The Search for the Criterion of Epistemic Privacy

Evan Jack
7 min readMay 31, 2022



In order to have anti-skeptical epistemic privacy, one needs more than non-propositionality. For if one has inferentiality, then the inference can be questioned, putting that which one claimed to have private justification for in limbo, in that, its actual epistemic status is now not only validly in question, but also in that inferential links could be denied, and if they were dismissed, then that which one has claimed to be privately justified is no longer justified as a valid defeater has been put forward against it. One could argue that we could claim private justification for the inference itself. If we did such a thing, then the premises from which the conclusion one is claiming to be privately justified was inferred could be questioned, thus restarting the whole predicament lined out above. But, could one not repeat the same argument? Could one not claim private justification for these premises? They certainly could, but if the premises are not non-propositional, then they have committed a category error in claiming private justification unless they fall back onto a non-propositional “premise” (e.g., intuition) to justify the propositional premise in question privately. But, if they do such a thing, the issue repeats again, for all the questions regarding inferentiality for this non-propositional “premise” could be asked. In other words, if the thinking subject that claims private justification for a propositional premise, for example, falls back onto a non-propositional intuition, then we can still ask about the inferential status of such an intuition. If the intuition is inferential, then the whole issue repeats again, and if it is non-inferential, the regress ends. This means that all things with private justification must be non-propositional and non-inferential, meaning it must also be foundational. The reason it must be foundational is precisely because of the fact that any claim for private justification of infinitism or coherentism is problematic. Now, this is not a regress argument for foundationalism precisely because regress arguments act as disjunctive syllogisms, therefore meaning that regress arguments assume that one of the options within the disjunctive syllogism must actually be conducive to being the case, which is to say, it supposes that one of the options must be justified. Therefore, we are not doing a negative proof that demonstrates it must only be a specific option in virtue of the fact that it can be no other. Instead, we already have obtained justification by way of a claim to epistemic privacy. Therefore, there is no supposition that one of them is justified, as we already have justification. Now, the reason coherentism is not tenable is, again, because it either regresses, in which, according to its own standards, it has no justification, or, because it falls into circularity. But what is the issue with circularity? First, it is both practically and effectively indifferentiable from illogicality, and we are heading toward a solution to the logocentric predicament, therefore meaning what we value is logic. This means that we have a moral ban on circularity that functions as an epistemic ban by way of epistemic norms which are still nonetheless norms meaning normativity enters into the mix. Now, this is currently a supposition we hold, the supposition that we ought to head toward logic. This is all encapsulated by the problem of alternatives which is a problem to deal with another time. In fact, the problem of alternatives is the problem that will be dealt with immediately after this battle for epistemic privacy has been and this search for the criterion of epistemic privacy has been finished. Now, what if one heads toward infinitism? What if they are not resting their epistemic status on a vicious circle, but on a bottomless hole which they fall down forever. I want to quickly address an understanding of infinitism I see a lot and which I just indirectly referenced in the sentence before this one. The infinitist, or, at least, the Jackian infinitist (if such a thing even exists), is not claiming that this continual falling, this infinite regress, is not problematic for justification on the first-order, but rather that this “string” of infinite non-repeating reasons itself serves as the second-order basis for which first-order claims can rest. In this sense, they are meta-foundationalists. In another sense, there is still a problem here. If both the coherentist and infinitist claim private justification for their respective epistemologies, it will be predicated on something that is non-inferential, therefore meaning it is foundational, and, by definition, not regressive — thus, infinitism is out of the count. By specifying non-inferentiality as a criterion for epistemic privacy, are we not begging the question against infinitism? Not necessarily. I say not necessarily because we need to understand this: if something has not been justified, then it has not been justified. This does not mean that it is not justified. Furthermore, we are not saying that just because something has not been justified publicly, it has not been justified, for this is not a valid inference because of the genuine existence of epistemic privacy. Rather, if something has been claimed to have obtained justification by way of epistemic privacy, then this can still be contested, by presenting a defeater, and this defeater can be public and/or private. Because it can be public, epistemic publicity has been brought back into the mix. No matter how much the thinking subject claims private justification, if the premises of their proposition which they are claiming private justification for is predicated upon are denied or in question, the proposition itself is denied or in question. Being validly in question, it is not necessarily unjustified, but it is not necessarily justified either, not because it hasn’t been shown to be justified, but instead because its justification has not been confirmed to the thinking subject who has claimed private justification themself. Certainty is a requirement for knowledge and fallibility is not a defeater for such a thing, but if the “private” warrant is inferential then an issue certainly arises, because certainty is not at all obtained. Furthermore, if a reason is inferential, then it is, by definition, public, because inferences are demonstrations. For example, if I infer that A=C, from the premises, A=B and B=C, then all I am “supposing” is inferential transitivity. If I rationally intuit the law of transitivity to be privately justified, then my inference is justified. But if I have no justification for such a law, then do I have justification for my inference? This question, already answered, and put to the side, leads us to ask the essential question of all things pertaining to the relation between inferentiality and epistemic privacy: “If I can start a regress, does that render that which is at the start of the regress (the first thing questioned) unjustified?” The answer: yes. This is because one is not entitled to their inferred conclusion without inferential entitlement. The inferential entitlement does not have to be publicly demonstrated, but if it is not based on epistemic publicity, then it is not based on an inference. Here is what I mean: If I claim epistemic privacy EP for inference i by way of intuition n, then, without a public defeater, (n EP(i)) = T.[1] The inference itself, if it is without an intuition, is therefore without inference, in a certain sense. If something is inferential, then by definition, its justification rests upon the status of the inference in regards to its capacity for conferring the property of being truth conducive. Thus, either, (a) the inference is unjustified, therefore genuinely leaving the conclusion unjustified and validly in question or (b) the inference is justified by way of itself in its full assertive content, in which, the inference is itself non-inferential, in that it is not inferred from anything. One could argue that the justification of the inference is predicated on the two premises, and I would agree. But in the case that someone is violating the principle of epistemic honesty to the highest degree and claiming the contrary, that the inference is justified in itself, we must let them have this and cannot deny this, but if this is the case, the inference itself is not inferred, the regress is stopped and the criterion of non-inferentiality is established! Let us understand that because of the fact that the justification of the inference is infinitely regressing, it is not established, therefore meaning all those inferential warrants and entitlements are left, in the dust, without justification. Again, the inference can be privately justified, but if the inference is inferential, then it is without justification until the inference is justified, by definition. Again, this new inference can too be privately justified, but the inference will just continue. Now, because inferentiality has been explained away over the course of these five pages, let us understand that circularity, then is out of the question too, thus throwing away circularity-based models of coherentism, but not because of our moral ban, which too prohibits all such models. I claim this because of the fact that key to circularity is inference. The claim is always pp or pp. But, if non-inferentiality is a criterion, then there can be no “⊢,” nor can there be “→.” Thus, all we have is p, therefore meaning p is basic/foundational. Therefore, metafoundationalism/foundationalism is the only epistemology compatible with a claim to epistemic privacy, which is the only way to escape skepticism and garner justification and thus a genuine and, therefore, true claim to knowledge. Now that we have established all that we have set out to establish thus far, let us continue.

The next criterion is epistemic atomism, i.e., the thing in question must contain all the content for its justification in itself, for if it does not, then that thing must be questioned. For example, with divine intuition, its truth conduciveness is not held by the intuition itself, but rather by way of God. Obviously, one cannot divinely intuit God’s capacity to confer an epistemic status on another thing, not because that would be circular, but precisely because we can question God again, as this new divine intuition arises. Therefore, on threat of regress, epistemic atomism is a criterion for anti-skeptical epistemic privacy (the only epistemic privacy that matters).

Is there any further criterion that can be elaborated?


[1]: To clarify T means true.



Evan Jack

How sweet terror is, not a single line, or a ray of morning sunlight fails to contain the sweetness of anguish. - Georges Bataille

Recommended from Medium


See more recommendations