The Inescapable Reality of Necessity
Recently, I’ve come to believe that some version of hard determinism is likely true. Here, I’ve carried together an overview of the subject and why it favours an incompatibilist view on Free Will. Although most people find this view depressing, I have found it oddly reassuring, since it encourages a kind of acceptance of life, the universe and everything, and the belief that everything happens exactly as it must.
I deeply dislike the term “Free Will”. Free Will, in my opinion, is an incoherent concept, and a very loaded word, philosophically, politically, morally, and theologically speaking. Everyone has a strong pre-conception of the term going into any discussion of it, making progress almost impossible. Many abstract concepts like this are nebulous enough to look and behave quite differently in each of our minds, and that hinders investigation because we aren’t always talking about the same things, though we use the same words. Even the careful definitions of analytic philosophy, which seek to solidify the abstract, can only do so much to set in stone the immaterial, and even less to ensure that the minds of others are well-behaved.
As a source, I rely heavily on various entries in the Stanford Encyclopedia of Philosophy, which contains a much more thorough treatment of this subject. I also heavily recommend the YouTube channel of Dr. Sabine Hossenfelder.
Instead of directly discussing Free Will, or how all actions or events are determined by prior actions or events, it is more prudent to say that the state of the world at any given time, plus a given set of natural laws, necessitates the state of the world at any other time. This holds for both the past and future, as the direction of causality need not be from past to future, necessarily.
In other words:
State of the world at time T + Set of physical laws => Set of all states of the world at all times T*.
Note that we make no claim about what the precise set of physical laws actually contains, nor how much we know about them. The only assumption is that the laws themselves exist and that they are consistent and universal. In general, this assumption is safe to make.
The reason why this formulation makes more sense than talking about specific events, or actions, is that events and actions are not well-defined. Events and actions are concepts that carve up the world (and time) in a somewhat arbitrary way; events and actions, just like objects, may have a fuzzy ontology, or will simply break down into their most fundamental parts (like some set of sub-atomic particles, or strings, or whatever may be the case). Thus, we shall begin with this general description of determinism and not one involving objects, choices, will, or knowledge. Another point of note is that this set of physical laws and the state of the world must not be necessarily predictable, but can nonetheless be perfectly deterministic. More on that later.
Those who are squarely in the empiricist camp, following in the footsteps of David Hume might say that any claim about universal physical laws cannot be made. All we have are historical patterns of behaviour of physical systems, which we assume will continue in their regularity. Human beings themselves are part of the physical world and therefore a complete description of physical laws as historical patterns must include our own past behaviour. A complete description of the “laws of nature” therefore, could only be given at the end of time, because it must include everything that has ever occurred. Let’s ignore these objections for now. It seems more plausible to assume that there is a set of physical laws than to suggest otherwise.
The more popular objection to this sort of determinism relates to quantum mechanics and its apparent indeterminacies. “At the quantum level, the universe really is indeterminate!” the cry of the people goes. Not quite.
It is not at all clear that quantum mechanics is fundamentally non-deterministic; it depends on the interpretation of the measurement problem. The problem is that particles can be described by the wave function as existing probabilistically in certain locations. However, the particles in question appear in a concrete location whenever a measurement is made. The wave function is described by the Schrödinger equation, which is perfectly deterministic. The issue is that a measurement is not part of the equation, and is in fact incompatible with it.
Throughout the history of the field, most physicists have professed to hold to the Copenhagen Interpretation of quantum mechanics, which makes the famous indeterminacy claim as related to the Measurement Problem. We are not supposed to ask about where the particle is before measurement, and the assumption is that it exists probabilistically in several positions at once, rather than assigning probabilities to where it might be (an important philosophical distinction). Further, Copenhagen claims that the measurement (or its device) is not governed by the same quantum laws as the thing being measured. If it were, it would also abide by the Schrödinger equation, and there wouldn’t be a problem. But, if the measurement cannot be described by the same physical laws as the system it is measuring, then this interpretation is fundamentally non-reductionist and incomplete.
When a measurement is made, the wave function collapses and the particle is observed to be in a specific place, rather than many different places. The collapse of the wave function is not a well-defined process at all and is somewhat clunky and ad hoc. This has led some, including Einstein, to propose that there must be more information in the system that would explain the position and or velocities of particles, which we are simply ignorant about. These alternatives are known as hidden variable theories, postulating that the fuzziness of particles is merely the result of insufficient knowledge.
Bohmian Mechanics (proposed in the 1950s) can achieve the exact same predictions as quantum mechanics (given the right conditions), without invoking any dubious collapses of the wave function. It stipulates that the particles are concrete things at all times, which are guided by the wave function (the pilot wave), much like a beach ball floating on an ocean, thus making quantum mechanics fully deterministic. The pilot wave here is the hidden variable that Einstein was so keen on.
An oft-cited criticism of Bohmian Mechanics is that Bell’s Theorem supposedly proves the impossibility of hidden variables, by showing that quantum mechanics is irreducibly non-local. This, however, is a mistaken criticism, as Bohmian Mechanics is non-local as well, and J. S. Bell himself became one of the main defenders of this approach.
A third interpretation is that of Many Worlds, in which the wave function never collapses, but we merely find ourselves in one universe or another, in which the particle happens to be in the position we measure. This runs into the same problem as before, namely that the detector is not compatible with the system it is measuring and cannot be derived from it. We are supposed to use the wave function to determine all possibilities in all worlds as to where the particle might be, but when making a measurement, we are supposed to only make it in one of those worlds (the one we are in). But that means we are merely moving the problem of the collapse of the wave function to the detector since we are not interested in all possible detector outcomes. It seems like cheating to consider infinite possible worlds for the particle, but only one world for the detector.
In summary, quantum mechanics is guided by deterministic equations, unless a measurement is made, which we cannot account for in the Schrödinger equation. The interpretations which are supposedly indeterminate are, in the eyes of most physicists and philosophers, either incomplete or flawed. Most physicists appear to agree that the Copenhagen Interpretation cannot be the final answer, as it is extremely unsatisfying, and Many Worlds has many problems as well. Bohmian Mechanics, on the other hand, which preserves determinism, is experiencing a renaissance in the philosophy of physics, if the Stanford Encyclopedia of Philosophy is to be believed, anyway.
I should mention that there are attempts to show that even in classical mechanics there are indeterminacies, but these tend to heavily rely on the types of assumptions that mathematicians like to make, i.e. frictionless materials, infinite accelerations, perfect vacuums, etc. Thus, examples like the “space invader” problem where, in backwards-looking causality, it seems as though an object can appear out of nowhere based on infinite acceleration, seem a bit silly.
Predictability vs Determinism
A common mistake made by people in the Free Will debate is a confusion about systems which are unpredictable with systems which are indeterminate. Chaotic systems, like the Three Body Problem, or various instances of Conway’s Game of Life, for example, are inherently unpredictable, yet perfectly deterministic at the same time. This confusion stems from the conflation of an ontological question with an epistemic one. Non-linear and dynamical systems, such as those found in Chaos theory are deterministic as they follow a unique evolution. Given a complete description of their initial conditions, they are in theory perfectly repeatable. We can think of a seed as somewhat analogous to computer science nerds who deal with random number generators. As long as the seed is fixed, the program is repeatable with the same results, even if those results are not predictable (see Conway’s Game of Life for example).
One of the main features of chaotic systems is that they display a Sensitive Dependence on Initial Conditions (SDIC). The classic example here is that of the butterfly flapping its wings, which causes a hurricane to appear at some future point in time, in some other part of the world. The word “cause” here merely stipulates that the hurricane would not have occurred without the butterfly. This illustrates that tiny changes in a system can be hugely amplified later on.
Some may want to argue that quantum mechanics, once again, comes to the rescue of indeterminacy. Even though it is not clear to what extent, if any, quantum effects influence the macroscopic world, one could argue that SDIC would lead to an amplification of any effect at all, thus violating the unique evolution principle. However, this comes with a few caveats. As mentioned above, it is unclear, philosophically, whether quantum mechanics in fact contains any indeterminacies. Let us assume that it does. Would this then mean that the universe is non-deterministic, or that Free Will is perhaps saved somehow?
Not quite. There is a question of scientific realism here. If we imagine that quantum effects are in fact somewhat indeterminate and that due to SDIC, they would eventually have macroscopic consequences, we must also make a realistic assessment of the timeline and scope of these effects. For example, it may be that the various microscopic quantum effects require more time than the existence of the universe to work their way through the system of amplifications to become significant or noticeable. If that is the case theoretically, we must ask ourselves practically, whether these types of models cohere sufficiently with the world to be true. Thus, we cannot categorically state that the indeterminacies of quantum effects (should there be any) necessarily lead to indeterminacies in the macroscopic universe. This question would have to be addressed on a case-by-case basis.
Regardless of whether quantum mechanics is fundamentally deterministic or not, the notion of Free Will cannot be saved, because it is incoherent. What does it mean, after all, to have Free Will?
Typically, Free Will is said to mean that choices could have been made otherwise; that no matter what we decided to do, doing something else was a real possibility. In the present, it means that any choices we face now or in the future are genuinely real and can change the course of our lives, history, or the world. However, as we’ve discussed above, the state of the universe at any given time t* follows from any other state at any other time. This is incompatible with the idea that different choices could have been made. The unique evolution of the universe contains all of your actions and choices as a matter of simple physics and causation.
Does an indeterminate quantum mechanics come to the rescue here? No. Although some have half-heartedly tried to make a connection between some quantum randomness on the sub-atomic scale, and being able to genuinely choose which outfit to wear today, it is a generally futile pursuit. No defender of Free Will is going to accept that some level of randomness equates to real choices.
Even if we disregard our entire opening about deterministic physics and imagine we inhabit a dualistic universe or one in which, magically, the ‘soul’ is not encumbered by causal laws and can choose freely, the problem arises. After all, why does the ‘soul’ want to make one choice over another? Presumably, it has some prior reason to prefer X over Y, thus repeating the question and returning us to square one. We would have simply exchanged ‘physical laws’ with ‘reasons’, but we are no more free. This is a sentiment often summarized in the saying “you can do what you will, but you can’t will what you will’. At some basic level, we simply want the things we want, whether that is food or water, love, or digital watches. These wants are not directly under our conscious control.
Another set of arguments about Free Will comes from theology, which typically claims that we have the power to choose. But to read this as anything other than a moral parable is highly questionable. For if the religious person tried to actually assert a metaphysics of freedom based on religious doctrine, then he would be instantly thwarted by God’s omniscience and its consequences. For if God can know both past and future, then this is simply another version of determinism with the added layer of divine foreknowledge. If God cannot see the future, however, then the monotheist will still face the above problem of where the ‘Will’ in ‘Free Will’ comes from.
Ultimately, it appears that a universe in which Free Will exists is not subjectively distinguishable from a universe in which it doesn’t, regardless of whether we are materialists or not, because Free Will isn’t a coherent concept.
Morality and Determinism
The first objection to the presence of determinism and the absence of Free Will is that it has catastrophic implications for morality. Specifically, it implies that moral blame and praise are incoherent because they require choice, or the ability of the agent to have done otherwise. I agree that this is a problem and that the efforts of compatibilists are somewhat dubious.
Compatibilists will argue that freedom merely consists of the absence of compulsion. In other words, if an agent can do whatever they please, they are free and therefore the appropriate target of praise or blame, even if determinism is true. This sort of solution essentially splits freedom into two concepts: the type of freedom we recognize in everyday life, and freedom from the laws of causality. Since the latter is impossible, it makes no sense to draw any kind of moral consequence from it, and one must therefore focus on the former. This is rather unsatisfying because it feels like the philosophical version of a shoulder shrug.
Many people are tempted to reject determinism based on this moral quandary, but this is deeply dishonest. If all reason and evidence suggest that determinism is true, then rejecting it on moral grounds is nothing short of self-deception (although I suppose the people doing this don’t really have a choice in the matter).
The inability to have done otherwise may have negative consequences for backward-looking moral judgements, but it may not have much effect on forward-looking moral prescriptions. In other words, our ability to justify praise and blame appears compromised, but our ability to say that such and such an action would be good in the future remains intact. This is especially true when taking a conditional view of morality, in which we try to answer moral questions based on pre-defined conditions.
For example, IF we want outcome X, we ought to do Y. These conditional moral statements may be true or false, but they are not encumbered by determinism at all. Whether or not the world is determined, the truth or falsity of these conditionals remains the same, and it is irrelevant why we want outcome X.
Perhaps a re-formulation of praise and blame is in order. Instead of demanding that people could or should have done otherwise, we can instead interpret praise and blame as adding to the social and cultural knowledge which helps us properly formulate these conditionals.
The Mongols did a lot of bad things, but in a deterministic universe, they had no option to do otherwise. In blaming them, we are really just educating ourselves and others about moral conditionals such as “IF you want a peaceful society, don’t do that” (simplified, obviously). These types of conditionals form the basis of our cultural understanding, which in turn leads to better moral outcomes, because the reasons for acting are, at least in part, determined by this understanding.
The fact that we are social and moral creatures who are determined to do this in the first place, merely suggests that functional accounts of ethics have more sway than perhaps generally accepted. The rules of morality which influence our actions flow in some way from the rules of physics. The rules of physics lead to the evolution of complex life on this planet, which then lead to a successful social species, which has in turn adopted those rules.
We can do nothing but hope that our universe is a benign one; that its unique evolution contains less suffering than other possible universes. In light of such a doom, it is perhaps tempting to suggest that nothing matters. But this is incorrect for the same reasons it would be in a non-determined world. All our choices and actions still matter and still change our lives and our world exactly as before. The only difference is the realization that we could not have done otherwise.