<Thus we should not rule out the possibility that many of the older,
more disrupted lunar craters could have formed supernaturally and are
not the products of natural causes. (Or, pace Moorad: We never
saw the impact which produced Bessel crater, so why couldn't it have been
made by elves? Science can't apply to any crater formed before 10000 BC,
because there weren't modern humans or instruments to record it. Let it
forever be an enigma!)>
.end speculations here...>>
>I don't know why you threw in the concept of the supernatural in your
>moon example, which I suppose was meant to address my statement. Did
>I say anything about the supernatural? No. So why do you?
That was more of a dig at Moorad than you. However, I do not see why
the idea of descent with modification becomes more problematic the further
one goes back in the history of life. Perhaps if one goes very far back,
say around the eubacterial, archaebacterial, eukaryotic split, that
question may apply, but not for most of what has happened since then.
Tim: <Genetic variation is the major change agent. What locks a particular
change into a population or affects its frequency in a population includes:
natural selection, neutral drift, and other factors. It is actually a hot
debate about which of the two, selection or drift, are the most influential
agents of change. Certainly neutral mutations accumulate in a genome faster
than selective ones.>
>While I can understand the mechanism of selection, I find it difficult to
>understand what drives neutral drift. I'd appreciate it if you would explain
>it. It seems to me to be a radical departure from natural selection. What is
>the significance of their "faster accumulation in the genome" that you
Nothing 'drives' neutral drift. It's simply variation that accumulates
in the absence of selection. For example, point mutations in the third
position of many codons are effectively neutral because the changes often
don't alter the amino acid sequence of proteins made. Similarly, there can
be changes made to proteins which have little or no effect on the immediate
viability of the organism. These are changes which can occur "under the
radar" of selection and pools of such variations can be supported in
a population with no immediate cost. These variations accumulate in the
population at approximately the rate of mutation (the majority of mutations
being neutral or negative), and that average rate is far greater than
rate of accumulation of positive mutations through selection.
Where does neutral drift fit in? One necessary factor for evolution
is the existence of variation within a population. Evolution won't happen
if species can't "move about in genomic space". Selection can "drive"
systems to particular locations (optima) in these "spaces", or neutral
drift can allow systems to "drift" across the landscape in "non-directed"
steps. One of the questions is: How much of the change which accumulates
during evolution resulted from neutral mutation and how much can be
attributed to selection?
[Note: Sometimes, as conditions change, variations which might once
started out neutral may actually have some affect on survival.]
Tim: <<Reversibility is irrelevant and in fact, many of the changes are not
reversible at the genetic level. Remember that there is a gaping chasm
between genotype and phenotype. The mapping between the two domains is
anything but linear in all but the rarest situations. Point mutations
can confer antibiotic resistance. However, given the odds of hitting that
exact site again to exactly reverse the point mutation, it's more likely
that a mutation that eliminates resistance will occur in another
portion of the genome. Thus a phenotypic reversal (short-term or otherwise)
does not equate to a genomic "reversal". And we have a recently described
case where a bacterium acquired streptomycin resistance in a "long-term"
manner such that loss of the resistance, even in the absence of the
antibiotic, was selected against (A secondary mutation occurred which
effectively "locked in" the first one). So, we can say that ratchet-like
evolutionary progression has been observed even over short timescales.>>
>One case doth not a theory make.
But one case doth destroy a blanket refutation...
(Further, the emergence of second-site suppressors is pretty commonly
observed. Some examples have become common in bacterial genetics
> Moreover this is not a ratchet-like
>evolutionary progression, as you claim. It is step one. Has anything
>been observed that builds on step one to produce step two? Unless it
>has you do not have a progression. You have merely a locked-in
The case I've described involved two-steps. First, the bacterium
acquired resistance to streptomycin. This is wonderful for such strains
when there is streptomycin around: All the wild-type strains which would
otherwise compete for food die. But the resistance carries a slight
penalty when no streptomycin is around. Strains carrying that mutation
tend to grow slightly slower and are thus handicapped against their
wild-type counterpart in an antibiotic-free environment. As a result,
these streptomycin resistant strains were not expected to persist in
the "outside" environment, where there is little of the antibiotic
However, there is a secondary mutation which can occur that eliminates
the growth handicaps of the resistant strains. This "suppressor" mutation
allows the bearers to not only grown when streptomycin is around but
also to grow as fast as the wild-type strains when the antibiotic is
Interestingly, this second mutation, by itself, also inhibits growth
rates. Only when teamed with the streptomycin resistance mutation
do the two mutations persist in the population. Strains missing either
one of the two mutations tend to be eliminated. This is an example
of lock-in (or ratcheting).
Tim: <<I suspect that true reversibility is a relatively uncommon
phenomenon. Should we be surprised? I think not. In the mind-bogglingly
huge morphology- or genetic-space available to organisms, it's not
likely that one could take more than a couple steps in any direction
and then return to exactly the same starting place.>>
>I suspect you are overstating your case. Does your "bacterium [that]
>acquired streptomycin resistance" that you mentioned earlier have
>"mind-bogglingly huge morphology- or genetic-space available to it"? I
>doubt it. If so, tell mme what it is.
E. coli has a genome of about 4.6 megabases. At four possible nucleotides
per base, that's a genome-space of about 4^(4.6E6) elements. But let's
start with one particular genome and calculate some odds of exactly
reversing a single point mutation. That's one in 4.6 million. What are
the odds of hitting anything else? Those are = 1 - 1/4.6E6 (or ~0.99999978).
So, long before you've reversed a _single_ mutation, you've probably hit
millions of other sites in the genome. That is the nature of genetic
Because most phenotypic traits (such a beak lengths) can be modified by
more than one gene (comprising thousands of bases/gene), odds are that
any phenotypic "reversals" aren't going to be genetic reversals.
Tim; <<Nobody is failing to acknowledge the hierarchical pattern of life and
discontinuities. These are what led to the conclusion of descent with
modification in the first place.>>
>Your first sentence in the above paragraph is simply not true. Read the
>statement that Dick Fischer presented and see if you can find hierarchical
>organization and discontinuities in it. Moreover, evolution abhors a
>discontinuity. Many evolutionary biologists are busy trying to iron them out.
From that statement:
"Nonetheless, no scientific hypothesis other than common descent
with modification can account for and make predictions about the
unity, diversity, and properties of living organisms." [...]
^^^^^ ^^^^^^^^^ ^^^^^^^^^^
Unity: hierarchical pattern of life
Diversity and properties: variations = discontinuities. There are
bears, dogs and cats. They are distinct entities now disconnected
from each other by breeding barriers.
Perhaps we are talking about slightly different things when use the word
"discontinuity". I think of it as the _appearance_ of gaps between related
species, rather than the _actual existence_ of gaps. True, evolution would
have problems with actual gaps. But the appearance of gaps (where there
really are none), is not unexpected, as Keith Miller explains elsewhere.
Tim: <<Discontinuity is a product of speciation (a mechanism of evolution)
Keith Miller described why (partially) in a recent post and many times in
the past. See: http://www.calvin.edu/archive/asa/200112/0080.html
Reiterated: Speciation reduces genetic exchange between populations
of organisms. After speciation, the separate populations may travel
on separate trajectories (i.e. diverge) and the amount of visible
divergence tends to increase over time. Also "intermediate" species
become extinct along the way. So while bears and dogs may appear very
different today, that was not the case for their progenitors. Similarly,
while extant mammals are clearly distinguishable from reptiles today,
the earliest mammals where not easy to differentiate from related
Tim: <<Such discontinuities tend to manifest more in the
morphological arena than the biochemical one. Also, hierarchical
organization is a manifestation of the rates of change and the nature
of speciation and organismal genetics. Understanding the timing and details
of specific steps behind evolution is real challenge, IMHO.>>
>Although these words may make sense to you, they don't to me.
Sorry, I wrote that too fast...
1) Common descent with modification is a process that generates
2) Organisms often (but don't always) differ more on the surface
(morphology) than they differ from a biochemical viewpoint. E.g.
Mammals may look different and live in different environments but
the underlying "hardware" is pretty similar.
3) The apparent distances between branches of the phylogenetic
tree of life reflect variations in the rates of change over
time. As long as the rates of morphological and genetic changes
are not too great, patterns deduced from morphological comparisons
can be expected to be consistent with the patterns created on the
basis of genetic comparisons.
And finally: Understanding the timing and details of specific steps
behind evolution is real challenge in the study of evolution.
>I think our brief discussion illustrates my point that the statement
>presented by Dick is incomplete, and I add, even misleading. It suggests
>that evolutionary theory is a done deal. I continue to hold that descent
>with modification needs a robustly supported mechanism (which in my view it
>doesn't have) and in addition, an unambiguous refutation of intelligent
>design, to substantiate its claim to be the all encompassing theory it wants
Descent with modification has an extremely robust mechanism and can be
paraphrased as follows: "If your parents didn't have children, neither
How the variations accumulated and what mechanisms fixed a particular
variation in a population are definitely open questions.
Regarding the unambiguous refutation of intelligent design: That is
simply impossible. One may as well ask: "What could an intelligent
designer not do?"
Tim Ikeda (firstname.lastname@example.org)
mail2web - Check your email from the web at
This archive was generated by hypermail 2b29 : Tue Dec 11 2001 - 16:54:29 EST