Information and Genetics

O hope at some point you are going to answer teh points I have raised in other threads. Until those points have been answered there are very good grounds to doubt that the genome does contain information in Gitt's sense.Likewise the idea that the genome contains CSI in Dembski's sense is just an opinion without any solid foundation. We're not in a position to even try to work that one out at present.

No, I don't misrepresent Gitt - I quoted directly form the AiG article to support my case. Gitt's article contains a quote claiming that semantic information cannot be mechanically translated. If Gitt does nto agree why did he include the quote ? If he does agree then how can there be semantic information in DNA ?Likewise you do not deal with the point I raise. No intelligence generates the message to say that a packet failed to arrive intact. The message is generated mechanically anf it originates information in the Shannon sense, but not in Gitt' sense.(And I would add that Chaitin-Kolmogorov information also does not represent meaning for the same reasons as Shannon. Shannon information can certainly apply to the transmission of genetic information - but it can only represent fidelity, not the phenotypical effects of any changes).

The original use in this thread referred to an insistence that expert opinion was the only acceptable evidence on the point in question. When apparently neither side knew what the expert opinion might be.Pointing out that it may be valid to rely on an appropriate authority when that authority's opinions is available is not relevant to that particular issue - since no such authority has been cited on either side.

Can you explain to me why, instead of answering the point I did make (that according to Gitt semantic information cannot be mechanically translated - as DNA is) in favour of answering a point I never made (that intelligence must be present every time information is copied) ?
That these are different points can be clearly seen by considering that in the reproductive process DNA is not simply copied - the information is extracted and used to produce a new individual. And this is the only information in DNA that we know of. The reproductive systems (in a wide sense, including eggs) *are* the receivers, so there seems to be no intelligent receiver either. Surely these facts are a major obstacle to the conclusion that DNA contaims Gitt information and I would appreciate it if they were answered instead of my points being misrepresented.And even when you have answered those you face the problem that the identification of prgamatics and apobetics requires showing that there is an intent. To claim that DNA has either without evidence of intent begs the question. Any argument which tries to conclude that life has an intelligent source based on the idea that DNA contains Gitt information cannot make much use of these levels since any argument that DNA does contain them would work just as well to support the conclusion that DNA has an intelligent source without the concept of Gitt information at all.

Rather than calling my points "inaccurate" without explanation and referrign me to Gitt's book I suggest you try to actually discuss the issues.It is a fact that the information in DNA is mechanically translated. It is a fact that Gitt argued that semantics could not be mechanically translated. Neither of these points is in any way inaccurate. If Gitt's book denies either then Gitt's book is inaccurate.
So your response to these is not only useless - you don't sya what you consider inaccurate - but obviously incorrect, because the points you object to are known truths.Your response on the "purpose" of life rests an equivocation of purpose (as well as assuming that DNA is the original genetic material for life - a view that is very likely false). There is no indication of *intent*.So my points remain unanswered and my conclusion that DNA does not contain Gitt information remains unchallenged.

Didn't you follow the link I provided ? Are you really telling me that you haven't even READ the points that you are suppoed to be answering ?In the post you were meant to be replying to I provided the quote from GItt's essay and the link to show that Gitt did indeed argue that semantic information could not be mechanically translated. It is not even an unusual view - Searly for instance makes it a centrepiece of his arguments against Artificial Intelligence. If the reproductive systems, complex as they are can indeed understand semantics then that would conclusively demolish Searle's arguments in showing that not even intelligence was required.And why would you not believe Gitt's own words, regardless of where they appear ? Are you suggesting that Gitt might misrepresent his own views ? If so then how could we trust the book ?If you are claimign that apobetics has no requirement of conscious intent then I accept you are not equivocating. Yu are destroying Gitt's arguments instead - any self-perpetuating system would present "apobetics". Even one resting solely on the laws of physics and chemistry.

Well lets start with the fact that Gitt's essay is not just a web publication - it appeared in AiG's "peer reviewed" Technical Journal. AiG EssayI suggest that you read the section on "theorem" 9 yourself. The thrust is clear - semantics must represent meaning beyond the physical properties - there must be a mental dimension. That indeed is the basis of "theorem 10". But where is the need for understanding DNA in any sense other than understanding how it functions in it's proper context ? An understanding that applies equally to examples such as tree rings.

Theorem 10 rests on the idea that semantic content has a mental dimension, so certainly Gitt does assert that semantic content has such a dimension.And I suggest that you remember that the point I am criticising is the idea that DNA contains information as Gitt defines it. This is certainly very much in question. It's not even clear what you think Gitt's "error" is - do you really think that theorem 10 is a complete mistake ? If not, then what ?The genetic code (meanings of codons) is syntatctic, not semantic. Nor does the active role of DNA change the fact that the process it is involved in entirely follows the laws of physics and chemistry. In this respect it is like the tree rings.

I note that the arguments requires DNA to include (Gitt) information including all the levels from semantics on up. I still see no clear argument that any of these levels are to be found in DNA (the closest has been a set of assertions that certain aspects of the "decoding" represent these level without clear explanation. Moreover, as I have pointed out assigning these levels contradicts statements made by Gitt in a paper from AiG's top journal concerning semantics (similar views are also held by at least one signficant figure - Searle - in modern philosophy working on the nature of intelligence) and beg the question of an intellignet origin at the levels above that.

THere is a lot I could take issue with in your post, but lets start with a simple objective issue.What do you mean when you say that Avida uses "truncation selection" and what is your evidence that this is the case ?

quote:

---

1) GAs employ truncation selection. For example, in Avida see BIRTH_METHOD, Choose max Age/Merit. This is severe truncation selection! Problem is, there is no evidence that truncation selection occurs in nature.:

---

This is only ONE of the available methods. The DEFAULT method is to choose the oldest "creature" out of the "mother" and those in the eight neigbouring cells. Even the method Fred refers to does not directly implement truncation selection, since instead of using a fixed cut-off for merit it uses age modified by merit and chooses the worst (highest) score. (And all methods choose empty cells over occupied cells).[ADDED in Edit] The most famous use of Avida - the paper "The Evolutionary Origin of Complex Features" by Lenski et. al., showing the evolution of IC functions, in fact used the CHOOSE_RANDOMLY birth method (selecting either the "mother" or one of the 8 adjacent cells to be replaced). Avida clearly does NOT rely on the max Age/merit method.The paper is available at - and the information is in the "Experimental Conditions" section.
http://myxo.css.msu.edu/papers/nature2003So Fred how did your investigation into Avida manage to miss all these facts ?

quote:

---

2) GAs assume that a rare number of beneficial mutations exist (the variety where the mutated-type is more fit than the parent type in a normal (typical) environment). There is little or no evidence of this in nature.

---

That beneficial mutations have occurred is documented. I suspect that this point is nothing more than a quibble on "usual environment".
As it stands it is clearly false.

quote:

---

3) GAs set the hypothetical Beneficial mutation rate too high. There is no evidence of this in nature.In fact, evidence points to a mutation rate that given enough time will lead to extinction of all life ( 404 Not Found )

---

This is just nonsense. If it were true life would be extinct. A higher proportion of favourable mutations simply speeds up the process of evolution.

quote:

---

4) GAs do not permit extinction, ie they do not disqualify themselves but will instead run ad infinitum if need be. If realistic mutation rates were used they could not produce anything at all.

---

This presumably means TOTAL extinction. In principle Avida running with a DEATH_METHOD could result in that. But it is hard to see how this is a serious objection, unless total extinction is assumed to be near inevitable.

quote:

---

5) GAs are useless unless intelligence is there to cause the output to be realized. This is especially true of commercial GAs, as Rei calls them.

---

This clearly does not apply to Avida, Tierra and the like. Even for commercial GA's it is because they run in software rather than attempting to breed the hardware directly - which is what Fred seems to be demanding. This objection is obviously silly even when it does apply.[This message has been edited by PaulK, 09-30-2003]

No Fred it is not a strawman.You claimed

quote:

---

2) They do not emulate evolution in nature because they typically invoke truncation selection (the GAs I have looked at are Tierra and Avida)

---

If the only support for your claim relies on a single option then you certainly cannot claim that Avida typically relies on "truncation selection". Which raises the question of how you can come to a general conclusion when you have not even examined one of your *two* examples sufficiently to support your claim.I explained how Avida was used for that particular paper. If you think that implies that there is no selection then you either don't understand Avida or how the behaviour of Avida results in selection. Either way it shows that your knowledge of Avida is very superficial - yet you list it as one of the GAs that you DO know. Which raises the question of whether you have only a single example - or none at all.I will repeat that it is clearly stated that the selection of cells for replacement was random. And that you have not even shown how the age.merit example produces truncation selection. That is enough to demolish your argument.And if you wish too claim that the description you gave implies truncation selection then it is for you to explain how. It looks to me as if you are simply assuming that success relies on truncation selection which begs the question. I can't refute an argument that isn't even given.(I will point out that it explicitly states that the particular line was the most successful in the run - and so it follows that it will show a high proportion of beneficial mutations, because lines with fewer would be less successful ! So evaluating the efficiency of selection based on this line alone is "selective reporting" - and the fact that even this line included a significant number of deleterious mutations - 18+39 = 57 - shows that selection was certainly not able to eliminate all of the deleterious mutations)Indeed it seems that you assume that evolution cannot work and then dismiss any evidence form GA's that the processes underlying evolution do work as "rigging" because they contradict your assumption. A closed circle of reasoning for a closed mind.As for your final comment

quote:

---

Is ‘evolution’ guaranteed to succeed? Perhaps
this will help some understand my reason #4 why GAs don’t represent reality: GAs do not permit extinction, ie they do not disqualify themselves but will instead run ad infinitum if need be. If
realistic mutation rates were used they could not produce anything at all. Thanks for providing a classic example

---

it is just another example of question begging. You fail to deal with my point that you need to actually SHOW that total extinction is so likely under a "realistic" scenario (ignoring the prima facie evidence that life has existed for billions of years without total extinction) - as well as equating "high probability" to certainty.

Well let's start by pointing out that Fred's original "straman" assertion has been already debunked by direclty quoting what he said.

quote:

---

Paul then gets the irony of the month award for this statement:

​

quote:

​

it is just another example of question begging. You fail to deal with my point that you need to actually SHOW that total extinction is so likely under a "realistic" scenario (ignoring the prima facie evidence
that life has existed for billions of years without total extinction)

​

If you fail to see Paul’s own question begging in the sentence where he accuses me of the same, by golly this is a goooood sign your dogma and blinders are pegging the ‘severe’ meter!

---

Well it is certainly a fact that life has been around for a long time and the scientiifc consensus (based on evidence!) is that it has been around for billions of years. We have evidence that life is evolving now and evidence that is has evolved in the past. So on the face of it we do have prima facie evidence that life HAS been evolving for billions of years.Fred on the other hand won't even explain WHY it is "unrealistic" not to allow total extinction as he claims.And he still doesn't. So we can scratch that argument by Fred and a pure fabrication on his part.On to the serious stuff.

quote:

---

This is yet another strawman, because truncation selection doesn’t mean you necessarily will eliminate all of the deleterious mutations. You can have both truncation selection and deleterious mutations present. More importantly Paul refused to answer how the application of the GA in the paper he cited avoided truncation selection while producing such rapid results of accumulated beneficial mutation. No support was offered to defend his claim, yet I am the one arguing frompersonal incredulity? Go figure!

---

This is an admission by Fred that he does not understand how Avida works. If you remember Fred claimed to have studied Tierra and Avida and those studies were the basis for his claims about truncation selection. When challenged on Avida Fred could only point to a single birth option which he claimed produced truncation selection - and could not even explain that.Because Fred can't argue from knowledge or understanding (proving that his study was clearly inadequate) he has decided to set up his unsupported assertions as the default and demand disproof.In doing so he sets up more questions. If truncation selection is so unrealistic then why would it have to be AVOIDED ? Where is the reason to suspect truncation selection at all ? It seems to be based on the view that truncation selection is necessary to get good results - and absolutely nothing else. But that assumption has yet to be supported - the burden then is on Fred to either support hisassertion or to show that there somehow the Avida run requires *unrealistic* truncation selection. Until he does it constitutes prima facie evidence against his assertionFred also fails to mention that the only results he reported were those for the MOST successful line, and even these included a number of deleterious mutations which obviously were not immediatately selected out - the most successful line by definition has done well above the average. He also accused the authors of the paper of fraud with no grounds at all other than his refusal to accept the results.[edited]
As for Fred's "strawman" assertion well THAT is ironic. Fred claimed that the selection was very efficient and - as stated above - I pointed out that the number of deleterious mutations found in the MOST successful line indicated that the selection was not that efficient after all. In short it was never about truncation selection. So Fred's strawman accusation is itself a strawman. Beat that for irony, Fred.Fred, it is up to you to support your assertions. If the best you can do is claim that the counter evidence doesn't amount to proof - while not even offering evidence to support your own claims - then you simply don't have a case. Simply trying to gloss over that fact as you do in your comments on my posts won't work.[This message has been edited by PaulK, 10-08-2003]

And Fred, I'm quite willing to explain how selection happens in the particular Avida scenario under discussion. All you have to do is admit that you don't know.(Really it's quite easy to work out from a basic knowledge of evolution).