
SINGLE STEP VS. CUMULATIVE SELECTION: A DEMONSTRATION

Michael E. Mills, Ph.D.
Psychology Department
Loyola Marymount University
Los Angeles, CA 90045


In his book "The Blind Watchmaker" Dawkins makes a very important
distinction between "single step selection" and "cumulative
selection."

In single step selection, no information is passed between
generations. Without surviving information, each generation must
start again entirely from scratch, without any "assistance"
provided by ancestors.

In cumulative selection, with every generation any changes, or
"mutations," that are adaptive are retained. Nonadaptive changes
are discarded. In biological evolution, this means that only
those genetic sequences that promote reproductive success are
"sieved" into the next generation. Each succeeding generation
starts with the result of the last sieving process. This
process, repeated again and again over millions of generations,
ultimately produces the incredible complexity of living
organisms. Complexity is virtually always achieved through a
long process of many generations of cumulative selection.

[press the space bar to see the next screen]



This is a reverse entropy process. Via cumulative selection,
information and complexity are increasing. This is in contrast
to the rest of the universe, which is slowly losing information,
or undergoing entropy.

By the way, this too is how science generally works to produce
knowledge. Both science and evolution: a) perform experiments,
and b) pass the results of those experiments on to succeeding
generations in an ever accumulating fashion. For example,
scientists perform experiments and pass the results of those
experiments on to the next generation of scientists to build on.
Theories that do not explain observed phenomena well are
discarded. "Mutant" theories are developed, a few of which will
be "adaptive". Both evolution and science are the ultimate
pragmatists: they only keep "what works" and pass that
information on as the starting point for the next generation.

[press the space bar to see the next screen]






Let's illustrate the difference between single step and
cumulative selection. Suppose you had a monkey sitting at a
typewriter, randomly pressing keys. How long do you think it
would take the monkey to type the sentence: "Methinks it is like
a weasel"?

By analogy, another way to think of this (in a biological
evolution sense) is that the letters of the above sentence are
the "genetic code" that will produce an organism that is
optimally adapted to a particular environment.

Let's say that a particular ecological environment is rather
stable. And let's say that the best adapted form for that
environment is a series of genes that have the follow letters:

 "Methinks it is like a weasel"

[press the space bar to see the next screen]





Directly analogous to the question, 'How long would it take a
monkey to type the above sentence?', is the question: 'How many
generations would it take to produce an organism with the above
genetic code, given that random mutations are retained when they
are adaptive?"


This program will illustrate why cumulative selection is vastly
superior to single step selection. Our goal: to produce the
sentence "Methinks it is like a weasel." We will begin with a
series of 28 random letters that we will use as a starting point
to "breed" successive generations, keeping adaptive mutuations,
until we reach our goal.

[press the space bar to see the next screen]
