From A Priori Knowledge and Miracles
by IBD Vice President Matthew J Coombe
Essentially there are two types of knowledge,1 a priori (“from the earlier” or “before the senses”) and a posteriori (“from the later” or “after the senses”). The distinction between the two types is purely epistemic in nature. The most telling difference between the two, as described by Kant is, a priori knowledge is independent of all experience,2 and includes propositions such as, “all bachelors are unmarried males,” and “7 + 5 = 12.3”
Classically, many philosophers have accepted a priori knowledge but recently it has received much scrutiny and even some have attempted to reduce it to pure linguistics.4 The skepticism results from the failure to distinguish a clear and coherent account of the classical conception of a priori knowledge from a general theory of knowledge.5 To properly distinguish and apply a priori knowledge one must ask two questions, 1) what is the primary target of the analysis? 2) Does the analysis of the primary target presuppose a general theory of knowledge?6 If the target requires a theory of knowledge then it cannot be considered truly a priori. Some epistemologists, such as Phillip Kitcher, not only argue that propositions have no intrinsic meaning,7 (and therefore there is no a priori knowledge) but that a priori knowledge is experientially indefensible8 and therefore cannot even be proven at all.
A priori knowledge, if it exists, must be known independently of experience. Several issues are quickly brought to attention, is it possible for man to have knowledge independent of experience? If the answer is “no” then Hume’s a priori rejection fails. If the answer is “yes,” then it must be discerned if the type of knowledge utilized by Hume is indeed free from experience. So then, the issue before the house is. is knowledge contingent upon experience? If one incidence of knowledge can be determined to have occurred apart from experience then a priori knowledge is possible.
Consider a primitive alien planet. On this planet, Gog and Harry (two aliens) are looking at a pile of roldals (the closest equivalent to these on earth are apples). Gog places two roldals on the ground; though to delineate “two” Gog does not utilize the word “two” but rather “glue.” Likewise Harry grabs “glue roldals” and places them next to Gog’s roldals. Now they want to figure out how many they would have if they were to combine them. They designate a stick to be a “+” sign and a rock to be an “=” sign. Further they decide the number four shall be known as “horse.” Therefore: glue roldals, stick, glue roldals, rock, horse roldals— two plus two equals four.
Did experience aid in the formulation of the previous conclusion? It would seem that certain mathematical claims are universal and necessary. James R. Beebe examines if things like mathematical proofs are indeed “putatively a priori necessities.9” Beebe argues that due to pervasive nature of empiricism (or rather a posteriori knowledge) there exists an inherent tendency away from a priori knowledge;10 this tendency is not prima facie against a priori knowledge but rather for a posteriori. Beebe’s thesis is to investigate if this penchant is indeed justified. Instead of focusing on basic principles, most of the skepticism applied to a priori knowledge is concerned with impractical linguistic word play syllogisms. For example: 1) If I know that 2 + 3 = 5, then I know that I am not involved in any subject whose a priori beliefs are massively and constantly in error due to skeptical circumstances. 2) I do not know whether or not I am affected by such skeptical circumstances. 3) Therefore, I do not know that 2 + 3 = 5.11 Beebe sites Wittgenstein and Descartes as those who pose the type of skeptical circumstances that prevent such knowledge. For example, Descartes argues in his Third Meditation that it could be possible for God or some deity to deceive every instance of reliable knowledge12 (such is the basis for the previously mentioned syllogism). The problem is, even the proposition, “I should be skeptical concerning a priori knowledge because my environment could be fake, contrived, and/or deceiving,” is in fact an instance of a priori knowledge. Further, even if the most hyper-skeptical environment existed, it would in no way negate the veracity concerning mathematical principles. Consider the aliens once again, but in this instance suppose that they are in a completely computer generated environment and everything around them is fake. Even in this skeptical and false world, the number of rodals, when properly added, will always remain the same. Ultimately, Descartes viewed hyper-skepticism as a menace that restricted intellectual discourse and argued that it was not a useful epistemic tool.13
To determine the connection between a priori knowledge and miracles we need to question the relationship between “before the senses” knowledge and experience. While a priori knowledge can be independent of a hyper-skeptical world, this in no way ensures a unilateral connection between the two—a connection would be required if miracles were either to be accepted or objected via a priori knowledge. Confusion surround the nature of a priori thought further conflates the issue.
Merely because something is a priori, this does not entail that it must necessarily be devoid of any a posteriori components in order for it to remain an instance of a priori knowledge. For example, Plantinga argues that if there were five passengers in a car crash and two survived the crashed, we can know the number who died in the crash a priori is three. The a posteriori knowledge factors in when one considers, what a car crash is or what it means to “survive,” or “die.” Since the combination of the two types of knowledge do not contradict each other, then the combination in no way invalidates the instance of a priori knowledge.14
Consider four scenarios, in each of these scenarios there is a car crash of which there are five passengers, three victims, and two survivors. In the first scenario, the accident took place in a movie, the second was reported on during the evening news, the third was told from a friend who had witnessed the accident (though this friend is known for embellishing) and you witnessed the fourth scenario. The ability to obtain a priori knowledge in each of these scenarios is in one sense contingent upon the a posteriori and in the other independent from it. The veracity of a priori knowledge (in these incidences) is dependent on the efficacy and types of source accounts and evidence.15 In the movie example, the a priori knowledge is sound, but the event is not actual. The other three scenarios likewise result in consistent a priori knowledge iff the sources are accurate.
Therefore, a priori knowledge can be predicated on a posteriori knowledge and the burden of proof in these scenarios is on the a posteriori and not the a priori.16 Further, there is a link between experience and a priori knowledge. This link is not necessary, but when it does occur the veracity of the a priori knowledge is not contingent upon itself, but rather the veracity of the a posteriori presuppositions and justification is the contingency. Thus, to question certain instances of a priori knowledge is merely to question the justification for the a posteriori. It is because of this justification requirement that some epistemologists have argued for the superiority of a priori knowledge.1718
The final question concerning the link between the two types of knowledge is to answer the question if experientially justified a priori knowledge is capable of being defeasible or to question if it is able to be overturned. Epistemologist Joshua Thurow argues that if a priori knowledge is possible then it is defeasible by non-experiential evidence (due to its very nature). However, if it is defeasible by non-experiential justification then it would likewise be overturned by experiential evidence.19 This conclusion however seems unmerited. To determine if a priori knowledge is defeasible or not, the nature of the claims associated with it should be detailed to a further degree.
All necessary truths are incidences of a priori knowledge,20 but not all incidences of a priori knowledge are necessary. For example the law of excluded middle is a necessary truth that can be known a priori; it is necessarily the case that something cannot be both true and not true in the same sense at the same time and one is able to know this without examining anything in physical reality. For example, someone cannot rightly deduce, “I cannot know truth,21” because if true, her premise fails, and if false then truth can be known and would result in an instance of a priori knowledge. Further, this would be an example of an a priori truth that is also necessary.
Some a priori truths might be instances of knowledge but not necessarily true. The only types of a priori truths that are not necessarily true are those that are true by definition. “All bachelors are unmarried males,” is driven by a definition that need not necessarily be true—there could be a possible sub-culture where “bachelor” means a newly married male. What makes such an instance actually a priori knowledge is because “all bachelors are unmarried males” is true by definition and therefore requires no research or experience—thus, a priori.
Some have attempted to reduce necessary truths to pure linguistics.22 But as argued concerning the example of necessarily or axiomatic truths (of which to deny is self-refuting) such examples cannot be reduced to pure linguistics. While some necessary truths can be reduced to tautologies, this by no means entails all necessary truths are reducible to merely linguistically significant statements.2324 Even Thurow eventually concedes that instances of necessary truths which are defeasible by experience are of the “true by definition” variety and therefore the linguistic factor could affect the truth value, but, iff the definition were equivocal.25
In conclusion, a priori truths can be necessary or true by definition and either can be justified depending on the usage or if is predicated on some a posteriori truth. Necessary a priori truths are irrefutable but are limited in scope. True by definition a priori truths are contingent upon language and in some cases limited by a posteriori foundations (such as the example of car accident). Even before examining Hume, it seems unlikely he would consider his objection as “necessary” therefore, if his objection has merit it would have to be a true by definition or a posteriori contingent a priori truth.
1Alvin Plantinga, Warrant and Proper Function (Oxford University Press, USA, 1993).2
2Jonathan Dancy, Ernest Sosa, and Matthias Steup, A Companion to Epistemology (John Wiley & Sons, 2009). 1
3Plantinga, Warrant and Proper Function. 3
5Albert Casullo, “Analyzing a Priori Knowledge,” Philosophical Studies 142, no. 1 (January 2009): 77–90, doi:http://dx.doi.org.ezproxy.liberty.edu:2048/10.1007/s11098-008-9302-5.
7Philip Kitcher, “Knowledge, Society, and History,” Canadian Journal of Philosophy 23, no. 2 (June 1, 1993): 155–177, doi:10.2307/40231815.
8Philip Kitcher, “A Priori Knowledge,” The Philosophical Review 89, no. 1 (January 1, 1980): 3–23, doi:10.2307/2184861.
9James R. Beebe, “A Priori Skepticism*,” Philosophy and Phenomenological Research 83, no. 3 (2011): 583–602, doi:10.1111/j.1933-1592.2010.00488.x.
12René Descartes, Meditations on First Philosophy (NuVision Publications, LLC, 1960). 12
13Harry M. Bracken, Descartes (Oneworld, 2002).15
14Plantinga, Warrant and Proper Function. 3
15Casullo, “Analyzing a Priori Knowledge.” 89
17Darragh Byrne, “A Priori Justification,” Philosophical Books 48, no. 3 (2007): 241–251, doi:10.1111/j.1468-0149.2007.00447.x.
18Plantinga, Warrant and Proper Function. 8
19Joshua Thurow, “Experientially Defeasible A Priori Justification,” The Philosophical Quarterly 56, no. 225 (2006): 596–602, doi:10.1111/j.1467-9213.2006.461.x.596
20Albert Casullo, A Priori Justification (Oxford University Press, USA, 2003). 88
21Paul A. Boghossian, Content and Justification: Philosophical Papers, Text is Free of Markings (Oxford University Press, USA, 2008). 177. Boghossian is less inclined to believe something so specific can be known via a priori knowledge, but he is willing to allow statements such as, “I am currently entertaining a thought,” which equally argues for my point concerning propositions that are necessarily true because to deny the proposition ultimately affirms it.
22E. D. Klemke, “The Laws of Logic,” Philosophy of Science 33, no. 3 (September 1, 1966): 271–277, doi:10.2307/186275..
24Alfred J. Ayer and Sir Alfred Jules Ayer, Language, Truth and Logic, 2nd ed. (Dover Publications, 1952). 16
25Thurow, “Experientially Defeasible A Priori Justification.” 600