Probability-based arguments

I am new to EvC Forum, but not new to the debate. I have an idea for an interesting discussion that, in my opinion, is deserving from both viewpoints.I have seen many arguments which invoke some form of probabilistic reasoning to back a viewpoint.Example 1: This takes place over tens of millions of years, and is therefore possible (implying a number of trials over which random mutations may operate)
Example 2: The odds of a jawless fish becoming "jawed", with fully operating nerves, etc., are found to be 1:10^some big number, and is therefore not possibleEach of these seem reasonable on the surface, and they are repeated by camp residents over and over. However, there are a number of basic flaws.There is a vast difference in this discussion between our ability to model something _a priori_, based on first principles (thinking about a six-sided die), and our ability to model a process _a posterior_, based on observations of actual data, preferable of several different experiments (repeatedly rolling that die). Is the die loaded? Does it prefer certain long-term arrangements (is it non-random?)Since with the evolutionary discussion points, we have, at best, inferential supposition, neither modeling techniques is even marginally well founded. We have some localized data, but certainly not enough to calculate something as momentous as whether or not life could evolve spontaneously.It would seem then, that we should examine whether any of the probabilistic arguments are even valid at all, mathematically speaking. This is much more difficult than almost all folks consider.What say ye, people of the discussion?

Thanks for the welcome!I am not I explained my question well enough, but you raise a great point of explanation.Natural selection and selective breeding are well-documented and explained localized operating mechanisms, both being external operators influencing differential reproductive success. But, these topics are not central to what I was discussing, since they are always observed for a specific species in a specific situation.Rather, a closer situation would be to monitor, say, several species' morphological mutation rate in response to specific conditions, and attempting to infer a global "probability" (in time terms as well) that such a "successful" adaptation can occur for a generic species. This will prove difficult, but is what is required for the question at hand.Showing that an adaptation has occurred reveals nothing about its mathematically probability of occurrence except that it is nonzero (not even that it is nonzero, I suppose - God could have created it in violation of probability laws, but that is not what I seek). This is the core of the question.

Thanks Bluejay!I think you have hit on the key points of my question. I am well-versed in modeling, and think that my question involves the apparent lack of a good model or even awareness of modeling rules in the use of probabilistic statements. (BTW - thanks for the dBCodes tip)I don't believe I am asking for a 100% answer from the model. And, do not misunderstand, I am not dismissing any argument or technique. However, I am asking for rigorous review of statements and models, which I do not see happening.For example, I often pick on the astrobiologists' "Drake equation", which is a proposed model for the number of current planets with intelligent life. It suffers from the same problems as the "evolutionary probability" statements, and some models.You ask - "why don't you come up with something better?". That is what Science and constructive criticism does. I point out some of model deficiencies:
1. We have a sample set of one. We cannot infer generalities from this mathematically (i.e. a priori) - how can we expand our understanding to correct this?
2. Specific conditions, in this place and time, may not be applicable to other conditions - What factors should be applied to the model to correct for this potential difference?
3. Randomness or correlation is incorrectly accounted for in these models - Better understanding of phenomena such as self-organization and emergence would lead to better models.

Well, avoiding the potential spears by claiming evolution is true, I would rather just focus on a specific topic.Our model should be able to account for both internal and external effects. So, yes, external pressures will have some effects, although I don't know that I would use the term "directed". I have always considered mutations as random. They occur. They may occur more frequently due to an external situation. When an external factor provides a long-term reproductive success filter for a particular mutation-based characteristic, we can say evolution can occur, although it may not. However, neither the environment nor the individual directs this change. Also note that this change may not involve "progress" as we normally consider it.So, you have brought up some elements of a model!
Probability of a selection factor emerging
Probability of mutation occurring
Probability of stable outcome

Thanks for the friendly advice. I seem to appear to be anti-evolution. What is the word - how droll!Funny, I am not interested in the least in that debate. Anyone who wants to debate evolution vs creation is in it for neither science nor faith reasons - its about control. A scientist does not attempt to prove truth. Anyone wishing to stomp - have at it, its childish, and you lose scientific respect at an exponential rate. We should, however, hypothesize, test, revise, and move forward.If we are to truly be scientific, then we must accept all challenges and questions with an open mind and patience. If all we have is name calling, then we are lost.

Modulous, you have elucidated more good details in your replies. Thanks!Think with me (I am just exploring here): Could display behaviors be separable from other natural selection mechanisms in that is not necessarily due to a factor outside of the species itself? This is simply an attempt to identify the forms of any potential models. Excellent - now we can look at a modeling exercise. If a mutation rate is measured locally, and assumed constant (unclear), then we can use this as an independent, separable probability term in the model. However, if it is correlated to another variable, our model is a bit trickier. For example, if environmental radiation is the chief source of mutation, and a burst of radiation modifies the rates as well as exerting other environmental pressures such as a food source deprivation, the relationships may be difficult to ferret out.I appreciate what you explain about the difficulty of modeling evolutionary processes. I think you raise a good point about trying to keep things simple in the modeling arena.P.S. I don't believe you meant to say this. If this were true, the fact that there was a lottery win this weekend means there will be a lottery win next weekend, in fact, it means every drawing must result in a winner, simply because there was once a winner.

Thanks for pointing out my weak posting.First, case 2 is the hot air balloon which got me on this soap box. So, I am glad to hear that someone else agrees.Case 1 does need some definition. I think when someone does ask a question like "how did a jawless fish evolve into a jawed fish", the scientist owes a good answer, not just name-calling and deep-diving into Latin ponds. It is not enough to claim that this happened over "hundreds of millions of years" (in many cases, it doesn't seem to have). I give credit to Behe's book "Darwin's Black Box" for asking good questions, although I am not sure his analyses was thorough or that he conferred before release. I would have to give "the scientific community" (TSC) a solid C for its response to the book. Not that they don't know their stuff, but they tried to make him a crackpot. He asked reasonable questions. TSC did not give reasonable responses. They attacked him. Bad form, for scientists, but not unusual.Creationists, on the other hand, don't really know what Darwin's ideas were, and how it pained him to publish them. Darwin was truly one of the few great minds of the human race.P.S.
Some of my best ancestors were from Scotland! (Ferguson)

slow down, WK The propensity to make personal judgments is often a sign of weakness of intellectual position, but I digress. No, I don't give a hoot about Behe's thesis or intelligent design leanings. The complexity question is the only reason I bring it up. As far as examples of how the "millions of years" response is used, actually, I tend to see that in popular science quite regularly. Remember, my interest here is in probability. However, when I have looked at scientific literature the results are mixed. For some macro-scale elements, good results are found, such as the evolution of the eye. The jawed fish is relatively weak. The upright human debate is more speculation than anything else. There simply is not enough data available to explain, since inference is the chief tool. I admit to being less well-versed to the world of biochemical evolution, but the responses to Behe's book seemed more diatribe than substance, from a scientific viewpoint.However, an objective observer is not satisfied with merely a sparse fossil record to explain complex transitions. (Relatively recent publications re mammalian inner ear tracing do show some exciting lineage, and bear promise to real evolutionary proof rather than inference).In a probabilistic sense, the evolutionary scenarios I would pull from Behe's questions are the following:
1. How to account for the fact that an existing system can adapt to another purpose, in other words, each feature does not have to evolve from scratch in parallel and independently.
2. How to model the probability that parallel required system elements will be in place at the required time (the jawed fish example works at the macro scale, but several biochemical system work at the micro scale).
3. General Note: The power of the biochemical resources available for evolution is extraordinary. Simple probability which does not recognize the cumulative capabilities of this "system" and its self-organization capabilities can never capture the concept of evolutionary change.

Wounded King, you seem to understand the point of my question - but you get distracted very easily. Truly, I am uninterested in your debate over Behe's positions. It is like the tide fighting the sand - lots of action, but just wears after a while.Here are some items where I do not seem to be clear:
1. The probabilistic model I am investigating has nothing to do with proving/disproving a particular instance of evolution or evolution in general. It is a mathematical modeling investigation, and it is challenging to collect the models and data.
2. It is not clear that folks understand what a probabilistic model is versus a descriptive model. A descriptive model cannot predict whether something will happen again. A descriptive model may or may not be correct as to the occurrence it covers. A probabilistic model must work with existing examples, predict new instances, and give insights into operating mechanisms.
3. Think carefully about your statement "What possible reason would you have to do this?" If you understood developments in Physics and Chemistry over the past 150 years, you would have your answer. I admit, this is a daunting task. Complex is not equivalent to inexplicable (Faraday's Laws - deterministic) or unpredictable (quantum mechanics - probabilistic).As to the issues of "evolutionary proof", you have yourself stated what I have observed - inference is the only proof available for evolutionary change. At the macro level, scarcely more than sparse fossil records fuel that inference. As to whether this is quantitative proof, hardly. It is merely a reasonable inference. Any day a better explanation for the data could be offered, and this does happen. However, the record, the science, and the inferences all constitute data.My question is deals with all of that data. Can we use this data to build probabilistic models able to reasonably predict.Consider the example of the eye evolution (Details are not important for this question). Probabilistic models need multiple runs of experiments to collect statistical data (like rolling a six-sided die many times). The fossil record normally represents one such experiment. This is why the probability exercise seems futile - not for the reasons you suggest. But, should we be able to outline the fundamental elements required for the steps of eye evolution, this then would be like imagining the rolling of a that die given its structure and correlation assumptions. There are numerous examples of "lower" eye structures in living organisms which can serve as independent "experiments" in our development of the model. This is how the probabilistic model could be developed. Your "tornado in a junkyard" comment seem to express more the complexity of the task than an understanding of how such modeling could be done.The biological sciences has relatively recently been able to rise above qualitative modeling. Using such models allows for the ability to predict. This is the big difference between biological complex system sciences and those in the so-called "hard sciences".

OK, i will accept the challenge. Give me accessible references for evolution and I will see if they use the "millions of years" terminology.By "accessible" is meantSomething able to be accessed without cost or undue trouble - I am not spending much effort on this. Also, this should deal with evolution in a more popular sense - if it is a purely scientific journal then the vast majority of the population will not read it. When I published I knew that no more than 3 people read my papers, and I forced one person to read them.Keep in mind that I am a supporter - I am doing this to kill this ridiculous argument.

WK, there is something in the style and content of your reply that is troublesome. I quote directly: You left off the last line of the paragraph which does give a clear example of how quantitative and predictive modeling have been successful. Then, you call my response vacuous and unclear. Motive is questionable here. So, you want some specifics. How about radio wave emission from AC current in a wire (as modeled quantitatively by Faraday's laws). How about probability amplitude models (Schroedinger's equation) being able to predict the structure of simple atoms (predictive probabilistic model). This is what was referred to in the sentence you dropped.But, to deal with the topic, the point is that quantitative modeling is important, even if you don't understand why. The hard sciences have had enormous success with quantitative modeling, while the evolutionary sciences have been limited to mostly descriptive modeling (although significant progress has been made since computation power has allowed, hmm, complex problems such as genetics to be addressed).Predictive and quantitative models allow you to move forward. Descriptive models do not.

PaulK, you misread my reply.I was suggesting that you provide acceptable publications and I will find the references. It was an attempt to avoid the situation of providing references which you would claim are not credible because of source quality.Sheesh, you people are touchy. Must be the forum. Argument prone.