Irreducible Complexity: The Bacterial Flagellum and Its Critics

In 1996 a Lehigh University biochemist, Michael Behe, published Darwin's Black Box and gave the modern design movement its best-known phrase: irreducible complexity. His favorite example was the bacterial flagellum, a whip-like tail that many bacteria spin like a propeller to swim. Behe argued that some molecular machines need all of their parts at once to work, so they could not have been built up one small, useful step at a time. A designing mind, he suggested, is the better explanation.

The flagellum became famous far beyond microbiology, even in a federal courtroom in Pennsylvania in 2005. This article sets out what Behe claimed, what his critics replied, how he answered, and what the evidence now supports. HonestApologist.com argues that the world is the work of a wise Creator, but accuracy comes before apologetic convenience, so we rate each claim honestly (see how we work).

In brief

Behe defined an irreducibly complex system as one of several well-matched parts in which removing any one part makes the system "effectively cease functioning," and he argued that such systems could not arise by gradual Darwinian steps. Critics such as the cell biologist Kenneth Miller, Mark Pallen and Nicholas Matzke answered that the flagellum is not irreducible in that sense: a subset of its proteins forms a working secretion system (the type III secretion system), many of its other parts have relatives elsewhere in the cell, and only about half of the standard set of proteins seems to be universally necessary. They also argued that evolution often re-uses parts for new jobs (co-option or exaptation), which Behe's definition does not rule out. Design advocates replied, with some support from later research, that the secretion system probably descended from the flagellum rather than the reverse, and that no detailed, tested step-by-step pathway to the flagellum has yet been established. Our assessment: the strict argument (remove one part and it fails, so it cannot have evolved) has not held up; the more modest claim, that integrated molecular machines still lack detailed evolutionary explanations, is fair but is an argument from present ignorance that cannot by itself establish design.

Thesis: The flagellum remains a striking piece of engineering that is open to a design reading, but irreducible complexity, as Behe defined it, does not prove that it could not have evolved, and honest Christians should say so.

What Behe argued

Behe began from a test that Darwin himself proposed. In On the Origin of Species (1859) Darwin wrote: "If it could be demonstrated that any complex organ existed, which could not possibly have been formed by numerous, successive, slight modifications, my theory would absolutely break down."¹ Behe believed that biochemistry, unknown in Darwin's day, had supplied such cases.

His definition, as quoted by the federal court in 2005, reads:

By irreducibly complex I mean a single system which is composed of several well-matched, interacting parts that contribute to the basic function, wherein the removal of any one of the parts causes the system to effectively cease functioning. An irreducibly complex system cannot be produced directly by slight, successive modifications of a precursor system, because any precursor to an irreducibly complex system that is missing a part is by definition nonfunctional.²

To make the idea vivid he used an ordinary spring mousetrap: "If any one of the components of the mousetrap (the base, hammer, spring, catch, or holding bar) is removed, then the trap does not function."³ His main biological examples were the bacterial flagellum, the blood-clotting cascade and parts of the immune system.⁴

The flagellum suited the argument well. In the gut bacterium Salmonella it has a basal body anchored in the cell envelope, a rotor and stator that turn using the flow of ions across the membrane, a flexible hook, and a long filament, with dozens of proteins assembled in a set order. Scott Minnich, a microbiologist who testified for the defense at Dover, and Stephen Meyer described it as "Comprised of 40 parts that self-assemble into a true rotary engine."⁵ The cell contains other rotary machines too, such as ATP synthase.

Behe did not claim that intelligent design tells us how or when design happened; asked at trial about common descent, he said design "does not say how such a thing might have happened."⁶ He also located the positive case for design elsewhere: "The basis for the biochemical argument for design is the purposeful arrangement of parts. Irreducible complexity shows the difficulties for Darwinian processes in trying to explain these things."⁶

The critics' first reply: the flagellum is not irreducible

The best-known reply came from Kenneth Miller of Brown University, a Roman Catholic biologist, like Behe himself (Behe confirmed at the Dover trial that the two share the same faith).⁷ Miller pointed to the type III secretion system (often abbreviated T3SS or TTSS), the molecular syringe that some disease-causing bacteria use to inject proteins into host cells. Its core proteins are related to the proteins at the base of the flagellum that export flagellar parts. In Miller's words, "a smaller subset of the full complement of proteins in the flagellum makes up the functional transmembrane portion of the TTSS."⁸

Behe's definition said that a precursor "missing a part is by definition nonfunctional." The secretion system is missing most of the flagellum's parts, yet it works, so the definition, taken strictly, fails for its flagship example. Miller was candid about the limit of his own argument. He acknowledged the objection "that the TTSS does not tell us how either it or the flagellum evolved," and replied, "This is certainly true."⁸ His claim was narrower: the flagellum is not irreducible in Behe's sense.

In 2006 Mark Pallen and Nicholas Matzke widened the case in Nature Reviews Microbiology. They stressed that "there is no such thing as `the' bacterial flagellum": flagella vary greatly between species in structure, power source and behavior.<sup>9</sup> They noted that "Of the forty or so proteins in the standard flagellum of S.typhimurium strain LT2 or E. coli K-12, only about half seem to be universally necessary," and they listed relatives of flagellar proteins in other systems: for example, the motor proteins MotA and MotB resemble parts of the Ton and Tol transport systems, which use the same kind of ion flow for a different task.<sup>9</sup> On the secretion system they were careful: "Regardless of the conclusion of this debate, the existence of NF T3SSs is `proof of concept' that a flagellar subsystem can function for purposes other than motility."⁹

The same paper made two admissions. Pallen and Matzke wrote that even the reduced core flagellum "is still a challenge to explain," and that "the flagellar research community has scarcely begun to consider how these systems have evolved."⁹ In 2007 Renyi Liu and Howard Ochman proposed that the core parts arose by "the successive duplication and modification of a few, or perhaps even a single, precursor gene"; a Current Biology commentary called the scheme "perhaps too simple."¹⁰

The mousetrap drew a similar reply. The University of Delaware biologist John McDonald published drawings of traps that catch mice with fewer parts, arranged as a series in which each stage works.¹¹ Behe answered that for him "the only relevant question is whether the mousetraps he drew are physical precursors, or merely conceptual precursors."¹² On this narrow point Behe is right: an analogy cannot settle a biochemical question either way.

The critics' second reply: co-option and exaptation

The deeper criticism concerns how evolution is thought to work. Biologists do not picture a machine built one part at a time toward a fixed goal, but parts that already did one job being borrowed, duplicated and refitted for another: co-option or exaptation. Behe's definition looks at what happens when a part is removed from the finished machine today; it does not consider whether its parts once served other functions.

In his 2001 "Reply to My Critics" Behe wrote that his definition puts the focus on removing parts from a working system, whereas "[t]he difficult task facing Darwinian evolution, however, would not be to remove parts from sophisticated pre-existing systems; it would be to bring together components to make a new system in the first place."¹³ He said he hoped to repair the definition in future work. To his credit he named the problem, but the original definition, by itself, does not show that the flagellum could not evolve.

Laboratory work has since tested co-option directly in smaller systems. In 2006 Jamie Bridgham, Sean Carroll and Joseph Thornton reconstructed ancient hormone receptors. They stated the design challenge fairly: integrated systems seem hard to explain "because it is not obvious how any element's function can be selected for unless the partners with which it interacts are already present." Their reconstructed receptor showed that "long before the hormone evolved, the receptor's affinity for aldosterone was present as a structural by-product of its partnership with chemically similar, more ancient ligands."¹⁴ Behe called the result "piddling" when asked by the New York Times, saying a two-component hormone-receptor pair was too simple to count as irreducibly complex, which in his view needs at least three pieces.¹⁵ Both points are fair: the study shows that an interlocking pair can arise stepwise, and it is much smaller than a flagellum.

The blood-clotting cascade gave critics more evidence. In 2003 Yong Jiang and Russell Doolittle reported that in the genome of the puffer fish, genes for the "contact system" factors (factor XI, factor XII and prekallikrein) "could not be identified," although puffer fish clot their blood.¹⁶ If a working cascade can lack three factors, the cascade as a whole is not irreducible in Behe's sense.

Behe's replies

At Dover Behe made three main points, which deserve a full hearing.¹⁷

• A shared part is not a pathway. Finding that some flagellar proteins resemble proteins elsewhere does not show how the whole machine was assembled by random mutation and natural selection. As he put it, "the sequences of the proteins in the type III secretory system and the bacterial flagellum are all well-known, but people still can't figure out how such a thing could have been put together."
• The clotting argument was narrower than critics said. Reading from his book, he noted that he had written, "leaving aside the system before the fork in the pathway, where some details are less well-known, the blood clotting system fits the definition of irreducible complexity." The puffer fish lacks factors from the early part of the cascade, which he had set aside, and knocking out factors in that later part, such as fibrinogen and prothrombin, does break clotting. The court was not persuaded, finding it "not a scientifically warranted redefinition."⁴
• Design rests on purposeful arrangement, not on gaps alone. Irreducible complexity was meant to show a difficulty for one explanation; the inference to design comes from recognizing parts arranged for a purpose, as we do with human machines.

Minnich and Meyer added a point that later research partly supported. They argued that "if anything, the pump evolved from the motor, not the motor from the pump," and that the co-option argument "tacitly presupposes the need for the very thing it seeks to explain--a functionally interdependent system of proteins."¹⁸ In 2012 Sophie Abby and Eduardo Rocha analyzed more than a thousand bacterial genomes and concluded that the non-flagellar secretion system "arose from an exaptation of the flagellum."¹⁹ {{moderate: The best current evidence suggests that the injection syringe descended from the flagellum, so it is probably not the flagellum's ancestor.}} That weakens any popular retelling in which the syringe was a stepping stone to the flagellum. It does not restore the strict irreducibility claim, because a subset of the parts still does useful work, and Abby and Rocha concluded that their work "demonstrates how molecular evolution can convert one complex molecular machine into a second, equally complex machine by successive deletions, innovations, and recruitment from other molecular systems."¹⁹

Behe also tried to put numbers on the problem. In 2004 he and the physicist David Snoke modeled how long it would take point mutations in duplicated genes to produce a protein feature that needs two or more amino acid changes at once, and concluded that very large populations would be needed.²⁰ The evolutionary geneticist Michael Lynch replied in the same journal that their conclusions were "an artifact of unwarranted biological assumptions, inappropriate mathematical modeling, and faulty logic."²¹ Behe and Snoke replied in the same issue.²¹ The court at Dover noted that the Behe and Snoke paper does not mention either irreducible complexity or intelligent design.²²

Behe's later books, The Edge of Evolution (2007) and Darwin Devolves (2019), shifted to the limits of random mutation. Reviewing the second in Science, Nathan Lents, Joshua Swamidass and Richard Lenski agreed that gene-breaking adaptations are real; Lenski summarized their objection: "Behe largely ignores the ways that evolution generates new functions and thereby produces complexity."²³

The Dover trial

In Kitzmiller v. Dover Area School District (2005), parents challenged a school board policy requiring that ninth-grade biology students hear a statement presenting intelligent design as an alternative to evolution. Behe and Minnich were among the experts for the board; Miller and Kevin Padian were among those for the parents. Judge John E. Jones III ruled the policy unconstitutional.

On irreducible complexity the opinion was blunt. It found that Behe's definition excludes co-option "by definitional fiat," and quoted a National Academy of Sciences publication: "it is incorrect to assume that a complex structure or biochemical process can function only if all its components are present and functioning as we see them today."²⁴ It recorded Minnich's own concession that irreducible complexity "is not a test of intelligent design; it's a test of evolution."²⁵ On the immune system, Behe was shown fifty-eight peer-reviewed publications, nine books and several textbook chapters on its evolution and said they were not "good enough"; the court concluded that the argument depended "upon setting a scientifically unreasonable burden of proof for the theory of evolution."²⁶ {{very-strong: The court found that Behe's claim for irreducible complexity "has been refuted in peer-reviewed research papers and has been rejected by the scientific community at large."}}²⁷

Two cautions. First, a court decides a legal question on the record before it; it is not a scientific body, and design advocates have argued that the opinion misread their position. Second, the judge himself limited what he decided. He wrote that "while ID arguments may be true, a proposition on which the Court takes no position, ID is not science," and added, "Nor do we controvert that ID should continue to be studied, debated, and discussed."²⁸ The ruling is strong evidence of how the scientific community judged irreducible complexity in 2005. It is not a verdict on whether God made the world.

The strongest objections, stated fairly

Objection 1: the definition fails for its own best example

This is the critics' central point, and we think it correct. A working subset of flagellar proteins exists, and many flagella lack parts that others have. If the claim is that removing any part leaves nothing useful, the claim is false for the flagellum. Design advocates can reply that the remaining machine still needs its core parts, and that is true, but it is a weaker claim than the one first made.

Objection 2: it is an argument from ignorance

Critics say that a present lack of knowledge about how something evolved has repeatedly been turned into a claim that it could not have evolved, and that the immune-system episode at Dover illustrates the danger. Behe's best reply is that he does not argue from ignorance alone: we infer design from the purposeful arrangement of parts, as we would for an outboard motor. That reply has force as philosophy (see Design as an Inference, Not a Science-Stopper and The Design Inference), but it is a different argument from irreducible complexity and must be judged on its own terms.

Behe's counter-objection: shared parts are not a mechanism

This is fair. Shared ancestry of parts makes co-option plausible but is not a detailed, tested pathway, as Miller and Pallen and Matzke conceded. The honest summary: a detailed pathway is missing, but the specific reason given for thinking none could exist has failed.

How strong is the argument? Our ratings

• {{debated: The bacterial flagellum is irreducibly complex in Behe's strict 1996 sense.}} The evidence from functioning subsystems, variation between flagella and shared parts tells against it.
• The milder claim that no detailed, tested step-by-step evolutionary account of the flagellum yet exists: Strong, by the critics' own admission, though proposed schemes exist and research continues.
• The inference from the lack of a detailed account to the conclusion that the flagellum was designed: Debated. It depends on philosophical judgments about the best explanation, not on irreducible complexity alone.

This weakens a popular apologetic, and we would rather say so than repeat it. The flagellum is best presented as a marvel of order, not as proof that evolution is impossible.

An Orthodox Christian perspective

We know of no conciliar definition of the Orthodox Church about the mechanisms of biology (see Can a Christian Accept Evolution?). What the Church does teach is that creation, all of it, reveals its Maker. St Paul wrote that "since the creation of the world His invisible attributes are clearly seen, being understood by the things that are made" (Romans 1:20, NKJV), and the Wisdom of Solomon, part of the Old Testament as the Church received it in the Septuagint, says: "For by the greatness and beauty of the creatures proportionably the maker of them is seen" (Wisdom 13:5, Brenton).²⁹

St Basil the Great, preaching on the six days of creation in the fourth century, said that the world "was not conceived by chance and without reason, but for an useful end," since it is "the school where reasonable souls exercise themselves, the training ground where they learn to know God."³⁰ For Basil, the whole ordered cosmos is the evidence, not only the parts that resist explanation. If God's wisdom is seen only where science has a gap, each discovery shrinks the evidence for God. If it is seen in the order of nature as a whole, a natural history of the flagellum, if one is found, would not remove the Creator; it would show more of how he works. See also Has Science Disproved Christianity? and Does God Exist?

None of this tells a Christian that Behe is wrong to look for design in biology, nor that Miller is right about every detail. It means that the faith does not rest on the outcome of the flagellum debate. Orthodox Christians can follow the evidence calmly, give credit where it is due on both sides, and keep their confidence in the Creator where the Fathers placed it.

What we can say confidently

• Behe introduced irreducible complexity in 1996 and defined it as a system that effectively stops working when any one part is removed.
• A subset of flagellar proteins forms a working secretion system, and only about half of the standard flagellar proteins seem to be universally necessary.
• Behe acknowledged in 2001 that his definition did not address how components come together in the first place.
• The Kitzmiller court (2005) found irreducible complexity refuted and rejected by the scientific community, while expressly taking no position on whether design arguments are true.
• Leading critics (Miller in 2004, Pallen and Matzke in 2006) acknowledged that no detailed evolutionary history of the flagellum had been worked out.

What remains disputed

• The exact history of the flagellum and of its relationship to the secretion system (recent work favors descent of the syringe from the flagellum; others have argued for a shared ancestor).
• Whether co-option can account for the origin of the parts unique to the flagellum and for their ordered assembly.
• How much random mutation and selection can achieve in realistic populations (Behe and Snoke against Lynch and others).
• Whether "purposeful arrangement of parts" is a valid scientific inference, a philosophical one, or neither.
• Whether the Kitzmiller opinion described intelligent design fairly.

Conclusion

The flagellum is a beautiful thing: a self-assembling rotary motor. Behe was right that it deserves explanation and that homologies alone do not supply one. But his specific argument, that removing any part leaves nothing natural selection could use, has not survived the evidence, and we should not keep using it as if it had. The Christian case for a Creator does not need it. The Church confesses "one God, the Father Almighty, Maker of Heaven and Earth and of all things visible and invisible,"³¹ and that confession stands whether the flagellum's history is someday written out in detail or remains, for now, an open question. Wonder at the order of creation, and honesty about what we do not know, belong together.

Notes

1. Charles Darwin, On the Origin of Species by Means of Natural Selection, 1st ed. (London: John Murray, 1859), ch. 6, "Difficulties on Theory." Text checked at Project Gutenberg eBook #1228.
2. Michael J. Behe, Darwin's Black Box (New York: Free Press, 1996), 39, and "Reply to My Critics," Biology and Philosophy 16 (2001): 694, as quoted in Kitzmiller v. Dover Area School District, 400 F. Supp. 2d 707 (M.D. Pa. 2005), slip op. (Doc. 342) at 73. We quote the court's text; we did not open the book itself.
3. Behe, Darwin's Black Box, as quoted by John H. McDonald, "A reducibly complex mousetrap," University of Delaware, udel.edu/~mcdonald/mousetrap.html. Page in Behe not verified.
4. Kitzmiller, slip op. at 76-77.
5. Scott A. Minnich and Stephen C. Meyer, "Genetic Analysis of Coordinate Flagellar and Type III Regulatory Circuits in Pathogenic Bacteria," in Design and Nature II, ed. M. W. Collins and C. A. Brebbia (Southampton: WIT Press, 2004), abstract; open-access text at witpress.com.
6. Kitzmiller v. Dover trial transcript, Day 10, afternoon session (October 17, 2005), testimony of Michael Behe, TalkOrigins Archive, talkorigins.org/faqs/dover/day10pm.html.
7. Trial transcript, Day 10, morning session (October 17, 2005), Behe direct examination, talkorigins.org/faqs/dover/day10am.html.
8. Kenneth R. Miller, "The Flagellum Unspun: The Collapse of 'Irreducible Complexity,'" in Debating Design: From Darwin to DNA, ed. William A. Dembski and Michael Ruse (Cambridge: Cambridge University Press, 2004). Quoted from the author's pre-publication copy at millerandlevine.com; wording in the printed book not checked.
9. Mark J. Pallen and Nicholas J. Matzke, "From The Origin of Species to the origin of bacterial flagella," Nature Reviews Microbiology 4, no. 10 (2006): 784-790, doi:10.1038/nrmicro1493.
10. Renyi Liu and Howard Ochman, "Stepwise formation of the bacterial flagellar system," Proceedings of the National Academy of Sciences 104, no. 17 (2007): 7116-7121, abstract, doi:10.1073/pnas.0700266104; W. Ford Doolittle and Olga Zhaxybayeva, "Evolution: reducible complexity -- the case for bacterial flagella," Current Biology 17, no. 13 (2007): R510-R512, abstract, doi:10.1016/j.cub.2007.05.003.
11. McDonald, "A reducibly complex mousetrap."
12. Michael J. Behe, "A Mousetrap Defended: Response to Critics," Discovery Institute, July 31, 2000, discovery.org/a/446/.
13. Behe, "Reply to My Critics," 695, as quoted in Kitzmiller, slip op. at 73. The article's full pagination is Biology and Philosophy 16, no. 5 (2001): 683-707, doi:10.1023/A:1012268700496.
14. Jamie T. Bridgham, Sean M. Carroll and Joseph W. Thornton, "Evolution of Hormone-Receptor Complexity by Molecular Exploitation," Science 312, no. 5770 (2006): 97-101, abstract, doi:10.1126/science.1123348.
15. Kenneth Chang, "Study, in a First, Explains Evolution's Molecular Advance," New York Times, April 7, 2006.
16. Yong Jiang and Russell F. Doolittle, "The evolution of vertebrate blood coagulation as viewed from a comparison of puffer fish and sea squirt genomes," Proceedings of the National Academy of Sciences 100, no. 13 (2003): 7527-7532, abstract, doi:10.1073/pnas.0932632100.
17. Trial transcript, Day 10, afternoon, and Day 11, morning session (October 18, 2005), Behe direct examination, talkorigins.org/faqs/dover/day11am.html.
18. Minnich and Meyer, "Genetic Analysis," discussion section (p. 302 of the volume).
19. Sophie S. Abby and Eduardo P. C. Rocha, "The Non-Flagellar Type III Secretion System Evolved from the Bacterial Flagellum and Diversified into Host-Cell Adapted Systems," PLoS Genetics 8, no. 9 (2012): e1002983, abstract, doi:10.1371/journal.pgen.1002983.
20. Michael J. Behe and David W. Snoke, "Simulating evolution by gene duplication of protein features that require multiple amino acid residues," Protein Science 13, no. 10 (2004): 2651-2664, doi:10.1110/ps.04802904.
21. Michael Lynch, "Simple evolutionary pathways to complex proteins," Protein Science 14, no. 9 (2005): 2217-2225, abstract, doi:10.1110/ps.041171805, with Michael Behe and David Snoke, "A response to Michael Lynch," Protein Science 14, no. 9 (2005): 2226-2227, doi:10.1110/ps.051674105.
22. Kitzmiller, slip op. at 88 n. 17, which also records Behe's testimony about population size.
23. Nathan H. Lents, S. Joshua Swamidass and Richard E. Lenski, "The end of evolution?" (review of Darwin Devolves), Science 363, no. 6427 (2019): 590, doi:10.1126/science.aaw4056; Richard Lenski, "On damaged genes and polar bears," Telliamed Revisited (blog), February 22, 2019.
24. Kitzmiller, slip op. at 75.
25. Kitzmiller, slip op. at 72, citing trial transcript 38:82 (Minnich).
26. Kitzmiller, slip op. at 78.
27. Kitzmiller, slip op. at 79.
28. Kitzmiller, slip op. at 64 and 137.
29. Romans 1:20, New King James Version; Wisdom of Solomon 13:5, in Lancelot C. L. Brenton, trans., The Septuagint with Apocrypha (London: Samuel Bagster, 1851), text at ebible.org.
30. Basil of Caesarea, Hexaemeron, Homily 1.6, trans. Blomfield Jackson, in Nicene and Post-Nicene Fathers, Second Series, vol. 8 (1895), text at newadvent.org/fathers/32011.htm.
31. The Nicene-Constantinopolitan Creed, in the Orthodox Church in America's translation, "The Symbol of Faith," oca.org/orthodoxy/prayers/symbol-of-faith.

Bibliography

Primary sources

• Behe, Michael J. Darwin's Black Box: The Biochemical Challenge to Evolution. New York: Free Press, 1996. Later books: The Edge of Evolution (Free Press, 2007); Darwin Devolves (HarperOne, 2019).
• Behe, Michael J. "A Mousetrap Defended: Response to Critics." Discovery Institute, July 31, 2000.
• Behe, Michael J. "Reply to My Critics: A Response to Reviews of Darwin's Black Box." Biology and Philosophy 16, no. 5 (2001): 683-707.
• Darwin, Charles. On the Origin of Species by Means of Natural Selection. London: John Murray, 1859.
• Kitzmiller v. Dover Area School District, 400 F. Supp. 2d 707 (M.D. Pa. 2005). Memorandum Opinion, Doc. 342, December 20, 2005, pamd.uscourts.gov.
• Kitzmiller v. Dover trial transcripts, Days 10 and 11 (October 17-18, 2005). TalkOrigins Archive.
• The Holy Bible, New King James Version. Nashville: Thomas Nelson, 1982. Brenton, Lancelot C. L., trans. The Septuagint with Apocrypha. London: Samuel Bagster, 1851.

Orthodox and patristic sources

• Basil of Caesarea. Hexaemeron. Translated by Blomfield Jackson. Nicene and Post-Nicene Fathers, Second Series, vol. 8. Edited by Philip Schaff and Henry Wace. 1895.
• Orthodox Church in America. "The Symbol of Faith" (Nicene-Constantinopolitan Creed). oca.org.

Other scholarship (including scholars who disagree)

• Abby, Sophie S., and Eduardo P. C. Rocha. "The Non-Flagellar Type III Secretion System Evolved from the Bacterial Flagellum and Diversified into Host-Cell Adapted Systems." PLoS Genetics 8, no. 9 (2012): e1002983.
• Bridgham, Jamie T., Sean M. Carroll, and Joseph W. Thornton. "Evolution of Hormone-Receptor Complexity by Molecular Exploitation." Science 312 (2006): 97-101.
• Chang, Kenneth. "Study, in a First, Explains Evolution's Molecular Advance." New York Times, April 7, 2006.
• Jiang, Yong, and Russell F. Doolittle. "The evolution of vertebrate blood coagulation as viewed from a comparison of puffer fish and sea squirt genomes." PNAS 100 (2003): 7527-7532.
• Lents, Nathan H., S. Joshua Swamidass, and Richard E. Lenski. "The end of evolution?" Science 363 (2019): 590. With Richard Lenski, "On damaged genes and polar bears," Telliamed Revisited, February 22, 2019.
• Lynch, Michael. "Simple evolutionary pathways to complex proteins." Protein Science 14 (2005): 2217-2225.
• Liu, Renyi, and Howard Ochman. "Stepwise formation of the bacterial flagellar system." PNAS 104 (2007): 7116-7121. With W. Ford Doolittle and Olga Zhaxybayeva, "Evolution: reducible complexity -- the case for bacterial flagella," Current Biology 17 (2007): R510-R512.
• McDonald, John H. "A reducibly complex mousetrap." University of Delaware, web page.
• Miller, Kenneth R. "The Flagellum Unspun: The Collapse of 'Irreducible Complexity.'" In Debating Design: From Darwin to DNA, edited by William A. Dembski and Michael Ruse. Cambridge: Cambridge University Press, 2004. See also Miller, Finding Darwin's God (HarperCollins, 1999).
• Pallen, Mark J., and Nicholas J. Matzke. "From The Origin of Species to the origin of bacterial flagella." Nature Reviews Microbiology 4 (2006): 784-790.

Sources from the intelligent design movement

• Behe, Michael J., and David W. Snoke. "Simulating evolution by gene duplication of protein features that require multiple amino acid residues." Protein Science 13 (2004): 2651-2664. With their "A response to Michael Lynch," Protein Science 14 (2005): 2226-2227.
• Minnich, Scott A., and Stephen C. Meyer. "Genetic Analysis of Coordinate Flagellar and Type III Regulatory Circuits in Pathogenic Bacteria." In Design and Nature II, edited by M. W. Collins and C. A. Brebbia, WIT Transactions on Ecology and the Environment 73. Southampton: WIT Press, 2004.