the doctrine of divine simplicity requires not only that God's knowing is single but also that what God knows is "single. 19 This charge can be deflected. It is correct to say that the truth of (4.1), a proposition with 'God' as its logical subject, does depend upon the subject of its predicate being 'God' or an intersubstitutable name; (5.2) is false precisely because 'evil' is the subject of the predicate: what God knows must be God alone when 'God' is laid down as the logical subject of the proposition. But in propositions of the form '([God] (eternally knows)) IS]', of which (4.2) and (5.1) are true instantiations, the subject can be 'God', 'evil', 'other things', and indefinitely many terms. The objects of God's single and immutable epistemic activity can be multiple and even variable, though only knowledge of itself can be truly predicated of God. These considerations justify the charge that a classical theist like Aquinas does not find in God 'anything which could appropriately be called, knowing the suffering and evil which transpire in our worM' .20 But this does not mean that Aquinas and CT in general cannot endorse (6)
([God] (eternally knows)) all the [suffering and evil which transpire in our world].
Similarly, while it is correct to suggest that Aquinas cannot claim (7.1)
[God] (eternally knows that some particular woman has gone blind)? 1
it should be acknowledged that he and CT in general can accept (7.2)
([God] (eternally knows)) that [some particular woman has gone blind],
whether 'that' be construed in the conditional or conjunctive senses. Though it is surely wrong to allege that CT cannot in any coherent fashion maintain that God knows specific evils and sufferings, it is certainly true that CT cannot in any way suggest that God sufferingly knows them. For CT, God's knowing is purely active and in no way passive; sorrow and pain of their very nature cannot be found in God. 22 '(God eternally knows) [suffering]' is acceptable to CT, but '[God] (suffers)' or '[God] (sufferingly knows)' or '(God sufferingly knows) [evil]' are not. The question arises: does this preclude God from being identical to some one who sufferingly knew the evils of our world? The answer is best pursued by developing a view of the logic of the Incarnation congenial to classical theism.
3. Classical theism and the logic of the Incarnation (a) The basic Christological amphiboly. Of particular interest to CT is the mutual compatibility of two propositions sometimes taken as capsule expressions of Christian orthodoxy: (8)
Jesus is God
and (9)
Jesus is a man. 23
Working under an anthropology which might be described as broadly Aristotelian, CT should maintain that reflection on any human's essential animality requires that (10)
Whatever [is a man] (is mortal)
be recognized as necessarily true. Now, as '[God] (is God)' and '[God] (eternally exists)' are both necessarily true, the doctrine of the necessary co-extensionality of all positive divine attributes requires that both 'Whatever [is God] (eternally exists)' and 'Whatever [eternally exists] (is God)' be recognized as necessarily true. As 'Whatever [eternally exists] (is not mortal)' is also necessarily true, CT must endorse (11)
Whatever [is God] (is not mortal)
as a necessary truth. But, given (11) and (10), it follows that the basic
assertions of Christian orthodoxy parsed as
(8.1)
[Jesus] (is God)
and (9.1)
[Jesus] (is a man)
constitute an inconsistent set. Thomas Morris's recent proposal for overcoming the inconsistency does so by in effect alleging a lexical ambiguity in the phrase 'is a man'. 24 But, finding it most difficult on his account to square CT's defense of extreme divine immutability with the doctrine of the Incarnation, Morris affirms God as neither altogether atemporal nor absolutely independent of its attributes. 25 Thus, from the perspective of CT, this proposal only saves the doctrine of the Incarnation from absurdity by flouting some of CT's most basic theological theses. Accordingly, CT must pursue a different course in analyzing the logic of the Incarnation. In lieu of analyzing 'Jesus is a man' as lexically ambiguous, it can notice its structural ambiguity. CT can then avoid inconsistency by rejecting the singular (9.1), while affirming the particular proposition (9.2)
([Jesus] is)) a [man],
that is, some man - is such that Jesus is identical to it. Denying that 'is a man' is truly said of Jesus is sometimes regarded as a Docetic dismissal of Christ's genuine humanity. 26 But in rejecting (9.1), a Christian classical theist is not rejecting the claim that Jesus is a man, but only favoring one particular analysis of it over another. This is not an ad hoc tactic, but an extension of a strategy CT developed in articulating its position on God's relation to what is other. In holding fast to (10) and (11), CT knows that it cannot avoid inconsistency by accepting 'Jesus is God' parsed as (8.2)
(Jesus is) [God].
As (8.2) is equivalent to '[God] (is Jesus)', being identical to Jesus would emerge as a divine attribute. But this characteristic cannot contingently belong to God without violating the thesis that in God there are no accidents. Nor can it belong necessarily to God without undermining the freedom that Christian doctrine attributes to the Incarnation. Furthermore, the version of divine simplicity endorsed by CT stipulates that all God's essentially positive attributes be necessarily co-extensive: it certainly cannot be claimed, however, that whatever eternally exists is identical to Jesus. Only (9.2) and (8.1) remain as parsings of the basic Christological sentences that CT can endorse as not patently inconsistent. Now, (8.1) entails not only '[Jesus] (is not mortal)', but also the
10 particular (12)
[Jesus is] someone that (is not mortal). 27
Through the necessary truth 'Every [man] (is mortal)', (9.2) entails in Datisi the particular
(13)
[Jesus is] someone that (is mortal).
Standing in subcontrary opposition, (12) and (13) can both be true. It is only through a fallacy of amphiboly that it can be suggested that the particular (13) entails the singular '[Jesus] (is mortal)', a proposition which does implicitly contradict (8.1). These results necessitate a twist on that aspect of the communicatio idiomatum which requires that both divine and human properties be attributed to Christ Jesus. From the standpoint of surface grammar 'Jesus' is the subject of the underlying strings expressed in (12) and in (13). However, it is the logical subject of neither claim. Thus, only in terms of surface grammar does Jesus have the divine characteristic of being someone that is not mortal and the human property of being someone that is mortal. If in this context it is said that it is Jesus qua GOd who is someone that is not mortal and that it is Jesus qua man that is someone that is mortal, it must be understood that the former claim is true in virtue of ($.1), which has 'is God' as its logical predicate, and the latter in virtue of (9.2), which has 'man' as its logical subject. Reduplicative assertions of similar surface form are here compendious for a variety of claims intertwined in a network of diversely parsed propositions and necessary truths. This serves to immunize them against the usual criticisms .28 (b) The incarnation as a paradox. By unpacking the amphibolous structure of 'Jesus is God' and 'Jesus is a man' in the light of both a theory of divine simplicity and an anthropology cognizant of human animality, CT has deflected the charge that the doctrine of the Incarnation expressed in these sentences is patently incoherent. Lest this approach to understanding CT and dissolving Climacus' 'Absolute Paradox' be regarded as idiosyncratic, it is helpful to note its similarity to what might be described as an Aristotelian diagnosis of Russell's paradox. As originally presented, this paradox arises in Russell's effort to determine whether the class of all classes that are not members of themselves is a member of itself or not: Let us first suppose that it is a member of itself. In that case it is one of the classes that are not members of themselves, i.e., it is not a member of itself. Let us then suppose that it is not a member of itself. In that case it is not one of those classes that are not members of themselves, i.e., it is one of those classes that are members of themselves, i.e., it is a
11 member of itself. Hence either hypothesis, that it is or that it is not a member of itself, leads to its contradiction. If it is a member of itself, it is not, and if it is not, it is. 29 The essence of this reasoning involves the derivation of a contradiction from the case situation described by (14)
The class of whatever is not a member of itself is the class of whatever is not a member of itself.
It can easily be shown that the assumption (15)
[The class of whatever is not a member of itself] ((is a member of) [itself]) 3°
together with (14) parsed as (14.1)
[The class of whatever is not a member of itself] ((is the class of whatever is not a member of) [itself])
with the predicate of its predicate ref'med as '[is the class of] whatever (is not a member of)' yields a contradiction. Moreover, with the assumption of the negation of (15), namely, (15')
[The class of whatever is not a member of itself] ((is not a member of) [itself]),
(14) parsed as (14.1) with the predicate of the predicate refined as '(is the class of) w h a t e v e r [is not a member of]' also yields a contradiction. The bottom line is that it is (14.1) understood as asserting the conjunction of two different construals of the predicate of its predicate which must be assessed as necessarily false. The dissolution of Russell's paradox emerges with the awareness that (14) is amphibolous. For construed as the identity claim (14.2)
[The class of whatever is not a member of itself] ((is) [the class of whatever is not a member of itself]),
it is necessarily true: the 'is' here serving as the predicate of the predicate means 'is identical to'. Once the essence of the reasoning is uncovered, the paradox that the sentential string (14) is at once necessarily true and necessarily false is easily dissolved: as (14.2) it is necessarily true while as a suitably construed (14.1) it is necessarily false. As was the case with the 'Absolute Paradox', antinomy is engendered by the illusion that different parsings of the same sentential string must have the same meaning and truth value. 31 (c) W h y G o d is not identical to a man. If the above exposition of CT's
12 alternatives with regard to 'Jesus is God' and 'Jesus is a man' is correct, it is clear that a classical theist cannot accept the relational identity 'God is a man', whether it be parsed as (16.1)
[God] (is a man)
or as
(16.2) (God is) a [man]. (16.1) posits a relation in God to what is other than God, contains a positive predicate expressing an attribute obviously not co-extensive with God's essential attributes, and through (10) entails the false '[God] (is mortal)'. As (16.2) is equivalent to 'A [man] (is God)', it entails, through (11), 'A [man] (is not mortal)', a proposition which gainsays the obvious anthropological truth 'Every [man] (is mortal)'. Thus, 'God is a man' cannot encapsulate for CT a precisely correct position in Incamational theology. The contention that CT cannot endorse the two basic parsings of 'God is a man' is at odds with Aquinas's view that both 'God is a man' (Deus est homo) and 'A man is God' (Homo est Deus) expresses true propositions. 32 It will be profitable to explore an argument Aquinas gives for his conclusion to 'God is a man'. He begins by recalling the kenotic text wherein Christ Jesus is described as one 'who though he was by nature God ... emptied himself, taking the nature of a slave and being made like unto men, and in habit found as a man'.33 He then argues: he who is in the nature of God is a man; but he who is in the nature of God is God; God is therefore a man. 34 This argument is likely intended by Aquinas as the expository syllogism [B]
The one that is in the nature of God is a man The one that is in the nature of God is God God is a man
in which the copula 'is' functions syntactically with 'God' as the minor term and 'man' as the major term. Of course, this typically medieval view of the form of the argument must be abandoned in favor of a parsed expository syllogism with 'The one that is in the nature of God' as the wildly quantified middle term or in favor of the parsed Disamis [B*]
Someone that [is in the nature of God] (is a man) Everyone that [is in the nature of God] (is God) Someone that [is God] (is a man).
13 But such replacements would be unacceptable to CT. The conclusion of [B*] is equivalent to '[God is] someone that (is a man)', a claim which entails a conflict with (10) and (11), propositions which CT is unwilling to surrender. CT must diagnose the major premiss of [B*] as false. If parsed as '(Someone that [is in the nature of God] (is)) a [man]' it could be true; but this parsing could undercut the validity of the argument. It is important to note, however, that argument [B*] itself might be reworked so as to yield a possibly sound argument. This can be accomplished by integrating aspects of Trinitarian doctrine into the expression of the doctrine of the Incarnation. Chalcedonian orthodoxy professes that in the divine nature are three persons, each of which is an individual, that is, some one. The belief is that it is one of these three, not the total Godhead, who assumed a human nature. In this light, [B*] can be recast as~
[B**] Some [one] (that c [is a man] (is in the nature of God)) Every [one] (that c [is in the nature of God] (is God)) S o m e [one] (that c [is a man] (is God)). 35
The conclusion is equivalent to '((God is) some [one] that c [is a man])'. Though [B**] would still be valid if in the first premiss and in the conclusion 'that c' were replaced by 'thatk', the use of the conditional 'that c' indicates that being identical to a man is sustained by or necessarily conditioned by being in the nature o f God and by being identical to God. In fact, this construal conforms nicely with Aquinas's own insistence that the humanity of Christ is not juxtaposed to his divinity but is a living instrument united to the divinity, an instrument which itself freely acts even as it is acted upon by the divinity.36 But, perhaps more significantly, it reveals that CT can acknowledge the possible truth of such claims as '((God is) some [one] that c [suffered and died and sufferingly knew the evils of our world])', while still viewing '(God is) someone that [suffered and died]' and '[God] (suffered and died)' as impossible. 37 (d) The Incarnation as a free act of God. Some theists try to preserve the freedom of the Incarnation by implicitly claiming that while '[Jesus] (is God)' is de dicto and de re necessary, '[Jesus] (is a man)' is in both senses contingent. 38 For CT, however, acceptance of '[Jesus] (is God)' precludes acceptance of '[Jesus] (is a man)'. For it, the fact that '(Jesus is essentially identical to) a [man]' is consistent with but not derivable from '[Jesus] (is God)' is sufficient to preserve the doctrine of the Incarnation as expressive of a free act of God. Now, though '[Jesus] (is God)' is sometimes regarded as more basic
14 than '(Jesus is) a [man]' in articulating the logic of the Incarnation, the former itself derives from the relational identity '[Jesus] ((is) [God the Son])', wherein 'God the Son' is to be taken as a proper name referring to the Second Person of the Trinity. If 'S' abbreviates 'God the Son', '[Jesus] (is S)' entails '[Jesus] (is God)' through the necessary truth 'Whatever [is S] (is God)'. Notice the claim is not 'Jesus=S', that is, '[Jesus](S)'. In fact, to preserve the freedom of the incarnation it is necessary to leave open the possibility that God the Son could be what it is without being identical to Jesus. As Jesus ~ S or [Jesus] not (S), '[S] (is God)' neither entails nor is entailed by '[Jesus] (is God)'. It is thus appropriate to deny '[S] (is identical to Jesus)' while affLrming '[Jesus] (is identical to S)'. There is no relation in S to Jesus, though Jesus is eternally in a relation of identity to S and to God. (e) The logic of negative names. We have proposed that CT reject '[God] (is Jesus)', a proposition equivalent to (8.2). Now, '[God] not (is Jesus)' and '[Jesus] (is Jesus)' entail in Cesare '[Jesus] not (God)', that is 'Jesus God'. Thus CT posits both '[Jesus] not (God', a non-relational claim
denying identity, and '[Jesus] (is God)', a relational claim affLrming identity. By obversion, the negative proposition '[Jesus] not (God)' is equivalent to the affirmative proposition '[Jesus] (not God)'. However, if (17)
Every [not God] ((is not) [God])
is necessarily true, then the stance that we have urged CT take towards the Incarnation is a patently inconsistent one. It thus behooves CT to tarnish the credentials of (17) as a necessary truth. (i) As a first step in doing this, let 'X' represent any general name. Any proposition of the form 'Every [X] (is a X)' can then be intuitively recognized as a necessary truth. Those of the form 'Every [not X] (is a not X)' have the same intuitive appeal. But, as 'Every [not X] ((is) not a IX])' entails the at best unlikely 'Some [X] ((is) every IX])', it is not a form expressive of necessary truths. Argument [D] reveals this entailment: [D]: (1) Every [not X] ((is) not a [X]) (2) Every [not X] ((is not) a [X]) (3) Some [X] not ((is not) a [X]) (4) Some IX] ((is) every IX])
[As.] [1] 39 [2, Inversion] [3, Obversion]
It might be objected that in [D] the inference to (3) through the Full Inversion of a universal affirmative proposition is fallacious: the predicate term is undistributed in (2) and distributed in (3). This objection, however, is misplaced. Though it allows quantification within a complex subject or predicate, the understanding of propositional form used throughout this
15 paper rejects the notion of the predicate itself being quantified. In this regard it is helpful to note that Aristotle himself, not bewitched by the two-name theory of the text book tradition, neither understood nor justified the valid conversions of three of the four standard propositional forms by means of the quantification or distribution of predicate terms. Rather, he used indirect proof and proof by ecthesis. 4° (ii) The rejection through Argument [D] of the necessity of propositions of the form 'Every [not X] ((is) not a [X])' assumed 'X' to be a general name. But, as 'God' in (17)
Every [not God] ((is not) [God])
is assumed to be a wildly quantified proper name, Argument [D] is not decisive against its formal necessity. If, however, '[God] ((is) not a [not God])' is necessarily true by its form alone, (17) cannot be formally necessary. This is seen by considering Argument [E]: [E]: (1) (17) (2) (3)
[God] ((is) nota Every [not God] Whatever [(is) a Whatever [(is) a
[not God]) ((is not) [God]) [not God]] (not God) [not God]] ((is not) [God])
[Pr.] [As.] [1, Contraposition] [17, 2, Barbara]
Now, the simple converse of (E.3), 'Whatever [(is not) [God]] ((is) a [not God])', expresses a formally necessary truth: its complex predicate is the inverse of its complex subject. But as the complex subject of (E.3) does not formally entail its complex predicate, (E.3) itself is not a formally necessary truth. Assuming that (E.1) is formally necessary, CT is thus well within its rights in refusing to acknowledge (17) as formally necessary. While still accepting the formal necessity of 'Every [not X] ((is) a [not X])', CT can transfer its legitimate suspicions concerning the formal necessity of 'Every [not X] ((is) not a [X])' to propositions like (17) of the form 'Every [not X] ((is not) [X])'. A serious threat to the consistency of its assertions concerning the possibility of a freely effected divine incarnation has been blocked.
Notes
1. S6ren Kierkegaard, Philosophical Fragments, translated by Howard V. Hong (Princeton: Princeton University Press, 1962), pp. 61-67. 2. Accordingly, our revised Aristotelian logic will not view 'is' as a copula, that is, as a syntactical device connecting two names or noun phrases in a proposition. For arguments concluding to the inadequacy of Ockham's twoname theory vis-a-vis the doctrines of the Incarnation and the Trinity, see P.T. Geach, Logic Matters (Berkeley and Los Angeles: University of
16 California Press, 1972), pp. 289-301. 3. Francesco Petrarca, Rerum familiarium libri, I-VIII, translated by Aldo S. Bernardo (Albany: SUNY Press, 1975), pp. 38-39. 4. Fred Sommers (The Logic of Natural Language, Oxford: Clarendon Press, 1984, pp. 15-31) develops this Leibnizian notion of the 'wild quantification' operative in propositions with proper names or singularly referring personal pronouns as their logical subjects. He notes that as 'Every [Socrates] (is mortal)' is equivalent to 'Some [Socrates] (is mortal)', the quantifier in '[Socrates] (is mortal)' is naturally omitted. 5. If it were not suppressed as it is here, the main syntactical device would also he bolded. 6. The first recognition of 'that' as a syntactical device appears in George Englebretsen, Three Logicians (Assen: Van Gorcum, 1981), pp. 90-91, where it functions in effect as a 'wild quantifier' indifferently universal or particular. This is a legitimate construal, but one which will not be developed in what follows. 7. For statements of this interpretation of the doctrine of divine simplicity, see William E. Mann, 'Divine Simplicity', Religious Studies, 18 (1982), pp. 451-71; Norman Kretzmann, 'Abraham, Isaac, and Euthyphro: God and the Basis of Morality' in Hamartia (The Concept of Error in Western Thought), edited by Donald V. Stump et al. (New York: Edwin Mellen Press, 1983), pp. 27-50; Eleonore Stump, 'Hamart/a in Christian Belief', in Ibid., pp. 131--48; and Eleonore Stump and Norman Kretzmann, 'Absolute Simplicity', Faith and Philosophy, 2 (1985), pp. 353-82. 8. Thomas Aquinas, Compendium of Theology, I, translated by Cyril Vollert (St. Louis: B. Herder, 1947), ch. 50. 9. To avoid clutter, we shall sometimes only indicate the main subjects, predicates, and syntactical devices of a string. 10. Recognition of the difference between the relational 'is identical to' and the non-relational '=' is crucial. It is best captured by reflecting on the use of the former in a claim such as '[Plato] (is identical to itself)'. Here 'is identical to itself' or 'is itself' expresses a property necessarily possessed by every entity. One can confidently assert both '[Plato] (is identical to itself)' and '[Socrates] (is identical to itself)' as members of a consistent set. But the joint assertion of 'Plato = itself' and 'Socrates = itself' entails the absurd 'Plato = Socrates'. Accordingly, in expounding the Trinitarian convictions of Christian orthodoxy, CT would insist '[God the Father] (is identical to God)' and '[God the Son] (is identical to God)', while denying the Sabellian claims 'God the Father = God' and 'GOd the Son = God'. 11. Accordingly, the construals of the doctrine of divine simplicity mentioned in Note 7 above must be read with an awareness of the difference between these two senses of identity. It should also be noted that while CT need not accept the non-relational identity 'The Existence of GOd = the Eternal Love of God', it is committed to such non-relational identifies as '(The Existence of [God] is the Eternal Love of God)'. With the bracketed 'God' as its logical subject and the gapped 'The Existence of _ _ is the Eternal Love of God' as its logical predicate, the latter claims: God - is such that the Existence of it is identical to the Eternal Love of God. 12. See Mann, 'Divine Simplicity', pp. 462--64, and William F. Vallicella, 'Reply to Davies: Creation and Existence', International Philosophical
17
Quarterly, 31 (1991), p. 223. 13. Saint Augustine, The Trinity, translated by Stephen McKenna (Washington, D.C.: Catholic University of America Press, 1963), Bk. V, ch. 4, and Thomas Aquinas, Summa Contra Gentiles, translated by Anton C. Pegis (Garden City: Doubleday, 1962), I, ch. 23. Hereafter SCG. See also, Thomas Aquinas, Summa Theologica, three volumes, translated by the English Dominican Fathers (New York: Benzinger Brothers, 1947), I, 3, 6. Hereafter: ST. The Latin text and English translation of the Summa Theoiogiae contained in the various Blackfriars volumes published by McGraw-Hill have also been consulted and will be used from time to time. 14. Thomas V. Morris, Our Idea of God (Notre Dame: University of Notre Dame Press, 1991), p. 117. 15. Aquinas, ST, I, 13, 7c. See also ST, I, 45, 3, and Thomas Aquinas, On the Power of God, translated by the English Dominican Fathers (Westminster Md.: Newman Press, 1952), q.3, a.3. 16. ST, I, 14, 8c: 'Dicendum quod scientia Dei est causa rerum'. 17. It is Charles Hartshorne's failure (in The Divine Relativity, New Haven: Yale University Press, 1969, pp. 116-119) to distinguish propositions of the form instantiated by (3) from those of the form instantiated by the various construals of (3.1) that undermines his attack on CT's denial of contingent properties in God. 18. Aquinas, ST, I, 14, 10, and SCG, I, ch. 71. 19. Nicholas Wolterstorff, 'Suffering Love', in Philosophy and The Christian Faith, edited by Thomas V. Morris (Notre Dame: University of Notre Dame Press, 1988), p. 222. 20. Wolterstorff, 'Suffering Love', p. 223. Emphases are always those of the quoted author. 21. Wolterstorff, in 'Suffering Love', p. 223. 22. Aquinas, SCG, I. ch. 89. 23. See C.J.F. Williams, 'A Programme for Christology', Religious Studies, 3 (1968), pp. 513-524. Williams uses 'Christ' instead of 'Jesus', but for present purposes these names shall be interchangeable: 'Jesus = Christ'. 'Man', of course, is used in the gender inclusive sense. As the use of the indefinite article is crucial for subsequent discussion, 'Jesus is a man' is preferable to such locutions as 'Jesus is human' and 'Jesus is man'; it is also more convenient and less awkward than 'Jesus is a human' and 'Jesus is a human being'. 24. Thomas V. Morris, The Logic of God Incarnate (Ithaca: Cornell University Press, 1986), pp. 62-70. 25. Thomas V. Morris, Anselmian Explorations (Notre Dame: University of Notre Dame Press, 1987), pp. 124-138. See also, Thomas V. Morris, 'Metaphysical Dependence, Independence, and Perfection', in Being and Goodness, edited by Scott MacDonald (Ithaca: Cornell University Press, 1991), pp. 278-297, where, in opposition to CT, a 'mutual logical dependency' between God and God's properties is endorsed. 26. Williams, 'A Programme for Christology', p. 515. 27. 'Someone that' is a particular quantifier and a stylistic variant of 'Something that' and 'some'; 'that' here functions as a grammatical filler and is used neither in the conditional nor conjunctive senses. 28. For clear statements of these criticisms, see Allan B~ick, 'Aquinas on the
18
29.
30. 31.
32. 33. 34. 35.
Incarnation', New Scholasticism, 56 (1982), pp. 127-145, and Morris, The Logic of God Incarnate, pp. 45-55. Bertrand Russell, 'The Philosophy of Logical Atomism', in Logic and Knowledge, edited by Robert C. Marsh (London: George Allen and Unwin, 1968), p. 261. In (15), the many-worded name 'The class of whatever is not a member of itself' is to be taken as wildly quantified and individually referring. It must be noted that in contrast to (14.1) there is no reason to believe that (14) understand as (14.3) [The class of whatever is not a member of itself] ((is the class of) whatever [.is not a member of itself]) conjoined with (14.31) [The class of whatever is not a member of itself] ([is the class of] whatever (is not a member of itself)) entails a contradiction. What is distinctively important about (14.1) is that the wildly-quantified ' i t s d f is the subject of the predicate, a role it obviously does not play in either (14.3) or (14.31). Clarification of the above diagnosis can be effected by observing that the Grelling sentence "'Is not true of itself' is true of that and only that which is not true of itself' parsed either as "['Is not true of itself'] ((is true of) that and only that which [is not true of itself])" or as "(['Is not true of itself'] (is true of)) that and only that which [is not true of itself]" retains its aura of plausibility. But parsed as "['Is not true of itself'] ((is true of that and only that which is not true of) [itself])", it can easily be shown to be self-contradictory. Grelling's paradox like Russell's is engendered by a fallacy of amphiboly. Aquinas, ST, III, 16, 1 and 2. Philippians, 2:6-7. Aquinas, ST, 11I, 16, 1, Sed Contra: 'Et sic ille qui est in forma Dei, est homo. Sed ille qui est in forma.Dei est Deus. Ergo Deus est homo'. This valid inference relies on the principles: (P1) '[Every [S](P)] and (Some [S](Q))' entails 'Some [S] ([Q] and
(P))' and
(P2)
'Some [S] ([if P then Q] and (if Q then R))' entails 'Some [S] (/f [P] then (R))'. (P1) states that whatever is predicated a subject universally quantified can be conjoined with whatever is predicated of the same subject when particularly quantified. (P2) is a principle of the transitivity of implication for a conjunction of predicates attached to the same subject. 36. Thomas Aquinas, ST, III, 2, 6, ad 4, and ST, IIl, 7, 1, ad 3. See Paul G. Crowley, 'Instrumentum Divinitatis in Thomas Aquinas: Recovering the Divinity of Christ', Theological Studies, 52 (1991), pp. 451-475, where Aquinas's notion of instrumental causality is developed in the context of the problem presented by monothelitism, the doctrine that in Christ there is but one will, the divine. The medieval notion of a substance sustained by something distinct from it is developed in Alfred J. Freddoso, 'Human Nature, Potency and the Incarnation', Faith and Philosophy, 3 (1986), pp. 28-30. 37. The other claim that Aquinas accepted as true is 'A man is God' ('Homo est
19 Deus'). Since parsed as 'A [man] (is God)' it entails the false 'Some [man] (is not mortal)', CT must reject it. An argument that Aquinas gave for his position runs: "Romans states 'of whom is Christ, according to the flesh, who is over all things, God blessed for ever'. Now Christ, according to the flesh, is a man. Therefore the statement, 'A man is God', is true" (Aquinas, ST, 1II, 16, 2, Sed Contra: Blackfriars translation). It is likely that for Aquinas the essence of this argument could be displayed as the expository syllogism [C]
Christ is God Christ is a man A man is God
in which 'is' functions syntactically. As CT should reject the minor premiss parsed as '[Christ] (man)' or '[Christ] (is a man)', an expository syllogism modelled on [C] would for it be unsound. With the minor parsed as the acceptable '(Christ is) a [man]', there would be no valid inference. However, if [C] is transformed and rendered in Datisi as [C*]
Whoever [is Christ] (is God) Someone who [is Christ] (is a man) Someone who [is a man] (is God),
CT must accept the minor premiss which follows from the acceptable 'Christ is) a [man]'. Must it accept the major which has been used (by B~ick, 'Aquinas on the Incarnation', p. 128) as an interpretation of 'Christ is God' for the purpose of deriving a contradiction in the doctrine of the Incarnation? No, for '[Christ] (is God)' only entails "Someone who [is Christ] (is God)'. Pace Aquinas, CT must reject 'A [man] (is God)' and 'Someone that [is a man] (is God)', while again finding the Trinitarian based 'Some [one] (thatc [is a man] (is God))' acceptable. Aquinas's argument can be diagnosed as conflating either '[Christ] (is God)' with 'Whoever [is Christ] (is God)' or '(Christ is) a [man]' with '[Christ] (is a man)'. 38. Morris, The Logic of God Incarnate, pp. 52-53. It should be noted that Morris (p. 21) does not understand traditional Chfistology as regarding 'Jesus is God' as an identity claim but as a perfectly meaningful and true predication. For us, CT should acknowledge the relational identity claim '[Jesus] ((is identical to) [God])', but reject the non-relational identity 'Jesus -- God', that is, the directly convertible and wildely quantified '[Jesus] (God)'. In contrast to Morris, CT must see '[Jesus] (is identical to GOd)', '[Jesus] (is divine)', and '[Jesus] (is divinity)' as all logically equivalent. 39. As the predicate of (2) is the subalternate of the predicate of (1), 'Whatever [(is) not a [X]] ((is not) a [X])' is a formally necessary truth which can serve as the major premiss in the syllogistic derivation of (2) from (1). 40. Aristotle, Prior Analytics, in The Complete Works of Aristotle, Vol. I, edited by Jonathan Barnes (Princeton: Princeton University Press, 1985), Book I, ch. 2 (25a 14-25). The nature of ecthesis has proved notoriously difficult for scholars to unravel. But, by developing a suggestion of Gunther Patzig (Aristotle's Theory of the Syllogism, second edition, translated by Jonathan Barnes (Dordrecht: Reidel, 1968), pp. 161-164), we can understand it as
20 involving Aristotle's endorsement of (18) 'Some IX] (P)' is equivalent to 'Some [C] ([Every C X] and (Every C P))' as a logical law. ('Every [C](X)' and "Every [C](P)' are to be taken as the respective parsings of the conjuncts in the predicate of the right side expression.) What is significant about (18) is its pre-syllogistic employment of internal quantification of a complex predicate term without quantification of the predicate itself. The result is that the Full Inversion of universal affirmative propositions cannot be invalidated by failure to conform to a rule developed within the parameters of the standard two-name theory of the proposition.
Address for correspondence: Prof. Charles J. Kelly, Department of Philosophy, de Moyne College, Syracuse, NY 13214-1399, U.S.A. Phone: (315) 445 4494; Fax: (315) 445 4540