• Quick note - the problem with Youtube videos not embedding on the forum appears to have been fixed, thanks to ZiprHead. If you do still see problems let me know.

Evolution: Technically Random?

If some website listed dolphins under the heading of 'Fish', would you conclude that they're piscean?
 
Last edited:
Melendwyr said:
If some website listed dolphins under the heading of 'Fish', would you conclude that they're piscean?
Click to expand...

If.

Look, I don't dispute what you said. I just don't think it has serious bearing on my argument either way.
 
What's left of your argument? Evolution is not a random process because its results are not random relative to the environment. If you would like to say it's a process with a random component, by all means. But when you say it's a "random process," people misunderstand you.

~~ Paul
 
Paul C. Anagnostopoulos said:
What's left of your argument? Evolution is not a random process because its results are not random relative to the environment. If you would like to say it's a process with a random component, by all means. But when you say it's a "random process," people misunderstand you.

~~ Paul
Click to expand...

I don't think I've been misunderstood at all. Non-trivial functions of random variables are random variables. People seem to get that.

So if evolution can be reasonably described as a function (NS) of things, some of which are random (RM), then describing evolution as random is not wrong.

Not random relative to the environment... Not exactly sure what your claim means. If it is random at all, there is randomness, hence describing it as random is accurate. Why does it matter what the randomness is relative to?
 
I don't believe it is reasonable to consider the Earth's environment as a random process. It is not merely a family of random variables. Because evolution is constrained by the environment (selection pressure), it also is not a random process.

There are random components to evolution. But to call it "random" is tantamount to calling it a "random process," which it is not. Unless your intention is to confuse, I would not call evolution random.

~~ Paul
 
If someone heard you say "evolution is random", and knew nothing of the theory of evolution, then you would have given them a thoroughly incorrect picture of evolution: for they would take "random" to mean indifferent between outcomes, which evolution is not. At this point they might do one of two things:

(1) If they were sensible, they'd start asking how a "random" process can account for, say, the streamlining of fish, rather than making them in random shapes, and then you'd have to explain what you really meant, and tell them about natural selection.

(2) If they're some ignorant fundie, they'd grunt out agreement and laugh at the stupid atheist scientists who cling to the absurd dogma that "random evolution" can account for the appearance of design.

So I can see no point to saying "evolution is random" except to deceive and confuse the ignorant and to annoy the knowledgable.

It would be accurate to say that evolution is stochastic, and more accurate still to say that evolution is Darwinian.
 
I don't think you can reduce evolution to a simple mathematical formula. Mutations occur and the bad genes persist in the population though if they are bad enough and dominant then they may be eliminated. For non random evolution to occur there needs to be selective pressure on the population and still the bad genes may be preserved especially if they are recessive genes since there is no pressure to eliminate them due to lack of an effect when they are present heterogeneously.
 
Dr Adequate said:
for they would take "random" to mean indifferent between outcomes, which evolution is not.
Click to expand...

They might indeed do that, but that would be them getting confused on what random means, not me getting confused on what it means. :)

It would be accurate to say that evolution is stochastic
Click to expand...

Stochastic means random.
 
Paul C. Anagnostopoulos said:
T'ai, you're just being stubborn.
Click to expand...

Paul, there's no need for even tame name-calling here.

Do you agree that evolution is not a random process?
Click to expand...

I agree that natural selection is not random. Evolution, which is natural selection acting on things, some of which are random, can justifiably be called random.

You have said "There are random components to evolution.". If something is a (non trivial) function of random components, it itself is random.
 
Paul C. Anagnostopoulos said:
T'ai, you're just being stubborn. Do you agree that evolution is not a random process?

~~ Paul
Click to expand...
The way I look at it it is both random and not random. Randomly mutations occur and become part of the genome of a population. In addition randomly the genome of a population shifts without any selective pressure on it. However with pressure certain genomes are favored and become more prevalent (this I would consider non random). Also new beneficial mutations may be favored and increase in an non random manner without selective pressures.
 
T'ai said:
You have said "There are random components to evolution.". If something is a (non trivial) function of random components, it itself is random.
Click to expand...
It would have to be a function of random components only, but evolution is not. If what you say were true, then Ev would not evolve an increase in the information content of the genome:

http://www.lecb.ncifcrf.gov/~toms/papers/ev/evj/

~~ Paul

Edited to add: In fact, any simulation involving some stochastic factors would produce random results.
 
Last edited:
Paul C. Anagnostopoulos said:
It would have to be a function of random components only, but evolution is not.
Click to expand...

I don't believe so. With a (non-trivial) function, as long as it has some random components, the output is random.

Consider this toy example. I create a sequence for 100 'generations', as

f = rand(U) + t

where f stands for the fitness of some population of creatures or whatever, t is the time period, in this case from 1 to 100 years, and U is a standard normal random variable (that is normally distributed between 0 and 1).

(and no I don't think this toy example has any reality to it as accuratly describing evolution, nor do I know of any, just an example to illustrate the main point)

If I simulate and make a graph of f vs. t, it is definitely random but there is a structure, a basically 45 degree line. If you 'zoom in' to it, we see many ups, many downs.

If you ask me what f(100) is, I'd bet it is darn close to 100, but I am not able to predict f(100) with certainty because f is a random variable because of the U. From one simulation I get 100.34. Same if we speculate on f(5000), or however many 'generations' we want to go out.

The t, the non-random part may dominate, the fitness trend may be in general predictable, but the specific output is random.

I believe any (non trivial) function you give me would have, when zoomed in, these ups and downs.

I looked at that site, and I don't see how it contradicts anything I posted. They start with random sequences (the "U") and have some evaluating functions (the "f"). Over time a trend emerges, yet the outcome is unpredictable with certainty from the start.
 
Last edited:
Ah, so if the output takes a random walk around nonrandom values, you're calling the function random. That may be technically correct, although I thought a random function was more specifically defined than that. I'm fairly certain that a "random process" involves only a family of random variables.

Anyone know?

I've lost track of the point of this conversation. Why is this important?

~~ Paul
 
Paul said:
Ah, so if the output takes a random walk around nonrandom values, you're calling the function random.
Click to expand...

Calling the output random.

As far as I know, X is a random variable if X has a probability distribution.
 
Well, "random output" is a loosely defined term.

I looked up the definition of a stochastic process:

A random process. Common usage excludes essentially deterministic processes that are subject only to random errors.

Ain't that interesting? From the viewpoint of the organism, evolution is not a random process. From the viewpoint of the environment, it is.

What is the purpose of honing this fine a point on the issue?

~~ Paul
 
You must log in or register to reply here.

Back
Top Bottom