HXA articles

Stop Saying

The popular principle of freedom/non-coercion/voluntariness is unsoundly over-strong, and ultimately contradictory to any political form. It is theoretically incoherent and practically impossible. If we accept it, it would not allow us any political or ethical system, or even any structure at all.

4574 words (23 minutes)


The following book will be taken as an exemplar of non-coercion/voluntarism ideas, and a target for their general rebuttal:

"Against The State"; Sartwell; 2008 / ISBN-9780791474471 / book.

This is a pleasing enough book – informed, thorough, lively, and best of all, succinct. It is good as a neat summary of plausible justifications for political organisation. But the basic idea driving its criticism of those systems – a principle of freedom – is irretrievably flawed.

The root criticism against states is unsoundly over-strong, ultimately being contradictory to any political form. It works so forcefully only because it defeats everything. If we accept it, it would not allow us any political or ethical system, or even any structure at all.

Freedom is essentially opposed to structure – social, political, ethical; structure of any kind. The two trade-off. If you expand one, you diminish the other. And if you make one – freedom – the absolute sine qua non, you require the elimination of the other – structure. So it cannot make sense to make freedom the principle of political structure. A political system is a filtering mechanism, a way of deriving a single cohesive alignment from arbitrarily conflicting constituents. To only point at individual status as the ground of moral standing is to completely fail to understand and address the problem at hand.

That spurious principle of freedom, also going by the terms non-coercion, voluntarism, liberty, is quite widespread and a mainstay of much popular ‘libertarian’ stuff. So this article can serve as an accumulated and concentrated philosophical repository to break up that whole cluster of hokum. The intent of this article is neither to defend states nor oppose anarchism. It is to repulse the common superficial bad thinking about freedom/etc, and induce you to make better proposals and arguments.

There are two parts. First, a distillation of arguments against freedom/non-coercion/voluntary thinking. Second, an explanation of specific wrongnesses in the arguments for freedom/etc in the book. A very short summary of the main arguments rebutting the ‘freedom/voluntarism/non-coercion’ principle of the book is:

  • Coercion is intrinsic to the scarcities of the physical world.
  • Telling you what to do is essential to the purpose of political systems.
  • We are never free of the influence and constraint of others.
  • If there are no concentrations of power, there is no structure at all.

And a very short summary of some specific wrongnesses in the book's arguments is:

  • If freedom is a matter of degree, then freedom is now subservient to some other criterion, or it is always beaten by no rules at all.
  • If we are limited to instinctual/natural altruism, then we are limited to a primitivism of no artificial structure.
  • If the judgement is by freedom to join, then any system can be non-coercive, because nothing is specified about any particular system.
  • If ‘emergence’ justifies a system, then our current states are too.

Arguments against, distilled

Empirical: scarcity exists

Coercion is intrinsic to the scarcities of the physical world. So you cannot have any useful, effectual political system without coercion.


Consider this boiled-down scene, that presents a basic puzzle to voluntarism: what happens when two people's voluntary choices collide? What happens when my voluntary action blocks yours, and yours mine? If there is one almond, and you want it, and I want it, then one of us will go hungry. This is not solvable on any principle of voluntarism, because the realisation of one choice means the obstruction of the other.

Coercion is written into the fabric of the physical world as scarcity. All a political system can do is shift around / manage / optimise those scarcities, and hence the coercion inherent in them. ‘Non-coercion’ is impossible in the real world.


Is there a way of redefining coercion that will solve the puzzle? No. If you define coercion short of the very general idea of controlling people, then your political system cannot address the real problems of physical scarcity. It is useless in the real world. But, on the other hand, if you define coercion as including general control, and then reject that as coercive, you also render your system similarly useless.


Perhaps someone might bring up markets as a supposed counterexample: they are ‘non-coercive’ and exist, and operate well, right now. But of course they are coercive – that is exactly what the price system does, and how it works: markets constrain people's behaviour involving scarce goods. If the price the market sets for that almond is too high, you cannot pay and you do not get it. Markets control people's access to physical resources. That is the basic ‘raw material’ of coercion, and that is exactly their purpose and merit.


What if people agree? Is that a possible way out of the puzzle? No: just because you agree does not mean the coercion is gone: you still do not get the almond. If you look at the material facts, what is the difference? Having the almond taken from you, and you agreeing to relinquish it, both have the same result. How does you agreeing make things better? Why is agreement better?

This is a replacement of a material measure with some tenuous feeling of choice. Coercion measured purely by feeling, divorced from any material consequences, is meaningless.

What if half the population have a saintly submission and agree to anything. How does that improve things? You could say that (more) people get what they want. But this seems a feeble and dismal attitude: let us proclaim we are improving things, not because there is any real improvement at all, but merely because we have contrived to make people want less. You have replaced positive action with self-deception, by veering away from objective physical assessment.

Agreement, where and how much, is like a neutral measure of the structure. Agreement/consent is not good in itself. Consent is illusory here. To have any substantial rational meaning, this must in the end be a matter efficient distribution of resources.


To repeat, in summary, coercion is intrinsic to the physical world. You suggest a morality not infringing others' freedom? Then it cannot handle scarcity – which at bottom is the only substantial, real reason for moral/political systems anyway. If no-one can impair others' freedom, no scarcity can ever be used.

Logical: politics' purpose

Telling you what to do is essential to the purpose of political systems. So, again, you cannot have any useful, effectual political system without coercion.


What is a political system that does not tell anyone what to do, that has no rules, guides, structuring form? What does it do? It is empty of meaning. Morality/politics is essentially imperative, in its very purpose.

Maybe, you might respond, it could merely inform people of possibilities – it would not command, but offer. But it would have to offer you things you did not know of, otherwise be pointless. And those offers must result in changes to your choices and behaviour, also otherwise be pointless. But what is that but the thing you deny? What is changing people's behaviour by some influence outside themselves, what is that except controlling people?


Until you offer, as your political proposition, some non-voluntary non choice based structure you are not offering anything I can either agree or disagree with. If all you are saying is to ask ‘what do you want?’ you are waiting for me to offer structure, not offering any structure yourself.

You must suggest something, but then you are forced into voluntarism's dilemma of proposal.

Let you as a voluntarist propose any rules or rights you say are just, I will simply object that I did not volunteer for them. Now what? Either I will have brought down your whole system, or in order to override me you will be forced to recant your principle of voluntariness (which also brings down your system).

(You could insist that I can just go elsewhere, but that would be to assume away the basic problem here – contention of resources – and that voluntarism is inapplicable to the real world.)


Could ‘non-coercion’ be about seeking more consent/agreement? If we change the aim to one of degree of consent, will that make an escape from the dilemma? But is not more consent equal to less individual freedom? How can we be said to be more free, when we are more in agreement, when we conform more to others? We would not be free from each other, but then what else is there that we are free from of political interest?

The idea of an individual implies differences, which implies disagreement, which implies the need for a moral/political system, which implies constraint of individuals. The nucleus of morality is the elements of: general rules and individual agents. There is necessarily contention there, between the generality and the individuals.

Coercion is not only built into the fabric of physical fact, coercion is intrinsic to political structure in general. (And consent is never really an answer: it expresses no more than self-interest, which is exactly what morality, hence a political system, is meant to supersede.)

Freedom: is never absolute

We are never free of the influence and constraint of others. So, once again, any useful, effectual political system is impossible without coercion.


Here is a question for those who rely on references to freedom in their political opining. During pandemic lockdown we cannot leave the house much, but on the other hand we do not have to go to work – so overall, is this an increase or decrease in freedom? Why?

Can anyone answer credibly? Let alone lacking a measure of freedom, the two seem hardly even comparable. So we ought to be forgiven for exasperation at people principling their political philosophy on ‘freedom’, when they literally do not know what they are talking about.


We are never free. If freedom is to be unconstrained by others, then we are never in that condition. Our environment is made up of other people and their actions and effects. And every one of our decisions is shaped by our environment. Our actions are constrained by our environment, and our environment is made by other people's actions. We live surrounded with obstructions and compromises to our freedom.

Even if you can go into the wilderness, the fact that that wilderness still exists is only due to people's previous decisions and actions not to despoil it. Nothing you can do is free from others' influence.


But is not my freedom to read, or think, or look, not in conflict with anyone else? Are there not quite a few actions not involving scarcity, that can be done with no impingement upon any other's freedom?

Whatever you do limits my freedom to have the world another way. This seems an upshot of anything shared, such as, whatever we can talk about. You can argue (sometimes easily) who is wrong, but that indeed admits that there is a conflict. Whatever we know is the case excludes the possibility of it being otherwise. Similarly for what is proposed should be the case: whatever is proposed excludes the opposite proposition. We might happen to agree, but that says nothing of our freedom to.

What if I do not know what you are doing? Such a case escapes by being something that is not shared. What if I could never know, for example perhaps, what you are thinking? But that seems just an instance of something we cannot talk about.

The only way we can be wholly free is by each being isolated in their own world. But these cases would never be in debate, because we could not talk to each other. Wherever we can talk, we can disagree, hence are not free.


We can, a priori, never be free in any political system.

Really there can be no measure, or threshold concept etc, of freedom, because freedom is defined by the individual: freedom is an individual state/experience/assertion. You say whether you are free or not.

Whatever general measure or conception or criterion put forward for freedom, surely you would have to allow people to dissent from it. Surely anyone must be free to declare, on your measure, they are not free. They can say that the freedom your definition allots them constrains them from what they regard as freedom. If you allot them an amount of freedom by your measure, can they not reply that it constrains them more than they want? Freedom is necessarily individual/subjective, so it can never be a general principle for any moral proposition.

Any conception of freedom you made must amount to some bounds on what is allowed. But to any such rule I could say that it does not suit my choice. And surely you cannot argue that you deny my freedom, justified on the base that you are protecting freedom!

A political system principled on freedom must either contradict itself or say nothing.

Informational: structure constrains

If there are no concentrations of power, there is no structure at all. At the most fundamental level, organisation is intrinsically constraining.


Structure requires a shared causal source. Otherwise, how can the elements know what to do, how to relate to each other? Structure/organisation is about a correlation between elements, and that cannot be something each does by itself. The unavoidableness of this centralised connection seems just a kind of fundamental law of information.

And what is a concentration of power, but a shared causal source that shapes participants' behaviour? And what is that but the lack of freedom? So a freedom principle is at odds with structure and organisation at the deepest level.


‘Non-coercive organisation/system’ is at root a contradiction – or at best a trade-off. These are the argument's parts:

  1. Organisation is correlation,
  2. correlation implies shared information,
  3. shared information implies common causation,
  4. common causation is centralization,
  5. and centralization is coercion of individuals.

Two or more individuals, when uncorrelated, act randomly to each other – this is disorganisation. But to be organised those individuals must be correlated, and in not being random with respect to each other they must have some constraint between them. Correlation is not something each can do by itself, but something in common.

By acting in concert – joined by that constraint – something is shared, and the only thing that can be shared (copied rather than merely divided into pieces) is information. And the only non-infinitesimal chance of substantial information being the same in multiple places is by issuing from a single source.

So coercion, organisation, and centralization are inseparable: coercion is control from outside the individual (step 5), centralization is just a word for a single causative source (step 4), which is the only way to produce the same information (step 3), hence patterns of action (step 2), which to say organisation (step 1).


The most fundamental law of coordination is that it is defined by precisely that which is not decided by the individual. So to imagine a system ‘decentralised’-down only to individuals dealing with each other, makes no sense.

Could you decentralise property eg, removing any singleness underpinning it? Then it disappears: if each individual decides ownership, no-one owns anything. Property cannot be justified by appeal to the individual, because its domain is a contention of scarcity: every right to one is exactly a claim upon everyone else. Ownership, like any coordination, is not an act an individual commands, because every other individual, as such, will (in general) command the opposite.

It cannot be justified by freedom, since that is individual, but only about some common desiderata: it is a structure everyone must agree on, so it can only persuasively be justified by being something everyone benefits from.

Property, and any moral or political rules, are like language. There is no such thing as private property, or private political rules, any more than there is private language. And the same for any form of coordination.

Wrongnesses in the ideas for

A matter of degree

“Voluntary in the definition of anarchism above means, simply, uncoerced. As coercion is a matter of degree, so is voluntariness. Anarchism in this sense means that more voluntary arrangements are always preferable, other things being equal, to less voluntary ones.”
[§1.1.5 / p23]

“Anarchism is the view that all forms of human association ought to be voluntary, or it is the view that people ought to have maximal political freedom.”
[§1.1.5 / p23]


If freedom is a matter of degree, then either freedom is now subservient to some other criterion, or it is always beaten by no rules at all.


Such a political proposal needs to admit two things:

  • It has no cogent case unless clear about that degree: what is/not acceptable.
  • This is not about non-coercion/etc after all – since how much / what liberty/etc to allow is entirely dependent on that other criterion.

This is no longer about coercion: coercion is now subservient to something else. To decide what degree of coercion is acceptable, you have to appeal to some other prior model, and so that is now the principle.


‘Grab-what-you-can-world’ is a hypothetical moral/political system to elicit what ‘libertarian’s really believe. In Grab-World, there is only one rule: no-one is allowed to touch anyone else. Beyond that, anything is permitted: if you want to take or use or consume any part of the world, you can, with no restriction. Grab-World is liberty-‘optimal’ one might say, and a nicely simple kind of limit-case moral benchmark of the free-est possible political system.

Grab-World seems to ‘wrap perfectly tightly’ around physical freedom. If liberty is nothing more than freedom in disposing one's body – and what more can there be to liberty? – Grab-World only constrains liberty to the extent that exercising that liberty would constrain someone else's. There is a kind of zero-sum here in the basic physical circumstances, and Grab-World does not ‘add’ any more constraint: it only divides it up in a particular way.

So now we can say that, yes, like all morality, Grab-World affects individual liberty, but globally it does not add any more limitation. That sense of ‘liberty-conserving’ (or ‘efficient’) does seem to be significant: moral systems could vary here.

Property rules are certainly not so liberty-conserving as Grab-World – they do not fit so ‘tightly’, they add limitation in a negative-sum way: they restrict people's access to arbitrary pieces of the world, where no other person's bodily freedom would be imposed upon. So anyone defending property primarily on grounds of liberty must be driven to quit that alliance and take up with Grab-World instead – if they do not, we know that their intention is really about something other than liberty.

By posing Grab-World vs property, or indeed everything else, we can show that people do not think freedom is the main desideratum after all. And if they do accept Grab-World, we can counter-offer with a system of no rules at all: that must count as less coercive, and so win out. But with no rules at all, we are back to the problem of it being vacuous as a political system, and incapable of addressing real world scarcities.


If non-coercion is the absolute aim, comparisons drive you to the limit case of no political structure at all. If non-coercion is a matter of degree, you allow any political structure, depending on whatever other principle you have chosen as decider of that degree.

Instinctual altruism

“I aspire to an anarchism that: […]
(4) does not regard human beings as merely self-interested or incapable of altruism; […]
(8) is not an amoralism or an immoralism or a libertinism but is compatible with and encourages conscience and self-discipline;
(9) does not prescribe any particular set of social arrangements but leaves such arrangements to the voluntary decisions of individuals.”
[§5 / p99-100]


“Here, I do not intend to assume that people are basically benevolent.”
[§1.0.2 / p5]


If we are limited to instinctual/natural altruism, then we are limited to a primitivism of no artificial structure.


One approach is to appeal to voluntary association: people simply decide to alliances here and there as they will or not, and that is what the system is.

There is something of a trick here: the constraint must still be there to align behaviour, it has just been found internally in each person – they choose a particular pattern of behaviour, so they are constrained, but doing it themselves. Each could choose otherwise, but have here chosen to unify. It is ‘mathematically’ constrained, yet individually voluntary.

It has persuasiveness: the person decides themself: that decision is swayed by their moral intuition/capacity, but that is still them. There is a clear boundary here – around each person – and the decision comes from within it. We could raise the question that if one constrains oneself, is that not still a constraint? However, let us accept here that this will count as (personal) choice.


But the instinctive mechanism of constraint has practical deficiencies. Any external constraint is coercive, but internal constraint is voluntary. But what counts as external? Government? Certainly. Money? Most likely. Any rules? (As an exemplar) does property count? Locally, one could say a person has an intuitive feel for who owns what – no need for any external mechanisms like records. But that will not work at large scale. You could reply that one need not know the details, merely the custom, which you have intuitively: wherever you go, if people say they own something one accepts it. But now it seems to reach a problem: how can this be distinguished from government? When the police say you are doing something wrong, accepting that is a custom.

This expansion must be denied, otherwise any external constraint will be permitted by becoming customary, and the ‘instinctivism’ argument collapses. So there must be some low bound on any kind of custom. But then how can sophisticated organisation work – at scale, reliably? Detailed constraint must be represented in information somewhere. An intuitive custom will not suffice. Each person cannot have a record of every monetary status and transaction in the world economy, and that system only yields its benefit because it does capture such detail.


The purpose, value, of artificial, external, tools is to enable what humans cannot do themselves. So, the hard choice of primitivism: either you have internal, intuitive, customary, voluntary system, or you have a system that works at scale, but is external therefore coercive. Anything that humans are incapable of instinctively is exactly where external artifice helps to produce sophisticated systems.

Coercion is just the necessary accompaniment of artificial means of organisation. If you want sophisticated organisation you must accept coercion.

Free to join

“Bowling leagues are anarchist organizations; no one is forced to join and no one has to be forced to obey the rules. Rather, enjoyment of the activity depends on the fact that it is rule-governed and that by and large the rules are observed.”
[§1.0.7 / p11]

“Recently, internet events and approaches have provided anarchist models. The basic ideas of open-source software and wikis (websites anyone can edit at will) suggest that optimal orderings of tools and information, at least in some situations, emerge through the unconstrained collaboration of many people.”
[§1.0.7 / p12]


If the judgement is by freedom to join, then any system can be non-coercive, because nothing is specified about any particular system.


If you can choose to join a state, is that state then voluntary? But this says nothing about the state. It could be the most repressive, coercive arrangement possible, yet since it is free to join it thereby is rendered voluntary/non-coercive etc.

The magic addition of ‘free to join’ transforms any state into a voluntary one. This has happened because a trick has been played in what is being talked about.


‘Free to join’ just moves the criterion outside the actual proposal. A freedom to join an organisation is a freedom outside the organisation, not of the organisation itself. It says nothing about the organisation, and not only that it says nothing useful outside it either.

If the proposition is that joining the system is voluntary, then it is vacuous. This is because it has an implicit half – it is really saying: anyone can join this system, or do anything else unspecified and so unbounded. It says: you can follow the rules, or not follow the rules. So overall, this proposes a rule that permits anything, hence means nothing.

Furthermore, the ‘membership is voluntary’ idea would only work in a world of abundance: wherever there are scarce resources (the real world), it fails, because it does not address contention. To organise use of a supply of land, one cannot propose ‘you can join the property system, or not’. A property (or any other) system where one is free to not follow the rules is not a property system.


The idea of voluntarily joining groups, and in the event of any dissatisfaction, set up and join a new group – this is fantasy-land, predicated on being able to always expand into unoccupied, abundant space.

What happens when a member decides another's interpretation of the ‘rules’ is wrong? And they do not want to leave? When faced with conflict in the domain itself, voluntarism clearly solves nothing.

A moral/political system that presupposes abundance is otiose. Instead of answering how to arrange resources to give people certain desiderata, it says ‘everyone already has all they want’. The solution it offers is to propose that there is no problem in the first place – it begs the question.

A moral system must be either involuntary or incomplete – it cannot be both voluntary and complete. To cover the domain of concern, the rules have to tell people what to do, and then it is not voluntary. If the rules do not tell people what to do, you have not proffered a complete system.

Emergent is OK

“Within political philosophy, anarchism is the position that we should let go and see what happens.”
[§5.0 / p115]


If ‘emergence’ justifies a system, then our current states are too.


What we have now politically is what just ‘happened’ – we are living in an ‘emergent order’ right now. Where did this tangle of sub-systems come from? Who was the overall planner? There was none. This is self-organised.


And what does ‘letting go’ mean? If it truly means no constraints at all, it runs into all the problems surveyed in previous sections. Otherwise, it needs to explain what interventions it does propose.

The only substantial way to make the emergent argument is to distinguish its scene from the ‘what just happened’ of what we have now. You would need to specify either particular start conditions or an ongoing context and bound of activity. But then it would not be ‘letting go’ that defines the proposal, but those particular constraints.

‘Letting go’ defines nothing and says nothing.


Branding a proposal with liberty/freedom/non-coercion/voluntariness is bad because it is evasive or misleading. There cannot be a useful political system without some constraint of freedom. So to present one as being principally about that is to hide what is going on. Either you are proposing something surreptitious, at best tacitly assumed, or proposing nothing at all.


DC: {
   title: "Stop Saying 'Non-Coercive'",
   creator: "Harrison Ainsworth",

   date: "2020-10-25",
   date: "2020-07-17",

   type: "article",
   format: "text/html",

   language: "en-GB",
   subject: "philosophy, morality, political, anarchism, liberty",
   description: "The popular principle of freedom/non-coercion/voluntariness is unsoundly over-strong, and ultimately contradictory to any political form. It is theoretically incoherent and practically impossible. If we accept it, it would not allow us any political or ethical system, or even any structure at all. (4574 words)",

   identifier: "urn:uuid:6B09ADCF-535F-4229-B152-B03FABD71C89",
   relation: "http://www.hxa.name/articles/content/Stop-Saying-NonCoercive_hxa7241_2020.html",

   rights: "Creative Commons BY-SA 4.0 License"