On Thu, May 8, 2014 at 2:43 AM, Tim <tim@little-possums.net> wrote:
On Thu, May 08, 2014 at 12:05:24AM -0400, Richard Aiken wrote:
> Those favoring irreducible complexity have a point that a chemical
> process with ~23 steps (none of which do anything alone or in any other
> combination of each other)

They can stop right there.  Those steps *do* do something without the
full cycle.  I don't know when your text was written, so maybe it's
just obsolete rather than wilfully blind.


IIRC, the point was that parts of the cycle do not do anything useful to *that* organism. That is, the organism doesn't gain any benefit from any of the intermediate steps, only from the end product of the full process.

To use a car analogy, the transmission transfers power from the engine to the wheels. The transmission alone - even though if you turn it by hand it will still work - doesn't help you do anything.
 
> Those opposing irreducible complexity say that the individual
> components of such a process could be spread through a population as
> part of other functional processes but don't give any example of
> plausible previous processes.

Again, that could just be obsolescence.  Such examples are now known.

The counter arguments I've read so far don't address the specific example of the Krebbs Cycle, even though it was Michael Behe's (author of "Darwin's Black Box") main argument. They all just say that millions upon millions of chemical processes through millions upon millions of generations yielded photosynthesis. It exists, so it must have evolved.
 
> That lack strikes me as disingenous, since a *functional* process
> using specific components that could later form a different process
> contradicts someone's earlier point that chemical processes are
> incredibly delicately balanced - if any variable differs by the
> slightest degree, the process does not work the same or at all.

The closest I've seen anyone say that was myself, and it was in
relation to the processes of a *single* organism changing due to wide
temperature variations.  I said nothing about immutability of
processes across evolutionary timescales, nor about impossibility of
multiple processes operating in different temperature regimes.  In
fact I gave examples where processes had changed, and also examples of
organisms that do use quite different process in different conditions.

Noted. But - as you say - each such process only works in A particular way to produce A particular result under each set of circumstances. Any other arrangement of those particular chemical components won't work for that particular result under those particular circumstances. Other cycles aren't that cycle. You can't get incremental improvement of such cycles. It's all or nothing.

There may be another process which can give a photosynthesizing organism the same result as the Krebbs Cycle. But that process won't be a precursor of the Krebbs Cycle. It won't be a step in the cycle's evolution. It will be a different process.
 
> But the opponents also say that even if these future components
> didn't spread as part of a different process, they could still get
> carried forward as non-functional junk - which is true but depends
> on the future-valuable junk being linked to Something Else that is
> selected.

No, the latter condition does not hold.  All it requires is that the
pieces not be heavily selected against.

But they won't spread nearly as widely through the population, in that case. So you only get some smaller fraction of those millions upon millions upon millions of possibly mutating interactions.
 
> If not, it's prevalence would drop as per the quote above.

The quote above was giving examples of how a trait can persist for
quite some time even if it is selected against.  Traits that are
neither selected for nor against can persist for arbitrarily long time
spans.  They can be retained for longer than any one species that
carries them, via multiple mechanisms.

Noted. But see above. Not being selected for drops the chances of random assembly proportionately. 

 
In the case of chemical mechansism evolving in bacteria, there would
likely be more like 10^31 organisms at any one time (millions of
million million million millions).  Then multiply that by the number
of generations, so tack another "million million" or so onto that.
Also note that horizontal gene transfer is more viable in bacteria
than in the much more complex organisms that came later.

And you only get that full population for a trait that's been heavily selected. Then we need at least two different traits (probably more since there are 20+ steps in the Krebbs Cycle). We need all traits to be in the same bacteria. If either is part of a selected-for beneficial process, we need the genes controlling its reproduction to malfunction in a specific way. Then the various component parts of the complex cycle need to recombine in a specific way.

Sorry. The improbability still boggles me. But that's just me, I guess.

--
Richard Aiken

"Never insult anyone by accident."  Robert A. Heinlein
"A word to the wise ain't necessary -- it's the stupid ones that need the advice." - Bill Cosby
"We know a little about a lot of things; just enough to make us dangerous." Dean Winchester