Andy wrote:

>With a random process, every outcome is, of course, equally likely. This i=
s why
>it could be used to explain "irreducible complexity".

With a random outcome, every outcome is indeed equally likely, but
that doe=
sn't imply that certain *specified* outcomes are as likely as
unspecified o=
nes.=A0=A0An ace-high royal straight flush in spades is as likely (and
as u=
nlikely) as any other combination of five cards.=A0=A0But if you
shuffle th=
e deck and deal a hand, it is very unlikely that it will be an
ace-high roy=
al straight flush in spades.

In order for evolution to get started, there must first be
self-replicating=
 systems.=A0=A0They cannot be *perfectly* self-replicating, because
that wo=
uld make it impossible for change, and therefore evolution, to
occur.=A0=A0=
But if the replication is very imperfect, there will not be enough
stabilit=
y to permit characteristics to spread through a population.=A0=A0This,
as I=
 say, must be in place *before* there can be any talk of
evolution.=A0=A0It=
 is incoherent to suppose that such self-replicating systems
themselves "ev=
olved."

So it is reasonable to ask what the *simplest imaginable* such system
would=
 look like, just as we might ask what the simplest imaginable computer
prog=
ram for editing text might look like.=A0=A0Remember EDLIN, the little
text =
editor that was bundled with MS-DOS?=A0=A0The executable program was
only a=
bout 4K in size, if I remember right.=A0=A0And I'm sure it was not the
simp=
lest possible text editor.=A0=A0You could probably get something to
work in=
 a program half that size, or less.=A0=A0But at some point you'd
approach t=
he simplest possible editor, such that anything simpler wouldn't
function a=
s an editor at all.

Then we can ask about the likelihood of producing such an editor by
concate=
nating random strings of machine language instructions.=A0=A0How many
"deal=
s" would we expect to have to wait?=A0=A0The problem of getting an
appropri=
ate self-replicating system from random concatenations of prebiotic
molecul=
es (amino acids, nucleic acides, etc.) is not solved by saying that
all con=
catenations are equally likely.=A0=A0Furthermore, we have some idea of
the =
window of time within which this had to have happened.=A0=A0Current
observa=
tions indicate that fairly complex (i.e., bacteria) living things were
alre=
ady present 3.8 billion years ago.=A0=A0The planet only cooled
sufficiently=
 to have bodies of liquid water on its surface about 4 billion years
ago.=
=A0=A0Prior to that it was too hot to sustain any "broth" for the
"primal s=
oup."=A0=A0So that gives us about 200 million years to get from
prebiotic s=
oup to bacteria.=A0=A0And bacteria are by no means the simplest
imaginable =
living things, so that first self-replicator presumably appeared
rather ear=
lier.=A0=A0We cannot say how much earlier.

This is the reasoning that led Fred Hoyle, and John von Neumann, and
others=
 (Lee Spetner's book Not By Chance is a good source) to reject the
random o=
rigin of life thesis as mathematically too remote to be
believable.=A0=A0Ir=
onically, even Francis Crick accepts this conclusion, for exactly the
same =
reasons.=A0=A0This is what drives him to the "directed panspermia"
theory o=
f life's origin, according to which life had to have been implanted
here by=
 extraterrestrials.

Todd Moody
[log in to unmask]