Because of natural selection,
information cannot be used to calculate the probability that a gene will evolve.
Information is useful because it has a precise mathematical definition not because it can
answer questions concerning whether or not the evolution of a new gene is possible. Chance
is not in control if natural selection is guiding which mutations survive. Therefore,
relating the amount of information in a gene to a probability that it can evolve is not a
valid mathematical analysis.
Molecular knowledge is the minimum amount of useful
information required for a gene to have any function. If a gene does not contain molecular
knowledge, then it has no function, it confers no selective advantage, and it is not a
gene. Thus, before a region of DNA contains the requisite molecular knowledge, natural
selection plays no role in guiding its evolution. Chance controls which mutations survive.
Thus, molecular knowledge can be related to a probability of evolution. Figure 1 helps
illustrate these important concepts.
Notice that the first step that creates the required
molecular knowledge is vertical, and the subsequent step that creates molecular
information is sloped. This difference is important in that it is meant to show that
natural selection can help guide the second transition, but it plays no role in the first.
Thus, it is the size of the first step that determines whether or not a gene can evolve.
A simple example will now be introduced to help
clarify this concept. Consider the following sentences:
My 13 year old black lab, Bubba, likes to fetch a
tennis ball.
mi 13 yr od blk lab, buba, like fetch tenis bal.
kkde nng nnfkr skkzzzzzzzzzd.
Figure 1: Information in Life
Each sentence represents a gene. The first sentence uses the most
letters. The second sentence uses fewer letters, communicates the same points and is not
grammatically correct. The information in the first sentence is greater than the second
(it has more letters), but the useful information or knowledge is the same (the same
points are communicated). The last sentence is nonsense. It actually contains information,
but this information is not useful. If a person is asked to perform the job of natural
selection by optimizing each of the above sentences, then it is easy to see that they
might leave the first sentence alone and change the second sentence to read more like the
first. They will not change the last sentence into anything remotely similar to the first.
In fact, they will not be able to optimize this sentence because they do not know what it
should say. If they do happen to change it into a sentence similar to the first, then they
do so by chance. By analogy, natural selection cannot guide the evolution of genes that
contain no useful information.
If these sentences are composed by randomly selecting
letters from the alphabet, then the probability of spelling the second sentence is much
better than the first. So only this sentence and all sentences similar to it are useful
for assigning a probability to the evolution of this particular sentence. In this example,
the first sentence represents molecular information and the second one represents
molecular knowledge. If some concepts are not required, the second sentence may be
simplified further. For example, the sentence fragment, mi black lab, still
communicates some knowledge.
At least one biologist, Richard Dawkins, has suggested that
sentences like: "kkde fwegzzzzzzzzzd." still confer a selective advantage;
therefore, given time these sentences will evolve into something useful under the guidance
of natural selection.1,14 Dawkins ran many computer simulations with sentences
like the one above, and they all evolved into the desired result. But his programming and
logic are both flawed because natural selection cannot preserve or optimize a gene that
offers no selective advantage (the nonsense sentence above represents a gene that confers
no selective advantage). A gene must contain some useful information before natural
selection activates. Thus, chance and chance alone must create the initial knowledge.
Because the probability that chance can accomplish this goal is proportional to the height
of the first step in figure 1, the size of this step completely determines whether or not
a new gene will evolve. It is the goal of this book to characterize the size of this
initial step. If it is small then naturalistic laws explain the evolution of knowledge. On
the other hand, as the step size increases, the probability that chance will create the
required molecular knowledge approaches zero, and at some critical threshold, the design
inference becomes valid.
Large steps are associated with the evolution of genes that
are completely different from all other existing genes. For these genes, the probability
of chance finding an appropriate solution (even given 50,000 billion years) is very close
to zero. The origin of these genes imply design.
Next: Chemical
Evolution
Previous: Is Evolution Possible?
|
