Rebirth of Reason


Understanding Empiricism
by Joseph Rowlands

Empiricism, when defined, sounds somewhat reasonable. It suggests all of our knowledge comes from perceptual data. Seemingly, this is a rejection of a priori knowledge, or divine revelation, or any other form of knowledge disconnected from perception.   And if that's all it was, it would be perfectly reasonable.


It goes further, though. It excludes things like abstract knowledge and deductive knowledge. If you try to go from concretes to generalized truths, you are adding something beyond what you have perceived. You are making an assumption about identity, and not one that can be proven through perceptual data.


Consider the Law of Identity and how we support it. If we find examples where we don't grasp how it works, such as with Quantum Physics, do we conclude that there is no such things as identity/causality, or do we conclude that there is and we just don't have enough data? Some might treat identity as an axiomatic starting point, or an assumption. Others might say that it is the best theory to explain the evidence (all evidence!), and in that sense it can be thought of as a product of induction. But the empiricist would say that we can't perceive identity. They might conclude that identity is not true because it is apparently contradicted by evidence.


The empiricist approach to science treats all "knowledge" as simply a set of hypothesis. There is no justification for them. There is no way to say that the hypothesis is well-grounded. It is arbitrary, and treated indistinguishably from the arbitrary. Once you generate a hypothesis, you can begin to test it. Testing may prove the hypothesis wrong, but it never proves it right. The only real knowledge is perceptual data, and any attempt at explaining things is at best a hypothesis that hasn't been invalidated yet.


Consider something like the fact that an increase in minimum wage will increase unemployment, all things being equal. Empiricist economists would and do treat this as simply a hypothetical, to be tested by experiment or observation. If unemployment goes up, the hypothetical survives for now. If it ever goes down, the hypothetical is proven false.


A different economist would say that if unemployment went down, that it must be caused by another factor. The idea that minimum wage increases unemployment is not some arbitrary hypothetical. It is drawn from more fundamental facts that are well establish in terms of identity. We understand that when the price of a commodity goes up, the quantity demanded will go down (or possibly stay the same). We understand that this is a product of simpler truths. Given a set of preferences, a person optimizes his actions accordingly. If one of the options becomes costlier, it will rank lower in his preferences, not higher! He won't purchase more of it, and may very well purchase less of it.


The empiricist approach doesn't recognize identity or doesn't recognize our ability to identify. In either case, it rejects these deductions from general principles. The only way to validate an idea is to perceive it. They would ask how you can believe in this relationship between minimum wage and unemployment if you have never perceived it. You may perceive the data of each, but not the relationship itself. And if you can't perceive the relationship, it isn't grounded in perception.


If we take the vague definition of empiricism that all knowledge is ultimately derived from perception, then it seems reasonable. But if we take it how it is actually used and understood, the connection to perception is much narrower.


Ultimately the philosophical issue is how do we connect our abstractions to perceptual data. How do we generalize? How is induction used? How do we form concepts? Empiricism cannot find that connection, and so they dismiss or at least downplay the importance of abstractions. Perception is the only real source of knowledge, and abstractions are not perception. At best, they are arbitrary hypotheses that are consistent with perceptual data. But they are must always be suspect.


Without that connection, abstractions will always appear as arbitrary, groundless, and at best consistent with known data. But at any time, new evidence might show up. A classic example is that the sun rises every morning. But just because it always has, doesn't mean it always will. Tomorrow it might not. A generalization that it rises every day and so will rise again tomorrow is viewed as arbitrary. It may or may not. This is the problem of induction, but it is a reflection of the empiricist belief that abstractions are arbitrary and that only perception is real.

Sanctions: 17Sanctions: 17Sanctions: 17 Sanction this ArticleEditMark as your favorite article

Discuss this Article (0 messages)