Does randomness exist?

I thought I'd add another light to all of this, somewhat covered already. I'm going to basically paraphrase Dawkins here, since he has put succinctly both the reasons for calling it mutations random, and for not calling mutations random.Mutations have physical causes, for example X-Rays, so in this sense they are not random. Mutations are more likely to happen to some genes than to others, another kind of non-randomness. Mutation can be reversed. In some cases the probability of the mutation in either direction is equiprobable, however in some cases mutation in one direction is more frequent than reversing mutations in the other direction...leading to mutation pressure. This is describable as non-randomness.However, the great majority of mutations, regardless of their cause, are random with respect to the improvement (eg reproductive success) of the mutated organism/offspring. Naturally, this means that most mutations are neutral or bad since there are many more ways of breaking something than improving it.It is referred to as random mutation to highlight this. The mutations are not directed towards a specific goal, in contrast with natural selection which is goal orientated towards reproductive success.

I Propose that the universe (or "GOD", your choice ) does play dice (metaphorically) ....& they are loaded.
that way you get both a degree (not totally although IMHO) of randomness denoted by the physical realities of quantum uncertainty and an almost (again not totally but at a higher frequency than the random quality/factor IMHO) deterministic reality. This is my logical argument without empirical evidence so i would really appreciate some critical peer review so i can feel confident in adding it to my arsenal of Understanding Existance Efficiantly.I can only wonder the following if the above is factual
How many facets on the dice are there (this determines the starting states and values possible ) & where is the bias placed?
Was it rolled once and for all time or are the dice in a state of constant flux?
Is there more than 1 dice being rolled?

---

Become totally disillusioned & see everything for real.

I believe randomness is relative to ones view and understanding.

A problem with randomness and probability theory is that they are human inventions and must not necessarily have anything to do with reality. But probability theory makes it possible to make very good models of natural processes, which may also be simulated on computers.Simulated evolution - using pseudo random generators - is frequently used to solve very difficult technical problems. Such a generator may for instance go through all 12-decimal numbers between 0 and 1, one number at a time. After this the sequence will repeat itself. So, the sequence is not exactly random. But the Fourier transform of the sequence, giving the frequency spectrum, reveals that all frequency components all have the same amplitude. So, this is a very good approximation of so-called white noise.There will always be a mathematical model of the problem and the process has to climb a mathematical landscape, a peak being the solution to the problem. Now, if the landscape has only one peak the random climbing will hopefully arrive at the solution. And it will probably arrive at the same solution even if we use a quite different pseudo random sequence going through all 16-decimal numbers between 0 and 1 for instance. So, the solution depends solely on the landscape.But many landscapes may have an astronomical number of peaks. For instance, if we have 170 functions f1(x1), f2(x2), . , f170(x170) of the parameters x1, x2 . , x170 each having 3 peaks, then the number of peaks in a 170-dimensional landscape formed as the sum of these functionsf(x1) + f(x2) + . + f(170)will be of the same order of magnitude as the total number of atoms in universe.The natural evolution is supposed to climb a genetic landscape, the complexity of which will hardly be second to the former. Nevertheless, evolution has the ambition to find the highest peak, even though there is no guarantee. The reason is that it effectuates a simultaneous maximization of mean fitness and genetic disorder using the rules of genetic variation as a random number generator; see my contribution in Intelligent Design/A proof against ID and creationism/Evolution as intelligent design. This means that local peaks are smoothed out. And when the disorder increases, peaks are even more smoothed out and the hollow between them disappears.But since the number of individuals participating in the statistical optimization process is always limited the outcome is always uncertain. So, if there is evolution on some other planet, species may differ from our species even if the landscape is mainly the same, because the number of peaks is astronomical. But some species may perhaps look the same.Even if the climbing takes place at random, the landscape has an impact on the shape and behavior of organisms to evolve. It depends on the laws of nature, properties of RNA or DNA molecules whose origin are not known. I. E. I do not know if electro-magnetism has emanated from some random process.I also think that freedom of choice is not possible in a deterministic system and that randomness is the only way to get some freedom. The problem, however, is that randomness has no will. But an illusive will may originate from the ability of the process always striving towards a maximum or a peak. Besides, the random process may exhibit both imagination and creativity. It prefers higher peaks prior to lower, or better alternatives prior to worse. Anyway I am convinced that there is a lot of randomness in universe.

---

gkm

where is the randomness in the following" is sit at a table with a normal full 52 card deck of cards , i go to "RANDOMLY" draw one card from the deck , as my hand reaches out , the my moblie phone rings , i take the call , its urgent im needed else where asap , i get up and rush off , forgetting to pick a card from the deck .Question ...is the phone call and me being need else where and forgetting the cards a RANDOM result of me attampting to pick a card from the deck ..??is picking no card a result , or is the randomness locked into the deck and only when a card is picked is randomness in action ??

This is an interesting view. Though I'm not sure what you're asking, or what your opinion is.

quote:

---

Question ...is the phone call and me being need else where and forgetting the cards a RANDOM result of me attampting to pick a card from the deck ..??

---

I think you're confusing cause and effect. Sure the phone call came after you decided to pick a card, but I wouldn't say that the phone call was a "random result" of the card-picking.

quote:

---

is picking no card a result

---

Sure, why not?

I would say 'random' refers specifically to an outcome that is purely determined by chance alone.(The 'alone' is an important word here)Modulous' reply is a good one on the genetic signficance.
The dice example is perhaps potentially missleading because the outcome is determined from a very limited set of equi-probable alternative states. Randomness need not refer to a range of outcomes that are 'equally probable', but simply an outcome of an undirected process. Yes, if a 6 is rolled on a fair dice it is in some sense a random event, but a bird crapping on your head while walking thorough a field is likely also a random event (assuming no grudge on the part of the bird), albeit a much lower probability one. Thus both high and low probability events can occur randomly, but they won't be observed with the same ”frequency’. All dice rolls will be observed with equal frequency, even if the outcome of any single roll is random.Scientists can go to great lengths to achieve 'true' randomness in things like experimental design and it can be difficult to do because of the constant possiblities of unconscious bias affecting outcomes. Hence the best approach is usually to let a computer generate random numbers that can be used to randomize things like sets of replicates in an experiment.Often in biology, we can find out something is going on simply because results are not 'random'.
In these cases, a random outcome is the null hyptothesis and its failure to be supported by data is evidence that something is actually directing the outcomes in our experiment.An Example.I have 20 different varieties of wheat plants and I grow 1 plants of each in a big tray of soil. (I live in Kansas, ergo I grow wheat)
I plant 10 trays for sake of replication and I use a random-number generator to determine the exact placement of plants within each tray. (I want a ”random arrangement’ in each replicate, so that all replicates are not the same).
Next I place each tray in its own cage and I release 100 winged aphids in each cage.
(I like to play with aphids too )
I come back in 24 h and I count the number of aphids on plants of each wheat variety.
Here is the question. How do I know if the aphids land at random on the plants?Now I must assume ”randomness’ and test for departures from it.
The Poisson Distribution can be used to predict the exact probabilties of 0.1,2,3,4,5 etc. aphids on a single plant, assuming that aphids distribute themselves randomly. The numbers landing on plants of each type can be compared to this distribution to see if they depart from it significantly.So let’s say I get an average of 13 aphids on variety X, averaged across replicates. Could this be a ”random’ result, or is something else going on here?We first need the mean number of aphids per plant experiment-wide (let’s just say it’s 5) so we can look up a critical value in our table of Poission terms. I can then see that the probability of avaeraging 13 aphids per plant is only 0.0013, or only 13 times out of a thousand. So I am inclined to think maybe the aphids like something about variety X and are not responding randomly.(Of course what I would actually be looking for would more likely be a variety with fewer aphids per plant than expected by chance, but the same test works for that too.)Hope this helps your understanding of how scientists actually use the concept of ”randomness’.

quote:

---

Hence the best approach is usually to let a computer generate random numbers that can be used to randomize things like sets of replicates in an experiment.

---

Remember, a computer cannot generate a random number. All "randomly" generated numbers are created using a seed, which is just a starting number. (I usually use the internal clock value). Give it the same seed and it will always generate the same sequence of "random" numbers.Your Aphid experiment is interesting, and would create usable data. The problem is that any randomness that might occur is simply the result of not knowing the precise "mindstate" of the aphids, the appeal of each leaf of each plant to the aphids, the air temperature and who knows what else, etc, etc. My point is that your experiment just falls into my category of

quote:

---

...the culmination of simply too many variables acting on something to make easy predictions

---

In the OP, I suppose what I was really wanting to know about was the nature of Quantum mechanics and their (apparently truly) random nature.Chiroptera, 1.618, do you realize what this fact of quantum mechanics means? It means that there are things that happen without a cause at all. It means that at every picosecond, all matter in the universe is being manipulated by an infinite number of first causes.Ugh, Even though this has been proven by people much smarter than me, it just does not sit right with me. It never will.

I think this is disputable. At least, for experimental purposes, a computer-generated system of numbers is a valid way to avoid bias. Completely irrelevant. My point is that we can define and characterize behaviors that are definitely non-random, making a statistical definition of randomness useful in the demonstration of its absence. And as far as evolution is concerned, the 'mindstate' of the aphids is not directly relevant either - if you refer to their thinking about why they do what they do. They behave the way they do because of response or non-response to stimuli. If they respond to stimuli in a non-random way, it can be useful to know that. Now rememeber, responsiveness to stimuli has nothing to do with 'mindset' in terms of volition - insects simply respond to stimuli because their ancestors did, because these individuals were successful enough to have descendents, not because they make a conscious decision. I have news for you. 99% of experiments fall into that category to some degree. Science normally advances in small steps. If you can show that aphids choose plants in a non-random fashion you demonstrate they are responsive to stimuli particular to those plants that are either over- or under-colonized in the experiment. That is grounds for asking what these stimuli might be. The intra-plant variation effects and environmental factors you refer to are either controlled for in the exp. or can be factored out statistically because the experiment was replicated 10 times.The art of science is twofold: designing experiments to yield unabiguous results w/r/t critical variables, and then conceptualizing adjustments to theory in order to account for deviations from expectation in the observed data, assuming it is confirmed through replication. Ergo, theories evolve. Then maybe you should have been more specific.
QM does describe entirely random processes by delimiting them with probability theory. Bingo. Big light just come on? That is what we are trying to say about genetic drift, mutation, experimental error, etc. Science has to account for chance results all the time, and some branches of science deal with phenomena so unpredictable they are only definable via probability theory - the 'theory of predicting random events', if you will. Is any of this getting any clearer yet? This statement does not follow. It is overtly telological and in direct contradiction to what you said before it: Which is likely the smartest thing you have said yet.
Ever consider that you might be seeking directed processes where there are none?This message has been edited by EZscience, 04-19-2006 09:17 PM

I think you are confusing Drewsky's point, and to be fair to Drewsky I don't think it was at all unclear. I have just been to busy (lazy) to get round to addressing it properly.Probability theory has nothing to do with this, in that it can be applied to both situations that Drewsky is trying so hard to differentiate.Drewsky is trying to ask "are there events in the universe that are truly random?" which translated means - are there effects that cannot be traced back to a prior cause - "or are there only events which are unimaginably complex to predict based upon the number of variables involved and hence regarded as random"As has been mentioned by several posters, QM is the "obvious" but also only example of the former and the latter needs qualifying with examples of chaos.However, it's not quite as simple as that. I'll give my answer in reply to Drewsky's OP...

I get the feeling that "random for all intents and purposes" is not what you are interested in, and so we can forget complexity and chaos.Does quantum mechanics introduce "randomness"? Not really, no. Quantum mechanics has historically been split into two processes. Evolution of the wavefunction, and measurement.The former is the entire cause and effect of QM, and it is completely non-random. The latter is where we get so-called "collapse of the wave-function". It appears that this is a random process governed by probability, but these days we tend to think of it as just a continuation of the evolution process as the wavefuntion interacts with all of the wavefunctions of the measuring apparatus including the human observer if present.And so even this "random" element falls into your category of "too complicated to know so call it random". I know it has been mentioned that Bell's Theorem demonstrates an underlying randomness but this is a misunderstanding. It demonstrates an underlying entanglement but this is a property of the wavefunction, not of "collapse", and hence exists totally in the categorically determinstic regime.For a long time Hawking was convinced that there was an intrinsic randomness to quantum mechanics as the mathematics of Hawking Radiation seemed to show this, whereby information falling into a black hole would be re-emmitted in a purely thermal (random) state. This would be an example of non-deterministic evolution of the wavefunction. Most thought that this simply an artifact of the approximate calculations, and several years ago Hawking conceeded the point...And so it appears that we live in a totally deterministic universe. But this may change...This message has been edited by cavediver, 04-20-2006 06:35 AM

In mathematics (probability theory), "random" is an undefined concept with some rules of usage.In real life, I think it best to treat "random" as an indication that we are dealing with something that could perhaps be modelled by probability theory. Probability theory is often used when dealing with uncertainty or unpredictability.As a simple example, consider what happens when you make a network connection on your computer. There are two port numbers associated with the connection. In most cases, one of these is fixed by the application. For example, a web lookup uses port 80 to access a web page. The other port is often referred to as a "random port". But if you check (use the "netstat" command), you will find that the random port tends to increase monotonically, and is actually determined by an operating system counter that is incremented for each connection (or decremented in some operating systems). Practically speaking, we use "random" to indicate unpredictability. The world is a complex place, with many sources of unpredictability. Whether this unpredictability is real or apparent doesn't much matter in most cases.My personal opinion: The world is composed of a large collection of independent actors (quarks, or whatever quarks and other things are made of). The apparent regularities we observe are either human constructs, or are statistical trends in the interactions of these independent actors.Note that my personal opinion is subject to revision at any time. The metaphysical question of whether there is an underlying determinism is one that could never be settled with empirical evidence.

Take it from the computer scientists who program computers, and the cryptographers who rely on random sequences of numbers - there's no dispute. Computers do not generate random numbers. How could they? How could a deterministic computational device generate a nondeterministic output?What they generate are pseudorandom numbers; sequences of numbers that are statistically undeterminable from random numbers. That is to say - given a sequence of random numbers, and another sequence of pseudorandom numbers from a computer or another source (like a page from a one-time pad), there's no test you can perform on the sequences themselves to determine one from the other.Like I said, there's no dispute about this in the field of computer science and electrical engineering. The algorhythims used by CPU manufacturers to generate pseudorandom output are closely-guarded trade secrets, because if you could crack the algorhythim you'd be able to precisely predict what sequence of numbers a given computer would return from an input, and that would have extreme ramifications in the field of cryptography.I heard once in a CS course that Sun Microsystems used the output of a digital camera pointed at a trio of lava lamps to generate truly random output. That's basically what it takes.

OK. I can accept all that.
But for empirical purposes like my example, the use of such numbers to randomize experimental replicates is valid given they are 'statistically undeterminable' from truly random numbers.But we seem to have various concepts of 'random' emerging in the thread.1. a simple lack of predictability - the probability theory definitition (problem: system could contain directed processes or limit functions without our ability to discern them, e.g a chaotic attractor).2. a system whose behavior is uncertain because it does not appear to be directed by any underlying processes (problems: various types of regularity can emerge spontaneously in undirected systems causing them to display apparently non-random behavior. Also, our failure to recognize the influence of any underlying processes can never guarantee their non-existence)3. a system that is endogenously 'random' by virtue of immunity to all forms of external influence. (problem: not sure there is evidence to support the existence of any such system.)

Not everything that happens in a computer is deterministic. The time when a disk drive interrupt occurs, the time when a mouse interrupt occurs, the time when a keyboard interrupt occurs - these have the appearance of having a random component.In a number of operating systems, the kernel accumulates these apparently random events into an entropy pool. Then this pool is made available through a virtual I/O device ("/dev/random" in unix systems).