When you use the term orderly--you are talking about "something" (a force--the physical environment) that imposes order, right?
No. The order arises merely from the elementary constituents and the probability distribution function of their interactions.
Stuart Kauffman has an excellent example in
At Home In the Universe. I'll review it here as briefly as possible.
Take a box of buttons. Dump them on the ground. Mark the position of each button; you're going to be picking them up and putting them back where they came from.
Pick two buttons. Connect them with a piece of thread. Put them back.
Keep doing this.
Now, how does the size of the largest network of interconnected buttons vary with the ratio of the number of threads to the number of buttons?
One expects a relatively smooth upward trend. But that is not what one gets. What one gets is an S-curve; the size of the largest network varies not much (stays under about 10% of the buttons) until the number of threads is a bit short of half of the number of buttons; suddenly, each string you connect after that makes a larger and larger network, and this continues until just a bit more than half. at which point the size of the largest network is very close to all the buttons (90%). After that, the size of the largest network again doesn't vary much as you add buttons.
So between 45% and 55%, the size of the largest network goes from 10% of the buttons to 90% of the buttons. This is the critical zone for all randomly-defined networks. This is a tautology of the topology; it doesn't matter what the buttons represent, or what the threads represent- in topology, the buttons are called "nodes," and the threads "edges-" all networks constructed in this manner will show this behavior. It's a mathematical fact, and it's manifest in the real world.
Now, the environment doesn't matter to this behavior. Nothing matters but that there be something that can act as nodes, and something that can act as edges. As long as this is true, this is the behavior you will see. It's not yet been proven
ab initio, but empirically, Kaufmann has done many experiments of this type, and the behavior is always seen, on average. And the larger the number of nodes involved, the more closely the system adheres to this behavior.
The same sort of mathematical separation applies for statistical analyses of ensembles of elementary particles. It doesn't matter what the particles are, it doesn't matter what the interactions consist of, it doesn't matter what the environment is (because the environment is assumed in such analyses to have separate effects); if you have a probability distribution at the individual interaction level, you will see statistical behavior of the ensemble. The only things that matter are, what are the possible interactions, and what are their probability distributions. This was partially shown by Boltzmann and Maxwell, and conclusively proven as a mathematical theorem in the Fluctuation Theorem, which shows the mathematical underpinnings of the Maxwell-Boltzmann statistics that result in the Laws of Thermodynamics.
THAT'S what I mean by orderly. The outcome is certain, because of the probability distribution of the individual interactions; and the more interactions, the more certain the outcome.
When complexity comes from randomness it is due to physical aspects of the environment that impose order--right? This would be true if you are talking about the evolution of the galaxy or the evolution of life, right?
No. It is not due to the environment. It is due to the probabilities of the individual interactions. If those probabilities are altered by the environment, then you must allow for that in the probability distribution- and this will alter the ensemble behavior.
But the
base behavior is the base behavior- the environment is added in. And no environmental influence is necessary to see the complexity of ensemble behavior emerge from the chaos of individual interaction.
So how does this fit with biological evolution? Well, first, let's consider that if there were no environmental considerations, then not all genomes would be viable. Some genetic changes would interfere with an organism's ability to simply live. For example, if the genes that create the heart of a vertebrate during ontogeny are damaged, the animal will not live in any reasonably likely environment; or if the genes that allow the creation of the machinery of transcription are damaged, the embryo will never even get started. These types of mutations are universally fatal; it doesn't matter what the environment is. The question here is, is this a viable organism? Can it simply live? And all that complexity comes simply from the genes. It would not matter whether there was an environment or not; given sufficient time, we would see all viable lifeforms capable of procreating, if there were no environment. The processes of mutation and recombination would guarantee it.
That, I suppose, would be "random" as you seem to me to mean it.
Now, if we add environmental influences, then adaptation to the environment becomes a factor. Furthermore, we add competition, cooperation, predation, and all those other things that cause the organisms to coevolve; and asteroid strikes and Milankovitch cycles and all those other things that cause the environment to change separate from the effects of other organisms. And because of this, it would be incorrect to state that "evolution is random;" only those organisms not only simply viable and capable of reproducing, but those capable of surviving in this environment are now replicated.
Now, as far as I can tell, you mash these two sets of considerations together. But to a physical scientist, they are separate; the division may be hazy, but there are certainly regions where it is clear that
this mutation will affect the organism's viability or ability to reproduce no matter the environment, where as
that mutation will affect only the organism's interaction with the environment without affecting its viability or ability to reproduce.
Therefore, the complexity does not need the environment to manifest. The complexity is not the result of the environment; instead, the environment is a
restraint upon the possible organisms that are capable of surviving to reproduce, and the complexity is
reduced by its influence.
Would you define this "phenomena" that produces order as "random"?
Yes.
What I want to know is in physics and statistics would it make sense to define the "force" that brings order in terms of randomness? Or to call it random?
It is an inevitable consequence of the fact that the individual interactions are constrained by probabilities. If they were not, then we would see true chaos; but these constraints at the level of the interaction have inevitable effects on the ensemble behavior, and that leads to complexity. Without a great deal of mathematics, or empirical investigation, these complex ensemble behaviors are often difficult or impossible to predict. They are, however, ubiquitous, if not omnipresent.
Calling a system showing this type of complex, constrained behavior "random" in the ordinary sense of the word is wrong; that is, in the sense of "disorderly." It is not disorderly at all; anyone looking at the ensemble behavior will tell you it is not.
Is it meaningful to define that which brings order to the randomness to build complexity as being random in itself?
No, but selection is not the force that builds the complexity. The complexity is inherent in the probability distribution functions of the underlying interactions; given an ensemble of these interactions, we can predict the behavior, albeit not without a great deal of effort.
underlying We all agree that order can come from chaos and randomness. But as to the aspects that facilitate this order--how do you distinguish the random components in this "order bringer" from the randomness it acts upon? Or do you? [/quote]Mathematically. It arises of itself, not as a result of anything but the underlying probability distribution functions and the kinds and characters of the interactions.
How would you or how did Dawkins' communicate this so that you understood the difference between randomness that is acted upon and the randomness involved in the forces that cull from (or bring order to) the randomness?
I'd need to go back and read his description to fully recall; I cannot tell you for certain whether he differentiated between these two, but to the best of my recollection he did not.
Because that is the answer to the question isn't it. I really don't think I'm missing any of the definitions of random or why one could call evolution random.
I think you have missed the distinction between order that arises simply from the underlying probability distribution functions, and order that arises from interaction with the environment; and while the first is implicit in the general conception of "random" as used in the physical sciences, the second is not.
I'm really just trying to show why biologists think it's misleading and uninformative to categorize natural selection (the order imposer) with the same terminology as used in describing mutation (which is not even truly random in the strictest sense of the word.) Biologists do this because the many definitions of random make such phraseology ripe for abuse. It makes people go off on tangential semantic circles instead of just understanding the basics. And truly, it's this confusion that makes people think "evolution is impossible". But evolution is easy to understand. Finding the right terminology to convey it's basic premises without opening oneself up for semantic shenanigans is difficult as this thread beautifully illustrates.
I think that you've totally missed the point; I can't say much better for myself, until this post. We've been talking at cross purposes; but that's more my fault than yours. I can only plead that I had not considered the detailed underpinnings enough to realize fully that what I was talking about was not due to the environment.
What you need to understand is that this is the way that physical science thinks about randomness, so to say to a physical scientist, "evolution is not random," is totally non-informative. You are (and I was) lumping the environment together with the ensemble behavior, and that's wrong. That's not how it works. I think this is why I and others said biologists don't understand what random means; I think it's a common error.
Surely there must be a way to convey "that which brings order" so that it is not confused with that which it is bringing order to.
But which "that" are you referring to? The low-level constraints (probability distribution), or the high-level constraints (environment)?
A way that satisfies even the randomites--but is clear...and not subject to the "philosophy of science" and semantic obfuscation.
It is, as I said, a matter of terminology, and a complete understanding of the situation.
In any case, I hope the randomites understand how difficult this particular task is... From a biologists perspective, calling evolution random fails to convey that which brings order to the randomness--it allows for the creation mischaracterization of evolution and makes the concept of complexity from chance look unfathomable and an "intelligent designer" seem more plausible. Whatever words you use to describe evolution...if someone wants to know about the non-random parts or how the order "evolves"--use whatever words you would use to describe how complexity evolves from any sort of randomness.
Mmfff. I have to get back to work. I'll deal with this later.
