Evolution Not Provable ?

[Weirdmaskman]

Evolution not provable? The concept of evolution is possibly the most radical idea in human history. It challenges all nations {as we can see d debate here} of the permanence of biological form; what is "human today" may have been "ape" only a few million years ago and "worm" many millions of years b4 that. This, we'v heard many times, but in some ways, it defines common sense; how could something as complicated as a human possibly have had simple ancestors?

Evolution, being the cornerstone of biology, as one biologist remarked, "nothing in biology makes sence except in the light of evolution", has its evidences lyk fossil records, common ancestry, homologous structures, vestigial organ, embryology, biochemistry, et cetera but these and many more comes with its bottleneck and perplexing ideas.

[StringJunky]

Evolution not provable? The concept of evolution is possibly the most radical idea in human history. It challenges all nations {as we can see d debate here} of the permanence of biological form; what is "human today" may have been "ape" only a few million years ago and "worm" many millions of years b4 that. This, we'v heard many times, but in some ways, it defines common sense; how could something as complicated as a human possibly have had simple ancestors?

Evolution, being the cornerstone of biology, as one biologist remarked, "nothing in biology makes sence except in the light of evolution", has its evidences lyk fossil records, common ancestry, homologous structures, vestigial organ, embryology, biochemistry, et cetera but these and many more comes with its bottleneck and perplexing ideas.

 

It makes even less sense to me that something as complicated as a human could just 'pop' into existence as we are now in one go...if you really think about it it's quite absurd from a scientific and even 'common sense' perspective.

 

It's going to be a long haul but Evolutionary Theory is on the right track. IIRC it was dialogue between the evolutionists and cosmologists that helped the cosmologists determine the age of the Earth because the evolutionary biologists understood the minimum necessary timescales for certain biological events to occur and therefore the minimum theoretical lower bouind for it's age.

 

Evolution is a story that will take scientists a long time to understand and tell.

[Ophiolite]

IIRC it was dialogue between the evolutionists and cosmologists that helped the cosmologists determine the age of the Earth because the evolutionary biologists understood the minimum necessary timescales for certain biological events to occur and therefore the minimum theoretical lower bouind for it's age.

A small correction, but a necessary one if I am to defend my brother geologists. There were two dialogues. The first initiated by Hutton, who saw "No vestige of a beginning, no trace of an end", carried forward by Lyell (who was victorious in the Uniformity battle), then erroneously answered by Lord Kelvin, whose calculations failed to account for radioactivity. It was resumed in the 1950s when, for a brief period, geologists had established the age of the Earth as greater than the physicists and astronomers had determined that of the universe to be. The latter finally got it right.

[StringJunky]

A small correction, but a necessary one if I am to defend my brother geologists. There were two dialogues. The first initiated by Hutton, who saw "No vestige of a beginning, no trace of an end", carried forward by Lyell (who was victorious in the Uniformity battle), then erroneously answered by Lord Kelvin, whose calculations failed to account for radioactivity. It was resumed in the 1950s when, for a brief period, geologists had established the age of the Earth as greater than the physicists and astronomers had determined that of the universe to be. The latter finally got it right.

 

If it was the geologists and not biologists that engaged in the original dialogue I prostrate myself and humbly apologise to you and your professional brethren.

[Edtharan]

It appears to me that all you have demonstrated is that biological evolution could be identical to algorithmic evolution. You would also have to demonstrate that no other mechansims are present in biological evolution in order to state the case has been proven.

When you look at biology from an information technology perspective, all DNA is, is a program. A program being a set of instructions carried out by a Turing Machine.

 

The way that DNA is converted into proteins is a program. A sequence of 3 DNA letters forms a code that is translated by the ribosome into an amino acid which is joined to the growing protein.

 

The way a Turing Machine works is that it translates one set of symbols into another set of symbols according to a pre-set list of instructions (a universal Turing machine is a bit more complex, but is essentially the same thing, but one that the set of instructions is also subject to manipulation of the Turing Machine).

 

As you can see, the conversion of one set of symbols (DNA) into another set of symbols (amino acids) according to pre-set instructions (ribosome) is just what is happening. At its core, biology works of the same principals as information technology, and is thus also subject to proof via the same mathematics that govern it.

 

Of course, Turing Machines are the base of IT, and there has been much research and investigation into how they work at a higher level (how the basics combine into more complex structures and processes). This "System Analysis and Design" as it is known will also apply to other systems that use the same basic functions.

 

This can be shown because of the symbolic nature of information. As these symbols are abstract, and that the mathematics that govern it are not dependent on what the symbols represent, but how they are manipulated, by looking at the way the information within the system operates, we can determine the behaviour of the system.

 

The history of the development of this kind of analysis and mathematics goes back a long way. Algebra is where this mainly started. Once people realised that common mathematical processes could be generalised into just symbolic representations, we could dispense with numbers and use other symbols to do the same thing.

 

As 1 + 1 = 2 and that 2 + 2 = 4 share a patter ( if you add a number to itself, then the answer is equal to 2 times the number), this can be represented symbolicly as: a + a = 2a

 

But as times when on, it was realised that even these algebraic expressions could be represented symbolicly, and algorithms were invented. Computers are just devices designed to operate on, and create algorithms.

 

System Analysis is the next step up. It is a way of comparing algorithms and showing they are identical in function.

 

In practice, this is generally used to identify a process that occurs in the real world, such as the processing of forms for a business and creating the same process in the computer as a program. But, this could not occur if the program in the computer was not the same as the one that occurred on paper. Buy showing they were identical systems, the programmer can test that their software performs the same function as the paper system.

 

Now, using such analysis on biological evolution and algorithmic evolution, you can show that they are the same process, and that what would occur in one will occur in the other. More over, part of system analysis is identifying inputs and what data they act on and how they effect the processes going on.

 

If there is some unknown influence on the system, it can be shown to be either insignificant or irrelevent as the outcome of the algorithmic evolution, where all external influence are controlled and known is the same as biological evolution.

 

This means that either there are no external influences that create the process of evolution, or that any external processes are unnecessary to create the process of evolution.

 

IF there are external processes create the illusion of evolution, then they are operating exactly as the algorithmic processes would demand. And according to the principal of Occam's razor, if two or more explanations produce the same output, then you should use the simpler.

 

So which is more simple:

 

That biology, which has been shown to be capable of performing the functions necessary for evolution, is capable of performing those function without an external influence.

 

Or that biology, even though it is capable of performing the function of evolution has an external influence that prevent it from doing so, but then acting to create the illusion of the process of evolution.

 

Think about it. This is what creationists claim, they claim that something both prevents and then creates the illusion of evolution, despite the functions of biology having the capability of doing so unaided.

[Ophiolite]

I am not proposing that there are some external factors at work. I am stating that if their are wholly natural processes, currently unrecognised, that influence the path of evolution then these may stand outside any evolutionary algorithm. I repeat that what you seem to have done is demonstrate that aspects of evolutionary process can be described mathematically and thus proven, but you have not demonstrated that all aspects of the evolutionary process can be so described.

 

View this response as a placeholder while I ponder your suggestion more carefully.

[mooeypoo]

I am not as well versed as you guys on Evolutionary processes (I know the basics), so I apologize on the interjection here, but I had a thought.

 

I think you are debating more about semantics and perhaps argue across one another. I don't think the point was that we should be able to have *all* of evolutionary processes proven, but when the claim that "it's not proven" arose, the point (I think) Edtharan was making is that since we showed that we can replicate the process mathematically, we proved that the process exists. A lot of science is about "supporting" a theory, not quite 'proving' one. In the case of evolution, however, we *did* prove the process absolutely exists as a general principle. We managed to replicate it mathematically and use that as some form of general predictive measure.

 

Whether or not particular evolutionary steps occurred in certain species versus others is a different point. The theory of evolution is, unlike some other biological theories, mathematically proven.

 

This is more often said about physical theories (like gravity and relativity, etc) and not often about biological processes... so it puts evolution even more so in the same "rank" or "level" of "substantiability" as gravity, relativity, and other mathematically proven physical processes.

 

Am I making sense?

[MajorTom]

I think it depends on what is meant by "provable". As mooeypoo said, I think semantics are causing a lot of confusion.

 

If one means by proof a, "logical conclusion based on evidence", not only is evolution provable but it is proven.

 

If one means proof in the perpetual sense, evolution is no less provable than any other known fact. Evolution is a fact.

 

If one means proof in some other sense I'm not thinking of, then I'm not sure how to answer the question.

 

All that said, proof is a word that causes a lot of confusion, hence this discussion. Evidence is a better word. Let the evidence speak for itself.

[pantheory]

I think that by saying "there is no evidence on either side, its all speculation" is an incorrect statement concerning evolution. You might say there is no proof of the theory of evolution but there is almost countless evidence now that supports this theory, and its subset natural selection.

[Edtharan]

I am not proposing that there are some external factors at work. I am stating that if their are wholly natural processes, currently unrecognised, that influence the path of evolution then these may stand outside any evolutionary algorithm. I repeat that what you seem to have done is demonstrate that aspects of evolutionary process can be described mathematically and thus proven, but you have not demonstrated that all aspects of the evolutionary process can be so described.

 

View this response as a placeholder while I ponder your suggestion more carefully.

Because we can prove and understand the process as a mathematical construct, we can use maths to analyse a particular even of evolution and then see if any other influences are acting on it other than what are contained in the mathematical model. No extra influences have been found, thus either they are too small to detect, or don't exist.

 

If they are too small to detect, this means they are having a negligible effect on the process and can safely be ignored (in most circumstances - which is why scientists still look for them). IF they don't exist, then it is pointless to propose them.

[Ophiolite]

Nonsense. Cite an instance where we have sufficient detailed data of the event to be able to forumulate all of it mathematically. We see the results of the process in the fossil record and can infer, to some degree, the steps from genome analysis, but to suggest we can currently deliver the necessary precision to do what you are claiming is simply not the case.

[pantheory]

Edtharan...Because we can prove and understand the process as a mathematical construct

While there is a great deal of supporting evidence for evolution tis true, and some supporting genetic analysis involved, but math is not a good analog of evolution and probably never will be IMO.

Edited July 7, 2011 by pantheory

[Edtharan]

Nonsense. Cite an instance where we have sufficient detailed data of the event to be able to forumulate all of it mathematically. We see the results of the process in the fossil record and can infer, to some degree, the steps from genome analysis, but to suggest we can currently deliver the necessary precision to do what you are claiming is simply not the case.

You did know I was talking about the process of evolution, which is not only biological evolution. I made it a point to talk about both Biological evolution and Algorithmic evolution in order to separate that I was talking about both separately.

 

What I showed is that Algorithmic evolution is mathematically proven, but that Biological Evolution is just one instance of the algorithm.

 

We look to the fossil record to see if traces process of algorithmic evolution is left in the fossil record. Even if the fossil record did not show any evidence of evolution, it would not make a difference to my argument.

 

To reiterate my argument: Evolution, as a mathematical process has been proven. Also, biological systems have the same processes going on in them that would cause evolution to occur in them.

 

I am not saying that we are applying mathematical analysis to fossils to come up with the process of evolution to prove it is true. Even if biology did not exhibit evolution, my argument for the existence of Algorithmic Evolution would not be impacted at all. Trying to use the debate over biological evolution as proof that algorithmic evolution doesn't exist is a false argument.

 

What I did was to first establish that Algorithmic Evolution exists. Then once that is established, show that biological system perform the same processes and thus evolve. You seem to think I am trying to do the opposite: Prove mathematically that biological evolution exists, and thus prove that algorithmic evolution exists.

 

While there is a great deal of supporting evidence for evolution tis true, and some supporting genetic analysis involved, but math is not a good analog of evolution and probably never will be IMO.

I am not trying to use mathematics as an analogy. I am literally trying to show that Evolution as a mathematical algorithm is a fact. Then after that, show that biological evolution is just an implementation of that algorithm. In the first part, it does not matter at all if biological evolution exists. All I need to do is to show that there exists an algorithm that acts on data sets in the way evolution is supposed to. Only after that has been done do I try to show that biological systems show that same behaviour and that they do so through the same algorithm (set of mathematical processes).

 

All one needs to show to prove the first part is that computers will execute a program that implements algorithmic evolution. As this has been known about for almost as long as computers have exists, this is pretty much proven.

 

To answer the second part, all I needed to do was to show that biological systems can perform the processes required by the algorithm, and that these processes are linked together into such an algorithm.

 

As I said with Ophiolite above, I am not starting from biological evolution and trying to prove that mathematically. I am starting with mats, then showing that there exists a mathematically true process that produces the results we would call evolution (not biological evolution at this stage though). It is only once mathematical evolution is established that I try to show that biological systems perform this too.

[Realitycheck]

A mutation event occured around Scandinavia around 10,000 years ago, resulting in the blonde, blue-eyed genotype (not necessarily at the same time). The relative recentness of this is the primary reason why it is so rare. That's all the mathematical proof I need.

[pantheory]

Edtharan,

 

When I used the word "analog" above, it was unrelated to the word analogy. Here analog means an attempt to "approximate" evolution in some way. Concerning a math analog, for instance, the inverse square law is an excellent analog for both gravity and magnetism. But to provide a closer approximation, one would need to use General Relativity or Maxwell's equations respectively. Evolution concerning genes and DNA changes, is extremely complex since there are many "players" in this process. If you know of such an attempt at using math to approximate evolution I would like to see it. In my opinion it will be a long time from now, and still then of questionable value, concerning a useful computer program for the evolution process since it is very random involving the fickle finger of fate , or statistics in the language of math.

[pwagen]

Evolution concerning genes and DNA changes, is extremely complex since there are many "players" in this process. If you know of such an attempt at using math to approximate evolution I would like to see it.

Regarding your second sentence, genetic algorithms come to mind.

 

http://en.wikipedia.org/wiki/Genetic_algorithm

 

The problem is, this isn't very close to "real" evolution. For one, you evolve your programs with a goal in mind (optimization of some kind), which doesn't happen in the wild. The second problem is related to the first of your sentences I quoted here, the massive amount of genetic information and whatnot in a real world situation. Genetic algorithms can be seen as a decent approximation of evolution, at best, but to my knowledge this is about as far as we've come when it comes to "mathematizing evolution" (yeah, I just made up a word). At least when it comes to practical uses.

[pantheory]

pwagen,

 

A genetic algorithm is a search heuristic that mimics the process of natural evolution. This heuristic is routinely used to generate useful solutions .....

Quote from your link above, the key words here, I think, are "useful solutions." I, however, did not see the word mathematics in your link. It does say "Genetic algorithms find application in bioinformatics, phylogenetics, computational science, engineering, economics......." where it does say computational science that would imply math of some sort. Although maybe the most common use of the word algorithm is in mathematics, non-mathematical heuristic programmed algorithms might be used to assist in determining new approaches to alter genetics of plants and animals. Non-mathematical algorithms might be used to determine paths for future research such as genetic changes for agriculture, mathematical algorithms could be used to speculate how long particular observed changes might have taken to naturally evolve. So your statements have validity in that math can be used in genetic evaluations but I think that math, including algorithms and statistics, are no more than minor players concerning the study of evolution as a whole.

Edited July 12, 2011 by pantheory

[pwagen]

So your statements have validity in that math can be used in genetic evaluations but I think that math, including algorithms and statistics, are no more than minor players concerning the study of evolution as a whole.

Indeed. Genetic algorithms imitate some parts of real evolution, however I don't think they can be used to predict it, in any useful way.

 

And sorry for bringing in computational terms. I haven't really connected the word "algorithm" with maths (which is probably a reason I still haven't finished analysis), and didn't register you were talking about maths. So sorry about that, moving on.

[Edtharan]

Algorithms are a branch of mathematics. If you look at the history of mathematics, it started with simple statements like 1 + 1 = 2 and 1 + 2 = 3. Now, these simple statements can be generalise to statements like a + a = 2a and so on.

 

This first step of generalisation is called Algebra.

 

Interestingly algebraic statements can also be generalised into functions. And when you combine these functions together they are called Algorithms.

 

An example is trying to find the area left over from a square with a circle cut out of it that has the same diameter as the length of a side of the square.

 

You can do this by combining the Algebraic functions of both the area of a square and a circle, and subtracting the area of the circle from the area square.

 

So the algorithm is:

1) Find the length of one side of the square

2) Find the area of the square

3) Find the area of the circle

4) Subtract the area of the circle from the area of the square.

 

Written out in a Mathematical language:

1) a = Length of side of the square

1) b = a * a

2) c = pi * (0.5a * 0.5a)

3) d = b - c

 

Or as a computer program (psudocode):

 

Input: LengthOfSide

CircleRadius = LengthOfSide / 2

AreaOfSquare = LengthOfSide * LengthOfSide

AreaOfCircle = CircleRadius * Circle Radius * 3.14

FinalArea = AreaOfSquare - AreaOfCircle

Output: FinalArea

 

All 3 of these are identical. Even though the first is just an English language explanation of the process, it is none the less a mathematical algorithm. With algorithmic evolution, it too has a mathematical set of processes even though it can be expressed through English.

 

These should help:

http://en.wikipedia.org/wiki/Universal_Turing_machine

http://en.wikipedia.org/wiki/Computable_function

http://en.wikipedia.org/wiki/Turing-computable_function