Wednesday, February 18, 2009

Intelligent Design - Specified Complexity

The primary argument that ID advocates use is that Intelligent Design can be identified when seen. The classic example is the argument made by William Paley in his book “Natural Theology” more than two centuries ago (in 1802). That book was a favorite of Charles Darwin. Here’s that argument:

“In crossing a heath, suppose I pitched my foot against a stone, and were asked how the stone came to be there; I might possibly answer, that, for anything I knew to the contrary, it had lain there forever: nor would it perhaps be very easy to show the absurdity of this answer. But suppose I had found a watch upon the ground, and it should be inquired how the watch happened to be in that place; I should hardly think of the answer I had before given, that for anything I knew, the watch might have always been there. (...) There must have existed, at some time, and at some place or other, an artificer or artificers, who formed [the watch] for the purpose which we find it actually to answer; who comprehended its construction, and designed its use. (...) Every indication of contrivance, every manifestation of design, which existed in the watch, exists in the works of nature; with the difference, on the side of nature, of being greater or more, and that in a degree which exceeds all computation.”

I doubt that anyone would dispute Paley’s claim that anyone who saw a watch, even if only for the first time, would identify it as the work of an intelligent designer.

The challenge comes in finding a way to objectively quantify whether something is intelligently designed. A watch is a pretty obvious example. But not everything is so black-and-white.

We’ve all probably seen or heard of examples where someone sees an image of something or someone – often the Virgin Mary – in a reflection on a window, or in a bagel or even in a stain on a highway bridge. Those with more vivid imaginations claim that they are witnessing a miracle and that the image they see is divinely “designed” or inspired. Others often see no images at all.

Shown below is an example of such an image in a piece of toast. This toast was sold on the Internet for $28,000.

Some people see an intelligently designed picture of a woman’s face in the toast. Other people don’t. (I see a face myself but I doubt that it is intelligently designed.) How do you tell who’s right?

Another example is a rock formation on Mars that appears to be a human face when the shadows are just right. Here’s a picture of the rock taken at the “right time”.Some people see an Intelligently Designed face here and others don’t, believing that it is simply a product of shadows at a particular time of day.

Here is an on-line quote that shows how adamantly some people feel that this rock formation is an intelligently designed face[1].

"NASA can't hide the truth any longer, there's no mistaking the image of a Lion's face and ornamental headdress when the photo is viewed in its proper orientation."

"The raw NASA image was not altered. I colorized it to make the features more apparent. The impression of a face is very striking. It has all the qualities you would expect to see in a face. There is even an arched eyebrow on the left side above the eye socket! A pupil is plainly visible. There's also a very clear and distinct eye on the right. It's remarkably human-like. Yet it has a cat-like nose and even striations on the cheek like whiskers!"

Shown below is the “modified” picture that the author is referring to in these paragraphs.

To me it looks less like a face than it did in the previous picture. Alas.

A similar debate can be formed over features that we see in living things. To some they seem “designed”. To others they seem to be “jury-rigged” and not the result of intelligent design. But it is clearly all in the eyes of the beholder. There is no way of quantifying the artificiality of any of these things in order to determine if they are intelligently designed or not.

A mathematician (and theologian) named William Dembski has come up with a proposal on how to distinguish design from non-design. He calls the criteria that should be used to identify Intelligent Design “Specified Complexity”. In summary he claims that something can be identified as being intelligently designed if it is complex, performs a specific function (has specificity) and is unlikely to have been produced by purely random processes. One example that he uses is:

"A single letter of the alphabet is specified without being complex. A long sentence of random letters is complex without being specified. A Shakespearean sonnet is both complex and specified.[2]"

Dembski’s argument is rejected by the vast majority of mainstream scientists and mathematicians. One major argument is that it is impossible to calculate the odds of a particular thing happening by chance. How can someone look at the “picture” of the Virgin Mary in the piece of toast shown earlier and calculate the odds of that forming by chance? As I mentioned, someone paid $28,000 for that piece of toast. If “specified complexity” was a worthwhile concept, you should be able to use it to determine whether or not that was a reasonable purchase.

An even stronger argument against Dembski’s claims is that complexity isn’t really a sign of intelligent design. Instead artificiality is the only indication, and artificiality can’t be quantified.

A crack is a sidewalk can be very complex. But such cracks can be formed by simply having something falling on the sidewalk. Or they can be formed by nothing more than a tree root. On the other hand, a simple letter ‘A’ carved in the sidewalk would be relatively lacking in complexity. Yet it would be much more likely to be the result of intelligent design than would the much more complex crack. As a matter of fact, Dembski even uses the single letter example as something that is “specified without being complex” while obviously completely failing to see that even a single letter in the right circumstances is all that is needed to recognize ID – no complexity needed.

“Artificiality” is really the same thing that Dembski calls “specificity”. You can look at something and see that it seems to be designed with a particular purpose in mind.

The very best example that I can think of comes from the movie “2001 - A Space Odyssey” based on a book by Arthur C. Clark. In the book and the movie, a stone monolith is found on the moon by astronauts. The monolith is a large pure-black, three-dimensional rectangular slab. Its dimensions are in the ratio of exactly 1:4:9 to whatever precision humans are able to measure them. (Of course, 1, 4 and 9 are the exact squares to the first three integers.) The monolith is deemed, just as Paley’s watch would be, as the undeniable result of intelligent design, in this case the work of an alien intelligence.

While the movie and book are works of fiction, surely if such a monolith was actually discovered exactly as described in the book, it would be called a result of ID just as surely as it is done in the book.

But the shape of the monolith is not complex at all! In fact a rectangle is one of the simplest and least complex shapes possible. Even the color – pure-black – is less complex than any other color; certainly it less complex than any pattern of colors would be. But that ratio of 1:4:9 and color are very artificial. It is the artificiality that indicates ID, not complexity.

What’s even more interesting about this example is that it comes from a book written in 1968, well before the Intelligent Design movement had its roots. Yet the author theorized that if an alien civilization wanted to communicate with humans through an “intelligently designed” symbol, then the symbol would be simple rather than complex!

As a matter of fact, if you give the matter a little thought, it is obvious that complexity has absolutely nothing to do with Intelligent Design! Surely many intelligently designed things are complex; something like the space shuttle comes to mind.

But there are surely at least as many intelligently designed things that are not complex as there are examples of intelligently designed things that are complex.

As one simple, every-day example, consider what happens when someone weeds their garden. Removing the weeds reduces the complexity of the garden itself. So the gardener is using their intelligence to design something that is less complex than it would be otherwise. In fact, when starting a garden in the spring, a gardener often removes all pre-existing plants, considering everything in the garden to be weeds. Surely a piece of ground with nothing in it but dirt is the least complex patch of ground that is possible.

Another example from nature can be seen when we compare a typical forest to a Christmas tree farm. The forest consists of many trees of different species and ages planted in random patterns.
A Christmas tree farm consists of trees of the same species (or just a limited number of species) all roughly the same height illustrating that they were planted at the same time, arranged equally spaced in rows.

Yet it is the tree farm that is intelligently designed, not the regular forest. We can recognize that fact only because it is more artificial; the complexity or lack thereof is a non-factor.

One of the points of confusion is the meaning of the term “complexity”. A computer can quantify the complexity of data in a file by compressing it and then the number of bits (or bytes) in the compressed file can be used as a measure of its complexity. In more general, and possibly more practical terms, the complexity of something can be quantified by the number of words that are required to describe it. Something that lacks complexity can be described in relatively few words. On the other hand, something that is complex takes many more words to adequately describe it. A square is an easier shape to describe than is a trapezoid. So a trapezoid is more complex.

But here’s how one creationist argued against the arguments that I have just presented, focusing on my claim that “complexity” is not relevant:

> Specificity alone is insufficient. Drop five coins
> onto the table, and all five coins may land heads-up.
> That is quite specific, but not terribly complex,
> and doesn't particularly indicate intelligent
> manipulation.

> Drop a thousand coins and they land up approximately
> half heads and half tails; that's complex, but
> not specific, and doesn't indicate intelligent
> manipulation.

> Drop a thousand coins and they all land heads,
> that's both specific and complex, and you'd be
> highly suspicious that some intelligence had
> manipulated the coins somehow.

In fact, 1000 coins that are all-heads is not really more complex than five coins that are all heads. In either case, it can be described with this simple phrase: “all-heads”. With the larger number of coins, you have more data points but really no more complexity.

One way to understand this is to consider photographs. You could take a picture of a perfectly monochrome blue sky with both a low-resolution and a high-resolution camera. The high-resolution camera would have more data points (i.e. more pixels). But surely the two pictures would be judged by anyone as being equally complex. No one could look at the higher resolution image and claim that it is more complex merely because it has more data points. In fact, for all practical purposes you could describe either picture with a single “sky blue” pixel! That single point really describes, by itself, what the entire sky looked like.

Moreover with the picture of the blue sky, the additional data points don’t add artificiality any more than they add complexity. It is our perception that if one part of the sky is blue, there is an excellent chance that the rest of the sky will be blue as well. We’ve all seen completely blue skies. Nothing makes us view these pictures as unnecessarily artificial.

The only way to add complexity to the picture is to add additional items against the pure blue sky. If you added a cloud, an airplane or a bird to either image you would make it more complex. Similarly the pattern of coin flips becomes more complex as soon as some of the flips come up as tails. But as the pattern becomes more complex, it becomes less likely to be the result of intelligent design. So yet again, complexity tends to work against the recognition of ID.

As a matter of fact, all of the additional coins coming up only as “heads” are adding artificiality and not complexity. It is more artificial because we can calculate the odds that only “heads” will come up with each flip and at some relatively arbitrary point, come to the conclusion that the coin toss is “fixed” meaning that an intelligent agent is involved.

The odds that a fair coin will come up “heads” are one-out-of-two or 50%. The odds of “heads” coming up ‘N’ times consecutively are one in 2N.

The odds that heads will come up five consecutive times are 25 or one-chance-in-32. That’s about 3%. That might make a hardened skeptic suspicious, but many others would probably accept it as a reasonable chance even if it wasn’t very “likely”.

The next coin flip to come up “heads” decreases the odds by a factor of 2 to about 1.5%. The “heads” that comes up after that decreases the odds by yet another factor of 2. At this point, probably a few additional people are suspicious. As each coin flip comes up heads a continuously higher percentage of people become skeptics that the coin being flipped is honest. Surely after 1000 such flips, no one believes that some intelligent being didn’t do something to “fix” the coin so that it can only come up heads. But it is still just a matter of perception and the specific point where any particular individual becomes a skeptic will vary.

But a pattern of ten “heads” in a row isn’t more complex than a pattern of five “heads” in a row any more than a high-resolution picture of a perfectly blue sky is any more complex than a low-resolution picture of that same perfectly blue sky. It is simply more artificial because it is less likely to be the result of random events. In the case of flipped coins, we can calculate the amount of artificiality. We can’t do that with other things such as patterns on pieces of toast or faces in rock formations on Mars or the pattern of genes in DNA.

The only reason that people like Dembski use complexity in their arguments is that it allows them to bring lots of numbers into the argument. Over the years there has been a great deal of practical and theoretical work done, especially in computer science, towards quantifying complexity. Such studies provide algorithms that can be used to remove redundant information from messages allowing those messages to be sent more efficiently from one computer to another primarily over the Internet. If Dembski or some other ID advocate can convince people that complexity is a key factor in recognizing intelligent design, then they can apply these mathematical formulas and come up with numbers that have very many zeroes in them. Suddenly they can argue that something “specified” (i.e. “artificial”) could not have come about by chance because it is so unlikely.

But the argument is simply bogus. Complexity has nothing whatsoever to do with Intelligent Design. Artificiality is the only criterion that can be used to determine if something is intelligently designed. Artificiality can’t be quantified, except in very rare cases such as coin flips where specific odds of individual components are known.
[1], referenced on January 3, 2009
[2] William A. Dembski (1999). Intelligent Design, p. 47

No comments:

Post a Comment