We’re not in Kansas Anymore

You’re argument seems to be that since there is more than one fitness landscape, then there are no fitness landscapes

My own understanding is that there are no destinations, only survivors on the journey. But since Laughable made the Point, itatsi in demo mode, can be switched from one language landscape to another in the middle of a run, and the new fitness function will seamlessly begin shaping the population to look like the new language. Before anyone makes the obvious point, the language landscape is much more forgiving than the biological landscape. Nearly any random string has embedded substrings that appear in words. It's a game, not biology. But it makes the point that an evolutionary algorithm does not need a fixed target. Petrushka

Laughable,

To say that there is one ‘fitness landscape’ that the evolutionary search is working on is a bogus notion.

You're argument seems to be that since there is more than one fitness landscape, then there are no fitness landscapes...... Clive Hayden

DiEb @ 40:

d I don’t think that the definition of active information is very helpful, as it only looks at the size of an underlying set Ω but not on the functions – which are generally in the focus of the NFLT.:

WinstonEwert @ 45

The active information is calculated from the relative probabilities of search algorithms succeeding so I’m confused by your statement here.

In your paper, you are only looking at asymptotically perfect algorithms, i.e., algorithms which actually find the target. For those algorithms, the active information equals the endogenous information, a fancy way to state that it is lb(|Ω|) - so it's only depending on the underlying set and it's the same for all algorithms. DiEb

Atom, you must realize that in the material world the 'fitness landscape' varies significatly depending on place to place and even from time to time. The fitness landscape in say, LA, would be quite different from 500m underwater in the Atlantic ocean. Generally people, and say, small mammals, moths and bacteria, are the most fit creatures in a fitness ladscape of a urbanised area. Put them 500m under the ocean though and their fitness/survival rate drops to about 10 minutes. As a more realistic example of changing fitness landscapes that could be used in terms of early evolution, take for instance: (a) 1 mile under the sea (b) 1 foot under the sea, (c) floating on top of the ocean and (d) on the beach. All four of these places will have completly different fitness landscapes due to oceanic pressure, exposure to oxygen, exposure to light, etc. These fitness landscapes will also change depending on the availability of resouces, the presence of predators. The same life form placed at each of these 4 locations will 'hill climb' to the nearest local 'hill' over time to better suit that environment (if it survived at all). To say that there is one 'fitness landscape' that the evolutionary search is working on is a bogus notion. Dawkins used the Weasel program as a simple example of how a cumulative search works faster than a complete random search, he did not use the Weasel program as a complete model of how the biological world operates. Laughable

Petrushka, I've tried to make the same point. As a computer science research program, creating metrics for comparing search procedures may have some interest. But I don't see it edging closer to invalidating evolution as a process. What happened to the claim that evolutionary algorithms work as well as they do because an intelligent agent smuggled in the information? Which of the researchers whose names are on these papers will raise their hand and say, "I did it. That algorithm only works as well as it does because of me." Which of them did the smuggling when they ran Avida? If design is the thing to get to, eventually, why put so much importance on the fitness function? Does the Designer design in nature by the subtractive process of killing? If I was trying to prove design is important, I'd be focusing on the representation and the variation operators. Nakashima

...as Winston pointed out, these simulations and toy problems use carefully selected fitness functions amenable to the type of search we are performing. In short, we choose the type of landscape we need for success then assume nature has this exact type of landscape.

The only absolutely necessary characteristic of a biological landscape -- in order for some level of evolution to occur -- is that not everything that differs from its parent dies. Aside from that, the efficiency of evolution is irrelevant. We already know that evolution often fails to produce the changes necessary to avoid extinction. We also know that some seemingly simple islands are difficult or impossible to reach. Tame foxes with valuable, soft fur, for example. Which is why I question the metaphor of target. When children who are slightly different from their parents manage to live and reproduce, the population changes. It may be because the children are more fit, or it may be simply because they survive. Petrushka

Atom, Hans was a sock puppet that I've banned many times previously, so I banned him or her again. Sorry for letting him/her ramble on. :) Clive Hayden

Hans, I'm not trying to walk you through the paper. I'm only trying to help you ask the right questions. The paper deals with the Hamming oracle as an information source. It shows that this information can be extracted with varying levels of efficiency and that it is a LARGE source of information. That is the main gist, as Winston pointed out earlier. As to questions regarding biological systems: it is not our claim that these hand-crafted programs represent accurately how unguided nature is supposed to operate. Others make these claims. However, as Winston pointed out, these simulations and toy problems use carefully selected fitness functions amenable to the type of search we are performing. In short, we choose the type of landscape we need for success then assume nature has this exact type of landscape. However, the type of landscape we need to achieve success in a search cannot simply be assumed to be given to us. Many types of landscapes are possible; many of them (the majority) are not helpful. Now here's the point: narrowing down the set of possible fitness landscapes to only those that allow quick convergence requires target knowledge. We need to know which target we're searching for to select a fitness landscape that will work. You've grasped that in your last comment, asking "what is the specific target?" The correlation between fitness landscape and target location is essential. Furthermore, such a selection incurs an information cost. How much? Greater than or equal to the amount of active information you can extract from the landscape. Therefore, evolutionary algorithms, as presented in the toy problems, are not sources of unlimited, free functional information. They simply extract the information we provide, with varying levels of efficiency. That's the relevance to ID. Atom Atom

Will any and every fitness landscape / fitness function give an evolutionary search an advantage over blind random sampling? Or are some fitness landscapes better than others, in terms of convergence to a given target?

I think perhaps that "search" is the wrong metaphor. I wrote itatsi in part to see if a sufficiently knowledgeable oracle could lead toward islands of high fitness. One of the interesting observations that resulted is that different languages present different landscapes. French, for example, seems to have many words connected to other words by one or two character mutations. German does not. I assume that biological structures share this characteristic. Some are close and some distant. Behe seems to assert that some gaps cannot be traversed. My thought on this is that RMNS doesn't always find a target. Species go extinct, even though from a purely physical standpoint there would seem to be solutions. Perhaps something as simple as a change in behavior. I also wonder if some of the structures we do see might be accidents, that might never be repeated if the tape were rewound. the metaphor would be the lottery winner. Looking back, the odds of any particular person winning, or any particular structure evolving might seem insurmountable. If we attempted to replicate the evolution of a complex structure it might never happen. But that does not mean it didn't happen once. Petrushka

In all honesty, Winston has been rather lucid in his description of the core problem...

I'm having trouble understanding the core problem and how it relates to biology. It seems to mee that the biological oracle is omniscient in the sense that biochemistry dictates whether an organism lives after its DNA mutates, and the ecosystem selectively prunes populations. The relevant question is the one raised by Behe, and that is, can you get there from here. Assuming there can be two stable configurations, one having some enhanced functionality, are all the intermediate configurations viable? In biology, that strikes me as an experimental question. Efficiency doesn't seem to be the issue. Viability does. Petrushka

Hans, Thank you for answering my question. So one search, using a fixed set of parameters (mutation rate, population size, etc) and a specific fitness function can find a target quickly whereas a second search using the same algorithm and parameters but a different fitness function will not converge to the target at all. What differs, and makes the difference in search performance, appears to be having a landscape amenable to hill-climbing, with our specific target located near the apex of a hill. Now, we must ask, what fraction of all possible fitness landscapes have this form, namely, one suitable to finding our specific target? It isn't enough for it to simply be a smooth single hill landscape, because the maxima could be at a point other than our target, leading our search away from the target rather than towards it. So how many "good" fitness functions (in terms of fraction of total possible landscapes of fixed size) are there for a given target? This bears directly on whether we can just pick a landscape at random from the space of all (size-limited) fitness landscapes and expect to get one usable for our search. I will let you answer before continuing. Atom Atom

Atom, I'm not sure of the relevance of your question. Are you suggesting that the designer acts by changing the fitness landscape rather then interacting with the organism itself? To answer your question however, I would say "No" and "Yes". Hans Fritzsche

Hans Fritzche, Allow me to ask a preliminary question: Will any and every fitness landscape / fitness function give an evolutionary search an advantage over blind random sampling? Or are some fitness landscapes better than others, in terms of convergence to a given target? Atom Atom

Atom, At the risk of derailing this thread:

You don’t need the mutations to be directed, as long as you have enough offspring, a suitable mutation rate, and most importantly an information rich oracle (fitness landscape.) Hill-climbing, gradient ascent searches need a suitable fitness landscape to have any advantage over random sampling searches, and such suitable landscapes are quantifiably rare in the space of possible fitness landscapes.

There seems to be no requirement for an intelligent designer anywhere there. Where do you propose that the intelligent designer comes into play? After all, that is the point of your paper, this website and the claims made about the paper by some of the authors, i.e. that there is a requirement for an intelligent designer somewhere in the process. Where/when is that requirement satisifed? I've looked over the EIL but it's not clear to me why these papers are "ID supporting" papers. Could you clarify? Hans Fritzsche

Petrushka, You are free to argue about the "rather significant claims made on this website" with those who made them, on the threads in which they made them. I, on the other hand, would rather discuss the issues raised on this thread regarding carefully selected fitness landscapes and information costs associated with landscape bias. In all honesty, Winston has been rather lucid in his description of the core problem and I don't really need to add much to his points. It's just that I'm rather enjoying this thread so I'd rather allow it to stay focused on the current topic. Thanks for your cooperation. Atom Atom

Everyone here is focusing on the “information richness” (really, target proximity correlation) of the fitness landscapes/oracle.

Focusing on the oracle's intelligence ignores some rather significant claims made on this website, if not on this thread. Are you asserting that ID is completely comfortable with random variation plus selection, provided the selector is intelligent? I could have sworn that many ID supporters assert that there are occasions when living things produce a favorable mutation as a result of need. Either by computing the necessary adaptive mutation (front loading) or by preserving some environmentally induced adaptation (epigenetics?). Petrushka

Petrushka wrote:

So the itatsi oracle knows how close a random string is to a word, and it can also select the string from a population that is closest to being a word.

Which is exactly Winston's point. You've conceded the argument without realizing it. Directed versus non-directed mutations are a secondary issue. You don't need the mutations to be directed, as long as you have enough offspring, a suitable mutation rate, and most importantly an information rich oracle (fitness landscape.) Hill-climbing, gradient ascent searches need a suitable fitness landscape to have any advantage over random sampling searches, and such suitable landscapes are quantifiably rare in the space of possible fitness landscapes. This is the issue under discussion. Everyone here is focusing on the "information richness" (really, target proximity correlation) of the fitness landscapes/oracle. Re-read Winston's posts, you may see why your response is irrelevant to the discussion at hand. Atom Atom

Atom @51 and Winston, I believe you are crossing an abstraction layer with your point about location. As Nakashima says @47,

We don’t! We only reward what worked in the last incremental time step because we have no knowledge of whether a distant target exists or not. The inner loop of the algorithm doesn’t know what previous good population members looked like. It only knows the relative fitnesses of current members.

As part of the ..mechanism.. of the search, only the level of fitness is returned. This is the only thing evolution is concerned with. In a real evolutionary scenario, the fitness function is free to test any characteristic it chooses to, on a generation by generation basis. In a computer model, we could swap fitness functions each generation. This means you cannot know in advance, what will be most fit. Toronto

Winston’s claim is that Itatsi doesn’t just look at whether or not substrings occur, but also assesses fitness on the basis of whether or not those substrings appear in places similar to the places they appear in real words (in a statistical sense).

The cardinal rule of this kind of simulation is that variation is blind. There's nothing about the biological world that requires the oracle to be blind or unintelligent. Darwin compared natural selection to artificial selection, and artificial selection undeniably employs an intelligent oracle. So the itatsi oracle knows how close a random string is to a word, and it can also select the string from a population that is closest to being a word. But nothing is conveyed back to the mutation generator to help it make hopeful mutations. In fact, the actual population wanders around rather than homing in on any single word. In fact, the simple ploy of occasionally killing off the best word and selecting the second best prevents itatsi from getting stuck on a single target. It continues to make new words. Another thing it does is make pronounceable strings that aren't in its dictionary, or in any dictionary. If you google these strings you will find that most of them have been used somewhere by someone because humans love making new words. Not only does itatsi make novel words that are islands between dictionary words, it "knows" when a randomly generated non-word will be pronounceable and interesting to speakers of the language. It knows because the string has good genes. Petrushka

Petruska wrote:

I think you misrepresent what the oracle is doing. It communicates nothing about locations. It merely grades by fitness.

You seem to be subtly claiming that giving information about fitness can in no way communicate information about location, which is incorrect. If I am trying to select, out of the group of all Americans, a person who lives in my neighborhood, I can simply make fitness be a function of locality (namely proximity to my neighborhood), then select based on, guess what, fitness! That "fitness" information is really proximity information. Hamming oracles work in a similar way. Winston's claim is that Itatsi doesn't just look at whether or not substrings occur, but also assesses fitness on the basis of whether or not those substrings appear in places similar to the places they appear in real words (in a statistical sense). If that is in fact what Itatsi is doing, then Winston's criticism is completely valid. Remember, just because we claim that an oracle is simply returning "fitness" infomation, does not mean that the fitness is not strongly correlated to (or based on) some other trait that we are trying to optimize, like locality/proximity. Atom Atom

Mr Ewert, Having a useable fitness landscape is key. Key to what? Nakashima

Whoops, something in my wireless keyboard submitted a post while I was correcting it. Continuing where the error occurred: That is conceptually similar to differential reproductive success. The only thing the oracle does is choose which individual continues its line and which fail to reproduce. The children are all random variants of the selected child from the previous generation. Nothing about the oracle changes the way mutations occur. I do not think it is relevant to consider how "intelligent" an oracle is. The natural ecosystem can be pretty intelligent. Predators are pretty good at selecting the weakest. The only rule that can't be broken is the one forbidding the oracle from influencing how variants are produced. I think you might have misconstrued itatsi as a partioned search. Petrushka

I think if you want to have something that parallels a “natural oracle”, you need something which does not have a fitness function deliberately constructed to tell me about where in the space the targets live.

I think you misrepresent what the oracle is doing. It communicates nothing about locations. It merely grades by fitness. That is conceptually edifferential reproductive success, which Petrushka

That’s true, but I don’t see that it is conceptually unlike a natural oracle that causes differential success of phenotypes. The example you present, Itatsi, is rewarding correct pairs of letters taking into account the positions of those pairs. Put simply, it is rewarding looking statistically like the target. It is clever. For the analogy to work, it would be like having proteins which are rewarded for looking statistically like useful proteins. I think if you want to have something that parallels a "natural oracle", you need something which does not have a fitness function deliberately constructed to tell me about where in the space the targets live. Nakashima, The search algorithm has very little notion of state, but the fitness landscape needs to have the correct form for it to work. ONEMAX and WEASEL have really nice forms. A random fitness function will eliminate the advantage of hill climbing. A suitably defined deceptive fitness function will cause terrible performance on the part of hill climbing. Having a useable fitness landscape is key. WinstonEwert

Mr Ewert, The essence of hill climbing is that we can somehow measure whether or not we are moving towards the target. Hamming distances make that easy. But how are we to know are moving towards a target? We don't! We only reward what worked in the last incremental time step because we have no knowledge of whether a distant target exists or not. The inner loop of the algorithm doesn't know what previous good population members looked like. It only knows the relative fitnesses of current members. I agree that Weasel, like the OneMax toy problem used more commonly in EAs (and by Dr Dembski previously in MESA) is a problem where the variation operators and genotype representation combine to define a hill climbing behavior. But fitness landscapes do not need to be continuous, everywhere differentiable, etc. As you said, they can be deceptive (an area studied in depth by Dr David Goldberg of UIUC). Nakashima

In order to obtain your fitness function you are going to need to either analyze a dictionary (i.e. the list of targets or inject your own knowledge of common word fragments.

That's true, but I don't see that it is conceptually unlike a natural oracle that causes differential success of phenotypes. On the other hand, it is still an analogy, so I wouldn't make grandiose claims for it.

Additionally, I don’t think your proposal will actually work. If we reward common substrings we’ll probably get a word containing several of these substrings which do not go together and thus get stuck in a local optima.

I have an example program in mind. Petrushka

I don’t think that the definition of active information is very helpful, as it only looks at the size of an underlying set ? but not on the functions The active information is calculated from the relative probabilities of search algorithms succeeding so I'm confused by your statement here. Have a look at a third set of functions of size |?| which I will call eben’s spoilers: I think its been clearly established that NFLT wil break for many subsets of the possible fitness functions. What are you attempting to accomplish by demonstrating an example? Your fitness function has been engineered to take me to the target directly. Thats a pretty clear example of information being injected. I have to disagree. ES and other EA would still work better than random generate and test even in the presence of very noisy fitness functions Under NFLT that success only derives from making accurate assumptions about the fitness function. Noise isn't the issue here. The question is whether or not there will be a hill to climb. The hamming oracle puts a very simple hill in, which can be easily climbed. But in other functions there could be so many hills that attempting to climb them won't get you an success. Or perhaps the hills are deceptively placed and actually take you away from a good solution. The essence of hill climbing is that we can somehow measure whether or not we are moving towards the target. Hamming distances make that easy. But how are we to know are moving towards a target? What if, instead of a target, the oracle considered all dictionary words to have value, and selected children for survival based on having substrings that appear most often in actual words? What you propose is essentially measuring closeness to a collection of words. In order to obtain your fitness function you are going to need to either analyze a dictionary (i.e. the list of targets or inject your own knowledge of common word fragments. In other words to obtain active information you'll need prior knowledge about the targets embedded into the fitness function/oracle. Additionally, I don't think your proposal will actually work. If we reward common substrings we'll probably get a word containing several of these substrings which do not go together and thus get stuck in a local optima. We can fix these problems but only by providing more intelligent input into the algorithm or fitness function. WinstonEwert

#37:

Relative reproductive success is way different then measuring how close I am to a target.

What if, instead of a target, the oracle considered all dictionary words to have value, and selected children for survival based on having substrings that appear most often in actual words? You could consider substrings to be alleles, and the children that have the best set of alleles get the highest fitness score. Petrushka

I’m aware that review is far less stringent then would be the case if publishing in a journal. I raised this point with respect to the original post (since part of the argument in the post hinges on providing credentials for Bill Dembski, et al), but my comment was excluded by the moderators. I do think it is important to distinguish between the impact of peer reviewed publications in biological research journals and those in less stringent instruments with far less editorial participation by biologists (like the IEEE). spot48

Reviewers commented on the paper and gave recommendations concerning it, so yes it was peer reviewed. I'm aware that review is far less stringent then would be the case if publishing in a journal. Seeing as I have not yet had the honor of having my work published in a journal you should ask somebody else for any details on how exactly they differ. WinstonEwert

Mr Ewert, No. The hamming distance measures the distance from my position to a target. As Dawkins himself pointed out that is emphatically not what is going on in nature. Relative reproductive success is way different then measuring how close I am to a target. Great, I think we all agree that a paper criticising Weasel on the basis of its Hamming fitness function has focused on the least relevant part of Dawkins' trivial example of cumulative selection. The success of *-ES algorithms or FOOHOA is not based on any “specialness” to the algorithms. The success must be credited to the abilities of that oracle. I have to disagree. ES and other EA would still work better than random generate and test even in the presence of very noisy fitness functions. The specialness of these algorithms is their simplicity and how they model aspects of biological reality, not the particular fitness function of a specific problem. Nakashima

A usually, I'm not glad with the way the word information is used, and I don't think that the definition of active information is very helpful, as it only looks at the size of an underlying set Ω but not on the functions - which are generally in the focus of the NFLT.: We have seen that the char. functions of one certain element and the Hamming distances to a certain element both form classes of functions of size |Ω| Have a look at a third set of functions of size |Ω| which I will call eben's spoilers: Enumerate the N^L elements of Ω and define the function set of functions es(ω,φ) by es(ω,ω) = 0, es(ω,φ) = #ω if ω ≠ φ So, es(ω,...) tells you either: you found the value or you didn't find the value, but it is... . A "search" on this set of functions can be completed in one step - I suppose that's the optimal set of functions when you make a search. Obviously, this set of functions isn't c.u.p. (a minimal c.u.p. set of functions including the es(ω,...) would have (N^L)^2 elements - not that many in comparison with all functions on Ω taking values 0..N^L) DiEb

Mr. Ewert, are you under the impression that your "Hamming Oracle" paper published in the IEEE symposium proceedings was peer-reviewed? If so, please explain how the peer review process for symposium proceedings differs from the peer review process in mainstream scientific journals. Thank you. PaulBurnett

Nakashima-san, Take the next step. Is a Hamming distance oracle an apt model for the biological process of relative reproductive success?

Assuming not, then neither is Dawkins Weasel, worse, Dawkins WEASEL is maketed as how a blind watchmaker would work. The idea is to show how much info the oracle needs to make it perform better than blind search. Without that info, it is blind, and so will WEASEL be blind, and so will Darwinian selection in the wild. That means without an informed source to form the fitness function in the wild, Darwiniain evolution in the wild will not resolve structures that are likened to login/password pairs, and there are many such systems in biology like that. Consider that an un-informed hamming oracle will not solve the passwords to any given computer system. Similar problems emerge in biological systems where the fitness function must somehow have resources and foresight to resolve a password like structure. scordova

Is a Hamming distance oracle an apt model for the biological process of relative reproductive success? No. The hamming distance measures the distance from my position to a target. As Dawkins himself pointed out that is emphatically not what is going on in nature. Relative reproductive success is way different then measuring how close I am to a target. The problem is generally posed as: find the minimum of an unknown function f, knowing that f belong to a class of functions F. Yes, but by restricting F to some subset of all possible fitness functions I have gained some prior knowledge about the fitness functions I will actually have to optimize. That way I can gain performance in the functions I am interested in by reducing performance in the functions that I will never have to optimize. But doing so requires that I somehow know how the set is restricted. Picking F as a random subset of all possible fitness functions will probably produce a non c.u.p. set. So I can almost certainly find a better then random algorithm on that set. The problem is that knowing that F is a random subset does not help produce a smart algorithm. I need to know what is in F. In order to construct a better search algorithm, I need to know. WinstonEwert

Winston, Mustela Nivalis was a sock puppet that I have banned twice before, so this makes a third time. Clive Hayden

cumulative selection does work far better than blind search.

Not if the fitness functions are blind to the target. :-) Thus the "blindwatchmaker" hypothesis is not aided by WEASEL or any similar simulation. It's is therefore an illegitimate and misleading simulation clouding scientific understanding. The IEEE papers are helping set the record straight. scordova

Mr Ewert, The success must be credited to the abilities of that oracle. Take the next step. Is a Hamming distance oracle an apt model for the biological process of relative reproductive success? Nakashima

Suppose that I am given a fitness function from an unknown source and am asked to perform a search on it. The problem is generally posed as: find the minimum of an unknown function f, knowing that f belong to a class of functions F. Under the usual assumptions - i.e., everything is finite - there are more possible classes of functions F which are not c.u.p., than those which are c.u.p - so, generally it does make sense to look for an algorithm which works better than random search. DiEb

The NFLT are formulated for sets of functions which are closed under permutation. The set of characteristic functions of single elements on the set Ω is such a c.u.p. set - so, you can't outperform random search if the only values you get from your oracle are found and not found. The set of the functions of Hamming Distances to the elements of Ω is of the same size as the previous set, but it isn't c.u.p. So, the NFLT doesn't apply. DiEb

Mustela Nivalis, Our posts were hitting a variety of issues in each go, do you mind if I try to focus on one at time? Firstly, we should make sure that we have the same understanding of the No Free Lunch theorems. The actual theorems state that for any given performance criteria the distribution of values from that criteria are independent of the search algorithm when averaged over all possible fitness functions. The same is not true if we restrict or otherwise bias the set of possible fitness functions. For example, if we restrict the set of fitness functions to be only those functions that look like hamming distances to the target, we can clearly produce a much better algorithm as both WEASEL and FOOHOA demonstrate. Suppose that I am given a fitness function from an unknown source and am asked to perform a search on it. What are the implications of the NFLT? I can make no assumptions about the kind of function I am looking at. Under such a circumstance, the best I can hope for is to get the average performance. However, as NFLT shows this is independent of the algorithm chosen, so it doesn't matter what algorithm I choose. (It should be noted that in real life I may well be able to assume that certain types of problems are more common then others and use this bias to my advantage. But in that case I have some sort of information about the problem. I am assuming here that I have no such information) Do our understands align? WinstonEwert

Our intent was not to make arguments about the relative efficiency of hamming string extraction algorithms. The intent is to show how powerful the hamming oracle is. The different algorithms have different levels of success in extracting that information, but the oracle is shown to be very powerful. FOOHOA is very good at extracting the information from the oracle thus demonstrating how much information is available. The success of *-ES algorithms or FOOHOA is not based on any "specialness" to the algorithms. The success must be credited to the abilities of that oracle. WinstonEwert

WinstonEwert at 24, "the weasel algorithm knows, cumulative selection does work far better than blind search." But why? It's nice to see you back on the blog, Winston. What I was trying to determine in my conversation with scordova is why he was claiming that Marks and Dembski had made some sort of "case" against Dawkins' algorithm. Dawkins never claimed that it was more than a toy demonstration of cumulative selection, so there is no case to be made. That being said, your question is interesting on a couple of levels. Rather than repeat our previous conversation at: https://uncommondesc.wpengine.com/evolution/new-peer-reviewed-id-paper-deconstructing-the-dawkins-weasel/ I suggest we continue from my comment 85 there. Make sense? Mustela Nivalis

Mr Ewert, But why? There are several key differences - the comparison of entities against each other as well as against the oracle, and the preference based on that comparison. These points make the weasel algorithm an apt model (in these aspects, at least) for nature. This is the only reason to be interested in it, not because of its efficiency. Criticising Weasel on the grounds of its efficiency entirely misses the point of its aptness. Nature is not efficient. A more efficient algorithm, can easily be a less apt model. Nakashima

the weasel algorithm knows, cumulative selection does work far better than blind search. But why? WinstonEwert

scordova at 22, "What is it that you think the Marks and Dembski paper shows?" That Dawkins Weasel wasn’t a blind watchmaker, but an intelligently guided watchmaker. The selection in Dawkins Weasel was anything but blind, but on the contrary, intelligently designed. Dawkins never claimed otherwise, so this is not a particularly interesting conclusion. Dawkins describes the purpose and limitations of his toy program just a page or two after describing the algorithm: Although the monkey/Shakespeare model is useful for explaining the distinction between single-step selection and cumulative selection, it is misleading in important ways. One of these is that, in each generation of selective 'breeding', the mutant 'progeny' phrases were judged according to the criterion of resemblance to a distant ideal target, the phrase METHINKS IT IS LIKE A WEASEL. Life isn't like that. Evolution has no long-term goal. There is no long-distance target, no final perfection to serve as a criterion for selection, although human vanity cherishes the absurd notion that our species is the final goal of evolution. In real life, the criterion for selection is always short-term, either simple survival or, more generally, reproductive success. If, after the aeons, what looks like progress towards some distant goal seems, with hindsight, to have been achieved, this is always an incidental consequence of many generations of short-term selection. The 'watchmaker' that is cumulative natural selection is blind to the future and has no long-term goal. As anyone who has implemented, or even simply played with an implementation of, the weasel algorithm knows, cumulative selection does work far better than blind search. That's the only pedagogical point of Dawkins' much abused toy program and it is not contradicted by the Marks and Dembski paper. Mustela Nivalis

What is it that you think the Marks and Dembski paper shows?

That Dawkins Weasel wasn't a blind watchmaker, but an intelligently guided watchmaker. The selection in Dawkins Weasel was anything but blind, but on the contrary, intelligently designed. scordova

The "Hamming Oracle" article by Ewert, Dembski, et al, was not published in a "peer-reviewed IEEE publication" - it was published in a collection of papers presented at an IEEE electrical engineering symposium, which had nothing whatsoever to do with biology. "Papers presented at the (symposium) appear in a Proceedings Volume..." - http://www.ssst-usa.org/ The term "intelligent design" does not appear in the article. PaulBurnett

Sal, I would like to contribute, but the moderation policy is quite tiresome: my last comment on this thread is now kept in moderation for more than two hours, while a comment at another thread is still waiting for approval after ten hours. DiEb

scordova at 20, It would seem then that the result is Dembski and Marks paper were not so much wrong, but understated their case. Had they dealt with the real weasel, they would have had an even stronger case against Dawkins. What case? Implementing the weasel algorithm defined in The Blind Watchmaker is a trivial programming exercise and all correct implementations show that the algorithm finds the target phrase. What is it that you think the Marks and Dembski paper shows? Mustela Nivalis

Now, you have lost me completely. What do you think is the case against Dawkins? Perhaps you should re-read the paper in question - it can be easily found at the EIL. DiEb

So, I’ll state for you and the record: partitioned searches will find targets better on average than the real Weasel

It would seem then that the result is Dembski and Marks paper were not so much wrong, but understated their case. Had they dealt with the real weasel, they would have had an even stronger case against Dawkins. They were thus too conservative in their critique. scordova

Thanks DieB. scordova

And now to elaborate: the partitioned search as described in the paper provided an quite uncommon fitness function, which not only gave the number of correct letters (L - HammingDistance), but the position of the correct letters. Dembski et al. then gave a stochastic algorithm for which I calculated the expected number of queries to be E[Q]≈N*H(L), (N: size of the alphabet, L: length of the word, H(k): k-th Harmonic number). In the same paper, they showed that there is a deterministic algorithm which uses N queries at most (it's rather the Hangman game). So, I'll state for you and the record: partitioned searches will find targets better on average than the real Weasel. or, to be more correct: there will always be a deterministic algorithm which finds the target with a lesser number of queries than the Weasel algorithm does on average. DiEb

fbeckwith:

Perhaps I am mistaken, but is it not the case that Richard Dawkins has never published in a peer-reviewed philosophy or theology journal even though his entire book, The God Delusion, is a treatise on theology?

Professor Beckwith, Have you read The God Delusion? It's not a treatise on theology, but a critique of theistic belief.

And if that is the case, then why isn’t that book treated with deeper suspicion and ridicule by those who claim that in principle, and not a consequence of ideology, they maintain the sanctity of peer-reviewed publications?

I'm not aware of anyone who claims that every non-peer-reviewed publication must automatically be treated with ridicule and suspicion. The criticism of Dembski (and other ID proponents) is not that they publish non-peer-reviewed books or articles; it's that despite claiming that ID is science, they have an extremely poor publication record in the peer-reviewed literature.

Also, why is peer-review good? Is it because it contributes to greater accuracy and a closer proximity to the truth? That seems correct.

Yes.

But such goods only make sense when we think of human beings has the sorts of beings that a proper end to which they ought to strive. But the notion of “proper ends” depends on the idea of final causes that these folks reject.

Not at all. If scientists want to approach the truth, then peer review is a good idea. If scientists wanted to avoid the truth, then peer review would be a bad idea. As it happens, they seek the truth, and so peer review is considered a good thing. pelagius

yes. DiEb

Dieb, Does the parititioned search perform better in finding the target than the real (Dawkins) weasel. Surely you can answer with a simple yes. A simple answer to help those less familiar with the subject. :-) Thanks in advance. scordova

Perhaps I am mistaken, but is it not the case that Richard Dawkins has never published in a peer-reviewed philosophy or theology journal even though his entire book, The God Delusion, is a treatise on theology? And if that is the case, then why isn't that book treated with deeper suspicion and ridicule by those who claim that in principle, and not a consequence of ideology, they maintain the sanctity of peer-reviewed publications? Also, why is peer-review good? Is it because it contributes to greater accuracy and a closer proximity to the truth? That seems correct. But such goods only make sense when we think of human beings has the sorts of beings that a proper end to which they ought to strive. But the notion of "proper ends" depends on the idea of final causes that these folks reject. In that case, what could be the grounds for suggesting that Bill Dembski has not fulfilled his proper ends? If it is that the entire scientific infrastructure is a social construction, then the enterprise is a useful fiction--like the rules of Monopoly or Stratego--and thus carries no normative weight. So, Mr. Rosenau, here's your assignment: come up with an intellectually persuasive way to defend your normative judgments based exclusively on efficient and material causes. Good luck. fbeckwith

Not necessarily because if the partitioned search performs better than Dawkins real weasel then a critique that demonstrates the insufficiency of the partitioned search will by implication demonstrate the insufficiency of Dawkins Weasel.

I don't follow this logic. And my argument is that the given rationale for the paper to be pro-ID doesn't hold any longer. Of course, there may be other arguments how this paper is still pro-ID, I just haven't heard those yet. DiEb

The partitioned search as described in the paper Conservation of Information in Search does need quite a different fitness function (or oracle, if this pleases you more) from the weasel, and all the algorithms described in the later papers. The real weasel doesn't find targets well - in terms of the number of queries necessary: Dembski et al. have shown in their last paper that there are deterministic algorithms which are way more efficient. In fact, if we understand the oracle/fitness function thoroughly, we may be able to construct even an optimal algorithm: that's what the Search for a Search is about, I think. The advantages of the weasel are the typical advantages of evolutionary algorithms: 1. it is easy to program 2. it is easy to understand 3. it gets results even when we don't understand the fitness functions completely (think TSP) As I said earlier: it isn't about being the theoretically fittest, it's about being the fittest present - the story of the two hunters, the bear, and the pair of running shoes springs to my mind... DiEb

By the way DieB, can you state for the record whether you believe partitioned searches will find targets better on average than the real Weasel. :-) scordova

So, the authors didn’t critique Dawkins’s weasel – and so, they indeed weren’t making an argument that supports ID….

Not necessarily because if the partitioned search performs better than Dawkins real weasel then a critique that demonstrates the insufficiency of the partitioned search will by implication demonstrate the insufficiency of Dawkins Weasel. scordova

7 Barb Disparaging Christian opponents of UcD in that fashion essentially says that they are delusional when they indicate that they are orthodox Christians. It also shows a huge sense of superiority on behalf of some. Of course one wonders if this does not get close to the unforgivable sin that Jesus talked about. Someone commented that doing this kind of thing on a blog is just fine. Sure in comments from unknown authors one expects that kind of noise but in the main posts No not from a leader of ID and also not from a from a writer of a book about ID. Dave W gingoro

When will we see some ID-related research ? Yes, Dembski, et al have published all sorts of stuff that is anti-Evolution, but when will we see some results that tell us something about Design ? Just suppose it was universally agreed that Evolution is a dead duck, and every biologist in the world had chucked out all related publications, and the debate was over. Now, what can we say about Design ? Graham

gingoro: “Way to work in subtle digs” whereas at UcD the digs are not subtle but say things like EC/TEs are atheists or at best deists." Why is calling someone an atheist a "dig"? Richard Dawkins, Sam Harris, Daniel Dennett, and PZ Myers are outspokenly atheist and believe in evolution. Barb

The quality of some of the objections to ID on this thread are as saddening as the quality of some of Josh Rosenau's objections to ID when we presented our respective papers at St. Thomas University last November. Much like the mistake that Clive Hayden highlights here, Josh also tried to make hay out of the fact that a monograph by ID proponent Paul Nelson hasn’t been published yet, but Josh completely ignored publication of many important ID scientific books and papers by William Dembski (The Design of Life, The Design Inference, No Free Lunch), Jonathan Wells (Icons of Evolution), Stephen Meyer (Signature in the Cell), Michael Behe (Darwin’s Black Box, The Edge of Evolution)--and many others in recent years. (Indeed Paul Nelson was a co-author of “Explore Evolution,” but Josh failed to mention this as well.) Similarly, Josh charged that Bill Dembski has been "reduced to rewriting and analyzing programs originally written in 1980's.” I’m not sure exactly what that means, but it was telling that Josh's presentation failed to acknowledge that Dembski now works with the Evolutionary Informatics Lab, has submitted multiple research papers for publication, and had recently published a peer-reviewed article on evolutionary algorithms. (Dembski has since published 3 more papers since Josh's presentation.) But somehow Josh failed to note Dembski’s research productivity. Taking a similar approach, Josh's partner from the NCSE Peter Hess said in his presentation that ID “does not yet have a working research program.” (A tired objection which we all know is false -- I blogged about this here.) In any case, this all seems to be not just using “glass half-empty” thinking about ID. It's more like “take the glass, pour out all the water, then step on the glass, and then mock the lack of water” attacks on ID. But having spent enough time watching the NCSE's approach over the years, you sadly come to expect these kinds of misrepresentations. Needless to say, we were ready to rebut these misrepresentations and the many students I interacted with at the conference were not persuaded by NCSE's arguments. Casey Luskin

gingoro,

“Way to work in subtle digs” whereas at UcD the digs are not subtle but say things like EC/TEs are atheists or at best deists.

Digs on a blog are one thing, his were from a supposedly academic paper. Clive Hayden

"Way to work in subtle digs" whereas at UcD the digs are not subtle but say things like EC/TEs are atheists or at best deists. Dave W gingoro

As for Efficient Per Query Information Extraction from a Hamming Oracle: This paper is announced here at UncommonDescent as Deconstructing the Dawkins Weasel. Frankly, I have problems to see in which sense it does so... DiEb

off topic: This just up at ENV: Exploding the Darwin-Friendly Myth of Junk DNA http://www.evolutionnews.org/2010/04/exploding_the_darwin_friendly.html The paper: Human genome at ten: Life is complicated Excerpts: Non-coding DNA is crucial to biology, yet knowing that it is there hasn't made it any easier to understand what it does. "We fooled ourselves into thinking the genome was going to be a transparent blueprint, but it's not,,,,, as sequencing and other new technologies spew forth data, the complexity of biology has seemed to grow by orders of magnitude. Delving into it has been like zooming into a Mandelbrot set — a space that is determined by a simple equation, but that reveals ever more intricate patterns as one peers closer at its boundary.,,,, With the ability to access or assay almost any bit of information, biologists are now struggling with a very big question: can one ever truly know an organism — or even a cell, an organelle or a molecular pathway — down to the finest level of detail?,,,, "It seems like we're climbing a mountain that keeps getting higher and higher," says Jennifer Doudna, a biochemist at the University of California, Berkeley. "The more we know, the more we realize there is to know.",,,, The regulation of gene expression, for example, seemed more or less solved 50 years ago., Just one decade of post-genome biology has exploded that view. ,,, "Just the sheer existence of these exotic regulators suggests that our understanding about the most basic things — such as how a cell turns on and off — is incredibly naive,",,,, Even for a single molecule, vast swathes of messy complexity arise,,,, Researchers now know that p53 binds to thousands of sites in DNA, and some of these sites are thousands of base pairs away from any genes.,,,it seems wilfully ignorant to try to understand p53 on its own. Instead, biologists have shifted to studying the p53 network, as depicted in cartoons containing boxes, circles and arrows meant to symbolize its maze of interactions.,,,, "Now, we appreciate that the signalling information in cells is organized through networks of information rather than simple discrete pathways. It's infinitely more complex." ,,,, In the heady post-genome years, systems biologists started a long list of projects built on this strategy, attempting to model pieces of biology such as the yeast cell, E. coli, the liver and even the 'virtual human'. So far, all these attempts have run up against the same roadblock: there is no way to gather all the relevant data about each interaction included in the model.,,, In many cases, the models themselves quickly become so complex that they are unlikely to reveal insights about the system, degenerating instead into mazes of interactions that are simply exercises in cataloguing.,,, the beautiful patterns of biology's Mandelbrot-like intricacy show few signs of resolving. http://www.nature.com/news/2010/100331/full/464664a.html bornagain77

The question is: have those four peer-reviewed papers anything to do with Intelligent Design? When announcing his paper The Conversation of Information in Search W. Dembski wrote

Our critics will immediately say that this really isn’t a pro-ID article but that it’s about something else (I’ve seen this line now for over a decade once work on ID started encroaching into peer-review territory). Before you believe this, have a look at the article. In it we critique, for instance, Richard Dawkins METHINKS*IT*IS*LIKE*A*WEASEL (p. 1055). Question: When Dawkins introduced this example, was he arguing pro-Darwinism? Yes he was. In critiquing his example and arguing that information is not created by unguided evolutionary processes, we are indeed making an argument that supports ID

But now, we read in the errata section of www.evoinfo.org

In the paper Conservation of Information in Search: Measuring the Cost of Success we mistakingly referred to Dawkins' Weasel as a partitioned search. This interpretation of Dawkins's simulation was, however, wrong. Concerning generations, Dawkins writes “The computer examines the 'progeny' of the original phrases, and chooses the one, however slightly, most resembles the target phrase...” Since “phrases” is plural, Dawkins uses more than one child per generation. Thus, our initial interpretation was incorrect. See further discussion on the Weasel Ware research tool page.

So, the authors didn't critique Dawkins's weasel - and so, they indeed weren't making an argument that supports ID.... DiEb