Laplace’s demon is a thought experiment proposed in 1814 by French scholar Pierre-Simon Laplace. It was the first published articulation of determinism in conceptual form. It can be summarized by Laplace here:
We may regard the present state of the universe as the effect of its past and the cause of its future. An intellect which at a certain moment would know all forces that set nature in motion, and all positions of all items of which nature is composed, if this intellect were also vast enough to submit these data to analysis, it would embrace in a single formula the movements of the greatest bodies of the universe and those of the tiniest atom; for such an intellect nothing would be uncertain and the future just like the past would be present before its eyes.
—Pierre Simon Laplace, A Philosophical Essay on Probabilities
It must be noted that the word “demon” was never used by Laplace, rather he used the word “intellect”. The word “demon” was a later added for effect.
And of course with modern-day physics whether the universe is entirely causal or not is sort of “up in the air”. If we were to accept that it is entirely causal, meaning that the universe is deterministic (rather than indeterministic), does Laplace’s demon hold water?
The answer: In some regards yes, in other regards no. Namely, such a demon would be able to (theoretically) ascertain the future based on the variables, but I’d suggest Laplace was mistaken about it being able to ascertain the past. I know this seems counter-intuitive, that if one can ascertain all future events one should be able to ascertain past events the same way. This, however, is not the way causality works.
A cause must output a specific effect, but an effect does not necessarily have to have a specific cause. Read that sentence again because it’s a distinction about causality that can seem confusing at first.
To put it another way, a cause cannot have multiple possible effects, but different causes can lead to the very same effect. Chapter 16 of my book Breaking the Free Will Illusion for the Betterment of Humankind goes over a thought experiment to help you see why the latter is the case:
Picture a deterministic universe where some play-doh (a soft, pliable modeling compound) is sitting on a table. The play-doh is in the shape of a cube. Now picture almost the same universe except the initial conditions of the universe made it so that the same amount of play-doh is shaped like a sphere. Back to the first universe, picture a person grabbing the play-doh shaped like a cube, rounding it into a sphere, and placing it back down on the table. Out of luck the sphere he makes is identical in every way to the one in the second universe scenario (every particle is aligned perfectly). In the second universe scenario, the same person sees the play-doh but since it’s already a sphere he leaves it alone.
In both scenarios, a rogue machine then comes in, points its laser beam, and disintegrates the person into a pile of ashes, in which the ashes fall to the floor identically (atom for atom). A wind blows and both spheres roll off of the table and land in the exact same spot. Everything else in each universe is now aligned exactly, and the balls of play-doh are aligned. Even though there were events that had different causality, one in which someone shaped the play-doh into a ball, and another where it already was a ball, everything from that point on will be identical. There will be no difference in any future outcome if the causal structures now align and the universes are deterministic.
Now let’s play time in reverse. Both balls of play-doh fly up to the table. Gravity was the same for both of them. Both roll back to the spot in which they rolled from. The wind was the same for both of them. The person’s ashes materialize back to the person. The laser beam was the same for both of them. The person’s hands touch the play-doh and shape it into a cube. This cube shaping only happens in the first universe scenario and not the second. The causality is different at this point, yet when we play them both forward they both lead to the same output, and hence the same future output at any point afterward.
Due to this, it’s possible to think of two different past situations that could lead to the same future situation.
A -› X
B -› X
There is an important distinction between A necessarily needing to lead to X, and X necessarily requiring A as it’s cause. Though it is true that if A happens such would necessitate X (as I go over thoroughly in chapter 7 of the book), it is not necessarily true looking at X alone that it was caused by A.
To put this in the perspective of the smaller particle level, let’s assume a funneling setup. In other words, particles output a funnel (output molecules that bond to such a configuration) that other particles, when they go into it, output an exact forward momentum through the end tunnel of the funnel, no matter how they enter the tunnel. We can backward project that particle to the tube before the funnel, but we simply can’t backward project which direction that particle took to get there from that point on. Since the funnel would produce an exact effect no matter the direction the particle took to get to it (and it this thought experiment the exact same state of the universe), Laplace’s demon would simply be stumped if it only had all of the variables after the tunneling effect.
There are other problems with the idea of Laplace’s demon and prediction, including forward prediction, in a deterministic universe. For example, a proposed limit on the computational power of the universe, the irreversibility of thermal dynamics, and so on. I think there may be ways around such problems, and the Laplace’s demon thought experiment is, at the very least, one that is useful to our understanding of how free will is incompatible with determinism (placing indeterminism aside for now – which is equally as incompatible). I just thought I’d bring to light some of the problems with the assessment that backward causal predictability is the same as forward causal predictability. It’s not, for this very reason:
- A cause necessarily implies the presence of a specific effect (see chapter 7)
- An effect does not necessarily imply the presence of a specific cause (see chapter 16)
What people should really take out of the Laplace’s demon thought experiment is that, in a deterministic universe, events lead up to all other events, including our thoughts and decisions. Just because Laplace’s demon cannot see the past based on a complete understanding of the state of the universe, and even if there may or may not be limitations to the demon’s forward prediction, simply doesn’t matter in regards to free will not existing. After all, determinism does not necessarily imply predictability, but it does imply incompatibility with being able to have, of one’s own accord, done otherwise. The “free will” idea that has caused so many problems in the world. The idea that the Laplace’s demon thought experiment helps us understand, if you think about it.
As for any potential non-caused events (given an indeterministic quantum interpretation), such would be a huge stumbling block for Laplace’s demon, and if such an event has any say over our thoughts or actions, for our very own willfulness and coherence.
'Trick Slattery
Latest posts by 'Trick Slattery (see all)
- Semantic Shift Day, August 31 – REMINDER! Mark your calendar! - August 1, 2020
- Debunking Dennettian Diatribe - May 2, 2018
- Compatibilism: A Parable (comic) - December 7, 2017
3 Responses to “Laplace’s Demon and Past Prediction Problems”
Sorry, the comment form is closed at this time.
I think it is impossible to know exactly all the details of what happened in the past. This is why I find it stupid when people argue about whether certain historical events happened. They can argue forever about some things which are unknowable.
But doesn’t it work fine for predicting future?
Hi Rahul. Yes, in an entirely causal universe, Laplace’s Demon *theoretically* could predict every future event.