The veracity of Gettier’s argument against Justified True Belief being knowledge, in his article “Is Justified True Belief Knowledge" (Gettier, 1963, pp. 121ff) has been assailed from several quarters, since the publication of this article. However, many of the arguments advanced against Gettier have not been forceful enough to vitiate the strength of Gettier’s argument. Gettier has rejected the claim that knowledge can be defined in terms of belief, truth and evidence because the ‘sufficiency’ and ‘necessity’ requirements some scholars advance cannot guarantee the belief status to be knowledge. The tendency has been to refute Gettier by citing counter examples which are common but trivial in the sense that they do not touch the heart of the matter.
As a matter of fact, the knowledge – belief distinction rests on a certain number of classical examples. But Gettier would simply not be impressed by such examples. However, I do not agree with Gettier as will be made clear in the sequel. I feel the result of Gettier’s argument is an emphasis on a fundamental cleavage between knowledge and belief which is too extreme.
My interest in reviving this old debate about the knowledge and belief distinction is consequent upon the fact that the attempts made by some philosophers to refute Gettier’s claim as mentioned earlier, are not strong and stimulating enough. Although this is not to say that the objections of Gettier’s critics are obsolete and antiquated, yet they have not come out with sufficiently critical illustrations that will show the effectiveness of their denial of Gettier’s thesis. This throws the debate open. My attempt in this paper is meant partly to show by practice that philosophical discussions are never exhaustive because solutions to philosophical problems are neither closed nor absolute.
In Edmund Gettier’s ‘Is Justified True Belief Knowledge?’, a critical exposition of A. J. Ayer’s and R. Chsholm’s necessary and sufficient conditions for someone’s knowing a given proposition, it was argued that a person may have no justification in claiming that he knows a proposition p, given that
1. p is true
2. s believers that p
3. s is justified in believing that p (Sosa, 1964, p1. )
In other words, Gettier’s task was to show that ‘Justified True Belief’ cannot be equated with knowledge because the necessary conditions usually given in the classical definition are not jointly necessary and sufficient for belief to assume the status of knowledge.
In this paper, I shall not be trying to show that Gettier’s thesis is false in absolute terms. I shall rather suggest that its appeal is not as forceful as it seems and that sound examples could be given that would show that the ‘sufficiency’ and ‘necessity’ requirements do not provide adequate grounds for Gettier’s attempted refutations. Gettier presents us with two counter examples to show that conditions (i) – (iii) are not sufficient for knowledge. As put by Ernest Sosa (Sosa, p.1), Gettier’s first example is one in which a true ‘entailment’ is derived from a false proposition. The second is one in which S has good grounds for his belief that p, and from this deduces that pvq. However, unknown to s, (p. q. Thus the view that
1. p is true;
2. s believes that p;
3. s has evidence that p would not show adequately that S knows that pvq.
In each of Gettier’s counter-examples, a proposition which is in fact true and believed by S to be true on good grounds still fails to qualify S’s claim as knowledge claim since the grounds adduced by S for knowing that P are false.
Michael Clark, in his “knowledge and Grounds: A Comment on Mr. Gettier’s Paper", is also reported to have argued that there are non-deductive as well as deductive grounds and so Gettier’s examples are “stronger than they need have been" (Sosa, p.2). Clark is also believed to have said that granted S’s belief that P is consequent upon a reliable and honest friend’s report, we do not want to rashly conclude that his friend’s unqueried guess enhances S with knowledge. Furthermore, in an attempt to correct Gettier, clark argues that if we pursue the matter of grounds, we will reach a point at which such demands will no longer make any sense. Thus to avoid a kind of infinite regress, he considers that we must recognise a cognitive bottomline where it makes no sense to ask for grounds or evidence for a belief (Sosa, p.2).
My impression is that Clark’s contention against Gettier does not solve any problem. It can be conceded to Clark that we need to break the chain of grounds if knowledge is to be meaningfully accounted for, but his reformulation of the conditions of knowledge leaves the central matter of how to conceive of truth in relation to knowledge claims essentially untouched. One of the crucial points in Gettier’s whole argument is the point that a false proposition can entail a true one. I would like to illustrate this with Gettier’s examples (d) and (e) in his first illustration as follows:
1. Jones is the man who will get the job, and Jones has ten coins in his pocket.
2. The man who will get the job has ten coins in his pocket.
Clearly, (e) is an entailment from (d) Gettier sees (d) as false because although Jones has ten coins in his pocket he is not the man who will get the job; he sees (e) as true because Smith has ten coins in his pocket and he will get the job. In this case it can be said that a false proposition entails a true one. This concedes that ‘Justified True Belief’ can fail to be knowledge and as such the truth of any knowledge claim is relative to time and circumstances.
In a similar vein, T. Sorell in his “The Analysis of Knowing" has attempted to show the futility of equating Justified True Belief to Knowledge, irrespective of revisions or reformulations of the analysis of Justified True Belief.
Broadly speaking, Sorell distinguishes between two types of conservative reformulations of Justified True Belief: (i) conclusive reason analysis and (ii) the defeasibility analysis. The conclusive reason analysis fails, according to Sorell, because it can only work for mathematical knowledge; and as he contends, “the requirement that one’s reason guarantee the truth of one’s belief makes sense when what is in question is mathematical knowledge…". The defeasibility analysis rests on a ‘gap’ between the subject’s reasons for what he truly believes and the sum-total of his evidence as related to the belief. This analysis is problematic, says Sorell, because “it makes any further evidence that would weaken justification sufficient for taking away knowledge" (Shope 1963, p.21).
My impression is that Sorell’s proof, a proof of the impossibility of the non-equivalence of Justified True Belief to knowledge is not, and cannot be a non-equivalence and conclusive proof of Justified True Belief to knowledge. One can concede to him that there is a difference between apriori knowledge and a posteriori truth. However, my contention is that the traditional conception of Justified True Belief against which the venom of his attack is directed is not about mathematical knowledge, but ‘general knowledge’ as he chooses to distinguish. Thus his proofs against the defensibility of the Justified True Belief’s status is at the very best a reinforcement of an already stated proof by Gettier against Justified True Belief and which does not seem to offer us anything new that can destroy or repudiate the viability of Justified True Belief as being equivalent to knowledge.
Nevertheless, I wish to provide an illustration to show that there could be cases where ‘Justified True Belief’ can pass for knowledge and to comment on what this must mean for our understanding of the nature of knowledge. Suppose that X believes that
(f) Y is not in school today
Because X saw Y on his way to Lagos, and he was duly instructed by Y to inform the Head of Department that he would not be in school that day. Assuming that Y later informed X from Lagos on the phone that he is now in Lagos, confirming his safe arrival there. On this ground, X says that he believes with justification that Y is not in school. X then deduces the following from (f), (based on the evidence at his disposal about the absence of Y)
(g) Y’s lecture cannot hold.
However, unknown to X, Y was flown back from Lagos to the school with an urgent message for the Vice-Chancellor that same day and thus was able to hold his lecture after seeing the Vice-Chancellor. Here, we are permitted to say now both that X does not know that (g) and that it is not the case that X did not have sufficient evidence for (g). However, on the basis of the evidence available to him at the time when he inferred (g), one is inclined to argue that the circumstances of his claim entitled X to the knowledge that (g). It could be suggested that X’s knowledge that (g) was circumstantial; his belief was true at a particular time, say t1 because not until the return of Y, proposition (g) was true. Hence we argue that X knows that (g) up till time t2. With the arrival of Y, it could again be said that at time t2, X does not know that (g). This does not mean that X both knows and does not know because there is a time lag between the two knowledge claims being assessed.
Although it is arguable that the propositional sense of knowledge has the features of incorrigibility, infallibility and indubitability, yet at t1, X’s belief that (g) was true but was later at t2, overtaken by events. There is a sense in which a person can argue that the process of knowing is evolutionary and so what people know is in the process of evolution.
Something like the foregoing seems to have informed Karl Popper in his Objective Knowledge where he maintains that knowledge is never static (Popper, 1972, pp 31-49). Some would also say that the ‘truth’ of what is claimed to be known is relative (Hamlyn, 1986, p. 8). Truth may be relativised, considering the time and the circumstances under which a person claims to know something. The example of X in the foregoing helps to bring this point out. It cannot be denied to X that he knows that (f) since he has sufficient evidence for (f) and the evidence was true at t1. But events and circumstances proved (f) false. Thus, the knowledge of X that (f) was relative – relative to time and circumstances. The lesson to be drawn from this is that because the truth of beliefs cannot be assessed in absolute terms, claims to knowledge too cannot be assessed in absolute terms. The factor of the defeasibility of truth – claims in terms of truth – falsity eventualities should therefore not play as decisive a role as Gettier attributes to eventual falsity conceived in absolute terms, in our understanding of the general nature of knowledge. This may mean that although the criteria for knowledge are the same they apply flexibly with regard to time and circumstances, for example.
At this juncture, an illusion to Prof. Wiredu’s conception of truth as opinion and his critics’ argument may be apposite to our discussion.
Prof. Wiredu has attempted to refute any distinction between truth and opinion and has concluded that “there is nothing called truth as distinct from opinion" (Wiredu, 1980, p. 111). According to him, the objectivist view of truth as against opinion is that truth is ‘timeless’ and ‘eternal’. The objectivists were led to this conclusion by common-sense experience which has tempted them to hold some opinion as true at a particular time but which was later discovered to be false. For example, the objectivists have made a distinction between opinion and truth from the following observation:
at time t1, it can be raining, but may no longer be true at time t2 that it is raining as the rain may have stopped by then. (Wiredu, 1980, p. 114).
Wiredu disagrees with the objectivists’ theory of truth because as he put it, “if truth is categorically different from opinion, then, truth is, as a matter of logical principle, unknowable Wiredu (p. 114). Any given claim to truth is merely an opinion advanced from some specific point of view and categorically different from truth. Hence, knowledge of truth “as distinct from opinion is a self-contradictory notion" (Wiredu, 115). Wiredu’s main contention against the objectivists’ conception of truth is that truth should be understood as a cognitive claim from some ‘point of view’, which he identifies with opinion, and on this basis draws the lesson that truth is nothing but an opinion held from some “point of view", nothing but “opinion" (Bodunrin, 1985, p.12).
In criticizing Wiredu, Prof. J. Omoregbe has argued that “if as Wiredu contends, truth is nothing but opinion, then opinion would loose its distinctive characteristic which it has only if and when it is contrasted with objectivity. This applies to truth (objectivity) and opinion (subjectivity)" (Wiredu, p. 115).
This view of Omoregbe needs some modification. He seems to misconstrue Wiredu whose suggestion is that we cannot possibly separate a person from his opinion (Wiredu, p. 175), such that when it is said that a person has spoken the truth, the truth has come forth from his own subjective thought vehicle. When a person makes a truth claim, we should respect his personal autonomy, and consider him as being entitled to his own opinion as a rational individual.
Prof. Wiredu, in one of his replies to his critics distinguishes between the weak sense and the strong sense of opinion. According to him: “a matter of opinion … is a matter with regard to which criteria are unclear or even possibly non-existent or the evidence is scanty and there is, consequently, doubt and uncertainty (this is the weaker sense of opinion which is distinguishable from the proposition that two plus two equals four which is still an opinion, but) an outcome of a mental effort, the result of the mind’s activity of systematization and validation". In fact, to make it more forceful and emphatic, Wiredu concurs that (Wiredu, p. 175)
if an opinion can ever be conceived of as a thought advanced with full assurance with full assurance from some point of view, then there is nothing amiss philosophically in classing scientific and mathematical propositions alongside others as opinion. (Wiredu, p. 176).
If we want to be sincere, we cannot jettison the assertion that all of our thoughts, expressed in the propositions of mathematics and formal logic were formerly part of our individual subjective mind. A person cannot produce a thought outside his own thought system. It is like a child who denies the womb as his first environment before birth; such a child is like somebody destroying the foundation of an edifice, thinking that the whole castle will stand. Thus, Wiredu’s submission that “opinion is normally the outcome of rational inquiry and that the formation of opinion is governed by rules of evidence and of formal logic" seems to have some force of persuasion.
The little problem with Wiredu’s theory of truth is the difficulty in understanding the meaning of “point of view", with which he seems not to draw any distinction between ‘objectivity’ and ‘subjectivity’. It should be understood that the whole conception of truth is anchored on the cleavage between what is true and what is mere opinion. While the former can never be false without contradiction, the latter can be false without contradiction. Yet, it can still be conceded to Wiredu that we cannot arrive at the ‘objective’ (truth) without passing through the ‘subjective’ (opinion). It can be advised that one should be careful not to be overly critical of the possibility of the equation of ‘opinion’ to ‘knowledge’ as the case is with the critics of Prof. Wiredu’s conception of truth and opinion. Mutatis mutandis, the same appeal goes or justified True Belief and Knowledge. Justified True Belief can be knowledge given possible improved reformulation of justified True Belief.
We have seen that Prof. Wiredu has put truth in the subjective realm (cognitive point of view). He is claiming that whatever is considered as true is a judgement from the subjective realm. I have also tried to relativise truth by arguing that truth depends on time and circumstance. ‘Truth’ in both views is relative. For me, it is relative to time and circumstance, but for Wiredu, it is relative to an individual’s cognitive subjective state – what he calls his “point of view".
In conclusion, I like to submit that extreme skepticism can be dangerous. This danger is so great that rather than highlighting the cases where Justified True Belief appears not to be knowledge, we would do better to suggest alternative illustration that would make it easier to see Justified True Belief as knowledge with an improved conception on the nature and conditions of truth.
1. Gettier, ‘Is Justified True Belief Knowledge?’ In Analysis, Vol. 23, June 1963 (Copyright), Basil Blackwell, pp. 121ff: Reprinted in A. P. Griffiths, Knowledge And Belief, Oxford University Press, 1968.
1. B. Sosa, ‘The Analysis of Knowledge that p’, reprinted from Analysis, Vol. 25, No. 1, New Series, 1964, p. 1.
1. R. K. Shope, The Analysis of Knowing, Princeton University Press, New Jersey, 1983, p. 21.
1. K. Popper, Objective Knowledge, (An Evolutionary Approach), Clarendon Press, London, 1972, pp 31 – 49.
1. D. W. Hamlyn, The Theory of Knowledge, Macmillan, London, 1986, footnote 1, p. 8.
1. Wiredu ‘Truth as Opinion’ in his book Philosophy and an African Culture, Cambridge University Press, 1980. p. 111
1. P. O. Bodunrin, (ed. ), Philosophy In Africa: Trends And Perspectives, O. A. U. Press, Ife, 1985, p. 12.
Z. B. Ogundare
University of Ado-Ekiti, Nigeria
In partnership with
London Academy for Higher Education.