In Plato’s Theatetus, his characters debate over the nature and origin of knowledge. By the end of their discourse, though no conclusion is arrived at, a state of aporia is achieved. This sort of state does not necessarily signal the end of debate – rather, it could instead indicate that a reassessment of its foundational premises is required. In this paper, I show how Plato could arrive at a plausible definition of knowledge – given that he rejects that all knowledge is of a unified kind and accepts that knowledge does not necessarily require absolute certainty of truth. First, I outline the general discourse Plato lays out. Then, I proceed to modify his final definition of knowledge in a manner that forms a distinction between two kinds of knowledge – Knowledge of Deductive and Inductive natures. In the process, I craft a generally coherentist account of knowledge that is reasonably plausible.
The discourse in Theatetus:
Plato’s Theatetus presents us with a series of attempts at defining what knowledge in itself is. The underlying method he employs involves formulating a set of conditions that ultimately qualify knowledge as a unified kind. Perception is briefly considered, but is found to not be equivalent to knowledge, due to its inability to extend beyond instances in which it is had (Plato, 11). However, he notes that perception could instead be a stepping-stone to belief — his second definition. Reaching this point, it seems that since belief is fundamentally intentional — that is, directed at things or concepts, it could account for anything we could possibly claim to know. Still, beliefs are often false, and Plato does want to maintain that knowledge must be true. (Plato, 36) To Plato, then, belief (at least on its own) cannot be equivalent to knowledge. It seems that the belief had in knowledge requires some sort of justification to verify its truth.
The response to this notion brings us to the final set of conditions Plato explores for knowledge’s qualification — that knowledge is true belief with an account (Plato, 58). This rules out blatantly false belief as knowledge and makes justification a necessary condition in what constitutes knowledge. Still, Plato finds an objection to this definition — noting that the accounts given to justify belief themselves require accounts of justification, and that these, too, require accounts. It is clear that an infinite regress follows, and it seems like we may never be able to fully justify any belief — revealing a problem of inaccessibility toward knowledge that arrives as consequence of this definition (Plato, 66).
Knowledge, Modified:
The debate in Theatetus ends in aporia, and no further inquiry into knowledge persists. In this section, I hope to carry on the debate by revising the assumptions made by Plato in his exploration of knowledge. As mentioned earlier, my first revision involves splitting knowledge into two kinds — deductive and inductive knowledge. Inductive knowledge is grounded in empirical observation. If X seems to follow from Y in every instance observed so far, then we can induce that X follows from Y in general. Deductive knowledge, though, is independent of observation. Given a set of rules or axioms, these being either self-evident (in tautology, for example) or true by definition, we can deduce certain truths derived from and in relation to them. Though it is true that both deductive and inductive knowledge rely on a coherence between a set of proposed statements, I maintain that they are distinctly of different kinds by virtue of the nature of the statements that they are coherent with.
To show how deductive knowledge avoids the regress problem, there is no real need to drastically alter the argument already presented to us by Plato. Take the knowledge of mathematical principles as an example. First, justifying that “2 + 2 = 4” (or any equation, for that matter), requires nothing beyond reason — this claim must be true within the axiomatic framework that constitutes mathematics in the first place. “+” and “=“ perform the functions that they do purely in virtue of their definitions. Since “2 + 2 = 4” is coherent with these statements, we need not perceive that “two sets of two tokens add up to a total of four” to know that it is true. All deductive knowledge involves reasoning within a logical framework built upon axioms that must be true. As such, whatever statements built upon them within these parameters relate to one another in a way that can never be false. Claims of knowledge within these frameworks are self-contained truths, and require no accounts beyond the axioms they were derived from.
When considering inductive knowledge in light of the infinite regress problem, though, it is evident that, in a similar way to how axioms prevent infinite regress in the justification of deductive knowledge, we must formulate a set of conditions that reasonably satisfy what a foundational account for justifying inductive knowledge might look like. I believe this account is part constituted by knowledge gained through perception. This allows that in cases that have to do with justifying belief of sensible qualities like shape, size, and colour, perceiving those qualities would be sufficient justification for their truths. Knowledge that a particular chair is blue, for example, would be justified by the phenomenon of my perceiving that it is so. However, cases of illusion and hallucination cast doubt on the reliability of any belief formed by perceptual experience. Beyond this, perception on its own can only provide so much justification for belief— it is not the case that we perceive everything that we believe. This is true of things that are not directly accessible by the senses at the given moment (knowledge that Antarctica exists without perceiving it, for example), but also of entities that cannot be directly observed by our sensory capacities at any given moment (knowledge of unobservable entities like electrons, for example). As such, my second modification to Plato’s understanding of knowledge states that a coherentist account of justification is necessary to alleviate the problems faced by the current partial account.
In cases of hallucinatory and illusory perceptual experience, coherence of account through an appeal to public availability (that is, a coherence between interpersonal accounts) provides reasonable justification for inductive knowledge gained in this manner. If it sensibly appears to a subject that an object is right in front of him, and it sensibly appears to others to be the case, then there is reason to conclude that the subject’s sensory account of the world is coherent with others — thus justifying this belief about the world informed by perceptual experience. Similarly, in broader cases of beliefs grounded in induction, having an expansive framework of inductive claims that remain consistent with one another grants the whole body of knowledge, constituted by this framework, great epistemic power. Appealing to the way these claims relate to one another in a coherent manner reflects an approach to justifying knowledge similar to that of the manner in which deductive claims are justified. Still, inductive claims on their own are not self-evident — this is what distinguishes inductive knowledge from deductive knowledge. But, through appeal to coherentism, we allow for inductive beliefs about unobservable entities to be justified by use of scientific instruments that provide reasonable grounds for inductive inference. These inferences, when (and only when) in agreement with a greater body of inductive claims, display a coherence that seems sufficient for their justification.
Still, an objector to this account might reference the problem of induction — stating that inductive belief cannot allow for certainty of truth, and that if truth were necessary to knowledge, it cannot be the case that beliefs arrived at through induction could ever qualify as such. While this is true, it is an insurmountable limitation of induction. As such, we must weaken the definition of knowledge slightly with regards to truth’s importance within its constitution. So, the final account of inductive knowledge I reach in this discourse is as follows: “inductive knowledge is well-corroborated belief”. This requirement for corroboration brings to light one further distinction between inductive knowledge and deductive knowledge — Since the justificatory power of these claims lie in their consistency rather than their truth, it is possible to alter the framework itself (even drastically, in some cases). This framework is self-correcting in way that deductive knowledge is not and serves as reason to clearly distinguish these two kinds of knowledge from one another.
Conclusions:
In summary, Plato’s Theatetus provides great insight into modern epistemology, and the problems he raises are as relevant then as they are now. Bringing the debate to aporia, as plato had done, did a great deal of the philosophical legwork that modern philosophers now build upon with reference to the wider body of knowledge they now have access to. The account of knowledge presented in this paper is not perfect, but it does hope to explore some possibilities much in the same spirit that Plato’s Theatetus exhibits.
Works Cited:
Theaetetus. 1st ed. Project Gutenberg, 2008. Web. 1 Apr. 2020.
Written by,
Eryn Spicer (Social Media Manager 19/20′)
Edited by,
Joshua Chua (Editor 19/20′)
Editor’s Note:
Have any essays/works/ideas you would like to share? Contact us at our Instagram @nanyangphilosophy to submit your essays/works/ideas for publishing.
The Nanyang Philosophy Review 