# Thread: Can physical laws emerge from true randomness?

1. ## Can physical laws emerge from true randomness?

Is anything in the universe truly random? Wave collapse in the QM model is modeled mathematically as a random distribution, but why does that randomness conform to a predictable distribution?

Is it possible for physical laws to emerge from true randomness (i.e. randomness without underlying rules that determine statistical distributions)?

2. Originally Posted by Deckard
Is it possible for physical laws to emerge from true randomness (i.e. randomness without underlying rules that determine statistical distributions)?
I'm not sure what you mean by true randomness. I don't think a random variable can exist without having a statistical distribution. I mean, if we say "it can take any value. It doesn't fit a bell curve or any other random distribution shape" then that seems to me as though the variable will therefore take the uniform distribution. I don't see how a random variable could escape having a distribution.

By true randomness, I had thought of the distinction between the idealised concept of randomness, and the artificial randomness that is usually used in computers to simulate randomness. (Though I read a while ago that computers could, for example, use randomness within the temperature of components, to simulate something close to true randomness, as opposed to simulated randomness, which just appears random)

3. Originally Posted by scarydoor
I'm not sure what you mean by true randomness. I don't think a random variable can exist without having a statistical distribution. I mean, if we say "it can take any value. It doesn't fit a bell curve or any other random distribution shape" then that seems to me as though the variable will therefore take the uniform distribution. I don't see how a random variable could escape having a distribution.

By true randomness, I had thought of the distinction between the idealised concept of randomness, and the artificial randomness that is usually used in computers to simulate randomness. (Though I read a while ago that computers could, for example, use randomness within the temperature of components, to simulate something close to true randomness, as opposed to simulated randomness, which just appears random)
I should clarify the terminology I used. You're right that "true randomness" is generally used to distinguish pseudo-randomness derived from a predictable causal chain, and an inherently unpredictable event where the causal chain is broken.

In contrast, I'm using the term to distinguish randomness that conforms to a statistical distribution from randomness that doesn't. So, "true" in the sense that there aren't any rules governing the outcome, even statistically. It may be true that all randomness in the universe fits a definable distribution, but then why is that the case? Where did the rules come from that govern the distribution?

It appears complexity and order can emerge from randomness that follows a distribution pattern. But can it arise from randomness that doesn't fit such a pattern?

4. I don't mean to be pedantic, and I think this is away from your main point. However, if there was such a variable, surely we could then observe that variable, many times, and plot the outcomes. If we did that enough times, a shape would be formed. Whatever the shape is, we would say is an approximation for the true distribution. I'm struggling to think how a variable would not have a distribution. I would suppose the plotting of all the outcomes would need to not ever settle into one fixed shape. I suppose it's possible? I'm tempted to think that even such a thing could then be modelled as some kind of random variable whose distribution has a random element to it, and therefore (I would assume) has some structure to it.

5. Originally Posted by scarydoor
I'm struggling to think how a variable would not have a distribution.
I think this could just be due to the fact that all the random & pseudo-random systems we see follow a distribution, so it's what we're used to. I don't see any inherent property of randomness that requires it to follow a distribution, it's just that in our universe that's how it works.

I'm tempted to think that even such a thing could then be modelled as some kind of random variable whose distribution has a random element to it, and therefore (I would assume) has some structure to it.
I don't think this would fit the definition of "model" since it has no explanatory or predictive power.

6. I don't think you can try to analogize between the quantum world and the world of the various forces which define natural laws. Once you have an atom, you have gravity, nuclear strong and weak force, electromagnetism.

7. @Deckard

To echo scarydoor's point/perspective: I don't know what is meant by "true randomness". I'm not sure if you do either...?

The uniform distribution, for instance, is a description of a distribution with no discernable difference between the outcomes contained therein. But the fact that it contains no discernible probability difference between the outcomes is itself a piece of information, and from this we can derive higher order distributions that tell us things about what happens when we interact with such a variable.

Take a perfectly weighted coin: no difference between heads or tails. On the next flip, we can provide no model that predicts better whether a head or a tail will come up. But we can provide our model/run a simulation based upon the notion that we can't. And although its absolutely useless in predicting the next flip, its spookily accurate in determining the total number or heads or tails that will be arrived at during 100 tosses. And the aggregate outcome of a 100 repetitions of 100 tosses.

So i suppose you could say: No, the uniform distribution is not REALLY random, nor is the coin toss (determinist mechanics aside), because its bounded by some meaningful values. I want something totally unbounded by any meaningful values...

And then we're back in the realm where I don't know what you're talking about...

8. Also, distributions are just our lovely mathematical language to describe such things/patterns/phenomenon. When there are nice reasons for those distributions to be smooth and easily definable they tend to come out in pretty equations of greek letters and a smattering of subscript.

But we can define arbitrary distributions that describe arbitrary systems and distributions of probability: not considering practical considerations like how long it would take to define them and whether we can actually hold them in our heads/do useful math with them...that type of thing...

9. @ACow To illustrate, imagine if the position of waveform collapse didn't conform to a predictable distribution, that is, the collapse could happen anywhere with equal probability.

How would we discern higher order distributions from that kind of behaviour?

10. Originally Posted by Deckard
@ACow To illustrate, imagine if the position of waveform collapse didn't conform to a predictable distribution, that is, the collapse could happen anywhere with equal probability.

How would we discern higher order distributions from that kind of behaviour?
That's a uniform distribution of three variables (X, y, z).

#### Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts
•