Praxeology and a priori truth: an interview with David Gordon
By Richard Storey
The following is an interview I conducted with the brilliant philosopher, Dr. David Gordon, who has been described as the semiofficial reviewer of the libertarian community. There is hardly a better mind to help us understand ‘praxeology’ and the basis for Austro-libertarian thinking in the school of von Mises, Rothbard, Block and Hoppe et al. Please follow the links for helpful resources to understand the definition of some philosophical terms and to access useful materials mentioned by David.
R: David, you have divided up two schools of thought in philosophy and economics – the German school, which I think stems from the Vienna circle, influenced greatly by the British empiricists, such as Locke and Hume; and the Austrian school spearheaded by Ludwig von Mises who was influenced by Menger and Böhm-Bawerk. Now, Mises was somewhat of a rationalist; at least, he used the language of Immanuel Kant to show we need some sort of dualism when we engage in the empirical, physical sciences (which use the scientific method and the historical method). That is, in order to gain knowledge from the world, we must use rationalistic thinking as well. In that sense, would you say Mises’ thinking hearkens back to Aristotle by ascertaining those irrefutable axioms we can determine rationally for ourselves?
G: Yes, what you say sounds to me like what Mises thought. The empiricists think we get knowledge by extracting our concepts from the world; for example, you look at various tables and you see what they have in common and you come up with the concept of ‘the table’. But, Mises observed this doesn’t work for certain concepts and in this he was very much influenced by Kant. He noted people would be unable to determine what ‘actions’ are just by looking at the world. We could see movements of various kinds, but we wouldn’t be able to see actions unless we had the concept of action. So he thought there were some kinds of actions that were most important, ones that could not be derived from experience but, on the contrary, were needed in order for certain types of experience to make sense.
R: It’s important to bear in mind that the crux of Misesian thought is in these concepts. However, some people have criticised Mises for his discussion of the a priori. Please define a priori and what is a posteriori?
G: Well, this is a very controversial term. We have to distinguish between a priori concepts and judgements or propositions. An a priori concept is the kind we have already mentioned – it isn’t derived from experience and it is presupposed to understand a certain kind of experience. But, an a priori proposition, which is what most of the controversy is about, is one that you can know to be true without testing it by experience or anything else. It’s just one that you grasp immediately once you understand the meaning to be true. For example, 2 + 2 = 4; I don’t know this by taking two apples with two apples, I add them up and there’s four apples, and then I repeat this and make a generalisation – that doesn’t prove it’s right. It seems that I can take 2+ 2 = 4, expand on it and realise it’s true; that’s the a priori truth. It’s one that can be known independent of experience. However, you cannot know it to be true because you were told so by a reliable math teacher. Here, you don’t know it to be true because you thought about it, you’re just relying on another person’s word. That a priori proposition is one that you could know or that could be known just by thinking about it.
(It gets more complicated though; for example, known by whom? Is it known by anybody or by you? There are various theories on that.)
R: In summary, Mises’ view was that empiricism isn’t enough because we need the concept of action to make sense of our experiences. Is that a good summary of Mises’ thinking?
G: I like, very much, what you say because that is exactly his view – we need these concepts in order to understand experience. Actually, if you just stop there, you wouldn’t have to say that there are any a priori propositions. You could just say, all our knowledge turns out to be hypothetically true or yet to be tested. So someone could hold quite consistently the theory that there are a priori concepts but no a priori propositions. You could also hold it the other way, i.e. you get all your concepts from experience. Once we’ve got these concepts, we see there are certain relations between them and we just grasp it’s true. Mises held both these views and thought there were both a priori concepts and a priori true propositions.
Synthetic or Analytic a priori?
R: I think the debate about whether Mises was really talking about an ‘analytic’ a priori or a ‘synthetic’ a priori is unimportant. We only need to understand that we have these concepts which we can determine to be true before we test the matter, and these are irrefutable axioms. Would that be correct?
G: Yes, I agree with you entirely. I think there has been too much made of analytic and synthetic theories. Some people will say (and I don’t think they’re wrong to say) that Austrian economics consists of synthetic a priori truths, but Mises himself in Human Action never says that. Many say, ‘Of course he says that,’ but he never says that. Mises seems to think they’re analytic. Although, in his last book, The Ultimate Foundation of Science, he does give one argument: if you deny that there are synthetic a priori truths, isn’t that a synthetic a priori proposition itself? But, again, even here he doesn’t commit himself to saying the truths of economics are synthetic a priori.
I think I should say a little bit about what these terms mean. In terms of analytic and synthetic propositions as introduced by Kant, what Mises meant by these truths is somewhat different from what the Vienna School of Economic Thought meant.
What does ‘analytic’ mean?
Kant thought that the analytic proposition is one in which the predicate is contained in the subject. Suppose I say all bachelors are male. Well, it’s part of the concept of a bachelor that a bachelor is an unmarried male of a certain age; bachelors are male, so ‘male’ is contained in this concept of bachelor. That’s an analytic proposition.
What does ‘synthetic’ mean?
In a synthetic proposition, the predicate isn’t contained in the subject. So, all propositions are either analytic or synthetic because either the predicate is contained in the subject or it isn’t. The logical positivists have a rather different definition. The division they make is: analytic or empirical propositions. Their analytic proposition is the definition of a truth of logic or part of a definition. It is similar but not exactly the same in its concept. But an empirical one for them, is just one that can be verified by sensory observation; for example, I’m now talking to you on Skype. Perhaps there are observations that could verify that it wasn’t true. Perhaps our connection was disrupted; that wouldn’t be true then. But the main flaw with the logical positivist’s division between analytic and empirical is that it isn’t guaranteed that all propositions are either analytic or empirical, whereas, in Kant’s view, they’re all either analytic or synthetic. In the positivist’s view, there is still logical space for propositions that aren’t definitions and aren’t verifiable by sense experience but could still exist! Kant’s thesis was that there aren’t any such propositions.
R: I think that logical empiricism (or logical positivism) today seems to have descended into a sort of scientism. By that I mean that many thinkers, such as Richard Dawkins, tend to point to the scientific method as if this is the only way to derive knowledge about the world. But that very statement, ‘the scientific method is the only way to derive knowledge,’ cannot be determined to be true with the scientific method. In fact, they assume and use the concepts we are describing all the time in their work and in their everyday thinking, we all do. I think there’s something very intuitive about Misesian dualism. When I explain this concept to people, they say, ‘Yeah, I already know that.’ They seem to understand that is how they derive knowledge. Do you find that’s the case?
G: I think you have some excellent points there. It isn’t scientific to say that physical science is the only the way to know things; that would be a philosophical thesis about science. If physical science were the only way to know things, we wouldn’t be able to know that because that’s not a physical science proposition. One thing Mises stressed is that we shouldn’t confine science to the physical sciences. He said, just as you suggest, in the physical sciences, we ask, ‘How do we know what the matter is composed of?’ or, ‘What’s going to happen with the stars and planets?’ The way to find out is to observe things; we don’t have this inner grasp of how matter is moving. We can’t say, just by looking at certain physical elements, what they’re going to do. We would have to just watch and see what happens. But with the concept of human action, it’s different.
Each of us is an actor, not in the sense of someone who acts in movies or plays but someone that, as a human actor, does things all the time and is always acting. We have a grasp of action from the inside, we all know we are actors and so can go on that basis. Most importantly, we can build up a body of knowledge, just thinking about what’s involved in the concept of action. The whole point of praxeology, which is the term for this science of action, is that we can infer some very surprising truths about economics that people wouldn’t just know without something difficult to understand, like complex mathematics. We can take some very simple truths and come up with something great and very valuable. And that is the basis on which Mises operated. But, the positivists and Karl Popper said, ‘You’re asking, what are the criteria for science? Scientific propositions can’t possibly be classifiable.’ But they were not taking into account all sciences, particularly economics. Mises showed economics has a distinctive method of proceedings. If you want to talk about the criteria for a scientific statement, you should take into account economics as well.
R: So, in that sense, Mises was standing on the shoulders of Frege, in saying that we have logic, we have mathematics, which is a logical science, and he was saying that economics is composed of both and is a logical science. But Mises also recognised that we must use empiricism in economics too. Is that correct?
G: Yes. A key mistake, made particularly by Popper and his followers, is regarding the Misesian view of empirical knowledge. It doesn’t follow that all we mean by ‘empirical knowledge’ is something that can be doubted or something that is just a hypothesis, something always requiring more testing. We can know certain things to be true about the world – we’re actors – and these are not just hypotheses or guesses. I frequently find students make this mistake. It’s of course right that Mises was very interested in the philosophical foundations of economics and was much more philosophically well-informed than almost any other economist, but we shouldn’t confuse Austrian economics with philosophy. There is the problem of scepticism with philosophy. Something very sceptical might be, how do I know right now that I’m not a brain in a vat and all of the things that I’m experiencing and that I think I’m experiencing, I’m really not. There is a scientist plugging various things into my brain so that I’ll have to experience them. Another sceptical problem is that, even if I know there is a world out there I can experience in my own thought, how do I know that you’re thinking? Maybe I’m the only mind. Now these are very interesting philosophical problems but they’re not problems of Austrian economics.
Suppose somebody said ‘Are we going to have a recession next year?’ It wouldn’t be a good answer to say well you’re talking about a recession but we haven’t even established there are other people yet, how can you talk about a recession? In the sciences, and economics is one of the sciences, we’re taking for granted that the world exists, that people exist, we’re taking for granted the ordinary world of common sense because, in physics, you wouldn’t say, ‘We’ve got all these electromagnetic waves and various things showing up on our instruments but how do we know that there actually are instruments? How do we know that anything really exists?’ So when students see these philosophical terms like a priori, they tend to put the bar too high and ask, ‘How can Mises establish that people really exist?’ But, Mises wasn’t trying to do that, he was just taking a basic common sense view of things. We know that we act and what follows from that within the ordinary world of common sense.
R: I think that’s why Austrian economists or philosophers who may follow Mises and his epistemological dualism would then be accused of making metaphysical statements or things of that nature. But, Mises wasn’t setting out to show that the universe exists necessarily or anything like that, rather he was just trying to use logical science to determine those irrefutable axioms that we use as concepts, which then help us to make sense of the data we acquire with our senses. Similarly, learning a language enables us to both understand and to communicate with others.
G: Yes, I think that’s exactly right. We know that the concepts apply to the world because we’re in the world, we’re actors ourselves so we know that there are actions. There isn’t just, as the positivists might say, a concept implied by certain definitions, but that we don’t really know about the true world. Let me give you an example of a positivist criticism of Mises. One of the propositions you get in praxeology is that we always choose our most highly valued preference; if you preferred something else to talking to me right now, you would be doing that rather than talking to me, so if you’re talking to me that’s your most valued preference. Felix Kaufmann suggested we are just defining the most highly valuable preference as the one we choose. But, Mises isn’t defining your highest preference as the one you actually choose, just taking preference in the ordinary sense. For example, there are people in philosophy who say there can be cases like weakness of will, accuracy, and where you want to do something that is your most highly valued preference but you’re overcome by weakness and you choose something else, ‘I really don’t want to inject myself with heroin but I’m overcome to do it.’ But, Mises pointed out that this is just a person having a change of preference, to inject themselves with heroin a number of times.
R: Returning to the subject of axiomatic concepts, Rothbard seemed to disagree with Mises who viewed these concepts as a ‘law of thought’. Rothbard said ‘I would stand more in the Aristotelian camp in viewing them as a law of reality in that there’s something objective about them’, and he thought that Mises was viewing them very subjectively. Roderick Long wrote on this and he wanted to refine the thinking of Wittgenstein, showing how Wittgenstein sought to transcend this debate, saying that if our ability to apply logic or mathematics, or praxeology, for that matter, were to break down, it’s not a particular style of thought that we’ve lost but the ability to think all together. Does that make sense?
G: Well, if I have thoughts, I hope they don’t break down all together! I know Roderick very well; I think the point he was making can be illustrated with this question: ‘Do our concepts apply to reality?’ He thinks if you’re asking this question, you’re assuming there is some kind of separation between the picture we have of concepts inside our minds and the reality out there, but he says it’s not the case that the concepts are somehow constitutive of the world they’re in because they can’t be separated from the world. For example, logic isn’t a set of psychological laws of our thought (how we should think) but, rather, logic applies to the world; they’re not separate. He thinks there is a kind of ‘conceptual grammar’ we get through studying praxeology and I think that is a very useful way to think about it and it’s very similar to what Mises was thinking. We do need to have certain concepts in order to understand reality but that is not to say there’s some reality that exists apart from the concepts that is ungrasped, some kind of luminal world that exists without these concepts. No, this is the world to which our concepts apply – it is the world. There’s a very interesting German philosopher, Sebastian Müller, who says this was Kant’s view too; he wasn’t postulating some other world with applied concepts that we don’t know about, this is the way we’re grasping things.
Now Rothbard, he was an Aristotelian, as you say, viewing concepts as abstracted from the world. But, he thought that when we do this, when we abstract and get the concepts, we’re not limited only to certain continued propositions – ones that could be true but need not be. Suppose I say, again, that I’m talking to you now on Skype; that could have been false, I could have forgotten about the call and gone out, given that I’m an old man and very absent minded and sometimes do such things. But, Rothbard thought there were certain propositions we could grasp about the world that are necessary, that couldn’t be false and so he would talk more about necessary truths rather than a priori truths, and that sums up the difference here. Nevertheless, these two views are very similar in what they get from the axiom of human action; it’s just they have a slightly different philosophical argument.
R: Roderick Long points to Rand, saying she would agree more with Aristotle and Rothbard. Her view was that there is a unity to these things and, so, to have both the concept and the ability to garner data from the world around us is a whole, one unit, and not something that is divided up, as people at one time would have thought that the body and soul were two parts of one whole. Rand says it’s more like a computer; you have all the components and parts in order for it to perform a function, just as we need to have both our logical science and our empirical data. Do you think there’s a flaw in that thinking?
G: Rand had a very unusual view of the idea of concepts. You see, she thought that all the properties of an object were part of the concept so that if I have a concept of a table, it isn’t just a table as an item of furniture which we put things on, it is a sort of good rule of thumb definition. But, the concept includes every property of every table so it turns out that, if you could involve the properties in the definition, then all propositions turn out to be necessary (but, Rand made an exception for human free will). Suppose I say, light travels at 186,000 miles per second or something like that, that’s part of the definition and the concept of light, that it has that speed. So, if I can imagine a possible world in which light could travel faster or slower, Rand would say that’s how we determine whether it’s part of the concept of light. A lot of philosophers would call this a distinction between logical necessity and metaphysical necessity, but Radians don’t accept this. Now, I don’t think it’s very difficult to show that their view is logically false; it just strikes me as a very implausible view that nothing in the physical world could have been otherwise, that the laws of physical nature are ones that hold with absolute necessity. I think it would be a mistake for Austrian economics to saddle itself with that sort of metaphysical language.
R: What would you say are the best arguments against the a priori, as Mises defined it, and as it is applied by Austrian economists and philosophers like yourself? And how would you respond to those arguments?
G: Probably the best criticism of a priori truth is one derived from the philosopher, Willard Quine, who was very influential. His basic argument was applied to analytic and synthetic truth but would also apply to a priori and a posterior; he said that we don’t have any real way of coming up with non-circular definitions of analytics and synthetics. We could say that an analytic statement is true just by the meaning and the term. Alright, but what is something that’s true by its meaning and if we don’t have some way of understanding, we just get a circular thought process; we can’t say something is true by its meaning if it’s analytic because we don’t have a fixed way of understanding this. Related to that thought, he thought that all of the propositions we have are connected in some way so that we’re not testing an individual proposition by experience, it’s the whole body of knowledge that is tested by experience. For instance, if I don’t understand something, it isn’t that I would have to reject one particular proposition, I would adjust various things in this structure of mind beliefs to restore coherency. It isn’t that each proposition is tested separately, there are some that can’t be false and others that are; so, all are propositions and confront the world together so that we can adjust one to the other. Therefore, you can’t make a strict separation between a priori and a posteriori which just means ones that are known by experience and ones that are true and could never be revised. Quine disagreed and declared all are propositions, at least in principle, and revisable – there are some that are very unlikely to be revised that others are, but anything could be revised.
My response to this would be to go back to the point I made about Austrian economics and praxeology –it is not a philosophical discipline, it is part of science. We know in a very ordinary language sense that we act and various other truths, we have grasp on them; seemingly these are not going to be shown to be false by future experience. For instance, let’s say Quine was right and there are propositions that we think are true which could turn out to be false in another part of the universe or at a future time. Then one of his students asks, ‘Do you think it could ever be the case that two plus two doesn’t equal four?’ My answer would be: not for a long time! If Quine is right (and I’m not convinced he is), we can still say from the point of view of Austrian economics, we do know for all practical purposes that there is a truth and on the basis of that truth we can think about what’s involved in it deductively and that’s how we proceed in Austrian economics; we don’t judge whether that’s right, we only have to look at what’s involved in Austrian economics to see whether this tells you important truths about the world.
R: So, we’re not really saying that the axiomatic a priori concept of human action is necessary, that is, it has to be true in all possible worlds necessarily. We’re simply saying that it’s all that we can determine now. It’s something that we simply cannot revise – in attempting to revise the concept of human action you are what? You are a human acting; it would be a performative contradiction for a scientist to attempt to defy human action, praxeology. Is that right?
G: Yes, that seems right to me. I’m not committed to rejecting the view that axioms are true in all possible worlds where people exist but I think we don’t have to adopt that view, as you say.
R: Final thoughts from you, what recommended reading would you give if someone wanted to learn more and develop a good understanding of the Misesean a priori, and not just parrot these words, such as ‘synthetic a priori’ without really understanding the philosophical foundations?
G: I think Hans-Hermann Hoppe has a very good short pamphlet on the basic principles of praxeology titled, Economic Science and the Austrian Method. I would say of course the thing you have to read if you were going to understand it is the first part of Human Action, about the first 130 or so pages of Human Action. A lot of people find this difficult but if you wanted help to understand this I have an online course that’s available from Mises Institute which is a series of lectures I gave on this first part titled, ‘Human Action: Part One’.
R: I’m very grateful to have spoken with you today. I still don’t think I understand everything but I’m in a far better position than before I spoke to you.
G: Well thanks very much for having me, it was a great pleasure talking to you. You seem extremely well informed and I wish you a lot of success in your future.
R: Well, coming from you, that’s very encouraging! And I just want to say to the audience, I cannot see Dr. David Gordon at the moment so I have absolutely no way of determining whether he is in fact a head in a jar, but I’m sure time will tell.
G: I’m at least a jar!