The problem with TAG

0. Introduction

In this article, I will discuss the ‘transcendental argument for the existence of God’ (henceforth ‘TAG’). This forms the backbone of the ‘presuppositional’ approach to Christian apologetics, first formulated by Cornelius Van Til (1895 – 1987). At its simplest, it is a radical defense of the Christian position, which boldly tries to dismiss any counter-argument with the claim that the notion of argument itself presupposes the Christian position. If this were true, then the opponent of Christianity would have to assume the truth of the position they oppose when presenting any argument at all. Instead of the Christian being on the back-foot and trying to respond to the attacks of their opponent (say with archeological evidence or biblical contradictions, etc), TAG is an attempt to switch the weight of the attack back at the non-Christian (who has to justify their ability to present an argument of this nature in the first place). While this is an ingenious way of arguing, it is ultimately flawed, as I will show in this paper.

Usually, opponents of the presuppositional apologetic will make one of two mistakes; either they will take the bait of the argument and try to justify how they can ‘account for truth on their worldview’, etc, or they will try to show that the Christian position is unable to account for this itself. Neither of this tactics is advisable, on my view. This is not because I do not think that a non-theistic account of things like argumentation, truth, logic, morality, etc, can be had; on the contrary, I think that coherent non-theistic philosophical positions for all of the above can be formulated. Nor is it that I think that the Christian position is free from internal consistencies; on the contrary, I find the typically given Christian explanations of these things to be lacking in several key aspects (Euthyphro, problem of evil, lack of sufficient justification for biblical claims, etc). However, analysis of the situation reveals that even if one concedes both of these points to the presuppositionalist they have still not made their case. For even if I am unable to account for argumentation (or logic, morality, etc), and even if the Christian position (or ‘worldview’) is able to account for this, we still have not been given a demonstration that the Christian position is a necessary condition for the given feature. What this requires is a general refutation of not just my position, but of every alternative position, leaving Christianity as the only possible account. Absent this argument, TAG has not been justified. Essentially, when the presuppositionalist claims that the Christian worldview is the necessary precondition (or ‘presupposition’) for the intelligibility of human experience generally, the response should be: prove it. They have not provided any such proof so far, and for reasons I discuss below, it seems unlikely any such proof can be forthcoming.

1. Transcendental Arguments

Transcendental arguments try to establish their conclusion, Y, by showing that it is required for the truth of some other proposition X; one cannot have the X without they without the Y, and we do have the X, so we must also have the Y. Consider the following argument:

1. One cannot ride a bicycle without having two legs
2. John is riding a bicycle
3. Therefore, John must have two legs.

So if this argument is correct, we can say that having two legs is a ‘necessary precondition’ for riding a bicycle.

However, let us suppose that there is a way of riding a bicycle with one leg; perhaps a little-known technique is available whereby the one-legged person can operate a bicycle just as well as a two-legged person. If this were true, then the argument from above becomes unsound, as premise 1) would be false. It would also mean that having two legs is no longer a ‘necessary precondition’ for riding a bicycle; it would be a sufficient precondition, in the sense that having two legs will suffice for being able to ride a bicycle, but it is not necessary if there is a way for a one-legged person to be able to ride a bicycle. The correct argument, adding in the consideration of the one-legged technique, would be as follows:

1. If one can ride a bicycle, then either they have two legs or they know the one-legged technique.
2. John is riding a bicycle.
3. Therefore, John either has two legs or knows the one-legged technique.

This shows that the claim of Y being a ‘necessary precondition’ for X requires that the only way X can be true is if Y is true also. If there is another way that X can be true, say by Z being true instead, then Y is not a necessary precondition for X. So part of showing that Y is a necessary precondition for X means showing that there is no Z which is a different precondition for X.

My point with TAG is that it has not been shown that there is no Z which could account for logic, morality, etc. This holds even if the Christian apologist can account for it, and even if their interlocutor cannot. What must be shown is not just that the interlocutor cannot, but no possible interlocutor can.

2. TAG Itself

I shall use the words of Prof. Michael Butler as a foil to argue against. Butler is a professor of philosophy at Christ College, Lynchburg, Virginia, and himself a graduate of Westminster Theological Seminary, which Van Til helped to found and taught at for over 40 years. I use his words not because I think Butler is an example of a bad presuppositionalist; on the contrary, he appears to be a very insightful representative of the position. Indeed, the clarity with which he discusses the issue allows a clear formulation of what is wrong with it, and for that I am indebted to him.

According to Butler, TAG runs as follows:

‘It starts with human experience; such things as science, love, rationality and moral duties.  It then asserts that the existence of the Christian God is the necessary precondition of such experiences.  Finally, it proves this indirectly by demonstrating the impossibility of the contrary’.(https://presupp101.wordpress.com/2014/07/14/the-transcendental-argument-for-gods-existence/)

The ‘indirect demonstration’ referred to in the mention of the proof proceeds by showing that the alternative explanations of these experiences are all inconsistent on their own terms:

‘We must point out to them [i.e. atheists] that univocal reasoning [i.e. reasoning independently of God] itself leads to self-contradiction, not only from a theistic point of view, but from a non-theistic point of view as well. It is this that we ought to mean when we say that we must meet our enemy on their own ground. It is this that we ought to mean when we say that we reason from the impossibility of the contrary. The contrary is impossible only if it is self-contradictory when operating on the basis of its own assumptions’. – Van Til, A Survey of Christian Epistemology (Philadelphia: Presbyterian and Reformed, 1969), 204-5

This is the method of ‘internal critique’, where one assumes the interlocutor’s position to show that it leads to internal inconsistencies, as opposed to keeping one’s own assumptions fixed when analyzing the interlocutor’s position.

So two things need to be established:

1. The Christian worldview can provide a non-self-contradictory account of human experience.
2. The non-Christian worldview cannot provide a non-self-contradictory account of human experiences.

Talk of ‘the Christian worldview’ and ‘the non-Christian worldview’ is to be taken with a pinch of salt (although this will prove controversial later). Obviously, there are lots of different denominations of Christianity, including reformed Presbyterian, Lutheran, Catholic, Greek Orthodox, etc. Equally, there are many distinct non-Christian positions, including every denomination of every other religious worldview, plus every variation of atheist worldview, etc. If we take this plurality of worldviews into account, then the claim is that at least one Christian worldview can account for the intelligibility of human experience, and that none of the non-Christian worldviews can. This claim has not been demonstrate by presuppositionalists, and I will argue that we have reason to doubt that they can demonstrate this claim.

Although this will be dealt with in much more detail as we go along, here is my argument in a nutshell:

1. TAG is successful only if every non-Christian worldview necessarily entails a contradiction (or is ‘internally incoherent’).
2. There is a potentially infinite number of non-Christian worldviews.
3. Either:
4. a) There is one way to establish that all the non-Christian worldviews are internally incoherent, or b) One proof is not enough but there is a finite number of ways to establish that they are all incoherent, or c) There is an infinite number of ways required to establish that they are all incoherent.
5. No proponent of TAG has established a); and it seems easy to prove that b) cannot be established (given a plausible formalization of ‘worldview’ as a set of beliefs); and if c) then it is not possible for a finite being to prove TAG.
6. Therefore, TAG has not been established, and is likely to be unprovable.

Although this argument may seem complex, my basic point can be easily apprehended. While Greg Bahnsen often presents animated demonstrations that a particular worldview is inconsistent (as with Islam and the way the Koran says that the Bible is a previous message of God etc) there is no attempt to actually prove that every non-Christian worldview is inconsistent. Yet, without this proof, the whole of TAG falls apart.

What we have instead is a rough and ready way of fending off other typically encountered worldviews, the sorts of positions encountered while engaged in disputations with non-Christians. It is like a manual for new apologists to go off into the wider world armed with. But of course, the standards for what might prove useful enough for an apologist to do his job are different than the standards for a philosophical argument such as TAG being successfully proved.

So, how do we establish that every other worldview is internally inconsistent? Fundamentally, that is the problem that must be resolved for TAG and the presuppositional position generally to be viable.

3. First Try – The Michael Jordan example

Michael Butler first responds to the problem by giving an analogy he attributes to Greg Bahnsen:

‘Suppose a basketball player, say Michael Jordan, beats every worthy opponent in one-on-one basketball games.  He can justifiably claim to be the best individual basketball player in the world.  Suppose further that another jealous (and peevish) basketball player who was previously trounced by Jordan resents that he (Jordan) has titled himself “the best player in the world.”  His comeback is, “just because you have beat every current player does not mean that there is not another one coming who is better than you.”  Jordan’s response can be anticipated; “bring on my next opponent.”  The theoretical possibility that there may be another player better than Jordan is not a concern to him.  In the world of basketball, it is the one who is actually the best player, and not who is possibly be the best player, that is of importance.  In the practice of apologetics, things are similar.  What matters are actual worldviews not possible worldviews’. https://presupp101.wordpress.com/2014/07/14/the-transcendental-argument-for-gods-existence/

Now this ‘comeback’ is demonstrably wide of the mark. In an important sense it is possible worldviews and not just the actual worldviews that matters here, even for the practice of apologetics. There is a logical gap between Jordan’s unbroken record of success and the truth of his being the ‘best’ basketball player. It is not logically impossible for him to win every game he plays and yet not be the best player. Likewise, in the case of worldviews, there is a logical gap between the apologist’s unbroken record of being successful in debate, and the truth of the Christian worldview she defends. It is not logically impossible for someone to win every debate they ever engage in, and yet argue for a false thesis.

The reason the standards are so strict is due to the strength of the claim being made by TAG, point 2) from above. The claim is that the Christian God is a necessary condition on human experience, this means that if there were even one possible non-Christian worldview that could also account for human experience just as well as the Christian one, then the Christian view would lose its status as a necessary condition on the intelligibility of human experience. It might still be able to lay claim to being a ‘sufficient condition’, but it could no longer be claimed to be a necessary condition, which was needed for the proof to go through. This would stop the argument working backwards from the intelligibility of experience to the existence of the Christian God. If, for example, there were an atheist position, Z, which could account for human experience just as well as the Christian one, then one could only work from the intelligibility of experience to the disjunction of the two positions; either the Christian worldview is true or Z is true. This would be no proof of the truth of the Christian worldview then.

If we were to take seriously Butler’s claim that what matters are actual worldviews, then we would have to weaken the argument accordingly. So if we rephrased his articulation of the argument, but made it explicit that we were only talking about actually held worldviews, it would look like the following:

It starts with human experience; such things as science, love, rationality and moral duties.  It then asserts that the existence of the Christian God is the necessary precondition of such experiences.  Finally, it proves this indirectly by demonstrating the impossibility of any of the actually held contraries.

This basically just says: there cannot be a consistent non-Christian worldview because I have not seen one before. It is an elementary inductive fallacy. Even if all of the actually held worldviews were internally contradictory, this would not mean that there wasn’t a possible worldview which wasn’t. So the argument thus stated is invalid.

Stating that we are doing apologetics, where the standards are lower, is not a response. The purpose of apologetics is to provide a ‘reasoned defense of the faith’. An invalid argument is not a reasoned defense of the faith.[1]

To bring out the absurdity of thinking that the Michael Jordan type response is viable consider that if it was actually held worldviews that mattered, rather than merely possible worldviews, then one could prove that God existed by killing everyone who wasn’t a Christian, thus leaving Christianity as the only actually held worldview. This, of course, would not prove it to be correct.

Even though Butler says that Bahnsen’s analogy “hits the mark”, he also sees that it is inadequate, and states the problem clearly:

‘If there are an infinite number of worldviews and TAG only refutes a small slice of them, if one may speak this way, then it has not established that Christianity is the necessary precondition of human intelligibility.  That is, even granting that TAG demonstrates the absurdity of all actual worldviews, it does not follow that all possible worldviews are likewise absurd.’

Butler acknowledges that Bahnsen’s response is unsuccessful, on the basis that “winning the debate and proving that Christianity is the necessary precondition of human experience are two different things”. In this, Butler is quite right.

4. Second Try – Only Two Worldviews

Butler goes on to provide a second response by Bahnsen, and this time one he seems to endorse. Here is the quote in full:

‘But Bahnsen makes the further point is that this criticism misses the thrust of TAG altogether.  TAG argues for the impossibility of the contrary (the non-Christian worldview) and not the impossibility of an infinite number of possible worldviews.  TAG does not establish the necessity of Christianity by inductively refuting each and every possible non-Christian worldview (as finite proponents of TAG, this is an impossible task), but rather contends that the contrary of Christianity (any view that denies the Christian view of God) is shown to be impossible.  And if the negation of Christianity is false, Christianity is proved true.  In other words, the structure of the argument is a disjunctive syllogism.  Either A or ~A, ~~A, therefore, A.

At this point the clever opponent will simply deny the first premise.  He will contend that it should not be construed as a disjunction of a contradiction, but a simple disjunction.  The argument should thus be restated along the following lines: A or B, ~B, therefore, A. And once this move is made he will then contend that while the argument is valid, the first premise involves a false dilemma.  That is, he will grant that given A or B and the negation of B, A does indeed follow, but nevertheless maintain that the argument is unsound because the first premise (A or B) is not true.  The reason being that there are more possibilities than just A and B.  Given a true first premise, A or B or C or D … n, the negation of B merely entails that A along with the disjunction of other propositions besides B (C, D,…n) follows.

In order for this to be successful, it is incumbent upon the opponent of TAG to defend two claims.  First, he must defend the contention that the original first premise is not the disjunction of a contradiction and, second, he must show that there are other possible disjuncts besides B (what we can call the view that is opposed to the Christian worldview).’

The argument seems to be as follows: There is a confusion about what TAG actually claims. In this telling, the argument only says that the negation of the Christian worldview is inconsistent, which means that the options needed to be refuted have gone from seeming to be infinite to actually only being one case. This means the process of performing the demonstration that no non-Christian worldview is consistent should be easy in principle to do, as only one thing needs to be shown to be inconsistent, namely the negation of the Christian worldview.

If the reply comes from the opponent of TAG that it is not a question of refuting only one case but of refuting an infinite number of them, then Butler claims that it is “incumbent upon the opponent of TAG to show that this is the case”, i.e. that “there are other possible disjunctions of other propositions besides B”.

However, this is a shifting of the burden of proof; the proponent of TAG is setting out their premises for the argument, and they have to justify them. It is not incumbent on the opponent of the argument to justify their negation first.

Even though the burden is being fallaciously shifted, it is a burden that is easily met. All that has to be established (in order to show that it is B or C or D… rather than just ~A) is that there is more than one distinct non-Christian worldview. So here are two: the atheist worldview and the Islamic worldview. Each of these are treated separately by Bahnsen, indicating that he thinks of them as separate entities, deserving of distinct refutation. If there was only one alternative worldview there would be no need for separate refutations in these cases. The quote by Kant that Butler puts at the very start of his article explains this very point:

‘If, therefore, we observe the dogmatist coming forward with ten proofs, we can be quite sure that he really has none. For had he one that yielded . . . apodeictic proof, what need would he have of the others?’   –Immanuel Kant

The fact that Bahnsen offers a distinct refutation to the atheist as to the Muslim indicates that he considers them to have separate worldviews, otherwise he would only come with one refutation.

And there is no reason to think that there are only two worldviews, because there are combinations of philosophical positions that can form separate worldviews. An atheist can be a materialist or an idealist, a nominalist or a Platonist, a monist or a dualist, a determinist or an libertarian, an intuitionist or a formalist, etc, etc. So surely, it is quite evident that there is a huge number of potential worldviews, and quite possibly an infinite number. So even though I deny that I had to prove that there is more than one non-Christian worldview (due to the illegitimate shifting of the burden of proof), it is a burden that is trivially easy to meet, so Butler is back facing the issue of having to rule out all of the distinct alternative worldviews.

However, there is an even more pressing issue. And this relates to the other point on which Butler claims that the opponent of TAG has to justify, namely: the original first premise is not the disjunction of a contradiction. In a sense, it doesn’t matter if we take it to be the disjunction of a contradiction. Bahnsen’s claim was that all that needs to be shown is that the negation of the Christian worldview was inconsistent. But what does the ‘negation of a worldview’ actually mean? The usual set-theoretical operation that negation corresponds to is compliment; i.e. the result of removing something from a domain of discourse. So imagine a set, W, which contains all the possible worldviews (including the Christian worldview, W_Ch) as its elements;

W = {W_Ch, W₁, W₂, … W_n}.

If negation is compliment, then the negation of the Christian worldview (‘~A’) would be everything in W that wasn’t the Christian worldview;

‘~A’ = {W – W_Ch} = {W₁, W₂, … W_n}.

Therefore, ‘the negation of the Christian worldview’ is just ‘all the worldviews that are not the Christian worldview’; they mean the same thing. To put it in Butler’s terms: ~A = {B, C, D…n}. Therefore it doesn’t matter if we have to refute the negation of the Christian worldview, because this means the same as refuting all the non-Christian worldviews. This would mean that Bahnsen has just restated the objection and not countered it at all. This is a simple logical issue. It indicates a lack of logical sophistication in Bahnsen’s argument, and presuppositionalism in general, that this hasn’t been grasped before.

In fact, it seems easy to prove that there cannot be one method which disproves every non-Christian worldview, because there cannot be one contradiction that they all share. One natural way of understanding worldviews is that a worldview is just a list of propositions that an agent believes to be true. So if I believe that God exists, then the proposition ‘God exists’ is in my worldview set. The set of all possible worldviews is then the set of all sets of propositions. If there are no further restrictions on this then it is logically provable (and trivially so) that the intersection of all the worldviews is empty; i.e. there is no common element, say p, to every set, and thus no contradiction, say p & ~p that they all share.

One could put the point even more simply, as follows. The claim is that every non-Christian worldview is internally incoherent. If by ‘worldview’ we understand a set of propositions believed to be true by an agent, and by ‘internally incoherent’ we mean that the set is inconsistent (i.e. contains a proposition and its negation), then consider the non-Christian worldview that contains only one belief, i.e. {p}. This set is plainly not inconsistent. The retort will likely be that this ultra-simple worldview cannot ‘account for the intelligibility of human experience’. If so, what are the minimal conditions under which a set of beliefs could achieve this? It is not on the opponent of TAG to provide this analysis; all she has to do is point out that without this analysis the proof cannot be claimed to be established. The proponent of TAG needs to provide this analysis as part of the proof itself.

5. Last Try – Blame Autonomy

There is one final attempt at bridging this gap, which directly focuses on this issue, by which I mean it directly focuses on something that all non-Christian worldviews have in common, and in virtue of which they are necessarily inconsistent.

What, then, is the nature of the non-Christian worldview? Simply put, all non-Christian systems presuppose that experience can be accounted for on autonomous lines. The non-Christian worldviews share the common feature that experience can be made sense of independently of God and his revelatory word. Butler – https://presupp101.wordpress.com/2014/07/14/the-transcendental-argument-for-gods-existence/

Here then, seems to lie the root of the issue for the Van Tillian presuppositionalist. They argue that they can conflate every non-Christian worldview on the basis that they are all ‘autonomous reasoning’ worldviews, as opposed to the Christian one in which independent reasoning is subordinated to the word of God[2]. So the idea is that there is something wrong with the assumption that one can reason ‘autonomously’, which inevitably leads to self-contradiction. Butler continues:

‘From this we can see that Van Til is correct, “We have constantly sought to bring out that all forms of antitheistic thinking can be reduced to one.” Bahnsen elaborates on this important insight:

‘Despite “family squabbles” and secondary deviations among unregenerate men in their thinking, they are united at the basic level in setting aside the Christian conception of God. The indirect manner of proving the Christian position is thus to exhibit the intelligibility of reasoning, science, morality, etc., within the context of biblical presuppositions…and then to make an internal criticism of the presuppositions of autonomous thought (in whatever form it is presently being discussed) in order to show that it destroys the possibility of proving, understanding, or communicating anything.’’

It seems like progress has been made towards addressing how one could prove every alternative worldview is internally inconsistent, namely by positing ‘autonomous reasoning’ as the common element in every non-Christian worldview. However, really nothing has been put forward to get round the basic logical objection. The quote from Bahnsen that Butler cites as ‘elaborating’ on Van Til’s important insight assumes that Van Til’s insight is true, otherwise we are back to the Michael Jordan problem from before, in that refuting the position currently being presented is insufficient to establish that every worldview is inconsistent. This problem remains unless there is some way of establishing that the criticism being leveled also applies to every non-Christian worldview. We have been told that this can be done, and they way of doing it has something to do with autonomous reasoning. But what is it about autonomous reasoning that leads inevitably to internal inconsistency? Nothing in Bahnsen’s quote brings this to the fore. This is what needs to be demonstrated, otherwise an equivalent counterclaim can be made: that there is something inherently inconsistent about non-autonomous reasoning. If we do not have to say what it is about a certain type of reasoning that inevitably leads to inconsistency, then each side can level its own unsupported assertion. There must be more than a simple assertion that autonomous reasoning leads to an internal contradiction, otherwise this is no better than just saying that God somehow guarantees it. So it looked like progress was being made, but without some further argument we are left with unsupported assertion.

At this point, Butler seems to declare that the game has been won already, as the very next paragraph demonstrates:

‘Thus the Christian apologist may boldly assert that without an absolute personal being as the foundation of all things, there is no possibility of ethics. Without the ontological Trinity as the fount of all being, there is no possibility of unifying the particulars of human experience. Without the combined doctrines of the Trinity and man being God’s image bearer there is no possibility of predication and thus language. Without the doctrine of God’s sovereignty and providence there is not ground for inductive logic and science. Without a good and all-powerful God that creates both man and the natural realm there is no reason to believe that our senses are reliable. From these considerations it is clear why TAG is often described as an argument that proves the impossibility of the contrary.’

Each time the words ‘there is no possibility of…’ or ‘there is no ground for…’ occur, this claim has not been supported. At most, there is the Michael Jordan example, where no actually held worldview has managed to account for these issues, but as we saw this is insufficient to prove that there is no other worldview that could.

Thus, as the argument stands, TAG is unproven.

6. Conclusion

I have offered here an analysis of TAG that appears to show the following: TAG has not been established, and TAG probably cannot be established. Moreover, I offer what I think is the best way to counter the presuppositional apologetic; do not get tempted to respond by showing that your ‘worldview’ is internally consistent, or to show that the Christian worldview is not internally consistent. Merely ask for the proof that no worldview than the Christian one is internally consistent. If the proponent of TAG proceeds to try to undermine your particular worldview, then concede (for the sake of the argument) that your worldview is inconsistent, but still press for the general proof that no other worldview can be consistent. The initial idea with presuppositionalism was to move the burden of the argument from the Christian to the opponent of Christianity; however, for TAG to be established actually has a large burden of responsibility. All I am advocating is that we acknowledge this fact.

The best representatives of the presuppositional apologetic are trying to illicit a ‘Copernican’ shift in the way that the worldview is argued for. The worst representatives are not trying to do this. What they are up to is trying to confuse the non-Christian by demanding philosophical justification for abstract concepts like truth, logic, morality, etc, instead of addressing the actual arguments against Christianity. Any confusion on the part of the interlocutor is then pounced on as evidence that the argument has been won by the Christian. There is a conceit behind such tactics, in that the argument is not being offered honestly. Behind all this is the view that autonomous thought, which is basically thinking for yourself, is the problem; and it is going to be hard to win an argument with someone who thinks that independent thinking is morally wrong. Believing that independent thought is a sin seems to me to be a hallmark of brainwashing, and at the very least a rejection of the enlightenment values that are typical of philosophical argumentation generally. It seems to me that they don’t want to actually win an argument by being right, but instead they want to win an argument by exploiting a sense of philosophical vertigo and then using this vulnerability as an angle to hard-sell religious dogma as the answer.

[1] The essence of the ‘inductive’ fallacy is that an inductive argument is being offered as a deductive argument. However, one could simply construe the unbroken record of success of Michael Jordan as inductive support for the conclusion that he is the best basketball player. If the proponent of TAG wanted to take this move however it would be of no help. It would remove the charge that the argument was invalid, but only by construing it as an inductive argument. If this course was taken then one could never establish that the conclusion was a necessary precondition for human experience; no inductive argument could reach such a strong conclusion.

[2]There are two objections here: 1) autonomy with respect to reasoning is not unique to the Christian worldview (what prevents other monotheisms from claiming that they also subordinate their reasoning to their god?), and 2) there are Christian worldviews where the intellect is not subordinated to the word of God (there are autonomous Christian worldviews; in fact, almost all conceptions of Christianity apart from the Van Tilian presuppositionalist account do not explicitly subordinate the intellect to the word of God). So the equivalence of Christian worldview with non-autonomous reasoning fails in both directions.