What Multiverse Theory and Intelligent Design Have In Common

Hat tip to Matt Curtis for pointing me to this Jim Manzi post : The Multiverse, Intelligent Design, and Science, Much Ado About Nothing.

I have written on the multiverse theory before, basically making the same point as Manzi.  Manzi uses the term metaphysics to describe it.  In other words, it is a belief about what could be true about our universe, and what could possibly explain all the coincidences we see in the laws of physics.  It does not appeal to evidence, however.  It appeals to faith.  Yet, in the height of hypocrisy, scientists and science magazines continue to feature the multiverse as real scientific theory and dedicate ink and cover articles to it.

What would scientists promote non-science this way?

Simple.  Because it is the only alternative to allowing God in the door.  Therefore, the multiverse theory is more about worldview (i.e. scientism) than it is about making testable, falsifiable, empirically based observations.

Many equate science with scientism which is a category error.  One is a practice.  The other, scientism, is a philosophy.  Scientism is another word for materialism, a worldview that believes that matter is all there is, was and ever will be.  Scientific observation, in this worldview, is the final arbiter of truth.  God, as a spirit, does not exist in this worldview.

Christians, like myself, love science but battle scientism.

Manzi makes the point that intelligent design is a philosophical pursuit, and so is multiverse theory.  If you are going to diss one, then you need to diss both.  Therefore, he is exposing the hypocrisy of many scientists who embrace multiverse.

Manzi is right to point out their hypocrisy.

In my own case, I view all knowledge as valid.  I don't put some knowledge (empirical knowledge) on a pedestal higher than other knowledge (metaphysical knowledge).  Therefore, it is not a big deal to call something "science" versus "non-science".  The categories simply don't matter to me.  I am interested in truth and knowing it, not what kind of truth it is.

Those who ascribe to scientism, however, do care about classifications.  If something is scientific, then it has credibility and is considered a superior form of knowledge to something non-scientific.  That is why it is such a big deal to label intelligent design as pseudo-science or religion or metaphysics ... anything but science.

Well, if classifications are important, then they ought to be consistent and label the multiverse theory for what it truly is.  A form of reasoning that appeals to what we don't observe and can't know.  In other words, "gaps reasoning".  In other words, blind faith :)

Comments

[Paul]

Quoting a quote from the article:

"When I ask Linde whether physicists will ever be able to prove that the multiverse is real, he has a simple answer. “Nothing else fits the data,” he tells me. “We don’t have any alternative explanation for the dark energy; we don’t have any alternative explanation for the smallness of the mass of the electron; we don’t have any alternative explanation for many properties of particles."

It seems clear to me that we should refer to the multiverse hypothesis, rather than theory; there is enough there to warrant further study, but it's a long way from having the substance required of a theory. But it is trying to explain things that we currently see but cannot explain, as in the quote above, while remaining consistent with things we can explain.

That's where it differs from ID. While ID might be trying to explain things we don't yet understand, it does this by ignoring vast swathes of things that we can already observe and explain, in order to justify the conclusions it has already reached.

[matt curtis]

Paul,

You said:

"It seems clear to me that we should refer to the multiverse hypothesis, rather than theory; there is enough there to warrant further study, but it's a long way from having the substance required of a theory. But it is trying to explain things that we currently see but cannot explain, as in the quote above, while remaining consistent with things we can explain."

I have a very limited understanding of multiverse theory (or hypothesis) - in part because of its rather nebulous outlines. However, it seems to me that multiverse theory ("MT") essentially trys to explain the currently unexplained by explicitly rejecting consistency with those things we can explain. In other words, what we observe in our universe as physical constants (speed of light, etc.), do not have to be constant throughout the multiverse. MT appears to expressly reject "consisten[cy] with things we can explain."

"That's where it differs from ID. While ID might be trying to explain things we don't yet understand, it does this by ignoring vast swathes of things that we can already observe and explain, in order to justify the conclusions it has already reached."

I think you need to provide some support for this assertion. Specifically, what "vast swathes of things that we can already observe and explain" does ID theory ignore?

Finally, would you not agree that Darwinism often strays outside of objective, material science by making what are, at least currently, non-falsifiable assertions (e.g. that a reptile-like creature, with sufficient time and the right set of selective factors, can transition into a bird)?

[Paul]

Matt - I'll agree that MT does toy with things that we think we know (which much inquiry does, in its initial stages at least, and I'd concede that to ID gladly). And it's tricky in part because it can easily be a cheat - any fact you don't like can be dismissed by saying that it ain't so in another universe. But at the same time it is still dealing fundamentally in things we don't know, as with the speed of light; we know it's a constant (with qualification), but we don't know why, or why it is the speed it is. A possible explanation for at least part of that is that it can take any value, but once it's taken that value it's fixed. So in this universe it happened to hit upon its current value, but it didn't have to.

I don't want to get buried in defending MT, however, partly because I'm ignorant, and partly because I don't think it's gone far enough to deserve defending. I'd be much happier attacking it, in fact :)

An easy example of what ID ignores is, um, evolution. Specifically it claims that microeveolution is 'true', but that macroevolution isn't. To do this it doesn't demonstrate the mechanism that can act at a distance to keep two colonies of zebra from diverging sufficiently to speciate, but instead it denies the fact (and I use the word carefully) that mutations can create information, for example. Following on from this, common descent is denied, even though it is again a fact.

Finally, Darwinism is impressive in the degree to which it focused on objective, material science, and made very clear what was conjecture as opposed to conjecture. It got many things wrong, of course, because many things at the time were unknown, but I can't hold that against a 150 year old theory. If you can show me where Darwin asserted that birds came reptiles I'd be interested to read more.

[Mr. D]

re: "An easy example of what ID ignores is, um, evolution. Specifically it claims that microeveolution is 'true', but that macroevolution isn't."

ID, like evolution, is a bit of a weasel word. It means different things depending on who you talk to. Apparently, Paul thinks that ID equates to denying evolution.

This lack of understanding, unfortunately, is quite common.

Contrary to ignoring "vast swathes of data", design theory makes inferences based on what we do know ... that objects containing specified complexity require intelligent agents ... we see this truth every day.

What makes MT incredibly lame is that instead of making inferences about what we clearly observe ( fine tuning ), it makes conjecture about something we don't have a shred of evidence for ... the existence of a million ga-jillion universes. And this pish posh makes the cover of legitimate science magazines? It is laughable. It damages their credibility and legimitacy.

[Paul]

There are several problems with your argument, Mr. D. First, 'specified complexity' is an invention built on flawed assumptions, rendering it as technically meaningless as it is linguistically. I do quite like perhaps its most obvious corollary, though; items that do not have 'specified complexity' do not need a creator, hence there is no need to think that a deity created most forms of inanimate matter, which make up almost all of the universe.

Second, fine tuning is an argument in favor of evolution, not ID, though not for me a significant one. That life is so finely tuned to exist in the conditions it finds itself in is most easily explained by the constant iterations of evolution; if life was created by an omnipotent being it would have no need to be finely tuned.

Finally, ID does mean denying evolution, specifically the macroevolution I referred to. I'm happy to apologize if my understanding is wrong, however; can you point me to a notable ID source that believes in macroevolution?

I will agree with your final statement to some extent, though; it's an interesting idea, and it may explain some things that otherwise lack explanation, but it's extremely speculative and doesn't deserve a major place in the scientific literature. Fortunately it doesn't have one.

[tgirsch]

What Multiverse Theory and Intelligent Design Have In Common

I'll bite. Umm, neither one has any place in a high school science class? ;)

[Nick]

Finally, ID does mean denying evolution, specifically the macroevolution I referred to. I'm happy to apologize if my understanding is wrong, however; can you point me to a notable ID source that believes in macroevolution?

Hi Paul,

"Front-loaded evolution" is a speculative variation on ID accepts common descent with modification and most macro-evolutionary processes, while claiming that an ill-defined designer tinkered with the first eukaryotic cell. The main proponent online is the pseudonymous "MikeGene."

My current conclusion is that "intelligent design" is such a vague term that it is useless trying to discuss whether or not it is a valid hypothesis, or whether it is science or philosophy. Poll 10 ID proponents, and you will get 10 completely different answers ranging from front-loaded evolution to young earth creationism. The first thing to do in a discussion is pin down precisely what your interlocutor means by ID, because many variations are mutually contradictory.

Re: Mr Dawntreader's original post: Since I know some biology but very little physics, I will feel free to diss most variations of ID while completely ignoring MT. But I won't complain if he wants to diss both.

[Mr. D]

re: "Poll 10 ID proponents, and you will get 10 completely different answers ranging from front-loaded evolution to young earth creationism."

True. For me, personally, I think in terms of specified, complex information. This is the commonly accepted means by which most humans, from detectives to archaeologists to SETI researchers, infer intentionality (hence design).

re: "I will feel free to diss most variations of ID"

What variation of inferring intentionality do you choose to *not* diss?

[Mr. D]

"I'll bite. Umm, neither one has any place in a high school science class? ;)"

Yet one will get you fired and blacklisted, and one wouldn't generate anything other than a yawn if it were taught in a H.S. class. ;)

Go figure.

[Nick]

What variation of inferring intentionality do you choose to *not* diss?

The theistic evolution to which I subscribe, of course. Some people claim that TE is ID, others claim it's just as bad as full-on Dawkinsian atheism. If you would call it ID, then that's the ID that I don't diss.

I also don't have much of a bone to pick with front-loaded evolution. Its proponents accept most of standard evolutionary theory and are rather coy about the mechanism and timing of design. I don't find it particularly useful, but neither does it seem particularly pernicious.

[Mr. D]

Forget evolution for just a moment, are you familiar with the concept of specified complexity?

Do you think it is valid?

[Nick]

Forget evolution for just a moment, are you familiar with the concept of specified complexity?

familiar, yes. Expert, no.

Do you think it is valid?

Probably not. I lack the mathematical background to critique it in detail myself, but I have been more impressed by the arguments of the critics than the proponents. Are there a lot of examples (like more than two or three) where the calculations of specified complexity have been rigorously applied to specific organisms or biological systems?

Assuming that CSI is a meaningful concept, it's still not clear to me how that leads to a conclusion of design for organisms. Shouldn't the argument run something like this:

1. Human artifacts exhibit CSI
2. Organisms exhibit CSI
3. Nothing else exhibits CSI
4. Therefore organisms are human artifacts

That's obviously wrong, so we have two alternative possibilities:
4a. Organisms are artifacts of some other designer
4b. CSI can originate without a designer.

If there were many classes of objects exhibiting CSI, all with different designers, then 4a might be the obvious choice. But given that we only have human artifacts to work with (N=1), I don't see that the mere fact of CSI points towards one conclusion over the other.

Despite your suggestion to "forget evolution for a moment," could we break the tie by considering all the other evidence that organisms have evolved via Darwinian mechanisms. Perhaps CSI doesn't imply design, if the objects exhibiting CSI have the other unique properties of organisms. Or perhaps CSI is a bunch of hooey.

Here's a question that just occurred to me: does the CSI of a genetically engineered organism differ in any way from a "natural" organism.

[tgirsch]

Yet one will get you fired and blacklisted, and one wouldn't generate anything other than a yawn if it were taught in a H.S. class.

I'm not so sure about that. But in any case, the one that's more likely to get you fired and blacklisted is the one whose own early proponents readily admit is an attempt to bypass church/state separation and reintroduce religion into the science classroom. So it's not exactly an apples-to-apples comparison.

[Mr. D]

So it's not exactly an apples-to-apples comparison.

Wrong. It is an apples to apples comparison. They are both pseudo-science, yet when some pseudo-science is taught, it is no big deal as long as it support scientism as a world view. This exposes the canard that the issue is teaching pseudo-science in the classroom ... never was the issue.

It is all about the dominant worldview. As long as you toe the party line, you can teach pseudo-science. You can even get front page articles on Discover.

[Mr. D]

"Are there a lot of examples (like more than two or three) where the calculations of specified complexity have been rigorously applied to specific organisms or biological systems?"

I honestly don't know.

Shouldn't the argument run something like this:

1. Human artifacts exhibit CSI
2. Organisms exhibit CSI
3. Nothing else exhibits CSI
4. Therefore organisms are human artifacts

I have never seen it formulated that way.

I am more familiar with this syllogism.

1. Things we observe in our universe that have CSI have always been shown to have intelligent agency.
2. Information systems have CSI.
3. Therefore, it is likely that information systems have intelligent agency.

It is always possible that chaos can create an information system, though it has never been observed ... only assumed, in the case of the origin of biological information systems.

CSI seems to work well in SETI research, archaeology, anthropology, and forensics. I fail to see why we should suddenly slam the brakes on inductive reasoning when we venture outside those disciplines ... unless, of course, your worldview dictates that certain inferences are illegal.

does the CSI of a genetically engineered organism differ in any way from a "natural" organism.

Interesting question. Don't know. Not sure what it would prove if it did ... or did not.

[Nick]

Mr D:
I am more familiar with this syllogism.
1. Things we observe in our universe that have CSI have always been shown to have intelligent agency.
2. Information systems have CSI.
3. Therefore, it is likely that information systems have intelligent agency.

Hmm. I'm not sure that is so different from my statement, but I have a few questions:

1. Things we observe in our universe that have CSI have always been shown to have intelligent agency

In this statement, isn't "intelligent agency" synonymous with human agency? The set of "things that have CSI" apparently includes human products and natural organisms. Anything else?

2. Information systems have CSI
Does anything other than information systems have CSI, or is "information system" a simple synonym of "thing that has CSI"

So,
1. [Some] things that we observe in this universe that have CSI have always been shown to have [human] agency.
2 Information systems have CSI
3. ????

I dunno, your syllogism doesn't seem too much different than my statement, once we break it down, and it seems a weak chain of reasoning on which to hang a conclusion.

CSI seems to work well in SETI research, archaeology, anthropology, and forensics.

Well, archaeology, anthropology, and forensics aren't attempting to demonstrate intelligent agency. They're attempting to demonstrate human agency. Since we know humans exist, and we know their capabilities and limitations, the chain of induction is somewhat shorter.

As for SETI, we don't actually know that CSI works well. I guess we can talk again when the SETI researchers find something exhibiting CSI that they attribute to non-human intelligent agency. And of course, SETI researchers aren't looking for an information system that is a population of error-prone self-replicators under natural selection. Organisms have unique features, not shared by human artifacts or by hypothetical SETI messages, which seem highly relevant the issue of agency.

I fail to see why we should suddenly slam the brakes on inductive reasoning when we venture outside those disciplines ... unless, of course, your worldview dictates that certain inferences are illegal.

Or, you think that certain inferences are just wrong. Or you think that there aren't enough independent classes of things with CSI to support the inductive reasoning. Concluding that an argument is wrong or unsupported is not the same as claiming it is illegal.

I do agree with you that we should pay more attention to worldview, though. I'm immediately suspicious of arguments that claim to find unique evidence of intelligent agency in only some features of the universe. That would work if the intelligent agency is limited in the same fashion as humans--alien engineers. My worldview implies that we should find as much evidence of intelligent agency in the unique shape of the grand canyon as in a flagellum. If I can simultaneously hold that the grand canyon is the product of "impersonal" forces like erosion AND a creation, then I won't lose any sleep over claims that the flagellum is the product of similar impersonal forces. That probably makes me less inclined to support shaky ID arguments than someone who believes that appealing to erosion or mutation means that God is not involved.

[tgirsch]

It is an apples to apples comparison. They are both pseudo-science, yet when some pseudo-science is taught, it is no big deal as long as it support scientism as a world view.

I call BS. Try teaching phrenology (for example) in a high school science classroom, and I guarantee you there will be an uproar. Phrenology is pseudo-science with absolutely no religious component.

That aside, I don't pretend to know enough about the multiverse hypothesis to know whether or not it's fairly characterized as "pseudoscience." I do know that it's a fair characterization of ID and phrenology, however.

But it's good to see you admit that ID and Creationism are pseudoscience. I used to think you were an ID/OEC proponent. I'm happy to have been proven wrong.

[tgirsch]

Also, I believe your syllogism falls apart at Step 1. I don't think it's a true statement that "Things we observe in our universe that have CSI have always been shown to have intelligent agency." Setting aside the inherently circular and unscientific definition of CSI, there are plenty of complex things we observe in our universe that have unknown origins, or do not appear to be the product of intelligent agency. I'd say that these things outnumber the things for which the agency is known.

And, of course, it has already been pointed out that this line of reasoning is a trap, in that it follows that anything that lacks CSI was not necessarily created by an intelligent agency.

[Paul]

Nick - I'm not sure that whatever the mainstream of ID is would consider mikegene's ideas as ID, but in any case I'm very happy to apologize for my ignorance at just how nebulous ID might be :)

Mr D - I'd guess there would be concern if MT was taught in high schools, at least for more than a couple of minutes as a demonstration of some of the wild ideas that scientists play with. I'd be perfectly happy if ID got the same couple of minutes, though in that case as an example of what people come up with when they start with belief, and then pick evidence to suit.

It's true, though, that in the US at least it would cause a fuss, and that's because it's religion (as even its proponents have said), which the US takes admirably seriously.

[Mr. D]

Nick,

re: "'Things we observe in our universe that have CSI have always been shown to have intelligent agency'" In this statement, isn't "intelligent agency" synonymous with human agency?"

Or divine agency. As a Christ follower, wouldn't you agree that the Bible has earmarks of divine agency? Christ certainly seemed to think so.

re: "The set of "things that have CSI" apparently includes human products and natural organisms. Anything else?"

The Bible. Not only does it have language, which is CSI. It contains a record of fulfilled prophecy that was highly specified in nature ... that is a form of CSI. Then there are the miracles ... which are also CSI ... like being raised from the dead after prophesying it ... a very elegant form of CSI.

re: "Does anything other than information systems have CSI, or is "information system" a simple synonym of "thing that has CSI"

A good definition for information system is "system consisting of the network of all communication channels used within an organization". Nicely describes the information systems I support. Also describes the inner workings of the cell.

re: "Organisms have unique features, not shared by human artifacts or by hypothetical SETI messages, which seem highly relevant the issue of agency."

Expound.

re: "[Some] things that we observe in this universe that have CSI have always been shown to have [human] agency."

Good catch. I neglected a qualifier. It is not just things we observe, it is things we observe that have CSI and we know the origin.

Those would include, as you point out, human intelligence. It would also include divine intelligence as I have pointed out.

So please remove the word "some" and “human” ... unless you can find an exception.

"Concluding that an argument is wrong or unsupported is not the same as claiming it is illegal."

True. But we need to be honest enough to admit our biases … you and me and our atheist friends. I know you are not an atheist, so this comment is not about you. An atheist is precluded from certain conclusions because of his worldview.

"My worldview implies that we should find as much evidence of intelligent agency in the unique shape of the grand canyon as in a flagellum."

Why? I don't understand why this must be true in your worldview.

You seem to be implying that God has to always work the same way. For example, because God chose to use wind to part the Red Sea, then he always has to work within the laws of physics to perform his miracles. That seems to be line of reasoning you are implying in your example. I don’t see why God has to leave CSI on 100 percent of the objects in the universe. Why must this be the case?

"that is a population of error-prone self-replicators under natural selection"

This tips your hand a bit. Maybe this is due to our different vocations, but I marvel at the ingenuity of engineering information systems that self replicate, and am in awe over what it would take to create something like... yet you seem to disparage self-replicators as klunky and error prone. Am I guilty of reading too much into your comment?

[Paul]

Mr D - isn't there a pretty simple exception, which is the Bible you mention. We know it had human agency. We know that everything in there could have been written solely by human agents (I assume we both agree that the Koran was authored only by humans, for example and it has similar ambition and, I'm told, similarly beautiful language and imagery). What we're left with is your belief that it had additional authorship. So Nick's suggestion seems entirely valid; everything we see that seems to have CSI either has human agency, or has unknown agency. Of course, I'd argue that the reason it has unknown agency is because it had no agency.

[matt curtis]

"Of course, I'd argue that the reason it has unknown agency is because it had no agency."

Paul,

This assertion seems irrational to me. Of course, if we applied your argument to any particular hypothesized evolutionary transition, then we would have to conclude that any unknown natural selection mechanism implies no natural selection mechanism.

[Paul]

Matt: My point was that we have some phenomenon - say, a star - and we need a way of explaining how it came to be in the state it is (e.g. how did it start out, if/how it developed from there to its current state, etc). Starting with the assumption that it must have had some sort of agency behind it, particularly based on the fact that the things we know had some sort of agency behind them had some sort of agency behind them, seems like a logical leap too far to me. There's not much difference between that argument and:

1. Man makes fire
2. Star is fire
3. Man made star.

[Mr. D]

Paul,

I am fine with assuming natural causation as a starting point ... at least until it strains credulity to hold onto that.

If I saw seaweed on the shore in loosely the shape of a smile, I would naturally assume the waves did it.

If I saw seaweed on the shore in the shape of two dots and a smile ... like :) ... I would begin to suspect agency.

If I saw seaweed on the shore in the shape of the letters ... L O L .... then I would rule out natural causation and lean strongly toward agency ... even though I never saw an agent.

Wouldn't you do the same?

[Paul]

Mr D - Very nearly; if I saw those things I might assume human agency, which is the only agency (in the sense I think you mean the word) that we know of. But I can see something about as unlikely on Mars (http://www.nasa.gov/images/content/54799main_mars_smiley_face.gif) and honestly say I don't suspect agency at all. Do you?

[Mr. D]

I would assume an agent in the case of the seaweed ... and, I would assume the agent was human.

The differences in the LOL seaweed and the seaweed in the shape of a rough smiley face is one has high specificity ( the LOL ) and the other low ( the smiley ) ... and, one has complexity ( the LOL ) and the other does not.

In other words, the LOL seaweed has a high degree of CSI, and we naturally rule out the waves of the sea as our agent.

You and I do this intuitively ... CSI is just a technical term that explains why our intution serves us well.

By the way, think of an agent as a cause. The cause could be a person, or a force (like gravity), or an angel, or a demon, or the one true God.

I know all of those exist and are agents. One day, you will know too ... if not in this life, you will learn of it in the next.

The smiley face on Mars is another good test case. It has some degree of specificity ... because it kinda sorta looks like a smiley face, which means something to many humans. The problem is that a smiley face is not very complex ... not nearly as complex as LOL ... or "Hi Mom" ... or "Have a wonderful weekend Paul" ... so, it fails the CSI test. :)

By the way, a cellular information systems, like the one used in protein transcription, is a light years more specified and complex than a smiley face. It is an elegant design.

[matt curtis]

Paul,

Mr. D responded well to your point regarding the star and an assumption of agency, but I do want to address that a little more because I think you have misunderstood the argument regarding CSI.

My understanding is that the argument is more along these lines:

(1) We observe complexity and specificity in things such as a clock, a sentence in a book, a computer code, or a bee hive.

(2) We know that in the cases above, agency is involved.

(3) We observe other things that contain a high degree of both complexity and specificity (e.g. Mr. D's cellular information system) for which a natural cause (such as the ocean in the seaweed example) or an agent (such as a watchmaker or honey bee) is unknown.

(4) For all of those things which possess high degrees of CSI and for which we know the cause, that cause is always some agency.

(5) Therefore, in those cases in which we cannot identify agency or some natural cause, it is at least reasonable to assume some sort of agency. It's not proof of agency and it certainly doesn't mean that we cease looking for a natural cause. But it is certainly a strong argument for not outright rejecting the possibility of agency.

[Paul]

Matt - thanks for the post, I was wondering how to reply to Mr D and you prompted a thought.

Using the earlier example, a LOL is neither specified, nor particularly complex. For complexity, approximate right angles and circles are both pretty common in nature, so the only trick is for them to appear in order. That's unlikely, but not improbable. For specificity, LOL is not specific; it means essentially nothing to the majority of people on earth, and even less to most of the biomass on Earth. It only becomes specific because we impose specificity on it; if the message said OTX, for example, the specificity has decreased based purely on our pattern recognition abilities.

Putting aside the fact that Dembski's math is flawed, as are his arguments about probability, this pattern recognition is for me the major criticism of CSI. People build patterns habitually and unconsciously. Often there will be substance behind the pattern, but many times there isn't (for examples of this google 'old wives tales'). And once we have a pattern we often assign agency to it, again whether it's justified or not. That's why thunder is the gods arguing, that shadow at the end of my bed is a grue come to eat me, and Cindy is snubbing me at the office when actually she only sees me in the morning before she's had her coffee.

I don't outright reject the idea that the appearance of agency suggests agency; in fact that's trivially true. But I'm aware that supernatural explanations has been suggested for countless phenomena, and time after time we've found an explanation without requiring such agency (though there's almost always a cause). It's not unlike the creationist's (flawed, again) arguments from probability; if a trillion things have been explained 'naturally' that doesn't mean the trillion-and-oneth won't have a 'supernatural' explanation, but it certainly seems suggestive.

[Susannah]

I'm curious about Paul's statement that the math behind ID is flawed. I'm not a mathematician, and have no way of personally evaluating claims of improbability. Paul, you're saying it's not true that the probability of cells forming through random, natural processes is about 10 to the power of 40,000? I gather this has been revised? I guess I'm thinking that even if it were revised downward a bit, it wouldn't make the likelihood any higher statistically speaking... Would it?

[tgirsch]

Paul:
It only becomes specific because we impose specificity on it

This is almost perfect. But it's important to note that it becomes specific because we impose specificity on it after the fact (or post hoc to use the fancy Latin-speak). And this also answers Susannah's question about the problem with Dembski's logic. It's not so much his calculations as it is the flawed premise.

Dawkins has the best analogy to explain it. Suppose you hit a golf ball, and it lands in the middle of the fairway. Dembski's logic is akin to looking at where the golf ball landed, after the fact, and saying "how incredibly improbable that of all the hundreds of millions of blades of grass in this golf course, the golf ball should land on these few!" In truth, there's nothing remarkable about where the ball landed because after all, the ball had to land somewhere. Now if someone had spraypainted those few blades of grass white without the knowledge of the person who was about to hit the ball, and specified in advance that the ball would land there, then that truly would be remarkable. But that's not how ID views the world.

There's a corollary there of backward thinking. ID proponents tend to view the fact that we couldn't exist in our current form if even minor aspects of the universe were trivially different, and think that this must mean that the universe was specifically created so that we could exist this way. They ignore the converse possibility that we exist this way because the universe is the way it is. They've got cause and effect exactly backwards.

In the end, it's simply a case of taking the pattern-seeking nature of the human brain too far. We've evolved to try to find patterns in the world around us, so much so that we even tend to impose such patterns where none exist (see, for example, the now-gone Old Man of the Mountain, the face on Mars, and countless other such examples).

The biggest flaw in ID is its simplistic premise: if it LOOKS designed, it must BE designed. ID offers no coherent way to distinguish between apparent design and actual design. And, as Jeff would be quick to point out, it's anti-Christian, because it implies that there are things in the universe that were NOT created or designed.

[Mr. D]

Susannah,

I am interested in hearing Paul reply to his assertion that Dembski's math is flawed. I am curious to see how. Let's see what he replies with.

Incidentally, I have found that if a worldview prevents one from inferring that God is the cause of anything, then it doesn't really matter how improbable the material explanation is. The materialist will simply agree with you and say that something amazing and improbable happened. It takes a remarkable faith in luck, but when your worldview gives you no other option, you must take it ... whether the odds are 10 to the 40,000 power to 1 or whatever ... the degree of luck doesn't matter.

BTW, that used to strike me as arrogant or irrational. But, it doesn't any more. I believe that people are guided by their presuppositions.

The best I can do is point out what I believe to be true and the rational reasons I believe it to be true. There is plenty of evidence for design ... no need to try to explain it away as apparent.

I see no evidence whatsover for the multiverse theory. It is an irrational leap toward the unknown. I find it ironic and somewhat amusing that those who are supposedly grounded in empiricism are willing to take such a non-empirical blind leap in the dark.

[Paul]

It appears my reply was eaten by the internets, so I'll try once more :)

@Susannah. I don't know where the number you quote comes from, but off the top of my head:

1. We don't know enough to calculate such odds
2. Cells aren't the starting point of life
3. It assumes that cells are the required destination for evolution (or creation), which isn't the case
4. It doesn't account for the massive parallelism we find in nature
5. Appeals to large numbers aren't always a precursor to an argument from personal incredulity, but in this case it seems likely.

@Mr D - The wikipedia article on CSI has an overview of some of the main criticisms. I confess many of the more detailed arguments are beyond me (I tried and bailed on a math degree), but one of the more common ones seems to be his use of assertions as the building blocks of his argument (such as the universal probability bound) that are neither proven nor, in some cases, even clearly defined (the single most common comment I see from mathematicians is 'yes, but what about...').

As a simple example, the chances of tossing a coin 496 times and getting a specific sequence exceeds the 'universal probability bound' - hence that can't happen randomly (according to Dembski). Yet it could, of course, though we should be very surprised if it did. Obscured by the maths, of course, is the fact that life isn't a single arrangement of heads and tails, but one of a number of possible outcomes (we don't know how many, hence my earlier comment about calculating odds) that could lead to something that could be asking these questions. The probability of getting a particular sequence of coin tosses is reduced immensely if I can decide what sequence I want after the coins have been tossed.

[matt curtis]

Paul,

I am no expert on Dembski's argument, nor have I analyzed it in order to come to any conclusion one way or the other on its validity. However, it appears to me that you have isolated one portion of Dembski's argument and left out an important element. Dembski writes,

Another problem was taking the small probability of events as sufficient reason to rule out their chance occurrence without acknowledging that small probability by itself is not enough to rule out chance. What else is needed? In my theory of design detection, I argue that what's needed is a specification, that is, a type of pattern with certain mathematical and logical characteristics.

http://www.designinference.com/documents/2005.02.Reply_to_Henry_Morris.htm.

Perhaps the sequence of heads and tails in your example would satisfy the degree of specification Dembski speaks of, but I suspect not. In fact, my first thought on your coin example is that we're not really talking about anything particularly specific or complex (we have two possibilities with each coin toss: either heads or tails; we're constructing nothing more complex than a sequence 460 characters long containing only H's and T's that conveys no information whatsoever; and we're utilizing only a single component that must land as either an H or a T.). The difference in magnitude between the coin toss example and tossing the components of a simple mechanical instrument 460 times and having it once land so that the components are properly placed and connected so that the instrument works is significant.

Finally, I have to disagree with your final statement. The fact that you get to decide what sequence you want after the coin tosses, doesn't change one bit the probability of getting that sequence.

[Mr. D]

The probability of getting a particular sequence of coin tosses is reduced immensely if I can decide what sequence I want after the coins have been tossed.

Agreed. No one is claiming otherwise. You are raising a straw man argument and then knocking it down.

The pattern must be meaningful ... like a sequence of nucleotides ... or the combination on a lock ... or the arrangement of letters ... or else it fails the specificity requirement.

"Obscured by the maths, of course, is the fact that life isn't a single arrangement of heads and tails, but one of a number of possible outcomes"

What? Can you provide some empirical evidence to support such a claim?

I would argue that you are committing the same fallacy as the multiverse folks ... arguing from non-evidence to support your claim.

I am more comfortable arguing from what we can observe ... that life is carbon based ... that life requires complex carbon molecules arranged precisely into information networks of little machines called proteins. Change around the sequence of nucleotides, and therefore the amino acids, and you end up with one possible outcome ... goo.

Matt,

The coin toss example, as you point out, is more of an illustration of probabilities than an illustration of CSI. The coin toss example, and the golf ball example that Tom likes to bring up, are straw men arguments that have nothing to do with CSI because we are talking about things that have no meaning ... blades of grass ... coins.

Stick with the seaweed :)

If you walk along a beach and see seaweed arranged in the following pattern "Dembski's Universal Probability Bound" ... now we can talk about whether the wind, the waves, and the kelp randomly organized itself. Would the odds be difficult to calculate? Perhaps. But does that mean we must therefore default to the theory that the wind, the waves and the kelp just happened to wash up that way? Hardly.

[Paul]

@matt - Indeed, I wasn't trying to refute Dembski's entire argument in one blog comment. My point is that you can very quickly get to a stage where nothing is possible, statistically speaking, and yet things keep happening. That's because we're indulging in the sort of choice after the fact that Dembski dresses up with talk of 'mathematical and logical characteristics'; from our privileged place in the universe it's easy to claim that the patterns we see have some greater meaning than just our recognition.

You're right with your example of tossing the components of a mechanical instrument - that's incredibly unlikely, which is one of the reasons biologists don't think that's even vaguely what happened.

Finally, you're free to disagree, but only because I didn't expand on it enough; if I see what coins you tossed, and then pick a target sequence, I can be immeasurably more confident of the probability of a match.

@Mr D - I'm surprised you think that's an unusual assertion, to be honest. Trivially, you and I are 'a different arrangement of heads and tails'. More substantially, just about all life that we routinely see is a member of eukarya, leaving bacteria and archaea as alternate routes to what we would consider advanced life. We can speculate about silicon-based life forms or whatever if you wish, but that doesn't seem necessary, though it might be fruitful to consider less sweeping substitutions within specific proteins.

Getting back to the seaweed. Let's assume that I am a Chinese farmer on my first ever trip to the ocean. I've seen your seaweed message, and it's utterly meaningless to me - it lacks any specified complexity. And yet the seaweed and its arrangement hasn't changed. That's a problem with Dembski's argument; we don't recognize signs of design in life, we recognize signs of what humans see as design in life. Could we do otherwise? Hardly.

[matt curtis]

"That's because we're indulging in the sort of choice after the fact that Dembski dresses up with talk of 'mathematical and logical characteristics'; from our privileged place in the universe it's easy to claim that the patterns we see have some greater meaning than just our recognition."

Paul,

I don't disagree that we are looking at this after the fact, but then that is equally true of macroevolution, abiogenesis, archeology, and a whole host of other disciplines. We are trying to draw certain conclusions about what occurred in the past based upon present observations.

"Finally, you're free to disagree, but only because I didn't expand on it enough; if I see what coins you tossed, and then pick a target sequence, I can be immeasurably more confident of the probability of a match."

I stand by my position. All you have changed by choosing the sequence after all the coins have been tossed is your probability of choosing the correct sequence. You have not, however, changed the original probability of that sequence occurring. This is really quite easy to demonstrate. Let's say the probability of a particular sequence of H's and T's (heads and tails) before the coin is tossed 460 times is x . After the coins are tossed, you choose the sequence and, strangely enough, it matches the sequence actually tossed. If this exercise has somehow changed the probability of that particular sequence occurring (from x to 1), then you should be able to predict with certainty that tossing the coin another 460 times will produce the exact same sequence. But, of course, that's not the case. The probability of tossing the same sequence of H's and T's the second time is the same as it was the first time: x .

This is one of the areas where Dawkins's argument is so clearly deficient. He essentially assumes a materialistic cause (i.e. excluding any supernatural cause, such as a Creator), observes that we exist, and then concludes that because we exist a materialist explanation must be possible and have occurred even if it would seem improbable.

"Getting back to the seaweed. Let's assume that I am a Chinese farmer on my first ever trip to the ocean. I've seen your seaweed message, and it's utterly meaningless to me - it lacks any specified complexity. And yet the seaweed and its arrangement hasn't changed."

Are you suggesting that for something to be both specific and complex, it must be recognizeable as such by everyone? If an archeologist were to find an artifact with what appeared to be some rudimentary figures that he did not recognize as any language, yet another archeologist, seeing the same artifact, immediately recognized the figures as an ancient language, then wouldn't the figures still be an example of specified complexity?