Introduction
The concept of free will divides philosophers.
Determinism states that the future is completely fixed by the past and laws of nature. Compatibilism argues that free will and determinism can coexist.
The consequence argument states that if determinism is true then our actions are consequences of laws of nature and past events; we have no choice about what happened before we were born, and no choice about laws of nature, therefore no choice about the consequences of these things. Compatibilism argues free will does not depend on defying natural laws or changing the past but on acting according to one's desires free from external compulsion.
I argue compatibilism can be defended against the consequence argument. Frankfurt's thought experiments challenge the consequence argument. I show a stronger defence is possible, with free will rooted in human creativity and knowledge creation.
Section I explores Frankfurt's thought experiment, arguing free will does not require the ability to do otherwise. Section II examines Frankfurt's hierarchical theory, focusing on aligning desires. Section III investigates how knowledge creation (Popper and Deutsch) expands possibilities for free will. Section IV presents my levels of explanation argument, revealing why the consequence argument conflates different explanatory levels. These arguments collectively defend compatibilism.
I
Harry Frankfurt argues free will does not require the ability to do otherwise. Frankfurt's thought experiment: a man decides to raise his hand while a hidden device monitors his brain, ready to force the action if he shows any inclination to do otherwise. The man chooses freely, so the device never activates. Frankfurt concludes the man acted freely because his action originated from his reasoning and desires, despite lacking alternatives. What matters for free will is not having multiple options, but having actions flow from your mental processes.
The consequence argument responds that the man's decision was still determined by prior events and natural laws. His intention was shaped by earlier causes outside his control: genetics, upbringing, brain chemistry. If everything about his decision was fixed by the past, how can we say he acted freely? Even if the device never activated, the man's choice was still the inevitable result of forces beyond his control.
However, the consequence argument assumes all forms of causation are equivalent for free will. This is false. There is a difference between actions caused by our beliefs, desires, and reasoning (rational capacities) versus actions caused by external manipulation bypassing our mental processes entirely. When the man raises his hand through deliberation, he exercises capacities that constitute free will. These capacities represent free will even when operating deterministically, because they enable control over our actions. This control matters because it allows us to act on reasons we can reflectively endorse, making us the authors of our actions even if the authoring process is deterministic. Frankfurt shows free will is the ability to act through our rational capacities, not the ability to transcend causation.
II
Frankfurt shows mental processes matter for free will, but which processes constitute it? Frankfurt's hierarchical theory focuses on desire structure rather than causal origins. The theory defines free will as alignment between first-order desires (what we want) and second-order desires (what we want to want). Consider two smokers: one desperately wants to quit but cannot resist cigarettes (conflicting desires), whilst another has reflected and actioned their desire to smoke (aligned desires). For hierarchical theory, only the second smoker exercises free will because their smoking flows from reflectively endorsed desires.
The consequence argument challenges this as second-order desires are themselves determined by past events. If the smoker's endorsement results from deterministic processes like environment and brain chemistry, how does this differ from any other determinism? The alignment between desire levels might simply create an illusion of free will whilst obscuring that all desires, at every level, are determined by factors beyond our control.
The consequence argument treats all determinism as equivalent, overlooking what constitutes free will. The first smoker acts against their judgment, whilst the second acts by their values. Even if both smokers' mental processes are determined, only the second exercises genuine free will through reflective endorsement. What matters is not whether reflective processes are determined, but whether we possess and exercise our capacity for choice.
III
Beyond reflective processes, Karl Popper and David Deutsch argue human creativity introduces novelty through knowledge creation. Popper argues whilst natural laws are fixed, knowledge content is not logically deducible from initial conditions alone. When humans solve problems, we generate explanatory knowledge revealing previously unknown possibilities within natural laws. Deutsch develops this: humans create knowledge that fundamentally changes what's possible. Humans generate new information about manipulating natural laws, creating space for free will within determinism.
The consequence argument argues knowledge creation itself is deterministic. A scientist's discoveries result from genetic makeup, neural architecture, environmental influences. Even creative breakthroughs are complex outputs of deterministic systems. If belief formation and creative processes are mechanistic, how can their products constitute free will rather than complex determinism?
The consequence argument assumes that if a process follows natural laws, there cannot be free will. However, this conflates different explanatory levels. Even if knowledge creation operates deterministically at the physical level, it creates new content at the level of ideas and possibilities where human choice operates. When humans solve problems, we generate options for action that did not exist before. This expansion of possibilities demonstrates free will because it means our futures are not contained in our pasts; we add new content to the universe through creative thought. Even if mechanistic, this process grants choice: we navigate the deterministic labyrinth not by escaping it, but by creating new paths within it, instead of trying to escape natural laws. This knowledge creation operates at a different explanatory level than mere physical causation. The consequence argument illegitimately reduces explanation to the physical level.
IV
Traditional compatibilist responses provide insights but face challenges operating within the consequence argument's framework. The consequence argument conflates two questions: the causal question (are mental processes determined?) and the agency question (do determined mental processes constitute free will?). This creates false assumptions that all determinism is equivalent for free will. There is a distinction between determinism bypassing rational capacities (hypnosis, manipulation) and determinism operating through them (deliberation, reflection, reasoned choice; the right kind of cause). The consequence argument commits a category error demanding free will require being uncaused rather than being the right kind of cause (through deliberation, not bypassing it). When actions flow through beliefs, values, and reasoning, even if determined, we exercise free will because we act through rational capacities.
The consequence argument responds that this distinction is meaningless because both forms of determinism are products of prior causes beyond our control. If our reasoning, beliefs, and values are results of genetics, upbringing, and environmental factors, then appealing to "rational capacities" relocates the problem without solving it. How can determined processes constitute free will? The distinction between "right" and "wrong" forms of determinism appears arbitrary. Why should determinism through reasoning count as freedom whilst determinism through other pathways do not? Both are products of prior causes operating according to natural laws.
The consequence argument demands that free will involve being an uncaused cause, which is a logical impossibility that would violate natural laws and place agents outside the natural world. Free will does not require transcending causation; it requires being the right kind of cause operating through the right kind of processes. The distinction between different forms of determinism is not arbitrary but reflects a difference in the types of processes involved: when we deliberate, reflect, and choose, we exercise our capacity for rational evaluation and creative problem-solving. Consider the difference between a person who commits murder due to a mental dysfunction versus one who does so after deliberation. Both are determined, but only the second involves the capacities that ground responsibility. These processes constitute free will not because they are uncaused, but because they represent the exercise of our capacities. We are agents not despite being part of the natural world, but because of how we participate in it through reason, reflection, and creativity.
Conclusion
Many so-called paradoxes in philosophy vanish when you stop mixing up levels of explanation. "Does free will exist?" This is not a question physics can answer. The laws of motion do not refer to "free will". The confusion happens when you expect an explanation at the level of particles to tell you something about people. Deterministic physics does not contradict free will, because "free will" is not a concept within particle physics. They do not compete, they complement. So, when people say, “free will does not exist”, they have confused free will needing to be a fundamental force with free will not existing at all.