Shermer in The Matrix
by MikeGene
Jason Rosenhouse provides a nice summary of the debate between Bill Demsbki and Michael Shermer. What I most appreciate is Rosenhouse's summary of Shermer's talk, which apparently was organized into five basic points.
Being a selfish Gene, I'm curious as to how Shermer's arguments would stack up against The Matrix. Would Shermer's arguments pose a pointed and deadly attack in the Matrix?
Or would they, as arguments for the non-telic viewpoint, be soft and harmless?
Below the fold is my assessment:
"First up: Before we say something is out of this world, make sure that it's not in this world. The point was simply that you don't glom onto a fantastic, supernatural explanation when a natural one will suffice."
"Point two was that the burden of proof lies with the ID folks. If we are to accept ID as an explanation, some positive evidence in its favor is required. It is not sufficient to just make criticisms of evolution."
"This led into point three, which was that ID folks commit the either/or fallacy. That is, they act as if the only options are either Neo-Darwinian evolution or ID."
"But the bulk of the talk centered around point four, which is that evolution is etablished not by any one fact, but by a large collection of facts from disparate fields of study."
"The final point was the vacuity of supernatural explanations. Invoking unspecified, all-powerful designers just doesn't get you anywhere when you are trying to explain the natural world."
Summary:
Looks like someone's goin' need a new weapon.
February 18th, 2007 at 10:11 pm
Excuse me for appearing so daft, for surely I am missing something. But how exactly does a sword, a "balloon sword", five more balloon swords, and a popped balloon serve to refute your summary of Shermer's points?
Comment by jeff_alexander — February 18, 2007 @ 10:11 pm
February 18th, 2007 at 10:35 pm
Comment by thesciphishow — February 18, 2007 @ 10:35 pm
February 18th, 2007 at 10:50 pm
Yeah, I don't think Shermer's points would have any affect on the substance of my lecture this coming Saturday, either. I would agree with all of them, as I assume Mike's book does, as well.
Comment by Bilbo — February 18, 2007 @ 10:50 pm
February 18th, 2007 at 10:55 pm
I just don't see anything new coming from the evolutionary camp as far as philosophical arguments against ID/YEC paradigms. I've said it once and I'll say it again, a belief in evolution that keeps one from receiving the Truth of the Bible is a stronghold by the biblical enemy and his cohorts, and must be addressed in prayer as well as revelation of Truth to the individual who is being lied to. Be blessed!
Comment by Wayson — February 18, 2007 @ 10:55 pm
February 18th, 2007 at 11:22 pm
Are you planning to blog about this when its over?
Comment by MikeGene — February 18, 2007 @ 11:22 pm
February 18th, 2007 at 11:23 pm
What would you want to know?
Comment by Bilbo — February 18, 2007 @ 11:23 pm
February 18th, 2007 at 11:27 pm
I'd bet readers would be interested in the basic outline of your talk, how it was received, any questions you fielded, etc.
Comment by MikeGene — February 18, 2007 @ 11:27 pm
February 18th, 2007 at 11:30 pm
I sure would! §;o)
Comment by Joy — February 18, 2007 @ 11:30 pm
February 18th, 2007 at 11:40 pm
I can give the basic outline, now. That way you can all give me advice or criticism (even you ID critics) ahead of time.
I. Introductory remarks
A. Credentials: Little science, BA in philosophy, two full years of addiction to ID websites, plus 3 additional years of part-time addiction.
B. I believe that ID proponents should not try to get ID taught in public schools.
C. I accept common descent.
D. My hesitancy to accept Neo-Darwinian mechanisms for most of evolution is due more to Margulis than to Behe.
E. I don't see how science can be done without methodological naturalism.
F. Nevertheless, I maintain that biology can incorporate ID, just as other fields of science do.
II. Some fields of science where intelligent causation is used:
A. Cryptography
B. Criminal forensics
C. Archaeology
D. SETI. This last example is interesting, because we would conclude existence of ETI, without independent evidence of ETI. However, we would know the technology used, and ETI is considered plausible by the scientific community.
E. What sort of evidence is needed, when the designer is not considered plausible, and the technology is unknown?
It's late. I'll finish outline tomorrow.
Comment by Bilbo — February 18, 2007 @ 11:40 pm
February 19th, 2007 at 12:23 am
Mike,
I have again tried to purchase your book via Arbor vita. They now accept credit cards directly, but they rejected mine (which I've been able to use elsewhere recently). When it becomes available at Amazon, I probably will be able to purchase it.
I just thought I'd let you know, that I've kept trying to no avail. Sorry
Sal
Comment by Salvador T. Cordova — February 19, 2007 @ 12:23 am
February 19th, 2007 at 2:19 am
Are we talking about The Design Matrix, The Matrix, or both?
Comment by nickmatzke — February 19, 2007 @ 2:19 am
February 19th, 2007 at 2:34 am
Salvador wrote:
You should probably check your credit limit. Word may have gotten out that you're a YEC, which is never good for a FICO score.
Comment by keiths — February 19, 2007 @ 2:34 am
February 19th, 2007 at 3:36 am
jeff_alexander asks:
Jeff,
Mike isn't claiming to have publicly refuted Shermer's points. He's simply hinting that his yet-to-be-published book will do so.
Of course, the real question is not whether Mike's book can withstand Shermer's balloon-sword. After all, Shermer is a charitable guy, and I'm sure he chose the weapon that would do the least damage to ID proponents while still ensuring a decisive victory.
If Mike's argument is better than the prevailing pro-ID arguments (and I think we're all hoping it is), then the critics may have to trade in the balloon-sword for a Nerf sword, or even a cardboard one.
Comment by keiths — February 19, 2007 @ 3:36 am
February 19th, 2007 at 5:38 am
Hi Jeff,
It's not so much a "refutation" of Shermer's points as it is that these points are ineffective criticisms within the DM. For that matter, points 1, 3, 4, and 5 are impotent against the views I have long publicly laid out on my web page and here for years. For example, consider Shermer's points against the backdrop of my views. POP! goes his balloon.
What's basically left is point #2 "“ the need for a positive approach that goes beyond criticism of evolution. Sure "“ it's call The Design Matrix: A Consilience of Clues.
Comment by MikeGene — February 19, 2007 @ 5:38 am
February 19th, 2007 at 5:56 am
Hi Bilbo,
Nice summary. As for the SETI example, some of my recent blogs about this might be useful (then again, maybe not). Here, here, here.
And you do an excellent job of framing the issue, ending with, "What sort of evidence is needed, when the designer is not considered plausible, and the technology is unknown?"
Exactly. The designer-centric approach mandates that we first establish the plausibility of the designer and specify its methods. But the design-centric approach begins by acknowledging these handicaps and asks how to proceed from there.
[BTW, consider the conventional view that life was spawned by geochemistry. No one has established the plausibility of this event, as evidenced by the fact that the range of opinion from mainstream scientists is that such an event was fantastically improbable (a luck accident) to inevitable (determinism). And of course, everyone is still clueless about the method/mechanism (a point recently underscored by Shapiro trying to get people to "back to the drawing board").]
Comment by MikeGene — February 19, 2007 @ 5:56 am
February 19th, 2007 at 10:20 am
MikeGene, "What's basically left is point #2 "“ the need for a positive approach."
To some extent I debate this point. Though for ID to be recognized, it does need to make a positive case. However, darwinism has been elevated by many to the status of "fact". Yet darwinism can provide no answer at all to first life, and cannot provide a good answer to a miriad of questions, such as IC. The status of darwinism: fact, theory, hypothesis, or error should be determinable independant of whether ID is established. Personally, as a complete explanation for all of life's diversity, I would not rate darwinism above hypothesis.
Comment by bFast — February 19, 2007 @ 10:20 am
February 19th, 2007 at 1:17 pm
I think that point 2 is actually the most dubious of the five. Consider that the appearance of design is so strong that even those who deny design typically don't do it consistently. Quite often the same person will at times say that Darwinism has disproven design, and at other times that it explains design (Shermer himself has earned a few frequent-offender miles on this count). And those who manage to avoid the "d" word typically still fall into using the "f" word ("function") which is the same thing.
Now, normally the best explanation for why something appears designed would be the most intuitive and straightforward one: that it is, in fact, designed. This is, after all, how we learn about human design, by noting that the results of human actions appear designed (It's not like we look directly into people's psyches through mind reading, and sense their thoughts as they design things). However, in the case of biology, Darwinists believe they have something that debunks design (or explains design, or whatever it is any particular ateleologist holds to at any given moment). The point of the negative ID arguments is to argue that the attempts to either explain away or "naturalize" design (again, depending on what the ateleologist holds to at the moment) don't work, so that the explanation defaults to the more intuitive explanation, which is that design is, well, design (or that design is real - again, depending on whether the ateleologist is attempting to deny or "un-design" design).
So the problem with the "no positive evidence" mantra is that it's really just the ateleologist taking the strong appearance of design and going "I refuse to count that at all, and I have a problem with anybody else who does!". It's an attitude that, if applied consistently throughout science, would be the end of the whole enterprise (All our evidence for the existence of physical laws, for instance, ultimately boils down to our intuition that things look like they were caused in a lawlike manner. Someone who refused to allow that as positive evidence could insist that physics was just one big "negative argument" against things looking that way by chance, and that there was no positive evidence for laws at all).
Now, where there is some trace of validity to the argument is that for any particular hypothesis about the method or implementation of design (such as front-loading, or whatever) you need some reason to favor that particular explanation over others, in order for the explanation to be rational. But as an attempt to show that there's categorically no reason to consider design in the first place, it's totally off-base.
Comment by Deuce — February 19, 2007 @ 1:17 pm
February 19th, 2007 at 4:49 pm
Deuce wrote:
This appears to be nothing more than a semantic misunderstanding with regard to the word "design."
There are actually two entirely different senses in which evolutionary biologists use the word "design," and much confusion has resulted from this.
"Design" may refer to the classical argument from design, in which case it is referring to something designed by a supernatural being, or by a "mysterious something" in the case of ID.
But "design" is used in another way, and that is in reference to evolutionary design. A Darwinist might say, "an owl's wing feathers are designed to reduce noise which could alert its prey." What he means is that the process of variation and natural selection over the course of millions of years has resulted in feathers which were gradually more effective at catching prey and which lead to present-day owl feathers. So this use of the word "design" is just a shorthand way of speaking about this process.
During a debate between a Darwinist and an ID advocate, there is going to be a nearly unavoidable collision between the two meanings. You are right that it may appear downright confusing or hypocritical at times, but nonetheless it is just a semantic difference in the use of the word.
Comment by jeff_alexander — February 19, 2007 @ 4:49 pm
February 19th, 2007 at 6:30 pm
This is the most common response, but the two senses are definitely not entirely different. Consider that it's not just the word "design" that that ateleologists use, but pretty much the entire teleological lexicon ("design", "function", "purpose", "meant for", "selection", etc) when it comes down to actually describing biological phenomena! There is obviously some reason for this correlation - some reason ateleologists cannot just use completely different language to describe their "completely different" concepts. At best, one might try to argue that their use of these terms is metaphorical rather than literal, but unrelated they are not.
The problem is, I could take any physical object, say an ordinary mound of dirt, and point out that this mound is the way it is today because of billions of years of gradual modification, and I could refer to this modification as, say, "natural choosing" or what have you, that led to the surviving dirt pile I see today. The "choosing" part would just be me using a teleological word or metaphor for a non-teleological reality, of course, but the same is supposed to be true of the "selection" in natural selection. Some genomic patterns end up in greater numbers, and some in fewer or no numbers, due to a variety of physical causes, but the ones in greater numbers aren't literally "selected" any more than the mound of dirt.
Of course, nobody would apply teleological language to the mount of dirt, even though it can be described that way, but they do for biology. Hence, we need a more robust, principled rationale for why the language is applicable in biology, but not dirt-mound-ology.
Here's why I think ateleologists use design language. When we look at the biological world, we get a strong intuition that biological things have functions; functions which are independent of our minds (that is, they aren't just imposed by us), and which therefore require an explanation independent of our minds. It's very similar to the intuitions that tell us things are behaving in a lawlike manner, resulting in the discovery of physical laws in other branches of science. This intuition is the very reason biology exists in the first place - it's what biologists are trying to explain. Ateleologists share the intuition, and likewise assume that it signals something essentially independent of our minds that requires an external explanation, but they reject the explanation that comes with it - that it is designed. They sort of accept the intuition as half-real. But because the intuition is inseperable from teleological concepts, they are stuck using teleological terms when describing the biological properties that they are trying to explain because of the intuition. The mound of dirt doesn't trigger this intuition, which is why we don't apply teleological terms like "selection" to the string of physical causes that "selected" the mound of dirt.
Comment by Deuce — February 19, 2007 @ 6:30 pm
February 19th, 2007 at 7:40 pm
Or it could be that humans are rather unimaginative and reluctant to invent new vocabulary when the present one suffices.
Also, as you mention, humans think in metaphors. Consider the language of mathematics: groups, rings, fields, normal groups, norm, normals, normalizers, orbits, stabilizers. None of these concepts have anything to do with their real-world counterparts, short of metaphor. And the metaphors do help in some cases, particularly when talking about orbits and stabilizers.
How about the language of programmers: objects, messages, hooks, virtuals, closures, continuations, construct, destroy. These also are names which evoke helpful images or metaphors, but have no mystical connection to their real-world counterparts.
So I take your point that "completely different" are not quite the right words. It's not a crucial point, but there it is.
Sure, you could set up a dirt-pile experiment. Take a hundred dirt piles and apply some random modifications like sprinkling each with water while wearing a blindfold. Then select 1 dirt pile based on some criteria, say, resemblance to Margaret Thatcher. Now make 100 dirt piles identical to the Margaret-Thatcherish one you selected (this would be difficult in practice). Repeat. After, say, 40 generations, you'll end up with a dirt pile which bears an uncanny resemblance to Margaret Thatcher.
This is the what Dawkins did with computer-drawn line figures for which he selected successive generations to look like insects. The key is to have selection criteria acting on faithful-but-not-perfect replicators. He also did a similar word experiment which yielded a Shakespearean phrase.
Your dirt pile example does not describe an evolutionary algorithm, however.
I would: "Here is a dirt pile which was designed, using an evolutionary algorithm, to resemble Margaret Thatcher. … Here are the 20th and 21th generations. See how the earlobes are being pushed up here? Look at how the nose is trying to get straighter, departing further from the 'Margaret Thatcher With A Broken Nose' which first appeared on the 15th generation."
"Oh, and here we have a human. This specimen was also designed, using an evolutionary algorithm, to survive on a planet which has an atmosphere composed mainly of nitrogen and oxygen…"
Comment by jeff_alexander — February 19, 2007 @ 7:40 pm
February 19th, 2007 at 11:26 pm
I like Shermer, as he is a rather likeable fellow. But his arguments just smack of, how shall I say it, …. SPIN.
Comment by chunkdz — February 19, 2007 @ 11:26 pm
February 20th, 2007 at 12:07 am
Jeff's "Thatcher piles" reminded me of a fascinating visual illusion, the Margaret Thatcher Effect.
Truly mind-bending.
Comment by keiths — February 20, 2007 @ 12:07 am
February 20th, 2007 at 11:54 am
Being a software engineer myself, this example hits home. There are some points worth making here. You are correct that objects, hooks, closures, etc do not exist in the physical world. They exist as abstracta, and we have representations of them in our minds. But they do have a connection to their real-world counterparts. We modify the physical world such that it represents the abstracta that we are thinking about, and in so doing manipulate it into acheiving our goals. That's what programming is: creating physical representations of abstract objects (objects, loops, algorithms, etc) in a computer's memory to get a computer to do what we want.
The abstract objects don't actually exist in the computer's memory - it's just silicon and electrons in various configurations. These physical entities simply represent the abstract objects to the programmer who configured them. Nevertheless, often times a different programmer can come along and examine the program written by the first programmer, and in so doing they can intuit the abstract entities (the algorithms, the objects, etc) that the physical reality was intended to represent, even if they never met the first programmer. That, in a nutshell, is what it means to detect design: Detecting design is when you examine something in the world, and figure out that it is the way it is partly (or wholly) because of representations in another mind.
Well, no, but neither does Darwinian evolution, or at least it's not supposed to. Algorithms, including evolutionary algorithms, are abstracta. They don't exist in the physical world, but we do have mental representations of them, and we can modify physical objects to represent them to us. The Margaret Thatcher dirt pile and Dawkins line drawing examples do involve algorithms. Both examples involve programmers with concepts of algorithms in their minds, modifying physical reality to represent those algorithms to them. The whole point of the original dirt pile example, and of Darwinian evolution, is that there were no mental representations directing their outcomes, and hence no algorithms. We can impose algorithms on both to try to approximate their history, but in neither case are algorithms actually supposed to have contributed to their history - nobody actually manipulated them to match their representations.
In the Thatcher dirt pile example we use teleological language because looking at the pile triggers an intuition that it was intentionally caused by another mind's representation of Margaret Thatcher. If we look at the previous generations it will likely trigger an intuition that it was also caused by another mind's representation of a genetic algorithm. Presumably, we'd be right on both counts. The design language in this case is quite literal, and is meant in the traditional sense of "design".
That you see this example as parallel to Darwinian evolution of humans sort of illustrates my point. The two senses of "design" are very related, even arguably identical. Biological entities trigger the same kinds of intuitions as the Thatcher dirt pile - intuitions of another mind's representations. Like the Thatcher dirt pile, these intuitions also tell us that an explanation is needed for how it was done. Darwinism officially disowns the representations, but nevertheless maintains the conviction that an explanation is needed for how it was done (I think this actually renders it incoherent, but that's an argument for perhaps another day). But since the whole idea that there is something in need of explanation in the first place is tied to the intuition of representation and intent, ateleologists are stuck using teleological language when describing the things they wish to explain in biology. Even if they did come up with their own vocabulary, it wouldn't make any sense unless they mapped the meanings of their new words to the meanings of teleological words, which would defeat the shmurpose:wink:.
Anyhow, I think this all fairly well makes my point about positive evidence for design - the positive evidence is the same intuitions that fuel the search for explanations in the first place, which have resulted in millenia of philosophical debate over teleology going back to the Greeks and before, and which compel even ateleologists to borrow the lexicon. The negative ID arguments are actually meant to be negative arguments against a negative argument against that ancient positive evidence. If the ateleologist simply flat-out refuses to grant that the teleologist can even count that as evidence, then there's no place on which real discussion can even begin between the two of them.
Comment by Deuce — February 20, 2007 @ 11:54 am
February 21st, 2007 at 2:07 pm
Deuce,
When I mentioned that your dirt pile example does not describe an evolutionary algorithm, I was merely making a technical point. Your example does not fit the criteria of an evolutionary algorithm. In particular, selection is not being used to weed out successive "species" of dirt piles.
Since you are a programmer, you might appreciate this. It is well over ten years old, but since I remembered the title it was the first thing for which I searched. The important thing to note is that the creatures invented their own strategies of locomotion; the programmer only set up the rules.
Evolutionary algorithms have a different flavor than the mundane and predictable algorithms to which we are accustomed. The "twist" is in redirecting the output back into the input, which in conjunction with heredity, variation, and selection, produces the most unexpected and marvelous results.
We are already familiar with the peculiarity of using the output as the input, and that is in the bootstrapping of a compiler. We have a compiler which generates machine code (or VM code, or whatever) based on human-readable program text. But how did the compiler come about? The compiler is itself executable machine code! The compiler creates executables, but who created the compiler executable?
The answer is that the first compiler was bootstrapped. You start by hand-writing a primitive assembler, then use the primitive assembler to make a better assembler, then use the assembler to make a primitive compiler, then use the primitive compiler to make a 10% compiler, then use the 10% compiler to make a 20% compiler, all the way up until you have a 100% compiler.
Of course, nowadays there are such a plethora of hardware that we can simply use one machine to make the compiler for another (a cross-compiled compiler). But nonetheless we should keep in mind that we are riding the coat-tails of those original self-lifting (bootstrapping) algorithms. It is the "origin of life" for those collections of instructions which we call computer programs.
Compiler bootstrapping is not an evolutionary algorithm, but I mentioned it only as an example of the unusual nature of connecting output to back into input, which a key part of an evolutionary algorithm. If you are familiar with Godel, Escher, Bach (and if you aren't then you have much mind-expanding fun awaiting you), it is a bit like the Strange Loops which arise when a system gains the ability to express itself (the compiler compiling itself), after which it falls into the abyss of incompleteness. I am of course referring to Turing's work, which is Godel's Theorem expressed in terms of computer science.
As for the rest of your previous post, you seem to be asserting that algorithms require a "mind" while drawing a distinction between the "physical world" and "mental representations" of it. That sounds like a philosophical position, not a scientific one. (In retrospect, I should have noticed sooner that "Telic Thoughts" might be more focused on philosophy than science.) A question you might ask yourself is whether your assertions can be falsified through experiment or evidence. If not, then it is philosophy.
I don't mean this in a pejorative sense. I am reminded of the interpretation of quantum mechanics, which is mostly philosophy. Anyone is free to accept or reject the Copenhagen interpretation in favor of some other one. In a sense it doesn't matter since the mathematics remain the same in each case. However it could matter if a particular interpretation somehow makes a prediction which the others do not.
Likewise, if someone wishes to declare that each replicator mutation was guided by an intelligent designer and/or that each selection was so guided, there is nothing anyone can do to disprove it. One could say the designer acts in a way indistinguishable from natural variation and selection. It is simply an interpretation of the empirical evidence of evolution.
So the task for the ID movement is to produce evidence which demands the ID interpretation. But since there are countless cases of previously unexplained phenomenon being subsequently explained by naturalistic means, it is not enough to point to particular cases which are not (yet) fully understood ("God of the gaps").
From here I'll refer you to my discussion with Bradford, which could only have ended as it did: "Produce the evidence."
Comment by jeff_alexander — February 21, 2007 @ 2:07 pm
February 26th, 2007 at 3:50 pm
Jeff, just getting back to this:
Even this sentence is riddled with teleology. "Criteria" are goals or desires that persons have. "Used to" is teleological also. Persons use things to accomplish goals. Ateleological things, on the other hand, simply happen by definition. They don't try to reach goals, and hence don't use things in order to reach them.
But anyhow, this is wrong. You could, in fact, draw up an evolutionary algorithm that accurately described the dirt pile's history. Each change to the pile could be abstracted as a "mutation". Every minute you could refer to the dirt pile as entering a "new generation", and each minute that it didn't change from the previous minute, you could say that the current environment, or the "selection criteria", favored it staying the way it was, and so on. Silly and counterintuitive, perhaps, but in principle no different from imposing an algorithm on anything else that wasn't made with an algorithm in mind.
Now, if there actually was an algorithm, so that you're not imposing one, but actually discovering someone else's algorithm, that's different. The debate between teleology and ateleology here is really a debate between realism and constructivism, where the teleologist holds that we're discovering and intuiting real representations (algorithms, functions, designs, etc) and the ateleologist arguing that we're simply imposing our own (at least if the ateleologist is familiar enough with the logic).
Ultimately, programs using evolutionary algorithms are extensions of ourselves (like all computer programs). We use them when we know what criteria we want, and we know how to get from point A to point B, and we know the criteria and the steps we must take well enough to automate it all, but we just don't have time to go through all the configurations and do the tuning directly. We use a computer to do the calculations for us by proxy, and to return the information to us in a way that will represent a correct answer to us.
Yes indeedy. Or rather, I think that scientific beliefs are ultimately a species of philosophical belief. Take any belief you have, be it scientific, historical, whatever, and ask yourself why you believe it. Take your answer and ask yourself why you believe that, and so on recursively. Eventually you will come to your basic beliefs, your first principles, things you believe because they appear to be self-evidently true, but which you don't have outside evidence for, which are therefore philosophy by your definition, and which all your other beliefs are based on.
In this case, I think it's pretty self-evident and non-controversial that algorithms are abstract, and that physical processes are described with algorithms rather than being algorithms. And also that there's a distinction between physical reality and our representations of it (otherwise, we would never have to modify our representations of reality, or adjust our algorithms).
Of course, you can dispute this, and claim that algorithms are physical. Aside from being nonsensical (what color is an algorithm, for instance? How much does one weigh, and where can I find one?), there's the fact that calling algorithms physical is itself a philosophical argument, just as much as saying that they are not. The only other alternative is to claim that algorithms don't exist at all, which is also a philosophical argument.
However, to talk about algorithms at all, you must use one of these definitions of the word. By your reasoning, that would mean that talking about algorithms at all is philosophy and not science. Do you want to defend that view?
Comment by Deuce — February 26, 2007 @ 3:50 pm
February 27th, 2007 at 9:13 pm
Again, this is not an evolutionary algorithm. The selection criteria must have the ability to ruthlessly kill off dirt pile mutations which don't measure up. That is how evolution is able to play the odds in its favor, sometimes yielding increased complexity. What you describe is exactly backward, with the mutations deciding the selection criteria.
Previously, I made it clear that I was correcting you on a technical point regarding what an evolutionary algorithm is. By disregarding my factual correction, by not taking a little time to understand what an evolutionary algorithm is, by making uninformed assertions ("But anyhow, this is wrong"), and by obliviously launching into another confused example, you present yourself to others as an abject crackpot.
To get started on these concepts, some essential books are:
Godel, Escher, Bach
The Selfish Gene
The Extended Phenotype
Especially take some time to read GEB; it will be well worth it.
Comment by jeff_alexander — February 27, 2007 @ 9:13 pm
February 28th, 2007 at 12:06 am
Jeff,
The "abject crackpot" comment is out of line - there's no need for this discussion to degenerate into name-calling. And the "go read some books" argument gets tired pretty quickly. If you're too frustrated to continue the argument politely, let it go.
Deuce,
I just can't see the strength of your argument. Essentially you're claiming that our intuitive view of complex structures such as living organisms as "designed" (which has carried through into the teleological language used to describe them) counts as actual evidence for a teleological viewpoint. Fair enough; this would have been a reasonable argument a couple of centuries ago, when intuition was all we had. But science has moved on since then - we now have actual evidence that can be used to help determine the origins of biological structures.
To me, once actual evidence starts to arrive, intuition becomes obsolete. This is just as true for biology as it is for, say, quantum physics, relativity, or atomic theory, all of which defy intuition to a mind-boggling degree (matter is mostly made of empty space? travelling fast affects your passage through time? objects can be in two places at once? as if!). In each of these cases we carried our intuitions with us until evidence started to arrive; when it did, scientists abandoned their intuition and argued their points based on the actual evidence.
You can make the point that relativity is better-established than evolutionary theory, and I would largely agree with you (although the inconsistency between QM and relativity indicates some major unresolved problems in at least one of those fields). Nonetheless, we are now in a position to argue the case for and against teleology in biology using solid physical evidence. Our intuitions have become irrelevant; they certainly don't count as "evidence" in the same way that the results of well-controlled experiments do. If we have learnt anything from the progress of science, it should be that the universe doesn't operate according to intuitive rules, and this is just as true in biology as it is in physics.
Comment by Mesk — February 28, 2007 @ 12:06 am
February 28th, 2007 at 7:30 am
Mesk,
My comments would have been out of line if they were given the first time I corrected Deuce, but not the second time. Let's be clear what I said: I didn't simply call him an abject crackpot; after listing several reasons, I said that he was presenting himself as an abject crackpot. That is absolutely true. Any professional colleague will immediately recognize the pattern of stubborn ignorance and consequent failure to participate rational conversation which characterizes crackpot behavior.
Deuce should be informed of how he comes across; let's be adults and not withhold vital information for fear of hurting someone's feelings. I'm sure Deuce can handle it.
Mesk, where is your intellectual honesty? If a colleague gets caught up in primitive tribal/xenophobic emotions and launches into a racist tirade, you would waste no time in pointing out his irrationality. Likewise, when a colleague gets caught up in stubborn ignorance (another strong emotion), you should also correct him. Let's not sacrifice rational dialogue by pandering to niceties.
I wish I had made this clear previously, but I used crackpot not in the "wacky idea" sense—wacky ideas do not make one a crackpot—but in the "stubborn ignorance" sense, e.g. the Dunning-Kruger effect.
Comment by jeff_alexander — February 28, 2007 @ 7:30 am
February 28th, 2007 at 7:44 am
Jeff:
You don't seem to have a good feel for a) the internet and b) human nature. You are advocating that we travel down the road toward flame wars. If we must go there, then please be intellectually consistent and tell us if you think Richard Dawkins is a crackpot.
Comment by MikeGene — February 28, 2007 @ 7:44 am
February 28th, 2007 at 8:50 am
Mike,
You might have a point if the sentence you quoted had not been preceded by the previous sentence which you did not quote. If someone makes racist comments, I will respond forcefully without worrying too much about niceties. I am trying to make the same case for intellectual honesty. If someone is clearly being willfully ignorant by refusing to learn the simplest terms of the discussion, we should not hold back our disapproval. Willful ignorance should be "on par" with racism with regard to scientific discussions.
This is in the context of the foregoing conversation which went something like this:
Person A: "An equilateral triangle is interesting because [etc etc...]
Person B: Aha, but take an equilateral triangle whose sides are different lengths, and we get [etc etc...]
A: "No, that's not an equilateral triangle. It must have sides of equal length."
B: "But anyhow, that is wrong. We can take an equilateral triangle and say one of its sides is longer than the others, therefore [etc etc...]"
A: "No, that's not an equilateral triangle. By persisting in your ignorance of equilateral triangles, you are presenting yourself to others as an abject crackpot."
Regarding the original conversation, "evolutionary algorithm" was a technical term which was being similarly misused, and it was crucial to the discussion.
What do you think the proper response for Person A should be? You might say that Person A should walk away, but this would not be the usual response if Person B was making racist remarks. Why should it be different in the case of willful ignorance?
I don't immediately see how Richard Dawkins is relevant here. In any case, I will repeat that I didn't call Deuce a crackpot. I said he was presenting himself to others as one. And I stand by my decision to bring this to his attention, as you all should.
Comment by jeff_alexander — February 28, 2007 @ 8:50 am
February 28th, 2007 at 10:11 am
Jeff,
There's no hard line in the sand, and I'm sure you can re-interpret your remarks to mean anything you like. But in my view it's best to err on the side of politeness. That means simply saying "I disagree with you, and here's why." If you reach your frustration threshold before convincing the person that they are wrong, state that the conversation doesn't appear to be constructive and walk away.
As you no doubt are well aware, as soon as the first insult is thrown on an internet discussion forum (however carefully phrased it is) the conversation rapidly degenerates into reciprocal flaming - all heat, no light. It's a pointless exercise.
These aren't mutually exclusive options. Indeed, in general the latter is required to generate the former.
Comment by Mesk — February 28, 2007 @ 10:11 am
February 28th, 2007 at 10:32 am
Jeff,
I don't care how you rationalize your behavior. Don't use language like that or find another blog to comment on.
Comment by Krauze — February 28, 2007 @ 10:32 am
February 28th, 2007 at 10:40 am
Mesk,
After I explained why Mike took my sentence out of context, why did you quote the same sentence?
Why have you also chosen to characterize me with emotive terms like "frustration threshold" It is completely obvious that I was responding matter-of-factly in every post here. What exactly do you believe I have "re-interpreted"
I will repeat once again that I made no ad hominem insults; I did not call Deuce a crackpot; I outlined several factual, concrete reasons why he was presenting himself to others as one.
Comment by jeff_alexander — February 28, 2007 @ 10:40 am
February 28th, 2007 at 10:48 am
Jeff: Let's not sacrifice rational dialogue by pandering to niceties.
Jeff, look at this from a practical perspective. Even though Mesk and I disagree more often than not, I read what he writes and consider it carefully. So do most others. It comes down to interacting with others in ways that advance an objective which in this case, I hope, involves a search for truth. Politeness is not weakness. It is usually a sign of strength indicating that one feels secure enough not to have to engage in fisticuffs even if provoked.
Comment by Bradford — February 28, 2007 @ 10:48 am
February 28th, 2007 at 10:50 am
Krauze,
Can you be more specific about what you find objectionable about my behavior, and what you feel I have "rationalized" What is the "language" to which you are referring? Honestly, I don't understand what is going on here; I am unable to heed your warning because I don't understand it.
Comment by jeff_alexander — February 28, 2007 @ 10:50 am
February 28th, 2007 at 11:13 am
Jeff is presenting himself to others as an abject ass.
Comment by chunkdz — February 28, 2007 @ 11:13 am
February 28th, 2007 at 11:23 am
jeff:
As I just said to KK ['el kleinarino'] in the ID=Creationism thread, these tactics simply won't work. The implied "or else" to this type of advice is not that impressive. As if, were we to take your advice or klein's on how to define ourselves, you would suddenly be disposed to take us seriously. If you can't do that based on what is presented here then your opinion is of no concern on this end.
Comment by Joy — February 28, 2007 @ 11:23 am
February 28th, 2007 at 11:28 am
chunkdz:
If you gave specific, concrete, factual reasons why I am presenting myself as such, then you'll make a fair point.
Everyone knows that "You are acting stupid," or "You are presenting yourself as stupid in this forum for these reasons," mean something utterly different than "You are stupid."
Comment by jeff_alexander — February 28, 2007 @ 11:28 am
February 28th, 2007 at 2:03 pm
Without descending to the level of childishness just put on display here, jeff, I will just point out that I know perfectly well what an evolutionary algorithm is, and that you are the one who is confused.
As before, your confusion is that you do not discern the difference between algorithms and the objects and events that are described with or represented by algorithms.
Let's say that I want to build a better widget, and I decide to write an evolutionary algorithm to aid me. I have an idea of what I want my widget to be like - requirements for the traits that the new widget needs to have. These requirements are my criteria for what I want selected, otherwise known as the selection criteria. I then formulate an evolutionary algorithm that is designed to arrive at my selection criteria, and I encode that algorithm into a computer program. Then I run the program. The program eventually results in a bunch of different colored pixels on my screen - in a pattern that represents to me a diagram of a widget that meets my criteria.
Now, the running program is not my algorithm. It's mostly just a bunch of electrons being rapidly pushed through a bunch of registers. However, the program represents my algorithm to me, and carries out my algorithm for me. The program's behavior can be described with my algorithm. Furthermore, the program doesn't have any selection criteria. The program couldn't care less what it's outcome will be. I'm the one who has criteria. The program (or parts of it) represents my selection criteria to me. It arrives at output that meets my criteria, because it was front-loaded to do so, but it isn't my criteria.
Do you understand the distinction so far?
Now let's say that another programmer has written a program to build a better widget. I watch the program run. It prints out a bunch of diagrams of different widgets, and then stops. The widgets in the successive diagrams become, gradually, more and more like the widget in the final diagram. Now, the program and these diagrams are not themselves an evolutionary algorithm. They're a bunch of electrons and ink. However, I could infer that an evolutionary algorithm was used by another programmer to make this happen. I could formulate my own evolutionary algorithm that describes the behavior that I see. In so doing, I could take traits of the widget pictured in final diagram, and represent them as corresponding to someone's criteria. I could represent the program as carrying out those criteria.
Perhaps the algorithm and criteria I would come up with to describe the program's behavior would, in fact, be similar or identical to the actual algorithm and criteria employed by the original programmer. In that case, I haven't so much constructed an algorithm and criteria as I've discovered an existing algorithm and criteria.
On the other hand, let's say that nobody programmed the computer. The program was written by a feat of luck when a magnet got too close to the computer's hard drive. In that case, there is no algorithm and no selection criteria. I could still construct an evolutionary algorithm that described the behavior. I could imagine some "selection criteria", and represent the program as if it were intended to reach those criteria, even though it really wasn't. My algorithm might correctly describe the computer's behavior, but both the algorithm and the criteria would be things that I constructed and imposed on the program to conceptualize it a certain way.
Do you still get the distinction?
Now, back to dirt piles and evolution:
No, in my example the "selection criteria" cannot "ruthlessly kill off" dirt pile mutations that don't "measure up". But the "selection criteria" don't "ruthlessly kill off" anything in organic evolution either, because in both examples there are no selection criteria in the physical environment.
Yes, organisms come into existence with varying traits. Some happen to get copied into new organisms, some don't, and this is largely caused by environmental conditions. But that doesn't mean that the environmental conditions are "selection criteria", anymore than the environmental conditions that allow the dirt pile to persist unchanged for several minutes are "selection criteria".
Rather, we can draw up an algorithm to model the behavior of the physical phenomena that we see, and we can represent or model the environmental conditions as if they were implementing some criteria. But we could also, if we were so inclined, represent the environmental conditions that allow the dirt pile to remain unchanged for several "generations" as if they were implementing some criteria being used to "select" the dirt pile.
Wrong. The mutations don't "decide" anything in either example. It's the person developing the algorithm who decides what to represent as selection criteria and what to represent as mutations. There's no hard distinction in biology either. DNA is part of the environment too, and can be modeled as part of the "selection criteria" when attempting to draw algorithms to describe phenomena. You're just annoyed because the things I chose to represent as selection in the algorithm I described are counterintuitive to you.
Finally, even as you get yourself worked up, you also continue to demonstrate my original point for me. You say, "The selection criteria must have the ability to ruthlessly kill off dirt pile mutations which don't measure up." Take note of the "ruthlessly". From an ateleological standpoint, some organisms reproduce, and others don't, and various environmental causes affect which ones do and which ones don't. It just happens. It makes as much sense to say that the environment is "ruthlessly" killing off organisms that don't meet its "criteria" as it does to say that the sun is "ruthlessly" warming the earth, or that gravity "ruthlessly" pulls masses together.
Ruthlessness is an intentional quality that persons have. In order for something to be ruthless, it must be a person, a mind, with goals - aka desires, aka criteria - who is willing to employ any means to achieve those goals.
By using the term "ruthlessly" you reveal that you yourself understand, on some level, the connection between criteria/evolutionary algorithms and persons/minds. You yourself conceptualize algorithms using teleological (aka intentional) concepts, not just teleological language. You then conflate these algorithms and criteria with the impersonal processes that can be described by them, which is where you go wrong, but the basics are there.
Comment by Deuce — February 28, 2007 @ 2:03 pm
February 28th, 2007 at 2:14 pm
Here is a past post of mine about evolutionary algorithms. Saying that evolution can explain the appearance of design is like saying that chemistry can explain the appearance of arson. In one sense, chemistry can explain the origin of the fire. But in another sense, if it appears as though arson has occurred, it will require more than just a description of physical causes in order to explain it.
Comment by macht — February 28, 2007 @ 2:14 pm
February 28th, 2007 at 2:47 pm
Jeff, don't argue, just don't say it again.
Comment by Guts — February 28, 2007 @ 2:47 pm
February 28th, 2007 at 3:28 pm
Hi Mesk,
I disagree that we've abandoned our intuitions in coming up with new theories that replace our old theories, even when the new theories seem "strange" in some way. What's really happened is that some of our intuitions have been superceded by other intuitions.
Actually, that's not quite right either. What has really happened is that we come up with new theories that are compatible with all of our intuitions instead of just some of them, unifying more and more intuitions.
If this doesn't sound very descriptive, let me give an example. Let's say that you believe the earth is flat. Your reason for believing this is that when you look down, you don't see any roundness. You don't discern anything but overall flatness. Absent any reason to believe that there is more to it than this, you just assume that the world is in fact flat. Your reason for thinking that it's flat is simply that it appears flat, as far as you can tell. This base appearance is what I mean by an intuition.
But, then one day you go to the beach. You see a boat slowly disappear over the horizon and come back. Suddenly, you put it all together. The whole world isn't flat as you'd assumed. It's just that the slope is too gradual for you to discern from limited distances. What reason to you have to believe that the world isn't flat? That the boat went over the horizon and came back. But what reason do you have to believe that the boat really went over the horizon and came back? Well, just that the boat appeared to go over the horizon and come back. In other words, your new belief that the world is *not* flat is also based on a base appearance, ie an intuition!
Now, prima facie, it may seem like a new intuition has replaced a prior, erroneous intuition. But that's not really what has happened. The appearance of flatness was not an illusion. It's just that we had previously extrapolated too much from it. It was always logically compatible with the appearance that the earth had a degree of non-flatness that was too subtle for you to notice over a short distance. In fact, if you conclude that the appearance was simply an illusion, that your senses were just lying to you, then you have no reason to trust the appearance of the boat going over the horizon either! Such a conclusion means plunging yourself into a radical skepticism and self-doubt, from which you cannot reach any conclusions about the earth's shape at all! What has actually happened is that you've come up with a conclusion that satisfies both of your intuitions, even though it contradicts some of the things you had extrapolated from the first intuition on its own.
And so it goes with all of science. As I mentioned previously, you can take any scientific belief, or really any belief at all, and ask yourself why you believe it. Then, ask yourself why you believe that, and so on. Eventually, you will get down to base appearances (intuitions), upon which everything else is based. Simply tossing our intuitions aside is very dangerous. We must assume that our intuitions tell us the truth about reality, or else we must give up science altogether. Without them, there is no science. In fact, there is no knowledge at all.
In fact, in your post, you provided one more good example of intuitions at work. You said, "the inconsistency between QM and relativity indicates some major unresolved problems in at least one of those fields". Now, how do you know that QM and relativity are inconsistent? Well, they appear to be inconsistent. You have an in-built intuitive sense of logical consistency, and from that you have a logical intuition that they contradict each other. And note that based on that intuition, you conclude that there is something about reality that we have wrong. You assume that your intuitions tell you the truth about reality!
Mike's recent post about evidence is instructive here. Raw data - mere objects - are not evidence by themselves. Data becomes evidence when it appears to someone and is interpreted by them. Evidence is a function of base appearance, and hence of intuition. If our intuitions can't be relied upon, then neither can evidence.
Comment by Deuce — February 28, 2007 @ 3:28 pm
February 28th, 2007 at 3:43 pm
Deuce,
Since you don't believe me, I would suggest describing your dirt pile example to a knowledgeable third party and ask if it fulfills the requirements of an evolutionary algorithm. He/she will say no. There are specific requirements, to which I've already alluded; it's not a matter of opinion. You are free to define some other classification, but you are not free to redefine what an evolutionary algorithm is.
Regarding "there are no selection criteria in the physical environment," etc, that appears to be an untouchable philosophical position. We may very well be cogs in a grand evolutionary algorithm which we have come to call the Universe, conceived in the mind of The Designer and implemented by His intentions. It remains to be seen whether this or similar ideas will make any practical contribution to science.
As I mentioned in my earlier post about the common use of metaphors to describe abstract concepts and properties such as mathematical groups, rings, and fields, I don't believe the use of metaphors warrants any special metaphysical or mystical interpretation. It's just a useful aid in conveying information. Indeed, I regard such over-analyzing of metaphors as a type of flawed reasoning by analogy.
Comment by jeff_alexander — February 28, 2007 @ 3:43 pm
February 28th, 2007 at 4:10 pm
Jeff:
But so would the counter-argument to this. Selection criteria in the physical environment is erroneously believed to exclude an intelligently devised evolutionary algorithm. If I understand Deuce correctly there is no way to distinguish whether the applicable algorithm describes the effects of blind physical forces or directed ones. For the purpose of imputing and detecting design it does not matter.
Biological systems have features that are unique with respect to other branches of science. It may very well be that reminding ourselves that there is an appearance but not a reality to design is counter to reality as well as results.
Comment by Bradford — February 28, 2007 @ 4:10 pm
February 28th, 2007 at 4:12 pm
I agree that the dirt pile itself doesn't fulfill the requirements of an evolutionary algorithm, it being a physical thing. My claim was that it should be possible, in principle, to tailor an evolutionary algorithm that would describe the dirt pile's history, by deciding in post-hoc fashion which things would represent our "selection criteria", and which would represent "mutations", and so forth.
Comment by Deuce — February 28, 2007 @ 4:12 pm
February 28th, 2007 at 4:35 pm
Perhaps, but only in the sense that "1 is not 2" is an untouchable philosophical position. By that I mean, there are no selection criteria in the physical environment by definition. The normal definition of "criteria" is a teleological one. A person's criteria are their goals, what they intend to achieve. So the problem with saying that there are selection criteria in the environment isn't primarily that it's wrong, but that it doesn't have any meaning. It's like saying that coffee is cheerful, or that the number 3 is bored. How am I even supposed to conceptualize that? On the other hand, saying, "we can develop an evolutionary algorithm, in which the selection criteria will be represented by the environment" - now that makes sense.
Or to put it another way, using the normal definition of the word, it only makes sense to say that the environment has "selection criteria" if you are conceiving of the environment as a person with intentions or goals. But it's precisely the point of the ateleological position to try to account for life without the need of a person with intentions or goals.
Comment by Deuce — February 28, 2007 @ 4:35 pm
February 28th, 2007 at 4:39 pm
The great thing about EAs is that you can see the difficulties of getting a large IC system from the accumulation of small random variations, because finding high fitness peaks in such a landscape is like finding needles in a haystack.
Comment by Guts — February 28, 2007 @ 4:39 pm
February 28th, 2007 at 5:38 pm
It has occurred to me that Evolutionary Algorithms and Evangelical Atheists have the some acronym. One of them is going to have to budge!:lol:
Comment by Deuce — February 28, 2007 @ 5:38 pm
February 28th, 2007 at 6:14 pm
Deuce,
Would it be fair to say that your arguments are proximate to the speculation that we are in The Matrix?
Comment by jeff_alexander — February 28, 2007 @ 6:14 pm
March 1st, 2007 at 2:29 am
Deuce wrote:
Deuce,
We can't assume that our intuitions tell us the truth about reality. They are self-contradictory. If intuitions tell us the truth about reality, then we are forced to concede that reality itself is ultimately unintelligible.
It's true that some intuitions must be accepted as fundamental, such as logic itself. Science could not proceed without it. But other intuitions must be shed, as Mesk says, for science to progress.
Two of my favorite examples of faulty intuition:
1. In a room containing just 23 people, what are the odds that at least two share the same birthday? Intuition balks at the idea that the answer is greater than 50-50, but it is.
2. Adelson's checker-shadow illusion. Intuition tells us that squares A and B cannot possibly be the same shade of gray, even if we are familiar with optical illusions and have learned to distrust our vision to an extent. (Be sure to click on proof and explanation to convince yourself.)
In each of these cases, we are tempted to go with our immediate intuitions. Further analysis reveals that to accept these intuitions as true would require us to reject intuitions that are even more fundamental, such as logic itself. Most of us wisely demur and stick with logic.
To a surprisingly large extent, humanity's advances in knowledge can be viewed as a competition of intuitions. Most are jettisoned in favor of views derived from a few fundamental intuitions, with the goal being a satisfactory and coherent view of the world based on a minimal set of intuitions.
There is some danger in tossing intuitions aside, as you say, but assuming their truth uncritically is even more dangerous, and ultimately futile.
Comment by keiths — March 1, 2007 @ 2:29 am
March 1st, 2007 at 9:00 am
Hi keiths:
For my opinion with regards to intuitions and their replacement, look at the little story I sketched out above about finding out the world isn't flat. I think the realization that you only need 23 people before it is statistically probable that two share the same birthday is a pretty good example of that. It's not that we simply toss out the typical intuition that it's 183. It's rather that that number tells us how many people are needed before it's likely that someone shares our birthday. We had previously concluded too much from it. What I think really happens in most cases is that we clarify our intuitions, and come up with conclusions that unify intuitions that only seemed to be competing.
Comment by Deuce — March 1, 2007 @ 9:00 am
March 1st, 2007 at 11:14 am
Hi Deuce,
You seem to be saying that every intuition is correct, but that we sometimes apply an intuition to the wrong situation. To me, that sounds like telling the math teacher that all of your answers on the exam were correct — they just weren't the answers to the questions the teacher was asking.
I can't buy it. The intuition in the birthday scenario is that the odds of a match are low with 23 people in the room. That intuition is wrong, and gets replaced by a completely different notion: the odds are better than even with 23 people in the room.
Even if were to accept your explanation for our intuitions about the flatness of the earth and the likelihood of shared birthdays, how would it address the checker-shadow illusion?
Comment by keiths — March 1, 2007 @ 11:14 am
March 2nd, 2007 at 12:33 am
Deuce, you wrote,
So, if I understand you correctly, your position is that the employment of a teleological lexicon and metaphors by biologists constitutes actual evidence of design.
If that is your position, wouldn't the use of nonteleological language and metaphors by biologists constitute evidence against design as well?
Comment by genomic — March 2, 2007 @ 12:33 am