Feedback for August 2004

Regarding your question about the origin of life, the best single reference for general readers I have is Iris Fry, "The Emergence of Life on Earth: A Historical and Scientific Overview" 2000:Rutgers University Press. Although this is only 4 years old, it is already due for revision to reflect significant recent results.

I have not read "A Case for Faith," but I am well familiar with the creationist's criticisms of the Miller/Urey experiment. A complete analysis exceeds the role of the feedback page. In short form, I would frame a responce as follows;

1) The Miller/Urey experiment was not, and is not the only basis for abiogenesis research. Thus disputing their result does not invalidate abiogenesis theories.

2) The geochemical research on the early oxidation state of the Earth's oceans and atmoshphere indicates that it was genereally reducing, but not to the same extent as the gasses in the original Miller/Urey experiment. However, there is additional evidence of Early Archaen highly reduced "oasis" and even for oxic "oasis." Consider that even today there are highly reduced locations on Earth such as methane seeps, hydrogen sulfide seeps, and hydrothermal vents.

3) The same experiment has been run with positive results (the generation of amino acids and other complex "organic" molecules) with different gas mixtures, and different energy sources.

4) The analysis of amino acids from meteorite revealed that they were in the same general proportion as the results of the Miller/Urey experiment. This makes the continued relevance of the Miller/Urey result self-evident.

"1.5x10¹¹ grams/year" is the same as "1.5x10⁸ kilograms/year." Tim does switch units, but uses the same value throughout.

The sampling in the study is convincing that only one line of mtDNA remains in the human population for exactly the reason that your friend mentions: the odds that another line exists but remains unsampled is very small. The conclusion is that any other lines have ceased to exist in the human population; the age estimate for "mitochondrial Eve" attempts to work out just how long ago the existing mtDNA line originated.

The following Perl code performs a Monte Carlo simulation on a population with mtDNA lines denoted by letters (A,B,..,x). An initial population is formed of equal numbers of each. In each subsequent generation, each "individual" contributes some number of like "daughters" or fails to "reproduce". (The maternal descent of mtDNA helps simplify things here, since sons can be ignored.) Each run continues until only one line has surviving members in the population. A report is made at the end giving an average number of generations until each line became "fixed". This is meant as a simple demonstration of fixation of a trait, so the number of offspring rule used is simple. The user can change the number of original mtDNA lines and determining number of offspring per individual.

Example output:

All A lineage 26 times, average number of generations: 128.2
All B lineage 31 times, average number of generations: 115.7
All C lineage 15 times, average number of generations: 80.2
All D lineage 28 times, average number of generations: 98.5

This was for an initial population size of 100 and 100 runs made.

Skip the Perl code.

#!/usr/bin/perl
# mtEVE goes to Monte Carlo
# Simulation by Wesley R. Elsberry, 2004
# Customization section ---------------------------------
$nruns = 100; # How many times to run the Monte Carlo
# simulation
# Lines: the original lines present in the population
@lines = ("A","B","C","D");
# Number of individuals per line in the initial
# population
$initperline = 25;
# Offspring distribution. For a more or less stable population
# size, the sum of the entries should equal the number
# of entries.
#             1 2 3 4 5 6 7 8 9 _ 1 2 3 4 5 6 7 8 9 _
@offspring = (0,0,0,0,0,0,0,0,0,0,1,1,1,1,2,2,2,3,3,4);
# Initialization section --------------------------------
srand; # Seed random number generator
# Run Monte Carlo ---------------------------------------
undef(%goodruns);
for ($jj = 1; $jj<=$nruns; $jj++) {
    # Initial population: individuals equally divided
    # into two lines
    undef(@pop1);
    for ($ii = 1; $ii <= $initperline; $ii++) {
        foreach (@lines) {
            push(@pop1,$_);
        }
    }
    $ii = 0;
    $done = 0;
    do {
        undef(@pop2);
        undef(%popcnt);
        foreach (@pop1) {
            $linetype = $_;
            $os = int(rand($#offspring+1));
            for ($kk = 1; $kk <= $offspring[$os]; $kk++) {
                # Add an offspring to the new population
                push(@pop2,$linetype);
                $popcnt{$linetype}++;
            }
        }
        $ii++;
        @pop1 = @pop2;
        foreach (@lines) {
            if ($#pop2 <= ($popcnt{$_}-1)) {
                $done = 1;
                $winner = $_;
            }
        }
        print sprintf("\b\b\b\b\b\b%06d\n",$#pop2);
    } until ($done);
    if ("" eq $pop2[0]) {
        print "The population crashed!\n";
    } else {
        print "$winner became only lineage in $ii generations\n";
        $popgen{$winner} += $ii;
        $wincnt{$winner}++;
        $goodruns{$winner}++;
    }
}
# Report on overall results.
foreach (@lines) {
    $ave = $popgen{$_} / $goodruns{$_};
    print "All $_ lineage $wincnt{$_} times, average number of generations: $ave\n";
}

Skip the Javascript code.

function drift(){
nruns = simulation.nruns.value; // get value from form with id="simulation"
                            // with input tag with id="nruns"
halfpopsize = simulation.halfpopsize.value; // likewise
simulation.output.value = ""; // clear the output window
goodrunsA = 0;  // track the number of runs which A gets fixed
goodrunsB = 0;  // track the number of runs which B gets fixed
popgenA = 0; // counts number of generations it takes to fix A 
             // for the average at end of function.
popgenB = 0;
for ( jj=1; jj<=nruns; jj++ )
{ pop1NumberA = halfpopsize;
  pop1NumberB = halfpopsize;
  ii = 0;  // counts generations
  
  do
  { pop2NumberA = 0; // for the next generation
    pop2NumberB = 0;
  
    for ( i=0; i<pop1NumberA; i++ ) // scroll through each A
    { if ( 0.5 > Math.random() )
      {} // No offspring for this individual A
      else
      { pop2NumberA = pop2NumberA + 2;  // Two offspring for this A
      }
    } // end for
        
    for ( i=0; i<pop1NumberB; i++ ) // This time scroll through each B
    { if ( 0.5 > Math.random() )
      {} // No offspring for this individual B
      else
      { pop2NumberB = pop2NumberB + 2;  // Two offspring for this B
      }          
    } // end for
        
    ii++;
    pop1NumberA = pop2NumberA;
    pop1NumberB = pop2NumberB;
        
  } while ( 0!=pop1NumberA && 0!=pop1NumberB );
      
  if ( 0==pop1NumberA && 0==pop1NumberB )
    simulation.output.value += "The population crashed!\n"; 
  else if ( 0==pop1NumberA )
  { simulation.output.value += "B became only lineage in "
                      +  ii + " generations\n";
    popgenB = popgenB + ii;
    goodrunsB++;  
  }
  else
  { simulation.output.value += "A became only lineage in "
                      +  ii + " generations\n";
    popgenA = popgenA + ii;
    goodrunsA++;  
  }        
} // end for (jj)
aave = popgenA / goodrunsA;
bave = popgenB / goodrunsB;
simulation.output.value += "Ave. generations to all A lineage: "
                  + aave + ".\n";
simulation.output.value += "Ave. generations to all B lineage: "
                  + bave + ".\n";
                  
}

1. Ted Steele has argued that retroviruses (viruses that can insert their genetic code into the DNA of a cell) play a role by transferring acquired immunity from a body cell to a sex cell via retroviral insertion.
   
   
2. Around 40% or so of a mammalian genome is typically of retroviral origin, but nobody has yet shown that any of it is expressed. This may be because it is unlikely to be of use to the host, or it may be that natural selection immediately reshapes it so it is. Either way we would not see evidence. Or it may be we simply haven't yet come across the evidence that is there.

Jeff asks us "where are the facts supporting what we are denouncing?" That strikes at the heart of the problem. What facts support creationism? If Jeff thinks he has these facts, then he is encouraged to post them to talk.origins. Though he is strongly encouraged to first search this web site to see if it has already shown what is wrong with the alleged "facts." Hundreds of claims of the evolution deniers and what is wrong with them can be found at the Index to Creationist Claims.

I edited Jeff's feedback to remove a copy and paste of The Fool Says in His Heart... by Rev. Dale Tedder. Feedback is not a place to violate copyright law. I am always amazed at how many people think that plagiarism is okay.

Notwithstanding all that, the June PotM is relevant to the creation/evolution controvery. It talks about rational religion vs. fantatical religion, which is very nearly the core of the controversy. Yes, it also talks about politics, but politics has been part of creationism since the Scopes trial at least, and under a president who wants to make scientific findings subject to political decree, it is now an inescapable part of science, too.

It's very easy to make vague, unsupported assertions about "fallacies." Because it is so easy, empty claims are not really noteworthy. It would be far more interesting if you could go through this archive's Age of the Earth FAQ and document several fallacies in it.

Second mistake: Barry Setterfield rather obviously failed in his attempt to demonstrate any time variability of the speed of light from the historical record, as have all other creationists who have tried. See The Decay of c-decay, in this archive.

Third mistake: The idea that "nothing" suddenly became the big bang has never been part of "big bang" cosmology. Creationists always like to parade that claim, but it is disingenuous at best, since no scientist ever actually believed such nonsense. The correct interpretation of the singular beginning of the universe in general relativity (the "bang"), is that general relativity is an inadequate theory to describe that early era of the universe. It has long been known that quantum physics must be taken into account here. In one promising line of investigation, amongst many, string theory opens the door to pre big-bang cosmology, by eliminating the singularity that is a necessary ingredient of a classical theory, but not a quantum theory. Another string interpretation allows for the colliding brane scenario in a cyclic universe cosmology. Cosmologists never thought that "something came from nothing", so the criticism is pointless.

Fourth mistake: No, the law of entropy does not dictate that anything ordered moves to chaos. No law of physics does, except as they are commonly misused by creationists. The law of entropy (correctly referred to as the second law of thermodynamics) does say that systems in thermal contact will come to thermal equilibrium. But it is true only for statistically large collections of objects. It is known from both theory & practice that this law is not necessarily applicable to situations dominated by quantum physics. See, for instance, Nanotubes may have no 'temperature' (August, 2004), Quantum Heat Engines, the Second Law and Maxwell's Daemon (February 2004), Statistical thermodynamics of quantum Brownian motion: Birth of perpetuum mobile of the second kind (November, 2001), or the Thermodynamics, Evolution and Creationism FAQs, in this archive.

If one is going to make "factual" claims, one should first be able to distinguish fact from either fiction or just plain fantasy. This is an exercise at which creationists are notoriously weak. There is a distinct and important difference between trusting a supernatural God, and trusting non-supernatural, human creationists. One of the two is rather more reliable than the other, and it is left as an exercise for the reader to figure out which.

If it [freedom] operates according to biochemical laws, it cannot actually be free, being wholly determined by the laws which govern its operation.

However, laws can never wholly determine anything. The Heisenberg Uncertainty Principle points out that the very act of determining something (such as, say, the position of a particle) mucks up other things (such as the particle's velocity) so that they become more uncertain. Thus absolute determinism is impossible in a purely physical system. One may argue that things are still determined at some metaphysical level, but that would be beside the point of your argument. Ironically, one must add supernatural entities, able to "know" things without physically looking at them, to achieve determinism.

The rate at which Earth spins on its axis is expressed as a length of day (LOD), which is 23^h56^m04.^s0989 with respect to the fixed stars (the sidereal day), or 24^h00^m00.^s0084 with respect to the sun (the solar day). The rate of change of that spin is expressed as a variability in the LOD. The LOD is in fact quite variable, on a daily basis due to friction at the core-mantle boundary, or variations in atmospheric winds, and on annual cycles related to seasonal changes in atmosphere & ocean circulation (International Earth Rotation and Reference Systems Service (IERS), IERS Rapid Service/Prediction Center for Earth Orientation Parameters, Historical Development of Earth Rotation Knowledge). There is also a long term trend of slowing imposed by the tidal interaction between Earth and the moon (as described in my own FAQ entry linked above). It is this last trend that is at issue here.

The short-term accelerations mentioned above are highly variable, albeit with relatively small amplitude. The long term trend imposed by tides is almost constant over the last 3 billion years or so, at least when averaged over fairly long periods, but increasingly variable before that. Paleontological evidence indicates that the length of day about 620,000,000 years ago (precambrian) was about 21.9 hours ( Precambrian Length of Day and the Validity of Tidal Rhythmite paleotidal Values, G.E. Williams, Geophysical Research Letters 24(4): 421-424, February 15, 1997). The rate of change given in the FAQ (0.005 sec/year/year) translates into 0.0014 sec/day/century, insignificantly different from the standard quoted value of 0.0015 sec/day/century (Earth's Variable Rotation - Geophysical Causes and Consequences, Kurt Lambeck, Cambridge University Press, 1980). But (0.0015 sec/day/century) x (6,200,000 centuries) = 9,300 sec/day, the accumulated time that the day has lengthened since then, assuming a constant rate. That's about 2.58 hours. Subtract that from 24, and you get 21.42 hours, which is not all that different from 21.9 hours, considering the uncertainties involved.

The rate of change of length of day, and the recession rate of the moon, are most rapid during the first few hundred million years of the Earth-moon system (i.e., Evolution of the Earth-moon system, Touma & Wisdom, Astronomical Journal 108(5): 1943-1961, November 1994), but slows rather quickly after that. The rate used above (0.0015 sec/day/century) is probably unusually high, due to the favorable size of Earth's ocean basins, which encourage resonance with the moon (i.e., A stochastic model of the Earth-moon tidal evolution accounting for cyclic variations of resonant properties of the ocean: An asymptotic solution, Kagan & Maslova, Earth Moon and Planets, 66: 173-188, 1994).

The standard creationist argument addressed by Dave Matson's FAQ (How Good Are Those Young-Earth Arguments?) assumes that the leap second added to the atomic time scale, to keep it in synch with the civil time scale, is a measure of the rapidly slowing Earth (on the order of 1 sec/year/year or faster). Of course, if Earth slowed that rapidly, then the day 620,000,000 years ago would be 1,722,000 seconds shorter, a hard thing to pull off when the day only has 86,400 seconds to lose. If the leap second really did represent the spin down rate of Earth, the creationists would have a point. But it is no such thing, it is only a symptom of the fact that the two time scales, atomic & civil, tick at slightly different rates creating an accumulating difference between them. The leap second is an arbitrary device designed to keep that accumulated difference from exceeding 1 second (see Leap Seconds from the U.S. Naval Observatory's Time Service Department).

[Webmaster's note: A reader noticed that sidereal and solar days had been inadvertently reversed. This has now been fixed.]

Ages of Stars

The age of a star is determined by comparing theoretical stellar evolution models with the observed characteristics of a star, as shown on the Hertzsprung-Russell diagram (commonly known simply as the HR diagram, or a color-magnitude diagram).

The stellar evolution computations work by starting with a computer model of a young star or a protostar, and then allowing the clock to tick, and the star to age, in accordance with the known & applicable laws of physics, of which there are quite a few in this case. The brightness and color of the modeled stars are tabulated as a function of time, and then compared to the observed, non-random pattern of stellar properties from the HR diagram (there is an example of this on one of my webpages: The Hertszprung Russell Diagram and Stellar Evolution).

Of course, this means that the confidence we place in the derived ages is based on the confidence we have in the correctness of the details of physics that go into the models, which are quite complicated, and the fidelity of the match between derived & observed stellar properties. The match in fact is quite extraordinary, in that the synthetic HR diagrams, built from the evolution models, reproduce even small details, typically gaps or clumps, in the observed distributions of stellar properties. This is extrordinary because the models simply let physics work over time, with total ignorance of any expectations. There is no particular reason why they should reproduce the real patterns on observed HR diagrams, unless those patterns are really representative of the change with time of the observed properties of stars.

Are the stars still shining?

All of the stars we see with our eyeballs are within a few thousand light years of Earth, where a light year is the distance over which light travels in one year, about 6,000,000,000,000 miles. Since stars usually (but not always) change vewry slowly with time, the few thousand years represented by that distance don't amount to a hill-of-beans in the lifetime of the star. So we can be confident that the stars are still there.

But not always. For instance, supernova 1987a went off in the Large Magellanic Cloud, some 169,000 light years away. But the explosion was observed in 1987, meaning that we saw that star for about 169,000 years, during which it was really not there. So it can happen, but we can tell from the type of star, whether or not there is a chance that it's not there any more. And keep in mind that stars don't just fizzle out. They go out with a bang if they are big enough, or with a planetary nebula if they are not (that's what our sun will eventually do).

And I agree, Terry Pratchett was one of the most significant things I learned from talk.origins.