A theme that has occupied me on and off ever since I resumed posting
to talk.origins in December 2010 is the probability of earth life
having arisen from abiogenesis on earth as opposed to having been
brought (or sent) here from another solar system by a spacefaring
species that lived ca. 4 billion years ago.
The latest revival is on the thread with "Intelligent Design Book
Meets Obstacle" in the subject line, with John Harshman the main
person debating with me on it.
Here, before we get too bogged down in minutiae, is a reply to John's
latest post on this fascinating topic. Another reply will be made on
the original thread--it's a long post, too long for a single reply.
On May 29, 9:39 am, John Harshman <
jharsh...@pacbell.net> wrote:
> pnyikos wrote:
> > On May 28, 8:03 pm, John Harshman <
jharsh...@pacbell.net> wrote:
> >> pnyikos wrote:
> >>> On May 28, 9:54 am, John Harshman <
jharsh...@pacbell.net> wrote:
> >>> Let's see: what would constitute an argument in your eyes?
> >>> 1. Nailing down actual figures, as in "I hypothesize that
> >>> (a) .00004 of all the planets where prokaryote-level organisms arise
> >>> result in an intelligent species and of those,
>
> > That's one in 25,000 seeded planets going on to intelligent species.
> >>> (b) half go on to a level of technology capable of interstellar
> >>> communication; of those,
>
> > So one in 50,000 going on to 21st century human level.
>
> >>> (c) half come to discover that they are alone in the observable part
> >>> of the galaxy and possibly of the universe; and of those
>
> >>> (d) one-fourth, on the average, undertake a panspermia project that
> >>> seeds at least a million planets with prokaryote or higher level
> >>> organisms."?
>
> > So one in 8 species coming to 21st century human level seed one
> > million planets apiece. So these fruitful species are each, on the
> > average, responsible for 20 first-generation 21st century human level
> > species. Dividing by 8 gives 2.5 intelligent species resulting from
> > directed panspermia for every one resulting from homegrown
> > abiogenesis.
>
> ....If there is any seeding at all. But by far the highest probability is
> that there is none. You have computed the probability that, if someone
> seeded the local area before us, we are not those people.
False.
> It seems to me
> that you have shown that the probability is higher that nobody seeded
> anything.
I'd like to see your reasoning, if any, for this counterintuitive
claim.
> >>> So done. There: I've given you a stationary target to ask questions
> >>> about and raise objections to.
> >> That's certainly close to an argument. But it's incomplete. What are we
> >> assuming about the initial life that "arises"?
>
> > "arises" wasn't the best choice of words, I see: I should have said,
> > "comes to be on a planet favorable for subsequent evolution". It
> > could happen one of three natural [as opposed to supernatural ways]:
> > homegrown abiogenesis, undirected panspermia, or placement by
> > intelligent creatures that arose elsewhere.
>
> > For the *initial* life form -- that one in a galaxy/universe
> > happenstance according to my hypothesis, it would be via homegrown
> > abiogenesis, although a lot of ingredients for it might have come from
> > elsewhere.
>
> > I'm not making any assumptions as to the biological makeup of any
> > intelligent creatures, including the hypothesized panspermists. But
> > for the sake of focused discussion, I've formulated four sub-
> > hypotheses as to their biochemical makeup:
>
> > 1. The "Xordaxian" hypothesis: very similar to ours, including a
> > genetic code close to identical to ours.
>
> > 2. The "Golian" hypothesis: also very similar, except their genetic
> > code involved fewer amino acids, perhaps as few as four.
>
> > 3. The "Throomian" hypothesis: ribozyme enzymes, proteins relatively
> > simple and mostly structural.
>
> > 4. The "para-3M" hypothesis: life as we do NOT know it, made up of
> > far simpler cells whose structure we cannot at this point even
> > imagine, but whose progenotic precursors were highly likely to exist
> > given pre-biotic conditions like those on the early earth.
>
> Each of these should produce a different probability calculation, so you
> have to specify.
Only in the sense that we are currently in the dark as to the
difficulty of producing intelligent life forms in categories 2, 3, and
4, and so those 1 in 25,000, etc. odds are less well grounded. Beyond
that, I don't see the relevance. Perhaps you can explain something
I'm missing out on here.
We are given that we are here, and that we evolved from prokaryotes.
As to whether those prokaryotes are due to minimal or no genetic
modification -- something neither you nor Meyer would count as
designing "a whole organism" -- as in the Xordaxian hypothesis, or
something both of you would call that, as in the para-3M hypothesis,
hardly seems relevant to the probabilities.
> But as far as I can see, only #1 makes any sense. Why
> should an intelligent species engineer life very unlike themselves?
A. Diversifying one's "sendings" so that if one is not well adapted to
the environment, another might be;
B. Diversifying the outcome. I did a long reply to Robert Camp, who
seems to have quit this "Intelligent Design" thread, touching on
this theme at length back in December 2010:
http://groups.google.com/group/talk.origins/msg/44a5c533d7f0c9f6
C. Speeding up the outcome, by starting with a more adaptable
organism. This is especially easier to see in comparing our
biochemistry with the Golian.
>They
> know that their sort of life worked, and achieved their goal of evolving
> into an intelligent species. Why mess with success?
Why would setting things in motion for another intelligent species
billions of years in the future be "messing with success"? You are
unusually hard to follow in this post, John.
[snip something to be dealt with on the original thread]
> >> Certainly that ought to
> >> change the probabilities somewhat. If we suppose that life to be similar
> >> to what we have on earth, that's one thing. If it's your RNA-based life,
> >> that's another thing.
>
> > I don't think the differences between various sub-hypotheses is a
> > profitable angle to pursue at this point. I do believe the "para-3M"
> > hypothesis entails that "prokaryote-level" be a multicellular rather
> > than a unicellular stage, otherwise the odds against it evolving to an
> > intelligent life form is considerably reduced.
>
> How can you possibly suppose that, given that you say you know nothing
> about them?
Their extreme simplicity, dictated by the fact that they are
supposedly very likely on a planet as favorable towards life as the
earth. For that to be true, they couldn't be much more complicated
than Ian Musgrave's HypUrCell:
http://www.talkorigins.org/faqs/abioprob/abioprob.html
> And how can you equate stages in completely alien forms?
I'm working on that (and much else relevant to this sub-thread) now in
a sci.bio.evolution thread, "Expanding the Drake Equation." You can
either wait until the thread seems spent, or join me there.
> > Of course, if someone [perhaps Mitchell, or Michael, or Mark] wants to
> > argue otherwise, I'm willing to listen.
>
> It makes a huge difference for you, since the probability of life
> supposedly increases as we go from 1 to 4,
Yes, but the probability of intelligent life is a huge unknown there,
as I've noted above. It stands to reason that there is a trade-off,
the comparatively easier abiogenesis leading to a steeper climb to
achieve intelligence. Golian life is obviously less diverse and
adaptable than Xordaxian, for example.
> and is apparently high for
> option 4. Now perhaps the probability of intelligence becomes lower as
> we go from 1 to 4, and all that balances out. But don't you have to
> think about it?
I've thought about it since before my December 2010 return, without
coming to any conclusions as to how the trade-off works, except for
what I've said just now.
Peter Nyikos
Professor, Dept. of Mathematics -- standard disclaimer--
University of South Carolina
http://www.math.sc.edu/~nyikos/
nyikos @
math.sc.edu