Knowledge, Fake Barns, and Sherlock Holmes

It’s one of life’s little ironies that every person on earth possesses knowledge and makes claims to knowledge while the criteria of knowledge remain so difficult to identify. Analytic philosophy has offered many claims on what are the conditions for knowledge: the familiar justified true belief formula; warranted true belief; William Alston’s epistemic desiderata approach; reliabilism; etc. Not one of these, or any other formula, is without controversy.

Beginning with Plato, something like the justified-true-belief paradigm reigned supreme until 1963 when an obscure philosopher named Edmund Gettier under pressure to publish something to satisfy administrative requirements at Wayne State wrote a three-page article, entitled “Is Justified True Belief Knowledge?” The result was arguably the most significant development in epistemology in the twentieth century (Gettier hasn’t published anything since). Gettier’s idea was simply that there are scenarios in which our justified true beliefs do not equal knowledge. Though he does not offer this example (it actually comes from Bertrand Russell), imagine you look up at a clock on the wall for the time, but, unknown to you, it stopped exactly twelve hours prior to your looking at it. It yields the correct time and you form a true belief which is certainly justified yet it does not count as knowledge because it’s based on a mere coincidence. It could’ve just as well happened that the clock read the wrong time. The conditions of knowledge must be repeatable, normative and non-coincidental. This example and others like it became known as Gettier examples.

Here’s another popular Gettier example: imagine you are driving through the countryside of Wisconsin and you see a picturesque pastoral scene – verdent fields, rolling hills, wild flowers, dairy cows and everything else that makes for the quintessential farm scene, including one of those hip-roofed, red barns. Impressed with such a serene landscape you drive further and see yet another ideal farm with a red barn. As you drive, you see barn after barn – or so you think. Unknown to you, in an effort to impress passers-by such as yourself, rural Wisconsinites have randomly erected barn facades which as far as you can tell are indistinguishable from the real barns. In a rapturous moment of wonderment, you exclaim, “What a fine barn!” It so happens that the barn you speak of is one of the real barns instead of one of the facades, and your belief is justified and true but it is accidental since you can’t distinguish the real barns from the facades.

What do Gettier examples tell us about knowledge? Some suggest that we just have to add a fourth criteria: ungettierized justified true belief. But of course this sets the bar rather high and concedes quite a bit to skepticism. In the Wisconsin barn example, should one suspend all belief about the barns until they’ve all been checked for authenticity? Gettier examples abound in life and we can’t check them all. So what should we do? However one answers what is necessary and sufficient for knowledge, the answer must allow that knowledge is accessible by all people and not just a select group of intellectuals. In my view, it will be broad enough to include animal knowledge as well.

But the complexity of defining knowledge raises a more interesting question than “what are the necessary and sufficient conditions of knowledge?” The question is: can we possess knowledge without knowing the conditions of knowledge? How one answers this question depends largely on what one believes about the nature of knowledge. Either knowledge requires an intentional synthetic creation that comes from amassing the right evidence with an awareness of how the evidence relates to one’s belief, or we are equipped with true-belief producing faculties that yield knowledge in a naturalized way. I think that possessing knowledge will generally look something like the latter explanation. But demonstrating that we possess knowledge will look something like the former explanation. It is the difference between a personalized epistemology and an impersonalized normative one. It is not necessary to know that one knows the conditions for knowledge in order to know anything in the first place. Otherwise, no one knows anything until we have figured out the intricacies involved in defining knowledge. There’s the further problem of infinite regress. If we require that before we can know anything we must know the conditions of knowledge this will not be enough to have knowledge in a particular case. For this requirement also means that one must know that one knows the conditions for knowledge. But we can’t stop here. One must know that one knows that one knows the conditions of knowledge, and so on ad infinitum. But clearly this is absurd as well as impossible. Knowing that one knows is different than knowing how one knows. Knowledge does not require exhaustive comprehension of the grounds for one’s knowledge, otherwise hardly anyone would ever know anything.

Perhaps another example is in order: Sherlock Holmes brilliantly solves the case and catches the criminal in The Sign of Four–Mr. Jonathan Small. But eventually he gets up in years and lays aside his deerstalker cap and his magnifying glass and is no longer investigating crime. One day, while reminiscing about their days of fighting crime together, Watson asks Holmes about the evidence in the “strange story of Jonathan Small.” But for the life of him, Holmes cannot remember what it was that proved he did it. Must Holmes give up his belief that Small did it unless he can remember the evidence? Or does he still have knowledge whether he can remember the evidence or not? If knowledge requires that he continually keep the evidence in mind in order to hold onto his knowledge then he can no longer be said to know that Small was the killer. But it would be absurd for Watson to tell Holmes that since he’s forgotten the evidence he can’t say that he knows that Small did it. Obviously, Holmes can know without remembering how he knows. So when Watson asks how it was that they ever solved the case, Holmes can simply respond, “Elementary, my dear Watson, elementary.” (Of course if he really wanted to remind himself of how he solved it he could always google it.)

A Quick Word on Presuppositional Apologetics 

In case you’re interested in how all of this relates to pressupositional apologetics, here’s a quick word. Essential to the genius of presuppositionalism is that apart from the truth of the Word of God unbelievers could not know anything. But obviously they know all kinds of things. But if unbelievers must have knowledge of the grounds of their knowledge then the presuppositional claim would mean that they can’t know anything unless they first know that they know only because of the truth of the Christian faith. But they can have knowledge only because they hold beliefs and reason in ways that make sense only because of the truth of the Christian faith which they suppress and deny. It is only through conversion and a conscious, willfull submission to Christ that believers can both know and know the grounds for their knowledge and thus be in a position to demonstrate how they know what they know.

        -John Fraiser