From: Brian D Harper <bharper@postbox.acs.ohio-state.edu>
>This post pertains to my recent discussion with Richard.
>Since Richard indicated he will be out of town I have
>sent a copy to his e-mail address.
Thank you, Brian. I'm back from my holiday, well-refreshed and ready to
re-enter the fray. ;-)
>I've decided not to respond point by point to Richard's
>last post as I felt this would just yield further convolutions
>without returning back to the point of purpose.
The main point of my last post in this thread was to get you to drop the
issue of "algorithmic randomness", which I felt was a red herring. I will
now consider that issue dropped. ;-)
>This is an important issue for me because I feel that
>there are certain people who manage to get Christian
>lay persons stirred up and then actively opposed to
>science by telling them that evolution reveals a world
>without purpose meaning or design. Now, if this really
>is the result of the science of evolution then this is
>justified. If it is instead certain scientists claiming that
>their theological speculations are actually scientific
>conclusions then I think its high time for some corrections
>to be made. One problem is, of course, that the most
>vocal elements on both sides of the evolution/creation
>debate agree on this point.
>
>My position is that there are no technical, objective
>definitions of randomness that say anything about
>purpose meaning etc. This being the case I am quite
>happy to pursue the definition that you provided.
I don't claim to have given a precise definition of randomness. I've
attempted to clarify what I mean by it, and my ideas have changed a little
as the discussion has proceeded. I will have another stab at clarifying my
position below. If you have a good definition of randomness available, then
I'd love to see it.
>So, let's define a random process as one that involves
>at least one random element.
I wouldn't call this a definition of a random process. It is simply the
criterion I would use to distinguish a random process from a deterministic
one. More on this below.
>Now I would like to go
>back to a statement you made in an earlier post in
>this thread:
>
>======================================
>"So, if a designer has been manipulating the events of evolution to produce
a
>desired result, then evolution is not a random process. On the other hand,
>if a designer set up the evolutionary mechanism and then let it run without
>interference, then the process of evolution is random, and is without
>purpose, plan or design." -- Richard
>=======================================
I'm not sure it's helpful to go back to such an old post, when we've already
moved on since then. I didn't state my position very clearly in that passage
(although it's clearer when taken in its original context), and I've already
clarified my position in some respects. But let's continue...
>Let's consider the first part "So, if a designer has been
>manipulating the events of evolution to produce a
>desired result, then evolution is not a random process."
>
>Suppose we have a case of random mutations giving
>rise to variations but with artificial instead of natural
>selection. I suppose we could just consider a dog breeder,
>but I'll consider a somewhat more powerful designer.
>This designer monitors all living things and selects those
>individuals (or species) which survive. The designer selects
>only individuals (or species) that are fit enough to survive
>on their own without further intervention. Otherwise, the
>selection is sometimes whimsical. The designer rejects
>one very fit individual because it has ugly ears. Another
>marginally fit individual is selected because she thinks
>it is cute.
>
>So, here we have a designer who is manipulating the events
>of evolution to their own purposes. Yet the process has
>a random element and is thus a random process.
When I stated that I would consider a process to be random if it included
any random element, I believe I was talking about distinguishing between a
deterministic process and a random process (I can't remember the exact
context). But, in the passage you quoted above, I'm talking about
distinguishing between a random process and an intelligently directed
process (or one that includes elements of both). To resolve this confusion
(mea culpa), I must first deal with some fundamental issues.
I propose that there are possibly as many as three categories of event, with
respect to type of ultimate causation:
1) Deterministic: the event has only one possible outcome, fixed by the
conditions prevailing at the time of the event.
2) Random: multiple alternative outcomes are possible, but the probability
distribution of these outcomes is fixed by the conditions prevailing at the
time of the event. (Possible example: radioactive decay.)
3) Directed: neither of the above applies, and the outcome of the event is
determined by the free will of a conscious being.
Depending on your view of nature and free will, these three categories might
collpase into two or even one. I believe Chris Cogan's view is that all
events are ultimately deterministic in nature, and that randomness and free
will are only apparent. That question is beyond the scope of this thread, so
I allow for all three possibilities.
I think of these categories as being a kind of hierarchy in terms of
significance, with deterministic events being the least significant. So I
refer to a process as deterministic only if it includes only deterministic
events, while a random process may include both random and deterministic
events. If a process includes any directed events, then I consider it to be
a directed process (but it may include deterministic and random events too).
This was the sense in which I used "random process" in the passage you
quoted above. In this sense, your example above is a directed process, not a
random one. However, I accept that this is not a standard sense, so your
confusion is understandable. ;-)
Statisticians are not generally concerned with ultimate causes, but with
whether the outcome of an event is predictable from the point of view of an
observer. An event which is unpredictable is considered random, regardless
of whether that unpredictability is inherent in the event or is merely a
result of the observer's lack of complete knowledge. I call the latter
"apparent randomness". In terms of apparent randomness, your example would
be considered a random process, because at least some of the events are
unpredictable by an observer.
So, if we're talking in terms of *apparent* randomness, then evolution is a
random process, even if it's intelligently directed. In other words,
apparent randomness tells us nothing about ultimate causes. I think maybe
this is the point that you've been making all along, and I agree with it as
far as it goes. However, the concept of randomness as used in the theory of
evolution is more than just apparent randomness. If, by saying that mutation
is random, we only meant that mutation is unpredictable by an observer with
limited knowledge, that would not be saying very much.
As I explained in my last post, I think that, in the context of the theory
of evolution, "random" really means "uncorrelated with utility". To remind
you, I wrote:
[start quote]
While dropping the red herring of algorithmic randomness will help, I think
we need to go further in clarifying what is meant by "random mutation". In
theory, a mutation event could be probabilistic but still biased in favour
of greater reproductive fitness, and this would not be in accordance with
the theory of evolution. Both you and Chris Cogan have brought up the point
that the random in "random variation" means random with respect to utility.
I agreed with this. However, on further thought, I don't think this is clear
enough. I think a better formulation would be "uncorrelated with utility",
i.e. the probabilities of different mutations are uncorrelated with their
utilities. (For the moment, I won't go into what we mean by utility, but
this is an important issue that I'll address later.)
This also clarifies the issue of selection. Natural selection *is* random,
as I've said before, because it includes a probabilistic element. However,
it is also correlated with reproductive fitness. So we can say that mutation
is uncorrelated with utility but natural selection isn't. This is a lot
clearer than saying that mutation is random but natural selection isn't.
So can we agree that, according to the theory of evolution, random mutation
is uncorrelated with utility, in the above sense?
[end quote]
If we can agree on this, then the question of what we mean by a random
process becomes unimportant.
>Now for the other case: "On the other hand, if a designer set
>up the evolutionary mechanism and then let it run without
>interference, then the process of evolution is random, and is
>without purpose, plan or design." -- Richard
>
>This corresponds to the design process that I discussed
>earlier. We must conclude then that a design process is
>"without purpose, plan or design."
I've already clarified my position on this point. I wrote:
[start quote]
I agree that there may be a purpose and plan behind the *result*. But I was
referring to the events making up the process. Each individual event is
unplanned and purposeless.
Perhaps the difference between us is whether we're talking about proximate
purpose or ultimate purpose. I suppose the random events, while individually
purposeless, are collectively contributing to the engineer's ultimate
purpose. So, in that sense, I suppose they could be said to have a purpose.
But I think that's a weak sense.
[end quote]
Finally, I've appended below a copy of my original post to this thread, the
one from which the passage in question is taken, so that it can be seen in
its original context, and because I think some of the material in that post
was quite useful, even if the concluding paragraph was not!
Richard Wein (Tich)
-----Original Message-----
From: Richard Wein <rwein@lineone.net>
To: evolution@udomo3.calvin.edu <evolution@udomo3.calvin.edu>
Date: 24 August 2000 12:25
Subject: Re: Piecemeal genetic differences as support for macroevolution,
etc.
>From: Brian D Harper <bharper@postbox.acs.ohio-state.edu>
>
>>As has been noted many times, random in this context does not mean
>>"without meaning purpose, plan, or design". Any "meaning purpose, plan, or
>>design" is not detectable with scientific instruments. For example, in
>>information
>>theory one measures the quantity of information in a message irrespective
>of
>>what it means. One would not conclude from this that messages have no
>meaning.
>>Also, consider the engineer that designs complicated mechanisms by
>mimicking
>>Darwinism, i.e. by random variations coupled with a selection criteria.
>>Would the
>>random variations be "without meaning purpose, plan, or design" in this
>case?
>
>I think we need to distinguish between design of the system and design of
>events occurring within the system. (I'll set aside the question of
>"meaning" which is rather different, and, for convenience, I'll lump
>together "purpose, plan and design" as simply "design".)
>
>I would say that "random" in this context *does* mean without design. If
>some designer is guiding the events of evolution, whether detectable or
not,
>then I say they are not random.
>
>To take your example of an engineer building a Darwinian system, I would
say
>that the system as a whole is designed, but the random events occurring
>within it are not. To take a specific example of human design, a roulette
>wheel is designed, but the outcome of individual spins is not.
>
>One further complication is that we need to distinguish between design and
>determinism. We may consider an event to be random even if it's outcome is
>predetermined, providing we have insufficient knowledge to conduct the
>determination. For example, it may be that a sufficiently detailed analysis
>of the roulette wheel would enable us to predetermine the outcome of the
>next spin. (Actually, a roulette wheel may be such a chaotic system that
>this is not the case, but bear with me.) We would still consider the spin
to
>be random, because we know that such a predetermination is not a practical
>possibility. On the other hand, suppose the wheel were rigged so the
>croupier can fix exactly what the next outcome will be, and does so. In
this
>case, we would *not* say that the spin is random, even though, from the
>punters' point of view, each outcome is still equally likely.
>
>So, if a designer has been manipulating the events of evolution to produce
a
>desired result, then evolution is not a random process. On the other hand,
>if a designer set up the evolutionary mechanism and then let it run without
>interference, then the process of evolution is random, and is without
>purpose, plan or design.
>
>Richard Wein (Tich)
>
>
This archive was generated by hypermail 2b29 : Fri Sep 29 2000 - 08:48:33 EDT