|
Author
|
Topic: Different kinds of 'complexity'
|
Evan
Member
Member # 164
|
posted 28. April 2003 12:51
I am sorry the moderator feels that “stereotypes and 'worn-out' discussions” have crept in here. Perhaps my response to the comment about common descent contributed to this, and for that I apologize.
There is a key issue here that I think should be focused on, and that I hope doesn’t fall into the above categories: that of probability calculations that are, respectively, ahistorical and historical.
Thomas summarizes well when he writes,
quote:
Interesting would be calculations for probabilities of organisms, having a DNA of that and that sequence at generation 0, having no flagellum to get a DNA of that and that sequence at generation 0 + n (conceded that usual calculations stating an impossibility for n = 1, i.e. a saltation, are valid and would imply a designer) coding a flagellum....
My point is that IDists still yet have no valid argument for stating an
impossibility for an evolution of 'type b'-systems, based on calculations for systems of 'type a' or regarding just one (or very few) generations of 'type b' systems.
The issue here is not one of design vs. non-design: the issue is whether the event in question (type a) was immediately and completed arranged as if by pure chance (which would, as Thomas says, imply both a designer and a mode of design - saltation or immediate creation) vs. an event (type b) which became what it is through a set of intermediate steps (which may or may not have involved intelligence agency.)
Thomas’s arguments is that pure combinatorial probability only pertains to type a events. Calculations that show that such events are highly improbable show that if things happened that way, design can be concluded.
If if things didn’t happen that way - if a series of steps was involved, then such calculations are irrelevant irrespective of whether design was involved or not.
So unless there is evidence that spontaneous and immediate creation of things actually takes place (and I think design would be a reasonable interpretation of such an event), purely combinatorial calculations don’t help resolve the issue.
IP: Logged
|
|
gedanken
Member
Member # 594
|
posted 28. April 2003 14:52
This is related to the logical fallacy of "bifurcation". A couple of possibilities are shown, one is then shown to be unlikely -- and an inference is claimed for the other. But this is not logically sound. The third possiblity here is the sequence of steps. That needs to be analyzed properly to even approach a sound probabilistic methodology -- if any is possible at all.
IP: Logged
|
|
Chronos
Member
Member # 693
|
posted 28. April 2003 21:19
Perhaps I can contribute to this discussion with a line of reasoning that is a bit different.
I have observed that naturalists, in my humble opinion, tend to base their conclusions on something less than the solid premises that science has contributed to the subject of complexity.
As example, it seems that naturalists, when considering the complexity that they posit results from macroevolution via common decent, tend to hypothesize that complexity yields higher complexity. Had common decent occurred, it would have necessitated that each, or most, speciations yield complexity as higher order forms evolve from lower order forms, i.e., mammals from reptiles. So complexity---->higher complexity---->even higher complexity, etc.
Darwin had his view on the subject. He proposed that one of the primary characteristics of evolving organisms is “complexity of structure.”
But the studies of chemist Ilya Prigogine in his work in far from equilibrium, open systems, suggests just the opposite occurs in nature in reality.
The order of complexity can only form from a chaotic system. And the more chaos that is in a system, the more that complexity likes it because entropy is lower allowing order to form. Prigogine mused that his dissipative structures (things that order in open systems, such as crystals growing, etc.) can form only when entropy is very low in the system and chaos elevated. The closer a system gets to order, the closer it is to perfect equilibrium and the less chance that complexity can result from it.
As example, many physicists postulate that our order will be maximum when our universe goes to maximum entropy and dies its heat death where everything in the universe is nothing more than a random sea of floating atoms. How could complexity ever form from maximum entropy? But complexity could easily have formed right after the big bang when entropy was minimum.
I enjoy the musings of Prigogine and here is a line of reasoning from him in my words: Chaos seems to be the moving mechanism by which nature develops constrained and useful randomness; as there can arise no order unless also is present a flow of matter and energy which decreases in intensity as a system moves toward equilibrium. This disordered flow is our source of order.
Thus I think we can conclude that science is on the side of the IDist when we observe: chaos-----> complexity, higher chaos -----> higher complexity, even higher chaos -----> even higher complexity.
Again, Ilya Prigogine observes: “We begin to have a glimpse of the road that leads from being to becoming.”
IP: Logged
|
|
Evan
Member
Member # 164
|
posted 28. April 2003 22:54
Chronos writes, “Perhaps I can contribute to this discussion with a line of reasoning that is a bit different.”
I’m afraid I don’t think Chronos’s remarks have much bearing on the discussion here, for two reasons:
1) The issue here, as both Thomas and I have pointed out, is not design vs. non-design, (or “naturalists” vs. IDists): it is whether things come together into some specific form in one fell swoop, so to speak, so that combinatorial probabilities are relevant (Thomas’s “type a”), or whether things come into their specific form via intermediate steps - that is, have some history (Thomas’s “type b”).
ID hypotheses of both type a and type b seem to exist. (My original foray at ISCID back in March of 2002 offered an ID hypothesis of type b, and Dembski seems to point to such in his essay “ID Coming Clean,” where he specifically disavows the type of “miraculous” interventions necessary for a type a event.)
So we’re not talking “naturalists” vs “non-naturalists” here. Both naturalists and IDists who believe that the designer(s) influences the world in a "type b" fashion might be interested in the topic of calculating the probabilities of a chain of biological events leading up to the existence of the feature in question.
2) The confusion about the definition of complexity again arises here. Chronos writes,
quote:
It seems that naturalists, when considering the complexity that they posit results from macroevolution via common decent, tend to hypothesize that complexity yields higher complexity. Had common decent occurred, it would have necessitated that each, or most, speciations yield complexity as higher order forms evolve from lower order forms, i.e., mammals from reptiles. So complexity---->higher complexity---->even higher complexity, etc.
Notice that this use of the word complexity is not in keeping with the definition used by Dembski, and under consideration in this thread: that definition being the reciprocal of the probability of something happening by natural causes. Chronos is referring instead to the more “commonsense” version of complexity - having more interactive parts working in a more sophisticated fashion.
But these two definitions are not equivalent. We are discussing complexity in the Dembskian sense, not in the 'common" sense. (In fact, there is no reason why complexity in the improbable sense might not be involved in the design of something that was fairly "simple" in the other sense. Natural processes seem, at least according to mainstream evolutionary theory, to sometimes build sort of jury-rigged systems which might be more complex than something that was designed to get the job done more expediently.)
Furthermore, the “naturalist” (who, as I have pointed out, is not really of concern here) would probably deny Chronos’s point in respect to either definition. If you use Dembski’s definition, he would say that complexity does not in fact exist, because the given organisms and their various parts are not improbable. However, he would probably also deny that evolution posits an inevitable ever increasing complexity of the other sort. Yes, we do see that some “more complex” creatures have arisen over time, but we also see that less complex creatures (bacteria, viruses, termites, etc.) are doing just fine, thank you.
So I don’t think that Chronos’s entropy/thermodynamics issues from the other thread do add much to the issues being discussed here.
[ 28. April 2003, 23:23: Message edited by: Evan ]
IP: Logged
|
|
Chronos
Member
Member # 693
|
posted 29. April 2003 01:38
I was responding to the initial opening post where both of the tenets I mentioned were brought up:
“IDists use complexity in different forms to show that there must be a designer. I find it useful to differentiate two kinds of 'complexity'.”
And: “The open question, if 'descent with modification' is able to generate 'true' novelties is no argument for any validity of considerations based on a) for b).”
The paradigm that I sought to add to the discussion was to expound on that last question, not really to contrast ID with naturalism although this contrast is certainly a tenet of logic that accompanies my musings on that question.
Complexity has been shown to arise from chaos, not other complexity. This is my point. Therefore I feel that to surmise “The open question, if 'descent with modification' is able to generate 'true' novelties” is a settled issue in the field of science. We simply do not find complexity springing from complexity. Instead we see complexity arising from simplicity and the more simplicity that is in a system the more complexity that can arise from it. Stalagmites do not form from other complex stalagmites, they form from the simplicity of a simple drop of water containing calcium.
Perhaps a concise and succinct one line definition of complexity would be in order here at this time. If I’ve missed it in the thread then please point me back to it.
IP: Logged
|
|
RBH
Member
Member # 380
|
posted 29. April 2003 02:17
Chronos wrote quote:
Perhaps a concise and succinct one line definition of complexity would be in order here at this time. If I?ve missed it in the thread then please point me back to it.
Well, while the OP didn't define it, Evan invoked Dembskian "complexity" in the second post of the thread. Ryan picked that up, and the thread has been mostly about the probabilities of which Dembskian complexity is the reciprocal and how they're appropriately estimated, since that was the OP's thrust.
Then there's the thread Micah started on Types of Complexity. Given that thread and the references therein, there's something like 45 or 50 definitions to choose from. ![[Smile]](smile.gif)
While I don't have time to pursue it now, I have to say I'm not in agreement with Chronos's assertion that "complexity" (however defined) can emerge only from simplicity, not other complexity. There's a level(s) of analysis issue there that if I had a few hours I'd explore. But the thrust would be that what is complex at one level of analysis may be "simple" in the sense required at another. I'd look at it in terms of Herbert Simon's "Sciences of the Artificial" analysis. However, that belongs in a different thread.
RBH
IP: Logged
|
|
Chronos
Member
Member # 693
|
posted 29. April 2003 02:26
"Then there's the thread Micah started on Types of Complexity. Given that thread and the references therein, there's something like 45 or 50 definitions to choose from."
Exactly my point and I feel I could add a few more definitions. But I'm satisfied with Erik's assertion that the thread is discussing Demskinian Complexity. But we ought to be able to offer a one or two line concise definition of this, as well.
IP: Logged
|
|
Rex Kerr
Member
Member # 632
|
posted 29. April 2003 14:52
Wolfram's book A New Kind of Science demonstrates that simplicity or complexity or randomness (in initial conditions) can breed simplicity, complexity, or randomness (in final conditions), in pretty much any combination.
In all of his results, the underlying algorithms (if not the results and initial conditions) are simple, but they can be made complex simply by composing them.
Therefore, upon observing complexity, I don't think it is safe to assume what the source is. Rather, we must check in detail.
This is also a cautionary note to those who wish to engage in Dembskian complexity / low probability calculations. If you get your simple rules wrong, you can predict that complex outcomes are unlikely when in fact they are not (or vice versa).
IP: Logged
|
|
|
Printer-friendly view of this topic