Home » Library » Modern Library » Kile Jones Agnostic Theology

Kile Jones Agnostic Theology

An Agnostic Theology (2010)

Kile Jones

1. Introduction

For years now we have thought that agnosticism and theology were incompatible, for lack of better reasons, because theology was thought to entail revelation. Revelation, in turn, was considered by some theologians as an adequate source of knowledge, and claims to knowledge of God were not considered ones of possibility. It seems as if there is no place for one to say “I believe God may probably exist” or “God might exist” or “You can never know if God exists, but it is possible.” I am going to argue that it is entirely plausible, even intellectually honest, to have an agnostic theology, for it allows one to provide reasons for possible belief in God while simultaneously allowing reasons for possible nonbelief in God. Just because our tools may be limited, it does not mean that what they cannot grasp is or is not actually there.

2. Agnosticism and Theology: Definitions

One of the problems with agnosticism, as with most other epistemic theories, is that of definition. For the sake of clarity, I will differentiate between weak agnosticism (WA) as the position which states that we do not know but it might be possible to eventually know if God exists, and strong agnosticism (SA) as the position which states that it is impossible to know if God exists or not. Russell (1953) uses both of these ways of describing agnosticism, only in reverse order: An agnostic thinks it impossible (SA) to know the truth in matters such as God and the future life … if not impossible, at least impossible at the present time (WA).[1] What I mean by know also means a great deal to the debate. What I mean by know is not certain knowledge, but a belief that is minimally justified, enough that it is reasonable to hold. What exactly justification is, and what amount is needed, is not the focus of this chapter, though it is important; rather, I want to investigate how agnosticism (both weak and strong) can fit within a theological framework and what the implications of such a theory are. This does not require an in depth analysis of what constitutes belief, knowledge, and justification, only a surface one. This is because agnosticism can be used in the colloquial and historical manner as it is found in most works in the philosophy of religion. This way I can venture, with a common understanding and meaning, into the question of whether or not agnosticism is compatible with theology.

I should at this point clear up what I mean by theology. I am going to use theology to refer to a very broad spectrum of spiritual worldviews, not necessarily just classical theism, even though this is normally where the debate is to be found. My point is that all supernatural postulates, from an agnostic position, are epistemically inadequate for full belief, in that they always leave room for doubt, disproof, and skepticism. Even though I believe all supernatural postulates to be epistemically inadequate, I nonetheless have to use materials that center on the issue of agnostic theology and most of these are written with a classical Western theism in mind. With this in mind we may now turn to the objective at hand.

3. Religious Knowledge: The Problem

There has always been a problem with claims to religious knowledge, not necessarily due to what is claimed to be true, but by the claim that it is knowledge. When someone says that he has knowledge of something or another, whether an answer to a mathematical problem or an empirical problem, we always seem to ask how he knows it. The usual answers involve providing ‘justification’, ‘evidence’, or ‘warrant’, hoping that what he has provided is enough to ease the question’s force. If the justification is enough, and if it is true, we usually conclude that it is knowledge.[2] Now knowledge has many different forms: possible, probable, and certain. You could even break this is up into a 1 to 100 percentage scale of likelihood. On top of this, there are first-order and second-order beliefs. A first-order belief is one that is direct, intuitive, and usually psychologically necessary to hold. A second-order belief (or metabelief) refers to and calculates the percentage of warrant first-order beliefs have. So for instance, I have a first-order belief that there is a mind-independent world, but a second-order belief that tells me that the first-order belief (in a mind independent world) is possibly, or even probably, true. You may also break these beliefs down to 1 to 100 percentage scales in terms of strength. The strength of your belief that two plus two equals four is 100, while the strength of your belief that your Mother will be back from work at exactly five thirty p.m. may be around 60. Furthermore, what we think constitutes justification is going to determine what we think constitutes knowledge. These divisions must be kept in mind regarding claims to religious knowledge, for they are clear ways of understanding and analyzing the weight and force of theological propositions.

A typical theological proposition, specifically made regarding classical theism, is “God exists.” This proposition is clearly a positive ontological assertion which aims at conveying a supposed mind-independent truth, for whenever someone says it they do not mean “God exists only in my mind,” or “God might exist,” or even “God most likely exists.” The usual intention of the speaker is to assert this as an objective, or mind-independent, truth and reality. Yet there are problems with theological propositions like these. One is that of warrant and another is that of belief. The issue of warrant is how one knows this proposition to be true and justified, and what percentage of likelihood this supposed knowledge is thought to be. Typical responses attempting to provide warrant include the classical (and new formulations of) arguments from Cause, Morality, Ontology, Teleology, Transcendental, and from religious experience. The agnostic finds these arguments wanting, not only because there are logical problems with them, but because of the percentage of likelihood these arguments are thought to provide in favor of theism.[3]

Yet just because the arguments for God’s existence are inadequate (from an agnostic position) does not entail that they will always be. It could be that philosophers have yet to formulate or discover arguments with more logical weight, though this seems highly improbable. Terence Penelhum (1971) shares this opinion:

In so far as philosophical scepticism about the possibility of theistic proofs has a basis at all, it is usually the result of the failure of the proofs that philosophers have encountered. The failure of attempts to prove God’s existence is clearly inadequate to supply us with a proof that God’s existence cannot ever be proved. If it supports any conclusion at all, however, it is a sceptical one.[4]

This “skeptical conclusion” is due to the failure of the supposed theistic proofs to provide adequate justification for belief in God’s existence. But the fact that the justification is lacking is not equivalent to furnishing a proof for God’s nonexistence. On a similar level, the agnostic finds the arguments against God’s existence wanting, for if anything, they can only act as defeaters to theistic arguments, and if they are found more convincing, then it only follows that God’s existence is improbable.

Now it should be one’s epistemic duty to proportion the strength of one’s belief with the amount of justification it has. So if the amount of justification is 50 percent, then the strength of the belief should be 50 percent. One problem with this ideal is that most of us do not have the time or resources to weigh all of the arguments and their counterarguments. Similarly, it may be that most of us base our beliefs on emotive, rather than rational, reasons. Yet our aim should be to proportion our beliefs with the evidence presently available, for if our beliefs are disproportionate to the evidence, then we are not fulfilling our epistemic duties.[5] From an agnostic position, those who believe in God with the strength of say 70 percent are not fulfilling such duties because the justification does not amount to something as high as 70 percent. Whether or not one can believe in the existence of God with the strength of 51 percent or higher and remain agnostic is up for question, and can only be answered by establishing what constitutes knowledge. It could be that one has a direct belief (not a first-order belief) in God’s existence with the strength of 51 percent or higher and simultaneously has a metabelief that the justification is 50 percent or cannot be known. In this case one can hope for the best and throw themselves, in a Kierkegaardian fashion, into the unknown, or at least the possibly false. This is the fideism that Pascal’s wager asks for, a proverbial hedging of all of your bets.[6] For the most part agnostics find this position untenable due to its blind and irrational character, but this is one of the only views of God and knowledge that can be called agnostic.

4. Religious Agnosticism

In “The Presumption of Atheism” (1976), Antony Flew described “religious agnosticism” as follows:

It is also possible to speak of a religious Agnosticism. But if this expression is not to be contradictory it has to be taken to refer to an acceptance of the Agnostic Principle, with either a conviction that at least some minimum of affirmative religious doctrine can be established on adequate grounds, or else a devotion to the sort of religion or religiousness which makes no very substantial or disputatious doctrinal demands.[7]

This “Agnostic Principle,” taken from Thomas Huxley, states that “it is wrong for a man to say that he is certain of the objective truth of any proposition unless he can produce evidence which logically justifies that certainty.”[8] It can be said fairly that this principle is both the historical and epistemological cornerstone of agnosticism. I would agree with Huxley that if anyone makes objective truth claims it is incumbent upon them to provide evidence which logically justifies the claim made. Yet this principle is mainly attacking those who believe in the certainty of God’s existence, asking them to provide the necessary evidence, and for that it is fine, but it does not take into account probabilities of belief.

If we add that “some minimum of affirmative religious doctrine can be established on adequate grounds” to the Agnostic Principle, then we must believe that such doctrine meets at least a minimal criteria for justification. That is, a religious agnostic must have some good reasons for believing the “religious” side of the label. It could be that these good reasons are constituted by the logical weight of the arguments for God’s existence, and that the agnostic accepts some of the critical analyses of those arguments. This “balancing out” of the positive and negative arguments for God’s existence can still allow for belief in God, as long as the belief is one of possibility. If the positive arguments outweigh the negative ones, then you can reasonably believe that God probably exists, though the strength of this belief cannot exceed the probability of its justification.

If we are going to add “the religion or religiousness which makes no very substantial or disputational doctrinal demands” to the Agnostic Principle, then we must describe exactly what this “religion” amounts to in terms of propositions about God. If the propositions are “God might exist” or “It is possible that God exists,” then we have to be careful in proportioning the belief with the justification. As Flew notes, only this sort of religion can properly be called “agnostic,” for it does not have strong propositional truth claims. But belief in the high probability of God’s existence is stronger and thus requires firmer justification. This sort of “religiosity” is not “agnostic” because it goes beyond the proposition that one “cannot know” (SA), or even “cannot know now” (WA), whether God exists.

5. The Onus and the Razor

One may justifiably wonder how a hypothetical religion that does not make strong propositional truth claims—an agnostic theology—handles the burden of proof (onus probandi) and Ockham’s razor. When one presupposes theism—that is, believes it at face value—then the evidence (whether empirical or logical) is interpreted through that presupposition; and likewise for atheism or a more comprehensive worldview. The question does not concern analysis of evidence after the fact, but rather the presuppositions we make when we analyze and interpret evidence, or even decide what counts as evidence. So which presuppositions are more rational—or put differently, who has the greater burden of proof? We should be able to answer this question if we grant that there are comprehensible, mind-independent truths. We would simply need to weigh the justifications of the relevant propositions. An agnostic should agree with atheists that theism has a strong burden of proof because it postulates that which cannot be verified empirically or demonstrated tautologically. But this isn’t a particularly important point; it merely informs us that theism has a greater burden at the outset, not that theism cannot meet that burden. Ockham’s razor—the thesis that the answer requiring the fewest additional assumptions is the most rational one—might pull us toward viewing atheism as the more rational answer. But only if one presumes that God cannot be verified empirically, or demonstrated tautologically, or that revelation (whether historical or experiential) is not an adequate source of knowledge.[9]

Of course, to challenge the latter assumption the theist needs to provide some justification for believing that revelation occurs. The usual attempts to establish historical revelation (through miracles, written records, or testimonials) may rely on showing that events recorded within particular religious texts correspond to archeological or geological facts. A Christian might point to the Pontius Pilate stone, recorded Jewish practices, or extrabiblical texts to support the veracity of the New Testament. An agnostic ought to have reservations about these kinds of arguments, for even an accurate rendering of broad historical events in the New Testament would not entail that any reported supernatural events occurred. Similarly, even if it seems as if a reported supernatural event is best explained by presuming its historicity, most historical events have completely natural explanations, so belief in a supernatural cause must be regimented; the majority of experiences indicate that positing natural causes is the most accurate method to arrive at knowledge. Consequently, any belief in historical revelation should be tenuous, for we should always proportion the strength of our beliefs to the strength of their justification. And since historical revelation is not how we normally acquire knowledge, it should not be given the same weight as beliefs which rest upon more reliable methods. I should not believe that God revealed himself to Moses by performing miracles with the same strength for which I believe that there is a table in front of me; the method of verifying the former is simply less reliable than the method of verifying the latter.

This “soft agnostic” approach to revelation also applies to claims of experiential knowledge of God. When someone says that he “experienced God,” “felt God,” or “heard God,” a reflective person must wonder how he knew that God was the source of that experience. A typical response is to invoke an immediate (or intuitive) knowledge of God—that one simply knows that God caused the experience. The question is whether these sorts of claims can be verified in any way. What do we make of a person’s report that he knew that God spoke to him, and that he revealed himself in the form of a giant unicorn who hates humanity? If an intuitive experience of God can provide reliable knowledge, then who are we to say that the report is false? There must be some independent way of judging the veracity of claims derived from different mystical experiences. This can be done by analyzing the propositions asserted in terms of their internal consistency and correspondence to reality. For instance, by my lights a misanthropic unicorn God is not likely to exist because it is not an accurate idea of God, for the very concept of God entails perfect attributes, and misanthropy and the form of a unicorn fall short of perfect. There are obviously more consistent pictures of what God would be like if it did exist, but how one knows whether or not God exists is the main problem.

There are also numerous natural explanations for mystical experiences, even those that command absolute conviction in the experient. Perhaps they are a sociological artifact of how religious groups behave, a psychological result of fear and the need for certainty, a biological firing of synapses simulating a supernatural experience, or the result of a general existential desire for something more than this life and the regular ways of knowing. Whatever their motivation, beliefs based on mystical experiences should not be held strongly because they do not have strong means of verification. One may claim that an ostensible experience of God seemed or felt true without going so far as to claim that it was or must be true. I don’t point this out to dismiss the experiences of people, but to admonish you to reflect upon your beliefs and believe them with the strength that independent justification provides. One may believe in God understood as a postulate, a possible explanation, and not as something that can be justified or known. The strength of one’s belief in God should be proportionate to the strength of the justification for that belief. Consequently, many should ask themselves whether or not the strength of their belief is disproportionate to the strength of its justification.


[1] Bertrand Russell, “What Is An Agnostic” in Russell on Religion: Selections from the Writing of Bertrand Russell ed. Louis I. Greenspan and Stefan Andersson (New York, NY: Routledge, 1999): 41-49.

[2] Gettier problems have been raised against the identification of justified true belief (JTB) with knowledge. But I am not convinced that Gettier or other problems actually undermine the JTB identification. See Edmund L. Gettier, “Is Justified True Belief Knowledge?“, Analysis 23(6): 121-123 (June 1963), and Keith Lehrer and Thomas Paxson Jr., “Knowledge: Undefeated Justified True Belief,” Journal of Philosophy 66(8): 225-237 (April 24, 1969) for a response to Gettier’s seminal article.

[3] For a good presentation of arguments for and against the existence of God, see Richard M. Gale and Alexander R. Pruss, The Existence of God (Aldershot, UK: Dartmouth Publishing, 2003).

[4] Terence Penelhum, Problems of Religious Knowledge (London, UK: Macmillan Press, 1971), p. 43.

[5] I believe that we have epistemic duties, and that awareness of them ought to motivate us toward critical and reflective belief formation. See Matthias Steup, Knowledge, Truth, and Duty (Oxford, UK: Oxford University Press, 2001).

[6] One problem with Pascal’s wager is that it does not hedge all bets, for it unjustifiably excludes alternatives to Christian theism against nonbelief. The exclusive salvific truth claims of other religions do not allow one to spread their belief chips all over the board. See Graham Oppy, “On Rescher On Pascal’s Wager, International Journal for Philosophy Of Religion 30(3): 159-168 (1990).

[7] Antony Flew, The Presumption of Atheism (London, UK: Pemberton Publishing, 1976), p. 33.

[8] Thomas H. Huxley, “Agnosticism and Christianity” in Science and Christian Tradition (New York, NY: D. Appleton and Co., 1898): 309-365.

[9] A common criticism of A. J. Ayer’s verification principle is that it is self-refuting because it cannot itself be verified empirically or tautologically, and thus is not a sound criterion for justification. While this may be true, how else could beliefs be justified? According to traditional theism, various forms of revelation are an adequate source of knowledge, and so Ayer’s principle is false. But this is suspect because the manner of verifying God’s revelation is much weaker than the way that we verify most truth claims in ordinary life (by Ayer’s principle or Hume’s fork—that all knowledge is either of “relations of ideas” or of “matters of fact”).

Copyright ©2010 Kile Jones. The electronic version is copyright ©2010 by Internet Infidels, Inc. with the written permission of Kile Jones. All rights reserved.

all rights reserved