Convergent Evolution and Design Reuse

Camera eyes, sonar and wolf-like skulls each arose more than once, but from different parts. Where the same amino acid changes recur, the silent DNA sites still follow the family tree. That fits separate solutions under selection better than one reused design. Appeals to convergence in anatomy and fossils still lack an independent test.

By AI (Claude Fable 5.1, reasoning effort 30) Draft:  September 2026

"If human craftsmen made in God’s image reuse the same techniques and technologies, it’s reasonable to infer that their Creator would do the same." — Fazale Rana, The Cell's Design, 20081

"when only synonymous substitutions are used, dolphin and cow are correctly grouped with 100% bootstrap support" — Ying Li, Zhen Liu, Peng Shi and Jianzhi Zhang, on the hearing gene prestin, 20102

Summary

The same complex feature often turns up in creatures that no family tree puts together.  Octopus and human both see with a camera eye.  Bats and dolphins both hunt by sonar.  The extinct marsupial thylacine had a dog's skull.  Evolutionary biologists call this convergence and credit selection.  Creationist and intelligent-design writers call it one designer reusing a design.  This article tests both readings against the primary papers.

  1. The repeats are real and common. Simon Conway Morris counts at least six separate camera eyes.3 Rowan Sage counts 62 separate lineages of C4 photosynthesis.4
  2. In most cases the likeness stops at the outside. The octopus eye has a different retina, different light cells and a lens made of a different protein. Arctic and Antarctic fish make nearly the same antifreeze from unrelated genes. The thylacine shares no more protein changes with dogs than any two mammals that far apart.
  3. Where the same amino acid changes do recur, they are few and sit where the work is done. Prestin has 11 in the dolphin and echolocating bats. The sodium pump has two.
  4. The silent sites do not converge. In prestin, in a plant enzyme and in snake mitochondria, DNA positions that do not change the protein follow the ordinary tree.
  5. The genome-wide claim failed. The 2013 report of nearly 200 converging genes in bats and dolphins did not survive two reanalyses in 2015. A bat and a cow share more.
  6. The design reading comes in several forms that predict different things. Simple reuse of the same module predicts the same parts, and the parts differ. Walter ReMine's and Winston Ewert's versions fit the data, but at these points they predict what selection predicts.
  7. The evolutionary reading has its own gap. Convergence is tested well where DNA is available. For anatomy and fossils it is still assigned after a trait conflicts with the tree.

Verdict:  the data favour separate solutions to the same problem, each built from a lineage's own parts, over one design copied between creatures.  This is the weakest ground for the design argument from similarity.  The stronger question convergence raises is why workable designs are so few, and no one has settled it.

Part 1: What the design-reuse reading claims

The reading is not one theory.  Its makers say different things, and each is quoted here before it is tested.

Fazale Rana is a biochemist with the old-earth ministry Reasons to Believe.  In The Cell's Design he argues from human practice:  craftsmen reuse techniques, so a Creator would.1  He then states what evolution should show instead:  "If biochemical systems are the product of evolution, then the same biochemical designs should not recur throughout nature."1  He sorts molecular convergence into five kinds and concludes that its growing number of cases "is unexpected if life results from historical sequences of chance evolutionary events."1  His evolutionary prediction is Stephen Jay Gould's, taken up in Part 6.

Paul Nelson and Jonathan Wells put it as module reuse:  "An intelligent cause may reuse or redeploy the same module in different systems, without there necessarily being any material or physical connection between those systems."5

Walter ReMine says something different.  In The Biotic Message the designer wants life to look like the work of one mind and to resist every natural explanation.  Too much identical reuse would look like genes being moved between species, which he calls transposition.  So convergent features must be alike but not the same:  "If characters are too similar then they look like transposition, but if they are too different then they fail to demand special (and formidable) rationalizations like the evolutionists’ convergence explanation.  This delicate balance requires a designer."6  He also expects different designs for one function:  "a designer must not indiscriminately use the same design repeatedly."6  And he ranks the patterns:  the nested pattern strong, "the pattern of convergence must be comparatively weaker."6

Winston Ewert borrows from software.  Species share modules the way programs share libraries.7  His example graph has an echolocation module:  "the echolocating species depend on the echolocation module."7  He fitted the model to which gene families each species has, not to sequences.  The site's reproduction of that result is in Phylogenetic Congruence.  In 2023 he applied a sequence version to prestin, discussed in Part 3.

The young-earth ministries state the argument more simply.  John UpChurch of Answers in Genesis:  "Human designers do the same thing—integrating similar parts in different things they design."8  Jonathan Sarfati of Creation Ministries International adopts ReMine's theory by name.9  Randy Guliuzza of the Institute for Creation Research attacks the rival:  convergence "is a declaration based on mental pictures."10

Todd Wood, the leading young-earth student of created kinds, is more careful.  Of Ewert's paper he wrote:  "this is not a paradigm-breaking, epoch-making breakthrough.  It's a good step."  Of his own unpublished work on the pattern of life:  "I've been dissatisfied with the projects."11

Part 2: Camera eyes

A camera eye has one lens that focuses an image on a sheet of light cells.  Vertebrates have it.  So do squid and octopus, which are molluscs.  Conway Morris lists more:  a family of marine worms, some snails, box jellyfish and certain spiders.3

The two famous eyes differ in nearly every part:

Part Vertebrate Octopus and squid
Retina Light cells face away from the light, behind the nerve layers Light cells face the light
Light cell Ciliary Rhabdomeric, a single type
Image processing Several nerve cell types inside the retina Done in the optic lobe of the brain
Lens protein No S-crystallin, which only cephalopods have S-crystallin, a modified detoxifying enzyme
Retina forms from A different embryonic source Infolding of the outer layer, the skin

The retina, light cell and processing rows are from a 2025 single-cell atlas of the bobtail squid, which also describes the lens proteins.1212  The vertebrate arrangement is examined in The Inverted Retina.  The octopus lens protein is a relative of glutathione S-transferase that has "almost lost" its enzyme activity.13  Conway Morris gives the embryology.3

What is shared.  The gene Pax-6 helps start eye development in both.  Squid Pax-6 placed in a fruit fly makes extra eyes grow on the fly's wing, and they are fly eyes.14  The switch is interchangeable.  The design of the eye is not in the switch.  Even the switch is varied differently:  vertebrates use four spliced forms of Pax-6, squid use five unrelated ones, and insects use duplicate genes.15

Ogura 2004.  Atsushi Ogura, Kazuho Ikeo and Takashi Gojobori sequenced the genes active in octopus eye.  Of 1,052 identified, 729, or 69.3%, are also active in human eye.16  This figure is often quoted as deep sameness.  The paper does not support that use.  The authors report that 1,019 of the 1,052 were already present in the common ancestor of all bilaterian animals, and that more than 760 are kept in flies or worms, which have no camera eye.  They concluded:  "not all of these genes were specific to the camera eye."16  They also note that the share rose from 15% to 69.3% only because the human list grew from 3,809 to 13,303 sequences.16  The number measures how many genes all animals share.

How independent?  Walter Gehring took shared Pax-6 to mean eyes arose once.  Dan-Eric Nilsson answers that shared Pax-6 fits several origins too, and that the two light cell types took up their role separately.17  So "independent" depends on the level.  The switch and the light-catching opsin proteins are shared across animals.  The lens, the retina's layout and the wiring are not.

On a design reading this looks like two engineering teams given the same brief and different stock parts.  It does not look like one blueprint used twice.  Why Eyes Use the Light They Use covers the physics that narrows the brief.

Part 3: Echolocation and prestin

Toothed whales and the small bats called microbats find prey by echo.2  Baleen whales and the large fruit bats do not.  Prestin is the motor protein of the inner ear's outer hair cells.

The finding.  In 2010 Ying Li and colleagues built a tree from prestin protein sequences of 25 mammals.  The dolphin landed inside the echolocating bats, not beside the cow.2  They found 11 sites where the dolphin and a bat branch made the same change from the same starting amino acid, against a chance expectation far lower.2  Gareth Jones, reviewing this paper and a second by Yang Liu's group, put the total at "at least 14 derived amino acid sites."18  At position 7, every echolocator examined has threonine and every non-echolocator asparagine.2

The functional test.  In 2014 Zhen Liu and colleagues measured prestin's electrical behaviour in cultured cells.  Putting threonine at position 7 of fin whale and fruit bat prestin shifted one parameter toward the echolocators' value.  A second change, I384T, shifted another.19  So at least two shared changes do a job.  The same paper found that other shared changes pushed function the opposite way, and concluded that "the molecular mechanisms of convergent phenotypes are much more complicated than originally thought."19

The silent sites.  The genetic code has spare spellings.  Many DNA changes leave the protein unchanged.  Li's group built the tree again from those silent changes alone, and the dolphin went back beside the cow "with 100% bootstrap support."2  The likeness between bat and dolphin prestin is confined to changes that alter the protein.

The genome-wide claim.  In 2013 Joe Parker and colleagues scanned 2,326 genes in 22 mammals and reported "signatures consistent with convergence in nearly 200 loci."20  Sarfati built his case on it:  so many genes changing alike has a probability of "Practically zero."9

Two reanalyses in 2015 removed the claim.  Gregg Thomas and Matthew Hahn counted the shared changes directly.  A microbat and the dolphin share them in 1,372 genes.  A microbat and the cow, which has no sonar, share them in 1,951.21  Only 19 of Parker's top 117 genes had any such change.21  Zhengting Zou and Jianzhi Zhang found the same:  the signal between echolocators is "no stronger than those between echolocating and comparable nonecholocating lineages."22

What the rebuttals leave standing.  Zou and Zhang exempt seven hearing genes already known from single-gene studies, and say prestin has met both tests, statistical and functional.22  In 2019 Amir Marcovitz and colleagues agreed that echolocators have no more shared changes than other mammals.  But the changes they do share fall in 18 genes that build the cochlear ganglion more often than chance allows, and no such cluster appears in the control species.23  So the signal is small, real and in the ear.

Ewert's reply.  Ewert's 2023 paper reads prestin as carrying two echolocation modules.  He objects that convergence is invoked "not because convergence is known to be evolutionarily feasible, but because this preserves the presumed phylogenetic tree."24  He meets the silent-site result directly:  "an echolocation module has no reason to introduce synonymous mutations that are unlikely to optimize the protein for echolocation."24  That is a fair answer.  Its cost is that the module now consists only of the working amino acids, which is what selection predicts too.  The rest of the gene still groups dolphin with cow, and the model must give that to another module.

Part 4: The thylacine and the dogs

The thylacine, or Tasmanian tiger, was a pouched mammal.  The last known animal died in 1936.25  ReMine opens his case with it.6

Form.  Charles Feigin and colleagues measured 113 marsupial and placental skulls.  Diet predicted skull shape even after allowing for ancestry.  Convergence had closed about 34% of the shape distance between thylacine and red fox, and 40% between thylacine and grey wolf.25  So the skulls are alike, but well short of the same.  Douglass Rovinsky's group later found the closest match is not the wolf.  It is "mid-sized canids such as African jackals and South American ‘foxes’."26  The elbow is that of an ambush hunter, not a runner.27

Protein-coding genes.  The same 2018 paper read the thylacine genome.  Shared amino acid changes with dogs were no more common than between other mammal pairs at the same distance.25  Eighty-one genes carried such changes, and "none of the 81 homoplasious genes were found to be adaptively convergent."25

Regulatory DNA.  In 2019 Feigin, Axel Newton and Andrew Pask tested DNA switches active in the developing face.  They found 10,910 evolving fast in the thylacine, 1,923 in the wolf, and 339 in both.28  The title calls this "widespread" convergence.  The count does not show it.  If fast change in one species had nothing to do with fast change in the other, about 358 switches would be fast in both by chance.29  The paper's evidence is that the 339 sit near genes of two bone-shaping pathways.  "Fast in both" also means the same switch changed, not that it changed to the same sequence.

In 2025 the same laboratory tested the 339 in a reporter assay.  The preprint, not yet peer reviewed, concludes:  "shared gene regulatory potential is not a feature of these convergently accelerated regions in the thylacine and wolf."30

The wider set.  Textbook charts pair marsupials with placentals:  mole with mole, glider with flying squirrel.  Jonathan Losos notes the other half.  Australia's platypus, koala and kangaroo have no placental twin, though their habitats exist elsewhere.31  How marsupials came to be concentrated in Australia is a separate question, treated in Evolutionary Biogeography and Creationist Biogeography.

Part 5: Molecular cases, and what chance supplies

The sodium pump.  Milkweeds and toads make poisons, the cardiac glycosides, that jam the sodium pump.  Ying Zhen and colleagues sequenced the pump in 14 insect species that eat such plants.  Twenty-four of 33 changes fell between positions 111 and 122.  Position 122 changed from asparagine to histidine in five separate lineages.32  Susanne Dobler's group confirmed by experiment that the changes give resistance.  They also report that the pump is at least 78% identical between insects and vertebrates while its silent sites have diverged to saturation.33  Beata Ujvari and colleagues found the same small region changed in resistant insects, amphibians, reptiles and mammals:  "replacements of 2 of the 12 amino acids" at the poison's binding site.34  Monitor lizards in Africa and Asia, where toads live, carry the resistant form.  Monitors in toad-free Australia carry the susceptible one.34

Antifreeze.  Antarctic notothenioid fish and Arctic cod make "near-identical" antifreeze proteins.35  The Antarctic gene derives from a digestive enzyme gene, trypsinogen.36  The Arctic gene shares no sequence with trypsinogen, has a different layout, and spells the same three amino acids with different codons.35  In 2019 its source was traced to DNA that coded for nothing.37

C4 photosynthesis.  C4 is a carbon-concentrating add-on to ordinary photosynthesis.  Sage lists 62 lineages, centred on dry regions.4  In grasses the key enzyme was recruited at least eight times, with 21 amino acids converging.  Trees from the third codon positions and the introns matched the accepted grass tree.38

Lysozyme.  Cows, langur monkeys and the hoatzin bird digest leaves with stomach bacteria and use lysozyme to break them down.  Stewart, Schilling and Wilson reported convergence in 1987.  Jianzhi Zhang and Sudhir Kumar retested it in 1997 and found two sites, 75 and 87, changed in parallel beyond chance.  Elsewhere, "the parallel or convergent changes probably occurred just by chance."39

Hemoglobin.  Across 56 kinds of birds, those living high had blood that holds oxygen more tightly.  Few reached it by the same amino acid changes.  "Thus, predictable changes in biochemical phenotype do not have a predictable molecular basis."40

The chance baseline.  Each protein site has at most 20 possible states, so "a few convergent or parallel amino acid changes may simply arise by random chance."39  Zou and Zhang checked every shared protein of 12 fruit flies and 17 mammals and found "no genomic evidence for higher-than-neutral levels of molecular convergence."41  Andrew Foote's group found more shared changes among land mammals than among the sea mammals they were studying.42

When convergence misleads the tree.  Todd Castoe and colleagues found that whole mitochondrial genomes of snakes and agamid lizards support a wrong tree.  Of 113 convergent changes, about 44 were more than chance predicts.43  The error was caught because nuclear genes, anatomy and the silent sites disagreed:  fourfold-redundant positions gave "almost no differential likelihood support" to either tree.43

Part 6: What each reading predicts

Test Same module reused ReMine, Ewert Selection under constraint Found
Same genes behind the same trait? Yes Sometimes Only where few genes can do the job Mixed: yes for prestin and the sodium pump, no for antifreeze, lens and thylacine skull
Same substitutions? Yes At working sites At working sites, more often in close relatives At a handful of working sites
Silent sites and introns alike too? Expected if a sequence is copied No, per Ewert No No, in prestin, grass PEPC and snake mitochondria
Does the trait follow way of life or classification? Open Open Way of life Way of life: toads, dry climates, diet
Amount of convergence High Weaker than the nested pattern, per ReMine No firm figure Common in form, rare in sequence

The first column is the reading of Rana, Nelson and Wells, and the popular articles.  It fails rows one and three.  ReMine's and Ewert's versions pass every row, but only by agreeing with the fourth column wherever the data can check them.

Constraint and contingency.  Rana's and ReMine's charge is that evolution did not predict convergence.6  For Gould that is fair.  He held that replaying life's tape would give a different world each time:  "Replay the tape a million times from a Burgess beginning, and I doubt that anything like Homo sapiens would ever evolve again."44  Conway Morris holds the opposite:  "the evolutionary routes are many, but the destinations are limited."3  George McGhee agrees:  the "number of evolutionary pathways available to life is in fact not endless, but is quite restricted."45  Losos stands between, and explains why close relatives converge most:  they give selection "the same genetic systems on which to work."31  A 2018 review by Zachary Blount, Richard Lenski and Losos reaches the same middle:  repeatable among close relatives, less so as lineages diverge.46

A theory whose leading writers predicted both almost none and almost everything made no firm prediction.  The same holds for ReMine, whose designer uses the same design in some places and a different one in others.

All of these authors accept ages of millions of years.  That does not affect the comparisons used here, which are between living sequences and measured skulls.

Objections and responses

"Engineers reuse parts, so shared designs point to an engineer."  Then the shared designs should share parts.  The antifreeze genes, the lens proteins and the thylacine's proteins do not.  An engineer may also build the same function anew for each platform.  That reply is reasonable, but it gives up the prediction.

"Two hundred genes changing alike is statistically impossible."  The figure was withdrawn in effect in 2015.  Chance alone gives a bat and a cow shared changes in 1,951 genes.  Sarfati's article has not been revised since October 2013.

"A designer has no reason to match silent sites."  This is Ewert's answer and it is sound as far as it goes.  It leaves two things unexplained.  The silent sites of prestin group the dolphin with the cow, which is the ordinary tree.  And some shared prestin changes work against each other, which a purpose-built module does not predict.

"Imperfect convergence proves evolution."  Richard Dawkins and Conway Morris both argue this.63  Imperfect likeness shows separate solutions.  Separate solutions are also what separately created kinds, each with its own parts, would show.  The step to common descent needs the tree, which is judged in Phylogenetic Congruence.

"Convergence is an excuse used whenever a trait breaks the tree."  For DNA this is no longer true.  Silent sites, introns and function assays give a test that can fail, and Parker's claim failed it.  For anatomy the objection holds.  A fossil has no silent sites.  When a skeleton conflicts with the tree, calling the likeness convergent is a decision, not a measurement.  Whale Evolution meets this with the mesonychids.

"Shared Pax-6 shows eyes arose only once."  It shows a shared switch.  The switch builds a fly eye in a fly.

Issues and unanswered questions

  • No young-earth author has published an analysis of any case in this article. The one fitted design model, Ewert's, is from the intelligent-design side and assumes fixed species. Wood says the work on kinds is not done.
  • No baraminology paper was found that deals with convergence between kinds. Wood's 2002 tutorial does not mention it. The method groups animals by likeness, so a thylacine-and-jackal likeness is a problem it has to face.
  • Young-earth models accept selection within kinds. Much of the molecular evidence here, such as the monitor lizards, may sit inside one kind and so tests nothing between the two readings.
  • Why workable designs are so few is unexplained. Conway Morris himself senses "the ghost of teleology."3 This is the design argument's best ground here, but it is an argument about the laws that limit life, not about reuse. The published article Organisms Appear Designed takes it up.
  • Showing that an eye was built twice from different parts does not show how it was built once.
  • The overlap calculation in Part 4 uses fractions taken over slightly different sets. Shukla 2025 is a preprint.
  • Not read: Liu 2010 and Stewart 1987 in the original, and McGhee except in short extracts.

Verdict

Repeated arrival at the same complex solution is a fact.  Read closely, the cases show the same job done with different parts, and the same few amino acids changed where only those will do.  The DNA around them follows the ordinary tree.  That is better read as separate problem-solving under tight limits than as reuse of one design.  The simple reuse argument should be retired.  ReMine's and Ewert's versions survive, but they have not yet predicted anything here that selection does not.

The evolutionary reading wins the molecular cases and has not earned the rest.  For anatomy and fossils, convergence remains a name given after the tree is drawn.  Genetic Redundancy examines a related claim, that unrelated genes come to share one task.

Sources

  1. 1.

    Rana, Fazale.  The Cell's Design: How Chemistry Reveals the Creator's Artistry.  Baker Books.  2008.  Rana is a biochemist at Reasons to Believe, an old-earth creationist ministry.  Read on archive.org, item cellsdesignhowch0000rana, by paragraph search of a print-disabled scan;  not stored.  Printed page is leaf minus 2, from the scan's own page map, checked at leaves 34 and 216.  Copyright.

    1. 1a.

      Page 204:  “If human craftsmen made in God’s image reuse the same techniques and technologies, it’s reasonable to infer that their Creator would do the same. So, if life stems from his hand, then it’s reasonable to expect the same designs to repeatedly appear throughout nature.” ↩ ↩

    2. 1b.

      Page 204:  “While repeated occurrences of biochemical designs logically point to a Creator, that’s not the case for evolutionary processes. If biochemical systems are the product of evolution, then the same biochemical designs should not recur throughout nature.” ↩

    3. 1c.

      Page 206:  The five kinds are functional, mechanistic, structural, sequence and systemic convergence.  Page 215:  “The explosion in the number of examples of molecular convergence is unexpected if life results from historical sequences of chance evolutionary events. Yet, if life emanates from a Creator, it’s reasonable to expect he would use the same designs repeatedly.” ↩

  2. 2.

    Li, Ying et al.  “The hearing gene Prestin unites echolocating bats and whales.”  Current Biology 20(2):R55–R56.  Jan 26, 2010.  Read as the NIH author manuscript, which has no page breaks;  the printed paper is two pages.  Paragraphs are counted from the start of the body text.  Copyright.  Mirror:  PMC

    1. 2a.

      Paragraph 1:  “Only microbats and toothed whales have acquired sophisticated echolocation”. ↩

    2. 2b.

      Paragraph 3:  “Surprisingly, all methods group the dolphin within bats”. ↩

    3. 2c.

      Paragraph 4:  “Indeed, the same misplacement of dolphin is observed in the Prestin tree reconstructed with only nonsynonymous nucleotide substitutions (Fig. S1B); but, when only synonymous substitutions are used, dolphin and cow are correctly grouped with 100% bootstrap support (Fig. S1C).” ↩ ↩

    4. 2d.

      Paragraph 6:  “We identified no convergent site, but 4, 1, and 6 parallel sites between the red branch and branches 1, 2, and 3, respectively”.  The red branch is the dolphin's.  A parallel change starts from the same amino acid in both lineages;  a convergent change, in the paper's strict sense, starts from different ones. ↩

    5. 2e.

      Paragraph 7:  “This site has a Thr in all echolocating mammals but an Asn in all non-echolocating mammals examined so far.”  “Although not all parallel changes identified may be necessary for echolocation, as some have reverted in some microbats or occurred also in some nonecholocating mammals”. ↩

  3. 3.

    Conway Morris, Simon.  Life's Solution: Inevitable Humans in a Lonely Universe.  Cambridge University Press.  2003.  Conway Morris is a Cambridge palaeontologist who accepts common descent.  Read on archive.org, item lifessolutionine01conw, by paragraph search of a print-disabled scan;  not stored.  Printed page is leaf minus 24, from the scan's page map;  pages 128 and 145 also match the pages at which these sentences are commonly cited.  The scan drops superscript note numbers.  Copyright.

    1. 3a.

      Section Preface:  “the camera-like eye has evolved independently at least six times”.  The preface leaf carries no folio in the scan's page map. ↩

    2. 3b.

      Page 157:  “The hallmark of the animals that evolved camera-eyes, be they squid, vertebrate, heteropod snail, alciopid polychaete, cubozoan jellyfish, or even spider, seems to be that all are active, mobile, and typically predatory.” ↩

    3. 3c.

      Page 151:  “Most notable are those between the relative position and detailed structure of the light-sensitive layer, the retina, which arise as a result of the different embryologies in vertebrates and molluscs. In the molluscs the retina is derived from the outer layer (known as the ectoderm, and effectively equivalent to our skin).” ↩

    4. 3d.

      Page 145:  “the evolutionary routes are many, but the destinations are limited”. ↩

    5. 3e.

      Page 128:  “and this in itself is as concrete a piece of evidence for the reality of evolution as can be provided. Even so, the frequency of adjectival surprise associated with descriptions of convergence suggests to me that there is almost a feeling of unease in these similarities. Indeed, I strongly suspect that some of these biologists sense the ghost of teleology looking over their shoulders.”  "This" is the fact that convergent similarities are seldom precise. ↩ ↩

  4. 4.

    Sage, Rowan F. et al.  “The C4 plant lineages of planet Earth.”  Journal of Experimental Botany 62(9):3155–3169.  2011.  Read from the publisher's page.  Copyright.

    1. 4a.

      Section Abstract:  “Here, 62 recognizable lineages of C4 photosynthesis are listed.”  “These centres tend to cluster in areas corresponding to what are now arid to semi-arid regions”.  The count treats each lineage as separate "on the assumption that each lineage arose from ancestors that were fully functional C3 species." ↩ ↩

  5. 5.

    Nelson, Paul, and Jonathan Wells.  “Homology in Biology: Problem for Naturalistic Science and Prospect for Intelligent Design.”  Darwinism, Design, and Public Education, 303–322.  Michigan State University Press.  Edited by John Angus Campbell and Stephen C. Meyer.  2003.  Read on archive.org, item darwinismdesignp0000unse, by paragraph search;  not stored.  Printed page is leaf minus 42, from the scan's page map, checked at leaves 348 and 360.  Copyright.

    1. 5a.

      Page 316:  “An intelligent cause may reuse or redeploy the same module in different systems, without there necessarily being any material or physical connection between those systems. Even more simply, intelligent causes can generate identical patterns independently”. ↩

  6. 6.

    ReMine, Walter James.  The Biotic Message: Evolution Versus Message Theory.  St. Paul Science.  1993.  Read on archive.org, item bioticmessageevo0000remi, by paragraph search;  the library holds other chapters only.  Printed page is leaf minus 8, from the scan's page map and confirmed against the book's contents page, which lists Convergence at 141 and Balance Between Patterns at 363.  Copyright.

    1. 6a.

      Page 141:  “A marsupial tasmanian wolf (a recently extinct animal from Australia) is remarkably like a dog, yet this similarity was not inherited from a common ancestor.” ↩

    2. 6b.

      Page 168:  “Evolutionary theory does not predict convergence. Evolutionists merely claim that extensive convergence is virtually inevitable.” ↩

    3. 6c.

      Page 349:  “the convergence is not total. The different lines of evolution betray their independent origins in numerous points of detail.”  ReMine quoting Richard Dawkins, The Blind Watchmaker, 1986, pages 94–95, on octopus eyes.  The original was not checked. ↩

    4. 6d.

      Page 351:  “This situation requires a delicate balance. If characters are too similar then they look like transposition, but if they are too different then they fail to demand special (and formidable) rationalizations like the evolutionists’ convergence explanation. This delicate balance requires a designer.” ↩

    5. 6e.

      Page 354:  “To successfully send the biotic message, a designer must not indiscriminately use the same design repeatedly. Therefore, the designer is forced to sometimes use different designs to perform the same function.” ↩

    6. 6f.

      Page 363:  “For the biotic message to be perceived: (1) the cladistic and phenetic patterns should be strong and durable; (2) the pattern of convergence must be comparatively weaker; (3) a widespread transposition pattern should be minor at most.” ↩

  7. 7.

    Ewert, Winston.  “The Dependency Graph of Life.”  BIO-Complexity 2018(3):1–27.  2018.  Intelligent-design journal.  CC BY.

    1. 7a.

      Page 1:  “We present the dependency graph hypothesis as an alternative explanation, based on the technique used by software developers to reuse code among different software projects. This hypothesis postulates that different biological species share modules related by a dependency graph.” ↩

    2. 7b.

      Page 3:  “All marine species depend on a marine module, and the echolocating species depend on the echolocation module.” ↩

  8. 8.

    UpChurch, John.  “Convergent Evolution or Common Designer?”  Answers.  Oct 1, 2016.  Retrieved Sep 29, 2026.  Answers in Genesis magazine, October–December 2016 issue.  Copyright.

    1. 8a.

      Section The Master Engineer:  “Human designers do the same thing—integrating similar parts in different things they design.” ↩

  9. 9.

    Sarfati, Jonathan.  “Echolocation ‘evolved in the same way’.”  creation.com.  Oct 3, 2013.  Retrieved Sep 29, 2026.  Creation Ministries International.  The page gives both dates:  published 1 February 2011, updated 3 October 2013.  Copyright.

    1. 9a.

      Section Common designer:  “The observed uniformity is consistent with a particular subset of intelligent design: the biotic message theory, as proposed by Walter ReMine.” ↩

    2. 9b.

      Section Non-evolution of echolocation:  “Second, it is remarkable that so many genes ‘independently changed’ in the same way. Statistically, what are the chances or probability of this happening in the correct manner? Practically zero.” ↩

  10. 10.

    Guliuzza, Randy J.  “Major Evolutionary Blunders: Convergent Evolution Is a Seductive Intellectual Swindle.”  Acts & Facts 46(3).  2017.  Retrieved Sep 29, 2026.  Institute for Creation Research.  Copyright.

    1. 10a.

      Section A Better Organism-Focused, Design-Based Explanation:  “Convergence is not an observation demonstrated to flow from objectively discernible causes but is a declaration based on mental pictures”. ↩

  11. 11.

    Wood, Todd Charles.  “Ewert's dependency graph paper.”  Todd's Blog.  Jul 23, 2018.  Wood is a young-earth creationist and the main developer of statistical baraminology.  Read from the saved page in the source library.  Copyright.

    1. 11a.

      Paragraph 4:  “First of all, this is not a paradigm-breaking, epoch-making breakthrough. It's a good step, and definitely one possible response to the evolutionary tree of life.”  Paragraph 1:  “I have yet to publish this work, mostly because I've been dissatisfied with the projects.” ↩

  12. 12.

    Gavriouchkina, Daria et al.  “A single-cell atlas of the bobtail squid visual and nervous system highlights molecular principles of convergent evolution.”  Nature Ecology & Evolution 9(7):1245–1262.  2025.  CC BY 4.0.  Mirror:  PMC

    1. 12a.

      Section Introduction:  “In cephalopods, light reaches photoreceptors directly; by contrast, the vertebrate retina is ‘inverted’, so that light must traverse several cell layers before reaching photoreceptors. Cephalopods have a single type of rhabdomeric photoreceptor along with pigmented ‘support’ cells”.  “while the vertebrate retina includes photoreceptors as well as additional neuronal cell types that process the visual signal before transmission to higher brain centres”.  Section Results:  “ciliary photoreceptors of the vertebrate retina”. ↩

    2. 12b.

      Section Results:  “secrete the lens composed of two classes of crystallins of independent phylogenetic origins: cephalopod-specific glutathione-S-transferase-derived lens proteins (σ-crystallins)”. ↩

  13. 13.

    Tan, Wei-Hung et al.  “Structure of a Highly Active Cephalopod S-crystallin Mutant: New Molecular Evidence for Evolution from an Active Enzyme into Lens-Refractive Protein.”  Scientific Reports 6:31176.  2016.  CC BY 4.0.  Mirror:  PMC

    1. 13a.

      Section Abstract:  “S-crystallin has a preference for glutathione binding, although almost lost its GST enzymatic activity.”  Section Introduction:  “S-crystallin in the cephalopod lens has been shown to have an amino acid sequence similar to that of glutathione S-transferase (GST)”. ↩

  14. 14.

    Tomarev, Stanislav I. et al.  “Squid Pax-6 and eye development.”  Proceedings of the National Academy of Sciences 94(6):2421–2426.  1997.  Read from PubMed Central.  Copyright.  Mirror:  PMC

    1. 14a.

      Section Abstract:  “Despite significant sequence differences between squid Pax-6 and Drosophila eyeless in the region outside the paired- and homeodomains, squid Pax-6 is able to induce the formation of ectopic eyes in Drosophila.” ↩

  15. 15.

    Yoshida, Masa-aki et al.  “Cephalopod eye evolution was modulated by the acquisition of Pax-6 splicing variants.”  Scientific Reports 4:4256.  2014.  CC BY-NC-SA 3.0.  Mirror:  PMC

    1. 15a.

      Section Abstract:  “We examined Pax-6 variations in the squid and found five types of Pax-6 splicing variants but no duplication of the Pax-6 gene.”  “Our study suggests that cephalopods acquired Pax-6 splicing variants independent of those in vertebrates”. ↩

  16. 16.

    Ogura, Atsushi et al.  “Comparative Analysis of Gene Expression for Convergent Evolution of Camera Eye Between Octopus and Human.”  Genome Research 14(8):1555–1561.  2004.  Read from PubMed Central.  The abstract gives 875 genes conserved between human and octopus;  the Discussion gives 879.  Copyright.  Mirror:  PMC

    1. 16a.

      Section Abstract:  “Comparing these 1052 genes with 13,303 already-known ESTs of the human eye, 729 (69.3%) genes were commonly expressed between the human and octopus eyes.” ↩

    2. 16b.

      Section Abstract:  “We found that 1019 out of the 1052 genes had already existed at the common ancestor of bilateria”.  Section Discussion, Had the Origin of the Gene Set for the Camera Eye…:  “As we estimated 1019 genes for ancestral genes for the camera eye, more than 760 genes were conserved in flies or worms. This conservation in flies or worms suggests that not all of these genes were specific to the camera eye.” ↩

    3. 16c.

      Section Discussion, Similarity in Gene Expression of Human and Octopus Eyes:  “Note that the increase of gene expression similarities between human and octopus eyes from 15% (162/1052) to 69.3% (729/1052) is caused by the increase of the EST data set of human eye from only 3809 ESTs in the database of BodyMap to 13,303 human-eye ESTs”. ↩

  17. 17.

    Nilsson, Dan-Eric.  “Eye evolution and its functional basis.”  Visual Neuroscience 30(1–2):5–20.  2013.  Copyright.  Mirror:  PMC

    1. 17a.

      Section The evolution and fate of rhabdomeric photoreceptors in bilateria:  “has been taken as evidence that eyes only evolved once”.  Section The association of microvilli and cilia with r- and c-opsins:  “It thus seems that directional photoreception, with screening pigment association and membrane stacking, evolved independently from different class I receptors with c-opsin and r-opsin, respectively.” ↩

  18. 18.

    Jones, Gareth.  “Molecular Evolution: Gene Convergence in Echolocating Mammals.”  Current Biology 20(2):R62–R64.  2010.  A commentary on Li 2010 and on Liu, Cotton, Shen, Han, Rossiter and Zhang 2010 in the same issue, which was not obtained.  Copyright.  Mirror:  Websites.umich.edu

    1. 18a.

      Page R62:  “dolphins and porpoises share at least 14 derived amino acid sites in prestin with echolocating bats, including 10 shared with the highly specialised CF bats”. ↩

  19. 19.

    Liu, Zhen et al.  “Parallel Sites Implicate Functional Convergence of the Hearing Gene Prestin among Echolocating Mammals.”  Molecular Biology and Evolution 31(9):2415–2424.  2014.  Read from the publisher's page.  Copyright.

    1. 19a.

      Section Results, The Parallel Change of I384T:  “These results provide the first direct evidence that the parallel sites of N7T and I384T are responsible for the functional convergence in the parameters 1/α and V1/2 among echolocating mammals”. ↩

    2. 19b.

      Section Discussion:  “During this study, we also observed that some parallel sites (except for N7T and I384T) led to the functional divergence in prestin among echolocating mammals”.  “the molecular mechanisms of convergent phenotypes are much more complicated than originally thought”. ↩

  20. 20.

    Parker, Joe et al.  “Genome-wide signatures of convergent evolution in echolocating mammals.”  Nature 502(7470):228–231.  2013.  Read as the author manuscript at PubMed Central.  Copyright.  Mirror:  PMC

    1. 20a.

      Section Abstract:  “Systematic analyses of convergent sequence evolution in 805,053 amino acids within 2,326 orthologous coding gene sequences compared across 22 mammals (including four new bat genomes) revealed signatures consistent with convergence in nearly 200 loci.” ↩

  21. 21.

    Thomas, Gregg W. C., and Matthew W. Hahn.  “Determining the Null Model for Detecting Adaptive Convergence from Genomic Data: A Case Study using Echolocating Mammals.”  Molecular Biology and Evolution 32(5):1232–1236.  2015.  Copyright.  Mirror:  PMC

    1. 21a.

      Section main text:  “We examined convergent substitutions in our data set and found 1,372 genes with convergent changes between microbat and dolphin and 1,951 genes between microbat and the nonecholocating cow.” ↩

    2. 21b.

      Section main text:  “we found that only 19 of the 117 loci identified by Parker et al. as convergent using ΔSSLS actually had convergent substitutions in our data set”. ↩

  22. 22.

    Zou, Zhengting, and Jianzhi Zhang.  “No Genome-Wide Protein Sequence Convergence for Echolocation.”  Molecular Biology and Evolution 32(5):1237–1241.  2015.  Copyright.  Mirror:  PMC

    1. 22a.

      Section Abstract:  “Here, we show that genomic signatures of molecular convergence between echolocating lineages are generally no stronger than those between echolocating and comparable nonecholocating lineages.” ↩

    2. 22b.

      Section main text, final paragraph:  “This conclusion, however, does not preclude the possibility that the convergent substitutions previously identified from the case studies of a few hearing proteins are important for echolocation.”  “in the absence of the seven hearing genes known to be subject to convergent evolution”. ↩

  23. 23.

    Marcovitz, Amir et al.  “A functional enrichment test for molecular convergent evolution finds a clear protein-coding signal in echolocating bats and whales.”  Proceedings of the National Academy of Sciences 116(42):21094–21103.  2019.  Copyright.  Mirror:  PMC

    1. 23a.

      Section Abstract:  “Like others, we find that convergent amino acid substitutions are not more abundant in echolocating mammals compared to their outgroups.”  “affecting 18 genes, regulates development of the cochlear ganglion”.  “Conversely, when comparing to nonecholocating outgroups, no significant gene set enrichment exists.” ↩

  24. 24.

    Ewert, Winston.  “AminoGraph Analysis of the Auditory Protein Prestin From Bats and Whales Reveals a Dependency-Graph Signal That Is Missed by the Standard Convergence Model.”  BIO-Complexity 2023(1):1–15.  2023.  CC BY.

    1. 24a.

      Page 1:  “Conventional analyses interpret this as convergence, not because convergence is known to be evolutionarily feasible, but because this preserves the presumed phylogenetic tree.” ↩

    2. 24b.

      Page 12:  “In particular, an echolocation module has no reason to introduce synonymous mutations that are unlikely to optimize the protein for echolocation. Rather, we should expect an echolocation module to only modify the gene in ways that contribute to echolocation.” ↩

  25. 25.

    Feigin, Charles Y. et al.  “Genome of the Tasmanian tiger provides insights into the evolution and demography of an extinct marsupial carnivore.”  Nature Ecology & Evolution 2:182–192.  2018.  Pages checked one at a time against the PDF, whose footer prints the range 182–192.  Copyright.

    1. 25a.

      Page 182:  “The last known thylacine died in captivity in 1936”. ↩

    2. 25b.

      Page 184:  “diet was a significant predictor of cranial shape”.  “we estimate that approximately 34% of the ancestral phenotypic space between the thylacine and red fox has been closed by convergent evolution and 40% has been closed between the thylacine and grey wolf”. ↩

    3. 25c.

      Page 185:  “the prevalence of homoplasy between the thylacine and canids did not defy expectations based on neutral evolution and was comparable to pairwise comparisons between non-convergent species with similar phylogenetic distances”. ↩

    4. 25d.

      Page 186:  “Interestingly, none of the 81 homoplasious genes were found to be adaptively convergent in the thylacine and canids.” ↩

  26. 26.

    Rovinsky, Douglass S. et al.  “Functional ecological convergence between the thylacine and small prey-focused canids.”  BMC Ecology and Evolution 21:58.  2021.  CC BY 4.0.  Mirror:  PMC

    1. 26a.

      Section Abstract, Conclusions:  “Our study finds little support for a wolf/dog-like functional ecology in the thylacine, with it instead being most similar to mid-sized canids such as African jackals and South American ‘foxes’ that mainly take prey less than half their size.” ↩

  27. 27.

    Figueirido, Borja, and Christine M. Janis.  “The predatory behaviour of the thylacine: Tasmanian tiger or marsupial wolf?”  Biology Letters 7(6):937–940.  2011.  Copyright.  Mirror:  PMC

    1. 27a.

      Section Abstract:  “Here we show that thylacines had the elbow morphology typical of an ambush predator”. ↩

  28. 28.

    Feigin, Charles Y. et al.  “Widespread cis-regulatory convergence between the extinct Tasmanian tiger and gray wolf.”  Genome Research 29(10):1648–1658.  2019.  CC BY-NC 4.0.  Mirror:  PMC

    1. 28a.

      Section Results:  “Analysis of evolutionary rates identified 10,910 thylacine accelerated regions (TARs) and 1923 wolf accelerated regions (WARs; FDR = 0.1)”.  “Three hundred thirty-nine of these elements showed accelerated evolution in both species”. ↩

    2. 28b.

      Section Methods, Identification of vertebrate-conserved elements:  “In total, 40,547 thylacine and 56,278 wolf orthologs of pcfCREs were represented.”  Section Methods, Characterization of TWARs by proximity to protein-coding genes:  “the 38,993 pcfCREs with representative and thylacine and wolf sequences”. ↩

  29. 29.

    A calculation made for this article from the paper's own counts.28  The fast share is 10,910 of 40,547 thylacine elements, or 26.9%, and 1,923 of 56,278 wolf elements, or 3.4%.  Of the 38,993 elements present in both, independence predicts 38,993 × 0.269 × 0.0342 = 358 fast in both.  The paper reports 339.  The two shares are taken over each species' full set and not over the 38,993, which the paper does not tabulate, so the figure is approximate.  The paper does not test the overlap count. ↩

  30. 30.

    Shukla, Navya et al.  “Dissecting functional regulatory convergence over 160 million years of therian evolution.”  bioRxiv.  Jul 11, 2025.  A preprint, not peer reviewed;  no published version was listed on 29 September 2026.  Its pages carry printed line numbers, removed from the text the quotation was checked against.  CC BY-NC 4.0.

    1. 30a.

      Page 2:  “Ultimately our findings suggest that shared gene regulatory potential is not a feature of these convergently accelerated regions in the thylacine and wolf.” ↩

  31. 31.

    Losos, Jonathan B.  Improbable Destinies: Fate, Chance, and the Future of Evolution.  Riverhead Books.  2017.  Read on archive.org, item improbabledestin0000loso, by paragraph search of a print-disabled scan;  not stored.  Printed page is leaf minus 20, from the scan's page map, checked at leaves 66 and 300.  Copyright.

    1. 31a.

      Page 48:  “I’ve already talked about Australia’s penchant for evolutionary one-offs—the platypus, koala, and kangaroo leading the way.”  Page 88:  “So where’s our platypus doppelganger?” ↩

    2. 31b.

      Page 98:  “Close relatives tend to be similar genetically, so selection is likely to have the same genetic systems on which to work.” ↩

  32. 32.

    Zhen, Ying et al.  “Parallel Molecular Evolution in an Herbivore Community.”  Science 337(6102):1634–1637.  2012.  Read as the NIH author manuscript.  Copyright.  Mirror:  PMC

    1. 32a.

      Section main text:  “Also, position 122 has a replacement of asparagine by histidine in five independent lineages.”  “24 of 33 substitutions occur at positions 111 through 122”. ↩

  33. 33.

    Dobler, Susanne et al.  “Community-wide convergent evolution in insect adaptation to toxic cardenolides by substitutions in the Na,K-ATPase.”  Proceedings of the National Academy of Sciences 109(32):13040–13045.  2012.  Copyright.  Mirror:  PMC

    1. 33a.

      Section Results:  “This gene region is highly conserved, with a minimum of 78% amino acid identity between the vertebrate and insect sequences”.  “Synonymous substitutions, on the other hand, had reached saturation, with a proportion of 0.942 between pig and insects”. ↩

  34. 34.

    Ujvari, Beata et al.  “Widespread convergence in toxin resistance by predictable molecular evolution.”  Proceedings of the National Academy of Sciences 112(38):11911–11916.  2015.  Copyright.  Mirror:  PMC

    1. 34a.

      Section Abstract:  “is mediated by similar molecular changes to the sodium-potassium-pump (Na+/K+-ATPase) in insects, amphibians, reptiles, and mammals”.  “Importantly, resistance in all taxa is mediated by replacements of 2 of the 12 amino acids comprising the Na+/K+-ATPase H1–H2 extracellular domain that constitutes a core part of the cardiac glycoside binding site.” ↩

    2. 34b.

      Section Results and Discussion:  “While basal Afro-Asian varanids possess the resistant genotype (L111 and R120), the Australian varanids are characterized by the susceptible genotype (Q111 and G120)”.  The authors read this as resistance lost in Australia, and grant that two separate gains "appears similarly parsimonious." ↩

  35. 35.

    Chen, Liangbiao et al.  “Convergent evolution of antifreeze glycoproteins in Antarctic notothenioid fish and Arctic cod.”  Proceedings of the National Academy of Sciences 94(8):3817–3822.  1997.  Copyright.  Mirror:  PMC

    1. 35a.

      Section Abstract:  “Antarctic notothenioid fishes and several northern cods are phylogenetically distant (in different orders and superorders), yet produce near-identical antifreeze glycoproteins (AFGPs) to survive in their respective freezing environments.” ↩

    2. 35b.

      Section Abstract:  “Arctic cod AFGP genes share no sequence identity with the trypsinogen gene, indicating trypsinogen is not the progenitor.”  “the AFGP genes of the two fish have different intron–exon organizations and different spacer sequences”.  “coding sequences are drastically different in the two groups of genes”. ↩

  36. 36.

    Chen, Liangbiao et al.  “Evolution of antifreeze glycoprotein gene from a trypsinogen gene in Antarctic notothenioid fish.”  Proceedings of the National Academy of Sciences 94(8):3811–3816.  1997.  Copyright.  Mirror:  PMC

    1. 36a.

      Section Abstract:  “We have found that the antifreeze glycoproteins (AFGPs) of the predominant Antarctic fish taxon, the notothenioids, evolved from a pancreatic trypsinogen.” ↩

  37. 37.

    Zhuang, Xuan et al.  “Molecular mechanism and history of non-sense to sense evolution of antifreeze glycoprotein gene in northern gadids.”  Proceedings of the National Academy of Sciences 116(10):4400–4405.  2019.  CC BY-NC-ND.

    1. 37a.

      Page 4400:  “Here, we report clear evidence and a detailed molecular mechanism for the de novo formation of the northern gadid (codfish) antifreeze glycoprotein (AFGP) gene from a minimal noncoding sequence.” ↩

  38. 38.

    Christin, Pascal-Antoine et al.  “C4 Photosynthesis Evolved in Grasses via Parallel Adaptive Genetic Changes.”  Current Biology 17(14):1241–1247.  2007.  Read from the publisher's page.  Copyright.

    1. 38a.

      Section Summary:  “we showed that grass C4 PEPCs appeared at least eight times independently from the same non-C4 PEPC. Twenty-one amino acids evolved under positive selection and converged to similar or identical amino acids in most of the grass C4 PEPC lineages.”  Section Results and Discussion, Congruence between Gene and Species Trees…:  “species relationships deduced from third positions and intron topology (Figure 2) were congruent with accepted species trees”. ↩

  39. 39.

    Zhang, Jianzhi, and Sudhir Kumar.  “Detection of convergent and parallel evolution at the amino acid sequence level.”  Molecular Biology and Evolution 14(5):527–536.  1997.  A scanned PDF;  pages read from its running heads.  It reanalyses Stewart, Schilling and Wilson, Nature 330:401–404, 1987, which was not obtained.  Copyright.

    1. 39a.

      Page 527:  “a few convergent or parallel amino acid changes may simply arise by random chance, as protein sequence evolution is a stochastic process with at most 20 possible states at each site”. ↩

    2. 39b.

      Page 534:  “two sites (sites 75 and 87) have experienced parallel substitutions in the evolutionary lineages of the foregut fermenters (cow, langur, and hoatzin)”.  “So, positive selection is not necessary to explain the evolution of stomach lysozymes except for sites 75 and 87. At some sites, the parallel or convergent changes probably occurred just by chance.” ↩

  40. 40.

    Natarajan, Chandrasekhar et al.  “Predictable convergence in hemoglobin function has unpredictable molecular underpinnings.”  Science 354(6310):336–339.  2016.  Read as the NIH author manuscript.  Copyright.  Mirror:  PMC

    1. 40a.

      Section Abstract:  “Convergent increases in hemoglobin-oxygen affinity were pervasive among high-altitude taxa, but few such changes were attributable to parallel amino acid substitutions at key residues. Thus, predictable changes in biochemical phenotype do not have a predictable molecular basis.” ↩

  41. 41.

    Zou, Zhengting, and Jianzhi Zhang.  “Are Convergent and Parallel Amino Acid Substitutions in Protein Evolution More Prevalent Than Neutral Expectations?”  Molecular Biology and Evolution 32(8):2085–2096.  2015.  Copyright.  Mirror:  PMC

    1. 41a.

      Section Abstract:  “Together, our study finds no genomic evidence for higher-than-neutral levels of molecular convergence”. ↩

  42. 42.

    Foote, Andrew D. et al.  “Convergent evolution of the genomes of marine mammals.”  Nature Genetics 47(3):272–275.  2015.  Copyright.  Mirror:  PMC

    1. 42a.

      Section Abstract:  “However, we found higher levels of convergent amino acid substitutions in a control set of terrestrial sister taxa to the marine mammals.” ↩

  43. 43.

    Castoe, Todd A. et al.  “Evidence for an ancient adaptive episode of convergent molecular evolution.”  Proceedings of the National Academy of Sciences 106(22):8986–8991.  2009.  Copyright.  Mirror:  PMC

    1. 43a.

      Section Abstract:  “We estimate that ≈44 of 113 predicted convergent changes distributed across all 13 mitochondrial protein-coding genes are expected to have arisen from nonneutral causes—a remarkably large number.” ↩

    2. 43b.

      Section Results, Site-Specific Support for the 2 Topologies:  “Four-fold redundant third codon positions, which do not alter the amino acid sequence when they change, provide almost no differential likelihood support between the 2 topologies”. ↩

  44. 44.

    Gould, Stephen Jay.  Wonderful Life: The Burgess Shale and the Nature of History.  W. W. Norton.  1989.  The library text has no page numbers.  Page 289 was fixed from the archive.org scan wonderfullifebur00goul, whose page map puts the sentence on leaf 291, printed page 289.  Copyright.

    1. 44a.

      Page 289:  “Replay the tape a million times from a Burgess beginning, and I doubt that anything like Homo sapiens would ever evolve again.” ↩

  45. 45.

    McGhee, George R. Jr.  Convergent Evolution: Limited Forms Most Beautiful.  MIT Press.  2011.  Not on archive.org.  Read only in Google Books search extracts, volume QwDSr1qdqXUC, which print the page with each extract.  Copyright.

    1. 45a.

      Page xi:  “number of evolutionary pathways available to life is in fact not endless, but is quite restricted”. ↩

  46. 46.

    Blount, Zachary D. et al.  “Contingency and determinism in evolution: Replaying life's tape.”  Science 362(6415):eaam5979.  2018.  Copyright.

    1. 46a.

      Page 2:  “On balance, the evidence indicates that evolution tends to be surprisingly repeatable among closely related lineages, but disparate outcomes become more likely as the footprint of history grows deeper.”  Page 2 of the PDF, which is the article's abstract page. ↩