Transcendental Argument Revisited

[Published in Speaking the Truth in Love: The Theology of John M. Frame (Philipsburg, NJ: P & R 2009), 460-88]. Van Til and Transcendental Argument Revisited1 The phrase ‘cognitive dissonance’ first fell upon my ears in the fall of 1994 while enrolled in Professor John Frame’s course on ‘The Christian Mind’, the first in a series of three courses in Christian apologetics required for Master of Divinity students at Westminster Seminary in California. According to conventional philosophical wisdom, when philosophers run headlong into this sort of dissonance in the course of constructing arguments, they typically seek to overcome it, either by making a distinction or by defining a new term. Following in the footsteps of his apologetics mentor, Cornelius Van Til, Professor Frame added yet a third option, namely, humbly acquiescing in the possibility that such dissonance may in fact constitute a philosophical testimony to the truth of the ‘Creator-creature distinction’ and the Christian concept of mystery it entails. Not all forms of cognitive dissonance were to be regarded, therefore, as something analogous to a ‘charley-horse between the ears’ capable of being massaged away by the powers of reason and the tools of philosophical logic.2 It is crucial to keep this biblical wisdom in mind when it comes to assessing the debate over the distinctive character of transcendental argument in Van Til’s apologetic approach. While Professor Frame and I continue to assess this debate differently, the passage of years and 1An earlier edition of this essay appeared in the Fall 2003 issue of The Westminster Theological Journal, followed by a substantially abbreviated version in Revelation and Reason: New Essays in Reformed Apologetics (ed. K. Scott Oliphint & Lane G. Tipton; Phillipsburg, N.J.: Presbyterian & Reformed Publishing, 2007). For the purposes of this volume a number of changes and clarifications have been introduced. The essay thus appears here for the first time in a fully revised and expanded form and is gratefully dedicated to Professor John M. Frame in appreciation for the model of Christian charity, graciousness, and scholarship he has embodied through many years of teaching. 2 Not that Professor Frame himself resisted ‘making distinctions’ or ‘defining new terms’. Anyone who has taken his courses or read his writings on apologetics will know that he was not averse to making distinctions at crucial junctures in philosophical and 1 further reflection on this issue have lead me to a greater appreciation for the semantically rich and complex character of the concept of presupposition in Van Til’s apologetic. This should come as no surprise, especially since the theological reality the concept is linked to in apologetic argument is nothing less than the triune God of Scripture himself. It is precisely because its semantic richness and complexity in apologetic argument ultimately derive from its subject matter, that is, God in his triune identity as Father, Son, and Spirit, that it will always resist linguistic appropriation in the form of artificially constructed languages to some extent. As far as the languages of formal logic are concerned, the concept of referential presupposition will doubtless remain elusive, especially in the context of Christian-theistic apologetics. In what follows, therefore, readers hoping to discover a comprehensive formal account of the concept of presupposition will find themselves disappointed. This is not to say, however, that it is impossible to give formal expression to certain semantic distinctions that obtain between the concept of presupposition, on the one hand, and the rules of inference in standard proposition logic, on the other, the latter of which often serve as vehicles for stating traditional apologetic arguments (e.g., modus ponens and modus tollens). Insofar as Van Til himself attempted to justify this distinction, he typically did so in terms of theological concerns rather than formal argument per se. While I am far from suggesting that Van Til’s theologically-oriented approach to this issue lacks merit, ongoing debate over these matters attests to the fact that a formal basis for distinguishing his approach from traditional approaches is needed. The present essay seeks to address this issue by drawing upon certain insights into the concept of “semantic theological argument, or for that matter, defining ‘new’ terms when necessary (speaking more precisely, the latter usually amounted to redeploying ordinary language terms for specifically theological ends). 2 presupposition” that have arisen within the tradition of analytic philosophy.3 To that end I will begin with a survey of some important theological concerns underlying Van Til’s commitment to the distinctive character of transcendental argument, then move on to briefly summarize what I will call ‘the reductionist objection’ to transcendental arguments. Against this backdrop I will then attempt to provide a formal motivation for the distinctive character of Van Til’s transcendental argument from predication while at the same time distinguishing it from the method of reductio ad absurdum.4 This essay will then close with a discussion of possible objections in order to address potential misunderstandings and introduce further clarifications. I. Van Til and Transcendental Argument Central to the apologetic approach of Cornelius Van Til is the claim that a truly Christian apologetic and transcendental argument go hand-in-hand. Back of this claim lies the conviction, oft-stated by Van Til, that the Christian theism disclosed in Scripture entails a distinctive apologetic method. For example, in the opening pages of A Survey of Christian Epistemology Van Til writes that “every system of thought necessarily has a certain method of its own.”5 Thus Christian theism, considered as a coherent whole, requires an apologetic approach that is methodologically distinctive. For Van Til, this in turn requires the Christian apologist to employ a transcendental argument for God’s existence, since “the only argument for an absolute God 3 See especially Peter Strawson, An Introduction to Logical Theory (London: Methuen & Co. Ltd., 1952), 174-179; Bas C. van Fraassen, “Presupposition, Implication, and Self-Reference,” Journal of Philosophy (1968): 136-139. For a concise statement of the concept of semantic presupposition, see R. Bertolet, “Presupposition,” The Cambridge Dictionary of Philosophy (ed. Robert Audi; 2nd ed.; Cambridge: Cambridge University Press, 1999), 735. 4 Readers should note that my purpose in this paper is to defend the distinctiveness of transcendental arguments on a formal level. Such a defense, if successful, does not entail the conclusion that other argument types or forms have no place in a presuppositional apologetic. In general I agree, along with both Bahnsen and Frame, that “there is no transcendental argument that ‘rules out all other kinds of arguments’.....either in general philosophy and scholarship or particularly in apologetics” (Greg L. Bahnsen, Van Til’s Apologetic [Phillipsburg, N.J.: Presbyterian & Reformed Publishing, 1998], 502, n. 64). 5 Cornelius Van Til, A Survey of Christian Epistemology (Phillipsburg, N.J.: Presbyterian & Reformed Publishing, 1969), 5, hereafter SCE. 3 that holds water is a transcendental argument.”6 Criticism of Van Til’s stance on transcendental argument has not been lacking over the years, especially among Christian apologists who remain committed to inductive and deductive methods of argument, or who prefer to adopt a more integrative and methodologically diverse approach to the practice of apologetics.7 In recent years the distinctive character of Van Til’s transcendental approach has also generated critical debate within Van Tilian circles.8 This debate is especially significant for apologists who share a basic commitment to Van Til’s apologetic approach, inasmuch as it raises the question whether transcendental arguments can be distinguished from the traditional argument forms of natural theology, and if so, on what grounds. Before turning to a discussion of the formal issues at stake, it is helpful to begin by surveying a few of the more prominent theological concerns that motivated Van Til’s commitment to the distinctive character of transcendental argument in apologetic method and practice.9 This should also help clarify some of the reasons why he found traditional approaches to apologetic argument inadequate. Foremost among these theological concerns are the two closely related doctrines of God’s aseity and transcendence. For Van Til, safeguarding these 6 Van Til, SCE, 11. By way of qualification, one should note that Van Til’s commitment to transcendental argument did not lead him to reject the use of inductive and deductive methods of argument per se. However, in keeping with his belief that Reformed theology entails an apologetic method that is distinctive, namely the transcendental method, Van Til called for the methodological reconstruction of deductive and inductive argument along transcendental lines (see Van Til, SCE, 8-11, 201). It lies beyond the purview of this paper to enter into the question why Van Til himself chose not to provide us with formal examples of such a reconstruction in his writings. 8 See John M. Frame, Apologetics to the Glory of God (Phillipsburg, N.J.: Presbyterian & Reformed Publishing, 1994), 69-88, hereafter AGG; Cornelius Van Til: An Analysis of His Thought (Phillipsburg, N.J.: Presbyterian & Reformed Publishing, 1995), 317-320, hereafter CVT; “Presuppositional Apologetics,” in Five Views on Apologetics (ed. Stanley N. Gundry and Steven B. Cowan; Grand Rapids: Zondervan, 2000), 220-221, n. 18; Greg L. Bahnsen, Van Til’s Apologetic (Phillipsburg, N.J.: Presbyterian & Reformed Publishing, 1998), 499ff. 7 4 doctrines in apologetic practice requires one to make use of a transcendental argument. To do otherwise is to fail to “take seriously” the absolute character of God’s being when formulating an argument for Christian theism.10 The problem with traditional approaches to “inductive” and “deductive” argument, argues Van Til, is that they typically begin with the assumption that certain axioms are more ultimate or epistemologically certain than God’s existence (e.g. the principle of causality), then proceed by means of “straight line” reasoning to derive or deduce God’s existence from such principles.11 In so doing they unwittingly assign to the concept of God’s existence a logically derivative rather than logically primitive status, 12 thereby compromising both his aseity and his transcendence. By way of contrast, a transcendental argument preserves the logically primitive and absolute character of God’s existence by starting with the premise that God’s existence is a necessary precondition for argument itself.13 In this way argument is made to depend upon God, rather than vice versa, since argument is possible if and only if God’s existence is true from the outset of argument itself. Thus in contrast to both deductive and inductive forms of argument, a transcendental argument allows the concept of God to function as a logically primitive rather than logically derivative proposition, thereby bearing witness to the non-derivative character of 9 For a brief overview of these concerns, consult Van Til, SCE, 4-13. It cannot be stressed enough that Van Til embedded the transcendental argument within a distinctly Christian worldview. Failure to reckon with the implications of this fact has lead to a great deal of misunderstanding on the part of Van Til’s critics over the years. 10 “It should be particularly noted, therefore, that only a system of philosophy that takes the concept of an absolute God seriously can really be said to be employing a transcendental method. A truly transcendent God and a transcendental method go hand-inhand” (Van Til, SCE, 11). 11 Van Til, SCE, 8-11. 12 To my knowledge, Van Til never stated the matter in precisely these terms or categories (“logically derivative” vs. “logically primitive”). Their usage here hopefully serves to clarify the point that Van Til is making when he contrasts the “straight line” reasoning inherent in deductive argument with the “presuppositional” reasoning inherent in transcendental argument. 13 “It is not as though we already know some facts and laws to begin with, irrespective of the existence of God, in order then to reason from such a beginning to further conclusions. It is certainly true that if God has any significance for any object of knowledge at all, the relation of God to that object of knowledge must be taken into consideration from the outset. It is this fact that the transcendental method seeks to recognize” (Van Til, SCE, 201). 5 God’s existence on an argumentative level. To state matters another way, in Van Til’s Christiantheistic construction of transcendental argument, the truth of God’s existence is not a deductive consequence of the premises of the argument, but rather the ontological and logical ground for the very possibility of the premises themselves.14 This is undoubtedly one of the reasons, if not the chief reason, why he believed that transcendental arguments were uniquely suited for the task of placing into sharp relief the non-deductive character of the truth of God’s existence. In the second place, it was Van Til’s conviction that only a transcendental argument could do justice to the clarity of the objective evidence for God’s existence, since its peculiar form is specially suited to the apologetic task of bearing witness to the necessary character of God’s existence. That is to say, its formal character is such that it does not require the Christian apologist to implicitly grant the possibility that God’s existence is falsifiable, and thereby “tone down the objective claims of God upon men.”15 Inasmuch as creation clearly testifies to the necessary character of God’s existence, it follows that a Christian apologist cannot do justice to the objective evidence for Christian theism unless he or she affirms the non-contingent character of God’s existence in apologetic argument. Van Til often buttressed this claim by means of an argument from predication, insisting that a transcendental argument, theistically constructed, begins all argument upon the premise that predication requires for its possibility the necessary truth of God’s existence. In this manner the possibility of justifying predication is made to depend upon God’s existence from the outset of argument itself, thereby precluding any future possibility of using argument to falsify God’s existence. Argument cannot proceed without “The best, the only, the absolutely certain proof of the truth of Christianity is that unless its truth be presupposed there is no proof of anything. Christianity is proved as being the very foundation of the idea of proof itself” (Cornelius Van Til, The Defense of the Faith [Phillipsburg, N.J.: Presbyterian & Reformed Publishing, 1955], 396). Subsequent references to The Defense of the Faith (hereafter DOF) are to the 1955 edition unless otherwise noted. 14 6 predication, and predication requires for its possibility the necessary truth of God’s existence. At this juncture it is important to note that Van Til has in mind predication which affirms that something is the case (e.g., it is the case that John is tall), as well as predication which denies that something is the case (e.g., it is not the case that John is tall). In terms of assigning truth values to propositions, this amounts to the claim that it is ultimately impossible (logically speaking) to predicate truly or falsely about any proposition in the context of apologetic argument without referentially presupposing the truth of the proposition “God exists”.16 II. Van Til and Traditional Argument In contradistinction to the logical semantics of transcendental argument, Van Til argued that traditional constructions of the “theistic proofs” compromise the necessary character of God’s existence. Responding to criticisms made by S. J. Ridderbos in this connection, Van Til reminds him that for an argument to serve as a “witness” to God, it cannot bear witness to any other God but the “living and true God.” Thus it must bear witness to God as he truly is, and that in turn requires that it bear witness to God as “the One who cannot but exist.”17 In other words, in order to be a truthful witness to the triune God disclosed in Scripture, the logical semantics of apologetic argument must be congruent with the identity of the subject matter it seeks to defend. Christian-theistic argument must therefore bear witness to the necessary character of God’s selfexistence, and this precludes it from embracing the contrary premise that God’s existence, while 15 Van Til, DOF, 197. On the referential function of semantic presupposition, see John F. Post, “Referential Presupposition,” Australasian Journal of Philosophy 50:2 (1972): 160-67. On this account of the concept of ‘semantic presupposition’, to argue that the concept X referentially presupposes the concept Y is logically equivalent to asserting that Y truly ‘refers’ (i.e, that Y requires an existential, in this case the actual existence of God). ‘Referential presupposition’ rather than ‘semantic presupposition’ best summarizes what Van Til means by the concept of presupposition, inasmuch his version of TAG does not merely make a semantic or meaning claim (e.g., the meaningfulness of predication presupposes the meaningfulness of the concept of God), but a referential claim (the truth or falsity of predication presupposes that the proposition ‘God exists’ is referentially true). 17 Van Til, DOF, 197. 16 7 true, could be otherwise. Thus in the context of apologetic argument, the concept of God’s existence must not be allowed to function on the level of logical contingency, for to do so is to effectively grant the possibility that God’s existence is falsifiable. It is true, of course, that Van Til would sometimes argue the premise that God’s existence is falsifiable in order to perform a reductio ad absurdum of the non-Christian position. His use of the reductio, however, was the second part of a two-phase apologetic strategy wherein he adopted the unbeliever’s argument solely for the sake of argument. For this reason Van Til’s “practical strategy”18 of adopting the unbeliever’s stance for the sake of refuting it should be distinguished from his transcendental argument per se.19 At this point advocates of the traditional apologetic methods might object that Van Til’s endorsement of transcendental argument overlooks the fact that Anselm’s version of the ontological argument also argues from the necessary character of God’s being, and as such would be capable of addressing Van Til’s concern. At least one apologist in the Reformed tradition has gone even further and expressed the opinion that Van Til’s so-called transcendental argument is merely a sophisticated version of the ontological argument.20 Despite formal resemblances between the two, however, transcendental and ontological arguments for God’s existence are not merely two sides of the same coin. One must bear in mind that on Van Til’s view of the matter, the ontological argument is ultimately incapable of doing justice to the uniquely revelational sense in which God’s existence is necessary, since it “proves” a God who 18 Frame, CVT, 320. Frame suggests that Van Til’s transcendental argument is “essentially a reductio” (Frame, CVT, 315, 319). For further discussion of the reasons why I regard this as misleading, consult section VI of this essay. 20 Cf. the debate over apologetic method between Greg L. Bahnsen and R.C. Sproul, titled “The Bahnsen/Sproul Debate Over Apologetic Method,” in which Sproul asserts that the differences between Van Til’s transcendental argument and the ontological 19 8 exists “by the same necessity as does the universe,” and thus a God who is no more than “an aspect of, or simply the whole of, the universe.”21 He was aware of the fact that advocates of the ontological argument, and Anselm in particular, make a distinction between two different senses of “necessity” in order to distinguish God’s existence from that of the universe. For Van Til, however, this distinction is fatally undermined by the initial starting point of the argument itself. The ontological argument begins by defining God’s being as that being “than which nothing greater can be thought,” thereby identifying God’s being with humanity’s highest thought. In other words, the ontological argument begins by identifying God’s being with an order of thought and existence that is, on a biblical worldview, metaphysically contingent upon the creative decree of God. Moreover, even if a logical transfer into the realm of necessary being were possible by means of the ontological argument, such a transfer would not leave us with the biblical notion of God, a point often noted by Van Til himself: “If we take the highest being of which we can think, in the sense of have a concept of, and attribute to it actual existence, we do not have the biblical notion of God. God is not the reality that corresponds to the highest concept that man, considered as an independent being, can think.”22 In the practice of apologetics, therefore, one must distinguish the logical semantics of transcendental arguments from the semantics of arguments that are deductive in form, in this case the ontological argument. Those who equate the ontological argument with the transcendental argument implicitly assume that transcendental argument are non-substantive. Audio files of this debate are available from Covenant Media Foundation, 8784 FM 226, Nacogdoches, TX 75961 (http://www.cmfnow.com). 21 Cornelius Van Til, The Reformed Pastor and Modern Thought (Phillipsburg, N.J.: Presbyterian & Reformed Publishing, 1980), 64-65. 22 Cornelius Van Til, An Introduction to Systematic Theology (Phillipsburg, N.J.: Presbyterian & Reformed Publishing, 1974), 206. 9 arguments are reducible to deductive arguments, yet typically fail to provide grounds for this assumption, thus begging the very point in dispute. III. Van Til and the Argument from Predication There is yet a third reason why Van Til believed that transcendental arguments were uniquely suited for the task of Christian apologetics. Unfortunately this aspect of Van Til’s transcendental argument has not been given the due weight it deserves, even though it is precisely this feature that allows us to distinguish it on a formal level from traditional approaches to apologetic argument. At this point it is necessary to focus upon a particular theological and apologetic concern of Van Til’s, namely, the ontological basis for the possibility of predication. In his writings he frequently stressed the need for apologetic argument to engage this issue from a Christian-theistic point of view. Consider the following statement from A Christian Theory of Knowledge, which occurs in the context of Van Til’s stated purpose “to indicate in a broad way the method of reasoning that is to be pursued” in the vindication of Christian theism: How then we ask is the Christian to challenge this non-Christian approach to the interpretation of human experience? He can do so only if he shows that man must presuppose God as the final reference point in predication. Otherwise, he would destroy experience itself. He can do so only if he shows the non-Christian that even in his virtual negation of God, he is still really presupposing God. He can do so only if he shows the non-Christian that he cannot deny God unless he first affirm him, and that his own approach throughout its history has been shown to be destructive of human experience itself.23 Here we are reminded that for Van Til, the ontological Trinity constitutes “the final reference point in predication,” which is but another way of saying that the triune God of Scripture provides us with both the ontological and epistemological basis for the possibility of predication. 23 Cornelius Van Til, A Christian Theory of Knowledge (Phillipsburg, N.J.: Presbyterian & Reformed Publishing, 1969), 13 (emphasis added). 10 We are also reminded that Van Til directly identified presuppositional (or transcendental) argument with the task of justifying this possibility on Christian-theistic grounds. In other words, by means of a transcendental argument from predication, Van Til sought to make definite the claim that all human predication, whether that of affirmation or negation, referentially presupposes the truth of God’s existence: It is the firm conviction of every epistemologically self-conscious Christian that no human being can utter a single syllable, whether in negation or affirmation, unless it were for God’s existence. Thus the transcendental argument seeks to discover what sort of foundations the house of human knowledge must have, in order to be what it is.24 The last two quotes highlight the central position occupied by the transcendental argument from predication in Van Til’s apologetic, and this prominence has been noted by John Frame as well.25 Often overlooked, however, is Van Til’s concern to emphasize that the argument from predication is not limited to cases of affirmation, but also extends to cases of negation. Might not this emphasis merit closer scrutiny, especially in light of the distinction he drew between his own approach and the methods of traditional argument? Implicit in Van Til’s argument is the criticism, albeit undeveloped on a formal level, that traditional methods of argument are inadequate because they proceed upon the assumption that at least some types of predication are possible apart from the truth of God’s existence. By way of contrast, Van Til sought to argue that predication itself is impossible, philosophically speaking, unless the proposition “God exists” is true in a referential sense. Christian-theistic arguments for God’s existence must therefore make it clear that since his existence is the basis for all predication, one cannot predicate truly or falsely about anything in an apologetic context 24 Van Til, SCE, 11 (emphasis added). 11 unless the assertion “God exists” actually obtains. In the words of the late Greg Bahnsen: Van Til’s stunning application of this feature of transcendental argumentation to apologetics is that the truth of the Christian worldview is established not only by theistic premises and opinions, but also by antitheistic beliefs and opinions. As Van Til said, ‘Antitheism presupposes theism’ (SCE, xii). Even if the unbeliever wants to start with the assertion that ‘God does not exist,’ a transcendental analysis of it would show that the possibility of its coherence and meaningfulness assumes the existence of the very God that it denies.26 Stated negatively, the argument utilized by a Christian apologist must not grant the non-Christian assumption that predication, either in part or in toto, can be justified independently of the truth of God’s existence. Again, for Van Til the peculiar form and logical semantics inherent in transcendental arguments are uniquely suited to address such concerns. The question naturally arises as to whether Van Til was justified in thinking thus. Is it actually the case that the traditional arguments fail to do justice to the true character of the relation that obtains between God’s existence and predication? On the other hand, what are we to make of his confidence in the ability of transcendental argument to succeed where traditional arguments have failed? Is his conviction in this regard something that can be justified or is it merely a case of misdirected zeal on his part? In light of the preceding discussion, it would seem that the answers to these questions are to be found in a more precise clarification of the presuppositional nature of Van Til’s transcendental argument from predication. Before entering into this project, however, it is necessary to briefly consider what I will call “the reductionist objection” to Van Til’s belief in the distinctive character of transcendental argument. IV. John Frame and the Logical Semantics of Reduction Frame summarizes Van Til’s approach to theistic proof as the belief that “all legitimate theistic proof reduces to the ‘proof from the possibility of predication.’ God exists, in other words, because without him it would not be possible to reason, to think, or even to attach a predicate to a subject” (Frame, AGG, 70). 26 Cf. also the remarks of Greg Bahnsen in this regard (see Bahnsen, Van Til’s Apologetic, 502, n. 63). 25 12 Objections to the unique character of transcendental arguments are of some vintage in the history of philosophy, going back at least as far as Kant. The decade following the publication of Kant’s Critique in 1781 witnessed a number of critical responses to the distinctive claims of Kant’s transcendental program. Indeed, a number of Kant’s German contemporaries insisted that insofar as Kant’s transcendental program constituted an answer to Hume, it was merely restating arguments that had already been voiced by the rationalist philosopher Leibniz.27 Such criticism paved the way for later, more sophisticated attempts to deny the distinctive character of transcendental argument. To take but one example, in a series of articles published during the latter third of the 20th century, Moltke S. Gram mounted a sustained attack on the notion that transcendental arguments are formally distinct from deductive arguments.28 On Gram’s view, statements of the form “p presupposes q” are reducible to statements of the form “p implies q.”29 Hence there is at least some justification for classifying arguments of this type under the title “the reductionist objection.” To be sure, the descriptive term “reductionist,” like the term “rationalist,” admits of a broad range of uses. However its application in this paper is somewhat restricted and refers primarily to someone who claims that a relationship of deductive equivalence obtains between the form of a transcendental argument and the argument forms of standard propositional logic (e.g., modus ponens and modus tollens). Not surprisingly, the debate has also spilled over into Van Tilian circles. The late Greg 27 See Henry E. Allison, The Kant-Eberhard Controversy (Baltimore: The Johns Hopkins University Press, 1973). Eberhard was a contemporary critic of Kant who argued, according to Allison, “that whatever is true in Kant is already found in Leibniz, and that wherever Kant differs from Leibniz he is wrong” (ibid., 9). Kant’s own response to Eberhard came in 1790 in the form of a short essay entitled “On a Discovery According to Which Any New Critique of Pure Reason Has Been Made Superfluous by an Earlier One.” Although Eberhard never argued that a formal equivalence obtains between transcendental and deductive arguments, he nevertheless opened the door to such criticism by questioning whether Kant’s transcendental philosophy differed in substance from the deductive rationalism of Leibniz. 28 See the helpful bibliography on Gram provided in Stephen Wentworth Arndt, “Transcendental Method and Transcendental Arguments,” International Philosophical Quarterly 27 (1987): 43, n. 1. 13 Bahnsen and John Frame, arguably the two leading successors to Van Til in the 20th century, have weighed in on different sides of the debate, with Frame arguing in favor of the reductionist objection in a number of articles and books.30 Central to Frame’s argument is the claim that “any indirect argument can be made into a direct argument with some creative rephrasing.”31 In support of this claim Frame begins with an abbreviated statement of Van Til’s transcendental argument, then goes on to argue that it translates into an argument that is basically deductive in form. Thus in the final analysis, argues Frame, “it doesn’t make much difference whether you say ‘Causality, therefore God’ or ‘Without God, no causality, therefore God.’”32 A closer look at Frame’s program of reduction indicates that it turns upon the deductive relationship that exists between two rules of inference in formal logic known as modus ponens and modus tollens. In order to see this more clearly, it is necessary to state Frame’s argument more fully. Let us begin with Frame’s abbreviated statement of the direct argument, namely “Causality, therefore God.” Spelled out more fully, this argument takes the form of modus ponens, or “the mode of affirmation”: If causality, then God Causality Therefore God (premise 1) (premise 2) (conclusion) How does one get from this argument to Frame’s abbreviated statement of the indirect or transcendental argument, namely, that “Without God, no causality”? By means of modus tollens, or “the mode of denial”: Moltke S. Gram, “Transcendental Arguments,” Nous 5 (1971): 15-26. Relevant bibliography is cited in n. 8 above. It should be noted that in response to arguments advanced in the 2003 version of this essay, Frame has both clarified and qualified his position in some respects. See John M. Frame, “Reply to Don Collett on Transcendental Argument,” WTJ 65 (2003): 307-309. 29 30 14 If causality, then God (premise 1) Not God (premise 2) Therefore not causality (conclusion) Here modus tollens functions as a reductio ad absurdum for God’s existence. It assumes the proposition “not God” in order to refute it by deducing a conclusion from it that is obviously false (i.e., “not causality”). This refutation then serves to clear the way, as it were, for a positive affirmation of God’s existence. As Frame himself puts it, “Since we are unwilling to accept the conclusion, we must negate the premise and say that God does exist.”33 As noted earlier, Frame construes Van Til’s transcendental argument as an “indirect argument” that essentially functions as a reductio.34 On my view this equation helps explain Frame’s tendency to construe Van Til’s transcendental argument in terms of modus tollens, a practice that in my opinion is misleading (see further below). The important thing to note here, however, especially with respect to Frame’s program of reduction, is that the argument form modus tollens can be derived deductively from the common major premise it shares with the argument form modus ponens.35 Using the rules of replacement for standard propositional logic, such a derivation might appear as follows: 1. 2. 3. 4. 5. 6. (p  q)  ~ q pq ~q ~pvq ~p [(p  q)  ~ q]  ~ p (assumed premise) (by simplification from 1) (by simplification from 1) (by material implication from 2) (by disjunctive syllogism from 3, 4) (completed proof from 1 - 5) 31 Frame, AGG, 76. Frame, AGG, 76. 33 Frame, AGG, 76. By way of further clarification, one should also note that the conclusion “God exists” is not true unless premise 1 is true as well. In other words, Frame’s reconstruction takes premise 1 as a given. 34 See n. 19 above. 35 For a list of the rules of replacement commonly used in propositional logic, see Irving M. Copi, Introduction to Logic (7th ed.; New York: MacMillan, 1986). 32 15 As the above derivation demonstrates, in a complete propositional logic where the form modus ponens is valid, the form modus tollens will be valid as well. Thus one can move inferentially, as it were, from the common major premise modus ponens shares with modus tollens to the same conclusion one would have arrived at by using the argument form modus tollens. As far as the semantics of standard propositional logic are concerned, therefore, Frame is essentially right in his claim that “most positive arguments can be put into negative form and vice versa, with some skill in phrasing.”36 However, Frame goes even further. Having argued a case for the methodological equivalence of transcendental and traditional arguments, he goes on to suggest that Van Tilians should rest content with “a presuppositionalism of the heart” rather than continuing to insist upon the distinctiveness of presuppositionalism on a formal or methodological level.37 While some Van Tilians may be uncomfortable with the conclusion that presuppositionalism’s distinctiveness primarily consists in a subjective attitude rather than an objective method, such a conclusion is difficult to escape once the validity of Frame’s reductionist program is granted. It should be noted, however, that a more conciliatory reading of Frame’s argument is possible when one takes into account the following concession on his part: “I do not deny in principle that spiritual concerns can have specific methodological consequences. I am only saying that Van Til has not succeeded in proving that his spiritual concerns directly entail his methodological proposals.”38 Here Frame seems to be willing to grant in principle the possibility that Van Til’s “spiritual concerns” may in fact translate into “specific methodological consequences,” though he is 36 Frame, CVT, 318. Frame, CVT, 320; AGG, 85-88. 38 Frame, CVT, 320. 37 16 clearly skeptical about the possibility of making a case for it. At the same time it is important to note that, despite his reservations on the matter, Frame himself never fully closes the door to the possibility of making a formal case for the distinctive character of transcendental arguments.39 Nevertheless, Christian apologists from both sides of the apologetic fence have questioned the validity of Frame’s deductive reading of transcendental argument. In a recently published collection of apologetic essays, William Lane Craig asserts that Frame “confuses transcendental reasoning with what the medievals called demonstratio quia, proof that proceeds from consequence to ground.”40 This is but another way of saying that Frame confuses transcendental argument with a search for the premises in a deductive argument. While I agree to some extent with Craig’s observation, it should be noted that Craig does not interact with, much less refute, the case that Frame makes elsewhere in support of his particular reading of transcendental arguments.41 Thus Craig’s objection, although on the right track, fails to truly answer Frame.42 Criticism of Frame’s program has also arisen from within the household of Van Til. In a series of lectures given at Westminster Seminary in California,43 the late Greg Bahnsen argues that Frame’s denial of the distinctive character of transcendental argument rests upon 39 Renewed awareness of this fact on my part, coupled with the reflections registered at the outset of this essay, have lead me to the conclusion that when all is said and done, Professor Frame and I are actually not that far apart in our understandings of the issues at hand. 40 William Lane Craig, “A Classical Apologist’s Response,” in Five Views on Apologetics (ed. Stanley N. Gundry and Steven B. Cowan; Grand Rapids: Zondervan, 2000), 233. 41 Frame briefly references the case for his interpretation of transcendental argument in Frame, “Presuppositional Apologetics,” 220, n. 18. 42 Craig also attempts to identify Alvin Plantinga, rather than Van Til, as the true exponent of the transcendental argument for Christian theism in the 20th century. While Plantinga’s arguments provide us with a devastating critique of naturalism on its own terms, to my knowledge he nowhere makes the transcendental claim, as did Van Til, that the very intelligibility of the naturalist’s claims, whether true or false, necessarily presuppose the truth of Christian theism. 43 These lectures were taped and later transcribed into a booklet entitled An Answer to Frame’s Critique of Van Til: Profound Differences Between the Traditional and Presuppositional Methods. The booklet contains no publishing information, but 17 equivocation with respect to the meaning of causality. Bahnsen points out that when a Thomist makes use of a causal premise in traditional argument, that premise speaks of nothing more than the mere function of causality (i.e., for every effect there is a cause). On the other hand, when a Van Tilian makes use of a causal premise in transcendental argument, that premise concerns not merely the function of causality, but the ground of its intelligibility. Hence Frame is allegedly guilty of confusing descriptive claims about causality (Aquinas) with normative or regulative claims about its possibility (Van Til), thus turning a premise about the mere function of causality into a premise about its intelligibility. Bahnsen’s reply, however, ultimately misses Frame’s point. For Frame the problem is not whether Van Tilians and Thomists mean different things when they appeal to the concept of causality in apologetic argument, but whether these differences find expression on the level of apologetic method and formal argument as such. If they do not, then the sharp distinction Van Til posits between his method and the traditional method collapses, along with all attempts to distinguish, on methodological grounds alone, Van Tilian and Thomistic uses of the word “causality.”44 The question Bahnsen must answer, at least as far as Frame is concerned, is how we are to go about detecting the presence of such equivocation solely on the basis of argument form.45 Thus Bahnsen’s initial reply to Frame, while plausible in some respects, nevertheless fails to penetrate to the heart of Frame’s argument. Before proceeding to a concluding summary of Frame’s arguments, it should be noted by apparently derives from a series of guest lectures Bahnsen delivered in 1993 at Westminster Seminary in California. The booklet is made available by the Westminster Campus Bookstore in Philadelphia. 44 While Professor Frame is willing to grant that Van Til’s understanding of causality differs at important points from that of Aquinas, he apparently believes that one must consult Aquinas’ writings on nature and grace, rather than his apologetic method per se, to make a solid case for the claim that Aquinas defined the principle of causality in an “autonomous” fashion (email correspondence from John Frame to the author). 18 way of anticipation that a more recent argument formulated by Bahnsen in Van Til’s Apologetic comes much closer to addressing the precise issue raised by Frame.46 To my knowledge Frame has not directly replied to this argument in print. An attempt to further clarify and extend Bahnsen’s argument will be undertaken in section V of this essay, while at the same time adding a further argument of my own that attempts to exploit formal developments in the semantics of presupposition within the tradition of analytic philosophy. Frame’s argument, then, may be summarized in terms of two claims. First, that Van Til’s method of apologetic argument reduces to the traditional method in view of the relationship of deductive equivalence that obtains between the two. Consequently, Van Til’s attempt to draw a methodological distinction between his position and the argument forms of natural theology fails, since no such distinction exists. The second claim is closely related to the first, namely, that Van Til’s presuppositionalism is best understood “as an appeal to the heart rather than as a straightforward apologetic method.”47 One might be inclined to concede the case for “a presuppositionalism of the heart”48 were it not for the fact that another interpretation of the concept of presupposition is available in the work of Peter Strawson, one that arguably makes more systematic sense of Van Til’s transcendental argument, especially his argument from predication. Moreover, when this concept is used to clarify what is meant by the term “presupposition,” a plausible case can be made for the claim that transcendental arguments are not deductively equivalent with (or “How do we know when an apologist is assuming that the universe is intelligible apart from God? Usually, not from the form of his argument as such” (Frame, CVT, 319). 46 See Bahnsen, Van Til’s Apologetic, 501-502. 47 Frame, CVT, 320. 48 The arguments that follow should not be construed as an attempt to deny that presupppositionalism involves a heart attitude as well as an objective method. I heartily concur with Frame’s insistence that presuppositionalism, rightly understood, requires a 45 19 reducible to) the traditional argument forms of natural theology. Ironically, what Frame’s program demonstrates is not that transcendental and traditional arguments are deductively equivalent, but that the logical semantics of traditional argument forms are inadequate when it comes to capturing the distinctive concerns of Van Til’s apologetic. V. Distinguishing Presupposition from Implication By way of preface it should also be noted that while Van Til himself never provided a formal defense of the proposition that transcendental arguments are irreducible to either deductive or inductive arguments, it does not follow from this that he was unaware of the reductionist objection to his position. In A Survey of Christian Epistemology, a book that traces back to the earliest years of his teaching career, Van Til speaks of the distinction that exists between the transcendental method on the one hand, and the inductive and deductive method on the other: To us the only thing of great significance in this connection is that it is often found to be more difficult to distinguish our method from the deductive method than from the inductive method. But the favorite charge against us is that we are still bound to the past and are therefore employing the deductive method. Our opponents are thoughtlessly identifying our method with the Greek method of deduction. For this reason it is necessary for us to make the difference between these two methods as clear as we can.49 This passage serves as a reminder that the reductionist objection to transcendental argument is not new, nor was Van Til unaware of it. Nevertheless, there is truth in Frame’s claim that Van Til himself never provided us with an actual argument for its distinctiveness.50 What follows is a particular “heart attitude.” However, one need not concur with Frame’s deductive reading of the transcendental argument in order to agree with him on this point. 49 Van Til, SCE, 9. This work represents the second edition of a syllabus originally circulated by Van Til in 1932 under the title The Metaphysics of Apologetics. 50 “The first thing to note is that in this discussion Van Til has not presented us with an actual argument. He has presented (1) a conclusion, (2) a logical model, and (3) a practical strategy. . . . I confess that I am not convinced that a trancendental argument for Christian theism must of necessity be indirect rather than direct. To my knowledge, Van Til never argues the point, but merely asserts it. But it is by no means obvious” (Frame, CVT, 315, 317). 20 tentative attempt to do so by making use of twentieth-century philosophical discussion of the concept of presupposition and the subsequent application of this discussion to transcendental argument. Admittedly this will involve making use of ideas that, strictly speaking, do not appear in Van Til’s writings. Nevertheless, I believe that the clarity they lend to the concept of presupposition in Van Til’s approach to transcendental argument, and especially his argument from predication, will eventually justify their introduction. Perhaps not all Van Tilians will find my argument convincing. At the very least, however, it should serve to suggest a new avenue of approach to the question which others may perhaps build upon. The failure of traditional argument forms to capture what is meant by the concept of presupposition points up the need for a more precise way of construing the semantic relation between statements related by it. The most promising option that has emerged is arguably that of Peter Strawson. According to Strawson, a statement A may be said to presuppose a statement B if B is a necessary precondition of the truth-or-falsity of A.51 Strawson’s interpretation of the concept of presupposition has been restated in succinct fashion by Bas van Fraassen as follows: A presupposes B if and only if A is neither true nor false unless B is true.52 This may also be stated as follows:53 (1) A presupposes B if and only if: a) if A is true, then B is true. b) if ~A is true, then B is true. “A statement S presupposes a statement S' in the sense that the truth of S' is a precondition of the truth-or-falsity of S” (Strawson, Introduction to Logical Theory, 175). 52 Bas C. van Fraassen, “Presupposition, Implication, and Self-Reference,” Journal of Philosophy (1968): 137. Cf. also the application of van Fraassen’s work to Kant’s transcendental argument in Gordon G. Brittan, Kant’s Theory of Science (Princeton: Princeton University Press, 1978), 28-42. 53 Since A has no truth value (i.e., is neither true nor false) unless B is true, the truth of B must be presupposed whenever A has a truth value (i.e., whenever A is either true or false). Thus van Fraassen’s definition may be restated in terms of the conjunction given in (1) above. 51 21 Van Fraassen’s formulation is helpful for two reasons. First, it enables us to articulate more precisely Van Til’s claim that “no human being can utter a single syllable, whether in negation or affirmation, unless it were for God’s existence.”54 Second, it provides us with more formal language by which to articulate the differences between transcendental and traditional argument forms. To illustrate this, let us begin by applying the semantic relation embodied in (1a) to the causal argument for God’s existence. Letting C = causality, and G = God’s existence, we translate as follows: (2) C presupposes G C Therefore G (premise 1) (premise 2) (conclusion) A comparison of this argument form with modus ponens makes it clear, as van Fraassen has noted, “that an analogue of modus ponens holds also for presupposition.”55 Formal differences become apparent, however, when we negate the minor premise in (2) as follows: (3) C presupposes G ~C Therefore G (premise 1) (premise 2) (conclusion) Note that in terms of the characterization provided by (1), the corollary principle (1b) shows that (3) is valid, whereas this argument would be invalid for implication. Moreover, while (2) may be legitimately construed as an “analogue” of modus ponens, (3) is neither an instance of modus ponens nor strictly speaking an analogue to it.56 In other words, the argument form represented by (3) is apparently unique to arguments based upon the semantic relation of presupposition. We are now in a position to identify a distinguishing feature of arguments based upon the 54 55 Van Til, SCE, 11 (emphasis added). Van Fraassen, “Presupposition,” 137. 22 concept of presupposition as we have formulated it here. That feature concerns what logicians refer to as “truth-functionality.” In arguments (2) and (3), the truth value of the conclusion is not a function of the truth value of the antecedent minor premise (i.e., premise 2), since the conclusion remains true whether C or ~C obtains. By way of contrast, in the case of traditional arguments formulated in terms of modus ponens or modus tollens, the truth value of the conclusion is a direct function of the truth value of the antecedent minor premise. In Van Til’s Apologetic, Bahnsen calls attention to this peculiar feature of transcendental arguments. He summarizes the matter as follows: To put it simply, in the case of “direct” arguments (whether rational or empirical), the negation of one of their premises changes the truth or reliability of their conclusion. But this is not true of transcendental arguments, and that sets them off from the other kinds of proof or analysis. A transcendental argument begins with any item of experience or belief whatsoever and proceeds, by critical analysis, to ask what conditions (or what other beliefs) would need to be true in order for that original experience or belief to make sense, be meaningful, or be intelligible to us. Now then, if we should go back and negate the statement of that original belief (or consider a contrary experience), the transcendental analysis (if originally cogent or sound) would nevertheless reach the very same conclusion.57 Given the lack of a truth-functional relationship between the premises and conclusions of transcendental arguments, some may be inclined to question whether they should be regarded as actual instances of argument at all. However it must be remembered that “truth-functionality” cannot be invoked as a normative criterion for defining what counts as “argument” and what does not, since arguments expressed in terms of first-order quantificational logic also lack a truth- Cf. the discussion of ‘superficial resemblances’ to the forms of modus ponens and modus tollens in Irving M. Copi, Introduction to Logic (7th ed.; New York: MacMillan, 1986), 296-97. 57 Bahnsen, Van Til’s Apologetic, 501-502. 56 23 functional character, though no one would deny their status as arguments for that reason.58 Hence truth-functionality per se cannot serve as a decisive criterion by which to define what does or does not count as argument in apologetic discourse. On the other hand, as Bahnsen rightly recognized, this feature does serve to distinguish transcendental arguments from arguments expressed in terms of standard propositional logic (e.g., modus ponens and modus tollens). Generally speaking, then, presuppositional or transcendental arguments may be distinguished from traditional arguments based upon standard propositional logic, for instance modus ponens and modus tollens, in terms of “the truth-functional relation of their conclusions to their premises.”59 In view of this distinction, the claim that traditional forms of the causal argument yield “a transcendental conclusion”60 becomes questionable. To qualify as a transcendental conclusion, the truth value of the conclusion in such arguments would have to be in some sense independent of the truth value of its antecedent premise (i.e., non truth-functional). However, both modus ponens and modus tollens, two classic forms in which the causal argument has been traditionally expressed, fail to meet this criterion.61 The same must be said with respect to the claim that we can reach a “transcendental conclusion by many kinds of specific arguments, including many of the traditional ones.”62 In the nature of the case, the truth of a “transcendental conclusion” does not depend upon the truth value of its antecedent premise, regardless of whether this premise affirms causality or any other principle, since a transcendental conclusion constitutes 58 Cf. the discussion of quantification in Mark Sainsbury, Logical Forms: An Introduction to Philosophical Logic (Cambridge, M.A.: Basil Blackwell, 1991), 182-85. 59 Bahnsen, Van Til’s Apologetic, 501. 60 Frame, “Presuppositional Apologetics,” 220-221. 61 Noteworthy at this point is the fact that Frame construes the traditional causal argument for God’s existence in terms of both modus ponens and modus tollens (Frame, AGG, 76). 62 Frame, “Presuppositional Apologetics,” 220. 24 the very ground for the proof of that premise.63 Formal differences between the concepts of presupposition and implication also emerge when we consider the analogue to modus tollens for presupposition: (4) C presupposes G ~G Therefore ~C Again, whereas this argument would be valid for implication as an instance of modus tollens, it is not valid when C and G are joined by the semantic relation of presupposition, since in the latter case C has no truth value unless G is true.64 Thus the possibility of assigning a truth value to C depends upon the logically prior question whether G truly ‘refers’ or actually obtains.65 Thus if ~G obtains, we have what Strawson refers to as “a failure of presupposition,” in which case the possibility of assigning a truth value to C does not even arise. John Frame has rightly noted that Van Til’s apologetic method “seeks to show that all intelligibility depends on, or presupposes, Christian theism.”66 In light of the preceding arguments, however, there is reason to question whether traditional arguments expressed in terms of modus tollens meet this criterion. In argumentative instances where causality (C) and God’s existence (G) are related by implication, ~C follows from ~G. In other words, an argument formulated in terms of modus tollens contains the implication, albeit subtle, that at least some types of predication are possible in cases where God’s existence fails to obtain. By way of 63 The classic definition of a transcendental principle in argument was given by Kant in his Critique of Pure Reason: “But though it needs proof, it should be entitled a principle, not a theorem, because it has the peculiar character that it makes possible the very experience which is its own ground of proof, and that in this experience it must always itself be presupposed” (Immanuel Kant, Critique of Pure Reason [trans. Norman Kemp Smith; London: MacMillan, 1958], B765). 64 Recall that in terms of way we have construed the presuppositional relation, the truth value of C (i.e., its truth or falsity) depends upon the truth of G. Thus if G fails to obtain, then the possibility of predicating a truth value for C also fails to obtain. 65 On the referential function of the concept of semantic presupposition, at least as I’m construing it here, see n. 16 above. 66 Frame, CVT, 314-315 (emphasis added). 25 contrast, Van Til desired to argue that even cases of predicational negation presuppose the truth of God’s existence. In light of this, modus tollens would seem to be incapable of sustaining the apologetically radical goal he was aiming at with his argument from predication. To sum up, the problem with traditional argument forms is that they do not allow one to argue the proposition ~G therefore neither C or ~C, nor do they allow one to argue its corollary proposition ~C therefore G. The transcendental character of these propositions is evident from the fact that both of them depend upon the assumption that G constitutes a necessary precondition for the very possibility of assigning a truth value to C. Since Van Til clearly sought to establish these propositions via apologetic argument, and since traditional argument forms do not allow one to do so--indeed, in at least two cases actually invalidate these propositions--it follows that traditional argument forms simply cannot do justice to Van Til’s apologetic goals. It also remains questionable whether there is any meaningful sense in which one may continue to say that traditional arguments yield “transcendental conclusions.” At best such a statement is highly ambiguous and ultimately misleading. As noted in section IV of this paper, the reductionist objection to the unique character of transcendental argument rests upon the assertion that “p implies q” and “p presupposes q” are deductively equivalent propositions. However, if the arguments developed above are valid, as I am inclined to believe, then it follows that semantic differences between the concepts of presupposition and implication do in fact translate into differences on the level of formal argument and method. As such it is not possible to reduce “presuppositional” or transcendental arguments to arguments expressed in terms of standard propositional logic (e.g. modus ponens and modus tollens), all of which is to say that the reductionist objection to transcendental 26 argument fails. VI. Transcendental Argument and the Reductio ad Absurdum Strawson’s concept of presupposition, as formulated by van Fraassen, also allows us to sharpen the distinction between the method of reductio ad absurdum and Van Til’s transcendental argument from predication. In a reductio, a position is refuted by deducing a contradiction from its premises. In Van Til’s transcendental argument from predication, the possibility of assigning a truth value--and thus by extension the very possibility of generating a contradiction--fails to obtain unless God’s existence is already true (i.e., truly refers). In other words, Van Til’s transcendental argument from predication makes a stronger claim than the claim generated by the reductio. The latter generates a contradiction from the non-Christian position, while Van Til’s transcendental argument from predication makes the more radical claim that contradiction itself is impossible apart from the truth of God’s existence. To state the contrast in slightly different terms, if God’s existence is a necessary condition for the mere truth of causality, then denying God’s existence while affirming causality results in contradiction. However, if God’s existence is a necessary condition for both the truth or falsity of causality, then denying God’s existence results in a failure to predicate anything at all.67 This points up an important reason why the transcendental argument should not be confused or equated with the method of reductio ad absurdum. For Van Til it was not enough to deduce a contradiction from the non-Christian’s position and leave matters at that. Indeed, had 67 Strawson argues that the logical absurdity involved in self-contradiction should be distinguished from the logical absurdity involved in a failure of presupposition: “It is self-contradictory to conjoin S with the denial of S' if S' is a necessary condition of the truth, simply, of S. It is a different kind of logical absurdity to conjoin S with the denial of S' if S' is a necessary condition of the truth or falsity of S. The relation between S and S' in the first case is that S entails S'. We need a different name for the 27 Van Til stopped there it is doubtful whether he would have ruffled as many apologetic feathers as he did. Rather, Van Til insisted on going further and making the transcendental claim that the very intelligibility of the non-Christian’s claims, whether true or false, necessarily presuppose the truth of Christian theism. If there is an apologetic equivalent to the “offense of the cross” in Van Til’s method, this would be it. To be sure, in the context of apologetic argument the reductio helps clarify the nature of the presuppositional relation between God’s existence and causality, and it does this by pointing out contradictions that arise in the non-Christian position when God’s existence is denied. Strictly speaking, however, the reductio does not establish God’s existence, since the possibility of argument itself already presupposes the truth of God’s existence in a referential sense, and it is precisely the latter claim that Van Til’s transcendental argument from predication seeks to demonstrate. Van Til’s transcendental argument from predication helps us to see that the most fundamental question in logic and argument turns out to be an ontological one, namely that of God’s existence.68 For Van Til, God’s existence is an ontological presupposition that grounds the very possibility of logic, and thus (by extension) argument itself. This is doubtless the reason why Van Til’s apologetic method takes very seriously the essential character of the relation between God’s existence and argument–so much so that on Van Til’s view of things, the negation of God’s existence renders argument impossible. How so? By rendering impossible the task of assigning truth values in argument (i.e., predication). In this way Van Til’s transcendental relation between S and S' in the second case; let us say....that S presupposes S'” (Strawson, An Introduction to Logical Theory, 175). 68 The primacy of ontology over epistemology in Van Til’s apologetic traces back to his earliest writings, as evidenced by the title The Metaphysics of Apologetics, a work originally written by Van Til in 1932 (cf. n. 49 above). Viewed from this perspective, Van Til’s apologetic method is both thoroughly anti-Kantian and anti-modernist. 28 argument from predication takes us beyond the analysis of particular arguments and raises the question of argument itself. VII. Further Objections A few closing caveats in anticipation of possible objections.69 First of all, one might object that while the preceding analysis serves to clarify the concept of presupposition in Van Til’s apologetic, it fails to provide an actual example of transcendental or “presuppositional” argument. By way of response, it may be helpful to examine more closely a few of the assumptions involved in asserting that the following is an argument: (5) Causality implies God Not God Therefore not causality while at the same time asserting that the following is not an argument: (6) Causality presupposes God Not God Therefore neither causality or not causality (i.e., neither C or ~C).70 Note that (6) constitutes a formally valid instance of argument, given the definition of presupposition stated in (1) in section V of this paper. Hence the ground for privileging (5) while rejecting (6) as instances of argument appears to be motivated, at least in part, by a refusal to allow the introduction and integration of the concept of presupposition into one’s definition of 69 I would like to thank John Frame and James W. Allard (Professor of Philosophy at Montana State University) for reading over an earlier version of this essay and offering many helpful criticisms and suggestions. Their critical interaction provided the catalyst for many of the clarifications I have made throughout this essay. 70 The conclusion expressed by (6) should not be read as a denial of the existence of causality per se, but rather as a denial of the possibility of predicating truly or falsely about it. To state matters more generally, as an instance of transcendental argument, (6) concerns itself with those preconditions which secure the truth condition under which meaningful talk about the truth value of causality is possible. Again, the argument does not concern itself with establishing the ontological status of causality per se. 29 what counts as “argument.”71 Christian apologists should resist the rationalism lurking behind the Procrustean notion that the only arguments worthy of the name are those whose formal character can be articulated solely in terms of the semantic relations of propositional logic. Such a notion effectively forecloses the possibility of expanding the semantic domain of our concept of argument to include arguments similar in form to (6). Van Tilians in particular should be wary of adopting a form of rationalist dogmatism that refuses to admit the limitations inherent in the argument forms of propositional logic. Popular discussions of the discipline of logic often proceed as though it were a pure, objective science to be sharply distinguished from the more speculative and subjective character of other philosophical disciplines (e.g., metaphysics, epistemology, ethics, and aesthetics). If nothing else, continuing debates within the field of philosophical logic teach us that logic as a discipline is not immune from the influence of the ontological commitments of its practitioners, and thus should not be regarded an independent and objective science, wholly isolated from the theological and philosophical outlook of scholarly communities and guilds. More importantly for the point at hand, such debates should also teach us that the semantics of standard propositional logic and its accompanying argument forms do not constitute an infallible canon that is above criticism and/or expansion. Controversies surrounding the concept of validity in propositional logic, or what Mark Sainsbury refers to as “P-validity”, may be taken as a case in point. Reflecting on the adequacy of this concept, Sainsbury writes: “No one has ever supposed that Pvalidity exhausts the notion of validity, nor even that of formal validity...what is at issue is The most successful attempt to integrate the concept of “semantic presupposition” into formal logic is Bas van Fraassen’s. See van Fraassen, “Presupposition,” 136-152; cf. also the remarks of John F. Post, “Referential Presupposition,” Australasian Journal of Philosophy 50:2 (1972): 161, n. 4. 71 30 whether P-validity gives a partial characterization of validity, or formal validity.”72 One may nevertheless object that an argument like (6) begs the question, inasmuch as it assumes that a certain semantic relation between God and causality obtains from the outset. However, other commonly accepted forms of argument, for instance arguments expressed in terms of propositional logic, also make use of semantic relations and concepts that inevitably involve certain assumptions, a number of which are notoriously controversial among philosophers of logic. Thus there appears to be no reason why, prima facie, an argument that begins with the premise “C presupposes G” should be assigned a lesser status than an argument that begins with the premise “C implies G.” Indeed, one may go further and raise the question whether finite creatures can begin any argument without making semantic assumptions of some sort or other. The real question is not whether initial assumptions can be avoided, but whether subsequent argument confirms their soundness. The latter observation provides a convenient opportunity to address a possible misunderstanding of the nature of the claim being made here. In defending the right of (6) to lay claim to the title of “argument,” the additional claim is not being made that (6) proves the whole of Christian theism in one argument. While in my opinion (6) constitutes a legitimate instance of transcendental argument, it obviously does so in an abbreviated form. For this reason it is more accurate, and for apologetic purposes more useful, to regard (6) as an abbreviated transcendental argument. Other arguments may and in fact should be utilized in order to demonstrate the necessary character of the presuppositional relation between causality and God expressed in the 72 Mark Sainsbury, Logical Forms: An Introduction to Philosophical Logic (Cambridge, M.A.: Basil Blackwell, 1991), 51 (emphasis added). 31 major premise of (6). What (6) serves to illustrate is that an argument may be objectively valid from a formal point of view, yet insufficient from a practical point of view. In the concrete world of apologetic engagement one is certainly bound to make use of a variety of arguments. Admitting this does not entail the conclusion, however, that transcendental and traditional arguments cannot be distinguished from one another on formal grounds, or that they are somehow deductively equivalent to one another. Finally, some might object that the arguments advanced in this paper entail the conclusion that Christian apologists, and Van Tilians in particular, are somehow obliged to stop using the argument forms of standard propositional logic. After all, if argument forms such as modus tollens are formally invalid when arguments like (6) are operative, what place remains for the argument forms of standard propositional logic in a presuppositional apologetic? At this juncture it is helpful to keep in mind that in terms of the concept of presupposition set forth in this paper, both modus ponens and its presuppositional analogue represent valid forms of argument. This arises from the fact that semantic differences between the latter two argument forms, while clearly present, nevertheless do not register themselves on a formal level when the concept of presupposition is operative. Thus in the particular case of modus ponens, there is in fact formal overlap between an argument expressed in terms of propositional logic and its presuppositional analogue. Commenting on this nature of this overlap, van Fraassen writes: “Thus presupposition and implication are not the same, but they have something in common. What they have in common is that, if A either presupposes or implies B, the argument from A to B is valid.” 73 Taken together these considerations indicate that in argumentative contexts where modus ponens 32 is operative, methodological differences between presuppositional and traditional approaches do not register themselves on a formal level. Thus it would appear that modus ponens is amenable to integration with the logical semantics of transcendental argument on a formal level, and as such places no stumbling block in the path leading to Van Til’s apologetic. Modus tollens, on the other hand, is obviously a more difficult case. Here we find ourselves bumping up against the issues mentioned at the outset of this essay, thus returning full circle to the place where we began. Alongside its traditional apologetic applications, modus tollens now functions as a useful index of the limitations inherent in the artificial languages of formal logic, thus warning us of the dangers inherent in the ‘rationalist dogmatism’ mentioned above. From the fact that we have not been able to construct a logical language capable of fully reconciling these traditional applications with the logical semantics of presupposition, it does not follow that modus tollens should be immediately retired from the service of apologetic argument without further adieu. To do so would be to fail to properly respond to the partial nature of the characterization of ‘validity’ at work in the logical semantics of standard propositional logic. As an apologetic argument for God’s existence, modus tollens clearly has limitations and these limitations should be recognized as such. One must be careful at this point, however, not to throw out the philosophical baby with the bathwater. There are no compelling reasons for ‘presuppositionalists’ to religiously abstain from making use of modus tollens in any and all apologetic contexts, provided that they have been properly sensitized to the limitations inherent in the languages of formal logic. It is entirely possible and permissible to make a fruitful use of modus tollens in the context of apologetic argument while at the same time recognizing that its 73 Van Fraassen, “Presupposition,” 138.. 33 logical form is simply not capable of accommodating an important feature of the inner relationship between God’s existence and human predication. In this way the argument forms of standard propositional logic, properly qualified and thus circumscribed, find their true function and proper place under the larger umbrella provided by the semantics of presupposition and transcendental argument. VIII. Conclusion The philosophical journal Nous featured a symposium on transcendental arguments in 1971. Among the contributors to that symposium was Moltke S. Gram, who began his paper as follows: “The problem about transcendental arguments is whether there are any.”74 Obviously the passage of some 40 years has not rendered this question moot by any means. Secular philosophers certainly have not reached anything like a consensus on this question. From this it does not follow, however, that Christian apologists are somehow bound to share in Gram’s scepticism with respect to transcendental arguments. On the other hand, one must also grant that neither van Fraassen nor Strawson have said the last word on presupposition, and there may in fact be better ways to construe the relation of presupposition. There may also be different, and even better ways of stating the case for the distinctive character of “presuppositional” or transcendental argument. At the very least, however, the arguments advanced in this paper call into question the assumption that the concept of presupposition lacks formal and methodological significance. A plausible case can be made for the distinctive character of Van Til’s transcendental argument, provided one keeps an eye on the concept of presupposition and the distinctive way that it functions in his argument. 74 Moltke S. Gram, “Transcendental Arguments,” 15. 34