Is Randomness Real or Just Complex Predictability?

[wuliheron]

Obviously I can no more prove that something is truly random than I can prove an undetectable pixie sits on my left shoulder. Again, this is a metaphysical issue which is by definition beyond the ability of science to prove one way or another. Nor should it be within the purview of science, in my opinion, which has more productive matters issues to attend to.

However, what science can address is the definition of terms including "random" and "supernatural". These, I assert, only have demonstrable meaning according to their function in a given context. When the context becomes so broad as to include life, the universe, and everything there is no demonstrable difference between the random and supernatural. Both are said to not obey natural law.

 

[DrRocket]

Obviously I can no more prove that something is truly random than I can prove an undetectable pixie sits on my left shoulder. Again, this is a metaphysical issue which is by definition beyond the ability of science to prove one way or another. Nor should it be within the purview of science, in my opinion, which more productive matters issues to attend to.

However, what science can address is the definition of terms including "random" and "supernatural". These, I assert, only have demonstrable meaning according to their function in a given context. When the context becomes so broad as to include life, universe, and everything there is no demonstrable difference between the random and supernatural.

In order for you to prove that sometning is truly random, you would need a solid mathematical definition of what is meant by "random". Lacking that there is no possibility of a proof.

 

[Pythagorean]

In order for you to prove that sometning is truly random, you would need a solid mathematical definition of what is meant by "random". Lacking that there is no possibility of a proof.

In an earlier post, I brought up the definition that refers to a situation in which all events are equally probable, then all events are random (even in the causative perspective) I think this is mathematically rigorous enough as a definition. I also tend to think it's highly unlikely that any real system has a chance of all it's states being equally probable because real systems can't be perfectly isolated from perturbation and entropy.

That's not say we can't well approximate it for our purposes. Dice and coins are the popular examples. If you do an experiment with a coin, and you do a lot of trials, you will find about a 50/50 split. But there's always error. What if you have a habit of always starting the coins head up, and you've flipped a coin so many times your muscles are more likely to put eight flips into the coin than any other number of flips, etc, etc. These irregularities in the statistics are causative factors. True randomness, a lack of cause, would mean the coin truly had no destiny to land heads up or tails up.

In physics, there's one particular behavior I can think of that has no known cause and happens in a statistically consistent way regardless of conditions:

atom decay

people are still arguing over whether it's truly random (has no cause) or not.

 

[jostpuur]

However, what science can address is the definition of terms including "random" and "supernatural".

So you believe that you have a definition for what "random" actually means?

(So rigor definition, that it can be used to deal with these claims about randomness being supernatural.)

 

[jostpuur]

In an earlier post, I brought up the definition that refers to a situation in which all events are equally probable, then all events are random (even in the causative perspective) I think this is mathematically rigorous enough as a definition.

Defining the concept of randomness by using the concept of probability is not the most satisfying strategy for finding a definition

 

[disregardthat]

In an earlier post, I brought up the definition that refers to a situation in which all events are equally probable, then all events are random (even in the causative perspective) I think this is mathematically rigorous enough as a definition. I also tend to think it's highly unlikely that any real system has a chance of all it's states being equally probable because real systems can't be perfectly isolated from perturbation and entropy.

Why does it have to be equally probable? Suppose you are throwing a dice with 5 blue and 1 red side. Is not the outcome (side facing up) random even though it is 5 times more probable that the red side faces up?

Probability in common language is always used when we lack the ability to predict. So if something is 'really' random, does that mean it is impossible to predict, no matter what information you have? I can easily imagine that we can come up with a sort of event for which there are quantum mechanical principles which disallows us to collect the necessary amount of information to predict. But does this mean the event was 'really' random?

It's important to distinguish between 'true' causality, 'true' randomness, and just causality and randomness in models. If a phenomenon was 'really random', but behaved according to certain tendencies, we can have causal models of it (e.g. thermodynamics, given that microscopic movement is 'truly' random). And the other way, if a phenomenon is causal, we can just as well have probabilistic theories of it. Any pseudo-random phenomenon is an example of this.

So models cannot determine whether a phenomenon is 'truly' random or 'truly' causal, and I suggest that these as intrinsic properties does not even make sense. When we speak of a phenomenon, we are not merely labeling observations, we are extracting generality from individual observations. The generality extracted is a way of thinking of the phenomenon, and it is in this way of thinking terms like causal and random really make sense, and these terms are only meaningful in the sense of our ability to predict. So it is meaningless to apply these terms as intrinsic to examples of phenomena in themselves. Does this not become a question of our own ability to think of phenomena? I believe Kant argued that causality is one of our cognitive categories in which we interpret all sensory experience.

EDIT: Oh, look; this is my 777'th post. How random.

 

[russ_watters]

You're presupposing random is real, when the question of this thread is if it's real or not.

Well I'll answer that simply: random is real [sic], as far as we know. It is not a difficult question, scientifically. Much of QM is dependent on probability, behaving exactly like that dice throw.

Regarding random vs supernatural, Evo is right that she doesn't have to back up a counter to a claim that someone else made - they have to back up their claim. However, for expedience: The existence of randomness in nature does not violate physical laws, so there is no need for every random event to be considered supernatural. The proposed contradiction between "random" and "law" does not exist.

And note, repeating a throw of a die is not just a technical impossibility (I'm assuming we're not just dropping it from a height of 1" here...), but it is in fact a physical impossibility. Getting the initial conditions exactly equal every time would be a violation of physical law because the concept of "exact" violates QM.

This isn't a philosophical question, it is a scientific question and it really isn't all that difficult of a question.

 

[russ_watters]

Why does it have to be equally probable? Suppose you are throwing a dice with 5 blue and 1 red side. Is not the outcome (side facing up) random even though it is 5 times more probable that the red side faces up?

Probability in common language is always used when we lack the ability to predict.

Correct. People tend to mistake probability and randomness. That may be part of the motivation for this thread.

The fact that you can roll a die a large number of times and get a 1, 1/6 of the time does not make the roll non-random: you have no ability to predict the outcome of an individual roll greater than 1/6 of the time.

 

[russ_watters]

Some of you may find this site instructional: http://www.random.org/randomness/

It covers basically the entirety of this issue.

 

[TubbaBlubba]

I see random as something fundamentally unpredictable, that with any given a priori knowledge, one can impossibly determine the result (e.g. the position of an electron).

 

[disregardthat]

I see random as something fundamentally unpredictable, that with any given a priori knowledge, one can impossibly determine the result (e.g. the position of an electron).

There are many events which we cannot possibly determine the results of (even in principle), but for which it is entirely possible that is caused by prior events. Furthermore, we have a good ability to predict random events as well to a high accuracy. And even predictions of causal events is only up to a certain degree of accuracy anyway. The 'ability to predict' criterion is not well-defined, and is not a satisfactory criterion to establish true randomness (not to mention the impossibility of establishing that it is true for any event in practice). It is entirely plausible that we will one day find a deterministic model of the electron (and accounts for behavior on the quantum level), but which deals in other terms than today.

As to problems such as predicting of the position of the electron; consider the following analogy: What is the position of a platoon of soldiers? How accurately can we measure it in space? Do you agree that it does not entirely make sense to consider the 'position of a platoon of a battalion of soldiers' as a point in space? But even so, it does make sense to consider the position as a 0-dimensional point on the map (and so also in space) for a military tactician.

The point is that we only speak in terms of our models (e.g. a map) of nature, not of the terms of nature itself (such terms cannot exist). The 'position of the platoon' is of course not a random event, but cannot either be measured to an exact accuracy. So, a questions such as the 'true' position of the electron does not necessarily make any sense as a sort of exact position which only can be measured to a certain degree of accuracy. Hence does the complete prediction of the position make as little sense. (The analogy goes further; as the platoon is advancing it is more spread out, so you can have the position to an even less 'degree of accuracy'.)

In fact, I would argue that no claim whatsoever of nature could be true 'intrinsically' to nature for the same reason, but I won't pursue that here..

 

[DrRocket]

In an earlier post, I brought up the definition that refers to a situation in which all events are equally probable, then all events are random (even in the causative perspective) I think this is mathematically rigorous enough as a definition. not.

Nope.

First, it is not a definition of "random" but rather a definition of "uniformly distributed". Uniform distribution is a potential attribute of a random variable, but says nothing whatever about the definition of random.

There are other distributions besides the uniform distribution and in many situations a uniform distribution is impossible.

Mathematics avoids actually defining the term "random" and a "random variable" is nothing more and nothing less than a measurable function defined on a probability space. In turn a probability space is simply a set with a sigma algebra of subsets and a positive measure that measures the whole space as 1. So, you see therein lies no useful test for "randomness".

Thus you still need a viable definition for "random".

 

[Pythagorean]

This was from wiktionary, which is hardly a reliable source, but it is fundamentally speaking about causality, which is more physical than mathematical:

Having unpredictable outcomes and, in the ideal case, all outcomes equally probable; resulting from such selection; lacking statistical correlation.

Here's what wolfram, a more reliable source, says about random numbers (i.e. random in mathematics):

A random number is a number chosen as if by chance from some specified distribution such that selection of a large set of these numbers reproduces the underlying distribution. Almost always, such numbers are also required to be independent, so that there are no correlations between successive numbers. Computer-generated random numbers are sometimes called pseudorandom numbers, while the term "random" is reserved for the output of unpredictable physical processes. When used without qualification, the word "random" usually means "random with a uniform distribution." Other distributions are of course possible. For example, the Box-Muller transformation allows pairs of uniform random numbers to be transformed to corresponding random numbers having a two-dimensional normal distribution.

So my definition is not completely off-base, but I think the first sentence is even more rigorous a definition.

Why does it have to be equally probable? Suppose you are throwing a dice with 5 blue and 1 red side. Is not the outcome (side facing up) random even though it is 5 times more probable that the red side faces up?

but you're kind of playing games, you're not confront the causality. If it was a truly random system, than each of the six sides would have equal probability of coming up. In that case, you would know exactly why blue is more probable than red (because it's a truly random die and more sides are painted blue than red). The randomness still only exists in the equal distribution of the probability of the faces turning up themselves. The distribution of colors in your system is no longer random (remember? you made blue more probable than red so they're not equally probable), but which face turns up still is.

the fact that you can roll a die a large number of times and get a 1, 1/6 of the time does not make the roll non-random: you have no ability to predict the outcome of an individual roll greater than 1/6 of the time.

I don't think anyone implied that. It depends, of course, what definition of random you're operating under, but the reason dice are non-random is because they're chaotic. A dice roll is classically deterministic, it just has a lot of figures to fiddle with in four dimensional variable space and n dimensional parameter space.

I'm still not sure though, whether your definition of random pertains to unpredictability of lack of causation. When I say that outcomes are equally probable, I mean fundamentally lack causation.

also, from your link:

When discussing single numbers, a random number is one that is drawn from a set of possible values, each of which is equally probable, i.e., a uniform distribution. When discussing a sequence of random numbers, each number drawn must be statistically independent of the others.

 

[disregardthat]

but you're kind of playing games, you're not confront the causality. If it was a truly random system, than each of the six sides would have equal probability of coming up. In that case, you would know exactly why blue is more probable than red (because it's a truly random die and more sides are painted blue than red). The randomness still only exists in the equal distribution of the probability of the faces turning up themselves. The distribution of colors in your system is no longer random (remember? you made blue more probable than red so they're not equally probable), but which face turns up still is.

You must not confuse the information of the system with the information of the results. In my example we still have no information whatsoever what the result will be. The point is that it is a random process. It is still random, even though the distribution is not uniform.

 

[Ivan Seeking]

I understand the reference to supernatural and I think it is applicable. I have made arguments related to this idea before. Part of the problem is the interpretation of the word. As has been mentioned, "supernatural" is often associated with specific concepts like God, ghosts, or magic. But those concepts are really secondary to the definition. We assume that a God would be supernatural, but the word supernatural is not limited to the concept of a God.

Again here are the primary definitions. from several sources.

Of or relating to existence outside the natural world.
Attributed to a power that seems to violate or go beyond natural forces.
http://education.yahoo.com/reference/dictionary/entry/supernatural [American Heritage]

1 : of or relating to an order of existence beyond the visible observable universe; especially : of or relating to God or a god, demigod, spirit, or devil
2 a : departing from what is usual or normal especially so as to appear to transcend the laws of nature b : attributed to an invisible agent (as a ghost or spirit)
http://www.merriam-webster.com/dictionary/supernatural

supernatural adjective /ˌsuː.pəˈnætʃ.ər.əl//-pɚˈnætʃ.ɚ-/ adj
caused by forces that cannot be explained by science
http://dictionary.cambridge.org/dictionary/british/supernatural

1.existing or occurring outside the normal experience or knowledge of man; not explainable by the known forces or laws of nature; specif., of, involving, or attributed to God or a god
http://www.yourdictionary.com/supernatural [Webster New World]

To say that true randomness is supernatural, is only to say that any underlying process eludes description. It is beyond the ablity of science to describe it. It is simply a matter of definition.

By definition, events in a truly random system could not be predicted; they defy description, so truly random systems would qualify as being supernatural.

 

[Pythagorean]

You must not confuse the information of the system with the information of the results. In my example we still have no information whatsoever what the result will be. The point is that it is a random process. It is still random, even though the distribution is not uniform.

So then you're talking about a definition of random that only pertains to your subjective state of knowledge. This is what I would call something "appearing random". I covered this definition already. I'm talking about the causality. You seem to be talking about predictability.

 

[Evo]

the random and supernatural. Both are said to not obey natural law.

Where have you provided proof of this? Things can happen randomly while obeying all laws of nature. I understand if your belief is that nothing is random. But making such a claim needs backing up.

 

[Ivan Seeking]

Since I believe that anything real can ultimately be described by science, I maintain that the word supernatural has no meaning. It is an arbitrary concept used to dismiss concepts subjectively defined not to be real.

 

[Ivan Seeking]

Things can happen randomly while obeying all laws of nature. .

Name one.

 

[disregardthat]

So then you're talking about a definition of random that only pertains to your subjective state of knowledge. This is what I would call something "appearing random". I covered this definition already. I'm talking about the causality. You seem to be talking about predictability.

It must be some counter-intuitive definition of randomness if my suggestion is not an example of a random event. I don't agree with it. It certainly does not just "appear to be random", thus confusing it with such things as pseudo-randomness which also appears to be random. At best it's bad wording.

 

[Pythagorean]

It must be some counter-intuitive definition of randomness if my suggestion is not an example of a random event. I don't agree with it. It certainly does not just "appear to be random", thus confusing it with such things as pseudo-randomness which also appears to be random. At best it's bad wording.

The only thing that's truly random about your system (in terms of causality) would be that it's an inertially symmetric system, so it will land, with 1/6 probability, on any of the six sides.

the fact that somebody painted five sides blue and one side red doesn't change that, or change the fact that this is the underlying source of the randomness (the inertial symmetry of the die).

i.e. if you remove the underlying uniform distribution, the randomness will go away. The colors are irrelevant.

 

[Pythagorean]

supernatural vs. random discussion:

Evo:
Supernatural, I think (I hope) is not to be taken so literally. I can see the comparisons that Ivan and wulheron are drawing. I get it. But I don't think it's a complete match.

Ivan, wulheron:
quantum mechanics has examples of randomness. I wouldn't call them supernatural persay (though many physicists did seem to think it was eerie originally. Not so much today).

Atom decay is another random event (in a sample of decaying matter, any particular atom may spontaneously decay. The spatial probability distribution of which atom decays is uniform.

I wouldn't call this supernatural. It may just be fundamentally random.

 

[Evo]

Name one.

I'm carrying some dishes, one slips through my fingers and falls to the floor.

I'm not talking about systems. I'm talking about random events. Some people think that nothing can happen randomly, that everything that happens is predestined. This is the category, I believe, that wuliheron falls into. To him nothing can be random, therefore random is supernatural to him.

Below is in response to the OP.

Random, generally, can mean one of two things:
1)Unpredictable, from a given point of view.
2)Uncaused, by a previous event.

The first one is easy, random in this sense is just a description based on either a simple lack of knowledge or the impossiblity of having enough knowledge. The former being like predicting what your girlfriend will wear, whereas the latter is like predicting the weather.

The second refers to an actual event that has no preceding cause. Whether this can exist is an open question, and even if they do exist, it would be unlikely that one could distinguish it from something that is simply unpredictable.

 

[Pythagorean]

Evo, I think that example can be classically determined. It's not random, it's chaotic.

 

[wuliheron]

Where have you provided proof of this? Things can happen randomly while obeying all laws of nature. I understand if your belief is that nothing is random. But making such a claim needs backing up.

Something that is truly random, and not merely unpredictable, by definition does not follow any natural laws. To assert that something that does not have any rhyme or reason somehow follows natural law is, therefore, to utter a contradiction.

 

[apeiron]

This isn't a philosophical question, it is a scientific question and it really isn't all that difficult of a question.

Err, no one seems troubled by probablistic issues. It is the causal question that is of interest.

Rephrasing the OP: do uncaused events exist? Can something happen which had no preceding trigger?

Pythagorean suggested a way of making possible sense of this suggestion - imagining a state so pefectly poised, so symmetric, that it could break either way.

This is the old pencil balanced on its tip idea. However, a pencil would still seem to need a vibration, an unmeasured tilt, or some other triggering event to send it in some direction. A truly perfectly balanced pencil in isolation might never tip (unless we invoke QM?).

Another example given was atomistic decay. This is modeled as the probability of jumping a decay threshold - a series of fluctuations, one of which is large enough eventually. A poisson process. So is this a causeless event?

Putting mathematical models to one side, are there any convincing exampes of uncaused events, even with QM?

 

[wuliheron]

So you believe that you have a definition for what "random" actually means?

(So rigor definition, that it can be used to deal with these claims about randomness being supernatural.)

As I already stated, Words only have demonstrable meaning according to their function in a given context. The idea that anyone definition of "random" supersedes all others contradicts this observation. What I am asserting is that because the context is so broad when discussing the truly random (a metaphysical idea) its meaning becomes indistinguishable from the "supernatural".

 

[disregardthat]

The only thing that's truly random about your system (in terms of causality) would be that it's an inertially symmetric system, so it will land, with 1/6 probability, on any of the six sides.

the fact that somebody painted five sides blue and one side red doesn't change that, or change the fact that this is the underlying source of the randomness (the inertial symmetry of the die).

i.e. if you remove the underlying uniform distribution, the randomness will go away. The colors are irrelevant.

So, are you suggesting that behind every random process there is an underlying uniform distribution regardless of our ability to conceive of it?

 

[Pythagorean]

No. It's the most commonly used definition of random in the sciences when no qualification is used, as I quoted from Wolfram. The more general definition was in the first sentence of the quote, as I said when I quoted it.

 

[DrRocket]

So my definition is not completely off-base, but I think the first sentence is even more rigorous a definition.

The Wolfram definition is pretty much useless, and at best circular. It speaks of selecting a number "at random" from some "specified distribution" neatly sidestepping the basic question as to what is meant by "random" and how without such a definition there can be any meaning to a "specified distribution".

You are going n circles. You still lack any useful definition of "randomm". That is not likely to change.