Teleology is taboo in modern science. It’s not hard to see why: purpose implies design, design implies a designer, and a designer is exactly what the predominantly atheistic scientific community does not want to admit. But yet, we use telic language–specifically the language of engineering–to describe concepts in biology all the time. Is it just because we know of no better way to describe systems than by analogy to those that we ourselves have designed? Or is there something deeper to the compelling similarity between “molecular machines” that we discover inside the cell and machines that we use every day? It is such questions that drive Mike Gene’s brave new book: The Design Matrix: A Consilience of Clues, which aims to refocus the neverending debate over purpose in nature away from the black-and-white arguments of days past into a careful investigation of the actual evidence.
Mike (Dr. Gene, I presume?) is at his strongest when describing the intricate details of molecular machinery, and in particular that of DNA replication, mutagenesis, and error correction. He clearly describes several features of cellular replication that are salient to the debate over teleology. For example, the genetic code that we see in almost all organisms today is remarkably optimized. It allows for redundancy that reduces the chance of DNA mutations causing changes in protein structure, but on the other hand, the most common mutations due to the intrinsic chemistry of DNA lead to amino acid changes that increase hydrophobicity, increasing the likelihood of secondary structure and protein-protein interactions. There is also no evidence for precursor codes, and the variants that do exist (such as the mitochondrial code) are better explained as divergent from the universal code rather than primordial remnants. The numerous mechanisms within the cell to ensure the fidelity of DNA replication, RNA transcription, and protein translation, when compared with what we know about codes in general, are also good evidence for teleology in biology. Even more intriguingly, some bacteria have a mechanism for increasing the rate of mutations in their genomes in response to certain stresses, which along with the general trend toward hydrophobicity in proteins suggests that evolutionary mechanisms may be coopted by organisms to increase their complexity and chances of survival. One minor quibble: I would have liked to see some discussion of the “RNA World” hypothesis and how it fits (or doesn’t) with the possibility of evolution of the genetic code.
Mike’s discussion of nanotechnology and its similarities to molecular biology is also excellent. I had not appreciated the degree to which scientists studying nanotechnology are turning to biology for inspiration. He makes an excellent point when he points out that there is a remarkable convergence between the direction of engineering (toward smaller and more complex design) and what already exists within the cell. It is certainly resonable to argue from this evidence that the “molecular machinery” within the cell exhibits characteristics of design, especially as we refine our own designs by studying ever more closely analogous cellular structures.
I also appreciated, on a completely different note, Mike’s discussion of what he calls “inductive gradualism” as a method of studying teleology vs. non-teleology in nature. He presents an “explanatory continuum” between “X could not have possibly evolved” and “X certainly evolved”, as opposed to what is the common approach to these debates: admitting only those two possibilities and not allowing for varying degrees of uncertainty about the telic content of nature. His continuum is certainly better aligned with how scientific research ideally ought to proceed: gathering evidence gradually until a conclusion approaches inevitability, but always being open to other clues leading in a different direction. Contrast that with what is often heard in “ID vs. evolution” debates: on one side, any evidence that remotely points toward “design” or seems to not admit evolutionary explanations is held to “disprove evolution,” while the other side adamantly refuses to accept any telic explanation. Such a dichotomy is, as Mike demonstrates, simply bad science–and more importantly, poor philosophy.
However, in the end, despite my hearty agreement with Mike on the evidence for design in the genetic code and the merits of an inductive approach to studying biology, his book has some serious weaknesses, ironically in the final section where he lays out his “Design Matrix”. He lays out four criteria for discerning design: analogy, discontinuity, rationality, and foresight, and suggests that by scoring various biological structures on these areas we can come up with a “score” that correlates with a probability of teleology. If the four criteria were equal to each other, and if they could be reliably scored by different observers, then perhaps the Matrix would be a valuable tool. But, I fear that it is not:
1. Analogy: Mike suggests that we can infer an increasing probability of design if a biological feature has analogous aspects to things that we know to be designed. On the surface, this seems commonsensical. If we found a molecular machine akin to a rotary engine (the F1-F0 ATPase), then we might conclude that it was designed as was the fascinating engine in my old Mazda RX-7. But, I think that Mike’s excellent observations about nanotechnology subtly undercut the argument from analogy. Our designs ever more closely converge upon nature. Are we unwittingly approaching the same minima in the multidimensional fitness landscapes associated with doing particular kinds of work? I wonder if our own refinements of designs (such as Mike’s example of the Chevy Corvette) mirror nature not because of design per se, but because we are converging upon the best solution to a particular problem.
2. Discontinuity: This criterion is entirely based upon Michael Behe’s concept of Irreducible Complexity. Mike admirably explains the evolutionary explanations that purport to explain away Irreducible Complexity. In the interest of time, I will not go into why I find the possibility of cooption to adequately explain away this teleological inference. If a biological mechanism can be found that has multiple components, which cannot be reconstructed by gene duplication or other mechanisms from components found in simpler organisms, then I suppose that IC would be tenable. But, until that point, scoring features based upon discontinuity with the past is entirely too subjective. Even the genetic code, which is the example nonpareil of discontinuity, may have evolved from a simpler RNA (or other nucleic acid) precursor. My biggest concern with this criterion, however, is not scientific but theological. Another word for a discontinuity is a gap, and I fear that we place the Designer squarely into those gaps when we rely upon this explanation of design.
3. Rationality: Mike lists six features of rationality, which merit separate consideration.
a. Efficiency: While I agree that efficiency–that is, “using the mininum number of parts to carry out an objective,” is a hallmark of good human design, to judge efficiency is tricky business. How do we discern whether a system contains “needless complexity” as a way of making it more likely to be designed? Often in biology, we’ve thought a component superfluous, only to discover its function later. Take, as one example, the ribosomal protein L11. It isn’t a structural component of the ribosome (which for the uninitiated is the “molecular machine” that translates RNA into proteins), but yet is critically important. Free L11 (not in place in the ribosome) acts as a signal to activate the master stress regulator p53, causing the cell to either arrest its growth or die when protein synthesis is disrupted. Yes, the actual number of parts of biological systems often exceeds a bare minimum (the ribosome is an excellent example), but we simply don’t know whether these other parts evolved (or were designed) to have additional functions that we’ve yet to discover.
b. Specificity: While it is true that humans typically design things with “precisely specified interactions,” we are learning gradually that the “messy” and “error-prone” systems in biology are not necessarily the worse for wear. For example, the HIV virus takes great advantage of its error-prone reverse transcriptase, which generates mutations at a rate exceeding the capacity of the human immune system to adapt. “Error-prone” DNA polymerases aren’t particularly good at putting the correct nucleotide opposite its cognate partner in DNA replication, but if they weren’t able to stick nucleotides into less than optimum conditions, replication opposite damaged bases (due to radiation, enviromental pollutants, etc) wouldn’t be able to occur (and we’d be in big trouble as a species). I guess what I’m trying to say is that specificity in design isn’t always a good thing, and being “messy and error-prone” isn’t always bad. The question is whether there’s something in the error-prone system that we’re missing when we judge it not to have been designed.
c. Robustness: Yes, this is a feature of design, but I’m not aware of many biological systems that don’t possess it. Our cells and bodies are remarkably robust.
d. Elegance: I’m not sure why Mike included this, as he admits that it is a subjective measure. As I spend more time studying the inner workings of the body and molecular biology, my definition of elegance changes. I find things that I once thought sloppy, such as the production of mutations in the human genome leading to cancer, elegant because the alternative: slowing down replication to make it more accurate, would make human life impossible.
e. Flexibility: See my discussion of efficiency. Because these two features are usually in balance, it’s hard to judge them either for or against design.
f. Coherence: Mike suggests that a balance of the other five features, that is, a coherent function, is a hallmark of design. I agree, but I question how we are to actually judge this. Do we have a more coherent design of a system to compare against? Does a system (such as mitosis in primitive eukaryotes) actually need to be as carefully precise as it needs to be in higher eukaryotes. If it doesn’t need to be, its “messiness” is not an indicator of incoherence.
In all, my beef with Mike’s hallmarks of rationality is that they’re all terribly subjective. I find that, in studying biochemical systems in great detail, I’ve always proceeded on an assumption of rationality as a way of hypothesizing what a particular feature does. I suppose I give “evolution” more credit than some people, or perhaps on the contrary, we biochemists use a “design inference” far more than we admit.
4. Foresight: To conclude this rambling review, I find Mike’s fourth criterion of design to be actually the only criterion really worth considering, because it is the only one for which there cannot be any non-teleological alternative. Evolutionary processes can give the semblance of rationality; they can give the semblance of discontinuity; they can certainly appear analogous to human systems. However, by definition, they cannot operate with any foresight. If a unicelluar eukaryote’s genome contains proteins that serve it no critical function but are essential for multicellular life, that points toward “front-loading.” If the universal genomic code is so optimized as to both minimize errors and promote beneficial mutations, as Mike suggests, then that points toward clear foresight, particularly as unicellular organisms are hardly as concerned about maintaining genomic integrity as are long-lived multicellular ones. I also find the intricate structure of the genome itself evidence of some foresight, as “junk DNA” is proved more and more full of treasure. Would the timeline of life, if replayed, produce the same results? Gould says no, Simon Conway Morris (in Life’s Solution) says yes, and I think that Mike and I heartily add our votes to the “yes” category.
Perhaps the major difference between Mike and I is that I still operate too much from the “black-and-white” category. But, I’ve found that our colleagues in science are unwilling to even consider “clues” for design if they can be cast in an evolutionary framework. Mike has done us a great service in laying out the evidence he sees for design and an interesting, if flawed, new way of assessing it, but in the end, only pointing the way toward foresight in biology is going to make any headway in convincing the majority of scientists to start considering a Designer. For, when all is said and done, the issue is not one of science at all, but of philosophy and theology. We see evidence of a Designer everywhere who look for Him; but, those who close their eyes will take much to convince. As Emile Zola said:
“Were I to see all the sick at Lourdes cured, I would not believe in a miracle.”