Thursday, July 16, 2009

Stephen Meyer's Honesty Problem

Like most intelligent design advocates, Stephen Meyer, director of the Discovery Institute’s Center for the Renewal of Science and Culture, has a little problem telling the truth.

I first encountered his dissembling at an intelligent design conference held at Calvin College in May 2001. Meyer had written in 2000 that "Systems that are characterized by both specificity and complexity (what information theorists call "specified complexity'') have "information content''."

The only problem is, information theorists don't use the term "specified complexity" and they don't refer to "specificity" when discussing information. At the time, there was precisely one mathematician who was pushing the term "specified complexity", and that was William Dembski, who tried (but failed) to create a new, mathematically-rigorous definition at information which (were it coherent) would be at odds with how information is defined by other mathematicians and computer scientists.

I went up to Meyer at the conference and asked him, "You wrote that 'information theorists' (plural) talk about specified complexity. Who are they?" He then admitted that he knew no one but Dembski (and Dembski himself is not much of an information theorist, having published exactly 0 papers so far on the topic in the peer-reviewed scientific literature).

So the use of the plural, when Meyer knew perfectly well that information theorists do not use the term "specified complexity", was just a lie - and a lie intended to deceive the reader that his claims are supported by the scientific community, when they are not.

(Another anecdote: while I was waiting in line to ask Meyer this question, I was behind a woman who couldn't wait to meet Meyer. She gushed as she shook his hand, saying she was so honored to meet the man who was responsible for recruiting so many people for Christ through his work. He smiled and thanked her. And they claim ID is not religious!)

Meyer was also caught dissembling about the "No Child Left Behind" education bill, falsely claiming that it obligated Ohio to teach about alternative theories.

Now Meyer is back with a new book, and an op-ed in the Boston Globe to help flog his book. In the op-ed, Meyer claims, "Information - whether inscribed in hieroglyphics, written in a book, or encoded in a radio signal - always arises from an intelligent source." But this is the same old bogus ID claim that is repeated endlessly and endlessly, and it's not true. At least it's not true if you understand "information" in the sense that it is understood by mathematicians and computer scientists. For example, in the Kolmogorov theory, any random source produces information.

But then again, Meyer, with his little honesty problem, doesn't seem too concerned with the truth. What's important is, as that woman ahead of me in line told him, saving souls for Jesus.

Martin Luther once said, "What harm would it do, if a man told a good strong lie for the sake of the good and for the Christian church...a lie out of necessity, a useful lie, a helpful lie, such lies would not be against God, he would accept them." It seems that Stephen Meyer would agree.

23 comments:

TomC said...

Jeffrey,

Perhaps this has been asked before, but, in your humble opinion, what are some of the better books out there explaining information as mathematicians or computer scientists understand it? I'd like a few technical references, actually, say, at the advanced undergrad or beginning grad level, if you have a few.

I ask because I have friends who think Dembski is great, and I would love to tehnically counter their arguments about his "specified complexity" crap.

Thanks in advance, and sorry again if this has been asked before.

Unknown said...

I prefer Nietzsche's take on the subject: “The most perfidious way of harming a cause consists of defending it deliberately with faulty arguments.”

Jeffrey Shallit said...

TomC:

To my knowledge there are no good book-length treatments of information theory for laymen. One book that's worth reading (although it is hard to find these days) is Renyi's A Diary on Information Theory. It is good, but Renyi unfortunately died before he could finish it. And it only covers Shannon information, not Kolmogorov information.

Wesley Elsberry and I wrote a chapter for Why Intelligent Design Fails that explains the basics of Kolmogorov theory.

TomC said...

Great! I actually bought that book not too long ago, but haven't gotten around to reading it yet. I will now, so thanks again!

RBH said...

Tom Schneider has a page of information theory resources on the web. I have not evaluated any of them. I suspect they're mostly Shannon-oriented. Jeffrey?

Jason Rosenhouse said...

There is an excellent book published by Dover that presents a remarkably clear intorduction to Shannon information for laypeople. It is called "An Introduction to Information Theory: Symbols, Signals and Noise," by John R. Pierce. The early chapters in particular are incredibly readable. The book was first published in 1961, but the Dover edition was released in 1980 and is decsribed as a second, revised edition.

Gareth McCaughan said...

David MacKay's book on information theory, learning and inference is beautiful and at roughly the level TomC is after. Much more Shannon than Kolmogorov, as usual. It covers lots of topics besides information theory. Very clearly and approachably written, but it does expect you to put in a fair bit of intellectual work to get the most out of it.

Anonymous said...

Re Jason's comment: Pierce (d. 2002) was a close associate of Claude Shannon at Bell Labs- he was also a well-known science fiction writer, using the name J. J. Coupling. (The pseudonym refers to a quantum mechanical phenomenon.)

Bob Carroll

Anonymous said...

For TomC: Marcus Hutter maintains a page on algorithmic information theory that has some tutorial links and book references. It looks like the page hasn't been updated recently and some of the links are broken, but it should still prove useful.

Gregory Chaitin, one of the co-developers of AIT, has written a number of books for a general audience. He has a very quirky style and likes to prove things by using LISP code, so he's not to everyone's taste, but you might try looking at his book "Exploring Randomness".

For a *very* concise introduction to AIT, you could also look at section 6.4 of Siper's "Introduction to the Theory of Computation".

Anonymous said...

I once found myself arguing specified complexity without knowledge of AIT. My IDer opponent, sadly, was a fledgeling computer scientist, like myself (I weep for my profession).

He may not have reproduced Dembski accurately, but regardless, his argument made no sense. He prattled at me for a minute or two about the Universal Probability Bound, and how anything less probable than that must be evidence of Design, before I stopped him and said (paraphrased):

"Say I have a random number generator, and one day it starts spitting out the Fibbonacci Sequence. Lets stipulate that it spits out N + 1 terms of the sequence, where the likelihood of my RNG producing more than N Fibbonacci numbers in sequence is less than your UPB. Why is that sequence of N + 1 terms evidence of design — why does it magically contain information — where N terms is not and does not?"

He didn't have an answer. Not that he admitted it; he dissembled, equivocated and reiterated until we both had to leave. Were both of us wrong, or was it just him?

Jeffrey Shallit said...

Anonymous:

I don't think you're wrong, exactly. The difficulty in determining whether you are is that Dembski has been very inconsistent in his definition of what constitutes "information" and "CSI". Sometimes he says to be "information" the probability must be smaller than the universal bound, and sometimes he says it doesn't. Sometimes it is measured in bits; other times it is just yes/no, it contains information or it doesn't.

This is discussed in my long refutation written with Elsberry, which I linked to in my blog posting. Nailing down exactly what Dembski claims is often difficult because of these inconsistencies. We show in our paper that no matter what interpretation is taken, Dembski's claims are bogus.

D. Swart said...

Jeff, your post triggered a memory.

Did I ever tell you about The time you assigned a busy beaver "contest" for 5 and 6 states?

I ran randomly generated FSMs and halted them automatically after what I felt was a sufficiently large but practical number of moves. I ran these simulations over the weekend.

The winner of my simulated contest used one state to return to the beginning of a tape (the ones column), and add one to a binary number on the tape. It used the remaining states to carry the one as necessary. It conveniently halted after counting to 32 (64?) on the tape.

My experience of seeing such cleverness arise by chance is a valuable cautionary tale. I use it whenever I'm tempted to rely on intuition about such things.

Anonymous said...

Thanks much Jeff, for your clarification. I consulted Wikipedia after the fact, which said much the same. My opponent stuck to one definition, which he repeated often without rephrasing. I don't know that he really understood it.

He seemed shaky in other areas as well; he claimed the idea of gradualism in evolution was a fallacy of composition. I have only suspicions as to why.

I mainly objected to his claim that unlikely events were impossible, barring intelligent intervention. However, my RNG thought-experiment seemed to give him the most consternation.

At the very least, that conversation led me to an interest in pursuing information theory as a hobby, though I think I lack the time and discipline to make a serious study of it.

MrFreeThinker said...

Mr Shallit. Given how particular you have been about quotations in your past interactions with ID proponents , I would suggest you provide some sort of primary source for that Martin Luther quote or perhaps remove it (so that Luther's name isn't slurred).
I've read some Martin Luther and that does not seem to be consistent with his philosophy and I can't find it in any commentary so I am unable to check the context.

Alex said...

It seems like a silly argument. Wouldn't even reflected photons be considered "information"? They get ejected from the sun, travel millions of miles, hit a leaf, or a rock, or what have you, and certain frequencies are reflected back. When we look at them with the right sensory apparatus we can determine the direction and colour of the object from which they were reflected. At no point in the "creation" of that information do we require any intelligence to step in and sort things out. The fact that we can SEE tells us that information is created all the time through purely natural causes - the only step that requires intelligence is the interpretation of that information, and WE are the ones doing the interpreting.

I don't know, maybe I'm missing something but it seems to me that his argument could be pulled apart by a third-grader.

Jeffrey Shallit said...

Dear Mr. FreeThinker:

You are too funny!

As if Martin Luther's name could be slurred more than he had done himself, with his antisemitism.

Anyway, the source from my quote is from an incident involving Philip of Hesse. Philip wanted to take multiple wives, and to do so, tried to get his second bigamous marriage approved by church authorities. Martin Luther arranged a letter declaring the second marriage legal (the "Wittenberg Declaration") and then pretended to know nothing about it. He wrote Philip of Hesse and counseled him to lie about the arrangement.

That is the context for the quote, and it can be found in Here I Stand: A Life of Martin Luther by Bainton, among other places.

Douglas Groothuis, Ph.D. said...

You commit a fundamental confusion. ID relies on no uniquely religious assumptions to make its case that some aspects of biology are better explained by design than by chance and/or necessity. Its conclusion has philosophical implications concerning the explanatory inadequacy of naturalism, but this conclusion is earned through arguments.

Moreover, Meyer states in The Signature of the Cell that the ID argument alone cannot tell us that much about the nature of the designer.

The fact that a woman thought that Meyer's work helped people come to Christ does not mean that (a) the books is unscientific or (b) that it relies on uniquely religious premises. It does neither. However, the book might open people up to nonreductionistic nonnaturalistic understandings of reality, of which Christianity is one.

Jeffrey Shallit said...

You commit a fundamental confusion.

No, I don't.

ID relies on no uniquely religious assumptions to make its case that some aspects of biology are better explained by design than by chance and/or necessity.

Not true, since the truth of ID depends on the existence of a being with sufficient causal powers to tweak the fundamental constants of physics.

But I'll admit it, you parrot the party line pretty well.

And you say nothing about Meyer's dishonesty. Why am I not surprised?

MrFreeThinker said...

"Not true, since the truth of ID depends on the existence of a being with sufficient causal powers to tweak the fundamental constants of physics."

And that claim isn't uniquely religious. Didn't Dawkins or some atheist say they were open to the idea that perhaps aliens from another universe designed the universe we are in now?

P.S. this post is a bit dated and you may want to change things now since Dembski has published a couple papers since then and gotten support from Robert Marks II and some citations and favorable reviews.

RBH said...

MrFreeThinker wrote

Didn't Dawkins or some atheist say they were open to the idea that perhaps aliens from another universe designed the universe we are in now?

Not to my knowledge, and I've read all of Dawkins' books and a number of his essays and talks. Be nice to have a reference for a blind guess like that.

Jeffrey Shallit said...

favorable reviews

Maybe Mr. not-so-Free Thinker would care to enlighten us about the "favorable reviews" of Dembski's work.

James C said...

"Didn't Dawkins or some atheist say they were open to the idea that perhaps aliens from another universe designed the universe we are in now?"

I think you're thinking of the bit in Expelled where Dawkins says something along the lines of:

"It's not impossible that life on earth could have been created & designed by aliens, but those aliens would have had to arise themselves by some evolutionary process somewhere else in the universe."

Take note: 'life' does not equal 'the universe'... Your creationism is showing!

Jeffrey Shallit said...

Notice that when I asked Mr Not-so-Free-Thinker to produce those "favorable reviews" he was claiming, he magically disappeared.