Physics has laws, the Universe might not

post by Shmi (shminux) · 2018-06-09T05:33:29.122Z · LW · GW · 23 comments

Contents

23 comments

Inspired by http://backreaction.blogspot.com/2018/06/physicist-concludes-there-are-no-laws.html, which dissed this article: https://www.quantamagazine.org/there-are-no-laws-of-physics-theres-only-the-landscape-20180604.

Epistemic status: very raw, likely discussed elsewhere, though in different terms, but feels like has a kernel of usefulness in it.

What does it mean for the universe to be governed by physical laws? What does the term physical law mean? It means that someone knowing that law can predict with some accuracy the state of the universe at some point in the future from its state at the time of observation. Actually, a qualifier is in order. Can predict the observed state of the universe at some point in the future from its observed state at the time of observation. So

laws => predictability

This is more than a one-directional implication, however. What does it mean for something to be predictable? Again, it means that, by observing the state of the universe at some point in time the observer can make a reasonably accurate prediction of the observed state of the universe at some point in the future. Notice the qualifier "observed" again. How can an observer make this prediction? They must have a model of the observed universe ("map of the territory") inside, and use this model ("trace the map") to predict the observed state of the universe at some point in the future. This model can be very simple, "Raarg hold rock. Raarg let go. Rock fall", or more complicated, "In absence of other forces all objects accelerate downward at 9.81 meters per second squared", or even more abstract, "The stress-energy tensor is proportional to the spacetime curvature." But it is a model nonetheless.

When is a model promoted to the status of a law? When it is useful for more than a single case. When the prediction can be made repeatedly in similar but slightly different circumstances using the same model.  There is a lot of complexity hiding under the surface of this "simple" statement, but at the end of the day, models are only useful if they can be reused, and thus become patterns, templates for the observers to predict the universe. Thus we have the implication in the other direction:

predictability => laws

Thus the two terms are equivalent, at least in this framework:

predictability <=> laws

Let's restate the definitions, which are admittedly only the first approximation, and may not looks standard:

Predictability: an observer inside the universe can infer the state of the universe at some future point in time, with the accuracy acceptable to the observer.

Physical (or other) laws: reusable models of the universe that are part of the observer and let the universe appear predictable to the observer.

I have been trying really hard to avoid, or at least to minimize, the mind projection fallacy, such as stating that the physical laws are the objective laws of the universe. They might well be, but that would be a next step in modeling the universe, potentially useful, but not minimal.

Returning to the title of this post, "Physics has laws, the Universe might not," what I mean by it is that Physics is one of the sciences that we humans, "observers" call a collection some of those reusable models, and so is in itself an aggregate model. The laws of Physics are the constituent models. On the other hand, the Universe, or "the territory" may or may not have something that Stephen Hawking once phrased as a question: 

"What is it that breathes fire into the equations and makes a universe for them to describe?"

The minimal answer might be that the cause and effect are reversed here: the universe just exists (assuming it does), and is somewhat predictable, and the equations are those physical laws inside the observers' minds.

Now, the above is, of course, another (meta-)model. And models are not very useful if they do not result in better predictions. Or, in the language of this site, the beliefs must pay rent. So, what does this one predict? Well, for example, if the universe has no "internal" laws, according to this model, then one could potentially generate a "toy universe," possibly with some form of predictability but without any preset laws and see if anything that could be called (toy) laws would  "emerge" in this toy universe, and under what conditions. It would be interesting to explore this approach further, but it requires a fair amount of decomposition and analysis to make sense in more than a handwavy way. Here are some questions that come to mind:

• Can one start with a sequence of random numbers as a toy universe, i.e. no order and get somewhere that way, just by finding spurious patterns in the sequence?
• An observer is a part of the universe, how would a sub-sequence of numbers represent an observer?
• A "physical law" in this toy universe would be a part of the observer, or maybe the even the whole observer, that can be considered a reusable template, the way our physical laws are. How might it be represented in this case?
• Can the above conditions be relaxed enough to apply to a toy model, but still be offer useful insights?

Should an experiment like that worked out, it would lend some credence to the above conjecture, that the physical laws are not some inherent property of the Universe, but a human attempt to make sense of it by creating reusable templates inside themselves.

23 comments

Comments sorted by top scores.