By James Fox
Since its publication Gettier’s Is Justified True Belief Knowledge? has become the seminal work in modern epistemology. This paper challenges the very assumptions of Gettier’s counterexamples and is therefore a radical alternative to both the proponents, and critics, of Gettier. By showing how knowledge is found, not in mere words or statements, but within the fundamental beliefs of the speaker, I expose the way in which ambiguity in language can mislead us into rejecting the traditional definition of knowledge as Justified True Belief.
“What is truth? said jesting Pilate; and wouldn’t wait for an answer.”
— Francis Bacon
Pilate’s question has haunted humanity, and its answer I shall not presume to give; suffice to say that it will be an answer both beautifully simple and excruciatingly complex. However, as a preface to any study of Truth and knowledge, this question, simpler and less frequently asked, must be considered: Where is Truth? If we are to discover Truth, where should we look? There are many sources that purport to be true – spoken words, manuscripts and books, scientific theorems and even the deepest feelings of men’s hearts. In this essay I intend to address two main questions. The first is the relationship between a true statement or belief and Truth Itself. Is a statement True by its own virtue, or by what it causes or by what it reveals, for instance. Secondly, I intend to defend the view that Knowledge is Justified, True Belief, one which has only recently been challenged and yet is now almost entirely rejected. I shall not attempt to define knowledge, only to break it down into its three components.
So, to our business of finding Truth. I must make it clear that I am not attempting to investigate the nature of Truth itself, its many complexities and misattributions. As Beauty is commonly used to describe many things which are not truly beautiful, merely elegant, picturesque, alluring or some lesser aesthetic quality, it is quite possible that our notion of Truth would include those which, if a thorough consideration were applied, would need to be demoted. However, this is not my aim. As it is acceptable, for the purposes of buying an oil painting, for instance, to use a wide and admittedly imperfect notion of beauty, I hope that it will serve my purposes to do the same for truth, and that you, my reader, will forgive my failure to think more clearly and precisely. Truth then, for the purposes of a preface only, we may take as corresponding to reality, as it is described by science and believed by men of common sense. Let us for now put metaphysical speculations aside and concentrate on the lives of men, leaving the Forms to their business.
“Lying lips are an abomination unto the Lord,” we are told by Solomon in the Book of Proverbs. Indeed, for most of us our first encounter with the truth is as a child, when we are told to tell the truth, and not to lie. Is this then where the philosopher should search for Truth? When we look at those who do not deceive, we see that what they are saying corresponds with what they believe. However, this is not Truth in itself. For one could imagine a lie, by coincidence, being factually accurate and therefore true. On the other hand, the thoughts of a liar, and his beliefs and opinions, may be true, while what he says is not. It seems that in this area, it is better to consider the issue of sincerity. A sincere man will say what he thinks, which may be true or not. To speak the truth, however, we must be more than sincere, we must have some knowledge of it, and our words must express this.
Let us then return to statements which are made, being able as we now are to ignore those which are insincere, and to concentrate our search for truth in those sayings and writings which accurately reveal the author’s beliefs. I believe that there are two ways to search for truth in a statement of language, and that it is the understanding of these which is the key to this question. However, with spoken expressions, we almost always have a fair amount of knowledge concerning the speaker – even if they are a stranger to us we know what they are wearing, their height, accent, and can usually gain a little of the motivation for what they say by considering the circumstances. Therefore our knowledge of the speaker influences our analysis of the statement itself. However, when reading, particularly when reading classics or other works where we are far removed by time and place from the author, we are forced to consider to a greater extent the words on their own merits. Here, then, is where we see the contrast most starkly.
A clear example of this is with the Holy Scriptures. For many of the books of the Bible, the authorship is uncertain, and what is known about the authors is severely limited. In addition, the writers lived in both places and times far removed from our own. Despite this, vast numbers of people, both inside and outside of organised religious traditions, have scoured its pages in a search for Truth. It is also certain that there is considerable disagreement as to how this resource is to be used. Broadly speaking, there are two main methods, although most readers will use both techniques and will not draw sharp boundaries.
There is one group who read books and apply them to their own lives, seeking to find truth which may not have been explicitly intended. This does not mean that each person’s interpretation is equally true, but it does mean that it is possible for there to be truth in a sentence which was not explicitly proposed by the author. This can be seen in Shakespeare’s use of irony, where what is said by the characters delivers an apparently unintended truth to the audience. Historians and others of a similar mind do not attempt this, but seek to discover what was intended by the author. In order to do this it is important to know the context, to make comparisons with other contemporary writings and to, as far as one can, enter the mindset of the author. As already mentioned, many will use a combination of these practices according to their objectives and their situation.
These two methods or comprehensions can be applied to any work. When applying oneself to Machiavelli’s Discourses, for example, one man may find it useful to read a biography of the Florentine, to compare his work to those of the period, to examine his letters and habits, so as to gain the greatest insight into the intended argument. When asked how useful and accurate the composition before him is, he will examine the arguments that the author made, and, if a fault can be found in them, will conclude that the book, while intriguing, is flawed, or relevant only to its particular period. Another man, however, may do none of this. Instead, he says to himself, “As this work has been preserved for many years, there must be something in it. Let’s give M the benefit of the doubt, and construct the best possible argument from what is written that we may use this to aid our understanding of politics.” When asked how useful the book is, the second man will recite what he has been able to glean from it, and if the argument he has produced is stronger, and more useful than that particularly intended by the author, he will not mind one jot. We cannot say that one of these modes of thought is superior to the other; they have entirely different aims. The first man is trying to understand Machiavelli’s psychology, his motivations, and his beliefs as a man and as a politician. The second may as well not know who has written the book at all, he is trying to gain knowledge of the subject for himself, and is more interested in the politics and philosophy than history and psychology. Indeed, it can be interesting to have both ways of understanding a book in mind when reading it, so one can gain knowledge of all of the above fields, and understand, if a stronger argument can be made, why the original writer did not manage to express this.
This distinction, however, gives us considerable insight into the relationship between spoken and written statements and Truth. Indeed, it shows that a statement’s truth depends on how it is read, and indeed, it is clear that, depending on the context in which it is spoken, and the reception that it receives, the same statement can be seen as being both true and false. For example, the statement “The princes in the tower” would seem true, if I were asked to name victims of Richard III. However, it would seem manifestly false if the question were “Who commanded the Armada?” From this, it can be seen that a statement is not simply true or false in itself; the truth we apply to it depends on the beliefs which underlie our utterances. Is this really that surprising? For, after all what distinguishes a statement from a clanging cymbal, or an inscription from mere gashes in a stone? Solely the fact that it can be interpreted. We cannot apply truth to mere noise, or knowledge to mere ink on parchment. A statement, as a string of words written or uttered, is neither true nor false.
We will find if we are being precise, it is wrong to call statements made using language true or false at all. The statement is a mere vehicle, an imperfect but usually reliable way of transmitting information about our beliefs. The statement itself does not correspond to reality. Instead, it is our belief that a state of affairs is so that either corresponds to reality or does not. Thus, only a belief, and not a written sentence or a remark which is spoken, can be said to be true or false. Thus, it is in our beliefs that we must look for truth, and be wary of falsehood.
This may seem like the mere splitting of philosophical hairs. After all, we do not know what another believes except by interpreting their statements and actions. If statements are all we can possibly know, then the above distinction, even if it is accurate, is not worth the paper upon which it is inscribed. However, I will attempt to show that this distinction is vitally important when we consider the question of knowledge, which is surely of great importance to philosophy. In particular, I wish to show that if the above is considered, the Examples given by Mr. Gettier, when he argues that Justified True Belief is not knowledge, do not cause a problem to the definition, and that, as this is the main stumbling block which has been encountered, the tripartite definition of knowledge is as far as we know correct.
The first example I shall consider is one of Gettier’s own1. In this situation, Smith and Jones are both applying for a job. Smith is justified in believing that Jones has 10 coins in his pocket. Smith is also justified in believing that Jones will get the job. The proposition The man who will get the job has 10 coins in his pocket seems therefore to be an accurate portrayal of Smith’s justified belief. However, against all expectations, Smith is offered the job. Unknown to him, he also has 10 coins in his pocket. Initially, it looks as if his proposition is true, and as it is also believed with justification, we are tempted to conclude that this is indeed a Justified True Belief. As, however, we would not like to say that Smith knew this; hence this definition of knowledge is inadequate.
In the storm that this example caused when first released, philosophers grappled with what seemed to be the most ambiguous part of the definition, namely the justification clause. However it is clear that in this situation S is justified in believing P. Now, in the light of our earlier discussion, let us take another look at the proposition. It is possible, by modelling the world as such, that “who will get the job has 10 coins in his pocket” can be entirely unambiguous – there is only one job, which only one man can get, we agree on a definition of coins and pockets, and of ownership of these. However, The man must be ambiguous, as in order for the example to work there must be two men, Smith and Jones. P is therefore true/false, depending on the identity of the man, which, I think we can safely conclude, refers to one, actual, concrete man rather than an abstract concept of any man. If we were determined to discover whether P is true or false, we would have to seek out Smith and ask him, which man did he mean? It seems clear that in this situation he would reply that, as he believed that Jones would get the job, the man refers to Jones. However, by substituting these now-identical terms, we have made the proposition false; Jones did not get the job. Smith’s belief, although justified, is now not true. It seems sensible that by clarifying the ambiguity in the language, we have not changed Smith’s belief about the state of affairs, so we must conclude that his belief was false all along, and he therefore never had knowledge of P.
Here the sceptic could object that Smith was indeed talking about a man in the abstract, not a concrete figure. For instance, he could, it must be admitted, simply believe that a man with 10 coins in his pocket will get the job, without identifying this in any way with Jones. However, if this is the case, how can it be that his belief is justified? He is justified in believing that Jones has 10 coins, but we have not seen Smith rifling through the pockets of an abstract form of a man!
It may however be accepted that this example fails, while still holding that some Gettier examples are conclusive. The original two situations are now rarely used to illustrate the criticism of Justified True Belief; they are usually replaced by more recent examples. I shall now take a more complex and persuasive example and show that the same method can be used to discount the criticism of the tripartite definition, before looking, in more general terms, at the failings of all such examples.
Imagine that I am sitting in my study, and I hear a siren from the direction of Broad Street. I know that, when one hears a siren of this description, it is produced by a police car, and therefore believe, with considerable justification, that there is a police car on Broad Street. However, I am mistaken; there is instead a prankster who is activating a siren in an attempt to cause a nuisance. There is in fact a police car on the street, driving quietly so as to intercept this disturber of the peace. Here the statement ‘There is a police car driving down Broad Street’ seems to be doubtless true. There is even a causal connection between the justification and the truth of the proposition. Surely here we have a Justified True Belief which is nevertheless not knowledge?2
Let us look at what exactly it is I believe when I make the statement, ‘there is a police car driving down Broad Street.’ I believe that there is an object in that location, that it is moving, that it is a car, and more specifically a police car, that inside are police officers, and that it is sounding its siren, and many other things about it. I must believe that it is sounding its siren, as if not, my belief cannot be said to be justified. Hearing a siren is not a justification for believing that there is a car driving silently nearby. I can therefore make many propositions about the state of affairs, all of which I believe, some of which are true. It is not true, however, that the police car has its siren on. In this situation, the term police car is ambiguous, it does not give me enough information to make a decision as to the truth or otherwise of my belief. When clarified, my belief, that there is an object in that location with the attributes listed above, is not true.
The obvious reply to this is that although my belief is a bundle of propositions, not all of them have to be true in order for the belief to count as Justified and True. However, let us consider what would occur if a different proposition was altered. It could be the case that I am correct in believing that there was a moving object carrying police officers in that location, sounding a siren. However, instead of being a car, the policemen in question are riding a pink rhinoceros down Broad Street, which has been conveniently equipped with a siren. Here we would not say that I am correct in believing that a police car was driving down Broad Street, with a siren and all of the other attributes above. By altering any one of the propositions which make up my belief, the belief is fundamentally altered. Thus, in order for our belief to be true, all of the propositions need to be correct, and any falsehood will remove the status of knowledge.
In view of this, we need to reconsider how we think about knowledge. Although the tripartite definition is usually referred to, saying that knowledge is a justified true belief, this is used to mean a proposition which is believed and also true and also justified, hence the term ‘propositional knowledge’. This is made explicitly clear at the start of Mr. Gettier’s paper, for instance. However, we have seen that beliefs are not single propositions, but whole bundles of them, which we could not hold independently, but only together in relation with one another. For propositions do not exist in the mind; only beliefs exist in the mind. It is the ambiguity of spoken and written language, and the desire for brevity at the expense of clarity, which leads us to think that a single proposition can be in itself a belief. Knowledge, being a relationship between internal beliefs and the external state of the world, deals in beliefs as they are, rather than as they are expressed.
It is important to note that this distinction in no way shows that Justified True Belief is knowledge. It shows, however, that the Gettier counterexamples do in no way argue that it is not knowledge, and so for the time being we should accept the tripartite definition. There is, I put to you, in every ‘Gettier case’ at least one term which is ambiguous as to its reference. This leads to the sentence being justified on one reading of it, and true only on another reading of it, while that word must refer to only one object in reality. Now if more evidence comes to light that this is not how we should define knowledge, this must be addressed and the definition changed. However, for the moment there can be no objection to defining knowledge as Justified, True Belief.
1 – E. Gettier, 1963
2 – For this example I am indebted to Peter Millican’s excellent lecture series on philosophy in Michaelmas Term 2006.
James Fox (’11) is a Philosophy and Politics major at University of Oxford.
Art courtesy of hussainking