Alvin Plantinga has developed a theory of knowledge called proper functionalism. In this article I will first explain what he calls “the central and paradigmatic core” of his notion of warrant, and second I will give counterexamples to his theory showing that his proposed criterion is neither necessary nor sufficient for knowledge. And lastly I will argue that his appeal to a supernatural designer is undercut by his own criterion of warrant. The main idea to my third point is twofold: first, according to Plantinga, intentional design according to a design plan is required for proper function and warrant, but God, because he is a person with cognitive faculties that were not intentionally designed according to a design plan, would lack function and the ability to produce warranted beliefs. Therefore, if God’s cognitive faculties cannot function properly and his beliefs are unwarranted, then it makes little sense to say that God designed our cognitive faculties such that ours function properly and produce warranted beliefs. And second, Plantinga’s appeal to divine knowledge is uninformative and any analogy between it and warrant breaks down.
Plantinga has developed proper functionalism as a nonevidentialist account of warrant. That is, instead of the justification of our beliefs coming from the internal access an agent has to good reasons, the nonevidentialist aims to provide an account of the external factors that describe how a true belief can become knowledge. For Plantinga, ‘warrant’ is “the quality or quantity enough of which, together with truth and belief, is sufficient for knowledge.” Proper functionalism states that a belief has warrant if and only if the following four conditions are met:
- The belief has been produced by cognitive faculties that are working properly
- The belief has been produced in a cognitive environment that is appropriate for the agent’s kinds of cognitive faculties
- The segment of the design plan governing the production of that belief is aimed at the production of true beliefs
- There is a high statistical probability that the belief produced under those conditions will be true.
The first and second clauses are the idea of “proper function,” or that a warranted belief must result from properly functioning faculties that are functioning as they were designed. The third clause is intended to rule out warranted beliefs that arise from wish fulfillment or other non-truth seeking aspects of a mental system. And the fourth clause is a reliability condition, similar to what is found in reliabilism.
These specifications are meant to provide an account of knowledge that avoid the Gettier cases of true belief that are true merely by accident; Plantinga argues that the Gettier cases are either beliefs formed not by virtue of proper function of the belief-producing mechanisms involved, or the source of the cognitive glitch is belief-producing mechanisms that are not functioning in an appropriate cognitive environment. While proper functionalism can indeed avoid the Gettier-style counterexamples, I will argue that, given Plantinga’s criterion, we can still generate counterexamples that undermine Plantinga’s notion of warrant as the necessary and sufficient conditions for knowledge.
1. The Lack of Sufficient Conditions
First, proper function is not sufficient for knowledge. I will focus on the “proper function clauses,” since those are unique to Plantinga’s theory. Consider Plantinga’s counterexample to Alvin Goldman’s reliabililst theory of knowledge:
[S]uppose I am struck by a burst of cosmic rays, resulting in the following unfortunate malfunction. Whenever I hear the word “prime” in any context, I form a belief, with respect to a randomly chosen natural number less than 100,000, that it is not prime. So you say “Pacific Palisades is prime residential area” or “Prime ribs is my favorite” . . . ; I form a belief, with respect to a randomly selected natural number between 1 and 100,000 that it is not prime. The process or mechanism in question is indeed reliable (given the vast preponderance of non-primes. . .) but my belief—that, say, 41 is not prime—has little or no positive epistemic status. The problem isn’t simply that the belief is false; the same goes for my (true) belief that 631 is not prime, if it is formed in this fashion. So reliable belief formation is not sufficient for positive epistemic status.
While proper functionalism may be able to deal with these kinds of malfunctions of cognitive faculties, it cannot deal with the possibility of similar (yet bizarre) cognitive systems that were intended to function in the same way. For instance, the designer could have mistakenly thought that the ‘prime’ belief-forming process was the best way to maximize true beliefs about non-prime numbers, thereby intending that his creatures have it as part of their design plan. If so, such creatures that form belief in this way would have cognitive faculties that function properly, in an environment that is appropriate for the their kinds of cognitive faculties, while the production of the belief would be from a segment of their design plan that is aimed at true beliefs, and would be produced with a high statistical probability that under those conditions the belief would be true. Here, Plantinga’s entire criterion of warrant is met; however, we would not consider their belief in non-prime numbers to be knowledge—their beliefs would lack positive epistemic status. Therefore, proper function is not a sufficient condition for knowledge.
2. The Lack of Necessary Conditions
Next, proper function is not necessary for knowledge. Suppose a person is the victim of a brain lesion, either due to injury, infection, exposure to certain chemicals, or an autoimmune disorder. And suppose the lesion results in an enhancement in the person’s cognitive abilities—the legion produces experiences in them that more clearly resemble their actual past experiences. That is, these lesion-induced experiences result in memory beliefs that have greater detail than the actual experiences. Under proper function theory, these memory beliefs would not be a case of knowledge because the person’s cognitive faculties would not be functioning according to their design plan. But this seems absurd. An extra clear belief of a past experience that is true and results from a brain lesion ought to still count as knowledge. It therefore seems possible for knowledge to exist without proper function according to a design plan; and so, proper function is not a necessary condition for knowledge.
It is clear, then, that having a warranted belief when one’s cognitive systems are functioning properly is an entirely contingent matter. Thus, there seems to be a distinction between something functioning as it was designed to function, and its functioning well; that is, if it is a well designed system, then warranted beliefs will result from proper function; but if it is not a well designed system, then there can be a divergence from what we consider knowledge. Feldman (2003) argues that a possible response from the proper functionalist could be that a cognitive system is functioning properly when it forms beliefs that are supported by the evidence available to the system. This, however, becomes problematic for the nonevidentialist: it undercuts their commitment to nonevidentialism, and is a return to an evidentialist theory. Feldman points out that proper functionalism was designed to be a rival to evidentialism, not a terminological variant of it.
3. God’s Lack of Proper Function and Warranted Beliefs
Plantinga’s appeal to a supernatural designer as the guarantor of our proper function and warrant faces a challenge. As we’ve seen, Plantinga sets out a criterion of warrant that includes proper function according to a design plan. However, God is said to be a person with his own cognitive faculties, but of the kind that were not designed and do not function according to a design plan. So it seems, according to Plantinga’s criterion, God’s cognitive faculties lack function, and therefore cannot produce warranted beliefs. This entails that God lacks knowledge, and so any design plan or intended function that God created in human beings will also lack warrant; or at least this difference gives us a reason to doubt the reliability of our own belief-producing mechanisms on the assumption of theism. Thus, we can conclude that if theism is true (at least the kind that Plantinga envisions), then our beliefs do not have warrant, or if they do it is inscrutable.
Plantinga attempts to defend his view against this objection in a footnote in his book Warrant and Proper Function:
Of course, God’s knowledge is significantly different from human knowledge: God has not been designed and does not have a design plan (in the sense of that term in which it applies to human beings). When applied to both God and human beings, such terms as ‘design plan’, ‘proper function’, and ‘knowledge’, as Aquinas pointed out, apply analogously rather than univocally. What precisely is the analogy in this case? . . . God has not been designed; still, there is a way in which (if I may say so) his cognitive or epistemic faculties work. This way is given by his being essentially omniscient and necessarily existent: God is essentially omniscient, but also a necessary being, so that it is a necessary truth that God believes a proposition A if and only if A is true. Call that way of working ‘W’. W is something like an ideal for cognitive beings—beings capable of holding beliefs, seeing connections between propositions, and holding true beliefs . . . W, therefore, is an ideal for cognitive design plans, and it is (partly) in virtue of that relation that the term ‘knowledge’ is analogically extended to apply to God.
There are several problems with this. First, it is not clear how we can appropriately apply concepts analogically with a cognitive being that is “significantly different” from human beings. I see two ways of interpreting Plantinga here. The first is we can take him to be making the strong claim that we can gain legitimate information about the ideal for human knowledge by contemplating God’s knowledge by analogy. This, however, becomes problematic given the significant differences in how human beings reason: consider, for instance, how we commonly form beliefs though inference. It is a mystery why God would ever need to infer anything, since an omniscient being would have all true propositions present before its mind in a single instance (or for all time). So there seems to be a breakdown in the strong interpretation in the analogy between human thinking and God’s thinking.
A second way we could interpret Plantinga is in a weaker sense: we may not be able to gain legitimate information about human epistemology by contemplating W, but rather W is more like an idealized concept of a way of thinking that human epistemology is a copy. So we can meaningfully speak of W, but we shouldn’t expect to learn anything about human epistemology, or divine knowledge, from an analogy to it. 
But this too faces an objection. According to Plantinga, human beings obtain warrant in the externalist sense; that is, warrant is had when conditions external to the agent’s mind are met (see the criterion above). Without these factors, warrant is not had. But for God, there are no external factors that either allow for, or prevent, warranted beliefs. As Plantinga says, God just believes all true propositions necessarily. So for God, the only two conditions for “knowledge” are belief and truth. This means, then, that warrant is not part of God’s knowledge! The so-called “ideal for [human] cognitive design plans” is that of a being who believes things solely due to his internal mental states.
So there appears to be a disconnect at a crucial point in Plantinga’s theory. I submit that it makes no sense to ground an externalist theory of warrant in a being whose beliefs are “known” by internal factors alone. By his own criterion, there is no resemblance to be found between human warrant and God’s knowledge. And apart from that, stipulating that God is omniscient doesn’t advance our understanding of how our belief-forming practices analogically relate to God’s, since we don’t know what it’s like for God to believe a proposition in the first place. The mechanisms or processes that God uses to believe anything is unseen by us, so any analogy to God’s thinking fails to tell us anything informative about it. So even on a weaker interpretation of Plantinga’s analogy, it is uninformative and the resemblance we need is not found.
Plantinga’s proper functionalism, as presented within a specific theistic framework, is insufficient at providing a plausible theory of knowledge. The criterion of warrant fail as sufficient conditions for knowledge, as in the case of bizarre design, and the criterion fail as necessary conditions, in the case of unexpected, yet better cognitive function. And lastly, Plantinga’s appeal to an analogous relationship between human being’s cognitive design plan and that of God’s knowledge is problematic. If a God exists, it remains significantly different than us—its mind does not function as ours does, since it doesn’t infer things, and doesn’t rely on external factors for warranted beliefs. In fact, given Plantinga’s criterion of warrant, God doesn’t have warranted beliefs at all. Merely believing all true propositions doesn’t provide an analogous resemblance or ideal for why our cognitive design is the way it is. The appeal to God’s omniscience reverts away from the external conditions that cause our warranted beliefs, so Plantinga’s theistic framework is of little help to the nonevidentialist in understanding how we have warranted beliefs.
 Alvin Plantinga, Warrant and Proper Function (Oxford University Press, 1993).
 Ibid, Plantinga, p. v.
 Dr. Watson, “4.5: Proper Function Theory,” video lecture (Arizona State University, 2017).
 Richard Feldman, “Review: Proper Functionalism,” Noûs, Vol. 27, No. 1 (Mar., 1993), pp. 34-50.
 Ibid, Plantinga, p. 35.
 Alvin Plantinga, Warrant: The Current Debate (New York: Oxford University Press, 1993), p. 265.
 Richard Feldman, Epistemology (Pearson Education, Inc., 2003).
 Ibid, Feldman, p. 104.
 Ibid, Feldman, p. 103.
 Ibid, Feldman, p. 104.
 Ibid, Plantinga, p. 236.
 Special thanks to Dr. Watson for bringing these points to my attention.