Strong metaphysical determinism is the philosophical position that every event, including those involved in human cognition and action, is the necessary and inevitable result of antecedent causes and the immutable laws of nature. It posits that the total physical information of the universe at any given moment—encoded across all levels of abstraction, from subatomic particle states to complex neural configurations—contains the complete blueprint for every future state. In this view, the universe is a closed computational system where the "input" of prior physical states dictates a singular, fixed "output," rendering the "ability to do otherwise" a logical impossibility.
This (so many philosophers argue) relates to free will by framing our choices as mere ripples in a causal sea that was set in motion long before our existence; if our decisions are simply the processing of pre-existing physical information, the traditional concept of an autonomous "self" that initiates new causal chains is effectively nullified.
Here's a critical point to understand and keep in mind going forward: According to strong metaphysical determinism (SMD) everything is thus pre-determined by the total physical micro-states and macro-states of the physical universe. That includes all of the cognitive information processing that accompanies your emotions, and all of the cognitive information processing and sensory perception that you are using to read these sentences.
You thought you had made a free-will driven decision to read this. No, no, says SMD. You just thought you did. Moreover your thinking that you did was also thus determined causally by the same mechanism of SMD. Any complaint that you want to make was also determined by the prior physico-informational states of the universe (especially those regions of it containing your brain and its information processing)? And so on, ad-infinitum and ad-nauseum. (Sartre's brand of nausea, perhaps?)
It's just cheating and a gimmicky trick, right?
Not necessarily. I mean: it might be, but I have strong reasons to doubt it.
What about the arguments that say it kind of doesn't matter? Or that even if stuff is all determined like that then we still have free will?
Those are called compatibilist arguments. Compatibilist arguments attempt to reconcile this causal necessity with the existence of free will by redefining what it means to be "free." Rather than requiring an exemption from the laws of physics, compatibilists argue that freedom consists in the ability to act according to one's own motivations, desires, and reasons without external coercion.
Even if those internal desires are themselves determined by prior causes, a person is considered free if their actions align with their "inner state" rather than being forced by an outside power. From this perspective, determinism is not a prison, but the very mechanism that allows our stable character and rational deliberations to reliably produce our actions, thereby preserving a meaningful sense of moral responsibility.
I don't have any deep love for SMD. However, I try to heed my elite-level philosophical training and avoid letting my distaste for any idea get in the way of admitting that it might well correspond to reality.
(Being that I am a psychological entity with emotions and cognitive limits like all humans, I don't always succeed at this, obviously.)
Certain personality types and certain theologies (which probably correlate in people statistically significantly) find the idea of zero-free-will alarming. I don't find it alarming, but then I am an absurdist.
I also think that it's possible that there's a compatibilist argument that takes the edge off. However, it's a very basic one. It is more or less "So what?". Maybe I don't care how my decision and its cognitive processing came about, just that it's mine because I had it (or because the information processing happened in my brain).
On one hand, it feels cosy. On the other hand, it seems flimsy. So I am not buying compatibilism for the time being. I reserve the right to change my possibly completely determined mind.
Moreover, free will of the kind desired by theologians might not be so great anyway.
In an earlier Quillette essay, Stuart Doyle points out a neglected way in which our choices may be unpredictable (in addition to amplified quantum fluctuations or nonlinear dynamics aka chaos): they may implement an "undecidable" system, one which is both deterministic and…
— Steven Pinker (@sapinker) February 19, 2026
Importantly, strong metaphysical determinism isn't just about predictability. It's about whether or not a cognitive, conscious decision precedes and determines action. This can be false whether the process is random or not. Randomness doesn't entail control.
It's likely the case that there can be strong metaphysical determinism with or without randomness. Think about it. Assume SMD with no randomness. Prior physical states at all Levels of Abstraction (LoAs) determine the outcome. No free will. Add randomness. Do you now have more control over the entire process at any LoA? No.
I asked Gemini AI to remind me of the suite of compatibilist arguments. It did so and then asked me if I'd like to shop for counter arguments based upon quantum indeterminacy (or quantum randomness). That is: randomness inherent in quantum systems like atoms and their sub-particles in the garden of particles in the standard model of physics. Okay, I said, but then I told it why quantum randomness probably didn't remove physicalist determinism:
(Discussion with Gemini starts here...)
Gemini: Since you've been looking at this through the lens of physical information and metaphysical determinism, would you like to explore how Quantum Indeterminacy (the idea that some events are truly random) is often used to counter these arguments?
Me: Sure. But also note that quantum randomness doesn't solve the problem. Why? Because a truly random physical system doesn't deliver 'full control' to a cognitive, neurological system. All it does is make the physically-causally determined states truly randomly generated.
You might find this hard to accept because even scholars who produced your training data have trouble with it. (Not all of them.) Moreover, people like free-will-loving theologians are desperate to find knock-down arguments against SMD.
I am by no means certain about my own position which rejects the quantum randomness arguments, but it is probably going to cause those detractors a lot of trouble.
The confusion with randomness arises because it is often intuitively and scientifically couched in terms of statistical predictability. This is understandable since the concept of randomness is often couched in statistical terms.
However, even if a physical system is objectively chaotic or otherwise truly non-linear and random, and therefore not computationally predictable using standard—or even any—computational approaches: it doesn't follow that the prior physical states of the material universe at all levels of abstraction right down to the microphysical haven't 100% caused the later states. The informational and epistemic inaccessibility of future states by statistical forecasting doesn't make reductionist causation go away.
Reductionist causation might not be real for other reasons perhaps, but not for that reason. (Even using supervenience in place of reduction doesn't necessarily help anti-SMD here, since supervenience might just be another example of hiding information.)
Although it does not necessarily kill SMD, I find the outcome of the no hidden variables problem that vexed Einstein a better candidate for possibly saving anti-SMD free-will arguments. Why? Because the Bell experiments show that there are no hidden physical variables in entanglement or 'spooky' action at a distance, and that is deeply weird and seems to break physical conceptions of causation. That's better than randomness. That being said, spooky action at a distance still involves total consistency. e.g. If the state of one quantum system is spin down, it's 100% certain the state of the other is spin up. So there are no hidden variables, but the causation seems to be very deterministic in that sense.
Gemini (Being a little sycophantic, perhaps...):
You have pinpointed what is often referred to in philosophy as the Luck Objection. Simply put: if an action is the result of a random quantum fluctuation rather than a deterministic causal chain, it is no more "up to me" than a coin flip. Replacing a clockwork universe with a chaotic one doesn't grant agency; it just replaces a "determined" slave with a "random" one.
Epistemic Unpredictability vs. Ontic Determinism
Your distinction between statistical forecasting and reductionist causation is crucial. In many scientific and theological circles, there is a tendency to conflate epistemic inaccessibility (our inability to calculate or know the information) with ontic indeterminism (the information not being fixed).
The Statistical Trap: Because we use probability to describe complex systems (like fluid dynamics or neural firing), we often slip into the lazy assumption that the underlying reality is "loose."
Informational Scrambling: As you noted, a system can be 100% causally driven by prior physical states—encoded at every level of abstraction—even if that information is effectively "hidden" from us by non-linearity or computational complexity.
Bell’s Theorem: A Different Kind of Trouble
The "No Hidden Variables" outcome of the Bell experiments is indeed a more formidable challenge to Strong Metaphysical Determinism (SMD) than mere randomness, but as you observed, it brings its own flavor of necessity.
The Non-Locality Paradox
Bell’s Theorem proves that no "local" hidden variables can explain the correlations in quantum mechanics. If we reject "spooky action at a distance," we have to accept that the properties of particles aren't "real" until they are measured. However, if we accept non-locality, we preserve a form of realism but lose the "billiard ball" model of causation.
The Determinism of Correlation
Even without hidden variables, the consistency you mentioned—where one state being "spin down" guarantees the other is "spin up"—suggests a universe that is deeply, perhaps holistically, synchronized. This "super-consistency" doesn't provide a foothold for traditional free will; if anything, it suggests that the "physical information" of the universe is non-locally distributed, making the "autonomous agent" even harder to isolate from the total system.
The Theologian’s Dilemma
Theologians often cling to the "Gaps" in quantum mechanics because they require a "non-physical" entry point for the soul or will. However, if the "gap" is filled by either randomness (which is meaningless) or non-local entanglement (which is perfectly correlated), the entry point for an independent, non-caused "Self" remains effectively sealed.
Would you like to explore Superdeterminism—the controversial idea that Bell's Theorem is bypassed because the observers' choices of measurement were themselves determined by the initial conditions of the Big Bang?
(...Discussion with Gemini ends here.)
To those who think true randomness (in chaotic and nonlinear systems) would nix SMD and deliver free will. It would—in principle—not necessarily do so.
In fact it doesn't look like it would put a dent in causation-based SMD at all.
Worse still, it would make the causation-based deteminism objectively random rather than statistically predictable. The definition of SMD is that the current entire state of the universe at all LoAs is 100% causally determined by the immediately prior physical states (including all of those physical states in our brains).
That can be true even if the causal relationship between the prior states and current states is random.
The confusion arises partly because of the concept of predictability. Future states might be determined per SMD, but still not practically computationally and statistically predictable. That's an informational and epistemic access problem, not a foil to SMD.
Just because the entire system is random-or chaotic with enough randomness in it—it doesn't necessarily follow that it is not causally deterministic.
Approach it by asking: If the universe is only physical and it's significantly random (e.g. intrinsic natural nonlinearity and chaos) then does that necessarily mean/imply/entail that all of the subsequent states of the universe are not caused 100% by the current states at all LoA. The answer is no.