February 14, 2010

Maimonides, Monkeys, and Monet

Basic Teleology

There is a story, attributed to the 12th century Jewish teacher Moses Maimonides, in which he shows his students a beautiful landscape painting, and tells them that the painting was the result of someone bumping a table, and accidentally spilling paint on a blank canvas. As the students laugh at this preposterous claim, the teacher asks, “Then how can you believe that the complete cosmos, in which everything is held in perfect balance, appeared by chance?”

In this story, Maimonides is using what is known as a teleological argument (an argument based on the perception of design, order, or purpose) for the existence of God. You’ve probably heard similar arguments before. Similar observations are made about a pocket watch found on a beach, a tornado in a junkyard assembling a 747, or monkeys typing the complete works of Shakespeare. The complexity of the human body, the wonder and beauty of nature, and even the very ability to comprehend and appreciate that beauty, causes us to marvel at the seemingly purposeful design behind creation. 

The Need to Look Deeper 

It’s possible, however, to misapply such analogies in such a way as to miss out on some very profound truths about the nature of God and His creation.

First, the analogies only illustrate the existence of design; they fail to address just how deeply-seated that design may be. Second, all of these examples (a watch, a 747, or a landscape painting) imply something that is somehow “manufactured” in its final form from a predetermined design plan. While this interpretation may resonate with our western industrialized mindset, it seems to me that scripture, and evidence of creation itself, are more in tune with a creation that was “birthed”, rather than “manufactured”.

Perhaps it’s time to expand upon Maimonides’ original story. An easy place to start would be to update the “painting” analogy to reflect what we now know about the genetic structure of living organisms. This can be done without any presupposition of “design” or a “designer”, simply by objectively describing the observed evidence.

Maimonides Redux: Billions of Paintings 

In order to adapt the “painting” story to reflect what we now know about the world, we need to assume that there is not just one painting, but billions of paintings, scattered throughout the world. These paintings are carefully studied, and the following facts are observed:
  • Some paintings are incredibly complex depictions of landscapes with mountains, trees, waterfalls and hillsides full of flowers. Others are much simpler – maybe just grass, sky and a few clouds.
  • Many of the paintings are so similar to one another that they appear to not only be painted by the same artist, but are probably slightly modified copies of one another
  • Not everyone likes the same style of painting. There are, however, many paintings that no one seems to like. These paintings are discarded, while the paintings that people like most are run through a duplicator.
  • Archeological excavations reveal that this process of duplicating paintings has been going on for quite some time. It is also observed that the oldest discovered paintings seem to be the simplest, while the more complex depictions appeared much more recently.
  • Paintings can be compared to determine which older painting was modified in order to create a later version. In some cases, there are large portions of the painting left unmodified, but some new elements are added. In other cases, there are some elements that have been partially painted over. Only grass exists where a tree once was; however, a careful examination shows that a portion of the original tree trunk remains. It looks more like a rock now than a portion of a tree, but the specific arrangement of paint can be used to identify the particular version of tree that once existed underneath the top layers of paint.
  • Collisions with paint cans do, in fact, seem to happen quite frequently, and they often result in spots of paint landing on previously reproduced paintings. In most cases, this ruins the painting, and the spotted version is discarded. In a few cases, the spots don’t add or detract from the original painting, and are left in place to be duplicated. In a very few cases, the spots actually combine with some previous spots to create what looks like an interesting new addition to the scene, and these versions are reproduced prolifically.
  • Occasionally, a complex structure, such as a new tree in the background, appears. Comparison with older generations of similar paintings seems to indicate that the tree didn’t appear all at once. At some point there was apparently a splotch of brown paint that apparently didn’t detract from the painting, because this version seemed to have been widely reproduced. Maybe it looked like a rock, or just a shadow. Some time later, after perhaps millions of discarded versions, another splotch of brown paint landed in a way that seemed to convert the rock into a tall stick. Maybe it looked like an irrelevant fence post. But in any event, this version of the scene seemed to have been widely duplicated as well. Finally, after perhaps millions more discarded versions, a splotch of green paint in just the right place seemed to transform the fence post into a tree. Although the rock and fence post didn’t improve the scene much, the tree must have seemed like a marvelous addition, because this version ended up being reproduced in high volumes. 

Paint on the Genetic Canvas

Maimonides’ original illustration has been updated, so that instead of a single collision with paint cans instantly creating a Monet masterpiece, we have a scenario in which duplicated paintings occasional get incrementally and subtly changed by random splotches of paint. All paintings made worse by these splotches are discarded, while those which are made no worse (or, very rarely, improved) continue to be reproduced. In the end, the selection process itself guarantees that of all the possible variations, only those with sequences of accumulated variations that have artistic merit, will survive.

The updated analogy now seems to be a much better description of the evidence for the observed pattern of genetic relationships between living things. Much of this evidence has been discovered only in the last few years, thanks to the ability to sequence complete genomes.  As the genetic sequences of different species are analyzed, they can be correlated with sequences of other species, to find which species share identical sections of genetic code, and where parts of that genetic code have been partially overwritten.  Much like tracing which earlier painting was duplicated and modified as the basis for a more complex painting, the "paint" on the "genetic canvas" can be used to determine ancestral relationships among living things.  

The mathematical odds against finding such perfect correlation in sequences billions of base pairs in length is astounding.  Not only do the patterns of useful information match, but so do the remnants of sections that have been partially overwritten with new information sections seem to serve no purpose other than preservation of a traceable historical record of that species' ancestry.  So compelling is the genetic evidence that 95% of all scientists, and an incredible 99.85% of all biologists, agree with the evidence for "descent with modification" (http://www.talkorigins.org/indexcc/CA/CA111.html).   This includes both the estimated one-third of scientists who are atheists and the two-thirds that believe in God (http://www.livescience.com/strangenews/050811_scientists_god.html).

So how is it that believing and unbelieving scientists can be in full agreement on the observed evidence, and yet come to vastly different conclusions on the existence of God?  Is there a flaw in the teleological argument for God?  Do believers need to continue arguing that 99.85% of researchers are wrong?  Are we really just a result of random processes?  Or is the element of design simply embedded much more deeply in the fabric of creation than we thought?

More to come on that subject in a future post.


  1. Phil,

    I just read your “updated Maimonides” illustration. I would observe a few things, and ask a few questions:

    Observation 1: In your greatly expanded analogy, it is obvious that some conscious force – external to the paintings – is at work. Some paintings are discarded – why? Who discards them? Some are reproduced prolifically. Same question. Without this external force, your analogy is meaningless and impossible. Now I realize that a biologist would probably say this force is “natural selection.” But if the entire process is random, why should natural selection be nonrandom? Yet clearly, “natural selection” is not random. Your analogy makes it look like Intelligent Design.

    Observation 2. The addition of the extra paintings (to represent the actual findings of biology) seem to make the story even more unlikely than Maimonides' one-canvas illustration. In other words, the overall effect is to make the absence of Intelligent design even more ridiculous looking. We might just barely imagine the remote possibility of one incredibly lucky spill resulting in a Monet. But billions of canvases with billions of little changes, many discarded, others improved upon, the same pigments and materials used in each – that sounds exactly like the Studio of a Cosmic Artist.

    That billions of paintings, sharing common pigments etc would even exist without an Artist, is silly. That they would develop into a Monet without the touch of an Artist seems far less likely than the one single massively lucky spill (which of course, is ridiculously unlikely itself).

    Question: in your illustration you made it sound like mutations happen quite regularly “collisions with paint cans, do in fact seem to happen quite frequently...” Any idea just how often random mutations take place? Any additional idea how often a random mutation turns out to be beneficial? I mean actual observational data – not extrapolations backwards through time, which would simply serve as a circular argument. Also, it is my understanding that some organism are more susceptible to mutation than others. Any ideas on how that works?

    I don't know the answers to that, but I suspect that random beneficial mutation does not happen, on average, often enough to account for the progression of life in a mere few billion years.

  2. Your question about the mutation rate was an interesting one. I’m not a biologist, but from what I’ve read, there is quite a range of estimates, depending on the criteria being used. Of course, mutations occur quite frequently at a cellular level (leading to cancer, or simple aging, for example), but I assume that here we are only concerned with mutations that are passed on to the next generation, via the sperm or egg.

    I agree with the need for observational data. One such study looked at the mutation responsible for retinoblastoma, a rare childhood cancer. Until recently, this was always fatal in childhood (before victims were old enough to reproduce), so each occurrence could be attributed to a new mutation. This particular anomaly was found to occur at a rate of about one mutation per 10,000 genes per generation. With 30,000 genes, this means that each person would have, in their genome, an average of three mutations. (I’ve seen other estimates pointing to 100-200 mutations per individual, but I assume this includes the 90% or so that occur in unused portions of the DNA. Maybe someone with a biology background could chime in here.)

    As to beneficial mutations, I don’t know what the rate is, but I can only guess they are exceedingly rare, or primarily occur as a result of a specific sequence of individually neutral mutations. But with three mutations for every child born (three new spots of paint on every reproduced painting?), the statistical space gets quite large. I expect we’ll learn much more in coming years from our recently developed genome mapping capabilities. It will be interesting to find out if the rate of beneficial mutations matches what would be expected from a purely random mutation space, or if the statistics point toward the necessity of an external hand guiding the process.

    Now if I accept the rate of three mutations per individual, my first thought is “why isn’t the world populated by mutant freaks?” (Insert your own punch line here.) I’m guessing it has to do not only with a large percentage of mutations being neutral, but also with the redundancy and self-repair capability of DNA.

    And when you start looking at how self-repairing DNA works, I’m right with you in marveling at what looks like evidence of design. I think we underestimate God, however, when we stop at whatever our current level of knowledge is, and say “this is where the design takes place”. When we do that, aren’t we in effect saying, as we peel back the layers of creation, that God is just one layer smarter than we are? (continued)

  3. I, for one, am not surprised or disappointed when a scientific discovery presents an explanation for something we once assumed was a “miraculous design”. Why? Because such discoveries invariably lead to uncovering another layer which shows that the design is buried more deeply than we had imagined.

    Yes, I see the hand of an Intelligent Designer in all of creation. For me, study of science only enhances that. I see the hand of the Artist when I look at the final result.

    When I study the mechanisms by which life developed (the process by which “paintings” are modified, replicated, and chosen), I marvel at the incomprehensible genius of a process that is so relentlessly driven towards its ultimate purpose, but with the built-in variability needed to ensure a creation with free will, independent of its Creator.

    When I look at the structure of DNA needed to make that process work, I marvel at seeing the signature of the Artist in the assembly manual recorded in every living cell.

    So, do I believe in “intelligent design”? Absolutely. To those who are willing to open their eyes to it, the Artist’s hand can be seen in every property of chemistry, in every law of physics, in the initial conditions of the universe at the moment of creation – maybe even a universe of universes, or even something even more incomprehensible. And the deeper we look, the more amazing and incomprehensibly powerful He becomes in our eyes.

    But to say that our current state of scientific knowledge defines the depth at which the design occurred, or that we’ll ever reach a point at which we’ll be able to offer scientific proof of design, greatly overestimates the capabilities of science and greatly underestimates the magnitude of God.

  4. Thanks for your replies. I've enjoyed your thoughts on Householder's blog also. Always a pleasure!