|
|||
 |
|||
|
|
|
|
|
|||
 |
|||
|
|
|
|
A Priori Knowledge One example of this is the concept of marginal utility. The idea is that the each additional unit of a resource is less valuable than the first unit. The reasoning is that we have many needs/desires. When we gain a unit of the resource, we can use it to satisfy our most urgent needs. So if you are penniless and given some money, you can use that money on the most pressing needs you face, such as finding food to eat so you don't die. Any additional unit of the resource, in this case money, will necessarily be used for less urgent needs or desires. And that means they will be of less value, since the needs they are satisfying are lesser. Why call this a priori knowledge? What purpose does it serve? The book explains that it is in contrast with empirically derived knowledge, such as seen in the physical sciences. The physical sciences look to the world to determine how things work. If you want to know what happens when you mix a set of chemicals, you have to perform the experiment to find out. You may have a theory for why it works, but you have to validate it with a test. And at anytime you might find some new evidence that overthrows your theory. Because of this possibility of a theory being overthrown, the physical sciences rest on a foundation of observation. Observation is king. In contrast to this view from the physical sciences, the foundational premises in economics appear to be different in kind. When an economist says that a person chooses between options based on which one will better satisfy his wants, this isn't some theory that has been tested in a lab many times and which we can now tentatively treat as a fact. We don't validate this premise through experimentation, but through logic and rational analysis. There is an important consequences. If these foundational premises were really experimentally derived, than when we combined them and deduced more complicated economic consequences, we would be obligated to test those through experimentation or observation as well. That would be the only way to validate the conclusions. But if our initial premises are seen as undeniably true, then the conclusions would be validated by checking that each logical step was valid. This works the same way as mathematics works. If we solve an arithmetic problem like (9+27)/6, we would confirm our results by checking the math. We wouldn't be forced to run an actual experiment in a laboratory to see that the results match. The foundation is viewed as so certain that only the process needs to be validated to validate the conclusion. I wholeheartedly agree with the methodological conclusions. Economics should be derived from foundational premises using a process of deduction. Conclusions are validated by validating the premises and deductions, not through experimentation. Experimentation would usually not work, as there are many factors involved and other factors that are unseen. For instance, when someone acts on their values, they act on the most important one, and we have no visibility in how that person may have prioritized the values that weren't acted upon. We don't even have visibility of what they are. While accepting the methodological conclusions, the distinctions and specific arguments made are open to criticism. One criticism possible is in the use of the term "a priori". It may be that the foundational premises are not the same as empirically derived facts, such as what the air pressure is at sea level. Maybe it is a different kind of knowledge, but what benefit is there in calling it "a priori"? Presumably the goal is to claim it is knowledge that is disconnected from any kind of observation of reality, thus making it immune to future observations or the need for additional testing. While this severing of ideas from reality may seem like a benefit, since it avoids the need for future testing, it's really a weakness. By denying it is based on knowledge gained through observation, it denies its roots. Consider the case of marginal utility. What kinds of things are taken from a wider body of knowledge? First, we recognize our own method of decision making. We recognize that we have many competing options, and that our resources are limited. We recognize that we act to satisfy the most urgent needs, and that our evaluation depends on the values created. We recognize other people also make choices, and we infer for a variety of reasons that their decision-making works the same way ours does. These premises are in fact conclusions, and conclusions based on a mountain of evidence and information. Rejecting that knowledge and claiming that the premises are based on nothing but deduction makes no sense, and loses sight of the actual foundations of the knowledge. It encourages you to ignore the reasons why you actually conclude something is true, and instead substitute a method that is vague in details. It has to be vague since it tries to make claims about the world while pretending to do it without knowledge of the world. Science needs to be grounded in reality. And that means induction. The physical sciences do this through a process of experimentation and observation. Economics also needs to be grounded. It doesn't need to be grounded in exactly the same way, though. In the physical science, it is common to start with a vague big picture and look to try to get a more refined understanding. This process requires a continuous comparison with reality as alternatives are tested or compared. Economics works in a different way. It starts with well-grounded premises, and can use deduction to derive more knowledge. Since it is deductive, the constant comparison with reality is unnecessary. This is why economics needs to be deductive. The comparisons don't work well. Variables are not constant, many of them are not observable in every situation, and they may be difficult or impossible to isolate. In terms of methodology, the reliance on "a priori" knowledge is unnecessary. The validity of the methodology doesn't rely on it. Since the method used is deduction, the validity of the conclusions is tied to the validity of the premises. And because of the nature of the subject, the method used is appropriate. These are enough to justify the method in both senses of the word. What if the premises are flawed? Then the conclusions would be as well. But there's no escaping that. Unless the premises are "self-evident" in the sense of automatic and reliable, something we directly perceive instead of something we have to conclude, there is a possibility of a mistake. It could be that important distinctions are not recognized, or one of the premises is not as universal as it first seems. But these don't damage the credibility of the deductive methodology. And pretending that the premises are not rooted in observed facts only obscures their foundation. The book contrasts these obvious and irrefutable claims with empirical knowledge. It describes a view in the philosophy of science that claims that knowledge is never confirmed, only falsified. Given that view, there's no way of saying that some fact is obvious and irrefutable. The strength of an argument is irrelevant. So from this perspective, the premises that are the foundation of economic theory could not be viewed as anything but tentative, unsupported hypotheses. But in fact the economic premises are so well supported, not just by evidence buy by logic and theory, that they are clearly true. Given this, it would make sense to view this kind of knowledge as separate from the straight empirically derived knowledge. I think there are two problems there. First, that philosophical view of knowledge as never confirmed is wrong. The entire approach is wrong. It solves the question of induction by simply rejecting it and claiming the only real knowledge is deductive knowledge. So building a theory of a priori knowledge to counter this flawed view is unnecessary. The second problem is that it creates a false dichotomy. On one hand, you have unsupported hypotheses that are constantly tested through experimentation but never confirmed. On the other hand, you have a theory that is so well supported by logic, theory, and evidence that it is viewed as absolutely certain. But isn't there a continuum? A theory can be supported not only by experimentation, but through integration with other theories or facts. The idea of a scientific theory being groundless and only validated through direct experimentation is wrong. A theory can be supported by logical consistency, by deductive connections to other knowledge, and by a mountain of evidence. If a single experiment seems to contradict such a well-grounded theory, the problem would be expected to appear outside of that theory. Evolution is an example. It is supported by a mountain of fossil evidence. It is supported by the theory and evidence of genetics. It is supported by the abundance of life and all of the variants. The idea that of natural selection is in many ways obviously true, as less fit species will be less likely to survive and reproduce. It is a theory that is so well supported that even contradictory evidence, if found, would suggest some other explanation instead of falsifying the theory. Not all of our knowledge is that well supported, but it also isn't so weakly supported that it can be treated as completely unsupported. There are many different degrees. So the biggest problem with the a priori "solution" is that it escapes the problem of explaining how these ideas can be considered irrefutable by simply asserting it. That may be a hard problem to deal with, but this approach is a dead end. It removes the question from inquiry, where it won't be examined and considered. And it offers nothing in return, except the quiet illusion that comes from pretending a problem doesn't exist.
Discuss this Article (2 messages) |
 |
 |
|
 |
|
