Philosophical Notes 88, A Response to Matt Zwolinski’s “Libertarianism and Liberty” Essays on (2013), by J.C. Lester

Share this



A Response to Matt Zwolinski’s “Libertarianism and Liberty” Essays on
J. C. Lester

Philosophical Notes No. 88

ISBN: 9781856376617
ISSN 0267-7091 (print)
ISSN 2042-2768 (online)

An occasional publication of the Libertarian Alliance,
Suite 35, 2 Lansdowne Row, Mayfair, London W1J 6HL

© 2013: Libertarian Alliance; J.C. Lester

J. C. Lester is a Senior Fellow with the Libertarian Alliance.
He is a libertarian philosopher and author of Arguments for Liberty: a Libertarian Miscellany (University of Buckingham Press, 2011)
and Escape from Leviathan: Libertarianism Without Justificationism, paperback (University of Buckingham Press, 2012).
His magnum opus is A Dictionary of Anti-Politics: Liberty Expounded and Defended (forthcoming).




Matt Zwolinski’s “libertarianism and liberty” essays are argued to have the following problems: taking libertarianism to be a “commitment” to the view that “liberty is the highest political value”; examining and rejecting the maximization of liberty without a libertarian theory of liberty; accepting a persuasive sense of “coercion”; misunderstanding liberty in the work place; conflating, to varying degrees, freedom of action and freedom from aggression and justice/rights/morals; focusing on logically possible clashes instead of practically possible congruence among utility, liberty, and justice—in particular, that “rule (preference-)utilitarianism” fits “rule libertarianism”; failing to distinguish liberty from license (and power) concerning slavery, and so-called “civil and democratic liberties” (and everything else); the idea that any coherent reference to a quantity of liberty requires precise cardinality; failing to see that the quantity of liberty has an inherently qualitative aspect; misunderstanding property as about limiting freedom; mistaking clashing Hobbesian freedom for non-clashing Lockean liberty; adopting G. A. Cohen’s confusion about freedom as the libertarian conception of freedom; assuming the—illogical—epistemology of “justification”;  not realizing that both allowing and prohibiting pollution “aggresses” and so “aggressions” need to be minimized; the failure of all six of his reasons for rejecting the non-aggression principle.

Introductory Remarks

It was suggested to me that it might be interesting to have my response to Matt Zwolinski’s “Liberty and Property” on  However, I see that his post is one of several that form what is more or less one long essay.  And so I take the opportunity to reply to more or less all of them.  I shall try to keep my responses short and avoid repetition except where it strikes me as desirable for clarity or emphasis.  Where I agree with Zwolinski, or do not significantly disagree, then I shall say nothing.  I keep Zwolinski’s titles for each section throughout.  For simplicity, I begin by simply inserting my comments below the relevant parts of the original text.  But in the final three sections I revert to a more normal essay style.

“Libertarianism and Liberty, Part 1: A Complicated Relationship”[1]

… what else could define a commitment to “libertarianism” other than a belief in liberty?  Other political ideologies pay lip service to freedom, and perhaps even hold it as one legitimate value among others to be balanced in the great political calculus.  But what sets libertarians apart is their belief that liberty is the highestpolitical value.  …

There are several immediate problems with the idea of having a “commitment” to the view that “liberty is the highest political value”.  1) We cannot be committed to any view.  We don’t decide what we believe is true or moral: introspection reveals our beliefs.  And so a perceived refutation can stop us holding a view at any time.  2) Strictly, though possibly pedantically, liberty is not a “value” (values only exist in people’s minds), but a concept or state of affairs.  3) Libertarianism is, in one sense, not “political” but anti-political in principle (at most, a minimal state is a necessary evil).  4) Libertarianism is not necessarily even the “highest” principle.  A libertarian principle might be held to be inviolable.  But even that does not entail that it is the “highest” principle.  If it is held for modus vivendi reasons, for instance (as I suppose it usually is to some degree, at least), then all participants might have other principles that they would personally rank higher, or value more, than the libertarian principle.  However, they realize that liberty is a safer way to promote those other principles than the use of aggression (i.e., flouting interpersonal liberty).

… let’s look at one popular and superficially plausible interpretation.  What it means to hold liberty as the highest political value, on this view, is to hold that liberty ought to be maximized.  …

Before jumping into issues of maximization, should we not first ask, “what is the best theory of libertarian liberty?” Otherwise, whether or how it can or should be maximized cannot be answered.  My own preferred theory of social or interpersonal libertarian liberty is an objective and pre-propertarian “absence of proactive impositions” (on people by people).  That is, I take ‘proactive impositions by other people’ to be the relevant aggressive constraints that fit what libertarianism requires to be avoided.  Consequently, where an absence is not fully possible—as is often the case—then a minimization will be the most libertarian option.  From this formula I have argued that we can derive self-ownership itself and all libertarian property, as well as solving various known paradoxes and newly arising problems.  But I can’t go into detailed explanations here.  See Lester 2012.[2]

… I am unable to think of a single libertarian philosopher who defends a position like the one I am describing.  …

I am inclined to the view that liberty ought to be maximized: why would a libertarian put up with less liberty if more were possible?  However, I also incline to a version of what I call ‘rule libertarianism’ (rather than ‘act libertarianism’): as a general rule, don’t violate liberty even where it looks as though greater liberty can thereby be achieved—because it won’t work in the long run.

…The standard libertarian response to such criticisms, of course, is to point out that employment relationships are voluntary, and so whatever restrictions employers impose upon their employees do not actually count as a violation of their freedom in the relevant sense.  Personally, I don’t find this response to be all that persuasive.  For starters, it simply assumesthat what libertarians believe oughtto be the case actually isthe case - viz., that employment relationships are entirely voluntary.  But this ignores the myriad ways in which coercion infests our present system, …

There is considerable confusion about “coercion” among libertarians.  Many of them use ‘coercion’ as meaning whatever is unlibertarian.  However, libertarians cannot be against ‘coercion’ as such in its plain English sense: roughly, interpersonal force and the threat of force.  They can only be against coercion that violates liberty.  They cannot be against coercion that enforces liberty or is voluntarily or contractually accepted.  And libertarians must also be against plainly non-coercive acts that violate liberty, such as fraud and most theft (some theft also involves coercion).

… often to the benefit of employers and to the detriment of laborers.  …

Yes, the state interferes and confuses matters.  But the crucial point must be distinguished and not lost: insofar as the state does not impose rules that flout liberty then “whatever restrictions employers impose upon their employees do not actually count as a violation of their freedom in the relevant sense”.  And then it cuts both ways: whatever restrictions employees impose upon their employers do not actually count as a violation of their freedom in the relevant sense.  Of course, it would clarify matters to have an explicit libertarian theory of liberty or freedom to apply here, and Zwolinski does not have one.

… Second, the argument assumes that whatever is voluntarily agreed to cannot be a restriction on freedom.  But this is either wrong or at least a very strange way of using the word “freedom.”  Suppose I ask you to lock me up in your dungeon and throw away the key, perhaps in exchange for your writing a check to my child who I would otherwise be unable to support.  However unimpeachable the contract may be on procedural terms, I am, once locked away in your dungeon, less free than I was when I was, well, free.  Libertarians might be right in thinking that there is nothing morally wrongwith the lack of freedom I now endure.  But to infer from this that it must not be a lack of freedom after all is an abuse of language and logic.

This is a mistake that no libertarian ought to make.  It comes from having no explicit theory of libertarian liberty.  The sense of ‘liberty’ or ‘freedom’ as the absence of mere physical constraint is completely different from the libertarian sense of not being aggressively constrained (or proactively imposed on) by another person.  Once voluntarily incarcerated, Zwolinski lacks physical freedom or liberty but he has suffered no loss of interpersonal libertarian freedom or liberty.  There is no “abuse of language and logic” in making this clear distinction.  The abuse is in conflating two conceptually distinct homonyms.

The fundamental problem with this line of argument is its reliance on what philosophers call a “moralized” conception of liberty.

Some libertarians are indeed confused in just this way.  But there is no need to mention morals at all.  Libertarian liberty has an objective content both theoretically and in its observance.  It is an entirely separate matter whether such liberty or its observance is moral.

… If, as many libertarians assume, violent criminals voluntarily forfeit their rights to liberty when they commit their criminal acts, then punishing them by imprisonment might not be unjust.  But, surely, this does not mean that the criminal is, all appearances to the contrary notwithstanding, perfectly free when the policemen handcuff him, throw him in their police car, and lock him away in a cell.

The criminal’s physical liberty has been reduced.  But his libertarian liberty has not been infringed to the extent that the judicial system was only engaged in rectifying his infringements of the libertarian liberty of others.  That ought not to appear paradoxical or unclear.

Freedom and Justice are both important values, and ones to which libertarians do and should give their allegiance.  But we should resist the temptation to suppose that they are the samevalue.  That they are not the same entails that it is possible, in principle at least, that they may in certain circumstances come into conflict.  This is a possibility that I will explore in a future post.

Libertarian liberty, and freedom of action, and justice are three distinct concepts and states of affairs.  But if what I call the ‘classical liberal compatibility thesis’ is true, then liberty will systematically be compatible with justice and human welfare in their practical applications.

“Libertarianism and Liberty, Part 2: Against Maximum Freedom”[3]

… What utilitarianism seeks to maximize is not the happiness of any particular person but of the aggregateof happiness across persons: the sum of my happiness, plus yours, plus hers, and so on.  But maximizing aggregate happiness is compatible with leaving some people destitute.  It is even, more disturbingly, compatible with makingsome people destitute.  …

It is also logically compatible with libertarianism.  Moreover, happily, it is contingently systematically compatible with libertarianism.

… So long as the misery of the few is sufficiently compensated by the happiness of the many, the treatment of fate of any particular individual is of no decisive relevance for the utilitarian.  It is for this reason that philosophers as distinct as John Rawls and Robert Nozick have both objected that utilitarianism fails to take seriously the separateness of persons.

In theory, but not in practice.  Analogously, libertarianism fails to take seriously the suffering of persons—in theory, but not in practice.  For the most part, the classical liberals (a broader church than—anarchist and minarchist—libertarians) did not see a clash between liberty and utility.  And they were right.  (Although I think preference-utilitarianism, as famously championed by R. M. Hare, is the best version: we often want real states of affairs as ends-in-themselves and would not want happy delusions or pleasurable mental states instead.)

The view that justice consists in maximizing liberty is subject to precisely the same objection.  Insofar as it is aggregateliberty that is to be maximized, this view will countenance the sacrifice of some persons’ freedom for the benefit of others, so long as the net result is positive.  Rather than freedom serving as a constrainton the ways in which others may permissibly act, freedom on this view serves as a goalto be maximized without any real constraint.

But if “rule libertarianism” is true, as I suppose, then the problem is not a practical one for libertarianism.

This has troubling implications for a variety of policy issues, such as (for example) questions involving the preventative detention of potentially dangerous individuals.  On the maximizing view, there is no principled objection to imprisoning an innocent person X merely on the grounds that X is deemed likely to commit some offense in the future.  Such preventative detention restricts X’s liberty, of course, but if it prevents X from acting in ways that would have restricted the liberty of sufficiently many other people (by killing them, or stealing from them, etc.), then, on this view, it is justifiable.

We are all “potentially dangerous individuals.”  But locking up everyone, even if it were practicable, would proactively impose more than it would prevent proactive impositions.  Hence liberty would be lessened thereby.  However, it is different if person X is known to be a significant and serious danger to others.  Such a person proactively imposes on us if he enters our private streets without our permission.  So we could at the very least exclude him from them in self-defense.  And if he is a serious-enough danger, then incarceration is theoretically possible.  But it is hard to test the issue critically here without concrete examples.

A defender of the maximizing view might argue that such trade-offs are unlikely to be beneficial in the real world as opposed to the world of philosophical thought experiments.  And there is undoubtedly some truth to this response.  I will simply note, however, that it is precisely the same response that a utilitarian might make to the charges of injustice we have leveled against his theory.  And so whatever reasons we have for finding the response inadequate in that context (and I think we have plenty), apply here as well.

But the utilitarian’s response is in practice adequate if he is a rule-preference-utilitarian who embraces rule-libertarianism as the right rule to that end.

… Libertarianism does not hold that people are morally free to do allthat they will.  The freedom of all individuals is sharply curtailed by the rights of others.  If I try to punch you in the nose, or trespass on your property, you may justly interfere with my doing so.

This, again, conflates libertarian liberty with freedom of action.  It also conflates what libertarian liberty is with whether or not that liberty is a right or is just.

It is easily imaginable that I can promote some people’s (libertarian) liberty at the expense of others’: for instance, the well-known libertarian thought-experiment concerning stealing a gun to shoot a murderer on a killing spree.  (In that particular case, I advocate stealing the gun but the owner being able to sue for damages if he wants.  Result: utility and liberty maximized—if temporarily and trivially flouted with respect to the gun-owner.)

“Libertarianism and Liberty, Part 3: It Doesn’t Add Up”[4]

… Abolition did, of course, increase the freedom of slaves.  But it also diminishedthe freedom of certain non-slaves.  Specifically, it diminished the freedom of slave-owners.

No it did not.  It diminished the license (the aggressive constraints or proactive impositions) of the slave-holders.  License is the opposite of (libertarian) liberty.  To free a slave is not to take any libertarian liberty from his slave-owner.

This sounds like a shocking claim.  But it shouldn’t be.  If we understand freedom in the way that most libertarians do – if we understand it, as Rothbard did, as the “absence of molestation by other persons” – then it is actually quite obvious.  Prior to abolition, slave-owners were able to do certain things to their slaves without fear of interference by other persons.  They could force their slaves to work, physically restrain them, beat them, and so on, all without the law doing anything to stop them.  After abolition, they could no longer engage in these activities without fear of legal intervention.  Before abolition, the law allowed them to do certain things.  After abolition, it didn’t.  Their freedom had been reduced.

No, the slave-owners’ power to restrict the liberty of others (to proactively impose on those others) had been reduced; they had not thereby themselves been proactively or aggressively imposed on.

… Freedom is one thing; justice is another.  …

Freedom of action is one thing; freedom from interpersonal aggression (libertarian liberty) is another; justice is a third.

… What is the “unit” of freedom on which our operations of addition and subtraction are to be performed?

This last question is especially important, and challenging.  To make it clearer, consider the following example from the philosopher Will Kymlicka’s critique of libertarianism.  Suppose we want to compare the freedom of people in London with that of people in pre-1989 communist Albania.  People in London have freedoms like the right to vote, …

Political voting is not a freedom but an attempt to oppress others in a majoritarian way.  (I have replied at length to Kymlicka’s hopeless “critique of libertarianism” elsewhere.[5])

… the right to practice their religion, and other civil and democratic liberties.

Many so-called “civil and democratic liberties” are licenses posing as liberties.

People in Albania, let us say, lack these freedoms.  “On the other hand, Albania does not have many traffic lights, and those people who own cars face few if any legal restrictions on where or how they drive” (143).  Kymlicka’s sense, which I share and I expect most of you do too, is that Albania’s lack of traffic regulations does not compensate for its lack of basic civil liberties.  It is, on the whole, a less free society than London.  But the question is: can we accountfor this judgment simply in terms of a quantitative judgmentabout the amount of freedom in Albania as compared to London?  …

Yes we can, but it’s a rough and ready quantity rather than a precise unit.  I can often tell you that one object is bigger than another—“more than twice as big”, for instance—without being able to give any exact figures.  And I can often do the same with liberty.  (One test of ordinal liberty in the current case, incidentally, is the direction of migration—if it is allowed at all.  People tend to move from areas of greater oppression to those of lesser oppression.)

… How would such a quantitative judgment be made?  …

By starting with a libertarian theory of liberty instead of freedom of action.

Should we count up the individual, particular action-tokens that are forbidden in Albania and compare them with the action-tokens that are forbidden in London?  If we discovered that, over the course of a year, red lights produce 18,623,545 instances of people being prevented from acting in the way they desire to act, whereas denial of the right to vote produces only 42,658 such instances, would that be sufficient to demonstrate that the red-lights are more freedom-restricting than the denial of political liberty?  Or should we be counting not individual action-tokens but more general action-types, i.e. “the right to vote” versus “the right to drive through intersections as one wishes”?  And whether we choose types or tokens, just how are we supposed to individuate actions in order to add them up?  Is the right to marry the person of your choice one action?  Or a shorthand way of describing an enormously large number of discrete actions?

Muddles about “liberty” aside, comparison of size simply does not entail that precise quantification is required.  We know that Albanians had even less libertarian liberty than we had, don’t we?

However we decide to count up actions, the whole exercise seems largely to miss the point.  For it assumes that what we care about when we care about liberty is (merely) the total number of actions allowed or prohibited.  But why think that all freedoms are of equal value?  Why should the freedom, say, to be governed by one’s own conscience in matters of religious belief count for no more than the freedom to count the blades of grass on one’s lawn?  Why believe that all that matters in assessing the freedom of a country is the numerical quantityof freedom allowed, and not the substantive qualityof that freedom?

The error here is in failing to understand that the amount of the lack of freedom relates to the extent that some infringement matters to the victim.  There is no full distinction between quantity and quality.  Pushing a passing person into a pond is a lesser infringement of their liberty than raping them if that person finds the latter to be worse.

Libertarians are right to believe that freedom matters.  …

How do you know when you don’t have a proper theory of what libertarian liberty is?

… They might even be right to believe that it is the highest political value.  But it is a mistake to think that these ideas, however true they might be, can be fleshed out in terms of a commitment to maximizingfreedom.  Morally, a commitment to maximizing freedom is inconsistent with libertarianism’s proper concern for individual rights.  …

Libertarianism is primarily about protecting liberty.  And more liberty is better than less.  Individual rights are a separate and subsidiary matter.

“Liberty and Property”[6]

… we shouldn’t allow the freedom-enhancing power of private property to blind us to its costs—or even to the fact that some of those costs are measured in the currency of freedom itself.  That property has the power to limit freedom as well as to protect it shouldn’t be surprising, really.  After all, imposing limits on others’ freedom is part of the pointof private property.  …

The point of private property is that it minimizes infractions of liberty.  It does this by limiting acts of license: acts that would proactively impose.

… The Hobbesian state of nature is a state of war precisely because and to the extent that each individual has the liberty to do “anything which, in his own judgment and reason, he shall conceive to be the aptest means” to the preservation of his own life—even if that thing is the crop you just harvested or your body itself.  The freedom of each person to do anything he wishes is a recipe for a life that is solitary, poor, nasty, brutish, and short.  Peace, prosperity, and stability are only achieved when each individual agrees to lay down some of this unlimited liberty and to respect the rights of others.

Hobbesian freedom is having unconstrained action, and it clashes with similar freedom among people.  Lockean liberty is the absence of aggressive constraints, and it does not clash with similar liberty among people.  They are completely different things.

I suspect that most libertarians will have no serious problems with what I have said so far.  My freedom to steal your bread and punch you in the face is, anyway, a pretty unattractive kind of freedom from a moral point of view.  So if private property simply places limits on thatkind of freedom, that’s a feature, not a bug.

That is only limiting Hobbesian freedom of action that is also license (proactively or aggressively imposing on others) and hence not limiting Lockean (libertarian) liberty.

… A property right in land is a right to control access to that land.  It is a “right to say ‘No’.”  But if all land is privately owned, and all landowners have a right to say “No” to all non-landowners, then non-landowners are not equally free with landowners.  They exist in a state of dependence.  Like feudal serfs or the most abject slaves, they live only by the consent of those in command.

However, this consequence is easily avoided by understanding and applying the correct pre-propertarian libertarian theory of liberty.  For to the extent that private property in land begins to proactively impose on non-landowners it is thereby not libertarian.  This, of course, is largely a theoretical possibility in the real world, but any putative empirical cases can in principle be dealt with in an entirely libertarian way.

… Now, the thing to note about Cohen’s argument, and Spencer’s for that matter, is that it is based on a perfectly ordinary understanding of what freedom is.  Cohen is not arguing that the poor lack “positive” freedom or “real” freedom or any other adjectival form of freedom of novel origin and dubious merit.  He is arguing that they lack precisely the kind of negative freedom that libertarians purport to be concerned with—freedom from liability to physical interference by other human beings.

This is wrong.  That is not the kind of freedom that clear-thinking libertarians are concerned with.  Libertarians are concerned with something more like ‘people not being aggressively constrained by other people’ (I formulate this as ‘the absence of proactive impositions’ for reasons of clarity and precision that I cannot briefly rehearse here).  And merely protecting one’s non-aggressively-acquired-and-held property (such as from would-be free riders) is precisely not to aggressively impose on the liberty of others.  I cannot completely blame Cohen for being so utterly confused when so many of the advocates of libertarianism are almost as confused as he is.

“More on Property, Freedom, and Coercion”[7]

… Why do I say that property rights limit freedom?  I start with the belief that to be free is to not be subject to interference by other people.  …

From a libertarian viewpoint, it would be clearer to say that “to be free is not to be subject to aggressive interference by other people.”  And then we need an abstract theory of such “aggressive interference” from which property is derivable.

… I then note that property rights are, at their core, socially and legally enforceable licenses to interfere with others.  If I have a property right in a piece of land, I get to physically interfere with anybody who tries to use that land without my consent, or call on the police to do my interfering for me.  So my having a property right in the land limits your freedom to use it.

On the contrary, it would be more accurate to say that “property rights are, at their core, socially and legally enforceable claims to stop others interfering with us.  If I have a property right in a piece of land, I get to physically defend myself from anybody who tries to use that land without my consent, or call on the police to do my defense for me.  So my having a property right in the land limits something more like your Hobbesian freedom-of-action to use it and not your Lockean freedom-from-aggressive-interference.”  Of course, all this is clearer with a more precise and explicit abstract theory of libertarian liberty than that we “not be subject to interference by other people”.  And then we can see how property is objectively derivable from observing liberty.

“Against Moralized Freedom”[8]

That the individual who owns no land is subject to interference by others seems undeniable.  But what the libertarian canperhaps argue is that this is not the kind of interference that renders him unfree.  Interference with the activity of another only renders that person unfree, we might say, when it does so in a way that violates his rights.  Someone who is enslaved against his will is unfree.  But someone who is forcibly preventedfrom enslaving another is not.  In both cases, the person’s actions are coercively interfered with.  But freedom is a moral term, the argument claims, not a neutral one.  And it is only interference that violates its target’s moral rights that counts as a genuine infringement of freedom, on this view.

It greatly confuses matters to conflate 1) an objective theory of libertarian liberty as the absence of interpersonal aggressive constraints (however that might be formulated) and 2) whether there is a moral right to such liberty.  Admittedly, it is a common confusion among libertarians.  However, it is a far worse confusion to view libertarian liberty as mere interpersonal freedom of action, as Zwolinski does.

“Locke and Nozick on the Justification of Property”[9]

Despite the title, I put Locke and Nozick aside—along with Zwolinski’s comments on them.  For the very idea that there can be a “justification of property”, or liberty, or anything else, is a fundamental epistemological error and one that holds libertarianism back.  This view is, however, probably more controversial and complicated than anything I have written above.  And so I intend to explain it in a little more detail than my previous responses.

If critical rationalism is true, as I believe, then the “justification” of any view is an epistemological impossibility.  And failing to understand this is a third main serious problem for most libertarians—along with 1) not having an explicit, objective, pre-propertarian theory of libertarian liberty, and 2) not fully appreciating how human liberty and welfare are systematically connected both conceptually and causally.  Therefore, I conclude my responses to Zwolinski by briefly outlining both critical rationalism and how I take it to apply to libertarianism.

No theory (or thesis, view, outlook, opinion, argument, proof, etc.) can ever rule out the possibility of a refuting counter-instance or counter-argument.  With our finite and fallible reasoning facing the infinite worlds of unknown matter and theories, we never know what we might have overlooked.  Therefore, theories cannot be justified (or supported, grounded, founded, based, backed, established, proven, etc.).  All theories remain conjectures (or guesses, assumptions, suppositions, and so forth).  We are obliged to use some conjectures for practical purposes; and not always unrefuted conjectures.1  But the only thing we can do with conjectures epistemologically is to criticize or test them as best we can.  If we cannot refute them, then they remain conjectures (but they might well be true, of course ).2  If we can refute them, then we learn something new.

However, refutations are themselves conjectural.  So no refutation is ever justified either.  Fortunately, there is a crucial asymmetry between a justification and a refutation.  A conjectural refutation is coherent.  It makes sense to say, if this observed phenomenon is a black swan, then “all swans are white” is refuted.  Or if this proof is a correctly derived inconsistency, then the theory from which it is derived is false.  By contrast, a conjectural justification is incoherent.  We could not observe all swans (everywhere and everywhen) being white.  Nor could we ‘prove’ the assumptions of an allegedly justifying argument without having an infinite regress, or circularity, or an arbitrary stopping point.  It is also worth mentioning that much that is very commonly mistaken for ‘justification’ is actually explanation (e.g., ‘Markets allocate resources efficiently by people bidding for them in proportion to their profitability, whether monetary or psychic’).  Such explanations may often be true or useful.  But they are themselves conjectures and usually incomplete.

How does this apply to libertarianism?  It is a conjecture that it is desirable to allow universal interpersonal liberty.  Libertarians think that this conjecture is not refuted by any criticisms.  All we can do is try to defend it by answering the best criticisms that we can find—and those of critics of libertarianism (occasionally these overlap).  What about rights?  One might conjecture that rights to liberty are the best rights, and then consider criticisms.  What about utility?  One might conjecture that liberty is the main cause of promoting utility (in terms of the satisfaction of spontaneous preferences, in particular), and then consider criticisms.3  I say “might” because neither of these views can support the universal theory of libertarianism and because a defender of libertarianism might offer different answers concerning rights, or utility, or whatever some specific criticism is about.  The point is to attempt the possible: to refute the particular potential refutation somehow (e.g., “Genetic tests indicate that this alleged black swan is really a new species of goose.”) It is not to attempt the impossible: to establish the universal conjecture (e.g., “Genetic tests show that every swan—everywhere and everywhen—is white”).

Of course, none of the above is intended to be a justification of critical rationalism and its application to libertarianism.  It is, rather, a conjectured explanation.  And for those interested, I have written at greater length on this matter in various other places.

“Libertarianism and Pollution”[10]

Zwolinski begins by telling us that “Libertarians generally believe that aggression against innocent persons is morally wrong, and that the only just use of violence is to prevent aggression by others.”  Somewhat less imprecisely, we might say that the only just use of coercion (using force and the threat of force against people) is to prevent or redress aggression.

He then says, “In this respect, at least, the liberal egalitarian philosopher John Rawls was on precisely the same page as his libertarian colleague, Robert Nozick” and quotes Rawls on “freedom” purportedly to that effect.  However, Rawls has an understanding of “freedom” that is inherently political and which sanctions much that libertarians would rightly see as itself involving aggression.  Consequently, Rawls and Nozick are certainly not “on precisely the same page”.

We are soon asked, “Suppose I aggress against you not by beating you over the head with a club, but by blowing tobacco smoke into your face?  The smoke-blowing, just like the clubbing, is a physical invasion of your body.  And it is a harmful invasion.”  And here we should note that the physical harm itself is irrelevant to liberty.  What matters is that the victim disvalues the invasion for whatever reason.  If an aggression were to improve the victim’s health, then it would still flout his liberty.

After such considerations, Zwolinski concludes that “The consistent application of Rothbard’s absolutist principle of non-aggression thus seems to require a prohibition on all forms of non-consensual pollution.”  And in its absolutist form I agree.  However, this overlooks something crucially important: it cuts both ways.  The enforcement of the prohibition would itself aggress against the people whose activities would produce the pollution (e.g., having fires for needed warmth and cooking).  So such prohibitions cannot be allowed either.  We have reached not one but two unacceptable conclusions and, more to the point, they amount to an inconsistency in the “absolutist” version of the theory.  Hence that form of the theory is a priori refuted.  (I know that Rothbardians try to introduce various points to solve such problems, but they look ad hoc and invalid to me.)

There are two main problems with the absolutist theory that lead to this result.  First, while liberty itself can be interpreted as the absence of aggression, the libertarian policy must be to minimize aggression when there is such a clash as that described.  Thus some sort of compromise is required, maybe with some damages being paid in one or the other direction.  Second, “aggression” understood in terms of violating property rights is only a rule of thumb.  “Aggression” can be more abstractly and accurately theorized as proactively imposing costs (such as both pollution and pollution prohibitions) on other people.  This pre-propertarian theory is required for consistently solving such property problems, paradoxes, and inconsistencies whenever they occur.


This alternative approach should become clearer and more cogent as I reply to Zwolinski’s next essay, in which he attempts to refute the “non-aggression principle” beyond any salvation.

“Six Reasons Libertarians Should Reject the Non-Aggression Principle”[11]

Zwolinski begins that “Many libertarians believe that the whole of their political philosophy can be summed up in a single, simple principle ...  the ‘non-aggression principle’ or ‘non-aggression axiom’ (hereafter ‘NAP’) ....”  And this is what he intends to refute.  Despite having criticized some interpretations of the non-aggression principle myself, I should here like to defend one version of it.  For I see no inherent confusion in using the NAP as a shorthand reference to how people ought to behave and what is necessary and sufficient for interpersonal liberty fully to exist.  However, in the event of a clash of liberties (e.g., I need to have a fire but you would suffer from my smoke) we need to resort to the MAP: minimization of aggression principle.  And it seems reasonable to interpret the MAP as an attempt to apply the NAP as far as possible.  Therefore, I view the MAP as practically implied by the NAP rather than as a separate and additional principle.  This should become clearer as we proceed.

Zwolinski continues that the NAP “holds that aggression against the person or property of others is always wrong....”  Except that, as we have seen, very often two parties cannot help impinging on the liberties of each other (for instance, whether pollution is allowed or prohibited: one side or the other side must suffer an interference/constraint/cost).  And in such cases “aggression” may not seem to be exactly the right word, though it will do.  And what is inevitable is not obviously “wrong”.

Zwolinski writes that “aggression is defined narrowly in terms of the use or threat of physical violence.”  I had rather say, somewhat less imprecisely, “aggression” is proactively interfering with another’s person or property (when these are not themselves the result of any proactive interference).  Is it true that “From this principle, many libertarians believe, the rest of libertarianism can be deduced as a matter of mere logic”?  Most libertarians appear to have supplementary or additional principles.  However, my own view is that only a pre-propertarian conception of libertarian liberty can fully allow that “the rest of libertarianism can be deduced as a matter of mere logic.”  But that is, indeed, a single principle and one that I wish to explain and defend.

As an implicit criticism, Zwolinski observes that “The libertarian armed with the NAP has little need for the close study of history, sociology, or empirical economics.”  That is surely a great virtue in a practical principle for everyone.  Moreover, this appears to overlook that, because study is bound to be finite, no study can support a universal theory such as the NAP—although it can test it and possibly refute it.  He continues that “With a little logic and a lot of faith in this basic axiom of morality, virtually any political problem can be neatly solved from the armchair.”  And such simplicity is clearly highly desirable.  Strictly speaking, no faith is required or possible: we do not choose what to believe.  However, any—necessarily conjectural—solutions can be derived.  And they are then ready for criticism and testing.

What is the philosophical significance of the fact that “On its face, the NAP’s prohibition of aggression falls nicely in line with common sense”?  Common sense is a fallacious criterion of truth or morality.  So it is similarly irrelevant to say that “it is far from common sense to think that its badness is absolute.”  But it is relevant to present “any other possible consideration of justice or political morality” as a criticism of the NAP conjecture.  It might seem that “There is a vast difference between a strong but defeasible presumption against the justice of aggression, and an absolute, universal prohibition.”  But in practice our, necessarily conjectural, theories are always open to potentially refuting criticism no matter how “absolute” we might think them to be.  Zwolinski approves of Brian Caplan’s view that “if you can’t think of counterexamples to the latter, you’re not trying hard enough.”  But counterexamples that are merely logical possibilities and unlikely scenarios are beside the point.  Real systematic refutations of the practical morality of the NAP/MAP do not appear to exist, as far as I can tell.

We then move on to the “six reasons why libertarians should reject the NAP.”  And we ought to note immediately that to refute one, dubious, interpretation of the NAP is not to refute every interpretation of it.

“1. Prohibits All Pollution”

Zwolinski asserts that “industrial pollution violates the NAP and must therefore be prohibited” moreover, even “personal pollution produced by driving, burning wood in one’s fireplace, smoking, etc., runs afoul of NAP.”  As I have explained, prohibiting pollution (for instance, coercively preventing someone from lighting his fire for needed warmth and cooking) also violates the absolute NAP.  Hence the MAP comes into play.

“2. Prohibits Small Harms for Large Benefits”

Zwolinski asks us to “suppose, to borrow a thought from Hume, that I could prevent the destruction of the whole world by lightly scratching your finger?” And here I would reply that the NAP is about the real world rather than about every logically possible world and thought experiment.  He goes on to “suppose that by imposing a very, very small tax on billionaires, I could provide life-saving vaccination for tens of thousands of desperately poor children.”  This is slightly less implausible but it is still not realistic.  We don’t need to tax anyone to develop new vaccines.  And the institution of any taxation would disrupt productivity immediately and then do cumulative damage as the economy has its growth slowed.  Moreover, that growth would probably have included new advances in vaccines sooner or later.  Zwolinski concludes by asking “is it really so obvious that the relatively minor aggression involved in these examples is wrong, given the tremendous benefit it produces?”  And the obvious answer appears to be that implausible assumptions do not refute a practical principle.

“3. All-or-Nothing Attitude toward Risk”

Zwolinski asks, “what if I merely run the risk of shooting you by putting one bullet in a six-shot revolver, spinning the cylinder, aiming it at your head, and squeezing the trigger?”  And the answer is that it is an aggressive act to take such a serious risk at someone else’s expense.  In monetary terms, the degree of the aggression is something like the amount of money that the victim would have to be paid to accept such a risk (I don’t mean to imply that everything can be reduced to money, of course).  Without such an agreement, you are using someone else’s property—his head—without his permission for your dangerous game.  Imposed risks are aggressions; actual damage is not necessary.  Otherwise, by analogy, you may as well say that coercing someone to do something at gunpoint only becomes an aggression if you actually shoot them when they fail to comply.

Zwolinski then observes that “almost everything we do imposes some risk of harm on innocent persons” and that “Most of us think that some of these risks are justifiable, while others are not” but our reasonable explanations “carry zero weight in the NAP’s absolute prohibition on aggression.”  And, again, this overlooks that there is aggression whether such risks are allowed or coercively prohibited.  But there is no insuperable problem with applying the MAP, as long as we have a reasonable account of what policy best deals with the clash in an unbiased way (it need not be perfect or admit of cardinal accounting).

“4. No Prohibition of Fraud”

Zwolinski asserts that “Libertarians usually say that violence may legitimately be used to prevent either force or fraud.”  Do libertarians “usually” use the word “violence”?  “Coercion” seems more likely and more appropriate.  He continues that “according to NAP, the only legitimate use of force is to prevent or punish the initiatory use of physical violence by others.  And fraud is not physical violence.”  This is easily answered.  A fraud is an aggression because it violates the property rights that the relevant agreement establishes.  All this talk about “violence” is merely confused.

“5. Parasitic on a Theory of Property”

We are told that “Even if the NAP is correct, it cannot serve as a fundamental principle of libertarian ethics, because its meaning and normative force are entirely parasitic on an underlying theory of property.”  In fact, it need not be “parasitic on an underlying theory of property.”  It is true that some NAP advocates argue along the following lines: “aggression” is the violation of legitimate property, and legitimate property is only established using assumptions that libertarians independently argue to be legitimate (self-ownership, labor-mingling ownership, etc.).  That is because they don’t have an abstract theory of liberty from which to derive property.  However, if we say that libertarian ‘liberty’ is ‘the absence of aggression’, then we can interpret this in a pre-propertarian way.  Property comes into existence in a libertarian manner when that property does not aggress on (i.e., proactively constrain or interfere with) other people.  For instance, I make and claim this spear, hut, rabbit stew, at no cost or loss to you: you are not worse off as a result.  And if there is some vestigial cost or loss to others (for instance, you cannot now use the very same natural resources that I did), then we again resort to the MAP.  I hope the gist of this view is clear enough (I have written at length in other places to deal with myriad details, but some readers become lost in their own inaccurate paraphrases of the details without first showing that they have grasped the basic problem or the basic idea of the solution).  In this way, respecting liberty—as the absence of aggression—can indeed be the fundamental principle of libertarian ethics.”

By way of illustration, Zwolinski asks us to “Suppose A is walking across an empty field, when B jumps out of the bushes and clubs A on the head ...  If it’s B’s field, and A was crossing it without B’s consent, then A was the one who was actually aggressing against B.”  It seems worth noting that a disproportionately large retaliation itself becomes a new act of aggression.  I won’t give a theory of proportionality here, but it is derivable from the NAP/MAP.

Thus we can readily agree with Zwolinski that “‘aggression,’ on the libertarian view, doesn’t really mean physical violence at all.”  And we can even agree that “It means ‘violation of property rights’”—as a rule of thumb.  But property rights themselves can be derived from whatever control of resources does not aggress, i.e., proactively constrain or interfere with others (or, in the event of a clash, what minimizes such constraints or interferences).  Hence, it is false to say that “It is the enforcement of property rights, not the prohibition of aggression, that is fundamental to libertarianism.”  As we have now seen, it is liberty itself—interpreted as the absence of interpersonal aggression—that is “fundamental to libertarianism.”  That conclusion should not be completely astounding.

“6. What About the Children???”

Zwolinski then tells us “the NAP implies that there is nothing wrong with allowing your three year-old son to starve to death, so long as you do not forcibly prevent him from obtaining food on his own.”  An analogy might help to answer this point.  A child will not swim in the pool without a lifeguard.  You volunteer to be the lifeguard, and as a consequence he gets into the pool.  Then to allow the child to drown flouts the claim to your protection that you have previously given him: it is thereby an aggression against the child (positive actions are not always necessary to aggress against the claims we cede to people).  In a relevantly similar way, a parent has assumed a duty of care for the vulnerable person that he has brought into existence.  Negligently to allow one’s own child to starve to death is to flout that duty and thereby commit an aggression against that child.  Therefore, one has a libertarian obligation either to feed him or to discharge the parental duty by finding someone who is willing to take it on.  (I have some reservations about this position, but I won’t discuss them here.)

Consequently, it is incorrect to say that the NAP “implies that it would be wrong for others to, say, trespass on your property in order to give the child you’re deliberately starving a piece of bread.”  As the starving child is having his given claims aggressed against, anyone has a right to come to his aid in his defense.  Any duties that we create by our behavior, including but not limited to explicit contracts, may be coercively enforced if that is what is necessary to minimize any overall aggression.

Zwolinski then sums up his position with a few observations.  He first notes that “There’s more to be said about each of these, of course.  Libertarians haven’t written much about the issue of pollution.”  Is that correct?  For what it’s worth, when I typed “pollution” into I saw “465 results.”  Then he observes that libertarians “can think up a host of ways to tweak, tinker, and contextualize the NAP in a way that makes some progress in dealing with the problems I have raised in this essay.”  And, indeed, the Rothbardians have already done this with their interpretation of the NAP.  But Zwolinski concludes that “There comes a point where what you need is not another refinement to the definition of ‘aggression’ but a radical paradigm shift in which we put aside the idea that non-aggression is the sole, immovable center of the moral universe.”  However, this overlooks a third possibility: one can have a paradigm shift within the interpretation of what constitutes “non-aggression” (or ‘liberty’).  And this is what I claim to provide.

Zwolinski’s concluding sentence is that “Libertarianism needs its own Copernican Revolution.”  The analogy is more apposite than he realizes (although, of course, Aristarchus of Samos long antedated Copernicus).  For the “Copernican Revolution” that we can have here is to stop trying to theorize “non-aggression” (or liberty) ultimately in terms of legitimate property and do the reverse: to theorize legitimate property ultimately in terms of non-aggression (or liberty).  And with this approach all six given reasons to reject the non-aggression principle can be comprehensively refuted.

Yet I fear that this ‘revolution’ is seen as ‘heretical’ by some libertarians—where it has been noticed at all—and this is compounded by my ‘incomprehensible’ rejection of all supposed justifications in favor of the critical-rationalist epistemology that I apply.  And so I should just like to emphasize that this position is not a criticism of libertarianism or any kind of compromise with non-libertarian principles.  Rather, it is intended to clarify and unify much currently diverse libertarian theory behind a single principle of liberty itself.  With that aim, at least, real libertarians ought to have some sympathy.


1: Newtonian mechanics are refuted but often useful approximations for practical purposes.  The conjecture that people can fly by the power of thought alone, by contrast, appears to be refuted and not worth trying in any dangerous context.

2: And either a theory or its negation must be true (assuming the logical law, or principle, of excluded middle).

3: I call this the “classical liberal compatibility thesis” in Escape from Leviathan.  But David Goldstone points out (private communication) that the “compatibility” of liberty and welfare makes it sound like an unexplained coincidence, when libertarians usually believe that liberty is a major explicable and testable cause of welfare promotion.  I am partly inclined to agree.  In that case, “classical liberal causality thesis” is possibly clearer.  However, some of the overlap is for important conceptual rather than causal reasons, and only “compatibility” seems to cover both.


[1] Zwolinski, M., ‘Libertarianism and Liberty, Part 1: A Complicated Relationship’, 10th January 2013, retrieved 25th May 2013,

[2] Lester, J. C., Escape from Leviathan: Libertarianism without Justificationism, University of Buckingham Press, 2012 (2000).


[3] Zwolinski, M., ‘Libertarianism and Liberty, Part 2: Against Maximum Freedom’, 15th January 2013, retrieved 25th May 2013,

[4] Zwolinski, M., ‘Libertarianism and Liberty, Part 3: It Doesn’t Add Up’, 25th January 2013, retrieved 25th May 2013,

[5] Lester, J. C., ‘Kymlicka on Libertarianism: A Response’, Libertarian Papers, Vol. 4, No.11, 2012, retrieved 25th May 2013,

[6] Zwolinski, M., ‘Liberty and Property’, 4th February 2013, retrieved 25th May 2013,

[7] Zwolinski, M., ‘More on Property, Freedom, and Coercion’, 5th February 2013, retrieved 25th May 2013,

[8] Zwolinski, M., ‘Against Moralized Freedom’, 11th February 2013, retrieved 25th May 2013,

[9] Zwolinski, M., ‘Locke and Nozick on the Justification of Property’, 18th February 2013, retrieved 25th May 2013,

[10] Zwolinski, M., ‘Libertarianism and Pollution’, 1st April 2013, retrieved 25th May 2013,

[11] Zwolinski, M., ‘Six Reasons Libertarians Should Reject the Non-Aggression Principle’, 8th April 2013, retrieved 25th May 2013,