The Principle of Sufficient Reason & Creaturely Free Agency

Note the conjunction in the title of this post. It is meant to convey the double intuition that on the one hand we (and perhaps many other sorts of creatures) are free agents, and on the other that everything that happens must be sufficiently caused – must, i.e., be exhaustively caused, and tied in to all other things that happen with perfect coherence and logical consistency in a seamless ontological web, so that we have for our environment an orderly cosmos, rational and therefore intelligible: the Principle of Sufficient Reason (PSR).

It seems prima facie that the truth of the PSR rules out creaturely free agency. It does not.

Note first that we can, and usually do, and rightly do, mean two quite different things by “freedom.” Quoting the invaluable James Chastek:

Let freedom mean the non-determination of a decision. This means two things which are opposed as imperfection to perfection. A decision before it is made is pretty clearly both undetermined and imperfect relative to the same decision made; but a decision that depends on someone or something else to be made is less perfect than one made independently on one’s own, and independence is another way of not being determined. The first sort of freedom – call it (A) – is “not determined” decision since it is the privation of a fully made decision; the second (B) is a “not determined” decision [in that] it does not depend on anything outside the one deciding.

“Perfect” in this instance means finished, completed.

B indetermination is fairly straightforward, and is the sort we more usually mean to indicate: we are free if our precedents do not exhaust the schedule of our causes and reasons, so that our acts involve some degree of utterly untrammelled innovation, some degree of creativity, of novelty unanticipated in our past.

A indetermination is a bit less intuitively obvious, but a moment’s consideration opens it up: it is impossible to determine the full schedule of causes of an event that has not yet transpired – that is not yet completely actual, that is still, or not yet, in the process of becoming fully determinate, and so as incompletely actual, cannot be completely specifiable. Until an act has eventuated in a completed fact, we can’t say anything positive about it at all, let alone specify its causes and reasons. A thing not yet fully real is not yet fully definite, or therefore fully caused. So, the only sorts of events to which we can properly say that the PSR applies are those that are past: facts fully in act, rather than acts still acting.

OK then: the PSR does not rule out A indetermination. But does it rule out B indetermination? That’s a somewhat more difficult question.

Here I think it helps to discern two different denotations of the term Chastek uses: dependence. It can mean either causal dependence of a mechanical nature, or functional derivation; which is to say, formal relation.

The behavior of the nail under the impact of the hammer is an example of the former. The nail has no choice in the determination of its reaction to the hammer. If our choices are dependent in this sense, we have no agency; we are not actors, and so do nothing; which is to say, that we do not really exist.

That’s a perverse result. If it were true, we could not possibly reach it, and so we would not now be talking about it. So mechanical causation can’t be what we mean by dependence in our conversation about dependence in B determination.

Are things better – more congruent with our experiences of being and acting – if by dependence we mean to denote formal relations? Well, what do we mean by “formal relations”? We mean, relations among ideas or concepts. It is easy to see how relations of concepts can influence our acts; indeed, all human cognition involves awareness and comparison of concepts abstracted from experience; and all human action consists of enaction of propositions: of, i.e., systems of formal relations. Experience per se is of facts that exemplify concepts, so that they have properties. From experience, we pick out certain properties of some things as important or attractive or good or righteous or relevant right now, mutatis mutandis. Thus we choose whether or not to take next a bite of pizza or a swig of beer: we choose the feeling of a given form of act on the basis of our past sensory experience of such acts. Then we act. And to act is to enact some congeries of formal relations in a concrete fact.

Not all the forms and their relations that enter into a given decision derive from the immediate past. I might rather than choosing either a bite of pizza or a swig of beer decide instead to say something about last night’s game, or about Bonald’s latest post, that nobody has yet mentioned this evening. But that game and that post are inputs from a past more distant but otherwise just like their more recent cousins, and so are not sources of genuine novelty in the present moment of decision. Is there such a thing as an ingress to this world order of a genuinely novel relation of forms? Is it possible to say or do anything truly new under the sun, that is more than a recombination of forms already there expressed, and thus implicit in their logical relations?

Yes. For, the forms are an infinite library, of which there are infinitely many relations. Granted that all such relations must be logically coinherent in order for truth to be possible – in order for thought to happen, or for that matter any other sort of fact. All the infinitely many forms, in their infinitely many relations, must be, each and all, mutually implicit. The truths must all consistently cohere, if there is to be any such thing as a true matter of fact. This must all be granted.

But not all the infinitely many forms and their infinitely many relations have yet been expressed in this finite world, nor obviously might they ever be, in any such world, nay even in any number of such finite worlds.[1] So much we can readily see, in that things keep happening, none of which are like any of the other things that have ever happened, or ever will. There is time, i.e. For, there is continued novelty.

There is an infinite library of forms, and an infinite library of their formal relations. Because they cannot all be expressed in any finite set of facts, so no finite set of facts, howsoever great, can specify them all. Thus is it that, no matter how long history may stretch on, it can never specify all the things that the actual constituents of history might eventually instantiate – might someday, i.e., enact. But if no finite cosmic past can suffice to *specify* subsequent events, then neither can any finite cosmic history suffice to *determine* subsequent events.[2]

So, no past can exhaustively determine subsequent events ex ante. Decisions then are B indeterminate. All events are nevertheless exhaustively caused – so that the PSR is never violated – but only ex post. Ex ante, then, decisions are A indeterminate.

Thus both the PSR and our free agency hold true, and do not contradict each other.


[1] Worlds are finite by definition; for, they must be bounded to be definite in the first place, so as each first to achieve coherent being in and as itself, and second to achieve coherent being as just itself, and no other.

[2] This is all implicit in Gödelian Incompleteness. Indeed, I am probably remembering it from JR Lucas’ Freedom of the Will, without realizing that I am doing so; without remembering his arguments on that particular topic. Indeed, it seems to me now that I might already have made this argument myself at the Orthosphere. Hard to tell. Oh well …

20 thoughts on “The Principle of Sufficient Reason & Creaturely Free Agency

  1. Without agency in creatures God cannot have a real deep relationship with them. It’s like having a computer program imitate human personality. But it truth its always an automaton and in the end only running specified scripts.

    If let’s say a robot is programmed to do bad things. And is judged and punished for said bad things. Then that scenario is absurd and against Justice and Righteousness itself.

    Calvinism without accounting for free will ends up causing this absurdity.

    An automaton cannot be held responsible, unlike an agent with true agency.

      • Indeed. Automations are objects. And advanced tools in the end. An object will act like an object does according to the laws of physics of the molecules it is constituted of.

        Putting that object on trial is a total absurdity.

      • *Putting that object on trial is a total absurdity *

        The law of deodand did something similar when a moving object, animate or inanimate caused the death of a human being: “Whatever moves to do the deed Is deodand and forfeited

        It was abolished in 1846, after a coroner’s jury declared a railway locomotive to be deodand.

  2. Schopenhauer’s masterful “On the Fourfold Root of the Principle of Sufficient Reason” and his subsequent masterpiece “The World as Will and Idea” does the best job from a panentheist/pagan point of view to explicate a compatibility between human freedom and a world determined in time and space by the guiding reasons, “principiae occulti”. That freedom is, according to Schopenhauer, the freedom of the unmanifest Will itself, which is what motivates all objectivity, including us in our objectivity (as distinct from our subjectivity). The problem ofthe supposed “free will” as it had been known, for Schopenhauer, is that once the transcendental Will manifests as “the World”, forces, time and space, according to the principle of sufficient reason, ie, it does so according to its own laws, which determine fixedly all causal events.

    Worth reading and thinking about, for Schopenhauer is the greatest non-theist thinker of all time. If only Schopenhauer post-dated René Girard and had read him, he might’ve taken Aquinas more seriously.

    • Schopenhauer is a boss, the first philosopher whom I loved as little more than a boy. But it is decades since I read him, so I can count myself honestly no sort of Schopenhauvian. But I did recently read a respectful attempt to decode his metaphysics, and found myself nowise discommoded. So, yeah, I guess. Will have to go back and review. That might take a few years.

      The key thing to remember is that the world as either idea or idea willed is not yet the actual world, but rather only the potential of an actual world. There are an infinite number of potentials for the world – an infinite number of ideas about how it might be – so no given state of the already actual world can possibly adequate to the specification, or therefore to the cause, of any subsequent state thereof.

      So the will – whether perfectly general or merely particular and parochial – is ever and always, and indeed by definition, radically untrammelled, by any exogenous prior. Notwithstanding that, it is certainly constrained by its own internal requirements of coherence and consistency. Each entity must be first coherently – consistently, and completely – itself, and no other. There is no escape from the Law of Noncontradiction, or (by a straightforward implication) any of the other laws of logic.

      What determines what shall happen is, then, not what has already happened, but rather what might compossibly happen – and among what might compossibly happen, also what might best happen, mutatis mutandis and in aeternum – according to some Schopenhauvian wills (whether particular or general). And only ex post might any mind ascertain what any such will in fact has willed.

      I note here in passing that the Schopenhauvian Will – the general Will, as expressed also in each particular creature (imago dei, & so forth) – is Whitehead’s Creativity, which with the Many and the One constitutes his Category of the Ultimate. Schopenhauer’s Will I take also to be a denotation of the fecund aspect – the Creative Urge (but not yet in itself the demiurge) – of the SupraPersonal Godhead of Dionysius; and also of the Plotinian One, the Platonic Good, and Brahman.

      • Each entity must be first coherently – consistently, and completely – itself, and no other.

        That is true only of elementary particles. They are like the Primes in Number Theory, the building blocks of all the rest.

        Essentially, this is the atomic theory of Democritus.

        There is no escape from the Law of Noncontradiction, or (by a straightforward implication) any of the other laws of logic.

        Unfortunately, as the author of the Tractatus came to realise, they are perfectly empty. Within the system of propositional logic, no proposition or variable carries any semantic content. The moment any proposition or variable takes on semantic content, the problem of application arises, because semantic content takes us outside the system.

        The proper application of the rules cannot itself be summarised in a set of further rules. That is the lesson of that little brain-teaser of Lewis Carol, What the Tortoise said to Achilles

      • Each entity must be first coherently – consistently, and completely – itself, and no other.

        That is true only of elementary particles. They are like the Primes in Number Theory, the building blocks of all the rest.

        Is there anything real that is not completely what it is? I can’t think of any such thing. I cannot even conceive of any such thing.

        Within the system of propositional logic, no proposition or variable carries any semantic content. The moment any proposition or variable takes on semantic content, the problem of application arises, because semantic content takes us outside the system.

        To be sure. But from the undeniable fact that application is problematic – fraught with uncertainty, confusion, ambivalence, and error (as must be the case for partiscient intelligence such as ours) – it does not follow that application is impossible. Indeed, that there is an application problem in the first place indicates that accurate application is possible. If it were not, so that we could never tell whether any actual state of affairs made logical sense (and so could understand nothing, including symbolic logic and the application problem), there could be no *problem* of application. Then “problem of application” would be meaningless; so would “application.”

        There is in fact application, so there can be defective application. How do we tell the difference? Well, one way that application can be known to be defective is that it gives rise always and inevitably, soon or late, to logical inconsistencies – to contradictions and thus incompossibilities in the properties we apprehend in the reals under our consideration. When we detect such inconsistencies, or their merely possible presence, we know we must do more work in parsing the reals we hope to understand.

        From this we may conclude – may induce – that logical contradictions cannot be actually realized; that there cannot be actual incompossibilities in real life. Reals must be logically consistent, throughly and exhaustively – i.e., in every respect whatever – in order to cohere actually. And this is as true for any assemblage of mundane reals as it is for any one of them: they must all be consistent with each other in order for things to hang together coherently (however loosely they do hang).

        (Excursus: this is why the Pragmatic test can work … pragmatically.)

        Which is fortunate, because if it were otherwise, there could be no thought, understanding, plans, coordinate acts, animal life, or worlds.

        The proper application of the rules cannot itself be summarised in a set of further rules.

        Right. Life is difficult, and chancy, for partiscients. But fortunately, we don’t need a set of further rules to tell whether we have applied the rules properly. If we have applied them improperly, or messed up our categoreal sets (so that our premises are nonsensical – are meaningless, strictly speaking, insofar as they are meant to denote actual or possibly actual states of affairs), we’ll end up with a mess of logical contradictions.

  3. This essay prompted some rather random thoughts

    Consider the difference between predictions and expressions of intention.

    “I’m going to be sick” is, normally, a prediction; “I’m going to go for a walk” is, normally, an expression of intention.

    The difference lies in the kinds of reasons we might give to justify them; the reasons for believing something to be true and “reasons for acting” (which requires a good deal of unpicking).

    Again, consider

    “Why did you spill your coffee?” – “I thought I saw a face at the window and it frightened me.”

    Now, here we would certainly speak of the “cause” of the event. Without going into the vexed question of “causality,” I suggest that we would all agree this is one of the places where we do use the word “cause.” However, it is a “cause” of a rather unusual kind: the subject is able to give the “cause” of a thought, or a feeling, or a bodily movement, in the same kind of way as he is able to state the place of his pain or the position of his limbs.

    Again, a man cuts his nephew out of his will. He was angry at the latter’s behaviour (the object of his anger), whilst the “cause” of his anger was reading an account of it on Facebook. (I wonder what Hume would make of that sort of causality.)

    One can imagine a therapist asking “What produced this action, or thought, or feeling on your part? What did you see, or hear, or feel? What ideas or images popped into your mind that led up to this?” It would elicit one kind of explanation, but not a Procrustian bed, to which all accounts of actions must conform.

    • Yes. Causes need not all reduce to one sort in order for the PSR to hold. All that is required for it to operate – for, i.e., reality to be first rational and then intelligible – is that the entire schedule of all the sorts of causes of a given event suffice to render it rational and so intelligible, at least in principle (i.e., at least to Omniscience, if not in practice to our creaturely partiscience). But because we creatures are all partiscient, it behooves us to be efficient in our intellectual operations. Parsimony then is almost irresistibly appealing, so improper reduction is a perennial temptation to all thinkers. Thus, e.g., materialists desperately want to reduce all causes to the efficient sort, while idealists want with equal urgency to reduce all causes to the formal sort.

      Nevertheless I do think that all causes are reasons in the final analysis; and if this is so, then all reasons reciprocally are causes. Not to say that material or efficient causes *just are* formal or final causes, or vice versa, or any other such thing; but rather only that all the sorts of causes of things must be rationally intelligible; which implies that they are each, and all taken together, rational – i.e., orderly, coordinated, coherent, consistent, and so forth – in themselves. Thus it would not do, e.g. – in other words, the PSR would fail – if we were ever able to say truly something like, “I spilled my coffee when I saw Everest was *actually in* the coffee cup.” Not, to be clear, when Everest merely *appeared* in the cup phenomenally (and thus, perhaps, illusorily), but when Everest *actually was* in the cup, *so that* I then accurately saw it there. If Everest were of a sudden actually to be in the cup, the PSR would fail, precisely because it is not logically consistent that Everest should be in a cup. Given the natures and characters of cups and mountains, the actual presence of Everest in a cup is a logically incompossible state of affairs. That incompossibility entails that its actual presence in my cup could not be sufficiently caused.

      So may we see that the PSR is forecondition, not just of thought, not just of action, but of actuality. All things supervene the Lógos.

  4. Pingback: What we’ve learned about Female Agency | Σ Frame

  5. Nothing happens until its final prior causes have played themselves out. Assuming a world of perfectly reliable cause and effect, the final prior cause of a deliberate act is the act of deliberation that precedes it. That deliberation is performed locally, within the brain itself. Given an infinite string of prior causes, the only ones we really care about are those that meaningfully explain why the event happened and which we also might actually do something about. The Big Bang is neither a meaningful nor a relevant cause of any human behavior. If a person orders a dinner in a restaurant, the waiter will bring the order to that person, along with the bill holding them responsible for their deliberate act. The waiter will not bring the bill to the Big Bang.

    • I find myself in sympathy with your gist, but I would enter a terminological quibble. It seems that by “final prior cause” you mean “last – ultimate – in the series of prior causes.” Now, prior causes are either material, efficient, or formal (on Aristotle), whereas the final cause is the state of affairs that an act intends, and tends usually to bring about, should circumstances allow. Thus, the final cause of an acorn is an oak – or, to be more capaciously precise, a world in which an oak lives, which is a result of that acorn, unless a squirrel, a hog or a human eats the acorn before it can sprout and take root.

      I would enter also an ontological quibble. Deliberation is performed locally, to be sure – that is, it is performed by the agent itself, wherever and whenever it may be – but that is not to say necessarily that it proceeds *within the brain.* I would argue that it is more accurate, as giving rise to fewer intractable philosophical difficulties, to say that the brain proceeds within the deliberation. The idea here is that the brain as it changes its configuration is not the source and total of the deliberation, but rather the artifact and thus for us an evidence thereof.

      The waiter will not bring the bill to the Big Bang.

      Absolutely, yes: agreed. To do so would be to vacate every jot of agency from the diner; which would be to vacate the diner, per se, and thus with him the meal, the bill, the restaurant, and indeed the whole cosmic shooting match.

      This consideration relevates a general principle: no effects of act x → no x.

      • Nicely put. I think that the deliberation, which actually sets the intent upon some final object, establishes Aristotle’s “final” cause. It always confused me that he would call the initial cause the final cause. But the truest, and most meaningful cause, is the deliberation that formed the intention (aka “will”) that motivates and directs our subsequent thoughts and actions toward completing that goal. For example, we decide to have dinner at a restaurant. That chosen intent causes us to hop in the car, drive to the restaurant, walk in the door, sit at a table, open the menu, order a dinner, eat the meal, and pay the cashier on the way out. Having completed the intended goal, we move on to what we will do next.

        On the ontological issue, to me it would seem that the mind, a process running upon the neural architecture of the brain, is the source of agency and control. (That which gets to choose what will happen next is exercising regulative control). The body, with its physical needs, bumps certain problems up to the reasoning centers, allowing it to control how the body will go about meeting its physical needs.

        I think it may be safe to say that every event is both an effect of prior causes as well as the cause of new effects. This is just how everything, including free will, works. From my perspective, determinism never actually changes anything. What we will inevitably do is exactly identical to us just being us, doing what we choose to do. And that is not a meaningful constraint. It is basically “what we would have done anyway.”

      • It always confused me that [Aristotle] would call the initial cause the final cause.

        Whitehead parsed this by calling the final cause of each novel occasion of becoming its “initial aim.” Whether or not that aim is ever achieved depends upon the occasion itself – not just in terms of its own success in being what it is intended to be, but then afterwards in confluence, and in some cadged together agreement, with the achievements of its predecessors and contemporaries. If there be no such agreement, then can there be no cosmos. So each thing limits every other: one day tells its tale to another.

        … determinism never actually changes anything. What we will inevitably do is exactly identical to us just being us, doing what we choose to do. And that is not a meaningful constraint. It is basically “what we would have done anyway.”

        That’s a cool insight. The key difference when it comes to philosophical rhetoric is this: determinists insist that because “we” are going to do what “we” are going to do, therefore “we” don’t do it in the first place (but, rather, our causal antecedents do it for us, and we are just along for the ride, mere epiphenomena); whereas the realists retort that because *we* are going to do what *we* are going to do, therefore *we* do it.

      • The thing about philosophical rhetoric is that it’s claims are often based more on figures of speech rather than empirical descriptions. If we’re looking for truth, then we need to preserve the literal meaning of words, because every figurative statement is literally false. Which can lead to us stepping into paradoxes that are difficult to get out of. For example, the notion that we “could not have done otherwise” is a blunder that conflates what we “can” do with what we “will” do. All that determinism may safely assert is that we “would not have done otherwise”.

        To say that our prior causes are the “real” causes of our actions, instead of us, is a false requirement, because none of our prior causes can pass that test. Every prior cause of us also has its own prior causes, and no “real” causes could ever be found.

        And no prior cause of us can participate in our deliberations without first becoming an integral part of who and what we are at the time of our deliberation. They cannot bypass us and start causing things on their own without our knowledge and consent. Only the prior causes that have become us get to decide what we will do. So, it is only ever us actually doing the deciding.

        Except of course when a choice is imposed upon us, by coercion or some other form of undue influence (significant mental illness, manipulation, authoritative command, etc.). These are the only things that our choices need to be free of to qualify as “a freely chosen will” or “free will”.

      • Determinism and free-will are empty categories; they cannot be used to distinguish any sequence of events from any other

      • Free will is one category of causes. It distinguishes deliberate acts from accidents, coerced acts, insane acts, etc.

        But, as you suggest, determinism makes no distinctions between any events. All events are equally causally necessary. So, determinism offers us no useful information. If determinism is used to excuse one thing then it excuses everything. For example, if it excuses the pickpocket who stole your wallet, then it also excuses the judge who cuts off his hand.


Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.