At the most fundamental level of analysis, I would say the same thing.
First of all, he didn’t think in practice they were in opposition to mainstream ideas:
He must have thought that he was odd only in the respect that he was conscious of the method. It wasn’t as if he employed a new method, and everybody else employed something else. No, it was simply because he was aware of this situation that he was able to make his economics cleaner and so on. I imagine while everybody else was making a bunch of errors out of being confused about the method, he was above most of that strife, and able to go on his way without even entering into a lot of their controversies.
Now, why did he think that his epistemology (his awareness of the method) was different than the mainstream? He knew in the end, his method wasn’t different, but why did he think that his understanding of it was different? I’m pretty sure that it was, and I suspect that most people still haven’t even an inkling of the significance of his argument. After all of his effort, most economists still really have no idea what he was going on and on about in the beginning of Human Action. In a lot of respects, it’s not even their fault; I contend that Mises was systematically misleading in those opening sections, but I will leave it for another day to ask whether that reflects some sort of shortcoming of understanding on his part. For now, it’s simply important to admit that he systematically confused the hell out of everybody. Or at least that’s what I suspect.
But, to get back to the point, why did he speak as if the social sciences (praxeology) was in a separate magisterum altogether from the natural sciences? Well, first it’s important to recognize that he didn’t try to say that there would be no way to unify science ever: He simply said at this point we don’t know how to do that; I don’t want you (or anybody else) to think that he was some sort of nutcase who thought that the mind had nothing to do with the body. He said for now we need to aquiesce in some sort of methodological dualism. I mean, I contend that such a dualism would exist no matter what, but whatever. (I will get to that in a second.) I’m not even sure whether he would have disagreed with that. Did he mean that we could destroy the methodological dualism, or did he merely think that we could make it useless? Anyway, I’m probably starting to get pretty confusing, so I will simply move on to what the methodological dualism is.
As I said in my 2nd post in this thread, I think of the social sciences and the natural sciences (their opposition being the methodological dualism) simply as 2 different kinds of models for the same thing. In one case, you model the other people as conscious. And, in the other one, you just model them as 3D objects. In both cases, you’re simply making a mental model of the sensations that you’re experiencing, but it’s as if they’re 2 completely different “points of view” or something. In the 1st one, you model the cluster of sensations that you would call the other people’s bodies on yourself. And, in the 2nd, you model them as mere objects. Or whatever. In you everyday life, (I assume) you experience sequence of sensations, some being “in your head” and others being “in the outside world”. Well, the social science model uses some of the sensations in your head (visualizing stuff, feeling desire for something, and so on) to organize your thoughts about some of those in the outside world (the regularity of the 2D images on your visual field that you refer to as the other people), and the natural science one doesn’t do that. But, it’s almost impossible to overemphasize the point that both models are simply to predict what your future sensations will be (what 2D images will appear on your visual field, what bodily sensations you will experience, what sounds will surround you, and so on).
Anyway, that’s the methodological dualism, but now let’s talk about how (I assume) the mainstream didn’t understand that. In the social science model, you model your cluster of sensations that you call the “other people” on your cluster of sensations that you call your “thoughts”. To predict what you will experience in the future in terms of the first cluster (the ones making up the other people), you imagine that they have thoughts, feelings, and so on too. It’s not just you who thinks; it’s also the rest of the people. But, again, it’s not as if you actually see any of their thoughts; it’s just a damn useful model for predicting some of your sensations. For this exact point, it’s perfectly meaningless to ask whether animals are really conscious, or just act like it. If they act like it (if it’s useful to consider them as if), they simply are. When we say that the other people are conscious, we’re simply saying that it’s useful to act as if they are. If it’s useful to do that with animals, there’s absolutely no difference. Anything else would plunge us into a useless dispute of words.
But that means that we have to start out our investigation by turning our eyes inward. Before we can turn our attention to the other people in the outside world, we must get an understanding of how our action works (the “moving parts”, such as how desire works, how belief works, and so on), and that’s in our head. Before we can model other people on ourselves, we have to model ourselves. If we don’t have any idea how our desire, belief, and so on works, how the hell could we expect to predict anything on their basis? So we can’t just start going outside and seeing what other people do; we have to start out by sitting inside and seeing what we do. Only once we have an understanding of our own mind can we act as if anybody else has the same thing going on. We must build a model of ourselves before we build a model of anybody else.
So where did the mainstream mess up? Well, they didn’t realize that at all. Simply working off of their subconscious, common sense ideas of how their action worked, they started to try to “empirically” discover how people behave. But, as Mises stressed over and over, it’s not that they’re somehow forgoing the first step (building a model of their own action); it’s that they’re simply letting it stay in its unscientific, common sense formulation. I mean, this whole thing is pretty easy to understand when you take notice of the fact that you probably started modeling other people on yourself when you were less than 2. That machinery sunk into your subconsciousness many years ago. It’s takes an extremely self-reflective mind to dig up a program that you wrote so far in the past. So for most people, they take it so for granted, that they don’t realize that the program even exists. It’s as if they think that they really do see other people’s thoughts or something. And, I mean, the whole thing works pretty well for everyday life, but it’s certainly not rigorous enough for science. Or at least Mises didn’t think so.
So while the rest of the economists worked off of a model that they packed deep into their subconsciousness a few decades in the past, Mises worked off one that he was constantly revising and trying to understand. He understood that he really couldn’t see anybody else’s thoughts (see The Alter Ego in Human Action), so he knew that the first step in any economic theory was to build that model. It wasn’t an ultimate given to his experience; it was something that he had to construct out of a bunch of much more fundamental pieces. He understood that, and it gave him the tools to be an extremely reasonable, powerful economist.