Assessing Agnosticism

An essay on the agnostic’s position in the debate concerning God’s existence

Original photo by author

Agnosticism is a stance of uncertainty towards some domain of inquiry or some proposition. If someone S is agnostic about P, this means that S suspends judgment with respect to P. The term “agnostic” was coined by the British biologist and naturalist Thomas Huxley (a.k.a. “Darwin’s Bulldog”) in 1869, specifically to pick out a position (and the one which Huxley himself endorsed) in the debate concerning the existence of God.¹

With respect to the existence of God, an agnostic is someone who affirms neither theism nor atheism, but instead suspends judgment with respect to the question of God’s existence.² Agnosticism is, and has now for more than a century been, a very common position among philosophers, scientists, and other intellectuals. For that reason, it merits careful consideration and analysis.³

A reality too often ignored is that there are different varieties and flavors of agnosticism, such that two agnostics can be in quite different epistemic situations relative to one another. Also frequently overlooked is the fact that, although it strikes many as the natural (and, indeed, default) position with respect to the question of God’s existence, agnosticism is in fact difficult to defend as an epistemic stance.

Here, I will first identify the different possible varieties of agnosticism. Then, I will articulate and analyze what I take to be (accessible versions of) the strongest and most compelling arguments for the most interesting and substantive variety of agnosticism.

---

1. Conditional vs. Hardline Agnosticism

The first major distinction to be drawn between agnostics is that between (i) those conditional agnostics, who merely take their agnosticism to be the right position for them, given their own epistemic situation, and (ii) hardline agnostics, who instead take it to be the objectively (i.e., stance-independently) correct position that any informed, rational agent will settle upon, given adequate reflection.

It is the second position which is obviously the more substantive and philosophically interesting one.

1.1 Conditional Agnosticism

Conditional agnostics think that they currently lack decisive evidence either for or against theism. They make no definitive claim about whether such evidence exists. So, conditional agnostics can hold that, for all they know, there is, in fact, objectively overwhelming evidence for or against theism. They also allow that theists and atheists can be rational in their theism and atheism, respectively (provided that they’re right about the direction in which their personal evidence points).

Conditional agnosticism is inherently unstable—and given the significance of the question, “Does God exist?”, the conditional agnostic should actively seek more evidence, if only as a prudential matter. Although conditional agnosticism just is the epistemically correct position for someone whose evidence is evenly split between atheism and theism, there would arguably be something seriously irrational about simply remaining incuriously and indefinitely parked in that state.

1.2 Hardline Agnosticism

A hardline agnostic thinks that agnosticism just is, objectively, the correct position—at least for epistemic agents like ourselves—to take in the God debate. Such an agnostic will think that atheists and theists alike are in error and that, if they considered the evidence adequately, then they would recognize that assent to both theism and atheism ought to be withheld.

Clearly, this is the philosophically interesting and substantive variety of agnosticism. It’s the position which says we all ought, objectively, to be agnostic.

---

2. Agnostic Leanings

Agnostics can lean towards theism or away from it. Some agnostics may have no real leanings, and they can be sorted into two further subcategories.

2.1 Atheistic Agnosticism

The atheistic agnostic is the most familiar sort. This is the agnostic who, despite lacking a confident belief that atheism is true, nevertheless clearly leans in that direction—and likely towards some kind of broadly naturalistic ontology (i.e., an ontology consisting only of natural entities and properties).

An atheistic agnostic who leans strongly towards atheism might even count as a tentative atheist and experience some genuinely atheistic moments (given that attitudes are variable and fluid across time).

2.2 Theistic Agnosticism

The theistic agnostic, despite lacking a belief that theism is true, leans towards theism and likely has various pro-theistic attitudes. The mirror image of the atheistic agnostic, the theistic agnostic who leans strongly towards theism might qualify as a tentative theist and experience genuinely theistic moments.

Theistic agnosticism might even be compatible with active participation in some forms of theistic religion (namely, those—if there are any—which require something more like hope than outright belief).⁴

2.3 True Agnosticism

The true agnostic has no real leanings either towards atheism or theism. True agnosticism comes in two varieties.

2.3.1 Conflicted True Agnosticism

Conflicted true agnosticism is the condition of someone actively investigating the evidence for and against the existence of God but who finds that evidence approximately evenly balanced. Such a person seeks the truth, genuinely desiring to know whether God is at the foundation of things or not. In that way, conflicted true agnostics are, by their own lights, in a non-ideal state. They desire to have the cognitive closure attendant to affirming either theism or its negation, but their assessment of the evidence is such as to render that epistemically and psychologically impossible.

Conflicted true agnostics can be of either the conditional or hardline sort. That is, conflicted true agnostics may think that although they currently lack clear evidence one way or the other, this will change at some future time, as a result of further investigation. But such a conflicted true agnostic may also eventually conclude that clear evidence for or against God’s existence simply does not exist, thereby becoming a hardline agnostic.

2.3.2 Apathetic Agnosticism

The apathetic agnostic affirms neither atheism nor theism, leans in neither direction to any significant degree, and also is simply uninterested in the question of whether God exists.

I think this position may well be irrational, given the significance and practical importance of the matter of God’s existence, but I perceive that apathetic agnosticism has become extremely common in the modern world. Plausibly, the myriad material distractions and rampant relativism in affluent, liberal democracies do much to foster such an attitude of indifference with respect to the ultimate questions of existence.

---

3. Hardline Agnostics Do Carry a Burden of Truth

A line one sometimes encounters is that agnostics need no argument for their agnosticism, since the position is merely the suspension of judgment with respect to the claim that God exists. According to this line of thought, both theists and atheists affirm a proposition (“God exists” and “God does not exist”, respectively), whereas the agnostic, by definition, refrains from doing so and thereby carries no “burden of proof”.

We can state this No Burden of Proof Argument as follows:

1. The agnostic affirms no proposition in relation to the matter of God’s existence.

2. If one affirms no proposition in relation to some matter, then one carries no burden of proof in relation to that matter.

So,

3. The agnostic bears no burden of proof in relation to the matter of God’s existence.

So,

4. Agnostics need no argument for their agnosticism.

Whether 1 is true depends upon whether the agnostic is of the conditional or the hardline sort.

Conditional agnostics do not affirm that agnosticism is the objectively correct position—only that it is the correct position for their own epistemic situation.

Hardline agnostics do, however, affirm that one ought, objectively, to affirm agnosticism as opposed to atheism or theism. As such, they do carry a burden of proof: They affirm a proposition, and anyone who affirms a proposition needs to have reason for doing so.

So, in short, the No Burden of Proof Argument lends support only to the variety of agnosticism which is the epistemically correct position for one who is correct that their personal evidence for and against theism is evenly balanced. It lends no support to the person who thinks that theists and atheists are in a condition of epistemic error.

---

4. Arguments For Hardline Agnosticism

Arguments for agnosticism — that is, for the view that agnosticism is the objectively correct position to take in the God debate—are, by definition, arguments for hardline agnosticism. Here I will consider five such arguments.

4.1 The Argument From Disagreement

Perhaps the most obvious argument for agnosticism proceeds from the observation that rational, intelligent, informed people — experts — disagree about the existence of God. When experts about some topic x disagree amongst themselves with respect to x, it’s plausible to suppose that non-experts — and perhaps the experts themselves — should simply suspend judgment with respect to x.

We can formulate this Argument From Disagreement as follows:

1. Among relevant experts, there is widespread and pervasive disagreement concerning the existence of God.

2. Where there is widespread and pervasive disagreement among relevant experts, we ought to suspend judgment.

So,

3 We ought to suspend judgment with respect to the proposition, “God exists”.

So,

4. We ought to be agnostic about God’s existence.

Both of these argument’s major premises are contentious.

With respect to Premise 1: It’s not clear who are the relevant experts. The ordinary person, having no specialized knowledge of the relevant arguments, nor the training necessary for evaluating them, does not plausibly count as such an expert. (If this sounds elitist and wrong, and you want to count the ordinary person as a relevant expert, that won’t get you to agnosticism, but to theism, since most ordinary people who’ve been aware of theism have affirmed it [or at least have been disposed to express that they affirm it⁵]). Those who know the most about the relevant arguments, and have the training necessary to evaluate them, are professional philosophers — and yet most contemporary professional philosophers (the majority of whom are atheists and agnostics⁶) have made no special study of the argument’s for God’s existence and so are not plausibly experts with respect to them.⁷ Philosophers of religion (who tend to be theists⁸) on average will know most about those arguments, though even in the subfield of philosophy of religion, the practical necessity of specialization in results in relatively few people with specialized knowledge of all of the relevant arguments.

With respect to Premise 2: It’s not clear that suspension of judgment is the correct response to peer disagreement. Ironically, experts on the epistemics of peer disagreement disagree amongst themselves about the best response to peer disagreement!⁹

Given that with respect to the question of God’s existence it’s not even clear that there is a class of people amongst whom agreement, or lack thereof, would be epistemically relevant, it’s not clear that any facts concerning disagreement about the existence of God should lead anyone to become an agnostic.

4.2 The Argument For God-Specific Metaphysical Skepticism

Setting aside the empirics of agreement and disagreement, one might instead argue for agnosticism on the grounds that it’s simply in the nature of the God question that we’re unable to settle it; that it’s beyond our cognitive abilities in some way to know whether God exists.

We could formulate the Argument For God-Specific Metaphysical Skepticism as follows:

1. Whether God exists is a uniquely difficult metaphysical question; while we can answer other important metaphysical questions, we cannot hope to determine whether God exists.

2. When confronted with an intrinsically unanswerable question, the rational course of action is to suspend judgment.

So,

3. We should suspend judgment about God’s existence.

So,

4. We ought to be agnostic about God’s existence.

Premise 1 here is particularly vulnerable because it, in effect, begs the question (i.e., it presupposes the conclusion of the argument). What we would really need is a strong argument to the conclusion that the God question is, among all metaphysical questions, uniquely unanswerable. But I personally cannot see how such an argument could be constructed. All metaphysics is, by its nature, somewhat speculative; it’s not as if we can impersonally verify where in metaphysics we’re reliable and where we aren’t — and any attempt to single out the God question as radically unique will seem hopelessly question-begging.

Premise 2 is also dubious. Perhaps even if the God question is intrinsically unanswerable we ought, nevertheless, to wager on his existence, postulate his existence, or take a leap of faith, as such a diverse array of thinkers as Blaise Pascal,¹⁰ Immanuel Kant,¹¹ William James,¹² and Søren Kierkegaard¹³ have held.

4.3 The Argument From General Metaphysical Skepticism

Rather than hyper-focusing on the God question, one might argue for agnosticism from a more general metaphysical skepticism. Influenced in part by the work of David Hume¹⁴ and Immanuel Kant,¹⁵ many modern philosophers came to endorse just such a form of skepticism on the general grounds that we cannot trust that our perceptions — or, therefore, that the concepts which are formed from them — correspond to (or reflect the actual nature of) objects that exist in objective reality.

Such writers could state an Argument From General Metaphysical Skepticism as follows:

1. We can’t know whether our concepts correspond to items, which exist as conceived, in objective reality.

2. If we can’t know whether our concepts correspond to items which exist, as conceived, in objective reality, then we should adopt a general metaphysical skepticism, withholding judgment about what actually exists.

3. Such general metaphysical skepticism extends to the God concept.

So,

4. We should withhold judgment about God’s existence.

So,

5. We should be agnostic about God’s existence.

Critics of general metaphysical skepticism will reject 1, on the grounds that it entails various absurdities which no one really believes outside the seminar room. For example, it entails that you can’t really know whether other people and the objects of ordinary experience, like tables and chairs, actually exist. I assume that ordinary agnostics do not take their agnosticism about God to entail agnosticism about the existence of tables, chairs, and other human beings.

4.4 The Argument From Incredibleness

It may seem that any proposition relating to the ultimate origin or foundation of reality is bound to be so incredible that it should simply not be believed, on the more general grounds that incredible propositions shouldn’t be believed.

According to the Argument From Incredibleness,

1. The proposition “God exists” is incredible — though the proposition “God does not exist” is equally incredible.

2. We shouldn’t affirm incredible propositions (at most, we should entertain them).

So,

3. We should affirm neither (i) that God exists or (ii) that God does not exist.

So,

4. We ought to be agnostic about God’s existence.

I do think this argument captures a real and recognizable kind of sensibility—and one to which I myself have been prone. However, it’s clearly not a good argument.

For one thing, Premise 1 is ambiguous between two meanings—one on which theism and atheism both seem like they can’t be true and another on which they both seem extraordinary. On the first meaning, Premise 1 is of course false: Necessarily, either atheism or theism is true. On the second meaning, the incredibleness of atheism and theism is not of a kind which suggests falsehood or, therefore, any reason at all to withhold belief.

For another thing, the judgment that atheism and theism are both incredible is itself contentious: Many find at least one of those views quite obviously correct and more of a mundane than incredible truth.

More importantly, though, Premise 2 is just plainly false; if accepted, it would rule out believing much of what our best contemporary science tells us about the world, since so much of what it tells us is incredible (e.g., read about cutting-edge theoretical physics or quantum mechanics).

4.5 The Underdetermination Argument

The most serious and interesting argument for agnosticism, I think, is one which proceeds from the premise according to which both atheism and theism are, to approximately the same degree, underdetermined by the extant evidence. According to this claim, if one surveys all of the evidence — all of the best versions of the best arguments for and against the existence of God — one will find that the evidence is roughly evenly split (evenly enough, at least, so as to rule out full-fledged belief in either view).

We can formulate the Underdetermination Argument as follows:

1. Objectively, the extant evidence underdetermines the matter of God’s existence: That is, the evidence is split evenly enough that it does not clearly favor either theism or its negation.

2. When the evidence underdetermines some matter M, one should suspend judgment with respect to M.

So,

3. We should suspend judgment with respect to the matter of God’s existence.

So,

4. We should be agnostic about God’s existence.

The most important thing to note about Premise 1 is that it’s incredibly contentious — and denied by many (indeed, perhaps most) of those who are plausibly regarded as experts in the God debate, and certainly by most people generally. It is no less contentious — and no easier to defend —than atheism or theism but is, rather, an extremely bold and ambitious claim — and one which few, if any, are plausibly in a position to affirm. Indeed, it strikes me as nothing short of an epistemological nightmare to try determine whether one is in position to affirm 1 (because, after all, there could always be another argument out there which, upon encountering it, might, by one’s lights, tip the scale one way or the other).

Premise 2 is plausible, although as mentioned in 4.2, thinkers such as Pascal, Kant, James, and Kierkegaard would deny it.

I think it would be quite remarkable, given that theism or its negation must be true, if the evidence in fact turned out to be evenly split between the two. Arguably, this is unexpected on both views, so that agnostics have reason to at least suspect, or worry, that they are missing something.

---

Conclusion

In conclusion, my contention here has been threefold: (i) that it’s in fact a cluster term which picks out a number of importantly different, if closely related, positions in the God debate and (ii) that hardline agnosticism — the view according to which agnosticism is the objectively correct position to hold in the God debate — must be argued for, and (iii) that there is no obviously successful argument for that position. While I have not argued that hardline agnosticism is wrong or irrational, I have offered reason to think that it is, in the end, a very difficult position to to defend — no less difficult than theism or atheism.

A fruitful question for future investigation is this: If it turns out that hardline agnosticism is the correct position, is that fact more expected on theism or on atheism?

1. See T.H. Huxley, 1889, “Agnosticism”, in T. H. Huxley, Collected Essays (Vol. 5), London: Macmillan.

2. See my post, “Thinking About Theism”, for definitions of key terms in the debate concerning the existence of God.

3. Note that there is a distinct possible kind of agnosticism about God’s nature, rather than God’s existence. That sort of agnosticism is not the variety with which I will concern myself here, though I may discuss it in a future essay.

4. This is another idea to which I will likely return in a future essay.

5. I think there is often some distance between what people profess to believe and what they actually believe, given that our behavior sometimes makes little sense given what we say that we believe.

6. See https://survey2020.philpeople.org/survey/results/4842

7. Anecdotally, it’s extremely common for seasoned, well-published philosophers to have, in effect, a Philosophy 101-level understanding of arguments for the existence of God. My hunch is that this is because, in contemporary academia, selection effects result in mostly secular, non-religious people going into philosophy graduate programs. Such people tend to have little native interest in the God question, and they can easily make it through a Ph.D. without ever having to engage it in a meaningful way.

8. See https://dailynous.com/2015/01/30/why-are-so-many-philosophers-of-religion-theists/

9. See https://plato.stanford.edu/entries/disagreement/

10. See Blaise Pascal, 1670/1966, Pensées (A. J. Krailsheimer, trans.). Penguin Classics.

11. See Immanuel Kant, 1788/2004, Critique of Practical Reason, (T. K. Abbott, trans.). Dover Publications.

12. See William James, 1896/1979, “The will to believe” in W. James, The Will to Believe and Other Essays in Popular Philosophy (pp. 13–33), Harvard University Press.

13. See Søren Kierkegaard, 1843/1983, Fear and Trembling (H. V. Hong & E. H. Hong, eds. & trans.), Princeton University Press.

14. See David Hume, 1779/2007, Dialogues Concerning Natural Religion (D. Coleman, ed.), Cambridge University Press.

15. See Immanuel Kant, 1781/1998, Critique of Pure Reason (P. Guyer & A. W. Wood, eds. & trans.), Cambridge University Press.