Probability theory: The logic of science
by E. T. Jaynes.

It is available at
at http://omega.albany.edu:8008/JaynesBook.html

While it is a book for specialists, I think the first sections give an interesting introduction into subjective probability theory accessible for the non-specialist too.

The very point relevant to the discussion here is that justifies numerical probabilities as the appropriate apparatus for plausible reasoning, which has nothing to do with frequencies.

And it also clarifies some more philosophical/political issues: the frequentist interpretation has a strong empiricist background, so that subjectivist probability is also anti-empiricist, and, in fact, as the mathematics of plausible reasoning it can be considered as a natural part of praxeology.

I think you are reading far too much into what I wrote. I certainly didn’t say that Mises thought that human action and natural phenomena were identical in every way. What I said is true even for soft determinists and causal determinists. Human actions do not occur for no reason whatsoever. They occur whenever the actor believes that he can make himself better off by acting than by refraining from acting. So, when one says “human action is purposeful,” he is also thereby saying that there is a reason for the action to take place, and a cause(s) for the action. The causes of action lie within man, (i.e., his volition and subjective beliefs), but they are causes nonetheless.

I’m not trying to say that human actions are identical to natural phenomena. All I am saying is that human actions do not occur for no reason whatsoever. They don’t just randomly occur, in the way that Richard von Mises thought atomic particles randomly and schizophrenically move around for no reason. There is a reason why every single human action occurs, and there is also a cause for every single human action.

That’s all I was pointing out.

Cheers,
Mark

I take issue with your postscript: “L. von Mises was a hard determinist, so the phrases I cited are essentially equivalent for him.”

The phrases in question are:
(1) “Human action is purposeful behavior,”
(2) “Human actions have causes,” and
(3) “Human actions have a reason for occurring.”

I don’t think these are essentially equivalent. Consider the following propositions:
(1) “A pit bull is a dog,”
(2) “A pit bull is an animal,” and
(3) “A pit bull is a living organism”

Suppose I’m a “hard anima-caninist” and believe that all dogs are animals. While this would mean that (2) and (3) follow necessarily from (1), it in no way implies that (1) is “essentially equivalent” to (2) and (3).

In fact, by insisting that there is no essential difference between saying (1) versus (2) and (3), I eliminate the possibility of forming any scientific statements that may be unique to the relationship expressed in (1).

The implications of this become clear if we try to do a little pit-bull science on the basis of our foundational claim. Let’s say I have recently purchased a pit bull. What should I give it to eat? Well, it’s an animal, right? What do animals eat? Compare that method versus saying, well, it’s a dog, right? What do dogs eat? The latter line of questioning will clearly yield more precise and interesting results. The former might very well kill my pit bull.

One might object that all pit bulls are dogs but not all dogs are pit bulls, and therefore inquiring what dogs-in-general eat isn’t going to solve my pit-bull-diet problem. But remember that for Mises, human beings are the only things that act purposefully. So to make the analogy a better fit, we’d have to suppose that pit bulls were the only dogs we knew of.

That Mises, even as a hard determinist, couldn’t have thought the propositions were “essentially equivalent” is also clear from the second sentence of human action. He describes human action in a number of additional ways beyond “purposeful behavior,” including his contrasting it with “animal reaction.” Now, for a hard determinist, animal actions are caused no less than human actions are caused (and still less so the movement of rocks). Yet Mises seems to think that human actions are different in kind than the movement of rocks or animal reaction. And insofar as it’s distinguishable, the science of economics has a role in explaining the implications of this.

Human action, like natural phenomena, does not occur for no reason whatsoever. Human action occurs whenever a man values some end and uses some means to try to achieve that end. As is the case with natural phenomena, however, man is not omniscient, so he is thus not in a position to know how other men will act in the world. He does not know the state of other men’s wills, nor does he really know what ends other men value, or the means that other men think will achieve their ends.

IF MAN WAS OMNISCIENT, however, he would know how other men were going to act even before they acted. He would have advance knowledge of the goals each man subjectively valued, and he would have advance knowledge about the means each man intended to use in order to achieve his goals. For a man endowed with such omniscience, there would be nothing uncertain or probabilistic at all about human action. He would know in advance, AND FOR CERTAIN, how other men are going to act.

Man is obviously not omniscient, however, when it comes to either human action OR natural phenomena. He does not know all of the factors that cause a natural event to occur. He does not know the state of other men’s minds. He thus often has a need to measure how likely these events and actions are, given his ignorance of their causes. This is where probability comes into the picture for man. He uses probability to measure his uncertainty about the world.

Man does not use probability to measure something “out there” in the world, however. For, if man knew all of the various causal factors that influence natural phenomena, and all of the various subjective ideas that motivate human action, he would have no need for probability at all. He would know, in advance and for certain, what will occur. When he uses probability, man thus measures his own ignorance about the world.

The only way around this conclusion would be to argue, as did Richard von Mises, that certain things can occur in the world FOR NO REASON WHATSOEVER. He was an indeterminist, in other words, which opened the door to countering my argument. Austrians, on the other hand, are not free to argue that things occur in the world for no reason whatsoever, so this avenue of attack is barred from them.

I hope this clarifies my argument, so we can get back to discussing it, rather than debating something completely extraneous.

Cheers,
Mark

P.S. L. von Mises was a hard determinist, so the phrases I cited are essentially equivalent for him.

Now, because we cannot know that human action B was caused by prior event A and because we cannot know that the statement “human action have causes” was caused by a prior event, Mises and Hoppe both argue that human actions and praxeology are not governed by causality but by teleology, that we face the insurmountable chasm of methodological dualism, and that we must distinguish between case and class probability. And this exactly why your definition of subjective probability is incorrect.

So we would still have to answer the challenge, we must prove that any one of the statements below is true, without falling into a performative contradiction:

1) “Human actions have causes”
2) “Human actions are ‘causally determined’ ”
3) “Human action B is caused by a prior event A”
4) “Human actions have antecedent and time-invariant causes”

Or, let us quote a prominent Austrian who has written on methodology proving any of them.

Until then we have to conclude that your proof of subjective probability is incorrect.

On a lighter but related note:

“ ‘Human action is purposeful’ is another way of saying that human actions have causes”–Mark Crovelli

“I could have just as well have used the phrase ‘reason for occurring’ instead of the word ‘cause’ “–Mark Crovelli

So instead of starting his Treatise on Economics with the sentence: “Human action is purposeful behavior” Mises could have just written: “Human actions have causes,” or better still: “Human actions have reasons for occurring?”

Listen, I enjoy your LRC articles and I do appreciate the time you have spent reading and responding. Maybe my comments rung a bell with you. Maybe not. But your paper and comments made me re-anchor certain concepts of human action and causality and for that I am thankful.

You are way, way off base if you think that there is some sort of performative contradiction involved here, because the phrase “human action is purposeful” is another way of saying that human actions have causes (i.e., a reason for occurring or not occurring), and this is the very core proposition for Austrians.

You are getting hung up on my choice of the word “cause,” which is making you think that I mean something analogous in the human world as in the natural world. This is not how I use the word, however, as you can gather from my paper, wherein I explicitly describe that I use the term in the sense of “causal determinism.” This is in contrast to the “indeterminism” of Richard von Mises. I could have just as well have used the phrase “reason for occurring” instead of the word “cause.” But, I wanted to contrast my causal determinism (and Ludwig von Mises’s determinism) with the crazy indeterminism of Richard von Mises.

Cheers,
Mark

Here is my proof that they do not, that we have no way of knowing that human actions have causes. If human action had causes, as I wrote above, then what about the statement “human actions have causes” which is, presumably, a human action itself? How do you explain how you came up with that statement? In other words, what caused you to come up with the statement “human actions have causes”? Did you learn it from experience? From past observations? From future observations? From probability theory? From subjective probability, as in, there is an 80% chance that I’m right? A mystical experience perhaps?

The answer is none of the above because one cannot prove that human action have causes without falling into a contradiction.

All of the categories of human action (means, ends, choice, costs, profit, time, and even causality itself) are derived via the action-axiom. They cannot be observed nor predicted.

Human actions are governed by teleology. Everything else is governed by causality. Both realms of phenomena are mutually exclusive but complementary. This duality demands two separate methods of study, that is, methodological dualism, the orthodox Austrian view, one for human actions (praxeology and entrepreneurship), another one for everything else (the scientific method, probability theory, etc.). Probability theory can and is used for “everything else,” but it cannot be used to study or predict human action.

Until you can prove that human actions have causes, we have to agree, per your own admission, that we cannot use probability (RvM’s, LvM’s or yours) to predict and that therefore your proof is incorrect.

Stating “human actions do not occur for no reason whatsoever” or “I don’t know how many more times I can say this” or “isn’t this obvious to you?” or “you are an Austrian, are you not?” is not a proof that the statement “human actions have causes” is true.

So answer my question: prove that human actions have causes without falling into a performative contradiction.

If you cannot prove it yourself, can you at least offer a citation from Mises, Hayek, Hoppe, Huerta de Soto, Rothbard, Gordon, Long, or any other prominent Austrian who has written on methodology wherein they prove that human actions have causes, or better yet, wherein they prove that human actions have causes and that therefore we can use probability to study human actions?

Are you really trying to argue that human actions have NO CAUSE WHATSOEVER?

You seem to be under the impression that I am arguing that human actions are caused in the same way that natural phenomena are caused. I have never said this, and, in fact, I explicitly disavow this idea both in my article and above.

So, let me state this again, in the hope you will read it this time: human actions are caused by the intentions and volition of the actors themselves. Human actions thus have causes, and they do not occur for no reason whatsoever.

Cheers,
Mark

Answer my question: prove that human actions have causes without falling into a performative contradiction.

What matters is whether human actions have causes. Do you dispute that they do indeed have causes? Are you really saying that human actions have no causes whatsoever?

Cheers,
Mark