How Random Is Random? Limits & Predetermination

[arunbg]

Is there any limit to randomness?
Can anything in the universe be completely random at all levels of observation ?

I feel otherwise. But doesn't that also lead to the conclusion that everything we do is definite and predetermined in some way (including the fact that I am posting this thread and that you are reading it)?

 

[zoobyshoe]

Is there any limit to randomness?
Can anything in the universe be completely random at all levels of observation ?

I feel otherwise. But doesn't that also lead to the conclusion that everything we do is definite and predetermined in some way (including the fact that I am posting this thread and that you are reading it)?

"Random" is really a human perception based on human values.

 

[Danger]

Can anything in the universe be completely random at all levels of observation ?

Female logic...




<runs and hides from the Sisterhood>

 

[brewnog]

Arunbg, I found myself asking this question for an A level physics project some moons ago.

While I can't give you an answer (mainly because there are far too many people here who know far more about it than myself), you want to be thinking about issues with deterministic, but not necessarily predictable effects. The roll of a dice has a number of well understood physical interactions (distance thrown from, spin, air resistance, initial orientation, interactions with whatever surface it hits) which mean that theoretically, its outcome is predictable. But such tiny differences in any of these can affect the outcome in such a significant manner than to be able to model the system to get accurate predictions would be an extremely tall order. The same goes for why every game of snooker is different, and why we can't really predict the weather. Read up on chaos theory!

 

[pi-r8]

But Brewnog, for chaotic systems like a dice roll, it's clearly not random- it's just so complicated that we can't predict what will happen. Still, if we had a fast enough computer and an accurate measurement of the initial conditions, we could in principle determine exactly how the dice roll would land, or what the weather would be like 5 years from now.

For true randomness, there would have to be literally no pattern whatsoever. No matter how fast your computer, no matter how good your data, you would be unable to make any predictions at all.

For that criteria, I think human consciousness is the best bet. You could keep track of everything a person has done their whole life, and use an infinitely fast computer, and still not be able to predict (perfectly) what they would do in a situation. Free will makes it hard to find patterns.

 

[Anttech]

Is your question one of perception of randomness, or the mathematics behind randomness (sp?)

If I was to write a progam in C for example that produced a random number after you click a button. The result for you would be random, since you cannot predict the output because you cannot see the source... Off course you could take a data sample and backward engineer the result to find my algorthim, however if my algorithm was strong enough you would have to take many many many (approaching infinity) results...

 

[Mattara]

A computer program is not random. It follows a repetative cycle (only a big one). In real life, random is random. A quantum computer can make random numbers, and nature itself makes random "numbers".

 

[zoobyshoe]

For that criteria, I think human consciousness is the best bet. You could keep track of everything a person has done their whole life, and use an infinitely fast computer, and still not be able to predict (perfectly) what they would do in a situation. Free will makes it hard to find patterns.

"Free will" is a pretty primitive, and also religious, term for people's decision making processes. You could, in principle, create a program that ran on the same dynamics if you could actually pinpoint what those were. The trouble is there can easily be a massive amount of information needed to understand why a person makes one choice and not another, and the decision making processes themselves evolve as the person learns more and is altered by experiences. None of this is random, just unbelievably complex.

 

[Astronuc]

But Brewnog, for chaotic systems like a dice roll, it's clearly not random- it's just so complicated that we can't predict what will happen. Still, if we had a fast enough computer and an accurate measurement of the initial conditions, we could in principle determine exactly how the dice roll would land, or what the weather would be like 5 years from now.

No, not so.

Even in bounded chaos, one can have randomness (stochastic phenomenon) that is not predictable.

The best models and predictions are simply approximations, and always will be.

 

[Anttech]

The output of a program can be random from the view point of someone looking at it's output if you can not see/understand/decrypt the alogorthm used to create the stated output.

the output is percieved as random!

I never said a computer program was random!

(This is basic cryptography)

 

[pi-r8]

"Free will" is a pretty primitive, and also religious, term for people's decision making processes. You could, in principle, create a program that ran on the same dynamics if you could actually pinpoint what those were. The trouble is there can easily be a massive amount of information needed to understand why a person makes one choice and not another, and the decision making processes themselves evolve as the person learns more and is altered by experiences. None of this is random, just unbelievably complex.

It may be a primitive term, but that's because it's directly observable. We're getting into philosophy now , but I don't think it reguires believing in religion to recognize that you have control over your thoughts and actions in a way that, say, a computer does not. Science might be able to one day explain how the brain generates this free will phenomenon, but it can't get escape the fact that it exists. Although, I suppose "random" isn't really the right word to use... "non-deterministic" would be better. My main point was that, no matter how much you knew about a person's brain, you'd never be able to perfectly predict their decisions.

 

[pi-r8]

No, not so.

Even in bounded chaos, one can have randomness (stochastic phenomenon) that is not predictable.

The best models and predictions are simply approximations, and always will be.

I'm not sure what stochastic phenomenon means. But how can something which is completely determined by physics (eg, a dice roll) be really random? Nonpredictable, sure, because we're limited in our knowledge of both the system and of physics. But that's not the same as truly random.

 

[zoobyshoe]

It may be a primitive term, but that's because it's directly observable. We're getting into philosophy now , but I don't think it reguires believing in religion to recognize that you have control over your thoughts and actions in a way that, say, a computer does not. Science might be able to one day explain how the brain generates this free will phenomenon, but it can't get escape the fact that it exists. Although, I suppose "random" isn't really the right word to use... "non-deterministic" would be better. My main point was that, no matter how much you knew about a person's brain, you'd never be able to perfectly predict their decisions.

I agree with your last sentence.

"Free will", though, is a term that originated in religion and which retains religious connotations. The phenomenon you're referring to is just a matter of the minds sophisticated ability to sort through a tangle of desires, values, and information. The reason this would be so hard to predict is that it would be so difficult to simply uncover what they all are in a given individual, and also, to understand the complex dynamics by which a person sorts things out. On top of it all, since a person is constantly receiving information from the world around them all of the factors are in constant flux: their desires are modified, and their values shift.

 

[Astronuc]

I'm not sure what stochastic phenomenon means. But how can something which is completely determined by physics (eg, a dice roll) be really random? Nonpredictable, sure, because we're limited in our knowledge of both the system and of physics. But that's not the same as truly random.

Greek stochastikos skillful in aiming, from stochazesthai to aim at, guess at, from stochos target, aim, guess --
1 : RANDOM; specifically : involving a random variable <a stochastic process>
2 : involving chance or probability : PROBABILISTIC <a stochastic model of radiation-induced mutation>

Merriam-Webster online dictionary.

Now with respect to - "something which is completely determined by physics" - Nature is random and chaotic - Nature determines the 'Physics'. As far as people are concerned, physics is just the study of nature, and at best we develop approximate models as to how we think Nature works. That's all. Sometimes we do a good job, and other times, we fall short.

The result of a die roll or dice roll is still random.

 

[arunbg]

The output of a program can be random from the view point of someone looking at it's output if you can not see/understand/decrypt the alogorthm used to create the stated output.

the output is percieved as random!

I never said a computer program was random!

(This is basic cryptography)

Let me repeat my question.
Is anything in the universe completely random at all levels of observation?
The apparent randomness of your algorithm to an outsider
is because of his level of observation.On the other hand the output of the program is not random to the programmer.
I feel that the term random is unfortunate and should be replaced with something like undetermined (especially in case of probability and the like)
From a more scientific point of view, aren't randomness and probability the building blocks of quantum physics?
Doesn't the uncertainty principle refute our views on absence of absolute randomness (it is impossible to measure with absolute certainty the momentum and position of an object)?

 

[Cyrus]

I would think the diffraction experiment is random. We have no way to know where the electron goes. We just know a probable distribution of it from a statistical standpoint.

The uncertainty principle says that we cannot know with 100% certainty the position or the momentum of any object, because we disturb the object when we measure it. I do not see how this refutes absolute randomness; it just means we have no way to know exactly position or momentum. That's a measurement issue, which for the double slit experiment, does not help us any. We can see where the photos go when they make the diffraction pattern, (That’s measurement issues) we CANT say what happens along the way. It is acting as a wave; we don’t know where the photon is. When it does interact it no longer behaves as a wave, it interacts as a particle. Meh, my explanation sucks.

 

[zoobyshoe]

Let me repeat my question.
Is anything in the universe completely random at all levels of observation?
The apparent randomness of your algorithm to an outsider
is because of his level of observation.On the other hand the output of the program is not random to the programmer.
I feel that the term random is unfortunate and should be replaced with something like undetermined (especially in case of probability and the like)
From a more scientific point of view, aren't randomness and probability the building blocks of quantum physics?
Doesn't the uncertainty principle refute our views on absence of absolute randomness (it is impossible to measure with absolute certainty the momentum and position of an object)?

I don't think anything is random at any level of observation. I don't think cause and effect ever just cuts out and allows anything to perform any action without a cause. Things appear random to us when we don't understand the causes.

 

[arunbg]

The uncertainty principle says that we cannot know with 100% certainty the position or the momentum of any object, because we disturb the object when we measure it.

I quote my textbook(high school national syllabus text mind you)

It may be emphasised that the uncertainty principle is not due to any limitation of the measuring instrument

So then why is the object disturbed and is there no way to measure the magnitude of this disturbance ?
If there is indeed no way, then it becomes completely random and all our arguments are in vain.

P.S: I have never understood the uncertainty principle as taught in school.
Then again, it must be because I just don't know enough.

 

[zoobyshoe]

It may be emphasised that the uncertainty principle is not due to any limitation of the measuring instrument

What's the context of this statement?

 

[Cyrus]

Yes. What I am saying is that by using the instrument, you must disturb what you measure. It is a fundamental issue. It does not matter how 'good' your instrument is, it HAS to distrub/interact with what its measuring.

Here, perhaps this will help:

It is tempting to suppose that we could get greater precision by using more sophisticated detectors of position and momentum. This turns out not to be possible. To detect a particle, the detector must interact with it, and this interaction unavoidably changes the state of motion of the particel, introducing uncertainty about its origional state. For example, if we were to bounce shorter-wavelength photons off a particle to better locate its position, the larger photon momentum h/lambda would amke the particle recoil more, giving us greater uncertainty in its momentum. A more detailed analysis of such hypothetical experiments shows that the uncertainties we ahve described are fundamental and intrinsic.

Let's put it another way, suppose I am trying to measure the location of a particle. I try to do this by using a wavelength of photon. Well, when that photon meets the particle, it is going to interact with it, thereby changing the state of the particle, and thus its momentum in the process. So the better I measure its position, the less I know its momentum.

 

[arunbg]

Ok , so you can't measure the position and momentum of a body with absolute certainty.
But is it not quite certain that the particle indeed exists somewhere in some state which cant't be predicted through measurement or do we also have to take its dual nature into account ?

 

[Cyrus]

Ok , so you can't measure the position and momentum of a body with absolute certainty.

Correct.

For the second part, I *think* the following is true.

Consider a double slit experiment. In classical phyiscs, if we shine light through the two slits we would expect to see two columns of light shine on the wall on the other end of the slits. If this were the case, we would know that all the electrons that pass through the slight simply traveled straight through . So we could model its location with time by the simple displacement equations, x = \int vdt.

But this is not what we see. This means the electrons are not going straight through the double slits. In fact, we can only say there is some probability of where they might be in space. Notice I said where they might be in space, and I did not say how to detect where they are. What I mean is that I can't say, for example, y=f(t) is where the partice is as a function of time in theory. I can only say where it is probable to be.

Now let's throw in the uncertainty principle. I know that I can only estimate where the particle ought to be , so let's say I go out with my expensive detector and try to find it. The uncertainty princple says that if I have found the particle, (which I did not know with 100% certainty where it was in space), also I now have some intrinsic uncertainties due to trying to measure it because I have disturbed it in the process of measurement.

Meh, this is getting to be confusing. Let me put it this way, you don't know where the particle is for sure. In addition, if you happen to find a particle, when you detect it you are going to have additional uncertainties as well.

Maybe ZapperZ or someone smart can explain this to you better. I am doing a lousy job.

 

[pi-r8]

Merriam-Webster online dictionary.

Now with respect to - "something which is completely determined by physics" - Nature is random and chaotic - Nature determines the 'Physics'. As far as people are concerned, physics is just the study of nature, and at best we develop approximate models as to how we think Nature works. That's all. Sometimes we do a good job, and other times, we fall short.

The result of a die roll or dice roll is still random.

You seem to be saying that EVERYTHING in nature is random. Not just dice rolls, but even such apparently simple phenomenom as pendulum and planets orbiting the sun. I'd like to know why you consider those to be random.

 

[arunbg]

In classical phyiscs, if we shine light through the two slits we would expect to see two columns of light shine on the wall on the other end of the slits.

I have always thought that the phenomena of diffraction and interference of EM waves is explained by the classical wave theory of Maxwell and Huygens constructions.Correct me if I am wrong.

Although I haven't yet studied this phenomena with an electron beam, considering what you said about the impossibility of determining where the electron eventually lands , it seems like this is only due to lack of sufficient knowledge or information regarding the nature and interactions involved in the phenomenon.
Otherwise please clarify why it is impossible to determine the coordinates
(measuring errors apart).

A word on probability distributions.
Of the very few probability distributions that I have studied (like velocities of molecules of a gas) the theories come close to predicting the actual distribution and of course it is possible to obtain the exact distributions once we take the infinite or so parameters affecting into consideration.
So there is nothing "random" about it (again measurements apart).

 

[Cyrus]

The fact that you can't take into account the infinite parameters affecting the consideration is exactly why it is random.

 

[arunbg]

The fact that you can't take into account the infinite parameters affecting the consideration is exactly why it is random.

I thought you'd say exactly that as I was typing that sentence.
(which also means that the human mind is not as random as it seems )

But the fact still remains that at least theoretically nothing is random(even with the near infinite conditions).
The diffraction phenomenon is equally the same as any other such phenomenon that requires the infinite parameters.

 

[BobG]

I have always thought that the phenomena of diffraction and interference of EM waves is explained by the classical wave theory of Maxwell and Huygens constructions.Correct me if I am wrong.

Although I haven't yet studied this phenomena with an electron beam, considering what you said about the impossibility of determining where the electron eventually lands , it seems like this is only due to lack of sufficient knowledge or information regarding the nature and interactions involved in the phenomenon.
Otherwise please clarify why it is impossible to determine the coordinates
(measuring errors apart).

A word on probability distributions.
Of the very few probability distributions that I have studied (like velocities of molecules of a gas) the theories come close to predicting the actual distribution and of course it is possible to obtain the exact distributions once we take the infinite or so parameters affecting into consideration.
So there is nothing "random" about it (again measurements apart).

Even with a random dice roll, you have a pattern of distribution. There's more ways for two dice to add up to 6 or 7 than there are to add up to 12. That doesn't mean you can accurately predict that one die will be a 2 and the second a 5.

That's what cyrus is talking about with the double slit experiment (the single slit experiment is probably a better example, though). You can predict the pattern created by many photons, but you can't predict which path any given photon will take.

 

[arunbg]

Even with a random dice roll, you have a pattern of distribution. There's more ways for two dice to add up to 6 or 7 than there are to add up to 12. That doesn't mean you can accurately predict that one die will be a 2 and the second a 5.

Why do we always keep returning to sqare 1 ?
Let me repeat my question a second time (3 overall).
Is anything in the universe completely random at all levels of observation ?

The outcomes of the roll of a die may appear completely "random" to a casual observer but to an expert (and hardworking) physicist, if he takes all conditions involved, the outcomes are anything but random.They can be explained by a handful of fundamental laws of physics.
The same is the case of the electron beam.

My question is not regarding the relevance of probability, rather the limit to which probability is relavent.

 

[Astronuc]

You seem to be saying that EVERYTHING in nature is random. Not just dice rolls, but even such apparently simple phenomenom as pendulum and planets orbiting the sun. I'd like to know why you consider those to be random.

Well, in one sense everything is deterministic, however we just can't process all the information.

On the other hand, on a die roll, things like air pressure, air resistance, the temperature of the die are variable, hence random.

Cyrus makes a good point about 'uncertainty'. It is inherent in Nature.

I think the problem one is facing is that there has to be an outcome, but is the outcome necessarily random? Statistically, one must consider a population of outcomes, and from that determine if something is random.

 

[ianthow]

No, I don't think there is a complete random, when there are too many variables to consider, we conveniently put it as random or chaos.

If you look at human history and the development of science and technology, certain random and chaos in the ancient times became predictable and solvable with today's computing power as we can handle more variables. Something random at this moment in time might become predictable. So, is there a true randomness? I don't think so.

Just as the definition of "infinite", is there truly an "infinite" or is that the number is too big to comprehend, we again conveniently put it as infinite.

 

[Kazza_765]

What about a decaying nucleus? Two nuclei that are completely identical in every single respect. One of them decays, and one of them doesn't. This doesn't follow cause and effect. And its completely random, no way of ever predicting which will decay and which won't, no matter how good your measuring instruments.

And its also my understanding in the double slit experiment that the point of detection of any individual electron/photon is completely random. All we can ever know about the particle is its wave function, and all that gives us is probabilities. But where it actually hits is completely unpredictable.

Ok , so you can't measure the position and momentum of a body with absolute certainty.
But is it not quite certain that the particle indeed exists somewhere in some state which cant't be predicted through measurement or do we also have to take its dual nature into account ?

Nope, this is what Heisenberg was trying to point out. Its not the case that the particle has a definate position and we just can't measure it. The particle actually doesn't have a clear position and momentum. When you take into acount the dual nature, and you represent the particle as a wave packet, in order to localize that wave packet over an increasingly small area, you must necessarily increase the uncertainty in its momentum.

 

[Hanny]

Reviving an old thread

Firstly, I'm totally new here so what I say may be rubbish.

What about a decaying nucleus? Two nuclei that are completely identical in every single respect. One of them decays, and one of them doesn't. This doesn't follow cause and effect. And its completely random, no way of ever predicting which will decay and which won't, no matter how good your measuring instruments.

I think that the notion of nucleus decay being random is not due to the decay occurring at that exact moment by chance, but rather that we don't have sufficient knowledge to determine what the trigger for nuclear decay really is.

Take the example of a random number generator from a computer. Although the numbers that come from it are seemingly random to an observer, once the algorhythm for generating the numbers is revealed, any observer with an identical computer can "predict" the next number by punching the same input conditions into an identical machine. In the same way, if we knew the causes of nuclear decay and were able to measure them, it would be possible for a complex algorhythm for predicting the nuclear decay to be set up based on the input conditions (the state of the atom).

An even more interesting point that stems from this idea is the notion of predestination. If everything in the universe followed a fixed set of rules governing their behaviour, it would mean an infinite number of universes created from an identical big bang would follow the exact same path of development.

The implications of this would be that everything in this current universe is following a path of reactions that cannot be deviated from. I am aware, however, that there is a notion of human free will that would technically create "randomness" and destroy such a predestination. However, once you look at humans as simply a complex interaction of energy and mass this notion disappears; why should the reactions between the particles in the human brain be any less predictable than the reactions of the particles inside a cooking loaf of bread?

It seems that everything is predestined?

I know that this argument is probably hugely flawed somewhere.. or everywhere! I'd love for you guys to come and tear it apart :)

Cheers,
Hanny

 

[tribdog]

Go back to Schroedinger's Cat. The original experiment was designed to make sure it was random whether or not the particle decayed.

 

[tribdog]

the outcome of a roll of a die is also random no computer, no matter how fast and accurate can predict what it will be before it is rolled.

 

[rcgldr]

Any algorithm that calculates the "nth digits" of pi or similar constants generates random sequences, and there are programs to do this, so computers can be used to generate random sequences. Go to the link below, click on algorithms, then "nth digit ...":

constants.html

 

[TuviaDaCat]

so a chaotic pendulum has no physical value which has a measurable chance to happen?

 

[Pythagorean]

http://www.myphysicslab.com/pendulum2.htmlEdit: this is not my physics lab

 

[AO eye 5]

A small contribution for the thread.

Since this is still in general discussion, there was something interesting I recall on how a mathematican considers the random nature of pi; that is it not random, but displays every pattern possible.

I feel this thread needs to be where it should.