Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Does random exist?

  1. Jul 21, 2008 #1
    Not sure where to put this, or if this even belongs on these forums at all.

    But recently I was thinking about random, probability clouds, and the like.

    Does random actually exist or is it simply another way of saying “too complicated for us to know right now”

    Obviously in questioning a word you need to agree on a definition first, I used the first line from google”define:random”

    “lacking any definite plan or order or purpose; governed by or depending on chance; "a random choice"; "bombs fell at random"; "random movements"

    It just seems to me that everything we call random, is just something with a ton of variables, or something with variables so small we cannot measure them without changing them.

    Feynman constantly corrects himself in his lectures after saying random, with “very complicated”

    Please move this to the correct section if needed or simply delete if it has no business being here 
  2. jcsd
  3. Jul 21, 2008 #2


    User Avatar
    Science Advisor
    Homework Helper

    The time for an individual radioatactive atom to decay is random - or there is a mechanism we don't understand yet.!
    That the famous Einstein quote that "God doesn't play dice" - he beleived that there must be some mechanistic process we didn't know about.
  4. Jul 21, 2008 #3
    I know of plenty of "random" things we have declared in science, because at this time we dont fully understand it. So to progress any further in something that involves this "random"
    feature, we need to refer to it as random.

    But I guess my question was more "(should/do) science related people/fields believe (truly) in random, as it is in a sense giving up (even if temporarily)"

    If you know something isnt actually random, but is random enough to not be known as of now, you still know there is a reason for it, and may someday return to it to fully understand.

    But if you consider something truly random, there is no reason to ever fully understand it, as understanding it as random, is fully understanding it.
  5. Jul 21, 2008 #4
    Chaos is. I think on the quantum level fields act and react chaotically.

    Removing "random" from existence is implicit of design. The size, velocity, direction the earth blew out of the big bang, with all of the elements it has... life the universe and everything being "designed"... asteroids who's sole mission is to be chewed up entering our atmosphere...

    I would say chaos exists. It is really more philosophy than physics, as it is not at all provable... which is the only reason I bother posting. Physics is not my friend.
  6. Jul 21, 2008 #5
    I guess this is more of a philosophy question. :-\

    But I also think its relevant to sciences in that, if a scientist were to deem something random, they may be prone to never investigating the situation more.

    It basically boils down to an almost religious question. Is it possible for an uncaused cause to exist. We generally expect things to follow the cause/effect logic, but for the concept of true random/chaos to exist, it would have to break that logic.

    Sure a chaotic event could be cause, but that’s a much higher level. When the actual chaos is looked at, there would have to be effects without cause at some point.

    Or am I spouting broken logic now?

    I looked into chaos a bit more and found

    So it seems that even chaos is tightly rule-bound but very sensitive to its initial state.
    So its a game of working out all the variables, which is what science is all about right?
    Last edited: Jul 21, 2008
  7. Jul 21, 2008 #6
    Just a quick note; if interested in the mathematics of chaos and randomness, read the book "Chaos" by James Gleick
  8. Jul 22, 2008 #7
    Whoa! Chaos, in a technical sense, is very far from involving randomness!

    Chaos is always deterministic: if you know the current state exactly, you can predict the future state at anytime.
  9. Jul 22, 2008 #8
    awesome! Im glad I interprited that correctly. Thats what I was trying to get at the entire time.

    Anything random, or chaotic just seems to me as though it is very much dependant on the variables at the exact state ..

    Now, Im not sure if this is allowed on these forums or not, but I've always thought this also applied to people, decisions.

    Most science people I talk to about this, agree that nothing is actually random, random is just a useful word we use to describe something too complicated for us to know atm.

    But most disagree when it comes to people. "I can randomly punch you in the face right now"

    Isnt this still deterministic based on their current brain chemistry/dna/etc etc?

    (sorry this is getting bio, but I find that bio is just a higher abstraction of physics isnt it?)
  10. Jul 22, 2008 #9

    Claude Bile

    User Avatar
    Science Advisor

    You touch on the issue of determinism; prior to the 20th century (i.e. the era of "classical" physics), the standard attitude amongst physicists was that, provided we have sufficient information about a system (even one as large and complicated as our universe) then we can in principle predict the behaviour of the system for all time.

    The discovery of quantum phenomena challenged this long-standing philosophy. Some argued that quantum phenomena are truly random whilst others insisted that there is some underlying mechanism that produces results that are only seemingly random.

    It is now readily accepted amongst physicists that quantum phenomena are truly random. "Classical" randomness (like rolling dice), we now understand to be deterministic, and are only perceived as random through lack of information.

    The connection between quantum randomness and how it might seed our perception of "freedom of choice" so to speak on a biological level is very poorly understood.

  11. Jul 22, 2008 #10


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    Although the equations of motion in "deterministic chaos" are, well, deterministic, the point is that they act on real number quantities. And you can never know a real number "precisely". It takes an infinite amount of information. So the randomness in deterministic chaos comes about from the fundamental impossibility to know an initial state made up of real numbers precisely.
    In other words, the randomness doesn't come from the dynamics, but from the initial conditions, but is unavoidable.
  12. Jul 23, 2008 #11
    Lets say you had a collection of information. If you continue to break the information down, it seemingly becomes more and more random. Deterministic processes occur smoothly, meaning that you could break the processes down and assess all the mechanistic features allowing for the phenomenon to occur. Now if the process seems to occur by itself (meaning there are no mechanistic elements to it), it tends to become more random...technically, when we call something random, it is because we do not know how it works in a deterministic manner (if it does).
  13. Jul 23, 2008 #12
    I find this just about as disturbing as a religious person would find my argument.

    Ive been searching around and find a decent amount of opposition to "quantum phenomena is truely random"

    Feynman, while explaining his diagrams and paths of photons made sure to explain that using probability clouds/paths are our best means of making anything of what we are viewing, but that this does not mean these events/phenomena are actually random, but just out of the reach of our current instrumentation.

    I makes sense to me that it would be hard to measure the position of a baseball, by throwing baseballs at it. If we are trying to test the smallest known particles (some i dont think we are even positive exist, like gravitrons? again... not sure, im new here :D) I can swallow having a hard time, and confusing results when experimenting using machines made of the things we are measuring.
  14. Jul 23, 2008 #13


    User Avatar

    Staff: Mentor

    So is the Schroedingers cat dead, or alive?

    I thought Laplace's daemon was impaled with aspen stake by Heisenberg's uncertainty principle.
  15. Jul 23, 2008 #14

    Andy Resnick

    User Avatar
    Science Advisor
    Education Advisor

    A dynamic variable- position, direction, phase, etc. can indeed be random, and random has a well-defined mathematical sense as well.

    Consider the two-point correlation function of a continuous variable- a measure of how well we can predict field values are *there and then* given knowledge of it *here and now*. Note that this knowledge can be exact.

    An incoherent field has zero predictability, while a coherent field has infinite predictability. Real physical fields are partially coherent, meaning that our ability to predict future values is limited but non-zero.
  16. Jul 23, 2008 #15
    i think it was an excellent point that in science we determine things with a cause/effect relationship... and if we cannot determine a cause it may seem to be random but once we figure out the cause it is no longer random...

    heres an example...

    rain drops landing on the ground in certain locations may seem random when looking at the ground.. but when we find out where and when the rain drop starts to fall we can dynamically predict where it will land... and then you can say that raindrops land in locations, random, in relation to each other but if we figure out why the raindrops fall and what are all the exact causes, we may be able to determine where every raindrop lands and why!
  17. Jul 23, 2008 #16
    The Schrödinger’s cat experiment relies on true randomness of radioactive decay right?

    What I am questioning, is the randomness of this decay. Things seem to follow (based on past history and science) that things are random until proven. We use probability to estimate something unknown, or not-easily measurable at the given time, with out given instrumentation.

    Is it impossible to measure a photon's position without altering its future path? (I'm pretty sure right now it is) but would you go so far as to say its impossible period?

    Im not going out on the Disney limb and claiming "anything’s possible"

    But to me, science has always been awesome because it constantly proves itself wrong. To some this is disconcerting. To me, it simply shows that we are making progress. If we picked one thing, and stuck with it, without questioning it I think science would be considered a religion.

    An incoherent field by definition has zero predictability, this doesn’t mean it is in anyway real right? By talking about the concept of "truly random" just because we have a word for a concept, does not mean it exists, or could exist.

    I personally believe our ability to predict the future value of any given experiment is based on how much information we have about the starting state, and information about the rules of interactions that would occur during the experiment.

    Ie, tossing dice (material, height, velocity, angle, pitch ,yaw ,roll ,weight , floor material, etc etc, and the rules for interactions would be basic physics collisions)

    This is exactly what I was thinking, I just wanted to see how many other people on here felt the same.

    It seems like random is a cop-out for not accepting the fact that we dont know. Some people cannot stand not knowing things, and it would seem as though random came about from these types of people. There are also some people who are perfectly fine with not knowing.

    Feynman-"I can live with doubt and uncertainty and not knowing. I think it is much more interesting to live not knowing than to have answers that might be wrong."
  18. Jul 23, 2008 #17
    wow, I've never seen this argument before. :) Thats really cool!

    But I guess Im not fully understanding heisenbergs unceretainty principle then.

    "In quantum physics, the Heisenberg uncertainty principle is the statement that locating a particle in a small region of space makes the velocity of the particle uncertain; and conversely, that measuring the velocity of a particle precisely makes the position uncertain."

    but the laplace daemon implys position and velocity.

    as vanesch said, its impossible for us to know everything precicesly, which is why random seems so real.

    But this doesnt mean that random exists in any sense other than what vanesch said "In other words, the randomness doesn't come from the dynamics, but from the initial conditions, but is unavoidable."
  19. Jul 23, 2008 #18


    User Avatar

    Staff: Mentor

    Laplace had no idea about uncertainty principle, so he thought world is completely determinitic.

    Uncertainty principle tells that we can't know exact speed and exact position at the same time. It is not an effect of the low precision of our instruments. It is the intrinsic property of our world. The better we know the speed, the less we know about the position and vice versa.

    In other words, Heisenberg train moves with exactly 30 mph, we just don't know if it is on rails :wink:

    Sure uncertainty of train position is so small that it completely doesn't matter.
  20. Jul 23, 2008 #19
    right, but in the theoretical situation that we did know both, everything would be deterministic from that point on right?
  21. Jul 23, 2008 #20


    User Avatar

    Staff: Mentor

    This is a pure speculation, so we can get to any conclusions you want :smile:
  22. Jul 23, 2008 #21
    dang lol

    I guess Im wondering if there is a difference between finding something out, and knowing something.

    Im having a hard time understanding how you cannot know a particles position and velocity.

    I understand how it could be impossible/impracticle to find this out.

    In reading more, I guess I need to find out more about the whole wave collapse deal with quantum physics right? The whole thing you're talking about is that, in finding one or the other out, you've collapse the wave and changed the experiment. That it may not necessarily have a position or velocity until measured?

    or are you saying it does have a position and velocity, but we cannot know both.
  23. Jul 23, 2008 #22


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    Well, quantum theory tells you (that doesn't mean that that is the way nature IS! it only means that that is what the model we've set up of nature does. We cannot know what nature is, but we do of course know what a model tells you about it) that a particle doesn't HAVE a precise position and a velocity. The reason why you cannot KNOW it is that it doesn't HAVE it. At least, "standard" quantum theory. A particle with a well-defined position - which means, a particle quantum state that will give you that position with certainty - is MADE UP of SEVERAL velocity states. Not just one. Several. In fact, all of them. The particle "has several velocities in parallel" according to the quantum mechanical description. It is as if you had a bucket with a red, a green and a blue ball in it, and you insist on "wanting to know the color of the ball in the bucket". There IS no single color. There are 3 of them. If you "measure" it, you pick out one of them, which can then arbitrarily be green, red or blue, but that's no "measurement error". It is because the 3 colors are present. (however, be careful with this analogy: quantum superposition is more subtle than this).

    Again, this is what the quantum-mechanical model gives us as its best description - that doesn't mean that this is a faithful representation of the inner gears and wheels of nature.

    So, in as much as someone thinking in classical terms "cannot imagine how you cannot know a particle's position and momentum", in quantum-mechanical thinking, one "cannot imagine how a particle could ever have a precise momentum and position at the same time". Simply because in the quantum mechanical description, there's no such particle state that HAS a precise momentum and position. The mathematical proof of the Heisenberg uncertainty relationship is in fact nothing else but that: of all quantum mechanical states, none exists that is made up of position states that are within a certain spread, and is made up of momentum states also within a certain spread, such that both spreads violate the Heisenberg uncertainty relationship.
    In quantum mechanics, it is not an "uncertainty" relationship, it simply means that there doesn't EXIST any quantum state that HAS at the same time a momentum within a certain interval, and a position within a certain interval.

    Now, there are certain views on quantum mechanics, such as Bohmian mechanics, who DO assign precise positions and momenta to particles, and then use the quantum mechanical wavefunction as a kind of force field that guides these particles. In fact, Bohmian mechanics is classical mechanics with extra forces, determined by the quantum mechanical wavefunction. In THAT view (given that it is classical mechanics), particles DO have precise positions and momenta, and it is indeed the act of measurement, through the workings of the "quantum force" (which behaves in highly contrived ways in Bohmian mechanics) which changes the trajectories of the particles in a seemingly random way. But in this theory (which gives the same results as standard non-relativistic quantum mechanics ; there are serious problems with a relativistic version, although these can be overcome it seems), the state description of the particle is like in standard classical physics. It is not the same as in normal quantum mechanics, where the wavefunction is not a force field, but is the state description of the particles.
    Last edited: Jul 23, 2008
  24. Jul 23, 2008 #23
    Ok, im definatly going to do some more reading.

    Some quick questions though, as we got a bit side-tracked with my lack of knowledge and excitement that someone would sit down and explain this stuff to me :)

    as of now, am I correct in chalking a particles lack of position and velocity to the fact that particles are another form of energy? being that we can measure them using certain tools and what not, but we can only measure them using other particles which are also manifestations of energy as well?

    back to the main question though, even if its not position or velocity, the names for the variables are irrelovent. If one had all of the information/variables to a system (universe or closed off or w/e) wouldnt the system be deterministic from that point on?

    Im basically fishing for whether or not random is ever truely random, or random can only be an inverse value to the amount of information we have.
  25. Jul 24, 2008 #24


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    "are particles a form of energy" is a statement that comes up regularly, but I fail to see what it could even mean. Because of course, that begs the question: what is "energy" then, so that it can "make up" particles ? Some jelly-like all-pervading stuff ? Or a number ?
    Whatever is a particle (in a specific model!) is given by the model's definition. In classical mechanics, the theatre is set up (3-dim Euclidean space for starters - it gets actually more complicated than that because of Galilean relativity). We say that particles are "matter points in space, which move on a continuous trajectory".
    Now, would it make sense to say, in classical mechanics: "are particles a form of position" ?

    This question sounds to me as gibberish. Are particles "made up" of "position" ?
    Do particles HAVE a position ? In classical mechanics, definitely yes. THAT makes sense. But ARE particles made up of position ? Is hot water made up of temperature ?

    In classical mechanics, "energy" is a number that we can calculate for a certain dynamical state, and in many respects it is an *interesting* number, because it has certain properties (is conserved over time, for instance).

    In quantum mechanics, we can do 2 things: define a quantum system "from scratch", or "be inspired" by a classical system. Many people think that we need a classical system and then "quantize" it - even Landau says so in his books. But that isn't true. Quantum mechanics can be entirely defined "from scratch". For instance, particle spin is "quantum mechanics from scratch". There's no classical model for it, which has been quantized.
    But most quantum mechanical systems are "quantized classical models". The classical model then serves as a guiding principle for the setup of the state space (and its "interpretation").

    A quantum theory starts out by giving a "complete state space basis". That is, you have to give a list of possible states your "system" is going to be in, which you will be able to measure. That can be a discrete list, or a continuous list.
    The simplest non-trivial discrete list is a list of 2 states, call it "up" and "down". It is up to you to link these states to some operational procedure to "measure" these states, in other words, to give a physical interpretation to these states. That's YOU who puts this in. It's part of your model building. It is not quantum theory that is going to tell you.
    This means that your state space is going to be 2-dimensional, and that all possible states are going to be of the form u |up> + v |down>
    You might also want to define "other" measurements, which have another basis. That's entirely up to you, and it will depend on properties you want your system to have. You might for instance define a measurement procedure which gives you |left> and |right>. |left> is then equal to |up> + |down> and |right> is equal to |up> - |down> for instance. Whether that is a good idea, and whether that corresponds to one or other genuine laboratory measurement on your system, is again entirely your responsability. You might define things that way, but it might not work out that way in the lab: it simply means that you didn't build a good quantum model of your system and your measurements.

    Next, you will have to define a time evolution of each of these states in your list. If I have "up" at a moment t0, what state (of the form u |up> + v |down> ) will my system be in at time t1 We can do that in any way we like, but the time evolution needs to obey certain properties (like being unitary).

    Now, there's a big help, because it turns out that in most useful systems, there are eigenstates to the time evolution operator: that is: states that remain themselves up to a factor. States that don't mix with others. We call such states: energy eigenstates. Usually, this is the first thing we look for when setting up a quantum system: what are the energy eigenstates.

    This is very analogous to the "energy" concept in classical mechanics, which had its main usefulness in conservative systems where energy was conserved in time.

    So we see that quantum theory allows for a huge number of possibilities of models. We wouldn't know where to start, so to say. That's why there's a way to derive quantum models from classical systems.

    If you have a classical system, you have a so-called "configuration space". For instance, for a single point particle, that configuration space is just Euclidean space. For a 2-particle system, it is a 6-dimensional space made up of the 6 coordinates of the 2 particles (3 + 3).

    Well, the configuration space is going to serve as a "complete state space basis". To each POINT in the configuration space corresponds one quantum state of our state space basis. That already means that our statespace is going to be infinite-dimensional. Let us stick to the 1 particle classical system. Our basis is now the list of all possible positions of the particle (that list is nothing else but 3-dim Euclidean space). So we have now not 2 states (up and down), but a state for each point in space ! There's infinity of them.
    We interpret these states of course as "precise position states".

    In our classical system, we also have momenta. Well, it turns out that we have to take momentum states as special combinations of position states. Momentum states are like the "left" and "right" states earlier. In fact, a momentum state with momentum p is a sum over all position states, with a coefficient in the sum which is nothing else but exp(i p x).

    That didn't have to be so, but it turns out that if we do this that way, that the quantum models that we build do work.

    We find our time evolution by looking at the classical hamiltonian of the system. It tells us how to find an operator which is the time derivative of the time evolution operator (this time derivative operator is usually called "the hamiltonian" operator). This is nothing else but the Schroedinger equation.


    Now, note that there is something strangely perturbing to quantum theory: the time evolution of the quantum state is entirely deterministic ! Where's the randomness ? The randomness in quantum theory comes from the following rule:

    If you are going to do a measurement, to that measurement corresponds a certain set of base states. For instance, if you are going to do a position measurement, then our initial defining base (inspired by the configuration space of the classical system) is the set of measurement base states. The actual quantum state can be a sum of those base states. It doesn't have to be exactly one of those states. Well, quantum mechanics tells us that the probability of obtaining a certain result (which, in the case of position measurement, is a single, precise, position), is then given by the value squared of the coefficient of the corresponding base state in the sum.

    For instance, if our state at a certain moment is 1/sqrt(2) |up> - 1/sqrt(2) |down>, and we are going to do an up/down measurement, then there is 1/2 chance to find "up" and 1/2 chance to find "down".

    After the measurement, the quantum state has abruptly changed, in agreement with the outcome of the measurement.

    Well, again, in order to even be able to answer that question, you should tell me what model you use. If it is standard quantum mechanics, and we accept that the quantum state is "all there is to know" to a system, then there's no way to know what measurements will give us beyond the probabilities given by quantum theory. If you have another model, then depending on your model, you will answer that question this or so.

    For instance, there's a very simple model, which is old as the world, and which is entirely deterministic: "it is written in god's book". All events, past, present and future, are part of a big catalog. No "dynamics", no "time evolution", no "state of the system": just a huge catalog of everything that happened and will happen. Useless model, yes. But thinkable model.
    Our "laws of nature" are then nothing else but funny correlations between those events in that big catalog. It is a universal model, which can apply in just all circumstances.
    You can't get more deterministic.

    As I said, that's model dependent. It is always possible to consider that everything is deterministic, given that there exists a universal model of nature that works that way "god's list".
  26. Jul 24, 2008 #25

    Regarding the point about god's book..
    Is there a way in laymans terms, to describe how a system could work, without it being written in 'gods book'?
    Because that's what I'm leaning towards.
    It is impossible for humans to describe an uncaused event, because we will always be asking, well what happened before the event?
    Saying "it popped out of nowhere!" is certainly a suitable answer, and it works as a mind model, but it doesn't /explain/ anything.
    Also, if everything is in gods book, and everything is completely deterministic for the lifespan of the system, won't all randomness in the system be simply a lack of information regarding the event?

    Goes something like this I guess:
    1. Energy cannot be destroyed or created
    2. All known events come from an effect, and lead to an effect
    3. IF 2 is correct then: All known events must be deterministic, and predictable through time.
    4. IF 2 is NOT correct, then energy can either be created, or destroyed
    5. Since we know of no such mechanism, all events must be deterministic.

    Is it possible to say that if the energy in the universe is constant, it must be deterministic?
    I mean, if there's always constant energy/mass, and let's say there was a seemingly random uncaused event, then the event would either have to occur in a time scale unmeasurable by us(meaning it happened instantly, and for no physical reason)
    Like a car turning into a cow..
    Or else, it would have to be created by energy not in the universe before it.. The cow just popped out instantly.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook