pleading the fourth: where is thy warrant?

Three detectives walk into a court. The first detective presents a piece of Knowledge collected from the defendant’s home. The Judge asks, “Did you have a warrant?” But he is a metaphysical naturalist, and he says, “No.”

“Then it is not admissible.”

The second detective presents a piece of Knowledge collected from the defendant’s home. The Judge asks, “Did you have a warrant?” But he is a pantheist, and he says, “No.”

“Then it is not admissible.”

Finally, the third detective presents a piece of Knowledge collected from the defendant’s home. The Judge asks, “Did you have a warrant?” He is a theist, and he says, “Yes, it was signed by Yourself.”

It is clear and widely agreed that Knowledge requires Belief. For one cannot know something without in the first place possessing belief of its truth. It is ridiculous to say one knows the sky is blue when in fact he believes it is green. (Such a man is likely sick in body or mind.) But Belief alone is hardly sufficient to constitute Knowledge, for it is just as inappropriate to say that a man who in fact believes the sky is green therefore knows it is so. Indeed, another ingredient is the Truth of the matter. To know demands, then, belief in a true thing.

But is Knowledge merely the sum of Belief and Truth? It cannot be so: consider an ordinary man deep within an underground prison cell, where he is insulated from the whole outside world. There, he believes it is raining outside. Suppose it is indeed raining: even so, the insulated man, who cannot hear the rain nor feel its moisture, who receives neither news of the weather nor any other perception of the outside—does he know it is raining? In fact there are beliefs which are warranted and those which are not. But warranted beliefs cannot be formed in completely free fashion; indeed, common intuition tells us that genuine warrant must be formed in right or appropriate ways. For the man may believe it is raining outside due to the superstition that it always rains when he accidentally bites his tongue, yet such will not support the claim that he truly knows it is raining; his belief is hardly warranted. But if a man hears the pattering rain and see its droplets on the window, then his belief may well be well warranted. It is Appropriate Warrant which ties Belief of a True thing into Knowledge.

The precise nature of the notion of epistemic “appropriateness” governing proper and improper warrants is perhaps debated. (For instance, to avoid infinite regress, most contemporary epistemologists acknowledge that some beliefs—for instance, perceptual beliefs—must be properly basic, that is, warranted in themselves and not by virtue of having “reasons” for their truth.) In any case, even a cursory consideration reveals that notions of epistemic warrant are manifestly normative, that is, prescriptive: it pertains not to how beliefs are formed but how they ought to be formed.

There are severe implications, in the first place posing a severe problem for the place of Knowledge in metaphysical naturalism. For real Normativity itself finds no welcome in the home of the naturalist, to whom all phenomena are explicable in the terms of science, a descriptive discipline concerned only with what is but not what ought to be. Though the naturalist may grasp for ways to receive this desirable guest, none appear defensible. Ought is not reducible into Is. He cannot therefore accommodate epistemic warrant. Knowledge as such is an alien thing.

But, if one might for a moment imagine, what would it mean to reduce epistemic normativity to naturalistic terms? What about, for instance, a form of reliabilism based on statistical norms? A belief is warranted if it is formed by processes that give rise to predominantly true beliefs. Issues related to naturalistic explanations for belief aside, reliability is inadequate to constitute warrant as such reliably produced beliefs may be yet merely accidental. Or what if one supposed that natural laws give rise to epistemic warrant? Since all human thinking operates in strict accordance with natural laws, it must also operate in accordance with epistemic norms. Then all beliefs would warranted, even contradictory ones, inviting in all manner of irrationality.

What of _super_naturalist views? It must welcome Knowledge where naturalism turns it away by accounting for genuine, objective Normativity. But even supposing Normativity as such is permitted, how might one understand epistemic normativity specifically? Two or three analyses seem possible.

A deontological perspective situates epistemic norms in terms of ethical ones, that is, Duty or Obligation. Ethical norms include epistemic norms: one knows something if he believes it and has not flouted obligations in believing it (such as ignoring relevant evidence against it). Or perhaps epistemic norms arise instead from teleological normativity, from, that is, Purpose or Design. In the framework of proper function as advocated by Plantinga and others, beliefs are warranted only if they formed by processes or systems (for instance, cognitive faculties) which are functioning properly. Of course, “proper” implies that the faculty is intended or designed to function in certain ways and not others. Finally, consider virtue epistemology, which seeks to remedy reliabilism by demanding, in addition to reliable processes, exercise of intellectual virtues such as honesty, thoroughness, and fairness.

It should be clear that none of these species of normativity have any place in thorough-going naturalism. (Anderson, which we summarise here, catalogs various objections, all of which are unsuccessful.¹) What of non-theistic Platonism, which holds to transcendent, abstract universals in addition to the immanent, concrete particulars of immediate reality? But it is difficult to see how the epistemic duty of deontological normativity, the cognitive design of teleological normativity, or the qualities of virtue normativity could arise from impersonal abstracta. Indeed, each of these is a thoroughly personal notion, exhibiting features such as intellect and volition; the grounds for epistemic normativity must be personal.

One is pushed towards the theistic. What if all is divine, as in a pantheistic view? The problem is immediate: there is therefore no distinction between that which imposes norms and that upon which norms are imposed, the norms and the normed. Indeed, the adherents of pantheism seem to suffer from the same problem as those who would reduce epistemic norms to natural law: if everything is divine, then everything is normative or normal. In fact, Knowledge demands a division between the normative and the normal. If one additionally holds to monistic pantheism—not only is all divine, all is One, and metaphysical distinctions and differentiations do not really exist—then he is even worse off: there is not even a difference between warranted and unwarranted beliefs, nor the subjects and objects of belief.

One might additionally consider panentheism, which dispenses the immutable and impassible divine of classical theism for a mutable and responsive mutualism: the universe is not identical to God (as in pantheism) but is a proper part of it. God consists of the immanent, non-divine universe (the ‘body’) and the transcendent, divine aspect (the ‘mind’ or ‘soul’). Neither is absolute or independent from the other, and it is their mutual dependency and interaction that accounts for God’s temporal change. The impersonal panentheist faces problems similar to the non-theistic Platonist: if God is not personal, then from where do epistemic duties or cognitive design arise? But even if God is personal, one must establish the specific grounds for epistemic norms. Ostensibly, the immanent would be subject to norms arising from the transcendent. This leads to several difficult questions. Firstly, would not the latter’s dependence on the former have ramifications on the nature of the norms? Indeed, panentheists are reluctant to ground obligations in the character or will of the divine, which is constantly changing in response to the immanent. And if the transcendent is privileged in this way, would it not make more sense to call it alone ‘God’?

Indeed, the briefest of reflections makes clear that it is the classically theistic view which comfortably and defensibly accounts for Knowledge proper. For epistemic norms find their absolute, objective source in the character of a personal, supernatural God and his creation of human beings to reflect his image. It is natural that creatures should possess duties and obligations toward their Creator. Notions of proper function are perfectly sensible because God created humans to exercise rational cognitive and perceptive faculties. And epistemic virtues may reasonably reflect qualities valued by God. It is thanks to the sharp distinction between Creator and creature that the norm and the normed can exist in consistent and Knowledge-enabling relationship.

Then deny Him if thou wilt! But do not then say that thou knowst that God doth not exist, or that He is impersonal or universal or dependent on us; for where is thy warrant? Therefore we speak: if Knowledge, then God.