Welcome to the reading group!
How do these normally work? I was thinking that we would agree to read a certain amount of text in regular intervals, then discuss said material while continuing to read?
Sounds good 
How about 30 pages a week? It’s 120 pages in total, which means it would take us a month. Of course, we could stretch it out longer if you prefer. You might want to link the pdf version in your OP.
Edit: Or it might be better to go by chapter (or sub-chapter).
Ultimate Foundation of Economic Science (Mises Wiki) - Download links: EPUB; PDF; HTML
That sounds good. How about this reading scheme:
Week 1 (April 17-24): Read until 2nd chapter
Week 2 (April 24-May 1): Read until 5th chapter
Week 3 (May 1 - May 8): Read until 6th Chapter
Week 4 (May 8 - May 15): Finish
If there’s no complaints, then - “Berichten Sie in eine woche zurück!” (report back in a week)
cool beans -I’m in
Just now looking at the TOC. I think I’m really going to like this book.
Intro.1 - I read him as saying that we don’t need to get into philosophical problems about whether the world itself is constantly changing or eternally unchanging - the fact is that the world as we experience it as we act is constantly changing in time and the old phrase “you can’t step in the same stream twice” is the case for the particular states of affairs that comprise awareness as it relates to action.
Intro.2 - The hope for Utopia has motivated people to attempt to “transcend” the limits of “mere human” action and reason and attempt to get at the “real reality” behind everything - the Absolute as he calls it in I believe Anti-Cap Mentality - and ignore the fact that knowledge is merely a category of action.
Intro.3 - The economist or praxeologist has to be a generalist of human knowledge, otherwise he is bound to confuse the means (levers, gears, antibiotics, etc.) with the ends. A broad view of human knowledge is required to be able to clearly say “these are not human ends, these are merely means to human ends”.
Intro.4 - Mises identifies the shake-up of mathematical epistemology instigated by the discovery of non-Euclidean geometries as a major factor in the dethroning of the deductive method as a reliable tool of knowledge. However, he argues that the positivists have gotten carried away and thrown out the baby with the bathwater - just because the axioms of geometry are not “necessary truths” doesn’t mean that praxeological axioms are in any kind of doubt - you act, you know you act and that’s all you need to know to derive the rest of praxeological truth.
Intro.5 - Mises banishes the hand-wringing about whether the world is a function of the mind or vice-versa - from the point-of-view of praxeology it does not matter. Whatever the ontological relationship between the external world and the mind, the external world offers resistance to the ends which the mind seeks and it is this unalterable fact which is the concern of praxeological study.
Intro.6 - Mises basically points out that natural science is unconcerned with teleology but praxeological science is nothing but teleology. Natural science is the study of means qua means (without regard to ends) but praxeology is the study of means as they relate to their suitability to attain ends.
Intro.7 - This is my favorite part. “As among these elements of teleology is also the category of causality, the category of action is the fundamental category of epistemology, the starting point of any epistemological analysis.” Basically, all knowledge is a category of action. What Mises stops short of saying (but is directly implying) is that all science - including natural science - is a category of action, that is, is properly categorized under the study of human action. Natural science is the study of means qua means (how many Newtons of force must be applied to a lever of this length to lift such-and-such a weight? etc.) but the study of means as a body of knowledge is itself a category of action! So, if there is any proper way to categorize all of human knowledge into a single book, its title would have to be “Human Action”!
Intro.8 - This is an interesting tidbit… Mises suggests that “pneumatology” (study of the spirit) would be perhaps the best word in the English language to describe the sciences of human action. This is consistent with my usage of the term “soul” to describe the self.
Clayton -
What is the list of “Intro.s,” Clayton? And by ‘regularity,’ does Mises mean to say ‘time-invariantly operating causes’?
@TOG: I mean each section of the Introduction, for example Introduction Section 1, Introduction Section 2, etc.
I’m not sure what cite you’re referring to “regularity” from - regularities do not have to be time invariant and they do not have to be causal.
Clayton -
Oh, ok.
Mises introduces the concept of regularity are in The A Priori Representation of Reality and furthers its use in Induction. Here’s some context:
No thinking and no acting would be possible to man if the universe were chaotic, i.e., if there were no regularity whatever in the succession and concatenation of events. […]
There is only one point about which there cannot be any disagreement, viz., that they all can be reduced to the a priori insight into the regularity in the succession of all observable phenomena of the external world. In a universe lacking this regularity there could not be any thinking and nothing could be experienced. For experience is the awareness of identity or the absence of identity in what is perceived; it is the first step toward a classification of events. And the concept of classes would be empty and useless if there were no regularity.
If there were no regularity, it would be impossible to resort to classification and to construct a language. All words signify bundles of regularly connected acts of perception or regular relations among such bundles. This is valid also of the language of physics, which the positivists want to elevate to the rank of a universal language of science. In a world without regularity there would not be any possibility of formulating “protocol sentences.” But even if it could be done, such a “protocol language” could not be the starting point of a science of physics. It would merely express historical facts.
If there were no regularity, nothing could be learned from experience. In proclaiming experience as the main instrument of acquiring knowledge, empiricism implicitly acknowledges the principles of regularity and causality. When the empiricist refers to experience, the meaning is: as A was in the past followed by B, and as we assume that there prevails a regularity in the concatenation and succession of natural events, we expect that A will also in the future be followed by B. Therefore there is a fundamental difference between the meaning of experience in the field of natural events and in the field of human action.
This goes to my Madame Blavatsky thread (thinking of the Universe and everything in it as an acting being) - people used to think of everything in the Universe as if it were an acting thing. The trees meant to grow and streams meant to run downhill and this was expressed in part through the assignment of deities to govern and bring order to all these acting entities.
In reaction to the animism of prior ages, people began to think of the world as a “clockwork universe” and this metaphor expanded until included the human mind itself. The denial that humans have a nature is a foundation-stone of modernism about which Steven Pinker has written an entire book.
I argue in my Madame Blavatsky thread that there is a way to salvage the concept of action with respect to what we ordinarily think of as inanimate matter - electrons, protons, etc. Mises argues in HA and this book that the correlation between human choice and “pleasure/satisfaction” is purely formal - what we mean by pleasure/satisfaction is that which a person is aiming at. There is no reason this formal correlation cannot be applied to any entity without making any kind of claim about the consciousness or complexity of the entity. It can be used as a conscientious anthropomorphism. However, what is not clear is that thinking this way is of any advantage - which is basically what Mises is saying when he says that people realized it’s in vain (no use).
Discussing the role of a priori reasoning within natural sciences, Mises notes:
He’s discussing the discovery that non-Euclidean geometries are useful in theories of physics which was disconcerting at the time since Euclidean geometry had been ascribed an almost revelational status for close to two thousand years. The correspondence of non-Euclidean geometries to physical theory was like a “loss of faith” in the axioms of Euclid, as if humanity had been deluded for two thousand years and had not realized that we had chosen the “wrong” axioms. But Mises points out above that the axioms of Euclid are not wrong from the perspective of action as buildings and bridges and engines work just fine based on the axioms of Euclid.
The response of the empiricists to treat all axioms as “arbitrary choices” was a radical overreaction to this loss of faith, again motivated by the pretense to Absolute knowledge. There was only a crisis for those who had imagined that the Euclidean axioms were Absolute. The simple solution is to recognize that the axioms of Euclid described reality non-absolutely - they are subject to revision. Interestingly, Steven Pinker has something to say on this subject, as well. He has identified an “intuitive theory of physics” in human language. I think this casts a lot of light on the epistemic nature of geometry.
But the significant difference between geometry and praxeology is that geometry describes the objects of knowledge, that is, that which is known. But praxeology describes the subject, the knower himself. Hence, praxeological axioms are not provisional, they are not subject to revision and praxeology is not in danger of ever experiencing a crisis like that which occurred in geometry in the 19th century.
I want to bolt on to this discussion the fact that some important problems in epistemology (the problem of induction, the limits of human knowledge) have been solved (with mild restrictions) in the latter half of the 20th century. These resolutions are consistent with Misesean epistemology and make his arguments more rigorous.
Clayton -
@TOG: There is actually a way to formalize the meaning of the word “regularity” if we adopt one of several fairly natural metaphysical constructs for discretizing the world.
Basically, we say something like this: for any natural phenomenon which admits a description by a computing machine (perhaps to some degree of precision), we define its irregularity as the length of the smallest computer program which, when executed on a Universal Turing Machine, yields the description of that phenomenon.
To translate this to ordinary language, we can roughly think of this as simulation. But simulation is a little bit restrictive because perhaps the phenomenon is being exhaustively described, in which case it is not merely a “simulation”. This is particularly relevant for quantum computation which can be thought of as fundamentally identical to the unfolding of physical phenomena (see Seth Lloyd’s book Programming the Universe).
So, the regularity of some phenomenon is the size (in bits or bytes) of the program and initial conditions required to recreate the phenomenon in simulation. For example, only a simple simulation may be required to simulate the mechanics of levers and pullies. So the phenomena associated with these simulations can be thought of as very regular. But the simulation required to recreate aerodynamic turbulence in simulation may be very complex, both in terms of the size of the simulator itself and in terms of the size of the initial conditions required to set up the simulation. So, we can say that aerodynamic turbulence is in some sense less regular than the phenomena of levers and pullies.
To understand in a moment Mises’s entire epistemological disagreement with empiricism and positivism, simply imagine trying to write a computer simulator that simulates the behavior of all human beings on Earth. This is the pretense of knowledge, the fatal conceit. Economists speak as if they have some kind of “rough simulator” of human behavior and that this model is “very high-level” kind of like a simulator of the Earth’s core or its magnetic field… naturally, it abstracts away immense amounts of detail but at the scale at which the phenomena are described, the laws describing them definitely hold within the margins of error. But this metaphor simply doesn’t hold when applied to humanity - those immense amounts of detail being abstracted away are all relevant to the global phenomena of human behavior. At the global scale, the laws which are supposed to describe human behavior do not hold at all because - in abstracting away details - the entire character of human behavior is altered so as to be unrecognizable.
Returning to the issue of regularity, where it becomes particularly interesting is when you neglect the physical aspect and investigate the mathematics of regularity as a purely abstract matter. What you discover is that - surprisingly - there is no finite set of mathematical axioms that encompasses all mathematical truth. The hope of early modernism in mathematics (Hilbert’s initiative to formalize all of mathematics) was that we could derive a provably correct, finite set of axioms from which all mathematical truths could be derived. But Godel dashed these hopes in 1931 and Turing and Chaitin would further eviscerate any idea of encompassing all mathematical truths.
The idea of formalizing mathematics is that it should be easy, at least compared to formalizing physics. It seems that mathematical truths should have fewer exceptions or special-cases than physical truths. But if we can’t even formalize all mathematical truths, how can we ever hope to formalize all physical truths? And the fact is that you provably cannot formalize all physical truths. Let any physicist come forward with a theory of everything and I will construct a device whose long-run behavior his theory cannot - even in principle - predict.
What this means to epistemology is that any set of axioms describing mathematical phenomena or phsyical phenomena is necessarily provisional. We are never dealing with the Absolute because we would have to be able to hold in our mind an infinite amount of incompressible information, that is, we would have to know the infinite axioms of mathematical truth. On every count, the modernist epistemology is DOA.
Clayton -
I think this jibes with Clayton’s definition, but in more layman like terms:
Regularity is when we see a recurring pattern in the Universe. For example, the sun rises and sets every day. Which makes us think we can thus predict that the sun will rise and set tomorrow. This curious human faculty, of thinking that what happened before with great regularity will happen again, is what we call inductive reasoning.
Clayton is, I think, describing one possible way of measuring how regular something is.
This is more or less what I thought of Mises’ use of the term, Smiling Dave.
Clayton seems to provide a broader definition of regularity and includes its applications to mathematics and physics. It’s interesting that he brings up the work of Gödel (with which I’m somewhat familiar) regarding the existence of mathematical problems for which there are no solutions (his Incompleteness Theorem) and other mathematicians (with whom I am flatly unfamiliar). This proof poses a limitation on what can be known in the mathematical world and all the more so in physical and social sciences (where there are no constant variables to the degree that there are such variables in mathematics; social sciences are based on dynamic and random events amongst actors).
Originally, I thought that Mises meant, by regularity, causality. Of course, there is a paragraph in which he uses both terms in the same sentence which thrashed that idea- It wasn’t an idea that I held particularly firmly anyway. In the following excerpt, Mises seems to make it clear:
In proclaiming experience as the main instrument of acquiring knowledge, empiricism implicitly acknowledges the principles of regularity and causality. When the empiricist refers to experience, the meaning is: as A was in the past followed by B, and as we assume that there prevails a regularity in the concatenation and succession of natural events, we expect that A will also in the future be followed by B.
In this excerpt, Mises links time with experience and states that empiricism, despite what it claims in regards to rationalism, nonetheless must assume a priori truths (causality and regularity, particularly). There’s no other reason why B should follow A at time T2, then the knowledge that B did follow A at time T1. These two events cannot be linked by experience; they may only be linked by the human intellect and by recognition of certain phenomena that repeat (whether this is assumed by the actor in question or is in fact true) regardless of time. So by predicting that B will follow A at time Tk, once other variables are controlled and such has been the case in prior trials, the empiricist necessarily assumes a regularity in the universe between certain phenomena. This regularity cannot be observed and is therefore not empirical, but a priori.
This is one of the problems that I was alluding to as having been solved in the 20th century (under mild assumptions about the physical world). Specifically, the problem of induction (how do we know that the future will be like the past?) is basically a poor way of casting the problem. We can rephrase the problem in terms of the probability space over programs input to a Universal Turing Machine (again, under mild assumptions about the physical world) and the problem of induction simply becomes a probability statement - it is so much more likely that the Sun will rise tomorrow than that it will not because the most probable models of the Sun’s behavior show the Sun rising tomorrow rather than not rising.
The theory (Solomonoff induction) has only really been worked out in detail in the mathematical realm but its applicability to physical reasoning - under mild assumptions about the physical world - is straightforward.
Clayton -
Clayton,
Are you sure you aren’t hiding the problem behind the words “under mild assumptions about the physical world”? Because the assumptions of course are that the world isn’t random.
This regularity cannot be observed and is therefore not empirical, but a priori.
Just to make sure we are on the same page:
When we have a pattern, it is composed of elements [blue square here, then yellow square to the left of the blue, then another blue square to the left of the yellow square etc.] and of the elements being arranged in a pattern. The pattern can be described, for example, as “each color has the different color to its left, always”.
I think we can agree that the elements of the pattern can be observed [=the sun rose and set on Monday. It rose and set on Tuesday, etc]. What is a priori is grasping that there exists a pattern that the elements follow [=the sun rises and sets every day].
If that’s not what I said, then that’s what I meant.
No, we can admit randomness and still come to the same conclusion (probabilistic Turing machines admit all the same conclusions in this regard that their deterministic counterparts do). Of course, the Universe can’t be completely random but any explanation by virtue of being an explanation is assuming that the Universe is not completely random, else what is the point of trying to explain anything?
In any case, the mild assumptions have nothing to do with randomness, the mild assumptions have to do with whether the Universe can be exhaustively described with discrete states. Turing basically gives a metaphysical argument for the discretization of sense perception in his foundational paper, I recommend you read his original discussion of this (see section 9.I for his argument, it is not mathematized, so it is accessible to non-specialists).
In any case, quantum computing theory may obviate the need for even these mild assumptions. The way Seth Lloyd explains it, “the universe is indistinguishable from a quantum computer” and the same conclusions that hold for Turing machines have already been extended by theorists to quantum computers. This means that we can naturally migrate the conclusions regarding the limits of knowledge, formal measures of complexity, and Solomnoff induction as directly physical theorems.
Clayton -
A few thoughts on the first chapter:
Mises presents the concept of regularity as an a priori truth, calling upon the example, or rather its impossibility, of ice cubes setting a glass of water on fire. Such scenarios handle positivism well enough. Here’s my question though: doesn’t the assertion of natural regularity somehow involve a posteriori knowledge? In parrallel, mustn’t all scientific experimentation involve both a posteriori and a priori truth? Perhaps I don’t understand the definitions well enough but I think that a posteriori knowledge is that that we know because we’ve experienced it. A Priori is that that we know ahead of time because any alternative is inconceivable. So a posteriori does not need to be proven because it only needs to be recorded whereas the case is different with a priori, is that so?
This makes the whole priori/posteriori dichotomy confusing to me, since empricist economists always bash a priori methods but isn’t an experiment (even if no economic experiment is possible) also an a priori assertion, if one seeks to use it as a device of prediction?
He throws Marx away with the usual deftness, turning his own materialist philosophy against him. If I am right in understanding, the material productive forces alone decide the course of history, that there is no correct or incorrect only the march of history. This of course implies that Marx’s own writings have no significance. Someone correct me if I’m wrong on this.