• AI as an alternative to TO

    From MarkE@me22over7@gmail.com to talk-origins on Mon Dec 15 23:09:20 2025
    From Newsgroup: talk.origins

    An interesting conversation with ChatGPT 5.2:

    Q1. With biological evolution of say mammals there are multiple traits
    (there must be thousands) that are simultaneously being optimised or conserved. How does does this work with purifying, selection, population genetics, fixation etc?

    https://chatgpt.com/s/t_693ffa4b47c881919ebcff6447a49748

    A. This question goes to the core of modern evolutionary genetics and is exactly where na|>ve rCLsingle-traitrCY intuitions break down. The short answer is: evolution does not optimise traits one-by-one. Instead,
    selection operates on whole organisms whose genomes encode thousands of partially coupled traits, and population genetics provides the machinery
    for how this can work without contradiction.

    I will structure the answer from first principles, then address the
    apparent paradox of rCLthousands of traits at oncerCY.

    1. What selection actually rCLseesrCY

    Natural selection does not see traits independently. It acts on:

    Whole organisms

    Whose phenotypes are emergent outcomes of

    Genotypes interacting with environments

    Formally, selection acts on differences in reproductive success
    (fitness). Traits matter only insofar as they contribute (positively, negatively, or neutrally) to that success.

    So the unit of selection is not:

    rCLeye sharpnessrCY, rCLbone densityrCY, rCLenzyme X efficiencyrCY

    but rather:

    organism A leaves more descendants than organism B in environment E.

    2. Genomes are high-dimensional systems, not collections of knobs

    A mammalian genome has:

    ~20,000 protein-coding genes

    Millions of regulatory elements

    Extensive pleiotropy (one gene affects many traits)

    Extensive epistasis (gene effects depend on other genes)

    This creates a high-dimensional fitness landscape, where:

    Each genotype is a point in ~millions of dimensions

    Fitness is a scalar value assigned to that point

    Selection therefore performs gradient ascent in a very high-dimensional
    space, not independent optimisation of each axis.

    Key implication:

    Most dimensions are nearly flat most of the time.

    3. Purifying selection dominates most of the genome

    Empirically, in mammals:

    The majority of mutations are neutral or nearly neutral

    A substantial fraction are deleterious

    Only a tiny fraction are beneficial at any time

    This is why purifying (negative) selection is the dominant mode:

    Mutations that disrupt conserved functions are removed

    Existing functional structures are maintained, not continuously rCLimprovedrCY

    Mathematically:

    For a deleterious mutation with selection coefficient
    EYaa
    <
    0
    s<0,

    Its fixation probability is approximately:

    EYaa
    ree
    1
    reA
    EYaA
    2
    EYaa
    1
    reA
    EYaA
    reA
    4
    EYaU
    EYaA
    EYaa
    Pree
    1reAe
    reA4N
    e
    rCi

    s
    1reAe
    2s
    rCi


    which is vanishingly small for moderate
    reu
    EYaa
    reu
    reusreu in large populations.

    Thus, thousands of traits are rCLsimultaneously conservedrCY simply because:

    Any mutation that degrades them is efficiently removed.

    No active optimisation is required.

    4. Nearly-neutral theory resolves the rCLtoo many traitsrCY intuition

    Motoo KimurarCOs nearly neutral theory is critical here.

    Key idea:

    Many mutations have effects
    reu
    EYaa
    reu
    ree
    1
    EYaU
    EYaA
    reusreuree
    N
    e
    rCi

    1
    rCi


    Their fate is governed largely by genetic drift, not selection

    Consequences:

    Small-effect mutations can drift through populations

    Large-effect deleterious mutations are purged

    Large-effect beneficial mutations are rare

    This means:

    Traits can wander within tolerance bounds

    As long as performance stays within a viable envelope, selection is indifferent

    This dramatically reduces the rCLcoordination problemrCY.

    5. Fixation does not require global optimisation

    A common misunderstanding is thinking fixation requires a trait to be rCLglobally optimalrCY.

    In reality:

    Fixation only requires local advantage or drift

    The population is almost never at a global optimum

    Most fixation events involve:

    Slight modifiers

    Compensatory changes

    Context-dependent advantages

    Important point:

    Fixation is path-dependent and historically contingent.

    Evolution rCLlocks inrCY workable solutions, not best ones.

    6. Trade-offs and constraint: why perfection is impossible

    Many traits are antagonistically coupled:

    Bone density vs. weight

    Immune sensitivity vs. autoimmunity

    Metabolic rate vs. longevity

    Formally:

    Fitness gradients for different traits often point in different directions

    The result is movement toward constraint surfaces, not maxima

    This leads to:

    Pareto fronts, not optima

    Stabilising selection around workable compromises

    So when you observe rCLthousands of traits conservedrCY, what you are seeing is:

    populations remaining inside a feasible region of phenotype space.

    7. Why selection doesnrCOt rCLcollapse under dimensionalityrCY

    Putting it together:

    Selection avoids a combinatorial explosion because:

    Most dimensions are neutral or nearly neutral

    Purifying selection removes large disruptions efficiently

    Pleiotropy couples traits, reducing effective dimensionality

    Selection acts on organisms, not traits

    Drift absorbs small-effect noise

    Stabilising selection maintains systems near viable equilibria

    This is why mammals can persist with:

    Highly conserved core systems (development, metabolism, cell biology)

    Slowly drifting peripheral traits

    Occasional adaptive shifts when environments change

    8. A useful analogy (non-hand-wavy)

    Think of a modern aircraft:

    Thousands of parameters

    Most cannot change without catastrophic failure

    Some can vary within tolerances

    Rare redesigns improve performance under new constraints

    Natural selection is not rCLdesigningrCY the aircraft from scratch each generation; it is:

    Rejecting planes that crash

    Allowing minor tolerated deviations

    Occasionally locking in improvements when conditions demand it

    9. Where this intersects ID-style intuitions (briefly, since you follow
    the literature)

    The intuition that rCLtoo many traits must be coordinatedrCY is not wrong,
    but the resolution is:

    Coordination is largely inherited, not newly created

    Selection is primarily conservative, not creative

    Evolution is maintenance with occasional modification, not constant
    innovation

    This is precisely why:

    Evolution is slow

    Major innovations are rare

    Lineages show long periods of stasis punctuated by change

    Summary

    Thousands of traits do not need to be optimised simultaneously.
    They are:

    Mostly conserved by purifying selection

    Buffered by developmental and genetic architecture

    Allowed to drift within tolerances

    Occasionally modified when local fitness gradients align

    Population genetics handles this cleanly once you stop thinking in terms
    of independent trait optimisation and start thinking in terms of high-dimensional, constrained, organism-level selection.

    If you want, next we can:

    Walk through a toy population-genetic model with many traits

    Discuss genetic load and HaldanerCOs dilemma in this context

    Or connect this explicitly to Evo-Devo constraints in mammals

    --- Synchronet 3.21a-Linux NewsLink 1.2
  • From MarkE@me22over7@gmail.com to talk-origins on Mon Dec 15 23:12:35 2025
    From Newsgroup: talk.origins

    Q2. You say "Major innovations are rare", but these are the driver of macroevolution. But how do these "major innovations" occur via many
    small steps?

    A. This is the right pressure point. If major innovations are both rare
    and essential for macroevolution, then a purely verbal appeal to rCLmany
    small stepsrCY is insufficient. The question becomes:

    What is the mechanism by which qualitatively new biological capacities
    arise via quantitatively small genetic changes, without collapsing
    fitness along the way?

    Below is the modern evolutionary answer, stated mechanistically, not rhetorically...

    https://chatgpt.com/s/t_693ffb26a698819184d0d92224b045b0

    --- Synchronet 3.21a-Linux NewsLink 1.2
  • From MarkE@me22over7@gmail.com to talk-origins on Mon Dec 15 23:16:25 2025
    From Newsgroup: talk.origins

    Q3. Yes [take one concrete case (e.g. mammalian middle ear, placenta, or adaptive immunity) and walk through the actual sequence of small,
    selectable steps in detail, so this remains empirical rather than abstract.]

    A. https://chatgpt.com/s/t_693ffc103e388191b2668e65fcca355d

    --- Synchronet 3.21a-Linux NewsLink 1.2
  • From MarkE@me22over7@gmail.com to talk-origins on Mon Dec 15 23:22:22 2025
    From Newsgroup: talk.origins

    Q4. This describes a possible macro change in the configuration of bones
    and associated tissue. But, for this to have functional advantage, simultaneously we need increased nerve connection capacity between the
    and brain, and more sophisticated brain auditory processing. What you
    have given is an example of irreducible complexity.

    A. https://chatgpt.com/s/t_693ffd777ab08191b83d4bdb7283c93c

    This is a serious objection, and it is correctly framed. You are not
    pointing to a gap of imagination (rCLmaybe nerves changed toorCY), but to a coupled-systems problem...

    --- Synchronet 3.21a-Linux NewsLink 1.2
  • From MarkE@me22over7@gmail.com to talk-origins on Mon Dec 15 23:27:33 2025
    From Newsgroup: talk.origins

    Q5. Examine one of the genuinely hard casesrCofor example: Origin of
    hair-cell mechanotransduction.

    A. https://chatgpt.com/s/t_693ffeaf88808191ae6e17e4d5334efa

    Hair-cell mechanotransduction (MET) is rCLhardrCY for exactly the reason you are targeting: it is a multi-component nanomachine (tip links, gating
    springs, a mechanically gated cation channel, lipid/actin couplings, adaptation motors) that must work at microsecondrComillisecond timescales.
    We can, however, say substantially more today than we could 15rCo20 years
    ago, because the core molecular players are now identified and
    reconstituted to varying degrees...

    --- Synchronet 3.21a-Linux NewsLink 1.2
  • From MarkE@me22over7@gmail.com to talk-origins on Mon Dec 15 23:31:20 2025
    From Newsgroup: talk.origins

    Q6. rCLMinimum viable METrCY: what is the smallest set of components that
    can yield directionally sensitive, tethered mechanotransduction (even if
    low fidelity)?

    A. https://chatgpt.com/s/t_693fff94b4e881919740963d356fec8a

    A rCLminimum viablerCY directionally sensitive, tethered MET system (even
    low fidelity) requires three functional modules:

    a force-bearing extracellular tether with defined polarity,

    a mechanically gated cation pore at one end of that tether, and

    a coupler/scaffold that converts tether tension into channel gating (and
    keeps the complex in the right place).

    ...

    --- Synchronet 3.21a-Linux NewsLink 1.2
  • From RonO@rokimoto557@gmail.com to talk-origins on Mon Dec 15 10:11:40 2025
    From Newsgroup: talk.origins

    On 12/15/2025 6:09 AM, MarkE wrote:
    An interesting conversation with ChatGPT 5.2:

    Q1. With biological evolution of say mammals there are multiple traits (there must be thousands) that are simultaneously being optimised or conserved. How does does this work with purifying, selection, population genetics, fixation etc?

    https://chatgpt.com/s/t_693ffa4b47c881919ebcff6447a49748

    A. This question goes to the core of modern evolutionary genetics and is exactly where na|>ve rCLsingle-traitrCY intuitions break down. The short answer is: evolution does not optimise traits one-by-one. Instead,
    selection operates on whole organisms whose genomes encode thousands of partially coupled traits, and population genetics provides the machinery
    for how this can work without contradiction.

    I will structure the answer from first principles, then address the
    apparent paradox of rCLthousands of traits at oncerCY.

    1. What selection actually rCLseesrCY

    Natural selection does not see traits independently. It acts on:

    Whole organisms

    Whose phenotypes are emergent outcomes of

    Genotypes interacting with environments

    Formally, selection acts on differences in reproductive success
    (fitness). Traits matter only insofar as they contribute (positively, negatively, or neutrally) to that success.

    So the unit of selection is not:

    rCLeye sharpnessrCY, rCLbone densityrCY, rCLenzyme X efficiencyrCY

    but rather:

    organism A leaves more descendants than organism B in environment E.

    2. Genomes are high-dimensional systems, not collections of knobs

    A mammalian genome has:

    ~20,000 protein-coding genes

    Millions of regulatory elements

    Extensive pleiotropy (one gene affects many traits)

    Extensive epistasis (gene effects depend on other genes)

    This creates a high-dimensional fitness landscape, where:

    Each genotype is a point in ~millions of dimensions

    Fitness is a scalar value assigned to that point

    Selection therefore performs gradient ascent in a very high-dimensional space, not independent optimisation of each axis.

    Key implication:

    Most dimensions are nearly flat most of the time.

    3. Purifying selection dominates most of the genome

    Empirically, in mammals:

    The majority of mutations are neutral or nearly neutral

    A substantial fraction are deleterious

    Only a tiny fraction are beneficial at any time

    This is why purifying (negative) selection is the dominant mode:

    Mutations that disrupt conserved functions are removed

    Existing functional structures are maintained, not continuously rCLimprovedrCY

    Mathematically:

    For a deleterious mutation with selection coefficient
    EYaa
    <
    0
    s<0,

    Its fixation probability is approximately:

    EYaa
    ree
    1
    reA
    EYaA
    2
    EYaa
    1
    reA
    EYaA
    reA
    4
    EYaU
    EYaA
    EYaa
    Pree
    1reAe
    reA4N
    e
    -a-a-a-arCi

    s
    1reAe
    2s
    -a-a-a-arCi


    which is vanishingly small for moderate
    reu
    EYaa
    reu
    reusreu in large populations.

    Thus, thousands of traits are rCLsimultaneously conservedrCY simply because:

    Any mutation that degrades them is efficiently removed.

    No active optimisation is required.

    4. Nearly-neutral theory resolves the rCLtoo many traitsrCY intuition

    Motoo KimurarCOs nearly neutral theory is critical here.

    Key idea:

    Many mutations have effects
    reu
    EYaa
    reu
    ree
    1
    EYaU
    EYaA
    reusreuree
    N
    e
    -a-a-a-arCi

    1
    -a-a-a-arCi


    Their fate is governed largely by genetic drift, not selection

    Consequences:

    Small-effect mutations can drift through populations

    Large-effect deleterious mutations are purged

    Large-effect beneficial mutations are rare

    This means:

    Traits can wander within tolerance bounds

    As long as performance stays within a viable envelope, selection is indifferent

    This dramatically reduces the rCLcoordination problemrCY.

    5. Fixation does not require global optimisation

    A common misunderstanding is thinking fixation requires a trait to be rCLglobally optimalrCY.

    In reality:

    Fixation only requires local advantage or drift

    The population is almost never at a global optimum

    Most fixation events involve:

    Slight modifiers

    Compensatory changes

    Context-dependent advantages

    Important point:

    Fixation is path-dependent and historically contingent.

    Evolution rCLlocks inrCY workable solutions, not best ones.

    6. Trade-offs and constraint: why perfection is impossible

    Many traits are antagonistically coupled:

    Bone density vs. weight

    Immune sensitivity vs. autoimmunity

    Metabolic rate vs. longevity

    Formally:

    Fitness gradients for different traits often point in different directions

    The result is movement toward constraint surfaces, not maxima

    This leads to:

    Pareto fronts, not optima

    Stabilising selection around workable compromises

    So when you observe rCLthousands of traits conservedrCY, what you are seeing is:

    populations remaining inside a feasible region of phenotype space.

    7. Why selection doesnrCOt rCLcollapse under dimensionalityrCY

    Putting it together:

    Selection avoids a combinatorial explosion because:

    Most dimensions are neutral or nearly neutral

    Purifying selection removes large disruptions efficiently

    Pleiotropy couples traits, reducing effective dimensionality

    Selection acts on organisms, not traits

    Drift absorbs small-effect noise

    Stabilising selection maintains systems near viable equilibria

    This is why mammals can persist with:

    Highly conserved core systems (development, metabolism, cell biology)

    Slowly drifting peripheral traits

    Occasional adaptive shifts when environments change

    8. A useful analogy (non-hand-wavy)

    Think of a modern aircraft:

    Thousands of parameters

    Most cannot change without catastrophic failure

    Some can vary within tolerances

    Rare redesigns improve performance under new constraints

    Natural selection is not rCLdesigningrCY the aircraft from scratch each generation; it is:

    Rejecting planes that crash

    Allowing minor tolerated deviations

    Occasionally locking in improvements when conditions demand it

    9. Where this intersects ID-style intuitions (briefly, since you follow
    the literature)

    The intuition that rCLtoo many traits must be coordinatedrCY is not wrong, but the resolution is:

    Coordination is largely inherited, not newly created

    Selection is primarily conservative, not creative

    Evolution is maintenance with occasional modification, not constant innovation

    This is precisely why:

    Evolution is slow

    Major innovations are rare

    Lineages show long periods of stasis punctuated by change

    Summary

    Thousands of traits do not need to be optimised simultaneously.
    They are:

    Mostly conserved by purifying selection

    Buffered by developmental and genetic architecture

    Allowed to drift within tolerances

    Occasionally modified when local fitness gradients align

    Population genetics handles this cleanly once you stop thinking in terms
    of independent trait optimisation and start thinking in terms of high- dimensional, constrained, organism-level selection.

    If you want, next we can:

    Walk through a toy population-genetic model with many traits

    Discuss genetic load and HaldanerCOs dilemma in this context

    Or connect this explicitly to Evo-Devo constraints in mammals


    What did you learn from this?

    ID is still a bait and switch scam because selection was never what you thought that it was. How do you think that Behe's three neutral
    mutations as evidence for design works? He has to put a time limit
    because when you deal with reality you find out things like mice and
    humans have been evolving as different lineages for around 80 million
    years. Most of their proteins may be around 20% different, so for an
    average protein of 300 amino acids there are around 60 neutral mutations
    that have occurred between mice and humans at this time. These proteins
    still do the same things in both mice and humans. We can take human
    proteins and put them into mice and they work. If some mutation occurs
    that changes the function of that gene a lot of those neutral mutations
    may be involved in the new function. They might not have changed the
    gene function enough to be selected for or against, but they could
    combine with other mutations to do something different.

    Behe needs his time limit because of the way that protein genes evolve.
    Your new information is forming all the time, and is how new traits
    evolve and are selected for or against.

    Everything has to work with what is already working. Whales lost a lot
    of genes because they didn't need them anymore, or they did things that
    were counter productive in the new environment that the whales were
    existing in.

    Ron Okimoto

    --- Synchronet 3.21a-Linux NewsLink 1.2
  • From RonO@rokimoto557@gmail.com to talk-origins on Mon Dec 15 10:24:39 2025
    From Newsgroup: talk.origins

    On 12/15/2025 6:12 AM, MarkE wrote:
    Q2. You say "Major innovations are rare", but these are the driver of macroevolution. But how do these "major innovations" occur via many
    small steps?

    A. This is the right pressure point. If major innovations are both rare
    and essential for macroevolution, then a purely verbal appeal to rCLmany small stepsrCY is insufficient. The question becomes:

    What is the mechanism by which qualitatively new biological capacities
    arise via quantitatively small genetic changes, without collapsing
    fitness along the way?

    Below is the modern evolutionary answer, stated mechanistically, not rhetorically...

    https://chatgpt.com/s/t_693ffb26a698819184d0d92224b045b0


    What macro "major innovations" needed to occur between chimps and humans?

    We have all the same tissues and organs. We have the same brain parts,
    just brain size has changed. We have all the same bones, but they have changed their shape and size. We are bipedal with a bent and shortened
    hip, but that deformity may have initally been due to a single gene
    defect with later modifications selected over time. We are still
    brachiating apes with shorter arms and longer legs (check out gymnasts).

    Ron Okimoto


    --- Synchronet 3.21a-Linux NewsLink 1.2
  • From RonO@rokimoto557@gmail.com to talk-origins on Mon Dec 15 10:30:26 2025
    From Newsgroup: talk.origins

    On 12/15/2025 6:16 AM, MarkE wrote:
    Q3. Yes [take one concrete case (e.g. mammalian middle ear, placenta, or adaptive immunity) and walk through the actual sequence of small,
    selectable steps in detail, so this remains empirical rather than
    abstract.]

    A. https://chatgpt.com/s/t_693ffc103e388191b2668e65fcca355d


    Obfuscation and denial will never support your Biblical beliefs because
    no mater how these things evolved on this planet the Bible is wrong
    about the creation.

    Were you impressed by the answer? We have those fossil intermediates.
    What does that tell you about Biblical creationism?

    Ron Okimoto

    --- Synchronet 3.21a-Linux NewsLink 1.2
  • From RonO@rokimoto557@gmail.com to talk-origins on Mon Dec 15 10:35:29 2025
    From Newsgroup: talk.origins

    On 12/15/2025 6:22 AM, MarkE wrote:
    Q4. This describes a possible macro change in the configuration of bones
    and associated tissue. But, for this to have functional advantage, simultaneously we need increased nerve connection capacity between the
    and brain, and more sophisticated brain auditory processing. What you
    have given is an example of irreducible complexity.

    A. https://chatgpt.com/s/t_693ffd777ab08191b83d4bdb7283c93c

    This is a serious objection, and it is correctly framed. You are not pointing to a gap of imagination (rCLmaybe nerves changed toorCY), but to a coupled-systems problem...


    What did you learn from the fact that your coupled system issue did not
    exist for this example? Reptiles were likely already "hearing" sensing
    sounds with their jaw bones. It likely evolved due to the jaw having to
    be able to sense things and be subject to muscle movements.

    Ron Okimoto

    --- Synchronet 3.21a-Linux NewsLink 1.2
  • From RonO@rokimoto557@gmail.com to talk-origins on Mon Dec 15 10:45:01 2025
    From Newsgroup: talk.origins

    On 12/15/2025 6:27 AM, MarkE wrote:
    Q5. Examine one of the genuinely hard casesrCofor example: Origin of hair- cell mechanotransduction.

    A. https://chatgpt.com/s/t_693ffeaf88808191ae6e17e4d5334efa

    Hair-cell mechanotransduction (MET) is rCLhardrCY for exactly the reason you are targeting: it is a multi-component nanomachine (tip links, gating springs, a mechanically gated cation channel, lipid/actin couplings, adaptation motors) that must work at microsecondrComillisecond timescales. We can, however, say substantially more today than we could 15rCo20 years ago, because the core molecular players are now identified and
    reconstituted to varying degrees...


    and these known components all evolved from preexisting components.

    Why pick the ear? Why not the eye? You do know that mammalian inner
    ear likely evolved before the crop plants existed that were supposed to
    have been created before land animals? Identifying the existence of
    this designer would just mean that the Bible is wrong about creation.
    The entire basis for your continued denial of reality would be found to
    be based on incorrect information. Why do you think that the other TO
    IDiotic creationists quit the ID scam when confronted by the reality of
    the Top Six?

    Ron Okimoto

    --- Synchronet 3.21a-Linux NewsLink 1.2
  • From RonO@rokimoto557@gmail.com to talk-origins on Mon Dec 15 10:54:54 2025
    From Newsgroup: talk.origins

    On 12/15/2025 6:31 AM, MarkE wrote:
    Q6. rCLMinimum viable METrCY: what is the smallest set of components that can yield directionally sensitive, tethered mechanotransduction (even if
    low fidelity)?

    A. https://chatgpt.com/s/t_693fff94b4e881919740963d356fec8a

    A rCLminimum viablerCY directionally sensitive, tethered MET system (even low fidelity) requires three functional modules:

    a force-bearing extracellular tether with defined polarity,

    a mechanically gated cation pore at one end of that tether, and

    a coupler/scaffold that converts tether tension into channel gating (and keeps the complex in the right place).

    ...


    IC failed and amounted to nothing for the ID creationist scam. This
    system requires 3 functions, but each function could exist and preform a function without the others. Together they produce a function that
    requires all three functions. It is one of the reasons that IC failed to
    be something that could not evolve. You need something more like Behe's
    "well matched" or you need more parts. Behe failed to define well
    matched so that it could be measured and his systems could be determined
    to have enough of it to matter, and he never could determine the maximum number of parts that would be impossible because he could never rule out multiple part subfunctions getting together.

    Ron Okimoto

    --- Synchronet 3.21a-Linux NewsLink 1.2
  • From MarkE@me22over7@gmail.com to talk-origins on Tue Dec 16 08:41:37 2025
    From Newsgroup: talk.origins

    On 16/12/2025 3:11 am, RonO wrote:
    On 12/15/2025 6:09 AM, MarkE wrote:
    An interesting conversation with ChatGPT 5.2:

    Q1. With biological evolution of say mammals there are multiple traits
    (there must be thousands) that are simultaneously being optimised or
    conserved. How does does this work with purifying, selection,
    population genetics, fixation etc?

    https://chatgpt.com/s/t_693ffa4b47c881919ebcff6447a49748

    A. This question goes to the core of modern evolutionary genetics and
    is exactly where na|>ve rCLsingle-traitrCY intuitions break down. The short >> answer is: evolution does not optimise traits one-by-one. Instead,
    selection operates on whole organisms whose genomes encode thousands
    of partially coupled traits, and population genetics provides the
    machinery for how this can work without contradiction.

    I will structure the answer from first principles, then address the
    apparent paradox of rCLthousands of traits at oncerCY.

    1. What selection actually rCLseesrCY

    Natural selection does not see traits independently. It acts on:

    Whole organisms

    Whose phenotypes are emergent outcomes of

    Genotypes interacting with environments

    Formally, selection acts on differences in reproductive success
    (fitness). Traits matter only insofar as they contribute (positively,
    negatively, or neutrally) to that success.

    So the unit of selection is not:

    rCLeye sharpnessrCY, rCLbone densityrCY, rCLenzyme X efficiencyrCY

    but rather:

    organism A leaves more descendants than organism B in environment E.

    2. Genomes are high-dimensional systems, not collections of knobs

    A mammalian genome has:

    ~20,000 protein-coding genes

    Millions of regulatory elements

    Extensive pleiotropy (one gene affects many traits)

    Extensive epistasis (gene effects depend on other genes)

    This creates a high-dimensional fitness landscape, where:

    Each genotype is a point in ~millions of dimensions

    Fitness is a scalar value assigned to that point

    Selection therefore performs gradient ascent in a very high-
    dimensional space, not independent optimisation of each axis.

    Key implication:

    Most dimensions are nearly flat most of the time.

    3. Purifying selection dominates most of the genome

    Empirically, in mammals:

    The majority of mutations are neutral or nearly neutral

    A substantial fraction are deleterious

    Only a tiny fraction are beneficial at any time

    This is why purifying (negative) selection is the dominant mode:

    Mutations that disrupt conserved functions are removed

    Existing functional structures are maintained, not continuously
    rCLimprovedrCY

    Mathematically:

    For a deleterious mutation with selection coefficient
    EYaa
    <
    0
    s<0,

    Its fixation probability is approximately:

    EYaa
    ree
    1
    reA
    EYaA
    2
    EYaa
    1
    reA
    EYaA
    reA
    4
    EYaU
    EYaA
    EYaa
    Pree
    1reAe
    reA4N
    e
    -a-a-a-a-arCi

    s
    1reAe
    2s
    -a-a-a-a-arCi


    which is vanishingly small for moderate
    reu
    EYaa
    reu
    reusreu in large populations.

    Thus, thousands of traits are rCLsimultaneously conservedrCY simply because: >>
    Any mutation that degrades them is efficiently removed.

    No active optimisation is required.

    4. Nearly-neutral theory resolves the rCLtoo many traitsrCY intuition

    Motoo KimurarCOs nearly neutral theory is critical here.

    Key idea:

    Many mutations have effects
    reu
    EYaa
    reu
    ree
    1
    EYaU
    EYaA
    reusreuree
    N
    e
    -a-a-a-a-arCi

    1
    -a-a-a-a-arCi


    Their fate is governed largely by genetic drift, not selection

    Consequences:

    Small-effect mutations can drift through populations

    Large-effect deleterious mutations are purged

    Large-effect beneficial mutations are rare

    This means:

    Traits can wander within tolerance bounds

    As long as performance stays within a viable envelope, selection is
    indifferent

    This dramatically reduces the rCLcoordination problemrCY.

    5. Fixation does not require global optimisation

    A common misunderstanding is thinking fixation requires a trait to be
    rCLglobally optimalrCY.

    In reality:

    Fixation only requires local advantage or drift

    The population is almost never at a global optimum

    Most fixation events involve:

    Slight modifiers

    Compensatory changes

    Context-dependent advantages

    Important point:

    Fixation is path-dependent and historically contingent.

    Evolution rCLlocks inrCY workable solutions, not best ones.

    6. Trade-offs and constraint: why perfection is impossible

    Many traits are antagonistically coupled:

    Bone density vs. weight

    Immune sensitivity vs. autoimmunity

    Metabolic rate vs. longevity

    Formally:

    Fitness gradients for different traits often point in different
    directions

    The result is movement toward constraint surfaces, not maxima

    This leads to:

    Pareto fronts, not optima

    Stabilising selection around workable compromises

    So when you observe rCLthousands of traits conservedrCY, what you are
    seeing is:

    populations remaining inside a feasible region of phenotype space.

    7. Why selection doesnrCOt rCLcollapse under dimensionalityrCY

    Putting it together:

    Selection avoids a combinatorial explosion because:

    Most dimensions are neutral or nearly neutral

    Purifying selection removes large disruptions efficiently

    Pleiotropy couples traits, reducing effective dimensionality

    Selection acts on organisms, not traits

    Drift absorbs small-effect noise

    Stabilising selection maintains systems near viable equilibria

    This is why mammals can persist with:

    Highly conserved core systems (development, metabolism, cell biology)

    Slowly drifting peripheral traits

    Occasional adaptive shifts when environments change

    8. A useful analogy (non-hand-wavy)

    Think of a modern aircraft:

    Thousands of parameters

    Most cannot change without catastrophic failure

    Some can vary within tolerances

    Rare redesigns improve performance under new constraints

    Natural selection is not rCLdesigningrCY the aircraft from scratch each
    generation; it is:

    Rejecting planes that crash

    Allowing minor tolerated deviations

    Occasionally locking in improvements when conditions demand it

    9. Where this intersects ID-style intuitions (briefly, since you
    follow the literature)

    The intuition that rCLtoo many traits must be coordinatedrCY is not wrong, >> but the resolution is:

    Coordination is largely inherited, not newly created

    Selection is primarily conservative, not creative

    Evolution is maintenance with occasional modification, not constant
    innovation

    This is precisely why:

    Evolution is slow

    Major innovations are rare

    Lineages show long periods of stasis punctuated by change

    Summary

    Thousands of traits do not need to be optimised simultaneously.
    They are:

    Mostly conserved by purifying selection

    Buffered by developmental and genetic architecture

    Allowed to drift within tolerances

    Occasionally modified when local fitness gradients align

    Population genetics handles this cleanly once you stop thinking in
    terms of independent trait optimisation and start thinking in terms of
    high- dimensional, constrained, organism-level selection.

    If you want, next we can:

    Walk through a toy population-genetic model with many traits

    Discuss genetic load and HaldanerCOs dilemma in this context

    Or connect this explicitly to Evo-Devo constraints in mammals


    What did you learn from this?

    ID is still a bait and switch scam because selection was never what you thought that it was.-a How do you think that Behe's three neutral
    mutations as evidence for design works?-a He has to put a time limit
    because when you deal with reality you find out things like mice and
    humans have been evolving as different lineages for around 80 million years.-a Most of their proteins may be around 20% different, so for an average protein of 300 amino acids there are around 60 neutral mutations that have occurred between mice and humans at this time.-a These proteins still do the same things in both mice and humans.-a We can take human proteins and put them into mice and they work.-a If some mutation occurs that changes the function of that gene a lot of those neutral mutations
    may be involved in the new function.-a They might not have changed the
    gene function enough to be selected for or against, but they could
    combine with other mutations to do something different.

    Behe needs his time limit because of the way that protein genes evolve.
    Your new information is forming all the time, and is how new traits
    evolve and are selected for or against.

    Everything has to work with what is already working.-a Whales lost a lot
    of genes because they didn't need them anymore, or they did things that
    were counter productive in the new environment that the whales were
    existing in.

    Ron Okimoto


    Calling this a rCLbait and switchrCY presupposes that ID once relied on an incorrect understanding of selection and later changed its claims. That
    is historically false.

    ID does not deny natural selection, nor has it.

    Behe explicitly accepts mutation, selection, genetic drift, and as I've previously noted, common descent. His argument concerns the creative
    limits of these mechanisms, not their existence.

    BeherCOs position in DarwinrCOs Black Box, as far back as 1996 (which has remained consistent) is that selection explains optimization, not
    origination of complex molecular systems.

    --- Synchronet 3.21a-Linux NewsLink 1.2
  • From RonO@rokimoto557@gmail.com to talk-origins on Mon Dec 15 18:14:42 2025
    From Newsgroup: talk.origins

    On 12/15/2025 3:41 PM, MarkE wrote:
    On 16/12/2025 3:11 am, RonO wrote:
    On 12/15/2025 6:09 AM, MarkE wrote:
    An interesting conversation with ChatGPT 5.2:

    Q1. With biological evolution of say mammals there are multiple
    traits (there must be thousands) that are simultaneously being
    optimised or conserved. How does does this work with purifying,
    selection, population genetics, fixation etc?

    https://chatgpt.com/s/t_693ffa4b47c881919ebcff6447a49748

    A. This question goes to the core of modern evolutionary genetics and
    is exactly where na|>ve rCLsingle-traitrCY intuitions break down. The
    short answer is: evolution does not optimise traits one-by-one.
    Instead, selection operates on whole organisms whose genomes encode
    thousands of partially coupled traits, and population genetics
    provides the machinery for how this can work without contradiction.

    I will structure the answer from first principles, then address the
    apparent paradox of rCLthousands of traits at oncerCY.

    1. What selection actually rCLseesrCY

    Natural selection does not see traits independently. It acts on:

    Whole organisms

    Whose phenotypes are emergent outcomes of

    Genotypes interacting with environments

    Formally, selection acts on differences in reproductive success
    (fitness). Traits matter only insofar as they contribute (positively,
    negatively, or neutrally) to that success.

    So the unit of selection is not:

    rCLeye sharpnessrCY, rCLbone densityrCY, rCLenzyme X efficiencyrCY

    but rather:

    organism A leaves more descendants than organism B in environment E.

    2. Genomes are high-dimensional systems, not collections of knobs

    A mammalian genome has:

    ~20,000 protein-coding genes

    Millions of regulatory elements

    Extensive pleiotropy (one gene affects many traits)

    Extensive epistasis (gene effects depend on other genes)

    This creates a high-dimensional fitness landscape, where:

    Each genotype is a point in ~millions of dimensions

    Fitness is a scalar value assigned to that point

    Selection therefore performs gradient ascent in a very high-
    dimensional space, not independent optimisation of each axis.

    Key implication:

    Most dimensions are nearly flat most of the time.

    3. Purifying selection dominates most of the genome

    Empirically, in mammals:

    The majority of mutations are neutral or nearly neutral

    A substantial fraction are deleterious

    Only a tiny fraction are beneficial at any time

    This is why purifying (negative) selection is the dominant mode:

    Mutations that disrupt conserved functions are removed

    Existing functional structures are maintained, not continuously
    rCLimprovedrCY

    Mathematically:

    For a deleterious mutation with selection coefficient
    EYaa
    <
    0
    s<0,

    Its fixation probability is approximately:

    EYaa
    ree
    1
    reA
    EYaA
    2
    EYaa
    1
    reA
    EYaA
    reA
    4
    EYaU
    EYaA
    EYaa
    Pree
    1reAe
    reA4N
    e
    -a-a-a-a-arCi

    s
    1reAe
    2s
    -a-a-a-a-arCi


    which is vanishingly small for moderate
    reu
    EYaa
    reu
    reusreu in large populations.

    Thus, thousands of traits are rCLsimultaneously conservedrCY simply because:

    Any mutation that degrades them is efficiently removed.

    No active optimisation is required.

    4. Nearly-neutral theory resolves the rCLtoo many traitsrCY intuition

    Motoo KimurarCOs nearly neutral theory is critical here.

    Key idea:

    Many mutations have effects
    reu
    EYaa
    reu
    ree
    1
    EYaU
    EYaA
    reusreuree
    N
    e
    -a-a-a-a-arCi

    1
    -a-a-a-a-arCi


    Their fate is governed largely by genetic drift, not selection

    Consequences:

    Small-effect mutations can drift through populations

    Large-effect deleterious mutations are purged

    Large-effect beneficial mutations are rare

    This means:

    Traits can wander within tolerance bounds

    As long as performance stays within a viable envelope, selection is
    indifferent

    This dramatically reduces the rCLcoordination problemrCY.

    5. Fixation does not require global optimisation

    A common misunderstanding is thinking fixation requires a trait to be
    rCLglobally optimalrCY.

    In reality:

    Fixation only requires local advantage or drift

    The population is almost never at a global optimum

    Most fixation events involve:

    Slight modifiers

    Compensatory changes

    Context-dependent advantages

    Important point:

    Fixation is path-dependent and historically contingent.

    Evolution rCLlocks inrCY workable solutions, not best ones.

    6. Trade-offs and constraint: why perfection is impossible

    Many traits are antagonistically coupled:

    Bone density vs. weight

    Immune sensitivity vs. autoimmunity

    Metabolic rate vs. longevity

    Formally:

    Fitness gradients for different traits often point in different
    directions

    The result is movement toward constraint surfaces, not maxima

    This leads to:

    Pareto fronts, not optima

    Stabilising selection around workable compromises

    So when you observe rCLthousands of traits conservedrCY, what you are
    seeing is:

    populations remaining inside a feasible region of phenotype space.

    7. Why selection doesnrCOt rCLcollapse under dimensionalityrCY

    Putting it together:

    Selection avoids a combinatorial explosion because:

    Most dimensions are neutral or nearly neutral

    Purifying selection removes large disruptions efficiently

    Pleiotropy couples traits, reducing effective dimensionality

    Selection acts on organisms, not traits

    Drift absorbs small-effect noise

    Stabilising selection maintains systems near viable equilibria

    This is why mammals can persist with:

    Highly conserved core systems (development, metabolism, cell biology)

    Slowly drifting peripheral traits

    Occasional adaptive shifts when environments change

    8. A useful analogy (non-hand-wavy)

    Think of a modern aircraft:

    Thousands of parameters

    Most cannot change without catastrophic failure

    Some can vary within tolerances

    Rare redesigns improve performance under new constraints

    Natural selection is not rCLdesigningrCY the aircraft from scratch each >>> generation; it is:

    Rejecting planes that crash

    Allowing minor tolerated deviations

    Occasionally locking in improvements when conditions demand it

    9. Where this intersects ID-style intuitions (briefly, since you
    follow the literature)

    The intuition that rCLtoo many traits must be coordinatedrCY is not
    wrong, but the resolution is:

    Coordination is largely inherited, not newly created

    Selection is primarily conservative, not creative

    Evolution is maintenance with occasional modification, not constant
    innovation

    This is precisely why:

    Evolution is slow

    Major innovations are rare

    Lineages show long periods of stasis punctuated by change

    Summary

    Thousands of traits do not need to be optimised simultaneously.
    They are:

    Mostly conserved by purifying selection

    Buffered by developmental and genetic architecture

    Allowed to drift within tolerances

    Occasionally modified when local fitness gradients align

    Population genetics handles this cleanly once you stop thinking in
    terms of independent trait optimisation and start thinking in terms
    of high- dimensional, constrained, organism-level selection.

    If you want, next we can:

    Walk through a toy population-genetic model with many traits

    Discuss genetic load and HaldanerCOs dilemma in this context

    Or connect this explicitly to Evo-Devo constraints in mammals


    What did you learn from this?

    ID is still a bait and switch scam because selection was never what
    you thought that it was.-a How do you think that Behe's three neutral
    mutations as evidence for design works?-a He has to put a time limit
    because when you deal with reality you find out things like mice and
    humans have been evolving as different lineages for around 80 million
    years.-a Most of their proteins may be around 20% different, so for an
    average protein of 300 amino acids there are around 60 neutral
    mutations that have occurred between mice and humans at this time.
    These proteins still do the same things in both mice and humans.-a We
    can take human proteins and put them into mice and they work.-a If some
    mutation occurs that changes the function of that gene a lot of those
    neutral mutations may be involved in the new function.-a They might not
    have changed the gene function enough to be selected for or against,
    but they could combine with other mutations to do something different.

    Behe needs his time limit because of the way that protein genes
    evolve. Your new information is forming all the time, and is how new
    traits evolve and are selected for or against.

    Everything has to work with what is already working.-a Whales lost a
    lot of genes because they didn't need them anymore, or they did things
    that were counter productive in the new environment that the whales
    were existing in.

    Ron Okimoto


    Calling this a rCLbait and switchrCY presupposes that ID once relied on an incorrect understanding of selection and later changed its claims. That
    is historically false.

    I wrote that selection was never what you thought that it was.

    The ID scam started running the bait and switch scam with the
    understanding that they did not have what they claimed to have. They understood that they had nothing to teach that was equivalent science to
    what they wanted to deny. In that regard the only excuse that they
    might have is that they misunderstood what they were making claims about.

    The ID perps did not have to have an incorrect understanding of
    selection in order to be lying about having any ID science worth
    teaching. There is nothing historically false about what I wrote about
    the bait and switch ID scam. The ID perps made the claim that they had
    the ID science to teach in the public schools, but when it came time to
    put up or shut up, they started running the bait and switch scam that
    they are still running to this day.


    ID does not deny natural selection, nor has it.

    The whole point about the ID scam's prevarications about Darwinism is to
    deny that natural selection can be one of the explanations for the
    diversity of life that exists on this planet. They understood from the
    very beginning that natural selection was only one of the factors
    accounting for the evolution of life on earth, but they chose to focus
    their efforts in trying to deny that it could account for what they were claiming was intelligent design. Behe had to acknowledge that genetic
    drift could also be a major factor in order to make his stupid waiting
    time nonsense that he could never find examples of 3 neutral mutations
    needing to have occurred under his time limits. He already understood
    that biological evolution could occur by 2 neutral mutations because
    there were already examples of that occurring. His only claim was that
    3 would be such a low probability event that it was impossible, but even
    that is just wrong. If something does happen it doesn't matter how
    improbable it is, it still would have happened.


    Behe explicitly accepts mutation, selection, genetic drift, and as I've previously noted, common descent. His argument concerns the creative
    limits of these mechanisms, not their existence.

    As I have acknowledged his understanding is the reason for his junk to
    be nonsense. He knows that his claims are not scientific and do not
    support the ID scam because he only has the claims, and he doesn't have
    any proof that 3 neutral mutations ever had to occur within his time
    limit. They would obviously be very rare events, and he has found no
    such events. Behe refuses to look for them in his IC systems. He knows
    what has to be done, but he claims that it isn't up to him to verify his
    own claims. Without verification his claims cannot support anything.


    BeherCOs position in DarwinrCOs Black Box, as far back as 1996 (which has remained consistent) is that selection explains optimization, not origination of complex molecular systems.


    Natural selection is expected to optimize the reproduction of each
    lineage under the current environmental conditions, and the
    environmental conditions include other lifeforms as soon as the first
    life form reproduced itself. Optimization is not stagnation. Every
    change that works within what is already working that improves the
    chance of reproduction is expected to be selected for. That is why
    Dembski had to admit that natural selection could be the designer of
    what he was claiming was specified complexity. He started calling this
    lower level specified complexity, but it was still increasing specified complexity of the systems. Behe admitted that this was true when he
    claimed that natural selection could account for the selection of his
    two neutral mutations that accounted for a new function. He just
    claimed that natural selection would have never been given the
    opportunity to select for 3 neutral mutations, and that if he found such examples that, that would be evidence for design, but Behe has never
    found any examples of 3 neutral mutations. Both Dembski and Behe
    understand that there is no limit for single mutation changes in
    function being selected by natural selection.

    The ID scam is just a scientific failure. The ID perps knew that it was
    a failure before they started running the bait and switch. Behe
    admitted that some IC systems could evolve at the turn of the century.
    Mike Gene claimed to have attended the early ID perp conferences, but
    never became a fellow, or, at least, never admitted to becoming a
    fellow, and Mike Gene claimed that he had given up on teaching the ID
    science back in 1999, years before the bait and switch started to go
    down in 2002. You know that Nelson started admitting that they never
    had the ID science to teach, but that they were working on creating
    some. Both Nelson and Mike Gene kept supporting using ID as bait, and
    that is all ID has ever become. Mike gene did not quit the ID scam and
    admit that the ID science never had existed until 2007. Nelson is still
    an ID perp promoting the bait even though he has acknowledged that they
    don't have any ID science. All he claims is that they are trying to accomplish some valid science.

    There is nothing for an honest creationist to support. You have known
    that for over 2 decades because you watched the stupid ID scam progress
    to the point where they got Santorum to submit his "amendment" to the no
    child left behind legislation and even though it was not included in the legislation (it could be found in some type of appendix for the Bill
    that included it as something submitted at one of the conferences where
    the bill was written). The ID perp claims about that got a lot of
    creationist rubes interested in teaching ID. Ohio was just the first,
    but there were likely at least a half dozen others that needed to have
    the bait and switch run on them within half a year of the first bait and switch on Ohio. Before Dover the ID perps had a list of over 20
    examples where they claimed the rubes were still contemplating going for
    the switch scam. I don't think that any of them had done so at that
    time except for Ohio. I did not start calling the ID perps "perps" for perpetrating the bait and switch scam until the bait and switch had gone
    down 100% of the time on creationist rubes for nearly 3 years. By that
    time there was no doubt about what the ID perps were doing with ID, and
    Ohio was still the only bunch of creationist rubes that had bent over
    for the switch scam. I don't recall any other legislators nor school
    boards that bent over for the switch scam once the bait and switch had
    gone down at that time. No one seemed to be interested in teaching the obfuscation and denial if they could not lie to the students about the religious reason for the scam. Texas and Louisiana did not pass their
    switch scam junk until after the ID perp's failure in Dover and both of
    them had to have the bait and switch rerun on them when they tried to
    teach ID using the switch scam.

    Ron Okimoto


    --- Synchronet 3.21a-Linux NewsLink 1.2
  • From MarkE@me22over7@gmail.com to talk-origins on Tue Dec 16 13:14:03 2025
    From Newsgroup: talk.origins

    On 16/12/2025 3:24 am, RonO wrote:
    On 12/15/2025 6:12 AM, MarkE wrote:
    Q2. You say "Major innovations are rare", but these are the driver of
    macroevolution. But how do these "major innovations" occur via many
    small steps?

    A. This is the right pressure point. If major innovations are both
    rare and essential for macroevolution, then a purely verbal appeal to
    rCLmany small stepsrCY is insufficient. The question becomes:

    What is the mechanism by which qualitatively new biological capacities
    arise via quantitatively small genetic changes, without collapsing
    fitness along the way?

    Below is the modern evolutionary answer, stated mechanistically, not
    rhetorically...

    https://chatgpt.com/s/t_693ffb26a698819184d0d92224b045b0


    What macro "major innovations" needed to occur between chimps and humans?

    We have all the same tissues and organs.-a We have the same brain parts, just brain size has changed.-a We have all the same bones, but they have changed their shape and size.-a We are bipedal with a bent and shortened hip, but that deformity may have initally been due to a single gene
    defect with later modifications selected over time.-a We are still brachiating apes with shorter arms and longer legs (check out gymnasts).


    No. That assessment grossly understates the differences.

    For details, see my post "Chimp to human evolution - Sandwalk perspective".

    --- Synchronet 3.21a-Linux NewsLink 1.2
  • From RonO@rokimoto557@gmail.com to talk-origins on Mon Dec 15 21:18:47 2025
    From Newsgroup: talk.origins

    On 12/15/2025 8:14 PM, MarkE wrote:
    On 16/12/2025 3:24 am, RonO wrote:
    On 12/15/2025 6:12 AM, MarkE wrote:
    Q2. You say "Major innovations are rare", but these are the driver of
    macroevolution. But how do these "major innovations" occur via many
    small steps?

    A. This is the right pressure point. If major innovations are both
    rare and essential for macroevolution, then a purely verbal appeal to
    rCLmany small stepsrCY is insufficient. The question becomes:

    What is the mechanism by which qualitatively new biological
    capacities arise via quantitatively small genetic changes, without
    collapsing fitness along the way?

    Below is the modern evolutionary answer, stated mechanistically, not
    rhetorically...

    https://chatgpt.com/s/t_693ffb26a698819184d0d92224b045b0


    What macro "major innovations" needed to occur between chimps and humans?

    We have all the same tissues and organs.-a We have the same brain
    parts, just brain size has changed.-a We have all the same bones, but
    they have changed their shape and size.-a We are bipedal with a bent
    and shortened hip, but that deformity may have initally been due to a
    single gene defect with later modifications selected over time.-a We
    are still brachiating apes with shorter arms and longer legs (check
    out gymnasts).


    No. That assessment grossly understates the differences.

    For details, see my post "Chimp to human evolution - Sandwalk perspective".


    Been there done that, and found nothing. Why can't you state what macro evolutionary changes had to occur? No new bones, no new tissues or
    organs, no new brain parts. The hip was deformed, but how macro was
    that? It was just a survivable hip deformation that may have had a
    selective advantage in climbing narrow enough tree trunks like you see
    humans climbing coconut palms. We can't find any major brain changes
    except for the size of some brain parts. Humans remain brachiating apes
    with big brains. The original apes that walked upright and had the hip deformation still had ape feet and were likely mostly arboreal. Australopithicine had more human like feet, but they were just upright
    walking apes for the rest of their anatomy and still had arboreal limb proportions (long arms and short legs). Homo habilis had a larger brain
    and the Homo parabolic jaw instead of the ape jaw, but still had
    arboreal limb proportions. The parabolic jaw likely evolved with the reduction in the canine teeth because Homo no longer needed an extended
    face to bite with to get their canines into play. Homo erectus had a
    larger brain, and during the evolution of that type eventually evolved a
    brain size that was within the range of extant modern humans, and from
    the neck down were essentially the same as modern humans except for
    their ribs. Their ribs still flared out over the hips like an ape
    instead of tapering to the waste giving humans the hourglass figure
    instead of the fire hydrant physique of Homo erectus and Neanderthal.
    Where are the macro evolutionary changes, and why did they evolve by
    smaller changes over time?

    Ron Okimoto

    --- Synchronet 3.21a-Linux NewsLink 1.2