Can natural selection create: Part 8: Irreducibly complex. (See Figure 6)
BQ: We recall from our recent line of questioning that being irreducibly complex means that a particular item needs all of its parts to function, like a mousetrap that, missing any single part, simply does not function at all.
There is a counter-proposition that each part functions quite well as something else if we take the mousetrap apart. One scientist explained that all the parts of the mousetrap could be used as various things—a hook, a paperweight, etc.—and then happen to come together at the same time in a synergistic evolutionary co-opting. For some, this seems like the needed answer! Tthese parts can do other things! But looking at the idea holistically, it quickly becomes apparent that it's an appeal to, well, making an infinite number of fairy tales. "Perhaps this did this, this thing did this, and this did...and then BAM, they all became one." From our previous look at poly-functional DNA and other concerns, we can now soundly refute this sort of idea. It doesn't work that way in genetics, and it doesn't even work that way with a mousetrap! Furthermore, that proposition still requires simultaneous synthesis, which is the problem in the first place.
So let's think about an old example we used of a red wagon, which has a limited number of parts. We were hand-copying instruction manuals (DNA), and seeing if the wagon would ever become a space shuttle over billions of years of hand-copying and keeping only the best instruction manuals to copy from each time. It didn't of course, even with intelligence thrown in. But what if we delete a wheel on accident and make a tricycle? Well a 3-wheeled wagon is not a tricycle, it's simply a broken wagon! This is the problem of irreducible complexity: as complexity increases, the need for intermediary forms would require massive reworking of the instruction manual (GENEtic code) and component parts.
Now let's consider a gene. A single gene has about 50,000 component parts! That's more than a car, and yet we can't create a linear path to make a car piece-by-piece, with each stage doing something USEFUL. For example, a pitman arm and drag link do nothing by themselves other than flop around. Yet this gene needs to have 50,000 parts, and is itself a minute portion of irreducible complexity within a galaxy of irreducibly complex parts, all of which must be working in an orchestrated way. Combine that with what we know about natural selection's ability to create a gene, and it's clearly impossible.
(PN244)
Figure 6
Can natural selection create: Part 9: Nearly neutral positively kills. (See figure 7.)
BQ: In our ongoing look at what we have demonstrated as natural selection’s inability to create even a single useful new gene, we need consider something that is often not discussed.
In earlier writings, we saw that most harmful mutations are nearly neutral. That is, they are relatively insignificant in terms of the health of a being. If the mutation is not massively harmful (such as a genetic disease that kills a child within 8 months), other selection factors make it hard to remove from the genome, and it is passed on. Think of it like this: if you make a copy on the copier, the copy isn't as good as the original. It has slightly negative qualities. These "harmful" mutations, however, are not enough to make the copy non-viable. The copy will still be used, and itt would take a very bad copy to be rejected.
However, a minuscule proportion of the time (note: this tiny proportion is a mutation within an already-existing gene; to create a new, useful mutation is infinitely rarer to the point of impossibility), there is a chance for a positive mutation occurring. We need these beneficial mutations during the construction of our new gene. The problem we encounter is that they, like the harmful near-neutrals, are almost certainly going to have little impact. Let's do some analogous math using brain-friendly numbers.* For every 10,000 harmful mutations, 9,900 of them are nearly neutral—not good, but not too harmful. 100 of them are fatal. As a ratio, we'll say that for every 100,000,000 harmful mutations, we have 10,000 helpful mutations. Unfortunately, of those 10,000 helpful mutations, 9,900 of them are also nearly neutral—slightly helpful, but nothing entirely useful.
Now step back and look at this in light of our new gene: the vast majority of our new gene is going to be defined by these nearly-neutral mutations! The nature of these nearly-neutral beneficial mutations presents compelling evidence for genetic design, as there is simply no way to build a gene one nucleotide at a time through natural selection.
*For the hard numbers, see our previous looks at this topic.
(PN245)
Figure 7
Can natural selection create: Part 10: Bad mutations are real and other considerations.
BQ: In our look at creating a new gene, we've excluded deleterious (harmful) mutations so that we'd have a leg up. However, in the real world, deleterious mutations outweigh beneficial mutations by perhaps a million to one. Ouch! Now let's look at other considerations.
Muller's ratchet: We've mentioned this before, so I won't define it here. In a strand if DNA, recombination of material happens mostly between entire genes, and not between nucleotides. Mutations happen as individual nucleotides. This means that good mutations and bad ones can't be separated. Bad mutations, as above, outnumber good ones by an unconquerable amount. If we DID manage to get 100 good mutations, time would actually be our ENEMY in evolution, as good mutations would start to back-mutate into harmful ones while we tried to get enough good mutations to make a good new gene. Thus term, "Muller's ratchet." (See Muller, Loewe, Gabriel et al., Tishkoff, and Verrelli.)
Too much cost: If we could identify half of the population with the worst problems and kill them off, we could calculate the "selective cost" and remove bad mutants. We have seen before that this cost would actually destroy the species in the long run, as reproduction wouldn't keep up. If the selection cost to remove harmful mutants is too much, we certainly can't "pay" for the removal of them and ALSO have genetic "funding" to pay for progressive/beneficial selection.
Non-random mutations: Mutations actually aren't entirely random. In fact, some genomic areas are far more likely to mutate than others. This is actually a problem. While we can get a good mutation in a "hot spot," we have to wait for the cold spot to also produce a good mutation. Again, time is our enemy. The hot spot will back-mutate and, ironically, this non-randomness slows down progressive selection and is antithetical to novel genetic architecture.
(PN246)
Can natural selection create: Part 11: Wrapping it up.
BQ: Research, reason, and applied genetics show that natural selection cannot innovate new genes within the evolutionary timespan, even when harmful mutations aren't present. When harmful mutations are present, backward-mutations outnumber forward mutations by such an amount that eventual termination awaits every species. Information decay exists all around us, and no amount of counter-intuitive, hand-waving theory can change that fact.
Information theory (Gitt) and numerical simulation (Sanford et al) show that information within our own genome is eroding. We can witness the future of these mutations in inbreeding: recessive mutations that have built up in the genome are expressed. The inbred individuals have worse health problems, die sooner, and function less well. . Much like with cloning, this is a glimpse into the future of information decay and the future of our species.
In very simple terms, we are biologic machines. We have incredible processes to fix and repair damage, and to reproduce. But like every machine, we operate at less than 100% efficiency. This inefficiency is a measure of entropy, or the universal tendency of things to run down or degrade apart from intelligent intervention. While biologists have philosophically argued for decades that special qualities of natural
selection can reverse the biological effects of the second law of thermodynamics, making life effectively immortal, we have shown that this is not true.
If the genome must degenerate, and it must, then the assertion that natural selection + mutation=improvement must be wrong. Why is this? Because mutation/selection cannot prevent the loss of genetic information. Selection occurs on the level of the organism, not the molecular level. It's like trying to fix a computer with a hammer: the microscopic complexity of the computer makes the hammer mostly irrelevant. In the same way, the microscopic complexity of the genome makes selection on the level of the whole individual largely irrelevant. (Sanford)
(PN247)