The free will debate has a way of making people feel hopeless. You read about how the brain decides things before you’re consciously aware of deciding, about how your upbringing and genetics and neural wiring constrain every choice you’ll ever make. And you come away thinking: what’s the point of trying?

This framing is wrong. Not because determinism is false, but because it’s asking the wrong question entirely.

The Stoics saw this clearly two thousand years ago. And recent research in psychology and cognitive neuroscience is arriving at the same place from a completely different direction.

The Quick Version

Research published in Frontiers in Psychology confirms what Stoic philosophers called second-order desire theory: true freedom isn’t the absence of causal influence. It’s the capacity to regulate your first-order desires from a higher level of reflection. The Stoics called this the hegemonikon, or ruling faculty. Neuroscientists call it executive self-regulation. Both are pointing at the same thing: a real and trainable capacity that determines whether you’re governed by your impulses or capable of choosing otherwise.

What the Free Will Debate Gets Wrong

The standard free will debate sets up a binary: either you have some mystical, causally unconstrained will that breaks the laws of physics, or you’re a deterministic machine with no genuine freedom.

Most people, when they hear “the brain decides before you do,” implicitly take this as evidence for the second option. Determinism is true, therefore free will is an illusion, therefore you’re off the hook.

The Stoics rejected this framing from the start. Not because they denied causation (they were committed determinists). The Stoics believed in fate (heimarmene) as complete causal order. Nothing happens outside the causal chain. And yet, they were among history’s most demanding advocates for personal responsibility and self-cultivation.

How do you hold both? By locating freedom differently than the modern debate does.

Two Levels of Wanting

The philosopher Harry Frankfurt, in a 1971 paper that became foundational for contemporary philosophy of action, introduced what he called first-order and second-order desires.

A first-order desire is a want: I want a cigarette. I want to check my phone. I want to snap back at this person who just annoyed me.

A second-order desire is a want about your wants: I wish I didn’t want that cigarette. I’d like to be the kind of person who can sit with discomfort without reaching for the phone. I want to respond rather than react.

Frankfurt’s insight was that freedom isn’t about whether your first-order desires are caused (they are, they always are). It’s about whether you have the capacity to evaluate and regulate them from a second-order perspective. A person with this capacity has what Frankfurt called a “free will.” A person who can’t step back from their first-order desires, who is pulled around by them without any higher-level evaluation, lacks it, regardless of their metaphysical situation.

The interesting thing is that Epictetus described the same structure in the Discourses, without Frankfurt’s vocabulary, roughly 1,900 years earlier.

What Epictetus Actually Said About Choice

The standard English translation of Epictetus’s key concept is “things in our power and things not in our power” (the dichotomy of control from the Enchiridion’s opening lines). But the Greek word Epictetus uses, prohairesis, is more specific than “things in our power.” It refers specifically to the faculty of deliberate choice: the capacity to step back from impressions, evaluate them, and decide whether to assent to them.

Prohairesis is, in Frankfurt’s terms, the second-order faculty. It’s not about controlling external events. It’s about whether you have a governing perspective on your own mental responses.

Epictetus was clear that first-order reactions (what the Stoics called propathē, the automatic pre-emotional responses) are not under voluntary control. You can’t decide not to flinch, not to feel a flash of fear or annoyance. What you can cultivate is the capacity to not be governed by those reactions. To notice them, evaluate them, and choose what to do next.

This is why Stoicism isn’t about suppressing feelings. It never was. The Stoics weren’t trying to flatten first-order desire. They were trying to build second-order regulatory capacity strong enough to actually govern behavior.

The dichotomy of control isn’t “control outcomes vs. don’t control outcomes.” It’s “first-order responses aren’t freely chosen; second-order response to those responses is.”

Where Fate Comes In

The Stoic position on fate is subtle and often misread. The shorthand is: the Stoics believed in fate but also believed in responsibility. How?

Their answer was that fate happens through your character. External events are determined. But how those events flow through a person depends on what kind of person they are. Epictetus put it this way: the same storm strikes two ships. One captain panics; the other doesn’t. The storm was fated. The response wasn’t, at least not in the sense that matters.

This is sometimes called “soft determinism” in philosophy: the view that determinism and meaningful freedom are compatible, because what matters for freedom isn’t causal independence but the right kind of causal history. A choice is free if it flows from your reflective character, from who you’ve made yourself, rather than from unexamined impulse.

The Stoics went further: they said character itself is cultivated through practice. You’re not passively your character; you build it through repeated choice. Fate runs through you, but you is partly something you construct.

That construction is the whole project of Stoic philosophy.

What Frontiers in Psychology Found

Research published in Frontiers in Psychology examined second-order desire theory empirically and found that what Frankfurt described philosophically maps directly onto identifiable psychological and neural mechanisms.

The study’s finding: the capacity to regulate first-order desires from a higher-order reflective perspective isn’t just a philosophical concept. It’s a measurable psychological function with distinct neural correlates. People vary in this capacity. It can be trained. And it predicts outcomes that matter: emotional regulation, decision quality under stress, resistance to addictive behavior, what psychologists call “self-concordance” (the degree to which your choices reflect your actual values rather than your immediate impulses).

The research also found that this capacity connects directly to what cognitive psychologists call metacognition: thinking about your own thinking. Metacognitive skill is one of the strongest predictors of adaptive behavior in psychology. People who can observe their own mental processes have a qualitatively different relationship to those processes than people who are simply inside them.

Epictetus called this observational capacity prosoche: vigilant attention to one’s own impressions. The practice of prosoche is essentially metacognitive training in Frankfurt’s sense: building the habit of noticing your first-order responses before deciding whether to act on them.

The CBT Connection

There’s another line here worth tracing.

Cognitive Behavioral Therapy, the most empirically validated psychological treatment framework developed in the 20th century, was built on a similar structure. Aaron Beck and Albert Ellis (CBT’s founders) both acknowledged Stoic influence explicitly. Ellis named Epictetus directly as a precursor to rational emotive behavior therapy.

CBT’s core mechanism: you don’t change feelings directly. You identify the automatic thought (the first-order cognitive-emotional response), evaluate whether it’s accurate, and choose whether to act on it. This is the synkatathesis (assent) structure that Epictetus described. The automatic thought is the phantasia (impression). The evaluation is whether to grant it assent. The behavioral change follows from the evaluation.

The reason CBT has decades of clinical outcome data behind it is the same reason Stoic practice has two thousand years of philosophical endorsement behind it: the mechanism is real. Building second-order regulatory capacity over first-order cognitive-emotional responses produces measurable changes in behavior and well-being.

If you’ve benefited from CBT or something like it, you’ve already experienced the Stoic theory of freedom. You found out that you could be more than the sum of your automatic responses.

The Neuroscience of Second-Order Regulation

The neural architecture underlying this isn’t mysterious. The prefrontal cortex (specifically the dorsolateral and ventromedial regions) is the biological substrate of second-order self-regulation. It’s what allows you to hold a first-order impulse in view without immediately acting on it, to compare it against your values and goals, and to choose based on that comparison rather than the raw impulse.

This is exactly why Stoic practices tend to engage prefrontal systems and reduce limbic reactivity, as the research on Stoic practices shows. But the free will angle adds something the general neuroscience discussion misses: it explains why this matters philosophically, not just practically.

The question isn’t whether your brain is causal. It is. Every mental event has neural correlates; every choice emerges from prior states. The question is what kind of causal process your choices emerge from. Do they emerge from unexamined impulse, from first-order desires running straight to behavior without passing through reflective evaluation? Or do they emerge from second-order evaluation, from the prefrontal systems that can hold your impulses at arm’s length and ask: is this who I want to be?

Both are causal. But they are not the same. And the difference is what matters.

The Stoics said fate runs through your character. Neuroscience says behavior emerges from neural architecture. These are compatible claims, and both point toward the same practical conclusion: the capacity for second-order self-regulation is real, trainable, and worth developing.

A Common Misreading Worth Clearing Up

People sometimes read the dichotomy of control as a permission structure for passivity. “If it’s not in my control, I don’t need to worry about it.” Used this way, Stoicism becomes a retreat: accept everything, change nothing, disengage from outcomes.

This is a misreading. The Stoics were intensely engaged with the world. Marcus Aurelius ran an empire through plagues and wars. Cato fought for the Republic. Epictetus trained students who went on to lead active lives. They weren’t advocating disengagement.

The dichotomy of control is about where to invest evaluative energy, not where to invest action. Act fully on what you can affect. Don’t attach your sense of worth and stability to what you can’t control. The engagement is maximal; the psychological dependency on external outcomes is minimal.

The same structure appears in the Frankfurt framework. Second-order freedom doesn’t mean refusing to act on any first-order desire. It means acting from a position of reflective ownership rather than unreflective compulsion. You still want things, pursue things, engage with the world. You just do so as an agent who has evaluated what’s worth wanting, rather than as someone dragged forward by whatever desire happens to be loudest.

This is why the Stoics connected free will directly to virtue. A person who acts from second-order reflection, whose behavior flows from their values rather than their impulses, is acting as a moral agent. A person who acts from unexamined first-order desire is something less than that, regardless of what their behavior produces. For a deeper comparison of how Marcus Aurelius and Seneca each worked with this, see this piece on their contrasting approaches.

What This Changes About Practice

If second-order regulatory capacity is the mechanism that matters, Stoic practice looks more specific than “be calm” or “accept what you can’t control.”

The practices are specifically training second-order observation.

The pause before assent. When you notice a reactive emotion or impulse beginning, treat the noticing itself as the goal. Epictetus was explicit: the sage doesn’t eliminate propathē; they don’t assent to it automatically. The practice is inserting a gap, however brief, between the first-order response and the decision to act on it. This is Frankfurt’s second-order evaluation in real time.

The values check. Before any significant decision, ask not “what do I want?” but “what do I want to want?” Or more practically: “Is this the response I’d choose if I were operating from my best judgment rather than the present emotion?” This question is the second-order move. It doesn’t require a particular answer. It requires the habit of asking.

Journaling as second-order reflection. The Stoic journaling practice isn’t primarily about tracking what happened. It’s about reviewing your first-order responses to what happened. Which impressions did you assent to reflexively? Which did you evaluate? Where did second-order regulation succeed, and where did first-order impulse win? This kind of review builds the capacity over time.

None of these practices require believing in a metaphysically uncaused will. They require only believing that the difference between acting from reflective evaluation and acting from unexamined impulse is real. Which it clearly is.

The Honest Limits

Second-order desire theory doesn’t resolve every philosophical question about free will. Frankfurt’s framework has been debated extensively since 1971. His original “principle of alternate possibilities” (the claim that moral responsibility requires alternative possible choices) has been challenged by numerous counterexamples. Philosophy of action is not a settled field.

And the neuroscience, while supportive, isn’t a clean proof. Identifying the prefrontal cortex as the substrate of executive self-regulation doesn’t tell you whether that self-regulation constitutes “genuine freedom” in any philosophically loaded sense. The brain research vindicates the mechanism; the metaphysics is still contested.

What’s not contested: the capacity for second-order self-regulation is real, measurable, and trainable. Whether you call it free will, prohairesis, executive function, or metacognition, it’s describing something that exists and that makes a difference to how you live.

That’s enough for the Stoic project to stand on. The metaphysics can wait. The practice is available now.

---

This is one framework among several for thinking about freedom and responsibility. If you’re working through questions about agency and choice in the context of mental health challenges, please consider that philosophy works better alongside professional support than as a substitute for it. For the Frankfurt framework, the original paper “Freedom of the Will and the Concept of a Person” is available through most academic libraries. The Stanford Encyclopedia of Philosophy’s entry on free will is a rigorous and accessible overview of the landscape. For Frontiers in Psychology research on self-regulation and metacognition, their self-regulation section collects the relevant work.