From Metaphysical Freedom To Civil Liberties: A Cybernetic Account Of Natural Law

This essay was originally an intended submission to les journées de Rochebrune 2005, but it is both far too long and far overdue. Maybe I will finish it some day, and then have a condensed version of it published to a similar conference afterwards?

Note that this is a preliminary draft, subject to revision. Parts that haven't been redacted yet have been removed. Feedback is welcome.

Abstract: We clarify the meanings of ``liberty´´ and related concepts. We notably distinguish metaphysical freedom or liberty in fact, from civil liberty or liberty in law. We will examine the properties of these concepts and related ones such as free will or individuality, and the relationship of all these concepts with each other. Along we way, we disarm the most common fallacies about liberty and reframe the debate using arguments from Praxeology, Objectivism, Phenomenology, Algorithmic Information Theory, Cybernetics, Darwinism, and Libertarian Law Theory.

Ideas have consequences. The consequences of Ideas about Liberty are that institutions are established that will preserve those liberties we acknowledge and understand, subvert those liberties we acknowledge but do not understand, and crush those liberties we fail to acknowledge.

Certainly, the deep examination of the meaning of Liberty is something that few people care about, and any philosophical debate on that topic is bound to find but few interested parties. But it is those few parties who will serve as authorities to others; it is those few parties who will set the tone for the casual discussions of educated men; it is them who will shape the sense of life of influential intellectuals, teachers, authors and artists, and through them the culture of future generations. And thus do philosophical arguments about Liberty ultimately matter.

Errors and fallacies about Liberty ultimately lead to oppression and suffering. Understanding of Liberty, what it is and what it isn't, ultimately leads to freedom and happiness. Lies and distorsions about Liberty are the ultimate tools of all tyrants through the establishment intellectuals that are their minions. Truths and Rectifications about Liberty are the ultimate defense of free men and the legacy of the authentic free thinkers.

There are many quite distinct meanings commonly ascribed to the word ``liberty´´, and to variants of this word such as ``freedom´´ or ``free will´´. People who argue about ``liberty´´ and its variants often get confused about these distinctions and find themselves unable to understand each other, or even prone to fall into absurdities and fallacies due to some semantic shift between several related but nevertheless distinct meanings of the word.

We propose to examine those main meanings of the word, mutually contradictory or self-defeating as they may sometimes be. We will disarm fake problem settings and other common philosophical traps, and reframe the concepts of Liberty where they are relevant. We aim to offer a clear vision of what Liberty is, what it isn't, why it matters, and how it matters.

In a first part, we will examine some common philosophical arguments about Free Will in terms of Determinism, Soul, Transcendence, etc. We will explain that they are fallacies based on absurd premises, and why they altogether fail to address any actual attribute of man. To achieve this explanation, we'll provide sound conceptual tools to reframe the debate correctly: praxeology, phenomenology, cybernetics.

In a second part, we will propose a notion of Free Will compatible with our man-centered approach: Epistemic Freedom. We will use information theory to argue that man is indeed intrinsically free according to this notion.

In a third part, we will explain how this notion of Epistemic Freedom is relevant in the study of social interaction. This study will lead us to another, quite different kind of Freedom, Civil Liberty. We will argue why Civil Liberty is beneficial in terms of what we previously established about human nature.

The first concept of liberty that I'll tackle is what I'll call metaphysical freedom. Metaphysics is the domain of fundamental properties of things, that necessarily follow from their very nature and existence ^[1]. I am metaphysically free to do something if in fact I somehow can do that thing, that is, if there is a possible, conceivable future in which I actually do that thing.

Thus, say, I am not metaphysically free to jump to the moon, or to turn into an eagle, because they are things I can't possibly do. Neither am I free to stop breathing for several minutes and still continue living. Nor am I free to sell the transcript of this speech for three hundred million dollars to the first comer.

Or am I? After all, what is? What can I or can I not really do? Can't I possibly conceive that I would jump to the moon? In Cyrano de Bergerac, the main character imagines many ways to jump to the moon, including pulling himself up by his bootstraps. So can I conceive my jumping to the moon. Lemme focus on the idea... here I did, how pleasant it was! — Maybe you'll tell me that this idea was ridiculous on its face, an impossible fantasy. But how do you know? What would you have replied in 1661 to Joseph Glanvill who wrote:

How can you draw the fine line (or not so fine line) between the actually real, the merely possible, and the outright ridiculous? And what if I used the help of a spaceship to achieve this jump? Would it be cheating or would it be a valid reframing of the concept? You might object that I utterly lack the opportunity to ever have such a spaceship ever built for me. How do you know? Wouldn't you have told the same to a young Neil Armstrong, too? Where is the line between opportunity and possibility? Is this line actually well-defined, or are the limits of reality fuzzy?

And what if I'd change my career plan, so as to become an astronaut, or a great scientist, or a successful space entrepreneur, or a billionaire, or an evil overlord, so that I would create the currently unimaginable opportunity of my jumping to the moon? What if we'd been talking about something else than jumping to the moon? Aren't there things you believe impossible and that will nonetheless happen? What if the world were just a dream or an illusion, and things were actually very different from what they seem? What if some miracle suddenly happened or ceased? Does anything go? Isn't everything possible, or at least possibly possible?

``What if´´: Does the question have a meaning at all? Is there any ``elbow room´´ left for metaphysical freedom? Or on the contrary is it a concept so large and so vague as to be useless? Do we possess any kind of free will? If so, what is it? Does it matter?

One way in which the debate on metaphysical freedom has often been framed is that of Fatalism versus Free Will, and its many variants, from the ancient religious arguments about Predestination, to the modern discussions about Physical Determinism.

According to some philosophies, each man has a Fate already decided before his birth; this Fate governs his deeds and his lot; he cannot escape it. In ancient myths and religions, this Fate or Predestination may have taken the form of a Great Book of Life where ``everything is already written´´, or an otherwise Omniscient God aware of all events past, present and future. In modern parlance, this Fate takes the form of people being the mindless puppets of some external ``determinism´´, be it genetic, social or psychological, and ultimately physical.

Equal and opposite philosophies argue that man may have ``free will´´ only inasmuch as the above Predestination does not exist or has no strength: there is no Fate, or it can be convinced to change its mind; there is no Book of Life, or it is still to be written; there are no Determinisms, or they can be broken. Finally, absurdist philosophies just give up reason, and accept several incompatible variants of the above as simultaneously or alternatively true in what constitutes a ``Great Mystery´´.

With the advent of modern science, this problem of Fatalism has been formalized and displaced as that of Physical Determinism. According to this point of view, the important question is whether the equations of Physics that describe our Universe are deterministic or non-deterministic; that is, whether or not they are such that everything in the future is exactly determined by the state of things in the present.

As the argument goes, if the Universe is deterministic then considering the state in which the Universe currently is, all the future events that will happen mechanically follow and each of us is only free to do but what he will do, and nothing else. In other words, everything is already written ^[2]^[3], there is as little freedom as possible, that is, none at all, and any semblance of freedom is a mistaken delusion borne in ignorance. Only if it is non-deterministic, if future events are not ``already written´´, if several outcomes are possible, is there room for metaphysical freedom ^[4]

Such formalism brings little in terms of philosophically meaningful statements; clever as they may think they are, physicists are just rehashing the same age-old debate. However, formalism has the advantage that it makes it clearer what the claims of each side of a debate are, and easier to establish those that are tautologies and to reject those that are meaningless absurdities.

And so the question is, how do we know whether the Universe is Deterministic or not? Can we know? Was the Universe sold with an officially certified label including a checkbox ``deterministic´´ that is either checked or not?

The very discussion whether or not the universe is deterministic presupposes a conceptual setting of many possible universes being considered out of which ``our´´ Universe may be one. Hence, simply asking about determinism requires some ``freedom´´ at a metaphysically more basic level.

The above argument is one of performative contradiction: a hypothesis is established as true if any action that would claim to deny it would contradict itself; the denial would presuppose the denied assertion. This argument can be systematized as Praxeology, or the Ontology of Human Action: the study of things from the point of view of how they ultimately affect the way that men act. For example, if I were try to logically argue that logical arguments are useless, I would be obviously contradicting myself, and my argument would therefore be wrong ^[5].

Of course, entrenched proponents of determinism might say that all these considerations about the freedom implicit in human action are but illusions based on misunderstanding of man's ``ultimate´´ predestination, of his lack of ``absolute´´ liberty. Still, our actions establish that as far as we discuss, choose and plan, there is freedom, to some extent. There only remains to establish the nature of this prerequisite freedom implicit in taking the argument seriously. Our approach will be to elucidate mistakes made by people focused on physical determinism, so as, by contrast, to gain insight about the real issue.

Phenomenology, according to Heidegger and after him cyberneticians of the third generation [B22], consists in studying phenomena within the framework that makes them meaningful, excluding things that are irrelevant because they do not affect meaning. Such a framework is called an ontology. By definition, an ontology embodies all that matters about some phenomena, nothing less, nothing more, so that anything outside of it doesn't matter, and anything that matters is inside of it ^[6].

Thus, consider the universe as observable by people inside the universe; in the phenomenological perspective, this universe is actually an Ontology of Human Action, comprehending everything that humans may observe and act upon, and the feedback that may they receive from such action. Whatever matters to Human Action by definition is part of this Ontology, and thus ``exists´´ in the Universe, whereas whatever doesn't matter to Human Action by definition isn't part of this Ontology, and doesn't ``exist´´ in the Universe.

Note that in any ontology, a piece of information may matter only inasmuch as it is used in a discriminating way by some observing actor inside the ontology to select or otherwise affect his future behaviour. Thus, although some ontologies may initially try to focus on the study of passive structures such as abstract grammar and mathematical algebras, they must always include some kind of action or interaction at some higher level, such as uttering sentences or using mechanical devices. Therefore, ontologies are ultimately all rooted in the Ontology of Human Action.

Once it is understood that metaphysics, like any human science, can only have meaning within an Ontology of Human Action, we see that the question of physical determinism vanishes into irrelevance and inexistence: because the very same structure of events has many models that are equivalent as far as what humans observe in interaction with what they do, yet some models are non-deterministic while others are deterministic.

Indeed, you can always ``complete´´ a given non-deterministic model of the universe into a deterministic model, in quite a straightforward way: for each place where the model says there is or might be a non-deterministic choice, add to the description of the universe some variable that determines the outcome of said choice ^[7]. If you have probabilistic constraints, adjust accordingly the distribution of the variables and the way they combine into various outcomes ^[8]. Or if you prefer, you may keep but one global variable ``source of randomness´´ that helps resolve all this (probabilistic) non-determinism; the infinitely many possible decompositions of this global variable into factorings of simpler variables follow from the structure of the universe. ^[9] Conversely, you can always ``quotient´´ a deterministic model of the universe into a non-deterministic model, also in quite a straightforward way: identify all states of the universe that we can't possibly distinguish by observation, and as for the structure that relates these states, instead of the deterministic relations between the original states, consider the non-deterministic relations between these classes of undistinguishable states. Depending on how much you realize that you can't ``actually´´ or ``possibly´´ distinguish things in the universe, the amount of non-determinism in your model will increase; and since we are far from gods actually able to distinguish everything, there will be a lot of non-determinism ^[10].

As Daniel Dennett points out [B4], we may never know if the universe is deterministic or non-deterministic, or if it switches between deterministic and non-deterministic every other day: no possible human observation can possibly tell one from the other, and for good reason. Therefore it follows that to us humans, the question of whether or not the universe is deterministic is undecidable. And this undecidability entails that in the ontology of human action, the question is irrelevant ^[11].

To use a mathematical term, determinism is not an intrinsic property of the Universe: it is merely an arbitrary property of models of the Universe. The map is not the territory. Models are not the Ontology. The choice of a model is not intrinsic to the ontology. The deterministic or non-deterministic nature of various models is as relevant to the ontology as the feminine or masculine gender of the word ``sun´´ in various languages is relevant to the structure of the star that lights our world ^[12]. It is not even an intrinsic property of our universe that its models should be physical: Physics is not a be-all end-all paradigm that explains all that is relevant about Human Action; other non-physical representations of our Universe are thus not only possible but necessary ^[13]^[14].

As established, the question of Determinism is irrelevant to metaphysics; making this question prominent in a metaphysical context is a Categorical Error. By examining the discrepancy between the physical ontology in which Determinism does have meaning and the proper ontology that matters for metaphysics, we can fully deconstruct the logical fallacy beneath this error.

Despite its scientistic veneer the interpretation of metaphysical freedom in terms of physical determinism is an instance of the all too mystical fallacy of transcendence.

Transcendence is the self-contradictory concept that there can somehow ``exist´´ things that are at the same time outside a currently considered ontology, yet meaningful within said ontology. Since an ontology by definition includes everything that is meaningful, such a notion is absurd: either something is meaningful and part of the ontology, or it is outside the ontology and meaningless. Once you adopt the phenomenological point of view, transcendence appears as obviously bogus; yet, it is a kind of fallacy that keeps being repeated over and over.

Now, the debate on the deterministic nature of the Universe critically assumes a One True physical model of the Universe, a complete description of it in terms of mathematical equations ^[15]. Yet, no model of the universe, physical or not, can both exist within the universe itself and be complete at the same time: only some kind of superior entity outside the model could bless the model as being the One True Model by interacting with it; it doesn't matter whether this godly entity is a personal god playing with his toy, an impersonal computer running a physical simulation, or whatever supernatural entity makes the model meaningful in its own meta-universe; it can't be both inside and outside. If it's inside the model, then a yet superior entity is required to give meaning to the actions of this entity, and the model isn't actually complete; and if it's outside the model, then the model is admittedly not actually complete either. Consequently, to claim that a One True Model could serve as the basis for interaction by a transcendent god both inside and outside the universe is a clear instance of the fallacy of transcendence.

If any kind of superior entity observably affects the outcome of anything in the universe, then this god is part of the universe's ontology, and any model of the ontology would have to include a model of this god's interactions and therefore be out of reach of this god itself; thus, the god might be somehow superior to men, he would not be transcendent ^[16]. The hereto unsubstantiated hypothesis of such a superior god thus pushes back all the questions up one level, but doesn't solve anything: is this god's behaviour deterministic? If a model is necessary for some system to be meaningful, and if some entity superior to it is somehow necessary to make a model meaningful, then where is the meta-god who'll make it meaningful for god to watch the universe? And what about the meta-meta-god, etc.? Wherever we stop, we must assume an ontology that stands without the need for a superior entity; and if we do not stop, then this infinite regression is no explanation at all, involves no global model of the ontology, and doesn't achieve the chimerical transcendence.

The fallacy of transcendence can be found in many attempts to explain freedom. The theory of transcendent souls ^[17] is a simple mystic variant of it; coordinated quantum wave collapse ^[18] is a more elaborate pseudo-scientific version. The fallacy can be recognized wherever singular metaphysical sources of freedom are invoked; these sources supposedly act upon the observable universe without being acted upon by it in any observable way. Said sources of freedom are fake explanations; all they really do is give a name to that which they refuse to examine and claim is beyond examination. The proponents of this fallacy pretend to have explained what they have blindly assumed ^[19].

People who seek an explanation of metaphysical freedom in a transcendent source also miss the point in a crucial way: their arguments could possibly claim that ``the universe´´ is free, but they cannot claim that men are either free or not free within it — because their theory is completely oblivious of men.^[20] By connecting liberty directly to a transcendent source instead of relating it to the essential nature and structure of man, these arguments short-circuit the middle man; they leave man a puppet in the hands of outside souls, gods and other imaginary spirits. Inasmuch as a man's behaviour is influenced by some kind of soul, quantum wave coordinator, personal god or impersonal random event outside of observable nature, how can we affirm that this mystical essence makes man free? Quite to the contrary, it binds man to events that are external to him — and indeed external to nature itself. Inasmuch as man is influenced by these events and cannot control them, how can we claim that the sources of these events are constitutive of the nature of man? These events, if they exist, are at best parasites, accidents or noise that interfere with man's behaviour. ^[21]

Any metaphysical theory of man's freedom must include or start from a theory of man. If we are to discuss whether individual human beings are free or not, we need a theory of what an individual is: what makes him distinct from his environment, from his peers. Furthermore, any attribute that an individual may possess must fall within the criteria that distinguish the individual from his environment; otherwise, it fails to actually belong to the individual.

Most of our actions assume the notion of individual: they suppose that in the world we act in there are distinct entities called individuals. Each individual is physically separate from his environment and autonomous from it; this environment includes his peers, other individual entities. The distinction is relevant because it determines the success or failure of important interactions. Moreover, each of us, including you and I, is such an individual. For example, if I know a secret, I won't expect you to be able to act on it. And if a poisonous snake bites you, you won't expect that administering the antidote to me will save you. ^[22]

An individual acts as such when he follows an identifiable character, a systematic persona, constant traits, an internal structure, etc. ^[23] When his behaviour is governed by randomness, by a fluke, by external miracles and coincidences, then he is the victim of external influences rather than an individual — and it is precisely because that is not the case, because he does behave according to certain characteristic patterns, susceptible to observation and negociation through inter-personal interaction, that we like or dislike, trust or distrust, befriend or avoid a given person, and that we recognize his existence as an individual to begin with.

People do not indulge in random absurd acts without structure or suit but to make plans and follow patterns that follow their individual character, personality and interests. This is an objective fact and a praxeological truth ^[24]. It is according to their nature as individuals in the observable universe that we may find or fail to find in men any meaningful kind of liberty.

In the preceding section, we have not only dispelled some all too common misframings of the concept of Liberty, we have also laid down the basic elements upon which to restate the question: What does it mean to be free?

What we saw is that this question was lacking both a subject and an observer: on the one hand, we need to relate freedom to an individual who can be free or not; on the other hand, we need to find an actual point of view from which it may mean something that said individual be free or not. We must specify who is or isn't free, and from which point of view.

As far as points of view matter at all, the question can often be simplified: Who is or isn't free, from whose point of view? And that's how, starting from the point of view of Human Action, we see that the question that matters with respect to liberty is whether a given individual A is or isn't free from the point of view of given individual B.

Note that the answer to such question does not depend on the opinion of B as such. B may be right or wrong in assessing that A is or isn't free from B's point of view. Ultimately, inasmuch as we'll find a meaning for freedom, whether A is or isn't free from B's point of view is an objective fact; B may be right in assessing it, and make plans that depend on this assessment and succeed; or B may be wrong in assessing it, and his plans will be jeopardized when they depend on this mistake. That's where the expression ``B's point of view´´ can be misleading. What we actually mean is ``B's point of being´´, properties intrinsic to the relationship of things to B, as considered in the Ontology of Human Action; or equivalently, properties intrinsic to things as considered in the Ontology of Being B.

We'll use the shorter expression ``relative to B´´ instead of ``as considered from the Ontology of Being B´´; we may also use ``as far as B could possibly tell´´, where this ``possibly´´ includes all the interactions into which B may or may not enter in the quest to tell the difference and act upon it ^[25]. And this shall be well distinguished from ``in B's opinion´´, as B may have opinions that may be correct or incorrect, meaningful or absurd, and that may or may not adequately describe the world in which B nonetheless lives. B's opinion, in addition to whatever biases and limitations constrain it, may only take into in consideration the interactions that B actually already entered into.

To further justify this distinction, a lot of living beings live without having any opinion at all about a great number of topics that nevertheless are relevant to the way they live; many living beings are not even sentient, and the notion of ``relative to them´´ makes sense, but ``in their opinion´´ doesn't. A may be acting in ways very relevant to the future of B, and B may have a correct understanding of what A is doing, or B may have an incorrect understanding of it, or B may not even have an understanding of it, B may not aware of it at all, assuming B has some kind of consciousness.

If you have skimmed through the above paragraphs without paying much attention, be aware that they were not a mere digression, or an idle nitpicking; they pinpoint the essence of what we'll argue liberty is all about: the discrepancy between A's behaviour as it does affect B, and A's behaviour as B may know it — what I'll call Epistemic Freedom (of A relative to B) ^[26].

If indeed we go back to the initial question of metaphysical freedom, ``Can or cannot A do X?´´, and put it in the only valid metaphysical perspective, the perspective of Human Action, ``as considered from the Ontology of being B´´, then it follows that the notion of metaphysical freedom that matters is ``Can or cannot A do do X, as considered from the Ontology of being B´´. Or if you prefer, ``Can or cannot A do X, as far as B could possibly tell´´ and act upon, through whatever kind of interactions B may engage into to know.

Now, this elusive notion is not reachable, as it requires omniscience of all such possible interactions to give answers. Interestingly another approximate notion is definitely reachable: ``Can or cannot A X, as far as B actually thinks´´. And you get your answers simply by asking B's opinion, or better, by seeing what opinion B actually acts upon.^[27]

All these notions are related but distinct variants of Epistemic Freedom. More variants are possible, and all variants may have uses. The important point is that ultimately, humans, as actors in the universe, act and interact according to the finite extent of their knowledge.^[28] That knowledge encodes what may be known as to the future behavior of a given actor.

(XXX later: is there circularity between that CAN and that MAY the "relative to" is indeed somewhat tautological the "as far as I know" isn't. returning the things: freedom as the default, knowledge as the probability, sheafs, combo. ?)

So, what is the nature of the metaphysical freedom implicit in any discussion about freedom? What to reply to the objection to metaphysical freedom by radical determinists that the whole discussion is actually meaningless and that any a priori non-obviousness of this meaninglessness is an illusion?

The freedom implicit in this discussion is what we may call ``epistemic freedom´´, the freedom of things we do not know yet that we may discover through further interactions. And the radical determinists' objection is irrelevant, because nobody may possess perfect total knowledge that can be acted upon.

In a first, static approximation, we may equate epistemic freedom with ignorance: we consider something as free inasmuch as its outcome or behaviour is unknown. The freedom of phenomena, of people are thus a direct expression of our intrinsic ignorance of the nature and state of the universe. This ignorance is what praxeologists call ``radical´´ ignorance: an irreducible ignorance that is significant because it is consubstantial with the nature of man as an actor in the universe. This ignorance cannot be dismissed as a mere defect. It isn't a circumstantial mistake that may be corrected by a proper attitude, either scientific inquiry or religious faith. It is a fundamental characteristic of any actor of any possible world.

But from a dynamic point of view, epistemic freedom is quite opposite to ignorance. Epistemic Freedom is not the absence and oblivion of information, but quite on the contrary, it is the openness to further information, which implies the gathering and positive use of ever more information, yet accumulating but ever partial knowledge. As contrasted to ignorance, this openness doesn't increase when knowledge is weakened, but when knowledge is improved; it doesn't increase when interaction is reduced, but when interaction is widened. From the point of view of a mindless piece of rock, ignorance is at its zenith, but epistemic freedom is at its nadir: the world has the least epistemic freedom from its point of view, because it is oblivious to any signal, because any information is irrelevant to it, because it won't interact with its environment, because it will react passively to outside forces, because it will never know anything ever; the world does what it does to the rock, it reacts how it does, and nothing matters. From the point of view of a rational human being on the other hand, who can interact with its environment in intelligent ways, who can learn many things, epistemic freedom is at its culmination: there are a great number of things that he may learn.

Epistemic freedom is interesting at its frontier, which grows with Life: it is what a living actor is interacting with that hasn't been integrated, the external, the future.

NAAAH - Exchanging Knowledge on things that matter to us for Freedom on things that don't. The above is too much of a BULLSHIT ratiocination of "it's good" from an emotional point of view.

Notes

[1]: Metaphysics is called after a treatise by Aristotle, where he discusses these fundamental problems of existence, what he calls ``first philosophy´´. This treatise was a compilation of fragments put together as part of a reference edition of Aristotle's works by 1^st century A.D. scholars; said edition included this treatise after the treatise on Nature, and thus called it ``after nature´´, or (in Greek) ``metaphysics´´.

Here is how Ayn Rand summarizes what metaphysics is in Philosophy, Who Needs It:

Are you in a universe which is ruled by natural laws and, therefore, is stable, firm, absolute — and knowable? Or are you in an incomprehensible chaos, a realm of inexplicable miracles, an unpredictable, unknowable flux, which your mind is impotent to grasp? Are the things you see around you real — or are they only an illusion? Do they exist independent of any observer — or are they created by the observer? Are they the object or the subject of man's consciousness? Are they what they are — or can they be changed by a mere act of your consciousness, such as a wish?
The nature of your actions — and of your ambition — will be different, according to which set of answers you come to accept. These answers are the province of metaphysics — the study of existence as such or, in Aristotle's words, of ``being qua being´´ — the basic branch of philosophy.
As a clarifying response to the terseness of Aristotle himself, though, see Paul Graham's essay [B19].

[2]: Albeit in the language of a physicist God.

[3]: Note that it can be considered as either Determinism or Non-Determinism the point of view that says that not one outcome is determined, but a probability distribution of outcomes, with any possible ``choice´´ of observations being an external cursor exploring said outcomes according to said distribution.

[4]: Or maybe we can call it physical freedom at this point, since it follows the point of view of the physicists.

[5]: The most famous appeal to this line of reasoning in philosophy is of course René Descartes's cogito ergo sum. The general argument was formalized by Ludwig von Mises ; he called Praxeology the science of things can be established through this argument by virtue of the nature of Human Action [B15]. The first thinker who systematized the use of praxeological proof in philosophy is Ayn Rand [B2]. For modern accounts of the epistemological implications of praxeology, see in French the PhD thesis by François Guillaumat [B6], or in English, works by Hans Herman Hoppe [B13].

[6]: Note that it is this approach is not obviously well-founded: after all, the idea of ontology does not guarantee that the concept is valid, and it may well be that this definition leads to contradiction. Indeed, in foundational mathematics, logical formalisms have to take extra precautions about ``existence´´ so as to avoid contradictions such as the Russell Paradox; a typical solution consists in layering ``existence´´ into an infinite hierarchy of ever more abstract ``types´´ or ``categories´´, etc. We'll assume that some discipline to avoid paradoxes will be used when manipulating ontologies, and reject those theories that rely on hypotheses that lead to such paradoxes.

[7]: These variables might be said to have been previously ``hidden´´, to be revealed when and only when that choice happens and its effects are observed.

[8]: What the standard ``objections´´ to ``hidden variable theories´´ in Quantum Physics actually establish is that such hidden variables can't be local in space-time. Although another way of seeing things is that it tells us that such variables are local in space-time, by very definition, but that in quantum physics, locality doesn't match the expectations of classical physics; i.e. the topology of the bifurcating multiverse of quantum physics is different from the topology of the bifurcating multiverse of classical physics. Big fat DUH.

[9]: Indeed, the computational tools with which we explore our models casually simulate non-deterministic models with deterministic computers by ``rolling dice´´ using a pseudo-random algorithm that literally has hidden variables constituting its internal state!

[10]: Yet again, the computational tools with which we explore our models often approximate deterministic models with non-deterministic simplifications, fuzzy sets, and lossy algorithms that try to keep the relevant information while dropping irrelevant noise and expensive secondary details.

Also, sensitive dependence on initial conditions as well as precision errors introduce chaotic and anomalous behaviour in deterministic models that to a human observer may be indistinguishable from non-determinism.

[11]: Though he does implicitly invoke the argument, Dennett fails to conceptualize the general notion of irrelevance in his book. I haven't read enough of his works to say whether this is out of deliberate choice or out of ignorance either of the notion of irrelevance or of its own relevance. Despite this lack of abstraction, or maybe because of it, the book is nonetheless very relevant to the exploration of the notions of metaphysical freedom, physical determinism, Free Will, and many other things that we will discuss in this article. In a way, the re-construction of Free Will by Dennett illustrates the ideas that we try to formalize herein, and squashes a lot of fallacies along the way. We definitely recommend reading this book to people interested in the topic of Free Will. However, further in this article, we try to relate the topic to that of civil liberties, which Dennett doesn't address in this book, except in a quick completely unargued denial.

[12]: Computer scientists may view the ontology as the semantics of a considered system, whereas the model is one among many particular syntaxes that can be used to describe the semantics; or they may view the ontology as what can be inferred by observations using the interface of the system, whereas the model is but a particular implementation of the specified system (or part of it, if it's a partial model), chosen to fit a particular underlying formalism, which helps illustrate a point or simulate the system on a given hardware platform.

Intrinsic properties are those that can be described in terms of the abstract interface, or that otherwise keep their meaning across changes of implementation. Determinism is a feature of particular implementations of the universe, that isn't preserved by even trivial changes of implementation; it isn't intrinsic. Inasmuch as the universe may be described with equations, many different sets of equations may equivalently describe the same universe, some of these sets being deterministic and some being non-deterministic.

The phenomenological point of view consists in considering only intrinsic properties of a system, and rejecting extrinsic properties as irrelevant. A similar point of view serves as the original motivation using Category Theory instead of Set Theory, or Constructive Logic instead of Classical Logic, as paradigms on which to found Mathematics.

[13]: Note that though any extrinsic model necessarily does introduce some amount of arbitrary irrelevant details, having a model at all may still be necessary to reason about the system being modelled. The structure of the model may be easier for a man or a machine to manipulate than the reality they represent. Actually, when we think, we may only manipulate representations. But it is important to distinguish what pertains to the ontology being modelled, and what pertains to the arbitrary choice of a model and to the ontology of representation in general. Arbitrary choice of names in languages, or deterministic nature in physical models, etc., are among the many details that belong to the ontology of their respective kind of representation. Confusion between levels of representation is what Korzybski identified as insanity, a delusion all too natural to the human brain, that can be a source of creative inspiration and artistic delight, or a weakness through which people are cruelly deceived.

Valid truths that can be established by a model about a system cannot depend on these representation-specific details. Any proposition that depends on concepts that cannot be expressed in all complete models of a given system cannot be intrinsic truths of the system, though they may be proven as true in a given model: it is a representation artifact, not an intrinsic property of the represented ontology.

Moreover, intrinsic properties of an ontology that can be established as valid truths thanks to a given representation, can also be established as such with any faithful representation of said ontology. If two representations are based upon incompatible premises, of course they will lead to incompatible sets of conclusions, but these incompatibilities are irrelevant to the modelled system as long as they deal with artefacts of representation itself. The same body of information, if used correctly, will lead to equivalent representations. Hence, if a mathematical model of any kind can help reach a conclusion, the same conclusion can be reached by reasoning ``directly´´, that is, with the representation implicit in our language, without an additional artifact. Note that on the one hand, direct reasoning spares the difficulty of introducing the mathematical model and its specific tools. On the other hand, a mathematical model may be designed to allow effective and clear reasoning and avoid fallacies that may creep in by abuse of informal language. A proper mathematical model may help establish more properties, through shorter proofs, than can other models, in a way that makes it easier (at least to domain specialists) to reach and communicate conclusions. However, what remains is the very same body of information that was used to synthetize a model so that it would faithfully describe a given ontology could just as well have been used to reason through a different representation, or directly.

Finally, the adequacy of some representation to reality is always an implicit assumption in any use of this representation. Yet the question of this adequacy is often swept under the rug, a positive answer to it silently admitted; and relying on this kind of unwarranted assumption is the generic fallacy by which people are led to false conclusions and wrong actions. But whichever representation is used, the validity of conclusions reached depends on the assumption that the initial information is correct. If a representation is based on false premises, it will not correctly model its intended ontology, and be an incorrect basis for taking decisions. Assessing the correctness of a model thus remains an important item, and one must not be distracted by proponents of a thesis diverting one's focus exclusively on the content of their thesis and away from justifications for the validity or invalidity of their premise.

[14]: This doesn't mean that physical models have no relevance at all, or that other models will in any way contradict physical models. Inasmuch as physical laws are indeed laws, and they are as far as we know, there will always be physical models compatible with any observation we make with any other faithful model of reality. But the same laws and models based on these laws are wholly useless and irrelevant to the discussion of a wide variety of important human endeavours.

The statements that I owe you a dinner, that the Mandarin language has four tones and no tenses, that prices rise when demand increases or supply decreases, or that you enjoy listening to Mozart can probably be translated into constraints on the arrangement of a great deal of subatomic particles. But having to reduce such statements to physical models is obviously absurd: not only is such an endeavour out of our reach by the lack of known or conceivable technology to map mental states to atomic arrangements and the prohibitively expensive resources such technology would requires even if it were to exist, but all these efforts would be completely worthless if they were ever actually spent. They would never bring any insight on the fragile dynamic notions upon which such psychological statements reside, and would actually be highly dependent on preexisting such insight to properly encode high-level statements into low-level physical implementations.

Just because they are arguably ``elementary´´ doesn't make the events that physics is concerned with relevant in any meaningful way in the description of important (to us) phenomena that arguably emerge on top of these more elementary events.

[15]: Instead of ``One True´´, mathematicians and after them computer scientists would use the term ``canonical´´, means ``intrinsic, and uniquely distinguished´´. This meaning is an extension of the previous meaning of ``according to the rules´´ as popularized in christian theology while referring to the (supposed) One set of Holy Rules.

Textbook examples are K^n being the canonical n-dimensional vector space and the vector family ((1,0,0,...),(0,1,0,...),(0,0,1,...),...) being its canonical base. In each case, the mathematical construction suggests a natural simplest object having a given intrinsic property, and within the given representation framework; said mathematical object is called ``canonical´´.

[16]: In particular, a programming god would no more transcendent to the universe than programming men are transcendent to their computer software; programmers are superior to their programs, but not transcending the ontology of programming.

We can examine the common analogy between god and a programmer and the related paradigm of Digital Physics where the physical universe is seen as a computer program, and see that these points of views, as valid as they might be, still don't suppose or prove anything about the intrinsic or canonical nature of any particular model of the universe, and even less about whether such a model would be deterministic or not.

Even in the ontology of computation, models may be refactored in many different ways; actually, it is expected that a good compiler should and hopefully will silently refactor any input model, so as to ``optimize´´ things. And the resulting model, refactored or not, definitely will be developed into a different model that fits the underlying machine made to execute the model; which machine itself may be modelled or implemented in many ways. So the programmer's model isn't canonical; it's just ``current´´ — what he is currently using to interact with the machine. In other words, it isn't intrinsic to the program and its meaning; it is intrinsic to the programming and debugging interaction between the programmer and the development machine.

If the programmer ceases to ever modify the program again, then all that matters, the only semantics for the program, is what it does, as used by users, and the model is not any more canonical than any equivalent model, as far as the user is concerned. And if no one ever uses the program at all, then the whole program is irrelevant and may be optimized away.

[17]: The theory of transendent souls is found in established religions from the cults of Ancient Egypt to contemporary Christian churches.

[18]: The theory of coordinated quantum wave collapse as a source of consciousness is expounded in Roger Penrose's The Emperor's New Mind.

[19]: To quote Isaac Asimov: ``To surrender to ignorance and call it God has always been premature, and it remains premature today.´´ Or as David Brooks wrote in The Necessity of Atheism:

To explain the unknown by the known is a logical procedure; to explain the known by the unknown is a form of theological lunacy.

Any explanation of what is based on unknowable transcendent entity is a non-explanation, it is an appeal to irrational faith.

Sometimes, explanations are not needed for some facts to be affirmed. If I say ``I love chocolate´´, or ``I don't like chocolate´´, I make the statement of a fact that doesn't require any verification or explanation — I speak of something I know better than anyone and I would lie at my own dismay. It is lunacy indeed to claim ``I like chocolate because some invisible green fairies tickle my tongue when I eat some´´ or invoking any other kind of transcendent entities to ``explain´´ what is actually not explained but obfuscated.

Things that are not known and that are utterly unknowable, by definition, cannot affect us, do not matter, and are irrelevant. They are outside the ontology of Human Action; they do not really exist; at best they are an artifact of the way we represent our knowledge. And thus it appears that human knowledge is an essential aspect of the world as humans experience it, and that epistemology is thus an essential component of metaphysics.

As a comparison, consider an account of one's taste for chocolate in terms of the underlying physiological mechanisms. One could indeed argue that such an account is limited as an ``explanation´´ the fact, since it doesn't explain ``why´´ the various physiological elements are in place in humans in general, and expressed in particularly pronounced or unpronounced way in a given individual. But it is acceptable for an explanation not to provide an infinitely deep foundation to what is explained. No finite explanation ever can, and a claim to such infinite content is a sure sign of a fallacy. What makes such a physiological explanation relevant is precisely that it expands on the initial phenomenon, relates it to other facts, to other phenomena, and ultimately provides useful relevant information to someone — the chocolate lover who wants to understand the health issues and control her consumption, the bio-engineer who would like to invent substitutes for chocolate, variants of it, or ``cures´´ to taste or distaste of it. The ultimate utility of such a scientific understanding may be limited; it remains a great resource saver in comparison with the irrational worship of chocolate as found in precolumbian societies, and a great source of pleasure in comparison with the plain ignorance or fear of chocolate that it may replace.

But in any case, I do love chocolate, and for all practical purposes, you need no further explanation or justification than that statement to accept this particular taste of mine as being most probably true, because I'm the one who knows best, and I'm interested in telling the truth about it.

[20]: It is as vain to seek to explain freedom by some pixie dust that would be mixed into the fabric of the universe, as it would be to explain that a table is round or square because of a magic ingredient of ``roundness´´ or ``squareness´´ has been mixed into its constituents.

Instead, we must look for freedom as a property of how the basic elements that constitute a man are structured together, and how they relate to the rest of the universe.

[21]: Inasmuch as the forces that control man are indeed susceptible to feedback from the man's interaction with the physical universe, then they are not mystical forces, but observable physical forces indeed, and they are amenable to the same scientific studies as any physical forces. Calling these forces ``soul´´ and calling for the supernatural is but the irrational cop-out of those who refuse to confront their belief to the test of knowable reality. By definition, it is natural or it doesn't exist, and it is irrational and wrong to believe it exists without and against evidence.

[22]: It is always possibly to make mistakes and arguments as to whether some entity constitutes an individual or not, whether two phenomena are the expressions of a same individual or not, or what are the physical limits of an individual. For instance, one may mistake a dummy for a person, get mixed up between twins, be cheated by an impersonator, argue for or against the humanity of an embryo or of a clinically dead person, confuse a prop for a leg, shoot at a shadow, etc.

However, evolution in a social environment has sharpened our common sense for the common cases we face, and we can usually identify with great ease that a live grown up human is an individual or that two distinct humans are different individuals — to the point that the very sentences seem tautological. We also correctly identify the fact that fallen hair, clipped nails and severed limbs are not anymore part of the individual they came from (though surgery can sometimes reattach a limb).

The individuality of other humans is an objective fact, as is non-individuality of inanimate object, or the co-individuality of the many limbs of a given man. We ignore such facts at our own loss, causing our plans and actions to fail; or we acknowledge them at our benefit, causing our plans and actions to succeed.

[23]: Latter day cyberneticians would characterize an individual in terms of an autopoietic system: a system made of interacting processes such that each process is directly or indirectly connected to each other process in some feedback loop, and where this looping interaction ensures the continued maintenance of the overall system. Indeed, if we can separate a system into two disjoint sets of processes with respect to feedback or maintenance, then how can we say that these two sets constitute an individual, rather than separate entities?

[24]: Can you imagine someone who'd start eating, then interrupt their meal and pour paint in their cloth drawers, then stop for a break dance while listening to a newscast, then phone their best friend and speak in tongues in the middle of the conversation? You'd call that person seriously deranged (or, if prompted by my challenge, a dedicated prankster), and wouldn't expect any random person you meet to be crazy like that. The very existence of names for activities like ``having a meal´´, ``break dancing´´, ``phoning´´, etc., all presuppose regularity in behaviour of the individual, and in the existence of a somewhat reliable supply chain and education infrastructure for all the items and skills involved in such activities.

Once again, the very existence of individuals, of regular action, etc. is presupposed by these very discussions and constitutes a praxeological established fact that precedes any experimental test and can be but confirmed by any such test. The language with which we describe individual human action proves by its very possibility that human action exists, possesses individual coherence; and the existence of complex prosperous society around us, of which language is but an aspect, proves that this coherent individual action can be crowned with enough success to create and sustain said society.

[25]: The more sagacious reader may have detected that the previous statement is improper if taken as a definition of metaphysical liberty. Indeed, we're just transforming a ``can or cannot´´ into a ``may or may not´´. This may may quantify over a different space than this can, and may bring a useful change in conceptual change in representation. But it doesn't reduce freedom to a more basic concept that doesn't suppose freedom. We invite same sagacious reader to anticipate how we will eschew the need to solve this issue in next section.

[26]: Although the idea of Epistemic Freedom seems to be implicit in many contexts, I had never seen it explicitly articulated; when prompted to write about complex systems, I wanted to make this idea the main thesis of an essay on liberty in social systems. But no sooner had I started to write the current essay that I stumbled on an article by Eric S. Raymond, ``The Predictability Account Of Free Will´´ [B5], that states this idea plainly and clearly. If no one else published it before, credits should thus go to ESR for this philosophical discovery. Being scooped discouraged me, which explains the long time it is taking me to complete this essay. My bad for not publishing earlier, and my salute to ESR.

I would guess that precursors of this idea most probably have been published; whether the idea of Epistemic Freedom has been fully articulated before ESR, I don't know. If you do, I would be most grateful that you offer either citations for the statement of the idea, or information showing that prominent authors in the field demonstrated lack of awareness of it.

[27]: Indeed, because people are seldom both fully aware of themselves and fully sincere with either each other or themselves, asking for a public or private opinion of what one thinks often yields very different answers than observing what opinions one acts upon.

``Do you believe that salute in the afterlife trumps any and all pleasure in this life? — Sure!´´ will reply the member of most any established religion. But his acts reflect that he doesn't take said precept quite that seriously, though he may be quite steadfast in signalling his membership of the religious group. In more mundane cases, people's choice in professional career, food consumption, marriage, investment, etc., reveal actual preferences often different from what they would openly state. How much is due to deliberate choices vs irrational miscalculations, we don't have to double-guess.