Chrysippus’ cylinder: agency in a material universe

Do we live in a material universe governed by cause and effect? I believe so. Do we, then, have free will? It depends on what you mean by that term. The so-called problem of free will is one that keeps intelligent and well intentioned people arguing in circles forever, and there is of course a huge philosophical literature about it (see, for an introduction, this article from the Internet Encyclopedia of Philosophy). Heck, some misguided scientists even think they can, and indeed have, solved the problem experimentally! (This article by Adina Roskies explains why that’s problematic.) I will not defend the above assertions here, but accept them as given and proceed with what I think is a more interesting discussion. Please note that this post should be of far wider interest than just to people attracted to Stoicism in particular, as the messy issue of “free will” arises for any philosophical position.

The Stoics, I believe, had the right picture in mind, though they did get one major thing wrong (but it was reasonable at the time). So, what I’m going to try in this essay is the following: i) explain Chrysippus’ famous metaphor of the rolling cylinder to introduce the Stoic distinction between internal and external causes; ii) properly use the metaphor to understand why the Stoics (indeed, the ancient Greco-Romans in general) didn’t use the word “freedom” in this context, and were right to do so; iii) address the famous “lazy argument” that was meant to defeat the Stoic position, and that still rears its ugly head in some contemporary discussions; iv) summarize the original Stoic conception of Fate as it relates to universal causality; and v) introduce a “deflated” version of the Stoic doctrine, which I believe is more appropriate for modern Stoics.

I. Chrysippus’ cylinder

Cicero, in De Fato (43), presents Chrysippus’ metaphor of the rolling cylinder in this fashion: “‘In the same way therefore,’ he says, ‘as a person who has pushed a roller forward has given it a beginning of motion, but has not given it the capacity to roll, so a sense-presentation when it impinges on the will, it is true impresses and as it were seals its appearance on the mind, but the act of assent will be in our power, and as we said in the case of the roller, though given a push from without, as to the rest will move by its own force and nature.”

Dorothea Frede (who is my main source for the material that follows) beautifully and clearly sets the stage for a discussion of the cylinder metaphor in her chapter on Stoic Determinism published as part of The Cambridge Companion to the Stoics, a collection that is a must for the serious prokopton. Chrysippus thought that the cosmos are embedded in a universal web of cause-effect, and that nothing happens without a cause — including, of course, human decisions.

The cylinder is meant as a visual aid to conceptually distinguish external from internal causes: the first ones are exerted by the environment in which an object (or a human being) finds itself. The second ones have to do with the inner mechanisms defining a given object (which in the case of humans include our character, dispositions, and judgments). If you push the cylinder on a flat surface, it will roll. But the push — an external cause — is only part of the explanation. The cylinder will roll also because it is a cylinder, as opposed to, say, a cube. Cubes don’t roll, even if you push them. It is in the nature of cylinders to roll when pushed, but it isn’t in the nature of cubes to do so. Similarly, it is in the nature of humans to make judgments about things, but such judgments are not in the nature of rocks, or plants, or most other animals (so far as we know).

Consider, for instance, a case in which I offer a bribe to a policeman. The bribe is an external cause for his potentially corrupted behavior. But that cause will be efficacious with some policemen but not others (because they have different characters and arrive at different judgments) or even with the same policeman in one instance but not another (because either or both of additional external and internal conditions have changed). Crucially, though, under the same exact combination of internal and external circumstances the policeman will (will not) agree to be bribed by me.

II) Not free will, but what is “up to us”

One of the fundamental doctrines of Stoicism is often referred to as the dichotomy of control. As Epictetus puts in the Enchiridion (1.1): “Some things are within our power, while others are not. Within our power are opinion, motivation, desire, aversion, and, in a word, whatever is of our own doing; not within our power are our body, our property, reputation, office, and, in a word, whatever is not of our own doing.”

Epictetus then continues to say that if you understand and internalize the dichotomy, “no one will ever be able to coerce you, no one will hinder you, you’ll find fault with no one, you’ll accuse no one, you’ll do nothing whatever against your will, you’ll have no enemy, and no one will ever harm you because no harm can affect you.”

While this is referred to in modern Stoicism as the dichotomy of control, the standard Stoic phrase is that some things are “up to us” and other things are not “up to us.” But what could this possibly mean, within the framework of a philosophy that accepts the notion of universal causality? (Notice that I’m staying purposely away from using the common term “determinism.” There is a reason for this, and it will become clear by the end.)

The answer lies in the conceptual separation between external and internal causes that Chrysippus made. Mind you, they are all causes, so the distinction is not to be taken as somehow deeply metaphysical. It’s just that some causal mechanisms happen to be internal to the human being, and that we know we can work on them to alter them in a desired direction (by way of education, social pressure, reasoned discourse, or the threat of jail time).

As Frede puts it: “It is easy to see why despite the complexity of the inner processes that lead to human action, the Stoics upheld the tenet that persons always act in the same way under the same circumstances. Given the same impressions and the same inner dispositions, the individual will always consent to those impressions. … Responsibility does not depend on the condition that we are always capable of acting otherwise; responsibility depends on the condition that human beings have it ‘in them’ to make up their own minds on how to act. … [The Stoics, therefore] were concerned with the question of how to attain the right inner makeup that enables a person to comprehend the decrees of reason and to follow them in the right way.”

So for the Stoics there is no contradiction inherent in the triad of {universal causality} <> {not doing things differently under identical conditions} <> {moral responsibility for our actions}. Universal causality holds; and both external causes (our environment) and internal ones (our desires, inclinations, and especially judgments) are all part of that universal causal web; and our actions are “ours” precisely in the sense that they spring from causal mechanisms that are internal to us. If all the above is true, of course we would arrive at the same judgments provided that all conditions (external as well as internal) were the same.

Frede continues: “eleutheria [freedom] originally had political connotations and was not used in the debate of fate until fairly late. … In moral discourse, ‘freedom’ cannot mean the absence of any kind of influence from outside since no such vacuum exists. Nor can freedom mean the absence of any inner conditioning. There are no persons without character. … Perhaps the Greeks were wise to give preference to the term ‘what is up to us’ in the debate of moral responsibility.” In modern psychological parlance the corresponding word is volition, our ability to make willful decisions.

III) The Lazy Argument

But if everything is the result of cause and effect, in a sense things are “fated” to occur, regardless of whether we do something or not, right? For instance, if it’s my destiny to get better after an illness, then it doesn’t matter whether I take proper medicines or not. Que sera sera, so to speak.

This objection was raised early on against the Stoics, and is known as the lazy argument, presumably because it advocates laziness: you don’t actually need to do anything, because things just happen, they are fated to be. The argument also is, as it should be clear, intellectually lazy, and doesn’t really bite into the Stoic worldview at all.

Since everything, including all our actions, are part of the causal web, then you will get better if, and only if, you do take the medicine. Taking the medicine is simply one component of the causal chain that starts with you getting sick and ends with you becoming well again.

One still hears versions of the lazy argument today, usually brought up by supporters of so-called contra-causal free will, i.e., by those people who think that the only way to have free will is for us to somehow be exempt from the laws of nature. That, seems to me, is sheer magical thinking.

IV) Fate, determinism, and all that jazz

This is the bit where I get to tell you where, I think, the Stoics went wrong, if understandably so.

The ancient Stoics often talked of Fate and God, and it is important to understand exactly what they meant by those words. Their metaphysics was one of pantheism, i.e., they believed the God is Nature, meaning that it is embedded, immanent, in it. In a sense we are all parts of the God/Nature thing, not just humans, but all other living creatures, and non living matter as well. God is all there is, it is physical, and its doings are governed, of course, by cause and effect.

There is no sense in praying to such a God, since there is no s/he to pray to, no entity outside of space and time who was responsible for the creation of the universe, no consciousness outside of that of living creatures within the universe itself.

However, as Frede astutely observes, the Stoics were not panpsychists (a misguided, in my opinion, position that is enjoying a puzzling revival of late): “The Stoics were pantheists in the sense that for them the entire world is permeated by the divine pneuma. But this type of pantheism is not to be confused with panpsychism: the divine pneuma is not present everywhere in the same form and does not give consciousness and reason to all things.” Indeed, the Stoics recognized different types of pneuma: one that simply gives coherence and internal property to inanimate things; another that allows plants to sustain themselves; one that confers perception and mobility to animals; and yet another one that allows human beings to exercise reason. They would have been panpsychists only if they thought that one kind of pneuma — capable of allowing some form of thinking — permeated the cosmos, but they clearly didn’t.

What the Stoics did believe in, however, was a kind of Providence, not in the sense of the existence of a divine plan for the universe, but in the more intriguing sense, as Frede puts it, that “the world’s wisdom is immanent in the world itself. … For them the causal network is rational in the sense that there can be no better overall order.”

She continues: “If humans knew more than they do about the causal network of which they are a part, they would understand the rationale for seemingly senseless personal tragedies. Such cosmic optimism may not be to everyone’s taste. But this is what made the Stoic doctrine attractive to generations of adherents. … They clearly saw it as a more plausible theory than the purely mechanistic view offered by the atomists [Epicureans] or than the ‘partial teleology’ of the Peripatetics [who believed in local causation, with no cosmic coordination of what happens] — not to speak of the quietism recommended by the Skeptics who desisted from any attempt to make sense of the world.”

And that, I think, is where the ancient Stoics went wrong. While the idea of wisdom as immanent in a world permeated by the pneuma was both attractive and reasonable two millennia ago, it isn’t now. It simply doesn’t go well at all with the modern scientific image of the universe, an image that I accept as by far the best one on the table, threatened by no credible alternative, at the moment.

So what is a modern Stoic who doesn’t believe in the pneuma and accepts modern science got to do? I’m glad you asked.

V) A deflated view of Fate for the Modern Stoic

To see what should be acceptable to a modern Stoic we have to develop what Frede calls a “deflated” view of Stoic metaphysics. Remember the lazy argument and the idea that you taking the medicine is part of the causal web that makes it possible for you to get well? Here is how Frede comments on what the Stoics were trying to do:

“The Stoics countered [the lazy argument] with the contention that most cases depend on the fulfillment of certain causal conditions so that the outcome is ‘co-fated’ in the divine order. If we strip this justification of its unusual terminology, the explanation at first sounds quite trivial. For it amounts to no more than the claim that there are necessary and sufficient conditions for all causal connections. In that case the Stoic theory of ‘fate’ would seem to reduce to the belief in universal causality: everything that happens in a given context is ‘fated’ in the sense that all conditions will be fulfilled. … What separates the Stoics from modern determinists or causalists would then be only their peculiar habit of calling the causal factors ‘fate’ and to attribute a ‘divine’ nature to it.”

Frede correctly rejects this as the right interpretation of what the ancient Stoics were thinking. But it is a perfectly good (and not at all “trivial”) alternative for modern Stoics, who after all benefit from two millennia of intervening philosophy and science. This, I believe, is both a major, and pretty much inevitable, update to Stoicism, unless you reject modern physics in favor of wishful thinking.

Conclusion: the last essay you’ll ever have to read on “free will”?

Modern discussions of free will are often confused and confusing, even for professional philosophers. As is well known, the three major positions are hard determinism, compatibilism, and contra-causal free will. The first two take for granted that we live in a deterministic universe, and differ in that hard determinists reject the very idea of free will as nonsensical, while compatibilists try to reconcile a reformed version of the concept with their acceptance of determinism. The third position is indefensible by anyone who takes science seriously, and is therefore limited to theologians who need the “free will defense” to get around (unsuccessfully, in my opinion) the so-called problem of evil against the existence of an all powerful, all-knowing, and yet benevolent God.

The Stoics are often classed among modern compatibilists, which is indeed a decent first approximation. But I hope that the above discussion has made it clear that their position was far more subtle and satisfying, even once we abandon the idea of a universe endowed with immanent wisdom.

To begin with, the Stoic emphasis was not on determinism (a modern term) but on the existence of a universal web of cause and effect. The reason the latter is more satisfying is because we actually don’t know whether the universe in which we live is deterministic or characterized by an irreducible randomness in the form of quantum mechanical fluctuations.

(Note: the equations of quantum mechanics are themselves deterministic, but that does not translate into quantum mechanical events themselves being deterministic. They are fundamentally random, meaning that the randomness is intrinsic to the physics, not simply a reflection of human epistemic limits.)

A lot of ink is wasted on discussions of whether the world is deterministic or stochastic, by people who are under the impression that the answer would somehow settle the issue of free will. It won’t. If we live in a deterministic universe, then “free will” is out of the window. If we live in a stochastic universe then at most we get random will, which isn’t at all free will. I wish people would stop framing the debate in those terms, but the qm-word is apparently irresistible, partly because nobody understands quantum mechanics, so it’s easy to use it in order to pull some wool on other people’s (or one’s own) eyes.

The other element that is often invoked in these discussions, and that the Stoics also recognized for the red herring that it is, is foreknowledge. The idea is that if we lack free will then we ought to be able to predict people’s behavior with precision. There are two problems here. First, the Stoics themselves realized that that would require a kind of knowledge that is unattainable by mere human beings, as Frede explains: “In order to know whether a particular action will succeed, nothing short of the knowledge of the entire world order would suffice. Only if we possessed that kind of omniscience could we predict whether a certain action is ultimately ‘fated’ to be successful or not.”

Second, we actually can predict quite a bit of what people will do based on a combination of general knowledge of human behavior and specific acquaintance with individual human beings. For instance, my friends and relatives can make an inordinate amount of accurate predictions about what I will do in a variety of instances. That, however, doesn’t mean that my actions are fated. And if they are, that’s not the source of my friends’ and relatives’ foreknowledge of such actions.

So let me summarize all of the above as clearly as possible in a handy bulleted list to keep in mind for all future discussions you may have about “free will”:

  • The discussion about determinism vs randomness is entirely irrelevant to whether we have free will or not. Drop it.
  • Foreknowledge of specific events neither requires nor confirms the idea of fate.
  • “Free will” is a grossly misleading term (free from what? Causality? Ah!). We should use the more neutral and accurate term “volition” instead. Or, better, talk about what is and is not “up to us,” as the Stoics did.
  • Personal responsibility for our actions, especially those of moral valence, doesn’t come from magic (i.e., suspension of cause-effect), but rather from Chrysippus’ useful, pragmatic distinction between external and internal causes. We act based on our judgments. And those, broadly speaking, are “up to us,” in the sense that we can mold them or alter them by use of reason and practice.

20 thoughts on “Chrysippus’ cylinder: agency in a material universe

  1. Paul Braterman

    The lazy argument is paradoxical. For if it is true, why bother to formulate it?

    To deny that I am really choosing because my brain is physically determined seems to me very like denying that a computer is really calculating because the electron flow through its logic gates is physically determined.

    Which seems to leave some kind of compatibilism as the only sane choice. Good. I don’t want to be free of being me.

    Liked by 2 people

  2. saphsin

    I’ve always seen determinism to be a requirement for free will, the idea that an indeterministic universe would result in free will always struck me as bizarre. I mean imagine a universe in which an agent desires to do, to act a certain way in the world and influence it, only lead to consequences outside of your control & predictability. That seems to be the complete absence of free will if anything.

    Liked by 1 person

  3. Arnis Mazlovskis (@mazlovskis)

    To the specious arbitrariness of judgments: everyone can only judge and decide within the frame of one’s intellectual capacity, personal heredity, education and life experience. Contemporary determinists prefer to attribute these stalwart factors to luck.


  4. viennahavana

    Nice cheat sheet… I often refer to Frede’s essay and Becker’s book to understand Stoic fate. Here is my favorite selection concerning responsibility from Lawrence Becker’s A New Stoicism. (Princeton: Princeton University Press, 1998.) In my opinion, it explains how a Stoic would choose to assent to their own volition, because every human action (deliberative or not) is ultimately moral.

    “…agents are fully responsible for their acts if and only if they (a) are aware of what they are doing; (b) are aware of the causes of their actions; (c) assent to acting in those ways from those causes— that is, are acting in accord with norms they recognize as their own; (d) are aware of the causes of their assent— that is, the causes of their own norms; (e) thereby introduce new causal factors into the determination of their actions through their awareness of the causal conditions that shape it; (f) are aware of this iterative, self-transformative causal process; and (g) assent to that, in the sense that they recognize that this process is normative for them.”

    Liked by 1 person

  5. synred

    The cyl. seems a poor analogy. Nothing is up to it. All cyl. will make the same choice. Roll on a flat surface. Roll w/o push if the surface i it’s sloped. Bounce if the surface is bumpy

    Other wise I find the Stoic view point sensible and practical!



  6. Massimo Post author


    It’s a common misconception of the analogy. The point isn’t that something is up to the cylinder. The point is that the cylinder rolls not just because of the external cause but because of its internal nature. And our internal nature produces judgments.


  7. synred

    The point is that the cylinder rolls not just because of the external cause but because of its internal nature. And our internal nature produces judgments.

    I wouldn’t call a cylinders nature ‘internal’.

    My view is that ‘I’ am some sort of process and part of that process is ‘making decisions’ The process is ‘me’. The decisions are my will or things that are ‘up to me’. I take it this is something like compatibablist and pretty much consistent with stoicism.

    A cylinder is static. No internal processes occur. It has no internal nature.


  8. Massimo Post author


    Again, I think you are focusing on the wrong aspect of the metaphor. The cylinder has its own nature, distinct from, say, that of a cube. That’s what Chrysippus is talking about. Similarly, we have our own nature, distinct from that of, say, a plant. Just like it is in the nature of the cylinder to roll if pushed, but not in that of a cube, so it is in the nature of a human being, but not a plant, to generate judgments about things.

    But yes, it is essentially a compatibilist position. The improvement is that — because of its emphasis on separating internal from external causes — it makes it more obvious what it means when a compatibilist says that our decisions are “ours.”


  9. Marco Iacoboni

    Massimo, all your posts are lovely and this one is especially lovely. I am not convinced, though, that given a set of external factors and internal motives the decisions of a rational agent will always be the same. I agree with the logic of the argument but I think that same logic can support alternative scenarios. Here is what I think. We have rational agent RA1 that is dealing with a set of external factors and needs to make a decision. RA1 is facing a set of twenty potential decisions, DM1, DM2, and so on. The internal motives of RA1 weigh on DM7 has the decision to make, a preferred action compared to the other 19 alternatives. So far so good. That’s what you are describing. It’s also possible, however, that the internal motives of RA1 given the set of external factors weigh equivalently on a smaller set of potential decisions, for instance DM4, DM13, and DM17. Life demands to make an action in a timely manner. What to do? I guess one possibility is that a random action selection process picks one of these three decisions, let’s say DM13. Next time, given the same set of external factors and internal motives, RA1 may actually go for DM17, for no other reason that this decision is randomly selected among the three equivalent potential actions. So, the logic of the argument is the same, but there may not be a fixed relationship between a set of external factors, internal motives, and one specific action or decision.

    Liked by 1 person

  10. John J. Bonnice (@JohnJBonn)

    From an evolutionary POV it seems to me that just as we have evolved a capacity to reason we could have evolved a capacity to make choices. Further, reason or the ability to know the truths of man and the universe would seem a lot more problematic than the ability to choose the position of my arm.
    Also, the throw of the dice is completely determined in the causal web yet I am not going to welcome every outcome. No Amor fati. If the gods, God or the internal Logos is not determining things then the outcome cannot be trusted.


  11. synred

    Indeed cylinders, cubes, people and plants all have different natures (though Dan might object in case of ‘human nature’).

    However, Cylinders and Cubes don’t have any internal causes. There ‘behavior’ on any surface is a product of there external shape. Nothing is ‘up-to’ them.

    The separation between external causes is not so clear in the case of living things either. Life by it nature is continouslh interactomg with the environment — breathing, etc. This is, e.g., why the Schrodinger’s cat makes nonsense; a cat is never an isolated system even in a sealed box[a].

    Still practically speaking the rough distinction between internal and external causes makes some sense for living things. Is a plant that aims it leaves toward the sun making a decision? Everything is fuzzy!

    [a] See ‘Schrodinger’s Cat and the Law’ here
    A fable in which the ‘sealed box’ comes in handy.


  12. Shane Sullivan

    Marco, in that scenario, wouldn’t the the random selection also represent a cause? If it was different every time, then I think that would constitute a (slightly) different set of circumstances.

    Liked by 1 person

  13. Massimo Post author


    I have no problem with your analysis, the scenario you describe may very well turn out to be the case in some instances. But I don’t think it changes the big picture, insofar the Stoic take on agency is concerned.

    Liked by 1 person

  14. marco iacoboni (@marcoiacoboni)


    yes of course, that’s what I meant to say with ‘the logic is the same’


    not so sure it’s different. The way I see it is that in some well defined circumstances there is a process (the random action selection process) that operates equally in all those circumstances and yet the outcome may differ (because of the randomness of picking one of the equivalent decisions/actions)


  15. sobersummer16

    Great post, Massimo. I really enjoyed they way you explained the stoic perspective. I especially found value in your comment regarding why we might find ‘universal web of causality’ more satisfying compared to the modern term determinism- that’s something I have been trying to express myself for a while.

    Being the amateur philosopher/stoic that I am, part of the conclusion leaves me slightly confused:

    ‘We act based on our judgments. And those, broadly speaking, are “up to us,” in the sense that we can mold them or alter them by use of reason and practice.’

    How, in the universe that you described, can we meaningfully be the actors that mold our own judgements? From my understanding it would seem like someone who innately had the stoic/molder inclinations would be motivated to seek this out, but someone with passive inclinations would not.



  16. Massimo Post author


    We can be meaningful actors because our judgments and decisions to act come from causes that are internal to us, hence the metaphor of the cylinder. But of course we are part of the universal web of cause-effect, nothing magical or contracausal about human volition.

    Liked by 1 person

  17. Massimo Post author


    Neither Baggini nor I have ever claimed that science is “merely” a human enterprise. He claims, however, and I agree, that it is an inescapably human enterprise, and thus relying on human judgment, which itself is not reducible to mechanical or inescapable operations of reason.


    Others have already pointed out that your criticism has no bite at all. When we say that astrology is irrational it is clear what we mean: that the practice of astrology is thus characterized. Your hair splitting is really not contributing to the discussion, you can do better.


Comments are closed.