Education Icon Education
Free Speech Icon Free Speech
Intelligent Design Icon Intelligent Design

Do Car Engines Run on Lugnuts? A Response to Ken Miller & Judge Jones’s Straw Tests of Irreducible Complexity for the Bacterial Flagellum (Continued — Part II)

(Part II, Version 1.0)
By Casey Luskin
Copyright © 2006 Casey Luskin. All Rights Reserved.

The entire article can be read here

…Yesterday, I posted Part I of this response. To reiterate, there are three primary problems with Judge Jones’s ruling that Ken Miller refuted Michael Behe’s arguments that the bacterial flagellum is irreducible complex:

  • (A) Experts say the evidence suggests that the TTSS evolved from the flagellum, and not the other way around.
  • (B) Behe and other ID-proponents have long-acknowledged “exaptation” or “co-option” as an attempt to evolve biological complexity, and have observed many problems with “co-option” explanations.
  • (C) Miller has inaccurately characterized how one tests for irreducible complexity, thus refuting only a straw-version of Behe’s concept of irreducible complexity.

Yesterday I posted sections addressing parts (A) and (B). Today I will continue with the response, expanding on Part (C):

(C) Miller’s Incorrect Characterization of Irreducible Complexity

To repeat Miller’s assertion, he testified that irreducible complexity is refuted if one sub-system can perform some other function in the cell:

“Dr. Behe’s prediction is that the parts of any irreducibly complex system should have no useful function. Therefore, we ought to be able to take the bacterial flagellum, for example, break its parts down, and discover that none of the parts are good for anything except when we’re all assembled in a flagellum.”

Dr. Kenneth Miller Testimony, Day 1, PM Session, page 16.

The question becomes, “how is Behe’s argument different from that of Ken Miller?” Behe actually formulates irreducible complexity as a test of building an entire system. IC operates on a collection of parts, not each individual part. Even if a separate function could be found for a sub-system, the latter would not refute the irreducible complexity and the unevolvability of the system as a whole. To repeat Behe’s definition, Behe writes:

“In The Origin of Species Darwin stated:

‘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.’

A system which meets Darwin’s criterion is one which exhibits irreducible complexity. By irreducible complexity I mean a single system 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.”

Michael Behe, Darwin’s Black Box, pg. 39 (Free Press, 1996)

Thus, according to Darwin, evolution requires that a system, or its sub-parts, be functional along each small step of their evolution to the final system. Yet one could find a sub-part that could be useful outside of the final system, and yet the total system would still face many points along an “evolutionary pathway” where it could not remain functional along “numerous, successive, slight modifications” that would be necessary for its gradual evolution. (With regards to the flagellum at least 2/3 of the parts are not known to be shared with any other structure therefore might not be even a sub-part of another system at all.) (*** Note: this footnoted statement has been changed in the official version to reflect newly learned information. See below for details.)

Thus, Miller mischaracterizes Behe’s argument as one which focuses on the non-functionality of sub-parts, when in fact, Behe’s argument actually focuses on the ability of the entire system to assemble, even if sub-parts can have functions outside of the final system.

A Car Example for Illustration

To understand how Miller’s test fails to accurately apply to Behe’s formulation of irreducible complexity, consider the example of a car engine and a bolt. Car engines use various kinds of bolts, and a bolt could be seen as a small “sub-part” or “sub-system” of a car engine. Under Miller’s logic, if a vital bolt in my car’s engine might also to perform some other function–perhaps as a lugnut–then it follows that my car’s whole engine system is not irreducibly complex. Such an argument is obviously fallacious.

In assessing whether an engine is irreducibly complex, one must focus on the function of the engine itself, not on the possible function of some sub-part that may operate elsewhere. Of course a bolt out of my engine could serve some other purpose in my car. However this observation does not explain how many complex parts such as pistons, cylinders, the camshaft, valves, the crankshaft, sparkplugs, the distributor cap, and wiring came together in the appropriate configuration to make a functional car engine. Even if all of these parts could perform some other function in the car (which is doubtful), how were these parts assembled properly to construct a functional engine? The answer requires intelligent design.

Behe asserts that a system is irreducibly complex if the system stops functioning upon the removal of one part. This is the appropriate test of Darwin’s theory because it asks the question, “Is there a minimal level of complexity which is required for functionality of this system?” Clearly my car’s engine has a core set of parts necessary in order for it to function. The ability of an engine bolt to also serve as a lugnut does not refute the irreducibly complex arrangement of parts necessary to make the final engine-system functional. Behe never suggests that subsystems cannot play some other role in the cell–in fact he suggests the opposite. Rather, Behe simply argues that evolution requires that the total system must be built up in a slight, step-by-step fashion, where each step is functional.

Miller has mischaracterized irreducible complexity, and his test is a straw-test for refuting irreducible complexity. The test for irreducible complexity does not ask “can one small part of the macrosystem be used to do something else?” as Miller claims, but rather asks “can the system as a whole be built in a step-by-step fashion which does not require any ‘non-slight’ modifications to gain the final target function?” Any non-slight modifications of complexity required to go from functional sub-part(s), operating outside-of-the-final system, to the entire final functional system, represent the irreducible complexity of a system.

Even if Miller could find that every part of the flagellum existed somewhere else in bacteria (which he cannot–he only accounts for the basal body, which constitutes about 1/4 of the total flagellar proteins), Miller is no where close to providing a plausible account of the evolution of the flagellum until he has explained how all the flagellar parts might have come together to produce a functional bacterial flagellum. Only then that Miller claim that the flagellum is not irreducibly complex.

Other Authorities Agree with Behe

William Dembski captures the essence of the problem with Miller’s definition and treatment of IC in Dembski’s expert rebuttal in which Dembski writes:

“[F]inding a subsystem of a functional system that performs some other function is hardly an argument for the original system evolving from that other system. One might just as well say that because the motor of a motorcycle can be used as a blender, therefore the [blender] motor evolved into the motorcycle. Perhaps, but not without intelligent design. Indeed, multipart, tightly integrated functional systems almost invariably contain multipart subsystems that serve some different function. At best the TTSS [Type-III Secretory System] represents one possible step in the indirect Darwinian evolution of the bacterial flagellum. But that still wouldn’t constitute a solution to the evolution of the bacterial flagellum. What’s needed is a complete evolutionary path and not merely a possible oasis along the way. To claim otherwise is like saying we can travel by foot from Los Angeles to Tokyo because we’ve discovered the Hawaiian Islands. Evolutionary biology needs to do better than that.”

(William A. Dembski, Rebuttal to Reports by Opposing Expert Witnesses, at

Though Miller has accounted for the origin of only a fraction of the flagellar parts, Scott A. Minnich and Stephen C. Meyer also explain how mere availability of parts is insufficient to explain the evolution of a system:

“[E]ven if all the protein parts were somehow available to make a flagellar motor during the evolution of life, the parts would need to be assembled in the correct temporal sequence similar to the way an automobile is assembled in factory. Yet, to choreograph the assembly of the parts of the flagellar motor, present-day bacteria need an elaborate system of genetic instructions as well as many other protein machines to time the expression of those assembly instructions. Arguably, this system is itself irreducibly complex. In any case, the co-option argument tacitly presupposes the need for the very thing it seeks to explain–a functionally interdependent system of proteins.”

(Scott A. Minnich and Stephen C. Meyer, Genetic Analysis of coordinate flagellar and type III regulatory circuits in pathogenic bacteria, pg. 8, at

Similarly, philosopher Angus Menuge lays out a number of obstacles which must be overcome by “co-option” or “exaptation” explanations, none of which were addressed by Miller during the trial. Menuge writes:

“For a working flagellum to be built by exaptation, the five following conditions would all have to be met:

“C1: Availability. Among the parts available for recruitment to form the flagellum, there would need to be ones capable of performing the highly specialized tasks of paddle, rotor, and motor, even though all of these items serve some other function or no function.
“C2: Synchronization. The availability of these parts would have to be synchronized so that at some point, either individually or in combination, they are all available at the same time.
“C3: Localization. The selected parts must all be made available at the same ‘construction site,’ perhaps not simultaneously but certainly at the time they are needed.
“C4: Coordination. The parts must be coordinated in just the right way: even if all of the parts of a flagellum are available at the right time, it is clear that the majority of ways of assembling them will be non-functional or irrelevant.
“C5: Interface compatibility. The parts must be mutually compatible, that is, ‘well-matched’ and capable of properly ‘interacting’: even if a paddle, rotor, and motor are put together in the right order, they also need to interface correctly.”

Angus Menuge, Agents Under Fire: Materialism and the Rationality of Science, pgs. 104-105 (Rowman & Littlefield, 2004).

William Dembski takes a similar approach to that of Menuge. Dembski effectively combines some of Menuge’s criteria in order to develop a means of calculating the probability of constructing an irreducibly complex object. In calculating the probability of a “discrete combinatorial object” one must take into account the probability of originating the parts, the probability of localizing the parts all in once place, and the probability of configuring the parts together:

Table 1. Comparison of Dembski and Menuge’s required explanatory components for any exaptation-based account of evolution (Table based upon the descriptions in William A. Dembski, No Free Lunch: Why Specified Complexity Cannot be Purchased Without Intelligence, pg. 291 (Rowman & Littlefield, 2002)):

Dembski’s FactorDescriptionAnalogue in Menuge’s Criteria
PorigProbability of originating the building blocks for that objects.C1
PlocalProbability of locating the building blocks in one place once they are given.C2, C3
PconfigProbability of configuring the building blocks once they are given and in one place.C4, C5

It is clear that Miller has found that the probability for originating about 1/4 of the flagellar proteins might be “1/1” if the TTSS (or perhaps a similar structure) existed prior to the flagellum.

However he has not accounted for the origin of the remaining the flagellar proteins, nor has he addressed Plocal or Pconfig in his evolutionary scenario. In light of Menuge’s and Dembski’s criteria, it seems fair to demand answers from Miller to the following questions:

  • What function was performed by the paddle, rotor, or motor outside of the flagellum? (At trial, Miller explained the function for the basal body of the flagellum via the TTSS, but left the most complex and vital motorized portions of the flagellum unexplained.)
  • How did the parts become synchronized in the flagellum? (At trial, Miller never discussed this.)
  • How did the parts become localized within the flagellum at the same construction site? (At trial, Miller never addressed this issue.)
  • How did the parts become structurally coordinated so as to interact when assembled to produce the flagellar swimming function? (Again, Miller also never addressed this issue at trial.)

Thus Miller never answered any of these important questions at the trial. Rather, he presented a straw version of testing irreducible complexity whereby he convinced the Judge in a fashion which did not come remotely close to accounting for how the flagellum might have actually evolved.

A Final Analogy: The Arch

Miller’s treatment of the bacterial flagellum did not refute its irreducibly complexity, as Miller did not address questions about how the final flagellar systems might arise. The existence of other functions for the TTSS does not imply that the flagellar system would not still require large leaps in complexity (or to use Darwin’s words, non-slight modifications) in order to ultimately achieve a functional flagellum. To use a final analogy to show the deficiency of Miller’s explanation, consider an attempt to build an irreducibly complex arch (Figure A):

Figure A: An arch is irreducibly complex: if one removes a piece, the remaining pieces will fall down. (Note: For the purpose of illustration, I am temporarily ignoring the common objection that an irreducibly complex arch might be made using natural erosional processes. I am aware of no appropriate “scaffolding” analogy within the biological realm, but it is not the present purpose of this discussion to rebut that objection.)

According to Miller, if we can find a function for some sub-piece, then a system is not irreducibly complex. Now, let’s now break this arch into sub-pieces:

Figure B: Here an arch has been broken up into subpieces. Similarly, Miller has apparently found a flagellar sub-piece (the TTSS) which can perform some other function. The TTSS comprises no more than 1/4 of the total flagellar parts. Similarly, in this arch, there is one large sub-section (labeled “S”) which comprises approximately 1/4 of the total arch. Sub-section “S” can have a function outside of the arch (i.e. here, it can stand on its own). However, this exposes the fallacy of Miller’s test: the ability of sub-section “S” to stand on its own does not therefore dictate that the arch is not irreducibly complex. Thus if one were to removes the top piece (t), the arch crumbles, even if sub-section “S” can still remain standing (Figure C):

Figure C: Even if sub-section “S” can have a function (i.e. stand) on its own outside of the arch, this does not imply that the arch as a whole is not irreducibly complex — capable of being built in a step-by-step manner. Thus, the appropriate test of irreducible complexity asks if the entire system can be built in a step-by-step manner using small, slight-modifications. It is important to note that the system does not become “reducibly complex” simply because one part remains functional outside of the final system.

Correctly Testing Irreducible Complexity:

Miller’s test of irreducible complexity is a straw test. The correct test would have stated:

“Dr. Behe’s prediction is that an irreducibly complex system will go through some non-functional stage along any evolutionary pathway. Therefore, we ought to be able to take the bacterial flagellum, for example, remove a part, and discover that the system stops working.”

Miller’s testimony and Judge Jones’s conclusion is based upon a false test of irreducible complexity which focuses on the functionality of one-sub-part, not the functionality of the entire flagellar system.

If Miller could find functions for all flagellar sub-systems outside of the flagellum, he would admittedly be making progress towards an evolutionary explanation by satisfying Angus Menuge’s first criterion of “Availability” (C1). However, as Minnich and Meyer ask:

“[T]he other thirty proteins in the flagellar motor (that are not present in the TTSS) are unique to the motor and are not found in any other living system. From whence, then, were these protein parts co-opted?”

Miller has no answer to that question. As it stands, all Miller could provide as evidence for the evolution of the flagellum is the TTSS, and which would account for the availability about 1/4 of Menuge’s first step. Ignoring the fact that the TTSS probably evolved from the flagellum, and not the other way around, if Miller could account for the availability (C1) of all of the parts of the flagellum then Miller would simply have to explain the Synchronization (C2), Localization (C3), Coordination (C4), and Interface compatibility (C5) to account for the evolution of the flagellum; thereby eliminating the need for “non-slight” modifications along its evolutionary pathway.

But Miller did none of this. Despite the inadequacy of Miller’s explanations, Judge Jones decided that Dr. Kenneth Miller’s arguments and inaccurate characterizations of irreducible complexity were correct. This is ruling was made despite the fact that Dr. Scott Minnich, a pro-ID microbiologist and expert on the flagellum, testified extensively at the trial about how his own tests demonstrate the irreducibly complex nature of the flagellum Consider Minnich’s testimony which Jones completely ignored in the Kitzmiller decision:

“A. I work on the bacterial flagellum, understanding the function of the bacterial flagellum for example by exposing cells to mutagenic compounds or agents, and then scoring for cells that have attenuated or lost motility. This is our phenotype. The cells can swim or they can’t. We mutagenize the cells, if we hit a gene that’s involved in function of the flagellum, they can’t swim, which is a scorable phenotype that we use. Reverse engineering is then employed to identify all these genes. We couple this with biochemistry to essentially rebuild the structure and understand what the function of each individual part is. Summary, it is the process more akin to design that propelled biology from a mere descriptive science to an experimental science in terms of employing these techniques.”

Scott Minnich testimony, Day 20, pm session, pg. 105

Minnich explained how he mutated all of the flagellar genes and found that the flagellum loses function if even one gene is missing. Thus, the flagellum is irreducibly complex with respect to its gene compliment:

“One mutation, one part knock out, it can’t swim. Put that single gene back in we restore motility. Same thing over here. We put, knock out one part, put a good copy of the gene back in, and they can swim. By definition the system is irreducibly complex. We’ve done that with all 35 components of the flagellum, and we get the same effect.”

Scott Minnich Testimony, Day 20, pm session, pg. 107-108

Unfortunately Judge Jones chose to ignore this testimony.


Regardless of Judge Jones’ claim, the pro-ID arguments regarding irreducible complexity in the flagellum were never, as Jones said, “refuted.” Miller provided the Judge with a false characterization of irreducible complexity and a straw-method of testing it. Unfortunately, this ruling, which canonized Miller’s misrepresentation of irreducible complexity, will lead the scientific community and the general public to mistakenly assume that the evolutionary origin of the bacterial flagellum can be explained. Incredibly, despite Minnich’s testimony and the presentation of his experimental slides, Judge Jones still held that, “ID … has failed to … engage in research and testing.” (Kitzmiller ruling, pg. 89.)

Had the Judge not also accepted Miller’s fallacious claim that irreducible complexity is not a positive argument for design, but merely a negative argument against evolution, perhaps we might have seen some different findings in this case. Minnich and Meyer make this positive case for the design of the flagellum:

“Molecular machines display a key signature or hallmark of design, namely, irreducible complexity. In all irreducibly complex systems in which the cause of the system is known by experience or observation, intelligent design or engineering played a role the origin of the system. … That we have encountered systems that tax our own capacities as design engineers, justifiably lead us to question whether these systems are the product of undirected, un-purposed, chance and necessity. Indeed, in any other context we would immediately recognize such systems as the product of very intelligent engineering. Although some may argue this is a merely an argument from ignorance, we regard it as an inference to the best explanation, given what we know about the powers of intelligent as opposed to strictly natural or material causes.”

Scott A. Minnich and Stephen C. Meyer, Genetic Analysis of coordinate flagellar and type III regulatory circuits in pathogenic bacteria, pg. 8, at

In the final analysis, Judge Jones’s ruling on the origin of the flagellum should be disregarded as an example of partisan politics, not as objective or accurate scientific analysis.

The entire article can be read here


I would like to thank Alex Binz, David Klinghoffer, and an un-named biochemist in the University of California system for their help with this post. Any mistakes are my own.

*** Note added 10/27/06: Version 1.0 of this document (which is the version reprinted here) stated: “With regards to the flagellum at least 2/3 of the parts are not known to be shared with any other structure therefore might not be even a sub-part of another system at all.” The latest version of this document, Version 1.5, has changed that statement to reflect newly learned information as follows:

With regards to the flagellum, there is dispute as to whether many flagellar genes can even be said to be homologous to genes outside of the flagellum. This dispute is witnessed in the conclusion of pro-ID biologist Mike Gene who writes when critiquing Nick Matzke’s model attempting to evolve a flagellum: “The various dissimilarities (some very profound) listed above, along with the weakness of the criteria for inferring homology, is only rendered more problematic by the seemingly arbitrary nature of the chosen matches.” Matzke’s model may be found at Mike Gene’s critique may be found at Obviously there are different types of flagella, but Matzke admits that least ¼ of flagellar proteins have no known homology, and also acknowledges that “the flagellar research community has scarcely begun to consider how these systems have evolved.” See Mark J. Pallen and Nicholas J. Matzke, “From The Origin of Species to the origin of bacterial flagella,” Nature Reviews Microbiology, AOP, published online 5 September 2006; doi:10.1038/nrmicro1493

Nature Reviews Microbiology, AOP, published online 5 September 2006; doi:10.1038/nrmicro1493

Back up to to the main text.