I might have figured out how to test in a lab for the limit of how far evolution can go, and at the same time test one of the main arguments of the intelligent design movement: irreducible complexity.
I’m using this as an excuse to read up on evolution’s practical limit. I’m starting by making sure that the standard definition of irreducible complexity is good enough for what I want to do. If I find some corner case or technicality that makes the definition unusable, I’ll adjust it to close that loophole. While I’m at it, I’ll debunk a few common arguments against evolution’s limit.
Today’s argument is:
If any machine in a living thing has a part that’s also used for something else, the bigger machine is totally evolvable.
Michael Behe wrote that if you wanted to know if something could appear by evolution, you could use a knockout test (Michael Behe; Darwin’s Black Box; page 39). Knock out each part of, say, a ribosome – one at a time – and see if it still works. If it doesn’t, then that machine couldn’t have appeared by the “numerous, successive, slight modifications” that evolution requires.
One example he gave was of the flagellum – a tail that many single-celled creatures use to move. They’re powered by electro-chemical motors. Apparently, if you knock any one of the 30-ish parts out of them or their motors, they stop working.
But Ken Miller – a well-known proponent of evolution – thinks that Behe missed something important. What if some parts of flagella are also used in other parts of the cell?
In his case, he (or someone else) found that in some single-celled creatures, there’s something that has many similar parts: a type III secretory system.
It’s a needle that stabs neighboring cells and injects stuff into them – usually to kill them. Flagella and type III secretory systems both use a pump to help build themselves, and the designs of their pumps are apparently very similar. The secretory system also uses its pump to squirt stuff into neighboring cells. He argues that because they share many similar parts, flagella evolved from type III secretory systems. So flagella are totally evolvable.
Wait a minute…
The limit of mutation is such that a 1-line, self-copying program can only evolve in a couple of different ways. They get plenty of good mutations, but they’re more of the same types that came before. They don’t build on top of each other to make new machines.
Plus, as Casey Luskin points out, if one of these evolved from the other, the needle would’ve evolved from the tail. There are a lot more species with flagella than with type III secretory systems. So if one part evolved from the other, one would expect that the tails came first. It looks to me like flagella are still unevolvable.
This pump controversy raises an important point for my experiment: irreducibly complex systems can be made of other irreducibly complex systems. How can I tell if a machine is unevolvable? If something appeared by evolution, I should be able to run evolution in reverse, and get a working machine at every step of the way. Muller’s ratchet (devolution) would not cause mutational meltdowns.
- Engine block
- Single-celled creatures with flagella
Tree of Life Web Project
- Type III secretory systems