© Cary Cook 2003
All attempts at defining knowledge fall into either circularity or tautology . e.g. Newell defined knowledge as cognitive patterns that map to external reality. (Physical Symbol Systems. Cognitive Science, 4, 135-183, 1980) But what does cognitive mean? Recognizing stuff. What's the difference between recognizing and knowing? Well, actually recognizing is a subset of knowing. So Newell is defining knowledge as a part of knowlwdge - which makes his definition circular. Here are some more:
Knowledge is justified true belief.
What does justified mean? Known! Another circular definition!
But let's say you don't buy that. You might say justification is evidence. Ok, a proposition can be justified by evidence. It can also be justified by deductive reasoning, inductive reasoning, common sense, authority, tarot cards, etc. Regardless of what a proposition is justified by, it is not known if that justification is legitimate unless it is known to be legitimate - as in Duhhh. Therefore 'justified' boils down to 'known'.
Knowledge is that by which a mind accepts the existence of truth and its own ability to distinguish it from non-truth.
That tells you what it does, but not what it is.
Knowledge is necessary assumption.
That's the best I can come up with, but it's still pretty much of a tautology.
Unfortunately these definitions offer no criteria by which to distinguish truth from non truth, or true knowledge from false knowledge - that which is thought to be known, but isn't. Those definitions that attempt to offer such criteria fail, because a definition of knowledge would have to exclude:
2) mental certainty without actual correctness
Defining knowledge as "actual correctness with mental certainty, both based on the same correct reasoning" would work until someone asks the criteria for actual correctness, mental certainty, and correct reasoning. The quest for knowable criteria is protracted ad infinitum.
Defining knowledge as "non-coincidental correspondence between what is true and what is thought to be true" would work until someone asks the criteria for non-coincidentality.
Knowledge cannot be defined, but that doesn't mean it can't be described. Knowledge is undefinable for several reasons:
1. Knowledge is an epistemically basic concept. i.e. It is intuitively understood by nearly everyone, yet it cannot be defined without using terms which are themselves defined by it.
2. We expect a definition to tell us not only what a thing is, but how to distinguish that thing from all that is not that thing. In the case of knowledge, I can tell you what it is (with circularity), but I can't tell you how to distinguish it from that which appears to be knowledge, but isn't.
2) operation: e.g. I know how to swim.
3) distinction: e.g. I know apples from non-apples.
As distinction, knowledge exists in three levels: immediate, categorical, & rational knowledge.
1. immediate knowledge: that by which a mind is certain that it exists, thinks, perceives, and emotes. By the narrowest definition of knowledge, immediate knowledge is the only true knowledge, in that it is necessarily correct, necessarily certain, and requires no underlying reasoning on which to be based.
2. categorical knowledge: that by which a mind distinguishes categories. Primitive categorical knowledge is arbitrary. It cannot be shown correct or incorrect until reasoning is applied. The first distinction recognized by an embryonic mind is generated by the first recognized sensory experience. It could be called a pre-verbal "What was that?" experience, when a sleeping mind is bumped into consciousness. The second sensory experience generates the first categorical distinction. The mind thinks:
"This experience is like the previous one"
"This experience is unlike the previous one."
3. rational knowledge: that by which a mind includes or excludes members from categories, orders those categories, and acknowledges true vs. false categorization. It is always expressed in propositions. Rational knowledge is self affirmative and undeniable. It doesn't prove that you know anything until you first accept it as proof. And you can't deny it without first affirming it. But it can't prove itself.
Rational knowledge contains two parts: logic, and some mental faculty by which foundational premises are formed. Note that logic itself is not a foundational premise. You don't know that logic is reliable because logic says so, because that is circular reasoning, which is a logical fallacy. There is necessarily some mental faculty apart from logic by which logic is recognized as reliable.
That faculty is usually called common sense. (If you don't like this term, substitute one you like.) Logic is the coherent system by which true & false are identified, and common sense is that by which a mind accepts the system itself, and those foundational premises with which to begin using it. Common sense is especially necessary for a mind to accept inductive logic. But common sense is the most fallible part of knowledge. When a logical conclusion conflicts with common sense, then common sense is necessarily wrong, either about that conclusion, or about the foundational premises which led to that conclusion.
Common sense is not the only method by which foundational premises are formed. Sometimes people willfully choose to abandon common sense in favor of emotion. This practice sometimes helps an individual to survive or be happier, but it is nevertheless epistemically illegitimate. It bases epistemology on the foundation of axiology, which is backwards, regardless of the desirability of its results.
The only empirical knowledge consists of sensory impressions. You know that you see, hear, feel, etc. something, but you don't know that what you see, hear, or feel is what you think it is until you have processed it thru reason.