A Priori Knowledge

Central to views regarding a priori knowledge is how such beliefs gain their epistemic justification independent of experience. Central to Alvin Plantinga’s account of what it takes for one to have a priori justification is for one to see the truth of the proposition in question. According to Plantinga, to see that a proposition is true is to believe that it is true, and necessarily true, to form this belief immediately (not on the basis of other beliefs, memory, or testimony), to form this belief with a particular, hard to describe, phenomenology, and to do so while not malfunctioning. One can also gain a priori justification for a proposition by seeing that it follows from a proposition that one sees to be true. When one sees that a proposition is true or that it follows from a proposition that one sees to be true, and the proposition is true, then one has a priori knowledge of it.

There are a couple of problems with Plantinga’s account. First, it seems that Plantinga has not accurately captured what it is for one to see that a proposition is true. Plantinga claims that believing a proposition is a necessary condition for seeing that it is true, but this does not seem right. Propositions can be seen to be true to me even though I do not believe them (such as that the top line in the Muller-Lyer illusion is longer, or Frege’s Axiom V). Similarly, one can believe a proposition without it seeing that it is true, such as when one has seen a mathematical derivation for a theorem that is too complicated to seem true. Beliefs are typically formed on the basis of seeing that a proposition is true, but the seeing and the believing are distinct relationships one has to the proposition. As such, Plantinga gets the nature of seeing the truth.

A second problem with Plantinga’s analysis concerns the modal requirements for gaining a priori justification. Plantinga claims that to see that a proposition is true one must not just believe it, but believe that it is necessarily true. It seems that one can come to know a proposition in an a priori fashion while at the same time being ignorant of modal concepts like necessity or having mistaken views regarding necessity according to which the proposition in question is not necessary (I am assuming here, with Plantinga, that all propositions that are known a priori are necessary). For instance, I could justifiably believe that mathematical propositions, like 2+2=4, do not have their truth values necessarily. Even though I am mistaken in this regard, it still seems that I can see the truth of 2+2=4 and/or that I can know that 2+2=4 a priori. In addition, Plantinga’s appeal to believing that the proposition in question is a necessary truth threatens an infinite regress. Presumably, the belief that the proposition in question is necessarily true is one that must have a priori justification (if not, then it is hard to see how it could contribute to the a priori justification of the proposition in question). If so, however, then this belief too must be seen to be true but in order to see that it is true one must also believe that it is necessarily true. The ‘necessarily’ modifiers will quickly compound leading to an infinite regress and to propositions that are plausibly too complicated to be believed by human minds. This reveals another flaw in Plantinga’s account.

George Bealer relies on intuitions, or intellectual seemings, to provide the a priori justification required for a priori knowledge. For S to have an intuition that P is for it to seem to S that P. Thus, intuitions are conscious episodes. Bealer distinguishes intuitions from beliefs for the reasons mentioned above so he avoids one problem that encountered Plantinga. However, Bealer believes that it is rational intuitions that do the work for a priori justification, and according to Bealer, a rational intuition presents a proposition as necessary – it must seem to S that P must be true. Worries arise here, like above, since it seems as though one can have a priori justification without the modal concepts Bealer appeals to or if one had a mistaken view of the relevant modal concepts (as described above).

Another contemporary proposal claims that the requirements of concept possession can provide the needed a priori justification. Paul Boghossian claims that in order to possess certain concepts one must be disposed to reason in certain ways. For instance, in order to possess the concept ‘conditional’ one must be disposed to reason according to modus ponens. The claim is that such inferences are thus justified in virtue of their being requisite for the possession of certain concepts. Propositions can then be known in an a priori fashion when they are the conclusions of such justified inferences.

Several problems are apparent with this account. First, it seems doubtful that one must be disposed to reason in certain ways in order to possess certain concepts. There does not appear to be anything incoherent with the idea of a wholly passive mind that possessed concepts but was unable to do any mental acts such as infer. So, it is doubtful that having such dispositions to reason is indeed requisite for the possession of concepts.

Second, even if such dispositions were requisite, this fact does not epistemically justify their use. Doing the work in order to possess certain concepts may be rational in a means/ends sense, but it does nothing to epistemically justify or entitle one to make such inferences. We could imagine a case where S is offered some epistemically valuable end if S performs the inference from P or Q to P and Q. Performing such an inference would be beneficial for S, but this fact in no way epistemically justifies S in performing the inference.

Finally, even if the inferences were justified a significant problem remains for Boghossian’s account. If the conclusions of such justified inferences are supposed to be justified a priori, then we need to have premises that are justified a priori. All that Boghossian’s account even attempts to do is to justify the inferences, but this is inadequate to the task at hand – the task of accounting for a priori knowledge.