We seem to be going through a kind of online Christian revival — as one would expect, from historical patterns of American religiosity, which seems to follow a roughly 80-year cycle. Our current revival has a number of sources, and cannot be attributed exclusively to any singular force. But a large amplifier of the expanding Christian bubble has been the Christian advocates in the world of ‘internet blood sports’ — namely, Jay Dyer and Andrew Wilson.
Dyer and Wilson both specialize in meta-ethics, and deploy a form of a transcendental argument that does not (usually) directly attack the arguments in question, but shifts the focus to the presuppositions which must exist for arguments to take place — what is necessary to make an argument in the first place?
Let’s take a look at this argument.
The transcendental argument for god (T.A.G., for short) goes more or less as follows:
- Any argument — for or against a position — presupposes the knowability of things
- Knowledge is justified true belief
- “Justification” is logical validity
- THEREFORE all arguments and assertions of knowledge presuppose the transcendental existence of logic, as well as the possibility of knowledge
- The preconditions of knowledge are themselves not justified (and therefore not knowable) without some explanatory foundation of their own
- That foundation is GOD
The transcendentalist may also say that without justification, all ad hoc assertions are equivalent: “if you can ‘just say’ something seems unlikely, I can ‘just say’ the opposite.” Further, he might argue that to even participate in a debate is to presuppose the necessity of logical justification (what else could a debate possibly be for?).
On its surface, the transcendental argument is not only powerful, but humiliating — which perhaps might explain a lot of its appeal to its proponents, for whom its adversaries appear to be wrong and also deeply confused about where they are, metaphysically. Every argument appears to be a self-contradiction, because it is an argument (and therefore is an attempt at both logic and demanding logical justification) but it appears unable to justify itself.
The singular problem with the transcendental argument is that it fundamentally misunderstands language (in a couple of different ways). To look at the rough, syllogistic format of the argument presented above, consider the second premise: that “knowledge” is justified, true belief. Since we are making grand proclamations (the transcendentalist is claiming to deduce God from logic itself), let us be precise in our terminology, and examine it closely. Why is knowledge “justified true belief”? Is that how words work? We just hear a definition from Aristotle, and accept it on its face? For that matter, ti estin aletheia — “what is truth?”
Andrew Wilson often cites “words have meaning” as an example of a presupposition underneath arguments, since to state the contrary — that words don’t have meaning — would either be self-contradictory or nonsensical; it is a statement that hinges upon the words comprising the sentence itself having meaning. One might just as easily say “this sentence is not a sentence.”
But suppose we alter the phrase slightly, from the tautology “words have meaning” to “some words have meanings.” Does this change how the presupposition works, for the transcendentalist?
To say a word has a meaning implies a one-to-one mapping of a single word to a single “true” meaning, as in, “the meaning of ‘cat’ is this small furry pet.” But this is not how words actually work. Words acquire their meaning descriptively, by use, and often expand metaphorically (indeed, it can be argued that all words are metaphors), such that words can acquire several meanings — even dozens. The word “set,” for example, has 25 listed definitions as a verb, 24 as a noun, and 7 as an adjective (according to Merriam-Webster). Which of these is the true meaning? After all, ‘words have meaning.’ Which meaning?
Can the meaning of a word be wrong?
This is not some flippant misunderstanding of the transcendentalists’ point, because the purpose of the observation that ‘words have meaning’ is to argue that we can understand each other by means of language through the use of words, and that understanding presupposes meaning in the language. But multiple meanings shows us that meaning in language is insufficient to provide understanding. Moreover, the descriptive, metaphorical nature of language shows us that knowledge precedes language. Language can convey knowledge, but knowledge does not itself come from language.
This will become very important later.
So how does a common understanding of a word come about, of the sort that does make mutual comprehension possible? Put another way: how do words acquire meaning?
It is trivial to point out that the meaning of words are arbitrary, but arbitrariness does not get us to mutual comprehension via language. There must be some method by which we agree to meanings, which we associate with arbitrary sounds.
And the meanings of words can get us into a completely different set of problems.
In my essay Philosophizing with the Body (2025), I argued that the word “murder” — as used in arguments for objective morality — is guilty of a kind of intrinsic circularity:
If a philosopher appeals to the existence of objective moral values by pointing to “murder” as something that is universally and always wrong, we might ask if “murder” is meant in the popular sense or in the strict sense — for a killing on the battlefield might be considered “murder” if the adversary was surrendering, or was an unarmed medical officer, but might not be called “murder” if the killed is an enemy combatant. Similar questions surround circumstances of self-defense. A truly objective moral law, such as a prohibition on murder, must have some objective standard for what constitutes murder. And yet there is no such objective standard. There are certainly commonalities in what human societies tend to agree should and should not be permitted, when it comes to homicide. But descriptively and indeed, definitionally, “murder” is simply killing we have agreed is wrong. Thus, to employ the intuitive wrongness of murder as evidence of objective moral truths presumes what it attempts to demonstrate.
In the same way, to say “words have meaning” is a tautology, since a “word” is, by definition, a sound with meaning.
The transcendental insertion of a tautological presupposition before an argument, as a precondition of the argument, serves as a sneaky back-door to inject preferable understandings of very important terms which are preferable to the transcendentalist — words such as “knowledge,” “justified,” and “debate.”
Is “knowledge” really “justified, true belief?”
Does “justified” literally just mean “logically valid?”
Is two philosophers exchanging syllogisms the only kind of debate?
All of these are perfectly valid understandings of these words, but they are not the only valid meanings of these words, nor were they even the first meanings. “Knowledge” is an Old English conjugation of knowen, but has been used as a translation for many different Greek words (including those of Plato and Aristotle), such as eidenai, gignōskein, noein, manthanein, and sunienai. These were very often sense perception terms: “I saw” becomes “I know (because I saw).” However, we can know things by other ways. We can “know” something because we heard it from someone else. We can “know” something by divine revelation. We can even know something by instinct. And, of course, we can know things by reason. All of these are perfectly valid ways of using the word “knowledge.” But if we grant that knowledge is only reason, or only “justified true belief,” we actually negate other forms of knowledge.
This becomes a problem when sense experience runs into conflict with our logical reasoning. Consider, for example, the following syllogism:
- The tree is green
- What is green cannot be not-green
- Therefore, the tree is not not-green
The conclusion (3) logically and necessarily follows from the first two premises, and wouldn’t you know it, both premises are true… and yet somehow, the conclusion is false; there are parts of the tree that are, indeed, not-green, but brown or red or grey.
As with Zeno’s famous paradoxes, there are ways of eventually stumbling upon the problem with the syllogism (in this case, it is an equivocal use of the word “the” in (1), treating a part of the tree as if it were the whole). But one could much more simply follow the example of Diogenes and walk away, or look out and simply observe that the tree is not entirely green, thus falsifying the logically sound and valid argument.
(If one says ‘premise (1) is actually not sound, because the tree is not only green,’ then they might be correct, but this introduces the problem of precision into the matter, which plagues all grand, logical assertions about the world.)
Is the observation that the tree is not completely green a “justified, true belief?” Well, it’s ‘justified’ in the sense that it seems reasonable… but it’s not exactly a logical syllogism. And how do we know that what we see is ‘true?’ It seems to be as true as any logical syllogism… but if we’re going by ‘seems,’ then we must suppose that someone else is just as justified to say that it seems completely green (or completely purple) to them… and who are we to judge?
At least we can call it a ‘belief!’
“Justified” can be understood to mean ‘logically valid,’ but it can also mean ‘reasonable,’ not the sense of being ‘guided by reason’ with some couched and preferential, particular definition of ‘reason,’ but in the conventional sense of being sensible, or even common sense. To say “the tree is green, but not entirely green” is reasonable, and therefore justified without any appeal to logic, causality, or the satiation of explaining tautological presuppositions that we inject in front of the claim.
So what are we to do with these presuppositions? These supposed preconditions of knowledge?
It turns out, on the other side of a proper understanding of language, that most of these preconditions are completely unnecessary, or are themselves equivocal. To say that “speaking presupposes that words have meaning” is trivially true, but once this is bootstrapped into some specific idea of making an argument presupposes the possibility of knowledge, we suddenly have a sneaky side-door opening to a very specific notion of knowledge. Arguing, in fact, does not necessarily pre-suppose knowledge; one could defend a position one does not hold (as in, attempting to justify what one does not believe to be true). But more commonly, one can argue in defense of something that one knows, but not through logical justification (i.e., the tree is green).
It is useful here to explore where transcendental arguments came from.
Immanuel Kant famously said that David Hume woke him from his “dogmatic slumber” with his skepticism. Hume argued that our notions of causality and induction are derived from experience, rather than reason. This threatened the entire project of metaphysics, and the possibility of objective knowledge. To a certain kind of person, knowledge is not knowledge unless it is objective. Kant, being of such a kind, set his mind to work, and wrote a defense of pure reason disguised under the most immaculately backwards title, A Critique of Pure Reason.
I wrote a little bit about this in Descending Odin’s Throne (2024):
But Kant was not a skeptic. He believed it was possible to acquire what he called a priori knowledge (“prior to experience”). Such “pure” knowledge does not depend upon the senses, instead deriving true knowledge and the possibility of certainty by means of analysis and synthesis of identities and necessities. In philosophy, this approach is sometimes called transcendentalism, and has given rise to “transcendental arguments” which attempt to show that something is a necessary condition of something else. Kant’s famous categorical imperative – “act only according to that maxim whereby you can at the same time will that it should become a universal law” – is in fact an a priori synthesis of what objective moral law must be, based upon the nature of morality itself. The assumption, of course, is that morality isn’t really morality unless it is universal.
But in all possibilities of a priori knowledge, language is required to reach some conclusion. Language is in fact the mediating mechanism of thought, and the words that comprise language are not a priori knowledge. Even Descartes’ famous ontological argument for a priori knowledge of the triangle requires the language of mathematics in order to conceive of the relationship at all, since thought on abstract subjects like this requires language in order to make inferences or deductions. Identities required for analysis are inherently linguistic. And language – logical, mathematical or conventional – comes from experience.
What is ironic is that Hume’s skepticism still leaves room for empirical knowledge (and other kinds of experiential knowledge); it is only certainty and objectivity that his criticisms undermine. The transcendental argument which holds that arguing itself presupposes the possibility of knowledge engages in a kind of equivocation, where “knowledge” is used in the broad sense when we speak of presupposing (this would include knowledge gained by experience), but in a much narrower sense when they speak of “justified true belief” later on.
The transcendental argument creates what we might call a “transcendental triangle,” a sort of mesh-network of particular definitions which all support each other, comprising a sort of three-part circle of self-justification. These terms are:
- Knowledge
- Justification
- Debate
We have gone on at length about “knowledge,” and explained the distinction between logical justification and justification in the more colloquial sense. But “debate” requires a little bit of unpacking.
On the surface, it does seem like debate is — or ought to be — the adversarial exchange of logical arguments. And indeed, in some cases, this is the case. But in other cases, this is not the case. Indeed, in the vast majority of debates, what we see is not a contest of logical arguments, but a contrast of observations, with each debater making the case to the audience that their particular set of observations is the more relevant and significant. These arguments aspire to be “reasonable” — in the colloquial sense — but do not even attempt to be logical.
Does the absence of formal logic make these debates incomprehensible, or even ‘not really debates’?
Different styles of debate will be appropriate to different contexts; what matters here is not adjudicating which sort of debate is best, but that a debate is not ‘not a debate’ in the absence of formal logical syllogisms. Such an assertion would be a misunderstanding of language, in how we use the term ‘debate.’ One might attempt to argue that a philosopher (at least an analytical philosopher) ought only to engage in formal, logical debates… but one still could not assert that a debate isn’t a debate in the absence of the sacred syllogism.
The transcendental trap is that challenging the meaning of one of these words brings the debaters back to the other two. To challenge the assumed nature of knowledge seems to require “justification”; to challenge the narrow sense of “justification” demanded by the transcendentalist invites the charge of not engaging in debate at all; and to challenge the very notion of debate allows someone to ask how do you know what a debate is?
The circle would be obvious if it were only two items. If someone were to define a “boat” as a “sea-vessel,” and when asked what a “sea-vessel” is, to define it as a “boat”… we would understand that no actual information was being provided (even if the definitions were, technically, correct). But at three items, the bootstrapping logic manages to sneak past most people’s notice.
But what about logic itself? If language has shown us that it isn’t necessarily a precondition of knowledge, or of debate, isn’t it at least a fascinating example of a non-material thing that humans discovered, and did not invent? Does it not remain an example of an ordering principle which seems to point to some orderer?
The problem of language, in fact, reaches here too.
Allow me to reference Descending Odin’s Throne again, here in relation to mathematics, which is just logic for non-philosophy-nerds:
The number “five” does not exist in nature.
It is an abstraction of an idea, a quantity that is relative to another discrete quantity. We identify a single apple as “one,” and only by this assumption do five fruits from an apple tree constitute “five.” The statement “an apple is one” is a premise, but it is not “true” in any objective sense. It is an assertion of identity that is arbitrarily presumed for the sake of convenience. If apples were quantified by weight, rather than by discrete fruits, a completely separate system of quantification could be imagined. Arithmetic done in base-15 is just as valid as arithmetic done in base-10, or base-2. Non-Euclidean geometry is just as valid as Euclidean geometry. They only have different assumptions.
With this in mind, it is clear that mathematics is actually only objective so long as one’s assumptions are objective. The use of mathematics in argumentation today is largely the art of hiding the subjective biases of these assumptions, but this too is being pushed back on, even from within mathematics itself. Gödel’s famous incompleteness theorem demonstrated – mathematically – that a mathematical system cannot prove or justify its own initial assumptions, thus all mathematical systems are “incomplete.”
But we did not need mathematics to know this. We could understand this simply through language, and the recognition that mathematics is just a derivative of language. It is, in fact, language purged of all discrete subject matter; a pure grammar.
It turns out that logic (and mathematics) are actually not discoveries, but inventions. To be precise, they are descriptions. The law of identity (1 = 1) is no more a “discovery” than is the aforementioned “discovery” that murder is always wrong, or that words have meaning. All mathematical equations are tautologies, if they are done correctly; similarly, all logical equations are tautologies. As tautologies, they cannot tell us anything new, outside of their assumed axioms. They simply don’t appear this way to us, intuitively, because we have replaced the semantic content of the “sentences” with abstractions (numbers). The layers of abstractions make it easy for us to forget that mathematical statements are descriptions, and the unreasonable effectiveness of mathematics (or logic) is simply the unreasonable effectiveness of description.
“Is it not fascinating that everything we describe as brown is brown?”
The ability for mathematics to “predict” reminds us of the utility of description, and the practical usefulness of logic. As a tradesman, I use mathematics myself, fairly regularly, for measurement in cutting, bending, and calculating ampacity. But it is easy for people to get carried away with the utility of math. Every time someone scratches their head and says “it wasn’t supposed to happen like that…” is an instance of math or logic failing. Usually, this is not a failure of calculation, but the failure to account for a variable. For this reason, most of engineering is an art of approximation.
To put it all together, the transcendental argument — for god, or for anything else — is a language shell game, based upon selective and self-reinforcing definitions of terms, which allow the debater to lose his opponent in a maze of terminology and declare victory. But this brings him no closer to the truth, if “truth” is understood to be “what is.” What is, in the world, is not dependent upon how we describe it, and the coherence model of truth that all transcendentalist arguments depend upon relies upon descriptions for coherence to be achieved. The entire move of pointing to underlying preconditions of knowledge is, itself, a language game.
Is pointing this out a transcendental observation? To observe that a proper understanding of language is presupposed in debate? One could frame it that way. But we actually don’t have to. A discussion which makes use of language does not require us to take a detour down linguistics lane and pay homage to language itself, affirming certain principles of language in order for our sentences to make sense.
Indeed, the sense of our language precedes our understanding of why our sentences sense. And any asserted theory as to how or why language must work is not given a free pass on account of language working. It might be wrong… and language still works.
In the same way, the supposed preconditions of knowledge, debate, and the idea of justification itself asserted by transcendentalists might all be incorrect, and yet here we are, still discussing, and accomplishing tasks more or less effectively — demonstrating knowledge — in the real world. This does not prove their argument, but only makes the language circle they invoke an unfalsifiable wizard’s spell, which can be dismissed simply by insisting upon a clear and consistent concept of “justification.”
It is certainly possible for the transcendentalist to say that actually, logic refers not to structured language, but to principles of identity and stability in the universe itself, prior to language. So long as they don’t use these two senses of the word “logic” interchangeably, that’s all well and good… but our universe also changes. Our body changes. Parts of us stay the same, for a while, while other parts shift. Where does the principle of stability and identity begin? Actually, what is the principle of identity in the material world, where it isn’t just 1 = 1?
Claiming that knowledge presupposes certain logical principles of intelligibility begins to inject language back into the discussion — even where the transcendentalist is ostensibly talking about physical things. How do we know what these “principles” are? How can we derive such principles prior to experience?
To assert Y is a metaphysical precondition of X is to reach beyond what the epistemic hands of experience can touch… and the only tool to artificially extend beyond that reach is language. If the transcendentalist’s criticism of Hume is that his skepticism makes knowledge impossible, then knowledge is, indeed, impossible… and yet somehow, people seem to keep going about knowing things.
Does this disprove Hume? Or does it merely disprove a utopian notion of knowledge that was itself arrived at by language?
In short: debate does not presuppose anything other than disagreement. How, why, and over what that disagreement takes place are in dispute… and indeed, we actually don’t need an account of debate in order to debate, just as we don’t need an account of war in order to go to war. Such an account might be useful, and it might also be the opposite, causing us to look for categories, rather than seeing the reality, and only then searching for the right words to describe what it is we are seeing.
So let’s try to apply it.
Suppose you run into a Jay Dyer in the wild, who says something like this:
What about reasoning about reasoning itself? These are meta-logical questions, meta-ethical questions, meta-physical questions. When I start to ask questions in that domain, and I ask questions about reasoning itself, or words themselves — how do we know that words have meaning? how do we know that logic is logical? — you see, these are questions that you can’t actually be wrong about, because if you’re wrong in that domain, it really destroys the possibility of knowledge.
How do you respond?
Do you simply accept the tautology as tautological?
No!
What is the insertion of meta– doing in these cases, with logic and ethics and physics?
It is turning the subject linguistic, and it is the set-up for equivocation by using the same word to describe different things. Logical consistency pertaining to language is not the same as the “logic” we perceive in the stability of things in the world. The “reasoning” we allude to in language is actually not always the same thing as the “reasonableness” of a conclusion. The reflexive turning back of a concept on itself changes the thing, and it is only by language that we are tricked into seeing the ungroundedness of logical systems as somehow disproving the stability of physical objects (which is already a challengeable interpretation) without some ordering God.
You can simply point out that the sense of knowledge that is being destroyed (coherence) is not the sense of knowledge that most people think of as knowledge (correspondence), and the maneuver to try to point out that the latter is some sort of logical contradiction is just another equivocal word trick to bring you back into the coherence framework. Similarly, if a position is unjustified in a strict, logical manner, it doesn’t mean it is unjustified in all senses of “justified,” and it is not some doorway to say “well, I guess everything is on the table, and I’m just as justified to blindly assert the opposite of your position”.
At the end of the day, I don’t think the transcendental argument is actually persuading many people. It allows the apologist to abuse and defeat adversaries with clever word games, like Socrates destroying Thrasymachus in Book 1 of Republic, but leaving sincere and intelligent onlookers on the side watching unpersuaded, saying “do you want to seem to have persuaded us, or truly to persuade us?” It rallies and energizes people already convinced with the feeling that they’re on the ‘winning side,’ but this too is an illusion… and I don’t think it even demoralizes the opponents, who generally walk away, not feeling stumped by a serious problem, but as someone feeling tricked, in a manner they can’t quite untangle.
Their feelings are correct, and the sooner we — in the collective world of online debate — can untangle this word game, the sooner we can get back to more serious arguments.