Our thoughts and prayers are with you.

Sal Cordova recently speculated on the relationship between the concept of irreducibility as defined in mathematics and physics and Behe's irreducible complexity. Some link-following took me to a paper by Chaitlin describing the ramifications of algorithmic information theory (and its roots in the work of Godel and Turing) for mathematics, physics, and biology.

I'm not as certain as Sal that the work of Godel, Turing, and Chaitlin can be straightforwardly applied to support ID - or that the concept of irreducibility in AIT correlates well with Behes' Irreducible complexity[1].

However Chaitlin's paper certainly illustrates some trends in the sciences that are favorable to ID: the discovery that the world around us (not just in biology but also in physics) is fundamentally informational, and that it can be fruitful to think of biological and physical systems as computers which process that information. The fact that Chaitlin goes so far as to predict the convergence of theoretical physics and theoretical computer science shows just how deep the similarities go. These discoveries provide strong support, I think, that the workings of a mind are behind the phenomena studied by the sciences.

---

[1] There seem to be intriguing similarities to specified complexity, though. A system could be irreducibly complex without being algorithmically irreducible (that is, it might break if you remove any one piece, but not be the most efficient method for producing a given result). On the other hand, AIT irreducibility is strongly tied to minimum description length, which is an explicit part of specified complexity.

Update (2/19): I'm reading Dembski's paper Specification: The Pattern That Signifies Intelligence and just discovered that his formulation of specified complexity makes explicit use of the concept of minimum description length as described by Chaitlin (and others). So I'm a little late to the party. Shows how much I know :-)

I recently re-watched an old (April, 2006) Intelligent Design debate between Stephen Meyer and Peter Ward (you can find it online). The debate itself is not very balanced. Essentially Meyer gives a well-reasoned, thoughtful defense of ID for an hour and a half, and Ward interupts periodically to interject "It's not science". To racous applause, no less! It's not that Ward wasn't given equal speaking time, it's that he didn't really use his time to say much of anything with substance.

There is certainly a more convincing case to be made for evolutionary theory than Ward presents. What is really telling is the fact that he doesn't take ID seriously enough to think it worthy of any serious rebutal. Ward has serious academic credentials, but you would never know it from listening to the debate. He doesn't seem to think he needs to utilize them.

But what I really wanted to highlight is a clever argument which Meyer makes. He notes that Darwin and Charles Lyell (a geologist who had a major influence on Darwin) sought to explain their observations through "presently acting causes". The idea is that it is improper to invoke exotic causes to explain your observations. Observations should instead be explained by the forces which we see at work today.

Lyell was the first to explain geological formations by the accumulation, over long periods of time, of the processes of erosion which we observe today. Darwin sought to explain life the same way: as an accumulation of small, successive variations of the type observed by breeders (in fact, the very name "natural selection" was chosen because Darwin explicity compares it to artificial human selection).

Meyer takes this principle and extends it to information: we now know that living cells are the most complex machines on the planet, far outstripping anything that humans have designed and containing reams upon reams of digital information quite analogous to computer code. What, he asks, is the only presently acting cause of the sort of information that is found in computer programs? Intelligence! We observe that intelligence is capable of producing just exactly what we find in the cell. We observe that to get a complex new function out of an existing system, you must add code (information) via an intelligent agent. There is no evidence whatsoever that the near-limitless capacity for variation which Darwinian evolution requires is achievable through undirected processes. On those grounds we ought to prefer the presently acting cause, intelligent design, over exotic hypotheses about self-assembling molecules in the primordial soup, or the emergence of exquisite nanotechnology from random mutation.

William Bradford, over at Intelligently Sequenced, talks a lot about the evidence of design from cellular DNA repair mechanisms. The genomes of modern cells rely for their integrity on incredibly complicated methods of error-detection and repair. Focussing ID criticism here highlights a fundamental weakness in darwinian origin-of-life scenarios: In order for natural selection to improve a self-reproducing molecule, it must have a mostly reliable method of reproducing itself. Otherwise fit individuals would not be able to transfer their fitness to their descendants.

As we now know, it cannot just be taken for granted that a primitive cell (or even a pool of replicating rna) without repair mechanisms  could maintain its genetic information with enough accuracy to survive, let alone ensure that the descendants of fit individuals were likely to retain their parent's fitness advantage. This is a major obstacle to abiogenesis (as if spontaneously generating self-reproducing molecules wasn't hard enough :-).

Last year, I defined "isomorphic instantiation" as the phenomenon in which a complex technology developed by intelligence is subsequently found to exist in nature. In a sense, isomorphic instantiation is the inverse of biomimetics, since biomimetics is the discipline of making technology that mimics the marvels of nature.

We can see a example of isomorphic instantiation in a paper entitled "Biplane wing planform and flight performance of the feathered dinosaur Microraptor gui" published yesterday in the Proceedings of the National Academy of Sciences. If you don't have access to this paper, you can read news accounts. The paper suggests that the Microraptor had 4 wings in a configuration strikingly similar to the biplane. The authors conclude:

Aircraft designers have mimicked many of nature’s flight "inventions," usually inadvertently. Leading edge slats delay stalling, as does the alula of birds; birds’ feet act as airbrakes, and streamlining reduces drag. Now, it seems likely that Microraptor invented the biplane 125 million years before the Wright 1903 Flyer.

Isomorphic instantiation is a good indicator of intelligent design. We know that intelligent agents (namely, the Wright Brothers) invented the biplane. When a feature in nature, like the Microraptor's biplane wings, is strikingly similar to an invention already designed by intelligent agents, one must entertain the possibility that the feature in nature may have also been designed by an intelligent agent. We can't just automatically assume that complex technologies invent themselves. After all, did the Wright 1903 Flyer invent the biplane or did the Wright Brothers invent the biplane?

I recently ran across a paper with the intriguing title Evolution based on Design-by-Contract: Origin of Life through an abiotic double-stranded RNA world by Albert De Roos.

Design-by-Contract is a computer science methodology for keeping software easily maintainable and changeable by reducing dependencies. Essentially, it advocates treating functionally distinct blocks of code as black boxes. These code blocks define a specific set of methods (interface) which are "public". Other code blocks interact with it only through these specific methods. As long as the interface remains intact, the internals of the code are free to change radically without fear of introducing bugs.

The author draws an analogy to origin-of-life research, noting that while many people believe single-stranded RNA was the first self-reproducing molecule, the transition from single-stranded RNA to DNA faces significant problems:

It is difficult to imagine ssRNA to have a dual function since the catalytic properties of RNA depend on the three-dimensional structure, while the informational capacity would require a simple linear structure that can be replicated (Taylor, 2005). The folding of an RNA molecule would prevent its own replication, while for replication a folded ribozyme would be necessary... From a system perspective, a dual function in a single entity would create a dependency that would reduce evolvability of the system.

The author goes on to explain how double-stranded RNA might be changed into DNA through a process that maintains existing interfaces and requires fewer radical changes. Central to his thesis is the idea that RNA started out as an information-carrying molecule. This makes information a pre-requisite for any self-replication.

The paper doesn't make any mention of a designer, however. In fact it speculates about how self-replicating double-stranded RNA might have arisen on the early earth, and explains the gradual steps which the author claims are sufficient to transform double-stranded RNA into DNA. Does it count, then, as an ID paper? I think so. Modeling biological systems with engineering paradigms is an obvious application of ID. If this approach is fruitful, then ID will be a successful research program.

I've said before that the standards of proof which design theorists require for an evolutionary inference is wildly different from the standard accepted (and usually given) by evolutionists. I've just read a new paper by Pallen and Matzke, From the Origin of Species to the origin of bacterial flagella, and it perfectly illustrates the disconnect. My comments on some quotes from the paper:

By even the most conservative estimate, there must therefore be thousands of different bacterial flagellar systems, perhaps even millions... either there were thousands or even millions of individual creation events, which strains Occam's razor to the breaking point, or one has to accept that all the highly diverse contemporary flagellar systems have evolved from a common ancestor.

First of all, this is rather sloppy use of Occam's razor. All things being equal, a simpler explanation should be prefered to a more complex one. But are all things equal? That is the interesting question. Pallen & Matzke gloss over it, assuming the truth of the controversial premise. This is blatant begging of the question.

Secondly, I don't know if all bacterial flagella descended from a single ancestor by a process of natural selection acting on random variation. They may have. But it takes more than common ancestry, inferred from protein sequence similarities, to decide the question. Common ancestry in itself says nothing about the processes acting on the (simpler?) ancestral flagella which caused it to resemble a modern one over time - only that it did.

What would it take to establish such an evolutionary hypothesis? Real evidence (as opposed to inference from historical data based on naturalistic assumptions) that natural selection is capable of doing that sort of thing.

Thirdly, Pallen & Matzke are clearly relying on a theoretical principle rather than experimental data for the force of their argument. It is fairly obvious that they do not consider creation an allowable explanation. This is fine and good so long as the principle is subservient to the data. But principles can be wrong. They have been wrong. For example: at the beginning of the last century, almost all physicists believed in an infinite, steady-state universe. The concept of a universe with a finite age was actually repugnant to many of them, on purely philosophical grounds. When the Big-Bang theory was first proposed (by a Christian, incidentally - someone with a prior commitment to a creation event) it was opposed for that reason for quite a while before the evidence overwhelmed the sceptics. I wonder if the same thing is in the process of happening in biology.

...but if one accepts that all current flagellar systems diverged from their last common ancestor (the ur-flagellum), why stop there? All flagellins show sequence similarity indicitave of common ancestry... Therefore [on the basis of homology with non-flagellar proteins] the flagellar rod-hook-filament complex has clearly evolved by multiple rounds of gene duplication and subsequent diversification.

I really don't know what to say to this. If protein homology is all the proof of darwinian evolution that you needed, then we should shut down this blog and go home. But it seems clear to me that it is not. Citing homologies does no prove what Pallen & Matzke seem to think it does.

From the above discussions of sequence hommologies and modularity, it is clear that designing an evolutionary model to account for the origin of the bacterial flagella requires no great coneptual leap.

As I have argued before, the hand-waving about homology and possible precursor systems does not approach the level of detail needed to convince anyone familiar with the implementation of complex systems that it is as simple as it sounds.

the flagellar research community has scarcely begun to consider how these systems have evolved. This neglect probably stems from a reluctance to engage in the 'armchair speculation' inherent in building evolutionary models
....
it is no longer enough to say, for example, that bacterial flagella evolved and that is that. Instead, scientific experts have to engage with a sceptical public.

I am greeting you early since I won't be blogging during the holidays. Advice to all workaholics: take some time off!

There's an article in the Sept. 28 issue of Nature discussing the regulation of mating type in two yeast species, c. albicans and s. cerevisiae (regretably, the text is only available if you're on a university network with access to the Nature archives). I'll spare you the gory details, but while the two yeast share the same observable trait (phenotype), the underlying genetic mechanism is completely different.

What is most interesting to me about the article is the detail in which a hypothesis is presented for the evolution of both systems from a common ancestor. The authors list the primary modifications necessary to change one trait into another, note that only a few mutations in one particular protein binding site would be needed to effect one of the modifications, point out that a third yeast species seems to represent an intermediate step between the two, and even suggest future work which includes engineering a strain of s. cerevisiae with the c. albicans mating system, to study what the ancestral yeast might have looked like.

This hypothesis is not only detailed but phenomenally testable (ironically, it involves neutral evolution with no net advantage to the organism). Compare this to the evolutionary arguments for the origin of the bacterial flagella, which rely exclusively on distant (and some not-so-distant) homologies and a good deal of story-telling and hand-waving.

The difference is immediately clear. If all evolutionary inferences were held to the standard of proof present in this nature article, I would have no objections. What I object to is the assumption that natural selection acting on random variation must be sufficiently poweful to fill the gaps between whatever homologies we happen to find in nature.

Can someone give me a good formal definition for intelligence? John McCarthy, a founding father of AI (artificial intelligence) thinks that there does not yet exist a solid definition for intelligence that does not depend on relating it to human intelligence. Here is my proposed informal definition: intelligence is the ability to achieve goals by constraining natural chance processes.

Consider the example of a scantron multiple choice test. Jack totally forgot to study for the test, and so his only recourse was to fill in the blanks at random. Jill, however, studied hard and was able to select the correct answers. We can say that Jill is more intelligent than Jack. While Jack relied totally on chance, Jill was able to curb chance and select the right answers, in order to achieve the goal of getting an A on the test.

So in order to detect artifacts of intelligence, we have to find places in nature where chance was purposefully constrained. This is perhaps where Dembski's explanatory filter comes to play (i.e. if not regularity, then chance; if not chance, then design). We might have to modify the filter a little: if not regularity, then unconstrained chance; if not unconstrained chance, then constrained chance (= design).

Note that intelligence does not need to be the antithesis of chance; rather, intelligence can be the constraining of chance. An intelligent agent can sometimes use constrained chance as a useful tool (e.g. the sorting algorithm known as Quicksort).

We have some catching up to do on this blog! I hope everyone is enjoying the holiday season. As an early Christmas gift, here is a bag of design/evolution links:

• Check out the Time article, God vs. Science (HT: Steve). It's a dialogue between Richard Dawkins, the foremost spokesman for evolution, and Francis Collins, director of the National Human Genome Research Institute. Although Collins is a theistic evolutionist, he does seem to argue for cosmological ID and fine-tuning (see a similar argument in my fine-tuning article).
• New Scientist critically examines Biologic Institute, a lab that is friendly to intelligent design (HT: Evolution News).
• Evolution News has reported that Judge Jones, in the famous Dover intelligent design decision, copied 90% of his section on whether intelligent design is science from an ACLU document. While some skeptics are arguing that this is a widespread legal practice, there is at least one skeptical biochemist who is frowning on this copying.
• Catch a glimpse of some good design with this cell animation from Harvard (HT: TT).
  

We administered a one-question survey at the iDesign booth during the Anteater Fair. This is a highly unscientific poll, since the sample size is way too low to confer any statistical significance. We asked:

"Which Theory Of Origins Is Closest To Your Personal View?"

Here are the results:

A. Darwinian Evolution (35%)
B. Theistic Evolution (6%)
C. Intelligent Design (16%)
D. Creationism (10%)
E. Don't Know / Don't Care / Won't Share (32%)

I'd like to administer a more thorough survey sometime during fall quarter.

The suboptimality argument against ID is looking more forceful than I originally gave it credit for. On the surface, it does not seem worthy of much serious consideration - merely a theological objection meant to make certain theists feel uncomfortable about God's competence.

But if it is interpreted as an argument against the validity of ID as a legitimate research program, it gains considerably more weight. Here is the problem: Design theories must be predictive. They must explain features of the natural world better than their competitors, suggest fruitful new areas of exploration, and make verifiable predictions.

Is it predictive to simply draw design inferences from isolated biological structures, without explaining why life sometimes appears elegantly designed, sometimes incompetently designed, and sometimes undesigned (i.e., the result of evolutionary processes)?

Well, not so much. I don't mean to disparage the legitimacy of isolated design inferences, but they would gain much more weight if they could be placed in a framework that explains why life does not appear to be uniformly well-designed and (crucially) where we might or might not expect to find design in the future.

This is not a small problem. In fact, I think it is one of the central challenges of ID: not just to create a statistical test for design, but to craft a creation story that incorporates and supercedes the evolutionary one. A story that provides a framework for discovering precisely what was designed, where, and when.

This is not to say, by the way, that evolution does not face the same dilemma of reconciling the existence of both elegance and kludge in nature. It is not really taken seriously by evolutionists, however. Evolution is smarter than you when it produces elegance, and dumber than you when it produces jury-rigged contraptions. No further explanation is required or expected,  because time and natural selection are presumed capable of creating, well, whatever we discover.

This is why I think design theorists are in a better position to explain the disconnect than evolutionists. We take the problem seriously (or we ought to), because ID allows us to distinguish between the mechanisms responsible for elegance and kludge. That is, for ID the question is distinguishing chance from design, for evolution it is getting design from chance. Better to have a problem of sub-optimality than a problem of optimality.

How can design theorists address this problem? First of all, we need to stop accepting suboptimal design as a real but unimportant (to ID) feature of the world. The very thing that makes ID falsifiable is that it makes certain minimal assumptions about the designer - for instance, that he designed in such a way as to be empirically detectable (as opposed to simply mimicking chance). In accordance with what we know about the molecular machines in the cell, I think we must also assume that the designer is not only competent but a great deal more advanced than we are. If ID is to make predictions, it will not do to say that he alternated unpredictably between elegant and inefficient designs.

If instead there is a reason that some biological structures appear elegant and others appear inefficient, it should be discoverable. In order to discover it, however, we need a specific theory of design. Exactly when did design occur? In the first cell and at the cambrian explosion are hypotheses which I have heard, though I think some would see design as more pervasive than that. We need to know. Occam's razor will be helpful here - frontloading and fewer design events should be preferred over more pervasive design, all things being equal.

From the design events, we can infer a trajectory. How did design progress? In an analogous way to technological evolution? Not at all (that is, designed structures appear fully complete and do not improve)? Some combination of the two? And what role does (or doesn't) evolution acting on front-loaded information play?

All of this should help us pin down where sub-optimality arises. For instance, if designed structures appear fully complete and fairly early in the history of life, you would expect most highly-conserved structures to be elegant and efficient. Sub-optimalities should be relatively unconserved and recent in the tree of life - presumably due to the accretions of chance, and in principle distinguishable from design. 

A fruitful theory of design such as this, which explains sub-optimality and successfully predicts where it ought/ought not to appear, would go a long way toward legitimizing ID as a science.

I wish to welcome everyone to another year at UCI. Thanks to all those who stopped by the iDesign booth in the park today! We plan to put together some great events this year. Last year, we showed some documentaries and had some discussions and put together a well-attended colloquium on origins.

If you are curious about our club, please read the blue sidebar on the left (which includes the mission statement and the FAQs).

Note that our club is open to all students, whether one be a creationist, evolutionist, ID proponent, or undecided/unsure. We think that intelligent design, evolution, and creationism are interesting and important topics to discuss, and we attempt to facilitate this discussion in a friendly environment.

For those who are interested, here are some introductory articles about intelligent design that I have written:

• Darwin Or Design?
• Fine-Tuned For Life

Also, all of iDesign's members are serious about excelling academically. New students may want to read the following article:

• Ten Tips For Academic Success

Feel free to explore the archives of this blog. Best wishes on your academic endeavors.

Intelligent design suggests that at least one aspect of nature is likely a product of intelligence. In other words, there exists an aspect X such that X is in nature and X was likely to be designed. Some possible candidates for X are "fine-tuning," the "bacterial flagellum," or the "living cell."

Darwinism (in a broad sense) would suggest that everything in nature is a product of unintelligent stochastic processes. In other words, there does not exist an aspect X such that X is in nature and X was likely to be designed.

So it seems like a good approximation to say that Darwinism and Design are mutually exclusive and exhaustive. The weaker one concept is, the stronger the other concept becomes.

I found this at Uncommon Descent:

When we visited the zoo the other day, my wife snapped this photo just outside of the Panda play area. I guess when they put together the verbage for the sign, they neglected to consult Gould because I didn’t read “looks jury-rigged” anywhere on there.

Further down in the comment thread, Markus Rammerstorfer has posted snippets from a Nature article on the Panda's Thumb:

The way in which the giant panda, Ailuropoda melanoleuca, uses the radial sesamoid bone — its ‘pseudo-thumb’ — for grasping makes it one of the most extraordinary manipulation systems in mammalian evolution..... We have shown that the hand of the giant panda has a much more refined grasping mechanism than has been suggested in previous morphological models.
Endo, H et al. (1999): Role of the giant panda’s ‘pseudo-thumb’ Nature 397: 309-310

Sub-optimality is not, of course, a decisive blow to ID. Nevertheless, it's nice to see another example of how first assumptions based on evolutionary theory have been proven incorrect by further research. At the very least, it shows that sub-optimality claims need to be based on solid research, and not the ad-hoc "it is obviously not optimal" claim which passes for an argument against design in some circles.

There is an interesting exchange going on between Salvador Cordova of Uncommon Descent and Dave Thomas of Panda's Thumb.

First, Dave posts about his genetic algorithm for finding Steiner trees ("the shortest networks of straight-line segments connecting a given collection of fixed points"). He claims his algorithm (explained in detail in an earlier post) is not targeted because it does not contain the optimal solution, but selects for incrementally better ones randomly. He challenges ID'ers to come up with a solution to the 6-node Steiner Tree problem. I think his aim was to prove how much better his genetic algorithm was than the intelligent attempts of design proponents.

Next, Salvador responds with a genetic algorithm he designed, which correctly finds the sum of the first 1,000 integers. The algorithm does not have a direct target (i.e., it is not explicitly searching for the answer), but instead uses Gauss's algorithm - obviously targeted in a secondary sort of way, which he implies is akin to Dave's Steiner tree algorithm.

Finally, Dave insists his algorithm is not targeted and is in no way analagous to Salvador's, which is blatantly targeted, and Salvador replies that of course his algorithm is targeted, and that the smoke and mirrors in his algorithm are indeed similar to Dave's.

The contentious issue here is what, exactly, constitutes a "target". Salvador's point is that if your cost function is intentionally designed to approach the solution you are looking for, the absense of the actual optimum value is a technicality. Dave's algorithm is clearly only good at finding Steiner trees because he designed the cost function to reward better Steiner trees. As johnnyb points out in a comment to Salvador's latest reply:

The problem is that the fitness function cannot be directly related to the problem being solved. Natural selection simply says the animal must survive or be better at reproduction, but somehow guides processes such as eye formation. Therefore, in a fitting scenario the selection algorithm should not directly correspond with what you are searching for, in fact there should be a large disconnect.

 This is exactly right, and so succinct I wish I'd come up with it myself. There may or may not be a genetic algorithm for constructing a structure like the human eye (depending on the number and structure of interdependencies between its parts). If one exists, then its cost function will be specifically designed for producing an object like the eye, whether the "target" is explicit or not. There is no evidence that a vague cost function like "survive" is capable of doing anything interesting.

A recent article in Nature explored the means by which certain proteins called chaperonins work. Chaperonins are proteins which help other proteins fold correctly. Specifically, they provide a "cage" in which the target protein can fold without interference. Artificially growing or shrinking the size of the cage can make certain proteins fold more quickly or slowly, but the overall size of the cage seems to be optimal given the number of proteins which this chaperonin interacts with.

Because the cell is so crowded with large molecules, it is energetically favorable for newly transcribed proteins to fold compactly, reducing the volume they occupy. If it weren't for chaperonins, this tendency would cause partially complete amino acid chains to bind with each other (there are many of these chains in close proximity because mRNA is transcribed by many ribosomes at once).

The upshot is that chaperonins use the energetic effects of crowding to stimulate folding while at the same time eliminating the negative effects of crowding - promiscuous binding to anything in the area. This is a neat trick. In the final paragraph of the article, the author (who, of course, credits natural selection with chaperonin design) even exhorts human designers to follow the example of chaperonins:

It is a testament to the ingenuity of natural selection that the chaperonin cage not only combats aggregation caused by crowding outside the cage but also uses crowding to accelerate protein folding inside the cage. Nanoengineers trying to improve the yield of therapeutic proteins could profit from studying the tricks of the chaperonin nanocage.

PZ Meyers has an interesting (and a bit over-my-head) post on the evolution median-fins over at his blog Pharyngula. Snipping the biology and leaving the punchline, he says:

A story is beginning to emerge, though, that shows that midline fin development and evolution is a wonderful example of a general principle: modularity and the reuse of hierarchies of genes.... The redeployment of previously refined genetic modules is going to turn out to be a universal property of evolved systems, I expect.

Modularity (perhaps even hierarchical modularity) in biological systems is becoming more and more evident, and I agree that we should expect to see a lot more of it in the future. But which is a more likely source of good engineering principles like modularity? Stochastic processes, or intelligence? Might there be different sorts or degrees of modularity that could serve as predictions of one theory or the other? In particular, is there any sort of modularity that is not predicted by evolutionary theory?