Uncommon Descent Serving The Intelligent Design Community

PZ Myers Does It Again

Share
Facebook
Twitter
LinkedIn
Flipboard
Print
Email

PZ Myers has, once again, railed against something that he doesn’t understand at his blog Pharyngula. Hi PZ! Notice that he doesn’t actually address the content of Dr. Dembski and Dr. Marks’ paper, which you can read here: Conservation of Information in Search: Measuring the Cost of Success, published at the IEEE. Given his argument, he doesn’t know how to measure the cost of success, yet claims that Dr. Dembski doesn’t understand selection. A bit of advice PZ, the argument presented by Dr. Dembski and Dr. Marks is very sophisticated PZ, your mud slinging isn’t PZ, you need to step it up PZ. I know this new stuff isn’t ez, but you may want to consider a response that has actual content PZ. Your argument against this peer-reviewed paper is still in its infancy, or, more accurately, still in the pharyngula stage, embryonic in its development.

Since evolution of the kind PZ subscribes to cannot be witnessed, the argument has moved into genetic algorithms with the advent of computational abilities to determine the affair, and the IEEE is an entirely appropriate place to publish on that subject. We’re not going anywhere, we’ll give him time to catch up and educate himself to the tenets of the paper’s actual content. And if/when he does, maybe he’ll write another blog, and possibly write one with active information, that is, actual information, or else his argument will never reach it’s target.

Comments
Onlookers: I note that BillB has now fired off a half-cocked gun, leading to a hang-fire. (He seemingly cannot conceive that he might not know the full set of relevant factors before commenting adversely on a matter and on another person's character.) I noted earlier that no sensible rape complainant in Barbados shows up in court dressed in a micromini. BillB plainly thought that his had to do with my prejudices -- but I was simply a spectator, not a participant in the trial. My opinion had zero weight on the outcome of the case. (Let's just say that I sat in a waiting room across from a late 20's young miss dressed in a greyish, plaid micro mini, black sheer stockings and a low cut top, and chatted with a couple of friends in for a traffic violation. The topic came up as to why I was there, and I said that I was there because of a rape case accusation against a former student who 300 mi from home was on a rape accusation by another student. Imagine my shock when I got into the courtroom, to see the same young miss as the complainant! Her lawyer should have been fired from the bar for failing to advise his client on suitable court dress, especially with so weak a case as she actually had.) The actual case was stopped by the Judge as unsafe to further pursue when it turned out after a couple of hours of evidence, that the claimed crime occurred in a bed room in a rooming house full of students with no-one else in the house at the time reporting any untoward sounds or circumstances. A complainant who presented herself as she did, and whose case was as weak as that effectively asked for such a result. I trust BillB will learn a lesson. GEM of TKI PS: Dieb, you will see that the specifics of the filter used are critical as I discussed in the linked, app 7, point 24, as well as pop size. AND, observe that we see showcased cases in the Weasel 86 publications, showcased precisely because they show steady cumulative progress to target. detuning the same basic algor perhaps for video impact could easily account for the 1987 plainly non-latched case. Cf the actual run that shows implicit latching here, as previously linked. I believe this was the second run I did using Atom's adjustable weasel, so it cannot be that improbable. And note the parameters were not "strange" either: a 4% mut rate is about one letter on average per member of pop, and 50 - 500 are reasonable pops.kairosfocus
August 25, 2009
August
08
Aug
25
25
2009
11:38 AM
11
11
38
AM
PDT
And here you can find a graph of the reversal probabilities for different populations and rates of mutation.DiEb
August 25, 2009
August
08
Aug
25
25
2009
10:45 AM
10
10
45
AM
PDT
So, what's the probability for an execution of Dawkins's algorithm that at a generation is followed by a less fit one at least once, i.e., that a latching definitely didn't take place? I did the math for a mutation rate of 4%: 1. population size 10: 95,7% 2. population size 50: 0,0000026 % For a small population - as obviously used in the video of 1987 - reversals are very probable. And while the probability is small for a population size of 50, some lucky programmer may have observed it...DiEb
August 25, 2009
August
08
Aug
25
25
2009
10:05 AM
10
10
05
AM
PDT
Joseph#187
In TBW the way Dawkins describes and illustrates cumulative selection there is no other inference than one of a partitioned search.
Picking one word out of a two page description of the algorithm does not come remotely close to the level of detail provided in the website to which I linked. If that's all you have to support your claim, you should just retract it now. Again, please provide a detailed explanation of how Dawkins’ words could possibly be interpreted as describing a partitioned search. I don't believe you can; what Dawkins describes is clearly not a partitioned search.DeLurker
August 25, 2009
August
08
Aug
25
25
2009
08:21 AM
8
08
21
AM
PDT
Joseph, hhhmm, Dembski and Marks give a pretty good description of their algorithm and it doesn´t include building a population or selecting something from a population. In each generation new random values are chosen for each incorrect letter, that´s all there really is. I think my claim is absolutely valid. If we can´t agree on the meaning of these very simple sentences there is not much more to discuss. Or maybe you can elaborate a bit?Indium
August 25, 2009
August
08
Aug
25
25
2009
07:46 AM
7
07
46
AM
PDT
DeLurker, The reference is to TBW. In TBW the way Dawkins describes and illustrates cumulative selection there is no other inference than one of a partitioned search. "Cumulative" is the key word. Now all you and your ilk can do is try to say that cumulative doesn't really mean cumulative.Joseph
August 25, 2009
August
08
Aug
25
25
2009
07:41 AM
7
07
41
AM
PDT
Indium, Your reference doesn't support your claim. As far as you know the partitioned search wentt exactly how Dawkins program went- population wise that is.Joseph
August 25, 2009
August
08
Aug
25
25
2009
07:39 AM
7
07
39
AM
PDT
R0b,
Biological evolution also transfers information from the environment to biota. In both cases, the mechanism of information transfer is cumulative selection. The point of disanalogy is that WEASEL has a long-term target and life does not. So what was your question again?
My question was on randomness, how Weasel, a targeted search, a point of reference, a distant goal, evidences anything about life if life isn't like that, in the respect that there is no target, point of reference, or goal. It seems like you're saying that a mere transfer of information links Weasel to real life, but the program was not designed to evidence mere transfer of information. It was supposed to evidence cumulative selection that actually does something, namely, meets a target phrase. If there is no target phrase, then accumulation is anything at all, all types of accumulation, even nonsense phrases. Indeed, there would be no way to determine what was nonsense and what wasn't unless you had a target. Que sera sera, whatever will be will be seems to me to be the very thing you're saying when you admit that there is no target in life.Clive Hayden
August 25, 2009
August
08
Aug
25
25
2009
07:39 AM
7
07
39
AM
PDT
Joseph#182
There isn’t anything in TBW that would lead one to suspect cumulative selection is anything other than a partitioned search.
I have already referenced this site where a programmer went step by step through the entire description of the Weasel algorithm in TBW and implemented it directly from Dawkins' own words: http://www.softwarematters.org/more-weasel.html Please show a similarly detailed explanation of how Dawkins' words could possibly be interpreted as describing a partitioned search.DeLurker
August 25, 2009
August
08
Aug
25
25
2009
07:22 AM
7
07
22
AM
PDT
Joseph: "For the incorrect letters, we select 26 new letters and obtain OOT*DENGISEDESEHT*ERA*NETSIL. (21) Five new letters are found, bringing the cumulative tally of discovered characters to {T, S,E, *,E, S,L}. All seven characters are ratcheted into place. The 19 new letters are chosen, and the process is repeated until the entire target phrase is found." (Hope this works!) There is no population. It is just a simple partitioned search. Except for the correct letters the daughter generations inherit nothing else from the parents. There is no selection. The algorithm as described by Dawkins works in a completely different way.Indium
August 25, 2009
August
08
Aug
25
25
2009
07:16 AM
7
07
16
AM
PDT
DeLurker, There isn't anything in TBW that would lead one to suspect cumulative selection is anything other than a partitioned search.Joseph
August 25, 2009
August
08
Aug
25
25
2009
07:04 AM
7
07
04
AM
PDT
Indium, There isn't anything in the Dembski/ Marks paper that says they do not select from a population of multiple individuals.Joseph
August 25, 2009
August
08
Aug
25
25
2009
07:03 AM
7
07
03
AM
PDT
Joseph#177
There doesn’t have to be any explicit latching/ ratcheting. The program does it as a matter of course.
Yes, without explicit latching, with a population, and without mutating every incorrect letter every generation. Dembski and Marks have completely mischaracterized the Weasel algorithm.
The way Dawkins describes cumulative selection and uses the weasel to illustrate the only inference is one of a partitioned search.
You need to look up the definition of partitioned search and compare it to what Dawkins describes in The Blind Watchmaker. You will find that your statement is incorrect.DeLurker
August 25, 2009
August
08
Aug
25
25
2009
06:56 AM
6
06
56
AM
PDT
Joseph, the Weasel algorithm selects the next parent from a population of multiple individuals. The algorithm of Dembski and Marks does nothing of the sort.Indium
August 25, 2009
August
08
Aug
25
25
2009
06:50 AM
6
06
50
AM
PDT
DeLurker, There doesn't have to be any explicit latching/ ratcheting. The program does it as a matter of course. The way Dawkins describes cumulative selection and uses the weasel to illustrate the only inference is one of a partitioned search.Joseph
August 25, 2009
August
08
Aug
25
25
2009
06:48 AM
6
06
48
AM
PDT
The wording of "cumulative selection" along with the weasel illustration is more than enough to infer latching/ ratcheting takes place. BTW the weasel program is an example of a targeted search.Joseph
August 25, 2009
August
08
Aug
25
25
2009
06:46 AM
6
06
46
AM
PDT
Joseph#167
Latching takes place regardless of whether or not it was programmed in.
Yes! That's the whole point! Mutation followed by selection, without explicit knowledge of which letters are correct, results in significant improvement over random search. I already noted this in my comment #133. There is, however, definitely no explicit latching in the algorithm.
Just because Dawkins isn’t smart enough or honest enough to understand that isn’t our problem.
If you are going to accuse Dr. Dawkins of dishonesty, you should provide proof.
The way he describes cumulative selection and uses the weasel program it is clear that latching is a big part of the process.
No, it is not. This has been made painfully obvious by BillB and Indium in this thread and at the websites I have referenced. I note that both you and kairosfocus have avoided the core issue here: Dembski and Marks have misrepresented Dawkins' algorithm despite being repeatedly corrected on this point. The Weasel algorithm is not a partitioned search.DeLurker
August 25, 2009
August
08
Aug
25
25
2009
06:45 AM
6
06
45
AM
PDT
Indium, The weasel program starts with a "population" of one. Each generation starts with a "population" of one.Joseph
August 25, 2009
August
08
Aug
25
25
2009
06:43 AM
6
06
43
AM
PDT
Joseph: Fair point, a random walk is not a search at all if there is no target, and my use of language there was imprecise. A random walk is a targeted search if it is being used to find a target. A GA is a population based hill climber, WEASEL is a GA where the highest up the hill gets to generate a new population, which end up spread about on the hill close to the parent. The proximity to the parent is determined by the mutation rate. Dembskis algorithm proceeds by randomly sampling the search space, checking to see if any individual letters match, and if they do those letters are fixed, whilst the rest continue to randomly sample the search space. Whilst it is possible to measure the population of a village as one, the use of the term in the field of computational search strategies implies that a number of individuals are used concurrently. A population of one in this context renders the use of the word meaningless, and makes all search strategies population based. Your objection is a semantic red-herring - the English language is complex and nuanced, please try and pay attention to the context.BillB
August 25, 2009
August
08
Aug
25
25
2009
06:10 AM
6
06
10
AM
PDT
I calculated some probabilities for the explicit or implicit latching. -kf A propos explicit or implicit: the adjustable weasel works with an explicit mutation rate, commonly 4%, I suppose. Dembski's weasel is another animal, it has an implicit mutation rate which is much higher (my guess, an average 50% - 60%)DiEb
August 25, 2009
August
08
Aug
25
25
2009
05:58 AM
5
05
58
AM
PDT
Joseph: Oh, so you call 1 individual a "population" and choosing this one individual for the next generation "selection"!?Indium
August 25, 2009
August
08
Aug
25
25
2009
05:48 AM
5
05
48
AM
PDT
Joseph, The wording in the Blind Watchmaker gives no hint of latching. A video of Dawkins presenting the algorithm 1987 (same algorithm, different parameters) shows no latching. Dawkins says there is no latching. Latching is not needed for the algorithm to work. The algorithm is more complicated when it uses latching. Explicit latching is not something biologists would implement when modelling evolution: Mutation rate is supposed to be independent of the resulting fitness. The only real argument FOR latching I have seen is the fact that no mutation of correct letters is shown in the BW tables, which is easily explained by the fact that only the best members of a few generatios were shown. There is no reason at all to believe one should see fitness reducing mutations in this case. One other “argument” that I find particularly funny proposes that it is “consistent” with the Blind Watchmaker to assume explicit latching. Of course it is! It is also consistent with the BW that Dawkins and kf have 4 arms and 7 legs. The other funny thing is that the implicit latching is exactly what was supposed to be demonstrated by the algorithm: The combination of random mutation and selction can result in an extremely (almost partitioned) search. Somehow I get the feeling that we have been at this point before. Anyway: Did you notice that we have more or less moved on from the latching issue anyway?Indium
August 25, 2009
August
08
Aug
25
25
2009
05:46 AM
5
05
46
AM
PDT
Indium, You claim that Dembski/ Marks don't use a population. Yet they have at least one individual which mutates. All it takes is one to make a population. And I believe they also use selection-closest to the target. That is what a partitioned search is all about.Joseph
August 25, 2009
August
08
Aug
25
25
2009
05:37 AM
5
05
37
AM
PDT
BillB:
DAWKINS IS DESCRIBING A POPULATION BASED RANDOM WALK (A GA)
A targeted search is not a random walk. And Dawkins uses a targeted search. And he uses his example to show how cumulative selection is better than a random walk.Joseph
August 25, 2009
August
08
Aug
25
25
2009
05:34 AM
5
05
34
AM
PDT
KF, you are weaseling around. I am not talking about some software on the EIL website. I am talking about the algorithms as presented in the Blind Watchmaker and in the new article from the Drs Dembski and Marks. Do you now finally see that they are very different in important aspects or not? No population vs population No selection vs selection Complete randomization of incorrect letters vs inheritance of wrong letters. The list is even longer. So, from your previous posts I see that it is impossible for you to admit this error (which is mainly the error of Dembski and Marks). But I guess everybody else has now understood the differences well enough to make a judgement "on the merits". Cheers!Indium
August 25, 2009
August
08
Aug
25
25
2009
05:34 AM
5
05
34
AM
PDT
DeLurker, Latching takes place regardless of whether or not it was programmed in. Just because Dawkins isn't smart enough or honest enough to understand that isn't our problem. The way he describes cumulative selection and uses the weasel program it is clear that latching is a big part of the process. That was the difference between random and cumulative. Now Dawkins could have been clear and said that cumulative means that anything can happen but that would have destroyed his whole premise.Joseph
August 25, 2009
August
08
Aug
25
25
2009
05:30 AM
5
05
30
AM
PDT
kairosfocus#144
Dawkins says there is no latching.
–> The latching (as the onlooker can easily enough see for him-/her- self) is highly evident from the o/p and description c 1986.
Are you accusing Dr. Dawkins of lying?DeLurker
August 25, 2009
August
08
Aug
25
25
2009
05:14 AM
5
05
14
AM
PDT
Hmm, I realised I really couldn't let this piece of bigotry stand without comment:
BTW, the only rape accused I have ever known and attended the trial of, was having to deal with an older woman who threw herself at him, then cried rape when things went sour. She discredted herself in the courtroom by showing up in stilettos, stockings and a tight micro-mini. Believe it or not.
Do you honestly believe that when a woman wears clothes that you disapprove of, an accusation of rape made by her should be disregarded?BillB
August 25, 2009
August
08
Aug
25
25
2009
04:51 AM
4
04
51
AM
PDT
Indium: You have already been aswered in adequate details. And it is ATOM who has authored the EIL's adjustable weasel, as already linked. Good bye. GEM of TKIkairosfocus
August 25, 2009
August
08
Aug
25
25
2009
04:50 AM
4
04
50
AM
PDT
Onlookers: One last note. BillB has just now shown through further turnabout accusations that he is not engaging in a rational discussion on even the tangential matter to the primary issue: the advance in design theory marked by the publication of the paper on the implications of active information and associated ideas. Good day. GEM of TKI PS: Off-topic, but important enough to footnote, pardon Clive: despite much faux or manipulated outrage and many confidently parroted talking points to the contrary, if you the interested onlooker are interested in the truth on some of the less savoury but important historical influences of Darwinst science (including e.g. in the USA through the eugenics movement -- watch the fairground lights display very carefully), you will find the lecture here of some significant interest. We had better face and learn from history, if we wish to progress rather than regress. [HT: BA 77.]kairosfocus
August 25, 2009
August
08
Aug
25
25
2009
04:48 AM
4
04
48
AM
PDT
1 2 3 4 5 9

Leave a Reply