Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10 UW 5/3/83; site uw-june
Path: utzoo!linus!decvax!tektronix!uw-beaver!uw-june!gordon
From: gordon@uw-june (Gordon Davisson)
Newsgroups: net.origins
Subject: Re: codes,designs,creation,intelligence
Message-ID: <43@uw-june>
Date: Fri, 12-Jul-85 06:17:22 EDT
Article-I.D.: uw-june.43
Posted: Fri Jul 12 06:17:22 1985
Date-Received: Tue, 16-Jul-85 20:55:40 EDT
References: <32500041@uiucdcsb>
Organization: U of Washington Computer Science
Lines: 241

>     Merlyn Leroy, responding to a note [A. Ray Miller] had written on
>informational thermodynamics, writes:
>>     There is the Miller experiment, which attempted to recreated the
>> young Earth environment, ran an energy source through it (a spark gap), and
>> ended up with amino acids, some fairly long, in only two weeks.

>[A. Ray Miller]
>     There are many problems with Miller's experiments; I'll mention just a
>few briefly.
> [... (argument I don't know much about skipped)]
>                                                     Second, the destruction
>rate of the compounds is far higher than the production rate.  When you trap
>out the products to get around this problem, you also remove the products from
>their energy source, and further progress becomes impossible.  It's a catch-22
>situation.

I'm not sure I believe this.  If the amino acids are destroyed so soon after
creation(?!), how did they manage to form long chains?  (I assume that's what
Merlyn meant by "some fairly long".  Or is that right?)  Also, I thought
others had since done similar experiments without the trap, with similar
results.

>            Third, he generated no code capable of carrying information.

If he'd have done that, he'd pretty much have created (that word again!) life
in the lab, right?  What do you expect in just 2 weeks in such a small
container?  Given most of the surface of the earth, and a billion or so
years, the appearance of complex, self-reproducing molecules wouldn't be too
suprising.  (No!  I did not say that just because there was so many
opportunities, it must have happened.  I said it's much (*much*) more likely
to happen under those circumstances than in Miller's experiment)  And once
you have self-reproducing molecules complex enough to undergo nonfatal
mutations, it's almost certain to lead to something we'd recognize as life.

>Ignoring for a moment the problem of the D-form amino acids, he has (roughly) a
>random-letter generator (using a chemical alphabet).  What does he produce?  A
>sequence of words such as: kjemmp lma wwqnx z pr gmbv ytc d qhiojfs xa u bqop.
>This does not carry information.  Even if you generate short "words" such as
>"a", "i", or even "the" mixed in with the above, it has no information content
>since it is a meaningless sequence of characters generated randomly.  You will
>never be able to generate randomly a meaningful code in the lifetime of the
>universe, even one as short as "the theory of evolution".

Ah, but with self-reproducing molecules (yes, the first one of those has to
occur by chance) there's an editor (natural selection) that throws out the
bad stuff and keeps the good, and this is where the information comes from.

Consider as an analogy a computer program that learns to play chess by
playing against itself: It starts out with a not-very-good strategy (like
picking a move at random), and randomly produces new strategies, then tests
them by playing them off against versions of itself that don't use the new
strategy, and if the new strategy does not make the program play better,
trashing it and trying again.  Notice that although most of the strategies
will not work very well (or even make any sense), the program will gradually
become better and better, to an arbitrary level of expertise.  It'll just take
a long time.  But surely all of the strategies were generated randomly, so
where does the program's apparent expertise come from?  Does the just-
described scenario violate information theory in some way?

>                            [description of SETI ...], looking for "evidence
>of design".  This has several implications for net.origins.
>     First, the SETI group must feel that time, chance, and natural processes
>are not sufficient to produce a code capable of carrying information.

What do you mean, 'capable of carrying information'?  Just about everything
carries information, it's just that not all of the information is meaningful.
(Now try to define meaningful :-)  I am reminded of I-forget-who publishing
several groups of numbers.  Some of them turned out to carry meaningful
information (his birthdate, ARPAnet node address, etc.) and some of which
were just random digits.

Assuming that you mean 'codes that *do* carry meaningful information', they
do think they can occur by chance.  Also, that the most likely way for them
to occur is for life to 'occur', then send them.

>                                                                       In this
>case, the code is electromagnetic.  Anyone currently reading this note is
>looking at a 26 letter code and no one, I'm sure, thinks it was produced by a
>random-letter generator, or a bug, or any other form of time, chance, and
>natural processes.

I think I'll confuse things a little here by bringing up AI again.  In the
Computer Recreations column of a recent Scientific American (titled something
like 'the art of turning file literature into gibberish'), there was a
description of an AI program that looked at a bunch of text, figured out the
frequencies of different sequences of letters, then produced random text with
roughly the same distribution.  Using short sequences of letters, the output
looked like gibberish, but it wasn't completely meaningless.  For instance,
it was reasonably easy to tell French gibberish from English and German
gibberish, and many of the words really existed.  Even though they hadn't
occured in the input!  Using longer sequences, words came out correct most or
even all of the time, and it was possible to tell who the block of input text
was written by from the writing style.  I suspect that Mark V. Shaney (mvs@
alice.UUCP) is an AI using this type of algorithm with recent net.singles
postings as input.  Here's an excerpt from one of his postings:

	For a year, I lived with a girl, on a consulting basis.  Well we sleep
	together at home, why should we come if you had gone anywhere on St
	Denis Street and asked a woman whether she was going to consider you a
	whore or a telephone.  When you stop thinking of sex as something to be
	near him and have been fooling around for hundreds of years, at least,
	and the special time requirements it entails.  If you don't shove it
	under their roof, you will be able to share more of your stomach, your
	SO to your personal life and decide which is more trouble than it is
	quite likely that you are looking for.

It should be obvious that this carries some meaningful information, since
it indicated to me how this had been produced.  But it's the result of random
character generation!  (Biased random character generation, but random
nonetheless)  Consider also that the information it carries is about its own
(random!) origin.

>                    When we look at the DNA of *any* life form, it is also a
>code (of a 4 chemical alphabet) which is far more advanced than any babble I'm
>likely to produce.  The media on which the code is carried is unimportant.
>Why then do we say DNA was produced by time, chance, and natural processes?

How about because it also contains information about its own (random!) origin.
'The Evolution of Darwinism' (Scientific American, July '85; strongly
recomended reading for anyone involved in this debate) describes a type of
mutation called tandem multiplication, and gives sections of the collagen
gene in chickens and the immunoglobulin genes in mice as examples of genes
that show the mark of this type of mutation.

>    Second, SETI claims they can recognize a designed object, i.e., one which
>requires intelligence (the I in SETI).  Note that this is not due to any
>inherent properties in the object itself.  The designed object will be some
>pattern of electromagnetic frequency in a sea of random electromagnetic fre-
>quencies.  It must be, therefore, be due solely to the nature of the pattern
>itself, i.e., a code carrying some information.

Not just any information.  Information is meaningless unless you know how to
interpret it.  The SETI people are basically looking for something that might
be a representation of the message 'We are here!'

>                                                 Yet not a week goes by on this
>net that we don't hear evolutionists tell us they can't recognize evidence of
>design and intelligence.

Not quite.  We tell you that it can't be measured objectively.  We tell you
that the design you see in the natural world could be something else, too.
Karl Dahlke mentioned that when pulsars were first discovered, people thought
they were signs of alien intelligence.  So the evolutionists (and the SETI
people) can see design where it isn't, too.

>                          They tell us this, of course, only when it's
>convenient, in other words, when they're talking to creationists.  When
>they're working on SETI, or looking for arrowheads made out of rocks just like
>all the other rocks lying on the ground, or noticing the difference between a
>sandcastle on a beach and the patterns waves make on that same beach, then -
>well, even a child can recognize that which took creative thought and that
>which natural processes can produce.

We recognize sandcastles and arrowheads as manmade at least partly from
experience.  Rock briges, salt pillars, and snowflakes, we recognize as
natural from experience, even though they may look designed at first.  In
other words, we see design in things we know are artificial, and not in
things we know aren't.  Since you believe everything was created, it's only
to be expected that you'll see design everywhere.

Let me bring up another example of something the SETI people are confused
about -- red giants.  Freeman(?) Dyson suggested that a growing civilization's
power needs would grow to the point at which it was necessary to trap all
of its sun's energy, presumably in some sort of sphere (that's where the term
'Dyson sphere' comes from).  Then someone pointed out that Dyson spheres might
look quite a bit like red giant stars, of which there are quite a few around.
So, are the red giants natural or artificial?  Theories exist explaining their
origin exist both ways exist, so they can't really be counted as evidence one
way or the other.

>     (A footnote here.  Symmetry, such as that formed in a crystal like ice,
>provides no help for the evolutionists, despite comments by some on this net.
>One of the guys CNN talked to from SETI mentioned they got a symmetric pattern
>once - they had discovered a pulsar.  It contained no information, however, and
>although an important discovery, provided no hope of ever producing life.)

Wrong.  Pulsars were discarded as evidence for life because a perfectly good
alternate explanation was found.  If nobody had thought of any reasonable
alternate theory, people *would* be justified in counting them as evidence
of extraterrestrial life.

Note that an alternate theory (evolution) about the order in the natural world
has been put forward, and, despite all creationist screaming to the contrary,
shown to fit the facts very well.  So you don't get to count 'design' as
evidence for creation.

>     Finally, it is theoretically possible to translate the DNA patterns of
>E. coli into an electromagnetic pattern (DNA, of course, being based on a
>simple four character alphabet).  This is a simple mapping function, e.g.,
>these very words have been mapped several times into analog and digital elec-
>tronic values from when my fingers typed on 26 keys.  All are equivalent, of
>course.  If SETI were to pick up such a transmission of E. coli DNA patterns,
>it would be trivial to recognize,

Do you mean recognize as E. coli DNA patterns, or recognize as patterns?  If
you mean the latter, I agree.  They'd probably (I haven't seen them myself,
so I'm guessing here) be noticed immediately as not the sort of thing that
occurs as the result of any known astronomical process.  (as to why E. coli
DNA patterns aren't recognized as patterns when we find them in E. coli, they
are.  We recognize them as the sort of thing that occurs as the result of
known evolutionary processes).  If you mean the former, I disagree.
Someone'd have to notice that the patterns are the same sort of patterns that
show up in DNA, then check all known organisms' DNA against it until one
matched.

>                                  and no doubt the High Priest of Evolution,
>Carl Sagan,

He's an astronomer, not a biologist.  You creationists really do lump all
science together, don't you?

>            would say: "Aha!  We have evidence of an intelligent designer,
>which we have not seen directly, but must exist."

I guess you meant the former.  Oh, well.  I think your description of Carl's
reaction is wrong, too.  I think he'd say "It's too much of a coincidence;
there must be something we don't understand going on.  *This bears further
investigation*."

>                                                   So when that same Carl
>Sagan sees E. coli here on earth, along with vastly more advanced forms of
>life expressing codes we haven't even begun to decipher, must less design
>ourselves (simply expressed in a chemical rather than electronic alphabet)
>what does he say?  "Evolution is a fact - like apples falling off trees."

Like I said above, E. coli DNA in E. coli makes sense as the result of known
processes.  E. coli DNA transmitted from the crab nebula doesn't.  This is
an important difference.

>     For my part, I'll stick with a Creator and information theory, rather
>than with Sagan and wishful thinking.

I'll stick with science and evolutionary theory, rather than religious nuts
and wishful thinking.

>P.S.  I'll put the shoe on the other foot now:  Would some evolutionist on the
>net care to give us a definition of design which would allow SETI to recognize
>created patterns but would differentiate against those patterns found here on
>earth evolutionists claim to be produced only by natural processes?  Well?????

No thanks, it can't be done.  All we've got are heuristics.