This topic is going to be a bit more on the “for people familiar with knowledge of some philosophical jargon” level. In other words, this won’t be a typical post that I normally make for larger mass consumption. Perhaps at a later day I’ll create a whittled down version, but this one is needed to point others to who claim ontic probability and use this as a way to contrive free will.
I’ll be using words such as “ontic / ontological” and “epistemic / epistemological”, I’ll be talking a little bit about quantum mechanics, and I’ll be talking about the differences between “epistemic probability” and “ontic probability” and my problem with the latter.
If you can deal with a few jargon words and a post that may be a little more technical than normal, please stick with this, because it comes up more often than I’d suspect in free will debates. What am I addressing? It’s this idea that, according to some quantum interpretations (interpretations of the scientific findings of quantum mechanics, basically the teenie tiny particle level), some people assert that probability is ontological.
Ontic or ontological means that the probability really exists. For those unfamiliar with these words, just think ontic (or the ontological) addresses what “exists” or “is” and epistemic (or the epistemological) addresses what we can or cannot “know”. Ontology is the study of existence, being, what is “real”, etc. Epistemology is the study of knowledge (what is knowledge, how we obtain it, standards, and so on). This article is attaching these words to the word “probability”: ontic probability (or ontological probability) vs epistemic probability (or epistemological probability). I’ll be using “ontic” and “epistemic” for their brevity and a few other reasons.
Note to philosophy buffs: As an analytic logician, I don’t make any Heideggerian distinction between the words “ontic” and “ontological”. Since I will be addressing probability in physics, ontic also seemed a little more appropriate but for the most part no real distinction is being made.
So what is probability? Basically, it means that an event either has a specific percentage chance, or a specific chance is assessed. For example, a flip of a coin may be assessed with a 50% chance of landing on tails and a 50% chance of landing on heads. If we were to say this probability was ontic we are saying that those probabilities aren’t simply due to our lack of knowledge, but that this probability actually exists inherent in the structure of the coin and toss. If we are saying it’s an epistemic probability, we are saying we are just assessing a probability because we don’t know all of the variables that lead the coin to the result.
This gets more complex when we start talking about quantum mechanics and wave functions, but before we do that let’s keep with the coin toss just so we can understand how this works for causal events. We’ll imagine that the universe is entirely causal (every event has a cause), and talk about the two different ways (epistemiv vs. ontic) in light of the coin flip.
For epistemic probability, our knowledge is such that we can only assess a 50/50 probability for the coin toss, but in actuality, the coin must land on heads due to the way the flip happens, the velocity and trajectory of the coin, the weight of the coin, the atmospheric conditions, the gravitational pull, the shape and texture of the coin, the qualities of the surface the coin lands on, and so on. If we could know all of the different variables (causes) of the environment and coin, and all of the different variables surrounding the flip, we could theoretically understand that the coin didn’t really have a 50% chance of landing on heads or tails, but rather a 100% chance of it landing on a specific side (e.g. heads). The probability was epistemic only, meaning in our heads.
For ontic probability, the coin flip is somehow the case that, even given the variables, the coin could officially land on either heads or tails. Not just due to our lack of knowledge of the variables involved, but in actuality, the coin has a 50% chance of landing on heads and a 50% chance of landing on tails. This doesn’t just apply to 50/50 probabilities, for example, it might be that the coin is weighted so it has an 80% chance of landing on heads and a 20% chance of tails. It doesn’t matter. What matters is that this probability, if ontic, exists inherently in it, it’s not just an assessment.
Now right off the bat most people are going to recognize that, for a coin, this is most likely epistemic probability only. They will point out that this example is a classical physics example, but when we get into the itsie bitsie teensie weensie world of quantum particles, this particle behavior is unlike the coin. In other words, the assertion often is that, rather than being merely epistemic probability, that for quantum particles, this is ontic probability. That the behavior of quantum particles are probabilistic in the “such probability exists” sense, rather than the “such probability is assessed because we can’t know the variables, only a probability for the event” sense. In a moment I’ll be addressing quantum behavior.
First, let’s talk about the experiment and terminology of quantum mechanics that lead people to this sort of thinking. The most important experiment is the double slit experiment. To summarize just part of the experiment, particles are shot through two slits that are a certain distance from each other, and most end up on a screen that reflects where the particle ended up. What they notice when the particles build up is what’s called an “interference pattern”, where as there are bands, darker toward the center, and lighter toward the edges. Put simply, this means that the particle is acting like a wave, rather than moving in a straight line from point A to point B (which is called, behaving like a particle for some odd reason):
When we measure the location of the particle before the slit, however, the pattern converts to the more straight line formation, and no interference pattern happens.
Basically, as soon as we measure the location of a particle, it no longer behaves like a wave, but behaves like a “particle”. This has given the behavior of particles what is often referred to as “particle/wave duality” meaning that sometimes a particle will act like a wave and sometimes it will not act like a wave (act like a particle), given whether or not the particle was in some way interacted with. For some interpretations of quantum mechanics, when the particle is measured, it conversion from a wave to a particle behavior is called “wave function collapse”, and for others the wave function doesn’t really collapse.
The “wave function” in quantum mechanics is the function that assesses the probability for the entire system. That is because we cannot assess anything but a probability for the entire system. In other words, when we shoot an individual particle out of the gun, we can’t measure it, as the very act of doing so prevents the interference pattern. To determine where the particle will end up on the screen, the only think we can assess is that it there is a specific probability for each band, based on the wave function – but we have no idea where an individual particle will end up other than assessing a probability.
I don’t want to go too deeply here into the wacky realm of quantum mechanics, as it’s this wave function part that is often thought of as having an ontic probability rather than an epistemic probability. So we can, for now, ignore the detector scenario for now, and just understand that this problem (that we can’t measure where a particle is heading without changing it’s trajectory in the process) is part of the reason why we can’t know where it is heading.
Let’s stick with the understanding that where the particles will end up on the screen are assessed through the wave function. Another term for this is called the “probability wave”. It is this type of term that often leads people to think that the “probability” part is a real (ontic) thing.
This is when we get into what is called quantum interpretations, basically, the different ideas about what the happenings an the quantum scale could mean. There are many different interpretations, and unfortunately, one that is a favorite in physics in the Copenhagen interpretation. This interpretation is an indeterministic interpretation that basically suggests that there is no underlying variable that leads a particle to one specific band. In other words, this doesn’t truly have a cause for it, or at least there is no good reason to assert one.
Per Bell’s theorem, if it is accepted, comes the idea that for there to be a cause that leads the particle to the specific place on the screen, such would have to be non-local, meaning instantaneous action at a distance. Defenders of Copenhagen suggest that if we don’t see a cause for the behavior, and if we have to inject in something like non-locality which has it’s own counter intuitiveness, we shouldn’t be injecting something non-evident in such as a cause.
Then you have non-local hidden variable interpretations such as Bohmian mechanics which explain how particles instantaneously exchange information with other particles through a nonlocal mechanism. And to the more radical end you have a multiple worlds interpretation, which postulates that all possibilities exist, each splitting off into different universes never to be seen by the others.
There are many other interpretations, some deterministic (meaning entirely causal), others indeterministic (meaning some non-caused events), and some agnostic to both of these. If you are uncertain what the words determinism and indeterminism mean, read this article: “Determinism” and “Indeterminism” for the Free Will Debate
Each interpretation also treats the wave function (and it’s collapse or illusion of collapse) a little differently. What’s more important, is that no matter what quantum interpretation is postulated, they all are equally incompatible with ontic probability, and here is why:
The Problem with Ontic Probability
The main problem with ontic probability is that it is incompatible with the only two ways events can happen, causally or acausally (without a cause). It matters not if you are a determinist and think all events must have a cause, or if you are an indeterminist and think some events might not have a cause, both of these scenarios do not allow for ontic probability. To understand why, let’s assess the two types starting with acausality.
The first thing I’m going to address is the idea that something with a probability assessment can be due to there not being a cause (what I call “acausal” events), because I find this very idea very problematic. We need to understand what we mean by saying that the behavior of an existing thing (such as a particle) doesn’t have a cause for it. That a particle could end up in location A with such and such probability, or it could end up in location B with such and such probability. Let’s simplify down those bands on the particle screen to two, just to make this more simple to understand.
If we say that the particles could end up in location A with a 75% chance or location B with a 25% chance, what are we saying? Basically, we are saying that there is some sort of mechanism that is forcing more particles to A than to B. If we are suggesting that there is no forcing factor here, we have a problem as to what accounts for the probability distribution.
There are two problems that will be dealt with separately:
- An acausal event can’t have a probability distribution with a range.
- An acausal event can’t be caused by an existing thing.
1) An acausal event can’t have a probability distribution with a range. If there is no forcing factor for an acausal event (no cause), it will either come about at some time in some location, or never come about. In other words, if there is nothing causing it to come about at a specific time or specific location, then if this comes about it would have no spatial or temporal determinacy that would drive it to be weighted with a percentage chance. It would just be an event that “popped” into existence. As soon as we say that such an event must happen within a range, we are injecting in something causing that range. For an acausal event to have a 75% chance of being in location Y at time B, this implies some sort of variables that funnel this likelihood.
2) An acausal event can’t be caused by an existing thing. This should go without saying, if we are saying that something happens without a cause (variable) then we are saying that the object that is behaving acausally can’t be the cause of this behavior. If the object is the cause, then it’s not acausal.
This leaves us with causal probability. And this is where we run into even worse problems when people think that causality and ontic probability are compatible. The fact of the matter, however, is that they are no more compatible than free will is to causality.
The main problem is that ontic probability creates a cause with self-contradictory variables within it. If, for example, we say that a specific particle result has a 25% chance to be at location X and a 75% chance to be at location Y, the variables that cause this would be such that either they lead the particle result to X or they don’t lead to X (to Y instead). This is the same reason that an event (logically) can’t be both what causes X and what doesn’t cause X (but Y instead). It imposes a self-contradiction. For more information on that read here:
Keep in mind that this 75/25 example is simplified down for the sake of keeping this to only two different end results. In actuality the wave distribution assesses a number of different probabilities for the entirety of all variables, meaning that such an ontic claim is contradictory in a number of different ways. The variables within the particle of a causal event would be such that it would lead to X and the very same variables not lead to X, and lead to Y, and not lead to Y, and lead to Z, and not lead to Z, and so on…all with their own percentages (10% to get to X, 20% to get to Y, 40% to get to Z, and so on). The problem should be obvious, if there are causal variables that lead the result to Y, then those causal variables couldn’t lead to X or Z, in which case, ontologically, Y would be 100% and the others 0%.
Since we can’t know those variables, the probability is epistemic only. We don’t know where the particle will end up, we can only assess a probability of where it will end up. But that is only due to our lack of knowledge of the variables. Otherwise, asserting that this is causal and has ontic probability is no different than asserting a self contradiction (e.g. invisible visible square circles).
In conclusion, if a quantum event is caused, it’s logically impossible for it to have an ontological probability of two or more different directions without holding contradictory variables. If it is acausal, then that has other problems in that the particle itself can’t cause an acausal event and this kind of an event cannot contain a probability distribution other than “at some point in some location” or “never”. Ontic probability, like free will, is incompatible with the two possible ways events can happen.
But what of a third “probabilistic” option between an event being caused and an event being acausal?
Certain people will have you believe that probabilism is a third option. This, however, is not the case. In fact, causal events and acausal events are in opposition. If an event is caused, it is not uncaused. If an event is uncaused, it is not caused. There is no magical middle ground – this is a necessary dichotomy. To say that an event is probabilistic but not caused is to say it’s an uncaused probability (acausal) and to say that an event is probabilistic and not uncaused is to say it’s a caused probability. It’s a logical absurdity to say that something is probabilistic and neither caused nor uncaused.
What does this have to do with the free will debate?
Some people suggest that ontic probability gives a condition that lead to free will. They imagine the fact that if a particle has a true probability for an event, that our mind might work in the same fashion where each option we have might be a viable option, and since the ontic probability is part of the particles in your brain, it can be considered “you” choosing that option. Of course this is in itself illogical, as if you make the decision of option A over B, you couldn’t have chosen B unless there were variables that lead to B instead of A. And if we say there is no variables, we inject in an event without a cause. There is no way around this dichotomy. Ontic probability is just as logically incoherent as free will, no matter how many times we qualify it with the word “quantum”. Quantum mechanics does not sidestep logic as some would have you think, it actually depends on it.
But even if we accept the illogical magic that is ontic probability, this does not really help with free will. There would be no amount of willing that would drive a decision to a 25% probable outcome over a 75% and vice versa. This would just be the same as throwing a weighted quantum dice in which, somehow, the probability distribution is ontic rather than epistemic.
Illogical magic put aside however, there is NO ontic probability.
Latest posts by 'Trick Slattery (see all)
- Compatibilism: A Parable (comic) - December 7, 2017
- A Response to a Blog: “This Atheist Believes in Free Will” - November 30, 2017
- Free Will Exists!! (and Happy S-S-Day) - August 31, 2017