Chapter 19 Chapter 2 Economics and Knowledge-8

Eight I must now turn to the question of the conditions of the specific assumptions, namely, the conditions under which people are supposed to acquire relevant knowledge, and the processes through which they might acquire this knowledge.If it is clear what assumptions are commonly used in this respect, we must examine them carefully in two respects.First, we have to inquire whether they are necessary and sufficient to explain the movement towards equilibrium; and second, we must show to what extent these assumptions are justified by reality.But I'm afraid I'm currently entering a stage where it's hard to say exactly what these assertions of existence, trends toward equilibrium are, based on what we've made and what our analysis fits well in the real world. What are the assumptions made.I cannot kid myself that I have studied this very well so far.All I have done, therefore, is to raise a great many questions to which, one by one, we must answer if we are to make sense of the significance of our argument.

Economists seem to agree that the only necessary condition for establishing an equilibrium is "invariance of arguments."But when we realize the ambiguity of the concept of "argument," we are skeptical, and quite rightly so, that it does not help to discuss the matter in depth.Even if we assume - as we may have to do - that the word "argument" is used here in an objective sense (which, it must be remembered, includes the preferences of different individuals), this remains unclear, Nor does the "argument" make it clear that people are both necessary and sufficient for people to actually acquire the knowledge they need, which is meant to state the conditions under which people do so.Some authors feel the need to add "perfect knowledge" as an additional and independent condition, which is important anyway.In fact, as we shall see, the invariance of objective arguments is neither a necessary nor a sufficient condition.The reason why it is not a necessary condition is due to the fact that, first, no one wants to explain it in the absolute sense that nothing in this world must happen forever, and second, as we have seen , once we want to include changes that occur periodically, or uniform changes at a constant rate, etc., the only way we can define invariance is with reference to expectations.The whole implication of all these cases is that there must be certain discernible rules in the world that make it possible for us to correctly predict events.But this is obviously not enough to prove that one will learn to foretell correctly, and the same is true even of invariance in the absolute sense, in no small degree.The invariance of arguments for any individual in no way implies that all facts independent of himself are invariant, for in this sense only the tastes of others, and not the behavior, must be assumed to be invariant. Yes, since all those other people change their decisions as they gain access to external facts and experience of the behavior of others, there is no reason why these successive processes of change should always achieve their goals.These difficulties are well known, and I mention them here only to remind the reader of how little we actually know about the conditions under which equilibrium will be reached.However, I do not intend to go any further along this line of thought, though not because of the lack of possibility for people to gain experience on unsolved and interesting problems (the possibility is only that people's subjective arguments will fit each other , and consistent with objective facts), but because I think there are other more fruitful ways of solving the key problem.