In a recent discussion with one of the commenters over at M and M’s blog (see the interchange between myself and someone using the nickname “Heraclides”) it occurred to me yet again that there are people – especially on the internet – who frequently wander into arguments about what are essentially subjects in philosophy, who clearly don’t have a background in philosophy, who appear not to have done much (or any) reading in the area they are arguing about, who are at times not really familiar with some of the basic terminology involved (even though they are using it), and there’s nothing terrible about any of this so far – but then your realise that they are talking as though they are absolutely certain that they are experts in the field. You offer a little advice, but you are told by this obvious newcomer that you couldn’t possibly know what you’re talking about.
Take my recent encounter. I said that scientists treat theories as provisional, but they do not treat knowledge as provisional. Knowledge is, after all, warranted true belief, so a scientist only knows something if he has become convinced that it is true. The reply that I was promptly given was “Theories *are* knowledge 😉 This suggests to me that you don’t understand what a theory really is.” Oh, and as for the fact that knowledge is warranted true belief, this is what my zealous fellow blog visitor had to say: “Only a religious person would write “knowledge is warranted true belief”. This both shows that you don’t understand science (and thereby aren’t in a position to criticise it) and that you don’t understand the failing of insisting something is “true belief” either (it’s blind to any revision or new information).”
Rather than simply get further frustrated at the bleak intellectual scene that one often finds in the comments section at blogs out there (as illustrated by the above encounter), I have decided to put a little more energy into becoming part of the solution. I’m adding a new category to my blog. The category is called “nuts and bolts.” In this new category, I’ll add posts that spell out basic terms and concepts used in the various subject areas in philosophy. You might think this is a bit redundant. After all, there are plenty of online dictionaries and encyclopedias out there. And you’re right, there are. But the way I see it, the more good basic information is out there, the more likely somebody will be to stumble upon it. So here it is, the very first post in the nuts and bolts category.
What is knowledge?
What does it mean to say that you know something? The first thing about knowledge – the stuff that I know – is that it consists of belief. I have a lot of beliefs. Some of them count as knowledge, and some of them don’t. That’s why it makes sense, when someone’s not really sure about a claim I’ve just made, for them to say “Glenn, you might think that’s so, but you don’t know it to be so.” Likewise, if someone asks me “where’s my wallet,” and I have absolutely no idea where it is – no beliefs whatsoever about its location,” I’ll say “I don’t know.” You can’t know X if you don’t have the belief that we’re calling X.
So far, knowledge = belief
But knowledge is more than just a belief, right? Otherwise, a person who believed in the flying spaghetti monster (OK, so no actual such people exist, but just hang with me on this one) could truthfully say “I know that the flying spaghetti monster is real. I believe that he is real, therefore I know that he’s real.” Take another example: There’s a jar full of jellybeans on my table, and I ask you how many there are. You decide (somehow) that there are 154 jellybeans in the jar. You tell me “I believe that there are 154 jellybeans in that jar.” next, I pour out all of the jellybeans onto the table and we count them. There are 135. If knowledge was just belief, then you actually knew that there were more jellybeans in the jar than there actually were! That’s obviously not right, so we need something extra to add to the definition before it’s really a definition of knowledge. Knowledge requires that the belief actually be correct. When someone says to me “I know what your name is,” their claim can only be true if in fact they know that my name is Glenn, because saying that my name is something else would not be true.
So far, knowledge = true belief
But is this enough? Clearly not. What if, for example, someone had guessed wildly that there were actually 135 jellybeans in my jar? What if someone had just randomly pulled a name out of a hat when trying to determine what my name is, and amazingly, the name “Glenn” was the one they drew? Obviously neither of these two examples would be examples of genuine knowledge, it would just be good luck in both cases. The belief would be accidentally true. Thus far the facts stated are absolutely uncontroversial. All agree that any plausible definition of knowledge must include the stipulation that knowledge must be a belief and it must be true. But something extra is needed, and it is the extra ingredient that has generated the bulk of the discussion in the literature.
Is knowledge justified true belief?
The first candidate for this missing ingredient was justification. The idea of justification is much like that of a legal definition – to justify your actions is to show that you did nothing wrong. Likewise, to justify holding your beliefs is to show that you did nothing wrong in forming your beliefs. You’ve got to be able to show that you had good grounds for holding the belief. Here’s an example that I put together: You’re driving in your car, and you’re traveling at 50 kilometres per hour. You look down at your speedometer, which indicates that you’re traveling at 50 kilometres per hour. You therefore form the belief that you’re traveling at 50 kilometres. Here we’ve got three ingredients. Firstly it’s a belief. Secondly it’s true (you really are traveling at 50 kilometres per hour), thirdly, you’re justified in holding the belief – you checked with a speedometer, and under the circumstances of driving, this is how anyone could expect to find out how fast they were traveling. If knowledge is justified true belief, then the driver of this car now knows that he is traveling at 50 kilometres per hour (give or take a little).
For a while, and to most in the field of epistemology (the study of knowledge and belief), this seemed correct. This bubble was burst in 1963 in a short article of only just a little over two pages, titled “Is Justified True Belief Knowledge?” Without using the specific examples that Gettier used (I for one find them unnecessarily complicated), the heart of his paper is that some beliefs could be justified and yet still only accidentally true. Think again of the driving example I used earlier. I deliberately left something out. Let’s review the facts again: You’re driving in your car and in fact you are traveling at 50 kilometres per hour (so it’s true that you’re traveling at 50 kilometres per hour), you’ve just checked your speedometer, which indicates that you’re traveling at 50 kilometres per hour (so it’s justified to believe that this is your actual speed), and so you form the belief that you’re traveling at 50 kilometres per hour. So you’ve got a justified true belief. Now here’s what I didn’t tell you before: Your speedometer is broken, and it is stuck on 50. No matter how fast you had actually been traveling, your speedometer would have indicated that you were traveling 50 kilometers per hour. The observer who is aware of this can now say “he didn’t really know that he was traveling at 50 kilometres per hour after all. He was just lucky!”
Once we realise how “Gettier problems” work, it becomes clear that there could be thousands, perhaps millions of them. If I become blind and mentally unwell, and I hallucinate that there is a tree five metres in front of me, then I am not doing anything wrong when I form that belief, so it is a justified belief. And if it coincidentally turns out that there really is a tree five metres in front of me, then the belief is also true, but it is not knowledge.
Warranted true belief
Obviously the fact that a belief is justified isn’t quite enough to make a true belief into knowledge. Think again of that car speedometer. Had it been working correctly, then you would have known that you were traveling at 50. As it is, you were just lucky. The idea of justification needed to be replaced to take that difference into account. That replacement is what epistemologists now refer to as warrant. Warrant takes into account Gettier examples and avoids them. Warrant is like justification but more carefully qualified. It takes into account our environment, and requires that it be congenial to passing on reliable information (so it rules out things like faulty speedometers), and it takes into account the proper functioning of our belief forming faculties in a truth aimed way (so it rules out things like hallucinations). Technically stated, and after carefully exploring other options and rejecting them, epistemologist Alvin Plantinga summed up:
[T]he best way to construe warrant is in terms of proper function: a belief has warrant, for a person, if it is produced by her cognitive faculties functioning properly in a congenial epistemic environment according to a design plan successfully aimed at the production of true or verisimilitudinous belief.
Alvin Plantinga, Warrant and Proper Function, 278.
Knowledge, therefore, is now widely (and I think, quite correctly) defined as “Warranted true belief.” It is a belief that you actually hold which is true and which is formed in the right way: By belief forming structures that are working correctly in a truth-aimed way in an environment that is conducive to providing those faculties with reliable information.
For further reading:
Richard Foley, “Conceptual Diversity in Epistemology,” Paul K. Moser (ed), Oxford Handbook of Epistemology (Oxford: Oxford University Press, 2002), 177-203.
Peter Klein, “Knowledge, Concept of,” Edward Craig (ed), Routledge Encyclopedia of Philosophy volume 5
Jonathan Kvanvig (ed), Warrant in Contemporary Epistemology (Rowman and Littlefield, 1996).
Alvin Plantinga, Warrant and Proper Function (New York: Oxford University Press, 1993).