Enlightenment vs Post-Enlightenment Libertarianism

“Reason is man’s only absolute,” Ayn Rand famously declared. Similarly, Murray Rothbard thought that reason uniquely vindicated the non-aggression principle on pain of self-contradiction, for “the preservation and furtherance of one’s life takes on the stature of an incontestable axiom” (The Ethics of Liberty, 33). Mises claimed that that the problems of social policy should be sought “in the same ways and by the same means that are at our disposal in the solution of other technical problems: by rational reflection and by examination of the given conditions. All that man is and all that raises him above the animals he owes to his reason.” (Liberalism, 7)

What do these three figures share? They believe in the consensus-generating power of human reason. In other words, they think that human reason, carefully exercised, will lead to moral agreement either on the nature of the good life, that the non-aggression principle is entailed by the universal value of life on pain of contradiction or on the principle of utility and its proper exercise.

It is sometimes said that the “Enlightenment” conception of reason was “confident” because Enlightenment thinkers believed that reason alone could solve our major social problems and give us true moral knowledge. Faith obscures, reason illuminates. Enlightenment-era liberalism was based on this optimism in unaided reason, and that optimism, in my view, filtered down into 20th century libertarianism. “Mysticism” as Rand and Mises called it, must be avoided. If only we can set aside our biases, ignorance and vice, we can all see that libertarianism or something like it (Objectivism) is the correct governing ideology.

Hayek, perhaps alone among libertarian intellectual heroes, was deeply skeptical of the power of human reason to secure agreement and knowledge. He constantly emphasized the fragmentation of knowledge – that we could never come to agree on a common scale of values or know enough to centrally plan the economy. Human reason, Hayek thought, is necessarily based in the mind understood as a classification system that is subject to evolutionary conditioning. Libertarians have criticized what we might call this “post-Enlightenment” conception of reason, fearing it leads to skepticism or relativism.

Significantly, the later Rawls agreed with Hayek – the free exercise of practical reason would lead us to disagree rather than agree about life’s most important questions, including about how to live together. I have elsewhere outlined Rawls’s view.

Gerald Gaus argues that over the last fifty years, liberal political theory has transitioned from an Enlightenment confidence in human reason to a post-Enlightenment view that recognizes the challenge of reasonable pluralism. The post-Enlightenment view still aspires to show that our diverse reasoning can lead us to converge on public principles that protect human freedom, but its aspirations are chastened. The fact of reasonable pluralism explains why many liberals have become public reason liberals, because treating others as free and equal requires admitting that the free use of practical reason leads in many different directions.

One reason I think libertarians have often allied with conservatives is that they wish to preserve a belief in an objective moral order. This is a key emphasis in Rothbardian and Randian thought. In a certain way then, American conservatives and libertarians share an ironic wish to preserve Enlightenment conceptions of reason, where reason leads to convergence on the right and the good. Reasonable pluralism seems to challenge this fact, since reasoning plainly does not lead to agreement.

I believe in reasonable pluralism and the post-Enlightenment’s more chastened conception of the ability of reason to generate unforced agreement. And I think that libertarians often seem authoritarian and ideological because they do not recognize these limits.

The point of a public reason libertarianism is that there are mutually justifiable norms that allow us to get along despite the fact that we will always disagree about matters of great import. Public reason libertarianism, a post-Enlightenment libertarianism, does not deny moral truth, but, recognizing the interminability of disputes about moral truth, attempts to justify moral order in some other, more interpersonal mode.

Due to embracing this post-Enlightenment critique of reason, I have become increasingly unsympathetic to Rothbardians and Randians who insist that reason will completely vindicate their moral and economic views. They prattle on about “axioms” that aren’t axioms at all, and claim that their incredibly controversial views are rationally undeniable. Such claims cannot withstand rational scrutiny and so prove to be poor foundations for a libertarianism that can speak to a post-Enlightenment world.

Only the dogmatist can insist that there is a faculty, REASON, which demonstrates that you are right and everyone else is wrong in a way that no one could honestly deny. Perhaps you libertarian readers believe, as I do, that the best reasoning, arguments and information show that our personal views are justified and perhaps uniquely so. But we would do well to recognize that others reasonably disagree.

Hayek and Rawls have taught us that disagreement is the natural result of living in a free society. And this fact should affect the way in which we attempt to justify a free society. Hayek and Rawls obviously disagreed about the nature of a free society but they asked the right questions, which is why I think they will matter increasingly more in libertarian political theory than Rand, Rothbard and Mises, as inspirational and important as they may be.

Many of my blog posts have been devoted to explaining the structure and ground of a public reason libertarianism. And I think that embracing it on the basis of a post-Enlightenment conception of free reasoning will have three salutary effects on the libertarian movement.

(1) We will be better able to recognize bad arguments when they come along, as we will be less likely to think that the case for a free society is settled among rational and informed persons.

(2) We will be able to set aside the cults of personality that plague the liberty movement by peeling away the charismatic veneer from figures who “get it” and have the obviously correct view that others can only deny via wickedness or stupidity.

(3) We can learn to speak to a post-Enlightenment political culture, one where diversity of belief and practice is a permanent fact of life, without seeming authoritarian and ideological.

  • Onomatopoeia30


  • Excellent.

  • “Hayek, perhaps alone among libertarian intellectual heroes, was deeply skeptical of the power of human reason to secure agreement and knowledge.” And that is why Hayek was the man.

  • Great post, Kevin.

    Along precisely these lines, I would like to heartily recommend to BHL readers Jerry’s excellent (and short!) book, Contemporary Theories of Liberalism, which traces precisely the development of liberalism from an Enlightenment to a Post-Enlightenment view that you discuss here. I read it a long time ago and it really changed the way I think about a lot of these issues.

  • martinbrock

    Pardon my cynicism, but moral knowledge is never true as much as it is convenient for the person knowing it. Actions can have discernible consequences, and your suffering can be a consequence of my action, but I need not call this consequence of my action “immoral”. Even Rothbard calls certain acts increasing the suffering of others “moral”, if the others violate a propriety that Rothbard asserts.

    When one lion challenges another for territory, he does not behave immorally, and when one man challenges another for territory, even if he violates a Rothbardian, non-aggression principle, he behaves “immorally” only according to Rothbard’s assertion of a morality. If I share Rothbard’s faith in a particular assertion of propriety, is the assertion then rational? What does “rational” mean in this context?

    In The Ethics of Liberty, Rothbard challenges Hayek’s assertion that a proprietor’s exclusive governance of the only spring in a desert, after other springs have dried up, is not inviolable. Rothbard counters that the proprietor’s liberty requires an inviolable claim even under these circumstances. I agree with Hayek here, but I might as easily agree with Rothbard. Pure reason does not settle the question. Either conclusion requires essentially mystical assumptions.

    Morality is not observed. It is chosen. Ultimately, a chosen faith is the only source of morality.

  • Pingback: Can Public Reason Save Us? « Pileus()

  • Jason Sorens

    Thought-provoking. Some of those thoughts here:

  • It’s entirely possible to retain most of the Enlightenment slogans, such as “Reason is man’s only absolute” (Rand), and “All that man is and all that raises him above the animals he owes to his reason” (Mises) once we have a correct theory of what reason is and what its limits are. This theory can be found in the work of Karl Popper. Once we recognise that the there are no indisputable axioms we can retain logic as “the organon of criticism”, as opposed to the “organon of proof”, as Aristotle would have it. Reason can’t demonstrate you are right, but it can, fairly confidently, indicate you are wrong.

    Personally I see this as a rehabilitation of the Enlightenment project, rather than a substitute for it. Toleration was a key tenet of (most) Enlightenment thinkers, after all.

    • martinbrock

      I agree, but Popper addresses empirical science rather than ethics. if your assumptions logically imply conclusions contradicting my experience, some of your assumptions are objectively “wrong”, but an ethical assumption is not subject to empirical falsification similarly.

      If Ayn Rand tells me that she has a moral right to force others not to copy Atlas Shrugged without her consent, how do I prove her wrong?

      If Rand claims that a copyright regime has observable effects on intellectual productivity, I might challenge the assumption empirically, but Rand may not care about the effects of her imposition on intellectual productivity. The cost of her imposition, in terms of intellectual productivity, may be the just price of her imposition, because her imposition is just. It just is, justly so.

      Ethical precepts generally seem this way to me, including my own utilitarianism. Who says that human happiness is right and any contrary experience of cows and pigs is wrong? Well, I do, because I am human and not a cow or a pig. That’s all there is to ethics, for better or for worse.

      Repeat after me. Intellectual property (and everything else called “property”) is just, right, proper, noble, moral, ethical and all the rest.

      Samuel Goldman makes the point in The American Conservative this month.

      • Popper addresses empirical science, ethics, metaphysics and just about everything else.

        One point he makes in connection with empirical science is no scientific theory can contradict your experience. A scientific theory is a statement which can contradict only other statements. When you state what your experience is, you produce something that a scientific theory may contradict. But that does not necessarily show that the scientific theory is false. It could be that you have made a mistake about your experience. This is very easily done, because to make a statement about our experience we have to interpret it, and our interpretations are fallible. Besides, it is only scientific theories in psychology that would imply anything about your experience. On general scientific theories talk about observable physical phenomena. For all this, see Chapter 5 of ‘Logic of Scientific Discovery.’

        The point of going through all that is that we MAKE scientific theories testable by agreeing to accept, tentatively, specific observation statements. What is to stop us making ethical statements testable by agreeing a procedure for tentatively refuting them? Nothing. Popper (‘Open Society,’ later chapters of Vol 2) proposed to test ethical theories according to their consequences for human suffering. Someone dubbed this ‘negative utilitarianism.’ I don’t like that proposal myself. I think we should test ethical theories according to how much they would contribute human flourishing if we lived by them. But testability is a pragmatic matter: it is for us to agree suitable procedures, just like the scientists do in the face of the fallibility of observation statements and the Duhem problem. The difference between ethics and science is not essentially one of testability, it is essentially one of its practitioners having the determination and ingenuity to find ways of testing theories.

        • martinbrock

          I’m happy enough with “negative utilitarianism” myself, but if Rand defends her copyright regardless of its effect on anyone else’s happiness, how do I prove her wrong? What makes my utilitarianism right and her proprietarianism (or your flourishianism) wrong? That I agree with Popper and you two don’t?

          I can’t agree that empirical science is essentially subjective in the same way. Yes, we must agree on a correspondence between experience and theoretical statements, but I’m convinced enough that an objective reality is out there and that reasonable men experience it similarly, so this correspondence is not a problem.

          We can distinguish Newton’s Gravity from Einstein’s by observing stars near the sun on the celestial sphere. We cannot distinguish ethical precepts similarly, and this difference seems fundamental to me.

          • You can’t prove anyone one wrong (or right). Eddington’s eclipse experiment came out in favour of Einstein and against Newton. But that verdict depends on a whole load of other theories in addition to the two under test, theories which were needed to derive statements about observable phenomena from the theories being tested. It would be possible to save Newton’s theory by revising some of these other theories instead. I suspect there are some people who have taken that course, though their research programme has probably not been very successful, otherwise we would have heard about it. This is the Duhem problem. There can be no decisive experiments.

            Like you, I think there is a spatio-temporal world out there and that whether or not our physical theories are true depends upon how things stand in that world. But we can never know how things stand in that world, because we are aware of the world only through our theories. We can never know whether our theories are true. in the face of this, some people give up all claims to know. Popper does not. He sees what the scientists do: they work out practical ways of testing theories and arriving at assessments of theories according to their explanatory value. And the history of science has shown that they actually make progress, in that later scientific theories have greater explanatory value than their predecessors. But this does not mean that any of those theories are true. For all we can know, they are all false.

            Unlike you, I think there is a moral reality out there and that whether or not our physical theories are true depends upon how things stand in that moral world. But, although we might stumble upon the moral truth, we can never know that we have done so. The best we can do is to imitate the scientists’ example: agree ways of testing and assessing theories so that we can rate ethical theories as better or worse. If we are as dogged as the scientists have been we might eventually find that we have developed moral theories which are much better than the ones we used to have. Of course, we can go down this avenue even if we think there is no moral reality out there, just as some of the old positivists recommended the way of science even though they believed there was no external world.

        • martinbrock

          @Danny Frederick
          If we incorporate all of these other assumptions, constraining the correspondence between experience and theoretical statements, into the theory being tested, we can prove the entirety of the theory wrong. We need not isolate a single false assumption to falsify the theory. We only need to establish that at least one assumption is false.

          All scientific theories are falsifiable, and standard theories almost certainly are false, because they are formal systems constrained by extremely limited observation, but they posit universal relevance.

          Human beings share a common nature, so most humans will agree on some standards of propriety. Even a person violating one of these standards will often call the violation “wrong”. In this sense, an objective human morality exists, but the most common ethical assumptions are not the most interesting, and human morality is little comfort to cows and pigs.

          One may argue, like Rothbard, that cows and pigs lack rational faculties that are axiomatic prerequisites for moral rights, but this sort of axiom is just what I expect from a morality constructed by human beings for human beings.

          As Goldman notes at The American Conservative, Lockean propriety in the context of the American frontier is the sort of standard I would expect eighteenth century, European settlers of this frontier to accept. Other persons imposing other standards in other contexts require other standards to rationalize their impositions, and so they construct them.

          • It is true that if we identified every theory used in the derivation of a prediction that conflicted with an observation statement, we would have something that is inconsistent: an extremely long conjunction, on the one hand, and an observation statement, on the other. Where do we go from there? We might take a line inspired by Graham Priest and say that this contradiction is one of the true contradictions. Or we might reject the observation statement (since it depends on fallible interpretation). Or we might reject the very long conjunction. If we do the latter, the next step is to try to identify which parts of the very long conjunction should be rejected. Different people may attack that project in different ways. If I take Newton’s ‘Principia’ and write ‘It is not the case that’ in front of it, I will in all likelihood be saying something true. But all I am saying is that at least one sentence in the book is false. That may be a beginning of wisdom, but it is no more than a very modest beginning.

            I doubt that our common human nature gives us common standards. Feyerabend and Kuhn, drawing lessons from the history of science, have argued that our concepts are essentially mercurial: they are not defined by necessary and sufficient conditions, but are bendable this way and that. Feyerabend (‘The Conquest of Abundance’) has argued that all humans have essentially the same set of concepts; but because these concepts are flexible, different cultures (social or scientific) produce different conceptual schemes from them, but we can turn one conceptual scheme into another by making intelligible modifications, in the way that the relativistic conceptual scheme emerged out of the Newtonian – though, significantly, Whorf claims that some of the American Indian tribes had a relativistic conception of space and time. Similarly, linguists have argued that such conceptual transformation is a commonplace of natural languages, in which the meanings of words and sentential structures depends upon analogical extensions (see Ortony, ‘Metaphor and Thought,’ for an illumination set of essays). Again, as Hayek argued, our common human nature gives each of us the capability to learn a natural language, but which language we learn depends upon our surroundings: different linguistic communities have different standards of linguistic propriety. The same goes for ethics, as Hayek also argued.

            But the fact that different communities have different ethical standards does not entail that they are all equally good. Some ethical standards permit the communities who have them to thrive, and others do not – again, as Hayek argued. That is why I proposed that we test ethical theories by their consequences for human flourishing.

            I think Rothbard is right to link moral rights with rational capacities. But we cannot prove that it is so. It is a hypothesis which needs a defence. It also does not imply that we have no duties that constrain our treatment of animals.

            Yeah, you can construct whatever ethical theory you like. But that does not mean that it is any good. We need a way of testing it.

        • martinbrock

          @Danny Frederick
          “That may be a beginning of wisdom, but it is no more than a very modest beginning.”

          I understand your point, but the problem is not so daunting in practice. In the case of special and general relativity, Einstein himself demonstrated the correspondence between Newtonian dynamics and Newtonian Gravity and the special and general theories of relativity. For example, he showed that the classical equation for kinetic energy (E = m*v^2/2) is a limiting case of the relativistic equation (E = dM*c^2) in the limit v << c. Newtonian Gravity is similarly a limiting case of General Relativity. We know that Einstein's theory does not contradict Newton's theory where Newton's theory agrees with experiment.

          Bohr established a principle of complementarity when refining a standard model. Progress in science doesn't simply throw out the standard model. It replaces the standard model with a variation on the theme. The choice is not between the standard model and nothing. The choice is between the standard model and a proposed reform. A reformer must demonstrate that his reform agrees with the standard model where the standard model is known to agree with observations.

          A reformer specifies a variation on the standard model and an experiment distinguishing the reformed theory from the standard model, and he also demonstrates that his reformed model agrees with the standard model where the standard model agrees with observation. If experiment then agrees with the reformed model and not the standard model, the reformed model replaces the standard model. The variation is the only reconsideration. We don't need to reconsider everything else in the model.

          If you want to reform the standard model somewhere else, to account for the reformer's experiment, then you must propose a variation on the standard model and demonstrate that your variation also agrees with the standard where the standard model agrees with observation. It's not enough to say that the reformer's variation is not the only possible one accounting for the experimental discrepancy. You must specify another one, and that's not so easy.

          • I am afraid that most of what you say here is false. New scientific theories typically correct the quantitative laws of their predecessors. Einstein’s quantitative law of gravity, for instance, modifies Newton’s inverse square law, Newton’s theory corrected both Galileo’s law of freefall and Kepler’s laws. In all these cases scientists like to show that the old theories are retained as special cases; but that is just claptrap. In addition, the new theory often employs concepts different to those used in the earlier theory, so even where quantitative laws are approximately the same, they actually mean something very different. On Einstein and Newton, see Kuhn’s ‘Structure of Scientific Revolutions.’ On Newton, Kepler and Galileo, see Duhem ‘The Aim and Structure of Physical Theory’ and Popper ‘The Aim of Science’ (a paper reprinted in both ‘Objective Knowledge’ and ‘Realism and the Aim of Science’). On the whole issue, see chapter 10 of Popper’s ‘Conjectures and Refutations.’

          • martinbrock

            Newton’s Gravity does not correct Kepler’s laws of planetary motion as much as the former generalizes from the latter. Kepler never claims to describe anything but planetary motion, and Newton’s Gravity does not contradict elliptical orbits or equal areas in equal times, ignoring gravitational interaction between the planets. The two theories are consistent in the observational domain in which they intersect, but Newton’s description is more widely applicable. Equal areas in equal times becomes the more general principle of conserved angular momentum, applicable to dynamic systems beyond planetary motion, for example.

            Einstein also generalizes Newton’s Gravity. Newton’s theory is a special case of General Relativity as a matter of mathematical fact. Assuming General Relativity, one can demonstrate that Newton’s inverse square relationship holds in specific circumstances.

            Showing that a new model incorporates an existing model as a special case is hardly claptrap. The existing standard model is the standard model precisely because it well describes many observations. Any model presuming to replace it must also describe these observations and must further describe observations that the standard model does not well describe. That’s the principle of complementarity. A new model may be very different from a standard model it aims to replace, but the model’s architect must nonetheless demonstrate complementarity if he expects to replace the standard model.

            A model means the observations it describes and nothing else. I’m a devout positivist in this respect. Special relativity assumes an invariant speed of light, but this invariance is not its meaning. Its only meaning is that the assumption implies particular observations not implied by a standard theory without the assumption. Another theory, positing a variable speed of light in particular circumstances, may replace special relativity.

          • I think you need to reconsider what you say in your first paragraph. You admit that, once interplanetary gravitation is taken into account, the motions of the planets as predicted by Newton are different to the motions of the planets as predicted by Kepler. That is, you admit that the two theories give contradictory predictions about the observable motions. Yet you also say that they are consistent in the observational domain in which they intersect. You contradict yourself.

            How could Newton’s theory be a generalisation from Kepler’s laws when it contradicts those laws?

            Similarly, how can Einstein’s law of gravity be a generalisation from Newton’s when the two theories imply different predictions about the observable motions of the planets? Recall that one of the main pieces of evidence in favour of Einstein was the motion of Mercury, which Einstein’s theory got approximately right but which Newton’s theory got wrong. Einstein’s quantitative law of gravity is a different equation to Newton’s law of gravity (though a similar one): the two are inconsistent with each other. Further Newton’s law is a law about gravitational force, all-pervasive and acting instantaneously at a distance. In general relativity there is no gravitational force. Einstein’s law of gravity and Newton’s law of gravity talk about different things.

            You say: “Any model presuming to replace it must also describe these observations and must further describe observations that the standard model does not well describe.” But this is in fact not what happens. Newton’s theory predicted observable motions which contradicted the observable predictions of Kepler’s laws and also of Galileo’s law of freefall. That is one reason why Newton’s theory was superior. Similarly, Einstein’s theories predict observable motions different to the predictions of Newton’s theory, such as Mercury’s perihelion. Such differences help to decide between competing theories. Popper made much of this point: a better theory does not just explain what its predecessor did and something else besides; it also CORRECTS the earlier theory, i.e., explains the earlier theory’s inconsistencies with the observable facts.

          • martinbrock

            “You admit that, once interplanetary gravitation is taken into account, the motions of the planets as predicted by Newton are different to the motions of the planets as predicted by Kepler.”

            Accounting for interplanetary, gravitational attraction, within Newton’s theory, implies orbits deviating from Kepler’s laws, but Newton did not develop his theory because he observed planetary orbits deviating from Kepler’s laws. He developed his theory because it was consistent with Kepler’s laws ignoring interplanetary interactions.

            “That is, you admit that the two theories give contradictory predictions about the observable motions.”

            “Contradictory predictions” is a stretch here. Kepler never presumed to predict planetary motion in the context of interplanetary, gravitational interaction. He only reduced Brahe’s observations to a more compact form, and Newton proceeded from this more compact description. If Kepler imagined himself discovering inviolable laws, he was clearly mistaken, but this fact doesn’t concern me. Scientists may imagine their enterprise this way without contradicting my assumptions.

            “Yet you also say that they are consistent in the observational domain in which they intersect. You contradict yourself.”

            No. Newton didn’t have any observation of planetary motion contradicting Kepler when he formulated his Gravity, as far as I know, and he didn’t need any.

            Newton didn’t know about the precession of the perihelion of Mercury at all as far as I know. Neither did Kepler. Neither did Brahe. Of course, they didn’t develop any theoretical description of something outside of their experience. How could they? Even if orbital precession was observed at this time, Newton was long dead before anyone proposed that Mercury’s precession is inconsistent Newton’s Gravity. Newton himself never did.

            “Similarly, how can Einstein’s law of gravity be a generalisation from Newton’s when the two theories imply different predictions about the observable motions of the planets?”

            My “generalize” describes only the principle of complementarity. Your “generalize” seems to denote something else.

            Einstein’s Gravity implies Newton’s Gravity in particular circumstances, and these particular circumstances are the observational domain in which Newton developed his Gravity. Einstein knew of Le Verrier’s work on the perihelion of Mercury, and Newton didn’t.

            “… the two are inconsistent with each other.”

            If Einstein’s theory were completely consistent with Newton’s, Newton’s theory would still be the standard model.

            “In general relativity there is no gravitational force.”

            Curved space-time replaces a “gravitational force” per se in General Relativity, but the principle of complementarity applies nonetheless, because the principle only addresses observational implications of the theories, not their ideal forms. The ideal forms are useful fictions.

            “Further Newton’s law is a law about gravitational force, all-pervasive and acting instantaneously at a distance.”

            Newton didn’t like action at a distance, but he had no reason to incorporate a speed of propogation of his gravitational interaction, because the inverse square law accounts for Kepler’s laws ignoring interplanetary attractions, and Kepler’s laws were the compact form of the best observations of planetary motion available.

            “Newton’s theory predicted observable motions which contradicted the observable predictions of Kepler’s laws …”

            What planetary motion, observed by Newton, violating Kepler’s laws did Newton’s theory predict? Later observers could apply Newton’s theory this way, and Newton understood that his theory could be applied this way, but he didn’t formulate the theory to contradict Kepler. He formulated it to be consistent with Kepler.

            Kepler’s laws are not predictive at all. They are descriptive. They summarize what Tycho Brahe observed.

            Newton could have formulated a theory essentially to contradict or to correct Kepler, as Einstein reacted to the precession of Mercury’s orbit, but he didn’t as far as I know.

            “Popper made much of this point: a better theory does not just explain what its predecessor did and something else besides; it also CORRECTS the earlier theory, i.e., explains the earlier theory’s inconsistencies with the observable facts.”

            If one theory describes observations that are a superset of the observations described by a second theory, the first theory is “more correct” in this sense than the second. Apart from some reactionary dark age, observations grow monotonically. Later theoreticians always have more observations with which to constrain their theorizing. This fact does not contradict any assumption of mine. I only prefer “more descriptive” to “more correct”. “More correct” sounds too much like “more pregnant” to me.

          • Several times you refer to the reasons why Newton developed his theory. But these are not relevant to our dispute. We are talking about whether Newton’s theory is inconsistent with Kepler’s and with Einstein’s. You are just changing the subject.

            What you say about Kepler is false. He was not merely reducing Brahe’s observations to a more compact form. He offered an a priori Pythagorean justification for his laws. He was a mathematical mystic, not a positivist. See, e.g., A Koestler, ‘The Sleepwalkers,’ or T. Kuhn, ‘The Copernican Revolution.’

            What you say about Newton is false. Kepler’s predictions of the planetary motions were only an approximate fit. It was generally known in 1686 (i.e., prior to the publication of Newton’s ‘Principia’) that the planets did not move in ellipses. Indeed Kepler himself knew as early as 1625 that Saturn and Jupiter do not move in ellipses. See I. Lakatos, ‘Newton’s Effect on Scientific Standards,’ in Vol 1 of his ‘Phil Papers.’

            I don’t see why you are talking about what Newton knew of Mercury’s anomalous motions. We are talking about the logical implications of Newton’s theory compared with Einstein’s theory. The two theories predicted different things and Einstein’s were a better fit with the facts. Therefore it is false that, as you claimed, Einstein’s theory is a generalisation of Newton’s, for if that had been so, the former would have entailed the latter, not contradicted it. This, it seems to me, is true on your ‘principle of complementarity’ interpretation of generalisation.

            You insist that Einstein’s theory of gravity implies Newton’s yet you admit that they are inconsistent with each other. Logically, that can be so only if Einstein’s theory is inconsistent. But you are not saying that. So you contradict yourself again.

            If you interpret Newton’s gravitational force and Einstein’s curved space-time as mere fictions, you are ignoring what the theories actually say and are talking about something else. But even your re-interpreted theories are inconsistent with each other because their observational implications are inconsistent. That is why the perihelion of Mercury is regarded as supplying a crucial test between them.

            Your last paragraph appears to have missed the point of much that I have said. Observation statements get revised. This should be obvious just from the improvement in telescopes. But there is more to it than that. For example, the Astronomer-Royal, Flamsteed, kept reporting to Newton observations that contradicted Newton’s theory; to which Newton’s response was to teach Flamsteed about the refractive properties of the atmosphere, so that he could correct his observations in the light of this interference (see Lakatos, ibid.). Similarly, Einstein’s theory of light bending also means that previous observations had to be re-interpreted. Thus, even if Kepler’s ellipses had exactly fit all the observations (which they did not), they would still have implied false observation statements when the past observations were reinterpreted in the light of new optical theories. A better theory does not just incorporate the observations of the old theory and add some new ones; it revises some of the old observations.

          • martinbrock

            A dispute you imagine is evolving, but I’m not changing the subject. I’m addressing the same subject I always addressed. You address a different subject and insist that I also address this subject.

            Of course, Newton’s theory describes planetary orbits more precisely that Kepler’s laws. This fact has never been a controversy.

            What I say about Kepler is true. What Kepler imagined himself doing is a separate issue. Kepler reduced Brahe’s observations to a more compact form, even if he imagined himself discovering universal laws. His compression of the observations was not lossless. If it had been, it would have useless to Newton.

            What I say about Newton is true. Kepler’s description of planetary orbits in terms of three, relatively simple principles enabled Newton to theorize a Gravitational force without dealing directly with the complex measurements from which Kepler developed the principles.

            A planet under the influence of a Gravitational attraction from a point source does move in an ellipse about the source. That observed orbits weren’t perfectly elliptical is not relevant. Newton didn’t develop his Gravity to explain deviations from Kepler’s elliptical orbits. He developed a theory consistent with elliptical orbits in the two body scenario. He could then surmise that interplanetary interactions explain deviations from perfectly elliptical orbits.

            I talk about what Newton knew of Mercury’s orbit, because we discuss the observations constraining Newton’s theories. Of course, Einsteins’ theory fits the facts better. He had more facts to fit.

            Einstein’s theory is a generalization of Newton’s in precisely the sense that I describe above. This “generalization” does not imply that Einstein’s theory never contradicts Newton’s theory, only that Einstein’s theory is complementary in the sense of Bohr. I make this point repeatedly.

            I never anywhere suggest that Einstein’s theory “entails” Newton’s theory. Only you say it. You say that I say all sorts of things that I never say. The habit impedes communication.

            Again, Einstein’s theory implies that Newton’s theory describes interactions where massive bodies are much larger than their Schwartzchild radius. Assuming Einstein’s theory, one can easily understand why Newton’s theory describes planetary motion as well as it does. In other words, Einstein’s theory complements Newton’s theory in the sense of Bohr. This fact does not imply that Einstein’s theory is the Truth. It only implies that Einstein’s theory does not contradict Newton’s theory where Newton’s theory is already known to fit observations.

            Again, special relativity implies that the classical equation for kinetic energy gives practically the same value as the relativistic equation when a body moves with a speed much less than the speed of light, so we can accept the relativistic equation without contradicting successful application of the classical equation in the non-relativistic domain. We don’t need to repeat all of the non-relativist experiments that classical dynamics well describes, because we know that special relativity is complementary.

            I don’t say that a gravitational force and curved space-time are “mere fictions”. That’s you again. I say that they are “useful fictions”. I don’t ignore what the theories are saying. I don’t reinterpret any theories. I consistently acknowledge that the theories describe different observations where they diverge. Again, this divergence is the whole point of the principle complementarity.

            Observations are continually revised, both by improved measurements and in terms of new theory. I don’t miss this point.

            A new theory may account for observations that an older theory successfully described without repeating every experiment against the new theory. That’s the point of complementarity.

            We can reinterpret observations in terms of a new theory, but theories don’t revise observations. The bright spots on Eddington’s plates are where they are. Theory tells us how these observations correspond to useful fictions. As a mere mortal, I don’t pretend to know how the fictions correspond to universal Truth, but contemplating the fictions is nonetheless satisfying.

      • Jay_Z

        Intellectual property seems a curious place to make an axiomatic stand. A pre-Gutenberg Rand would have been begging people to make copies of her work. Copies cost hundreds or thousands of dollars/man-hours of work, they weren’t a source for profit. Intellectual property laws were a post hoc decision made by society after Gutenberg made making cheap copies possible.
        The Internet has changed the publishing world again. Publishing in the form of a blog has dropped the cost again. A Rand of today would need to blog to be heard, most likely. The barriers to entry of publishers and editors are going away.
        Making money with your content is another thing. Put your content behind a paywall, and 99% of the time people will just read someone else. Publishing is cheap, words are cheap, words are hard to sell today, intellectual property or no.

        • martinbrock

          Rand may make curious axiomatic stands where she likes, and if she persuades the brute squad to stand with her, I must stand aside. I see little else in anyone’s ethical assertions. They rationalize forcible imposition, and that’s all. That’s the point.

          If some supernatural force somehow banishes forcible imposition from human interactions (if that even makes sense), we stop talking about ethics altogether, because we have no more use for it.

  • Pedro Pereira Hors

    Great article, Kevin. I find myself reacting the same way you do to towards Rothbard’s and Rand’s “axioms”. Mises is different, I think. I don’t think he favored an axiomatic approach to ethics. His emphasis was in social cooperation. That which preserves it is good. That which doesn’t, isn’t. I see him closer to Hayek in this, in spite of what he said about reason. See Hazlitt’s “The Foundations of Morality”, which attempts to develop an ethical ethics based on Mises’ insights on the subject.

    See this, for example:

    “Man is inconceivable as an isolated being, for humanity exists only as a
    social phenomenon and mankind transcended the stage of animality only
    in so far as co-operation evolved the social relationships between the
    individuals.” (“Socialism”) So, “co-operation”, which he sees as the ultimate criterion for moral rules, evolved, that is, it wasn’t the product of someone coming up with a “rational” social system.

    Also, aren’t Rawls’ and Hayek’s approach different? Isn’t Rawls guilty of the same extreme rationalism as Rothbard and Rand? Doesn’t Hayek tend a little to a mild form of conservatism that plays down the role of reason and stresses the importance of cultural evolution?

    I recently read a good article by Matt on the subject:

    • Kevin Vallier

      Rawls and Hayek certainly differ, but the later Rawls is not entirely guilty of that sort of rationalism, in my view. There’s a bit in Political Liberalism held over from A Theory of Justice, but not a lot.

  • Helgard

    Dr Vallier, how familiar are you with the work of the “realist” movement in political theory? The likes of Raymond Geuss, John Dunn, William Galston, Judith Shklar, Bernard Williams, Stephen Elkin etc. (it is a diverse camp) that argues for a “politics-first” as opposed to “ethics-first” approach to political theory. They are very much concerned with empirical facts, historical context, role of passion and emotions, skeptical of liberal ideas around consensus and legitimacy and view politics as a separate sphere from moral philosophy – very much stressing the inherent nature of conflict and disagreement. They are mainly critical of Kantian ethic first approaches associated with the likes of Rawls and Nozick. (Not doing them justice in such short space)

    I ask, because my own current D project deals with answering the realist critique of modern liberalism by looking at Hayek’s brand of classic liberalism. These realist seem to focus almost exclusively on the Rawlsian, Dworkin and Nozick brands – and ignore the Humean Scottish Enlightenment / Austrian inspired liberalism of Hayek, specifically his social theory.

    Any way, was just curious to know if these “realists” has had an impact on some of the debates within the BHL circle and more specifically the public reason route?

  • Dan

    Hi Kevin —

    One of the big worries that I think a lot of people have had — certainly I have — with public reason approaches is that it seems very hard to come to any kind of convergent set of laws on the basis of the reasons people actually have. This is because the operative notion of “the reasons people have” is an internalist one, making reference only to the individuals’ subjective motivational sets or evaluative standpoints (their moral beliefs, commitments, desires, etc) — not anything “external” like, e.g. objective moral facts.

    As far as I understand it, the preferred response to this kind of worry is to rule out the influence of individuals holding views that might have the result of eliminating huge numbers of attractive laws, because those views are “unreasonable”. Certainly this is the kind of response in Political Liberalism — there Rawls rules out views that, say, deny the equal worth of all individuals; and my impression from a skim-read of Gaus’ new book (as well as his part in the Cato Unbound exchange a wile ago) is that he takes a similar line.

    OK: so now it seems like you need to rule out a bunch of at least possible and coherent views as unreasonable. But in doing this, aren’t you just being a dogmatist? Aren’t you just insisting that there is a faculty, REASON, which demonstrates that you are right and
    everyone else is wrong in a way that no one could honestly deny? Sure, you might think that reason speaks less univocally than die-hard partisans of particular views (e.g. Rothbardians) do, but aren’t you still just making the same move at the end of the day, albeit with a slightly larger in-crowd?

    • Kevin Vallier

      Dan, actually the skim-read of Gaus won’t do. He tries to avoid insofar as he can the term “reasonable.” In fact, Gaus criticizes political liberals for appealing to reasonableness on a number of grounds. Further, the word “reasonable” only occurs eight times in the book and it is in quotes every single time.

      Now, Gaus wishes to idealize to ensure that public justifications aren’t based on bad information and reasoning, but he argues that conceptions of full rationality and information face severe problems (see Section 13 for the critique and the positive view) so the relevant idealization is merely modest.

      Finally, Gaus does not insist that agents see themselves as autonomous in any robust or thick sense of the term (even Rawls’s political autonomy). So long as persons are autarchical and have the moral emotions, they’re members of the public.

      So the in-crowd on the Gausian forms of public reason that I prefer is huge. Yes, reason will delimit a set of proposals are potentially justified. But the Post-Enlightenment view never meant to denigrate reason completely, but rather to seriously chasten it.

      Incidentally, I’m working on a paper write now about whether the Hayekian elements Gaus employs in his work imply a view of reason that is *too* chastened, so if anything I think Gaus errs on the side of too much inclusion.

  • I agree with a great deal of what you say here, but not all of it. A sustained critique of dogmatic rationalism (both empiricist and ‘rationalist’ in the narrow sense) can be found in the works of Karl Popper. To put it crudely, the enlightenment replaced religious dogmatism with the dogmatism of reason. But, for Popper, the mark of rationality is a critical mindset, a preparedness to hold all one’s views open to question. Hence he calls himself a ‘critical rationalist.’ He explains some of this in his paper ‘On the Sources of Knowledge and of Ignorance,’ which is available online in two parts, here:

    and here:

    His views developed out of a criticism of the positivists and originally concerned scientific knowledge, but he then generalised them. What he showed, in most detail in connection with science, is that it is possible to reject all justification and still find a path to objective knowledge, thereby avoiding relativism and Pyrrhonian or Academic scepticism.

    That last point hints at where I disagree with you. For the point that it is impossible to justify anything means that reasonable disagreement is always possible. Indeed, for Popper it is always welcome. As he puts it: ‘orthodoxy is the death of
    knowledge, since the growth of knowledge depends entirely on the existence of
    disagreement’ (‘Myth of the Framework,’ p.34); and ‘There is a place, and a
    function, within the critical method of science, even for the lunatic fringe’ (‘Realism and the Aim of Science,’ p. 70).

  • Fallon

    It is misleading to equate moral and economic reasoning. Besides, Mises added modern post-18th century economic thinking– marginalist understanding, subjectivist value, and the propertarian basis of economic calculation- to his analysis of policy. Rationalism is not frozen in stone. Mises also had much to criticize about the Enlightenment, and knew that majorities could choose wrong ideas, such as a belief in inevitable divine inspired Progress.
    Btw, it was Hayek, your champion besides Gaus, that had good things to say about Pinochet. Further, he was pro conscription (slavery).
    It is true that Rothbard teamed up with some real jerks, racists and anti-market folks (Sam Francis,…). Rothbard was sadly racist– but I believe those ideas he subordinated to an anarchic propertarian outlook.
    My question is whether public justification is superior to property as means just society. But to get at an answer there must be correct presentation of ideas and people.
    Mises on the Enlightenment:
    “There is no room within a system of praxeology for meliorism and
    optimistic fatalism. Man is free in the sense that he must daily choose anew
    between policies that lead to success and those that lead to disaster, social
    disintegration, and barbarism.”

    • good_in_theory

      What property is there prior to public justification?

      • Fallon

        What makes a public? Two people working out rules to foster cooperation? Or every sentient human being now living?
        Existence affects every one else. This is an inescapable fact. Should everyone stop what they are doing- right now- and hold a big meeting so all of the political philosophers and politicians can tell them what they should do? Human kind would be doomed. Nobody is allowed to act unless it is PJ’d? How about PJ’ing the PJ imperative in the first place?
        How do we know that the PJers aren’t just trying to round up all the people wearing glasses and have them taken out back and shot? Khmer Rouge, Pol Pot i.e..

        • good_in_theory

          Non responsive.

          • Fallon

            Is this submittal of claims for review before or after I take previously unclaimed land? And submittal to who or what?

          • good_in_theory

            How did you know it was unclaimed? Why is submitting your claim from review distinct from making the claim in the first place? Do you just wander around assuming things are unclaimed and then claiming them for yourself, all in the privacy of your own head?

          • Fallon

            I was assuming “previously unclaimed land” to make a point. But it is good you altered the empirical scenario. Rival claims of varying natures are distinct possibilities requiring contingencies. Native peoples with no formal deeds could show up and relate stories of how their ancestors used this land for primary winter hunting and fishing. Now what? At any rate, would conflict be solvable without the convention of property in the first place? Can two people occupy the same time/space moment? And still the question emerges, PJ to who and what?

          • good_in_theory

            Where do you get your convention of property? You don’t get from possession to property without passing through justification to others.

  • TracyW

    I’d +1 this if my Google account was working.

  • They prattle on about axioms that arent axioms at all … such as?

    • Kevin Vallier

      The non-aggression axiom, for starters.

      • This is most definitely not an axiom for Objectivism and is why Rand vehemently opposed Libertarians.

  • Aeon Skoble

    Kevin, are you not throwing out the baby with the bathwater? Fine, so you think Rand and Rothbard are overly confident about their ability to axiomatize their systems – how is that a critique of reason? You say “Only the dogmatist can insist that there is a faculty, REASON, which demonstrates that you are right and everyone else is wrong in a way that no one could honestly deny” Well, then color me dogmatic: without reason, we cannot use reductio ad absurdum arguments, or for that matter any arguments about anything. I don’t know why you have the caps-lock on when you type “reason,” but if the point was to suggest that we don’t have such a faculty, well, as Aristotle notes, that’s self-refuting. If your point is that enlightenment thinkers sometimes overestimated their ability to axiomatize and construct whole systems from the armchair, fine, I’ll buy that- that’s a critique that plays well into Hayek, but I don’t see how it plays into Rawls. Parse the distinctions between TJ and PL all you like, but it’s still Kantian, which means he’s doing what you seem to be criticizing. There is reasonable disagreement about the good because the human good is pluralistic in nature, but that’s not a reason to be skeptical about reason.

    • Kevin Vallier

      My point was epistemological, namely that the free exercise of practical reasoning leads to systematic disagreement about what is right and good. I think there’s an Enlightenment idea of Right Reason that leads inexorably towards truth, so to reject that there is such a faculty is not to reject the reliability or importance of reasoning as such, but rather that there is no one reasoning faculty that leads, through its natural exercise, inexorably towards truth.

      Further, and I didn’t bring this out, Hayek thought our reasoning faculties were modular, that is domain-specific, so in that sense there is no one faculty of reasoning either.

      I wasn’t trying to take on Rawls’s entire political liberalism with all its bells and whistles, but rather to speak in favor of public reason views generally.

    • I don’t think we should deny that we have a faculty of reason, but we should admit that it is fallible. Even some reductio ad absurdum arguments may be invalid. Recall Russell’s paradox. The set of all sets that are not elements of themselves is either an element of itself or it is not. If it is, it is not. If it is not, it is. This is self-contradictory. Russell concluded, by reductio ad absurdum, that there can be no such set. He then adopted a theory of types which prevented such a set from being talked about in his regimented language. Graham Priest says, very plausibly, that this is ad hoc. His counter-proposal is to admit that some contradictions are true, and he argues for this on the ground that the system of logic that he contrives (in which some but not all contradictions are true) avoids Russell’s ad hockery and is simpler and more elegant to boot. I don’t pretend to know who is right about this. Russell and Priest are both expert logicians who would make mincemeat of me. But what this dispute shows is that even something as fundamental as the law of non-contradiction can be reasonably denied. Further, denial of it does not mean that we give up arguments. Priest incorporates some contradictions into his logic (which is a theory of the validity of argument), and he argues for his logic. So admitting that REASON is fallible does not entail giving up reason. It just means that there are no certainties, literally none.


        How about this for a candidate: from the (apparent) fact that I am now experiencing thinking/feeling, thought/feelings exist (a la Descartes)? Note, I make no claim about who/what is experiencing thought/feelings.

        • That is a good one, Mark. As you may know, Kant argued against Descartes that I cannot even know that I think, for there may be no enduring thing that is me; and Hume of course claimed that he could never find himself. So, the ‘I’ bit is questionable, but what about the ‘think’ bit? Or, as you say, feelings.

          Well, there is a metaphysical view known as eliminative materialism, which is advocated by the two Churchlands and was, I think, originally proposed by Paul Feyerabend and taken up by Richard Rorty. On this view no mental states exist, so there are no thoughts and no feelings. The only reason we think there are such things is that we have a biologically inherited conceptual scheme which, though useful to us in our particular environmental niche, actually misrepresents the world. Just as we laugh at the mediaevals and their ‘witches,’ and those tribal people with their animistic theories of the world, so future generations will laugh at us and our ‘mental states.’

          I doubt that eliminative materialism is true. But it might be!

          • MARK_D_FRIEDMAN

            Hi Danny,
            Thanks for the response, and, as always, a pleasure to hear from you. The problem I have with the skeptical response you describe is that having the sensation of thought or feeling seems to me to exist quite apart from how we describe it or whatever cause we ascribe it to. So, while it seems possible (if implausible) to deny that that “mental states” exist, it does not seem possible to deny that the sensation (perception?) of such states exist (even if “mistaken” or “confused”). So, perhaps I should restate the claim to be simply “sensation exists.”
            Putting aside the specifics of this case, knowing you as I do, it will already have occured to you that your statement: “It just means that there are no certainties, literally none,” will itself be subject to the same qualification. Thus, what do we do in cases like the above (or this example, “1 = 1”) when there is reasonable disagreement about whether the claim is certain or potentially subject to some future falsification or qualification)? In other words, why can’t I proceed on the assumption that 1 = 1, since it might not be the case that “there are literally no certainties”?

          • It is always a pleasure to hear from you, too, Mark.

            Yes, your response to the eliminative materialists is the usual one. Note that one of my own sentences begins: ‘The only reason we think there are such things [as mental states].’ That is self-contradictory as it stands, because I am saying there are no thoughts, we only THINK there are. But that is just because I am using our inherited conceptual scheme to claim that it is wrong. Once we get a new conceptual scheme from developments in neuroscience, I will, when using that scheme, no longer be talking about thoughts. I will instead be talking about physical processes that benighted people used to call ‘thoughts.’

            I don’t think there is a sensation of having a thought. But there seems to be, in Nagel’s phrase, something that it is like to have a thought. So your claim would be that eliminative materialism would eliminate the ‘what it’s like;’ but, since there is something that is what it’s like, the materialist will have some fact missing from his worldview. I think the materialist would say that you are begging the question. How do we know there is something that it is like to have a thought? It seems to us that there is because we operate with a particular conceptual scheme. But if that conceptual scheme gets replaced by a better one, we may come to see that, in fact, there never was anything that it is like to have a thought (just as there never was anything it is like to be a witch).

            As I said before, I am not endorsing eliminative materialism. But it does not seem to me to be obvious that it is false. For example, who in 1850 could have thought that simultaneity would turn out not to be absolute?

            Indeed: nothing is certain (not even that). The Pyrrhonians made that reflexive point.

            I think you should proceed on the assumption that 1 = 1. To insist that it is uncertain is not to deny that it is the best we have for the time being. This is the difference between Popper and the Pyrrhonians. The latter thought that nothing is certain and that therefore we could know nothing; and they tried to obtain solace from the relaxation possible once all the striving after truth had been given up. Popper, in stark contrast, demands no rest: we can rate theories as better or worse (in terms of explanatory merit), and we can try to improve theories or to create better ones; so we should constantly strive for better and never rest content with what we have achieved. If someone claims that a particular theory is justified, show them that it is not, either by refuting it or by replacing it with a better theory,

            Recognising that everything is uncertain (including that) does not prevent us from making progress or from holding those views which have so far stood up to criticism better than their rivals.

          • TracyW

            But if that conceptual scheme gets replaced by a better one, we may come to see that, in fact, there never was anything that it is like to have a thought

            No, we wouldn’t. If such a conceptual scheme is created, then we won’t be seeing anything at all.

          • MARK_D_FRIEDMAN

            I (quite literally) cannot conceive of how the Law of Identity could be falsified. Among other things, this would seem to make rational thought impossible, because the concept or thing that I have in mind for one thought would not necessarily be the same concept or thing for the next thought. So, I have no choice but to proceed on the assumption that the Law of Identity is not only true, but certain. If this turns out not to be the case, then I will happily eat these words. But, since the “I” eating the words and the “words” themselves may not be the same ones referenced here, this may turn out to be a rather pleasant experience. 🙂

          • You kinda seem to concede my point here. You say that you have to proceed on the assumption that the law of identity is true, but then you go on: ‘If this turns out not to be the case…’ For your reasoning purposes, you don’t need to know for certain that the law of identity is true. All you need is that it is currently accepted as better than its rivals, at least in the arguments or topics in which you assume it.

            But even if the law of identity is false, that does not entail that every instance of the law is false (only that some are), so it need not entail that MDF is not MDF.

          • TracyW

            The ‘eliminative materialism’ argument is one that the more convincing someone thinks or feels it, the less they should be convinced by it.

          • I am sympathetic to these objections. But the fact that it is difficult for us to conceive, here and now, how eliminative materialism could be true does not show that it cannot be true. As I said to Mark, going back to 1850, who could have conceived that simultaneity is not absolute?

          • TracyW

            But if it is true, then we can’t have thoughts or feelings like that it’s difficult to conceive how it could be true. This strikes me as a pretty fundamental difference between simultaneity and mental states (at least my mental state, I might be imagining you.)

          • “But if it is true, then we can’t have thoughts or feelings like that it’s difficult to conceive how it could be true.”

            That is correct. But it is not much of an objection because you are presupposing the correctness of a conceptual scheme that the eliminative materialist expects to be replaced in future. That is, we are all talking about metal states because that is part of the conceptual scheme we currently use to talk about these issues, just as scientists used to talk about phlogiston before the discovery of oxygen. They could not back then talk about oxygen because they did not know there was such a thing. In a similar way, the eliminative materialist may maintain that a future neuroscience will give us the concepts and terms to talk about the things we currently identify as mental states so that we can eschew mentalistic terminology altogether. But we are not there yet.

            This is of course merely promissory. But it does seem to me to be plausible as a possibility.

          • TracyW

            but your analogy isn’t close enough. Physics explains to us the every-day world, rather than explaining it away. Yes, oxygen has replaced phlogiston, but chemists still acknowledge that fire exists. Oxygen is a more-sophisticated explanation of the every-day view of fire. The theory of relativity explains the appearance of simultaneity by that we are moving so much slower than the speed of light. It doesn’t try to claim that if I say “Fred and Alice spoke simultaneously” that I’m wrong about the appearance of simultaneity. Or take objects – physicists state that everything that seems solid to me is really mostly empty space, and that the reason I can sit on a chair rather than fall through it is that electromagnetic fields emitted by the atoms do the job that at the macro-scale appears to me that it’s being done by solidity.

            The eliminative materialists are claiming not that our mental states will be explained as the result of more subtler causes than we previously thought, but that our mental states don’t actually exist. Chemists don’t claim that fire doesn’t actually exist, and physicists still sit on chairs and turn on the TV expecting to catch the start of The Big Bang Theory. So what eliminative materialists are talkin gabout is in an important way not similar to phlogiston or simultaneity.

          • MARK_D_FRIEDMAN

            You are expressing my objection to Danny’s argument far better than I did; thanks.

          • TracyW

            I’m blushing, thank you.

          • Why isn’t this similar to complaining that the theory of oxygen does not explain phlogiston, it explains it away; or that theories of mental disturbance do not explain demonic possession, they explain it away; or that Newton’s theory does not explain celestial spheres, it just explains them away; or that relativity does not explain the ether, it just explains it away; or that physical explanations for thunder do not explain godly anger, they just explain it away?

            The eliminative materialists (so far as I understand them) are saying that we will in future be able to explain people without reference to mental states, so that metal states will be explained away.

          • MARK_D_FRIEDMAN

            But Danny, even if it is true that we can explain people without reference to mental states, how do we explain away the existence of sentience (or consciousness), whatever its cause or source? To deny it, or to even think about it, is to confirm that it exists. Consciousness mighy be an artifact of some sort that we don’t now understand, but it exists. This is what must be explained away, and I can’t see how this is possible.

          • Imagine someone a few centuries ago asking: ‘how do we explain away the existence of demonic possession,
            whatever its cause or source? To exorcise it, or to even think about it, is
            to confirm that it exists. Demonic possession might be an artifact of some
            sort that we don’t now understand, but it exists. This is what must be explained away, and I can’t see how this is possible.’

            Not an exact parallel, I agree. But surely close enough?

          • MARK_D_FRIEDMAN

            I recognize that your last question is rhetorical, but I would still respectfully answer “no.” The debunkers of demonic possession were not attempting to explain away, i.e. to deny, the experience that its “victims” described as demonic possession, they just discredited its purported source. Back then, people did indeed go crazy, have hullucinations, etc. Here, the burden is on you to debunk the very existence of experience itself, without respect to its nature (how we choose to characterize it). The sole issue is whether experience, consciousness, or sentenience undeniably exists.

          • But the debunkers of demonic possession were not trying to explain away experiences: they were only trying to explain away demonic possession. And they succeeded (so it seems to be generally, though not universally, agreed). But just as they succeeded in explaining away demonic possession, so neuroscientists will eventually explain away experiences – or so the eliminative materialists maintain.

            It is true that the eliminative materialist research programme is in an early phase, because we do not yet have the detailed theories that will explain human behaviour AND explain away mental states. However, once we get those theories it will be easy to see how there are no mental states. In the meantime the eliminative materialist can only draw parallels with other scientific revolutions.

          • MARK_D_FRIEDMAN

            My last comment on this. For sake of argument only, let me grant that it is possible that, as you say, that “once we get those theories it will be easy to see how there are no mental states.” It is also possible that this research program (and all similar ones) will fail (as I believe you will acknowledge). Since I am convinced that the very idea of a research program (conducted by sentient human beings) to disprove the existence of sentience is a contradiction in terms, why am I not entitled to proceed on the basis that I have certain knowledge of consciousness? It seems clear to me, based on the above, that I have at least the possibility of being correct in my certainty, and that there is no basis for saying that the eliminative materialists are certainly right. At this point aren’t you insisting on mere semantics, i.e. we shouldn’t call anything “certain,” even though I may hold a view that is never falsified.

          • You are entitled to proceed on any basis you like, so long as you are prepared to defend your position against criticism with counter-arguments. You may even, in this case, legitimately (it seems to me) claim that your position is superior to that of the eliminative materialist, since he admits that he has all the work to do and is at present insisting only on the possibility that he will turn out to have a successful research programme. Why not leave things at that? Why do you feel that you have to add that you feel certain in your position? Or, more strongly, that you “have certain knowledge of consciousness”? What is the point of that?

            You go on to say: “I have at least the possibility of being correct in my certainty.” Does that even make sense? If you were (really) certain, why would you want to insist on ‘the possibility of being correct in my certainty’? To admit that you may be wrong is to admit that you are not certain. That is how we use the word ‘certain.’ You cannot dismiss the correct use of words as ‘mere semantics.’ You remind me of something Russell once said: ‘a pedant is someone who prefers his statements to be true.’

            I agree that it is not the case that the eliminative materialists are certainly right. I even agree with you that, in all likelihood, your view will never be falsified. In fact, I think we agree on all matters of substance here. But you want to hang on to the claim that you know some things for certain. Why is that? Why should that be important?

          • TracyW

            Um, that phlogiston, demonic possession, celestial spheres, ether and godly anger are hypotheses, not observations? While flames, and lights moving in the sky (stars, planets, etc), and thunder are observations? (Mental disturbance and ether are/were hypotheses, I think mental disturbances more likely to exist than ether, but it’s still a hypothesis).
            Of course, in some senses, flames, stars, planets, comets thunder are also hypotheses – I could be a brain in a vat being fed stimuli that makes me falsely believe that there are flames, stars, etc. But if we allow my senses the benefit of the doubt then flames, etc have supporting evidence that phlogiston, etc doesn’t.
            If the eliminative materialists are saying that in the future we will be able to explain people without reference to mental states, that is one thing. I’m somewhat doubtful of the hypothesis but I see how it could happen. But I don’t see how that will result in explaining away mental states any more than the theory of oxygen results in explaining away flames. Chemists can still light fires.

            And as Descartes pointed out (or I imagine that such a person did), the evidence to me that right now I am thinking is rather stronger than the evidence that I could be burnt by a flame.

          • I think you are assuming too hard a distinction between hypotheses and observations. All our observations are informed by theory. Indeed, the signals we get from our senses give us remarkably little, and very confusing, information. We only manage to make any sense of it at all because we come to observation equipped with an interpretative framework which imposes a largely preconceived interpretation on the signals. Hayek makes this point in ‘The Sensory Order,’ but it is a commonplace of cognitive psychology and has been for decades. I think the first philosopher to make a big thing of it was Kant. He was well ahead of his time.

            Our interpretative framework is partly inherited biologically and partly inherited culturally; but it is also open to us to change it. Our ancestors, it seems, actually heard thunder as godly anger: that was an observation for them. Perhaps some people (thought they) SAW demonic possession. Similarly, what is for us an introspection of a mental state may be a product of our inherited conceptual framework. If we get a new conceptual framework, we may see whatever we now see as a mental state as something else instead.

          • TracyW

            I don’t think I did assume that. Indeed, I explicitly stated that in some senses flames, stars, planets, comets, thunder are also hypotheses. That was me briefly referring to that the “all our observations are informed by theory” thing that you outlined, as I thought my comment was long enough already. I will now say that I explicitly agree entirely with what you say in your first paragraph, and I have already encountered such ideas. I heartily recommend to you Pinker’s How The Mind Works, for more detail on this than Hayek and Kant went into. If you still have any doubt that I have not encountered such ideas, please tell me what information I could supply that would convince you entirely that I already know about them.

            As for your example of our ancestors actually hearing thunder as godly anger, this is a surprising claim that I am rather skeptical about. What cultures are you talking about? When are you referring to? What’s your source for this statement?
            Secondly, it remains that we still hear thunder. Science has not managed to explain thunder away, what science has done is change how we observe thunder.

            If we get a new conceptual framework, we may see whatever we now see as a mental state as something else instead.
            In such a conceptual framework, we won’t be seeing anything at all. As I pointed out to you before. If we are seeing anything, then we still have mental states.

          • As I indicated before, either to you or to Mark, this objection is not charitable to the eliminativist. Imagine someone who says ‘Demonic possession is a mental disturbance.’ You object that, if his statement is true, then there is demonic possession, because demonic possession must exist if it is to be a mental disturbance or anything else. The problem with the objection is that the maker of the statement is making an ‘inverted commas’ use of ‘demonic possession.’ He is saying that what people conceptualise as demonic possession is actually nothing of the kind; it is a mental disturbance. Similarly, the eliminativist is saying that what people conceptualise as mental states are nothing of the kind but something purely physical, though he is not in a position yet to say what they are exactly. So, when he says that, when we get a new conceptual framework, we will see whatever we now see as a mental state as something else instead, his use of ‘see’ must be taken as an inverted commas use. It will have a counterpart in the new conceptual scheme, but we do not yet know what that counterpart is. Once we switch to the new scheme we will drop ALL the mentalistic concepts and speak about the world in a new way.

            I should not have made the reference to our ancestors. I actually had in mind the ‘tribal people’ studied by anthropologists. But when I hear about ‘tribal people,’ I think of teenagers or politicos. So I wanted to say ‘primitive people,’ but I decided not to because some people take offence at that. So I substituted ‘our ancestors.’ I knew I had made a mistake as soon as I posted it.

            Thanks for recommending Pinker’s book. I actually have a copy, though I have not read it yet. I am a little bit iffy about him, though. I read his ‘Blank Slate’ a while back; and, while I thought it was good on the whole, I found him a bit dogmatic and ‘inductivist.’

          • TracyW

            As I indicated before, either to you or to Mark, this objection is not charitable to the eliminativist.

            Perhaps not, but I can’t think of a way of understanding the eliminativist’s claim that makes it any more coherent. I am already being charitable by not complaining that they are speaking of a possible future advances in knowledge as if it was definitely going to happen.

            Imagine someone who says ‘Demonic possession is a mental disturbance.’ … He is saying that what people conceptualise as demonic possession is actually nothing of the kind; it is a mental disturbance.

            But what we are trying to explain, in this case, is the observation of behaviour (please insert in here all the scientific caveats about observation) for which we can see no rational reason. We can say that this behaviour is caused by demonic possession, or that it’s caused by mental disturbance, and I prefer the latter explanation as it’s vaguer and more reflective of our ignorance, but the phrase “mental disturbance” still presupposes too much – there are plenty of cases of people appearing (insert formentioned caveats) to behave irrationally to outsiders just because the outsiders lack some key information, like “Fred’s not insane, he’s just got a hands-free headset for his mobile.”
            So this analogy is not applicable, the sample statement is about replacing one “explanation”, “demonic possession”, by another “mental disturbance”. (Scare quotes because “mental disturbance” is not really an explanation, compared with for example “Jim is behaving irrationally becuase he had too many of Granny Smith’s special cookies.”, which is at least a potentially disprovable explanation.)

            In the case of the eliminativists, they are speculating that we will one day find something that makes me see that I don’t see. This is a fundamental difference to explanations of why people appear to behave irrationally, and I can’t think of any case in science where we’ve explained away what we oberve to the point that we no longer observe it itself (insert forementioned caveats). It’s also incoherent in and of itself – how can I be convinced if I am incapable of conviction? Saying that the word see must be taken as an inverted commas use doesn’t make this problem go away, it’s still me who would potentially be using the inverted commas use.

            The eliminativists are expecting me to believe that science will one day explain away my observation of my own awareness, even though I can’t think of a single observation (as opposed to theory and please insert caveats) that science has explained away like that. Trying to think of a charitable interpretation of this is giving me a sense of weariness and confusion, which in and of itself refutes their claim.

            On the final point, I don’t think The Blank Slate is Pinker’s best book, it was a bit more polemical.

          • Here is your objection.

            The following makes sense:

            Phenomenon: queer behaviour.
            Explanation 1: demonic possession.
            Explanation 2: mental disturbance.
            Explanation 2 is better than explanation 1.
            Conjecture: there is no demonic possession.

            In contrast, the following is self-refuting:

            Phenomenon: mental states
            Explanation 1: psychology.
            Explanation 2: neuroscience.
            Explanation 2 is better than explanation 1.
            Conjecture: there is no psychology.

            The reason it is self-refuting is that the conjecture contradicts, and is thus refuted by, the phenomenon.

            Here, as I take it, is the eliminativist reply.

            Your second scheme misrepresents the eliminativist proposal. It should rather be like this:

            Phenomenon: physical process X
            Explanation 1: psychology.
            Explanation 2: neuroscience.
            Explanation 2 is better than explanation 1.
            Conjecture: there is no psychology.

            The temptation to misconstrue this as your second scheme is understandable because we are currently enmired in a psychologistic conceptual framework, so the only way we have of representing physical process X to ourselves in any sort of concrete way is as a mental state. Thus, we cannot, at present, identify physical process X except by saying that it is what is presumed to underlie a specific mental state. However, once we have a fully developed neuroscience, we will be able to specify in detail what physical process X is and to identify it directly as physical process X. Suppose, for neatness, that physical process X is what we currently call ‘seeing.’ Then, when we have a mature neuroscience, we will even physical process X (VERB) physical process X (NOUN) as physical process X (NOUN).

            As I have said, I am not an eliminativist and I don’t share their confidence or bravado about what a future neuroscience will do, so I might not have offered the best defence of their position. The weak point in my defence, it seems to me, is that ‘as’ in the final sentence. Can we make sense of that in purely physicalistic terms? I don’t know. Perhaps we cannot now, but will be able to once we have a mature neuroscience. I do not think we can confidently rule out that possibility.

          • TracyW

            Um, I don’t understand your characterisation of my objection. “Mental disturbance” isn’t really an explanation of queer behaviour, it’s just a label for our ignorance.

            Nor do I see how neuroscience contradicts psychology, or how either contradict mental states. If I understand the terminology right, neuroscience is to psychology as chemistry is to biology, in that biology is based on chemistry, but I’d never conclude from that that there is no such thing as biology. So why should I, or anyone else who at least aspires to be rational, conclude that neuroscience and psychology contradict each other? I’ve read through my previous replies and I can’t see a single point where I’ve said anything of the kind.
            As for your explanation of the eliminativist belief, this is the first time I’ve heard someone call a mental state a sort of concrete representation! Perhaps this is just a disagreement about terminology?
            Nor do I see why being able to explain awareness in terms of physical states would mean that awareness doesn’t exist. The eliminativists appear to be saying “If we can explain exactly why and how something happens, then we can conclude it doesn’t happen”! I’m trying to be charitable here, but I can’t do better than this out of your description of their arguments.

          • Now I am having difficulty understanding your difficulties. I’ll go through point by point.

            For ‘mental disturbance’ just substitute some psychological explanation of the phenomenon.

            Neuroscience and psychology do compete to explain the same phenomena. At least, this is so if you regard psychological phenomena as having some kind of efficacy in the world; it would not be so if you were an epiphenomenalist (in which case physics explains everything and the mental is just a shadow). But that is neither here nor there. The eliminativist sees a mature neuroscience as doing away with psychology, in a way analogous to that in which oxygen theory does away with phlogiston theory.

            I didn’t actually say that a mental state is a concrete representation (though I might do, in some contexts). What I said was that, at present, our only concrete way of representing X is as a mental state. If I say that my only concrete way of representing my sofa is as a sofa, I am not saying that my sofa is a concrete representation.

            What you say in your penultimate sentence is what I tried to represent in the second scheme in my previous message. The answer to it is the same.

          • TracyW

            For ‘mental disturbance’ just substitute some psychological explanation of the phenomenon.

            Now I’m even more confused by your first scheme. So we’ve gone from some vague idea of “mental disturbance” to some even vaguer “some psychological explanation”, for an analogy that I found inapplicable in the first place. What’s this meant to show?

            The eliminativist sees a mature neuroscience as doing away with psychology, in a way analogous to that in which oxygen theory does away with phlogiston theory.

            This analogy does not work. Psychology is a field of study, not a single hypothesis, or theory. Every single psychological theory could be disproved tomorrow and yet it would still be quite possible to study psychology (apart from probably rather difficult to decide where to begin, under such an upheaval).

            I’m getting the distinct impression that the eliminativists don’t know much about science. They rather come across as a bunch of blowhards, don’t they?

            What I said was that, at present, our only concrete way of representing X is as a mental state.

            Okay, then, if you wish to go into exact words, then this is the first time I’ve heard anyone call a mental state a concrete way of representing X.

            What you say in your penultimate sentence is what I tried to represent in the second scheme in my previous message. The answer to it is the same.

            I’ve tried and I’ve tried and I can’t see the resemblance between my sentence and any of the schemes you gave. Unless we want to count things like “they’re written in English”.

          • As I said before, I am struggling to see why you can’t see. Again, step by step.

            1. Just switch the example to oxygen:

            Phenomenon: combustion phenomena
            Explanation 1: phlogiston theory
            Explanation 2: oxygen theory.
            Explanation 2 is better than explanation 1.
            Conjecture: there is no phlogiston.

            2. The analogy does work if you think there are no mental states.

            3. You are still twisting my words: ‘concrete way of representing X is AS a mental state.’

            4. You are insisting that the eliminativists posit mental states and then offer a non-psychological explanation for them. But that is false. They say there are physical processes which we currently describe and explain using mentalistic concepts that will in future be better described and explained using physicalistic concepts.

          • TracyW

            On your 1, fine. But notice that you haven’t explained away the phenomenon: combustion. My awareness is to me a phenomenon, like combustion, as I can observe my own awareness. As I said at the start, science hasn’t explained away any observations, it’s often explained observations, and often explained there’s far more behind them than we used to think, but it hasn’t explained them away in the first place.

            2. I’m sorry, I don’t see how the analogy works even if there are no mental states. And, it’s ridiculous to think that there are no mental states.

            3. My apologies for twisting your words, it was unintentionally caused by fitting your words into my sentence. My point remains. This is the first time I’ve heard anyone call a concrete way of representing X is AS a mental state.

            4. I don’t recall insisting that eliminativists do anything. Indeed, I have even held in my mind the possibility that they (and you) are a figment of my imagination. And I rather thought the whole point of eliminativists was that they didn’t posit mental states.
            Your statement of eliminativist’s views is also different from your earlier claim. Earlier you said:

            On this view no mental states exist, so there are no thoughts and no feelings.

            Now, according to you,

            They say there are physical processes which we currently describe and explain using mentalistic concepts that will in future be better described and explained using physicalistic concepts.

            The two theories are quite different things. Now it sounds like you are saying that eliminativists agree that there are mental states, but they will at some point in the future be better described and explained using physicalistic concepts.
            They’re rather over-confident, aren’t they, these eliminativists? Particularly given their ignorance of science (if they are as you describe them).

          • Very briefly, because I am dashing off in a moment.

            Science has explained away loads of observations. For example, when people observe the sun on the horizon, the sun is not there. It only appears to be there because the light from it is refracted by the earth’s atmosphere. More generally, the fact that each observation always includes interpretation means that each observation could be explained away by showing that the interpretation is false (which is not to say that all observations can be explained away at the same time). For the eliminativist, your ‘observation ‘ of awareness includes a mentalistic interpretation which needs replacing.

            My two sentences are consistent. The second sentence does not entail the existence of mental states. It says there are things which are not mental states that we describe and explain (and conceive) as mental states.

          • TracyW

            Science has explained away loads of observations.

            No, it hasn’t. Please re-read my earlier comment on this, the first time you made this assertion.

            For example, when people observe the sun on the horizon, the sun is not there.

            And science explains why we observe it on the horizon, even though it’s not there. And we still observe sunsets, even people with PhDs in Physics (or at least, they say things like “look at the beautiful sunset”, though perhaps said physics PhDs are figments of my imagination too).

            For the eliminativist, your ‘observation ‘ of awareness includes a mentalistic interpretation which needs replacing.

            Out of curiousity, given what you’ve described of them, why do you bother listening to them?

          • Ah, okay. Science has explained away loads of previously accepted observation statements. But these explainings-away still left a mistaken observation. Observation itself has not been explained away. At least, not yet. The eliminativist thinks it will be, eventually (at least in the sense in which observation is a mental state – unlike, e..g. ‘observations’ made by a pierce of equipment, like a camera).

            I think that eliminative materialism is an important hypothesis. It seems at least conceivable that it could turn out to be correct, and this prompts us to consider a wider range of possibilities. Some of the arguments over it could well turn up some interesting results. As Einstein put it: ‘If at first an idea is not absurd, there is no hope for it.’

            It came up in this discussion as a challenge to some kinds of dogmatism. Challenges to dogmatism often seem ridiculous at first, precisely because the views they challenge are held dogmatically. But they sometimes turn out to advance our understanding by leaps and bounds. The Copernican hypothesis seemed ridiculous at first, because it contradicted many common-sense, religious and even widely held scientific views (Aristotelian physics and its mediaeval developments). The tendency amongst people who saw its explanatory utility was to regard it as a useful fiction. It took painstaking work by at least three intellectual giants (Kepler, Galileo and Newton) for the Copernican view to be accepted as a literal description of reality. The story of this is told beautifully in Kuhn’s ‘The Copernican Revolution,’ which I recommend to everyone who has not read it.

            For what it is worth, I use eliminative materialism myself to challenge a familiar kind of argument which I regard as dogmatic in a recent paper available here:


            Such off-the-wall hypotheses prompt us to retain an open mind.

          • TracyW

            And we come back to my original objection: that the more convincing you find the eliminativist argument, the less convincing you should find it (because the eliminativist argument means that you can’t be convinced). I see no signs that the eliminativists have conceived a way of explaining mental states as not existing, after all they keep using the words “see”, “think” or “conceive” in discussing their potential future where they claim we won’t be seeing or thinking or conceiving anything at all.

            As for dogmatism, they laughed at Galileo, they laughed at Einstein, but they also laughed at Boso the Clown. If the best you can come up with for your argument is that they laugh at you, it’s not a very good argument.

          • MARK_D_FRIEDMAN

            I must vehemently object: the name of this famous clown is “Bozo,” and you should know better! I knew him well during his long career, and he was a hard-working, nice guy. His name should be revered, and not be mutilated.

          • MARK_D_FRIEDMAN

            Because this (interesting) conversation has continued to this point, I will briefly jump back in for one question. Danny, I am interested in how far you will take your claim that: “So admitting that REASON is fallible does not entail giving up reason. It just means that there are no certainties, literally none.” Do you deny that the following statement is certain: “Earth exists.” I note, paralleling the discussion regarding awareness, consciousness, sentience, etc., that our understanding of what “Earth” is has changed dramatically over the centuries, but in everyday experience we walk on it, touch it, dig into it, look at it, etc. And, in this sense, the statement that “Earth exists” seems incontrovertible, without the need to fill in a detailed explanation that Earth is a planet, part of a solar system, etc. Do you agree?

          • Taking your question at face-value, it seems plainly controvertible that Earth exists. It may be that there are only minds which have representations of Earth, no Earth itself.

            The reason I talk about taking your question at face-value is that there are some people who have an ambiguous (or have-your-cake-and-eat-it) attitude to the question. Kant, for example, would say that Earth is empirically real but transcendentally ideal, i.e., it exists as a phenomenon (an appearance) but not as a noumenon (an objective thing out there independent of us). And he then goes on to limit the use of ‘exist’ to the realm of phenomena. So he says ‘Earth exists’ (in his sense of ‘exist’) even though, on the level of ultimate reality, it does not exist. Phenomenalist positivists do something similar.

            But taking words in their ordinary meanings, I say that it is open to doubt whether Earth exists, because idealism (or even solipsism) might be true, or Descartes’s demon hypothesis.

            What is behind your question?

          • MARK_D_FRIEDMAN

            The point of the question was just to test the scope of your claim. There is an entirely trivial sense in which nothing is “certain.” If someone challenged me to “prove beyond all doubt” that China is a large country in Asia, I could show him a map, a globe, eye-witness reports, etc. But the skeptic could always say things like “people are often mistaken about “facts,” including those who make maps, etc.” At this point, I would not waste further time with this person, as he is either a fool or arguing for argument’s sake. I am not suggesting this of you.

            However, frankly, the claim that awareness is not certain or the Law of Identity may be wrong seems in some way similiar to my example. You are resisting certainty not because you can cite a plausible argument against the truth of these things, but because someday there might be one, although for the reasons advanced by Tracy, I do not even concede this possibility. Of course, any claim of certain knowledge can be defeated this way—there might be this or there might be that…But, how do we test and evaluate this claim for truth? How do we set this one up for potential verification or falsification? I just don’t see it. If a thousand years from now the eliminative materialists have still failed, then we can wait another thousand years, etc. You will always be “right.”

            This is not to say that our explanation and understanding of our empirical observations is not always subject to revision. But, I agree with Tracy that your argument involves not this sort of explanation, but the (promised but not delivered) explaining away of the phenomena in question.

          • There are three connected points I want to make in response to this.

            The first point is that, in referring to the possibility of idealism, solipsism or Descartes’ demon, I was giving the wrong kind of response to your question. Here is how I should have interpreted your question: do you think that science could ever discover that Earth does not exist? To that I would also answer ‘yes.’ It is very difficult to conceive how that could happen; but a science-fiction writer could come up with something (perhaps one of them already has done).

            The second point is to raise and answer the question: how should we respond to the infinitude of reasonable doubt? The dogmatic response is to say: deny it. The traditional sceptical response is to say: we cannot know anything, so give up trying. But we can find a middle path. We agree that we want to obtain better theories about how the world works, theories which might not be true, but which are illuminating (as Kepler’s, Galileo’s and Newton’s theories were false but illuminating). It then does not matter to us whether these theories might turn out to be false; indeed, we hope that we will eventually find better theories to replace them in any case. What matters is whether our current theories explain more than their available rivals. If they do, we stick with them (for the time being) The fact that sceptical doubts can be raised about them does not change the fact that they are currently better than any available alternative. Indeed, only someone who is obsessed with certainty would think that sceptical doubts present some kind of obstacle to knowledge.

            The third point is to raise and answer the question: what do we do about people who raise sceptical doubts? We leave them alone. If they are just being sceptics, they are harmless (at least, they are harmless from the point of view of people who are not trying to obtain certainty). But if they are raising sceptical doubts as a beginning of an alternative research programme, then we encourage them. We might not take them seriously until their research programme bears fruit (it it ever does). But recognising our fallibility, and also that major new discoveries have come from people pursuing apparently crackpot ideas, we are happy to see such people spending time on off-the-wall research. We cannot know in advance which (if any) of these research projects will turn out to be valuable.

            The approach I just outlined in points 2 and 3 is that of Popper. You might notice its similarity to Hayek’s knowledge arguments for open markets. Although there are differences between Popper and Hayek, their basic vision is identical, though applied in different areas.

          • MARK_D_FRIEDMAN

            Of course I agree with the thrust of #2 & 3. Our only disagreement is about what constitutes “reasonable” doubt. On this, I am afraid we will have to continue to disagree.

          • There is a fourth point, not much connected to the earlier three.

            The law of identity has been impugned by philosophers who have argued that it is inconsistent with diachronic change. This goes back to Heraclitus: ‘You cannot step into the same river twice.’ It is also raised in the paradoxes of Zeno. Of course some philosophers impressed with this sort of argument have inferred that change is impossible. Others have inferred that, since change is real, in changing an object breaks the law of identity. Hegel went along this second route. There is still a debate about these issues.

      • Aeon Skoble

        That Priest thinks Russell has made a mistake doesn’t refute the PNC, and indeed without the PNC, we wouldn’t know one way or the other. Aristotle’s defense of the PNC is much deeper than logic-chopping. I want to say next that there are some certainties, e.g. the PNC and 2+2=4 and the cogito, but if there’s no such thing as reason, then we can’t discuss it. I don’t mean anything hostile by that, I mean we literally can’t discuss it, for what would be the basis for the discussion? What would count as a good inference?

        • I agree: the fact that Priest thinks Russell has made a mistake does not refute the principle of non-contradiction. But it does show – or, at least, Priest’s dialetheic logic shows – that it may be that the principle of non-contradiction is false, because if dialetheic logic gives the correct account of validity, then some (but not all) contradictions are true.

          Dialetheic logic is only one of numerous alternatives to classical logic. Every significant principle of inference of classical logic (modus ponens, double negation, excluded middle, conjunction elimination, and so on) is denied in one or more of these alternative systems. So the question ‘What would count as a good inference?’ is genuinely up for debate. This is not to deny that there is a great deal of agreement amongst logicians over which inferences are valid. Alternative systems use classical logic to cover most cases, but then suspend it for special cases. But we never know whether some new problem or proposed solution to an existing problem is going to suggest some further suspension of the classical rules of inference, or where, So uncertainty pervades the whole of logic.

          Still, I think most logicians seem to stick with classical logic and regard the alternative systems as curios. We can afford to do that so long as classical logic gives the right answer for most of the problems we are likely to encounter, just as engineers use Newton’s theory rather than Einstein’s for practical problem-solving.

          Again, as I said before, giving up certainty does not mean giving up reason. We can agree to accept modus ponens as a valid rule of inference and argue accordingly – until someone can show us that, for the problems we are dealing with, it does not actually hold. We don’t need certainties to make progress. We just take the best that we currently have available and work with that until we get something better. That is what the scientists do.


    Aeon has pretty much spoken for me here, so I will just add this. In response to his excellent comment, you say: “My point was epistemological, namely that the free exercise of practical reasoning leads to systematic disagreement about what is right and good.” I’m not sure what you think follows from this. One could say exactly the same thing about the scientific process. Today’s seemingly certain truth is tomorrow’s discredited theory. Science is always in conflict, but from this it does not follow that the pursuit of objective scientific truth is a mirage.

    I’m sorry, but your “post-Enlightenment” understanding of “reason” seems to me to be the central planning approach to resolving political/moral disagreements. You want to entrust someone or something with the task of defining “reasonable” pluralism, thus excluding the “unreasonable.” But this assumes that you have sufficient, i.e. objective, knowledge of the truth. This seems contradictory. On the other hand, if you lack such knowledge, you risk excluding the actual moral truth from the reasonable consensus. Whatever happended to the vaunted marketplace of ideas? Like the one used by scientists.

  • Fallon

    Rationalism vs. Anti-Rationalism is a great topic. I find suspect the concepts of spontaneous order, evolutionary cultural selection and biologically based determinism of ideas that only result in asymmetric subjectivism. For one, why all the imperative certainty in these concepts– that require knowledge in the first place- just to reject rational processes. I can see why Popper and Hayek were so lovey dovey.
    But it seems to me that discovery e.g. of the market can and does result in consciously deployable knowledge, replicable social technology. “Hey, indirect exchange wins for me– and for you. Looks like the more that get involved in this cooperation, the better off we are.”
    Is to deny that humans act purposively opening the door to recrudescent Medieval absolutism? Although now it is not God, nor is it Marx’s historicism, but social Evolution.

  • Hume22

    I enjoy the irony of labeling Enlightenment thinkers “dogmatists” (Kant’s First Critique).
    One problem that gives me a headache, that slows me from becoming a (dogmatic?) public reason / justificatory liberal is this: If Betty’s conception of justice is the *correct* conception of justice, how can she possibly *wrong* Alf by successfully acting on it? (‘wronging’ someone being an act of injustice). I’m not saying this is the argument that takes down the Gaussian project, but it’s just a very basic issue at the foundation of public reason that seems counterintuitive.

    • Kevin Vallier

      As I understand it, the problem is that you are coercing others based on rules and principles that they lack sufficient reason to accept. My view is that justice contains a recursive element, which includes the principle of legitimacy (this is not well-known, but Rawls says the liberal principle of legitimacy is selected in the original position). So in one sense, if you think justice requires that you coerce others without a public justification, then you don’t have the right view.

  • Yes, reasonable people can disagree on some things; does this mean they disagree on everything? For example, do reasonable people disagree that using one’s faculties of reason to the utmost is desirable? If they don’t, then what are some interesting things that follow from this? Rand says that her philosophy follows from that, even when she acknowledges that there are reasonable people out there who disagree. Her response to this fact would be that reasonable people can nonetheless fall short in their understanding of her ideas. For example, reasonable people can fall short in understanding that at least as central to her philosophical thinking as anything is the principle that using on’es faculties of reason to the utmost is desirable. Instead, a lot of people attribute some other ideas to her as “her most basic philosophical principle.”

    But if that is her most basic philosophical principle, who reasonably and honorably disagrees with it? That doesn’t get you to all of her substantive philosophical conclusions but this is one place you simply cannot fault her. Further, tied up in her understanding of “reason” is methodology – be it philosophical methodology or just cognitive methodology in general. (Would Rand recognize a distinction there? I suppose she would, considering her views on the nature of the role of philosophy in human life as distinct from other disciplines, which are nonetheless dependent on philosophy.) This methodology gets considerable coverage in Sciabarra’s work and in Peikoff’s ‘Understanding Objectivism.’ Reason means taking into account and keeping the widest applicable context in framing whatever issue one is thinking about; it means respecting the hierarchy of one’s concepts and of one’s knowledge; it means a strict policy of integrating one’s conclusions in a coherent manner with one’s other ideas. Now, who reasonably or honorably disagrees with that?

    Are we yet at substantive philosophy, or are we still just talking method and some very general commitment to reason? I don’t know, really. These points seem very interrelated to say the least. But let’s say that we do get some substantive conclusion, such as: Living thoughtfully and intelligently is the best way for people to live their lives. Or is that still too abstract? What’s the content? The content is whatever you get when you fill out a eudaimonistic ethical theory with detail; Rand does that to some extent in ‘The Objectivist Ethics,’ and Tara Smith goes at it at great length in ‘Ayn Rand’s Normative Ethics’. And there’s a rich tradition of other eudaimonist literature out there that a plausible account of Rand’s ethics needs to engage with thoroughly. Aside from Aristotle’s ‘Nicomachean Ethics’ you have works like Norton’s ‘Personal Destinies,’ or Veatch’s ‘Rational Man,’ or Gewirth’s ‘Self-Fulfillment,’ or Fromm’s ‘Man for Himself.’ I contend that reasonable and honorable folks would have very little in such works to disagree with, in terms of their essential theses anyhow.

    And let’s say we get another substantive thesis out of this: if using our faculties of reason to the utmost is most desirable, implicit in this is an endorsement of some form or other of the “non-agression principle,” something that I’m assuming all BHLs, left and right, agree with in the abstract. Where reasonable disagreements arise are on the practical implementation of rights in regard to extrapersonal resources – what the background structure of such rights should look like. Here’s where Nozick and Rawls would diverge, but they seem to be very much in agreement up to that point. Rawls also gives independent consideration to egalitarian ideals as they pertain to things like opportunities and capacities and not just the minimalist classically-liberal ideas about equality under the law. Rawls aimed to meld the libertarian principle with an egalitarian one under a broad conception of what reasonable people would agree to.

    But let’s back up a step and thing about this in terms of the philosophical fundamentals and not so much the derivative applications. Let’s think of this in terms of the normative ethics involved, which is basically eudaimonist with all that entails in normative ethics. There’s a lot to be hashed out, but a eudaimonist ethics appears to cover all the bases that we want to see in a normative ethics. Say that we treat Rawlsian contractarianism not as a normative ethics – which would rightly piss off the Alasdair MacIntyres and others not wedded to a normative modernism – but as a proto-political ethos. I think that’s what Rawls would prefer to do (or would ideally prefer to do), leaving normative ethics open to plausible accounts like eudaimonism. Given the truth of eudaimonism, you’re still left with questions about the structure of the political order. But looky what we have here: a fairly extensive idea of what reasonable people can agree on, with Rawls providing his own political philosophy based on some plausible (*plausible*, i.e., admissible for consideration) conception of what reasonable people would agree about. Is Rawls doing something different than Rand or Rothbard in that respect? And once we run the analysis of Hayek, is *he* doing something different?

    But I got a bit sidetracked there – with a point to it; I said we need to back up to the normative-ethics part of things, and notice what we could all normative-ethically agree about: that the utmost use of one’s reasoning faculties is desirable – indeed, something that moral and political philosophers *qua* philosophers can’t exactly object to on pain of self-contradiction, plus the fact that human history bears out the desirability of as many people as possible living thoughtfully and intelligently (a central prescription of eudaimonism). Before we so much as approach what to do politics-wise, before we come to *that* project in agreement upon some basics as a prerequisite, just imagine the consequences of a more widely shared and reasonable agreement on the normative-ethical principle. Wouldn’t that *already* lead to enough desirable effects that “what to do politically” becomes less of a seemingly insurmountable challenge? Say that the social entailments of eudaimonism – and the social entailments of a widespread embrace of eudaimonism – already give rise to strong supportive social networks that we tend to think political institutions are supposed to provide absent effective private cooperative arrangements. But not just that: think of the advances in science, in learning, in productivity, in the degree of peaceful social relations (i.e. less crime), better raising of children (along with social encouragement and perhaps even aid in the improved raising of children), and the combined and snowballing benefits of things like this?

    The human mind is, after all, the source of so much wealth we see around us; shouldn’t the promotion of Rand’s basic principle – shared by the likes of Aristotle, I believe – be the way we want to make lives better, people more prosperous, etc., rather than using politics as a means of divvying up an otherwise (much?) smaller pie of basic goods? Doesn’t it just make sense that something more fundamental and deeply-structural in the human condition needs to be addressed before we hope for an improvement politics-wise? And seeing as it is these more fundamental and deep-structural matters are something that reasonable people *can* more readily and easily agree upon, what’s realistically standing in the way of our starting *there* and working/flourishing outward?

    Doesn’t it stand to reason that reason *is* the answer to human challenges in general, as described above? (And how is a healthy absolutism about reason at odds with the recognition of a wide scope given to pluralism? The “Dougs” provide an extensive argument for how Aristotelian philosophy is reconcilable with important modern values such as pluralism. Why aren’t the things the Dougs say firmly ingrained in the minds of, well, everyone? It’s all so sensible.) This wholehearted, unqualified embrace of reason as an absolute in human life is not merely an Enlightenment principle, but a Classical one as well; it’s the point from which so much can be accomplished, if we would just open our eyes enough to see what’s right in front of our noses. 🙂

  • Graham Marco

    Postmetaphysics anyone? Habermas’ communicative rationality:

  • a) Surely it was the medieval era that was surpassingly confident about reason, and the Enlightenment that was comparatively (and I think excessively) skeptical about it. b) Getting consensus on public-reason libertarianism seems at least as hard as getting consensus on any other kind of libertarianism. I know that public-reason libertarianism doesn’t theoretically require that there be consensus on the public-reason approach itself, but I’ve never understood why it doesn’t. It seems to me that public-reason approaches are trying to find a shortcut around a fact that there is no shortcut around: namely that any position worth defending is going to encounter resistance from some intelligent and well-meaning people.

  • Pingback: Enlightenment- and Reformation-based arguments for economic liberty | Alan P Reynolds()

  • adfhaf

    Arto Farto Gentleman!!

  • Pingback: kalyan matka()

  • Pingback:

  • Pingback: angara fahise()