|
Author
|
Topic: A Catalogue of Criticisms of Irreducible Complexity
|
Micah Sparacio
Member
Member # 6
|
posted 29. May 2003 13:27
Following in the footsteps of this thread: Cataloguing Criticisms of Specified Complexity and the issues raised by the recent discussion on The Evolutionary Origin of Complex Features I thought it might be fitting to discuss criticisms of Irreducible Complexity.
Here is an encyclopedia entry on the term for reference: Irreducible Complexity
One concern that I've heard raised, as of late, is that there is no single definition of IC to target for criticism. Rather, it is a moving goalpost. And yet, I think that the basic concept of IC is strong enough in most of our minds to carry this discussion forward. If not, then use the above encyclopedia reference and pick one of the various instantiations of IC seen there.
So, here are the rules for this thread:
1. Be concise
2. No links
3. In your own words
4. Label your criticism as C* where * is a sequential number
5. No comments on the criticisms of others unless it is essential to developing your own criticism.
A new thread will be opened later on to discuss these criticisms. But, for the time being, this Brainstorm should be constrained by the above guidelines.
Here's are my initial criticisms:
C1: The current definitions were developed within the context of three dimensional biological systems, which constrains the concept to 3D mechanical systems. For example, what if we have a structurally simple system that performs a complex function? The current definitions, with their dependency on "parts" makes the definition ungeneralizable to various computer simulations, etc.
C2: Why do we apply the complexity criteria to structure rather than, say genetic sequence? This seems to imply that the "hurdle" to jump is the organization of physical matter in space, and NOT the organization of the information which specifies an organization for the physical matter in space.
IP: Logged
|
|
Argon
Member
Member # 276
|
posted 29. May 2003 16:08
Let's not forget a recent topic, "A sequence of tests for IC", started by Gedanken with the following proposal:
"In this thread I propose that varying tests and definitions of “Irreducible Complexity” or “IC” be stated."
IP: Logged
|
|
yersinia
Member
Member # 324
|
posted 30. May 2003 00:25
Here are my main ones:
C3: There remains ambiguity about whether "evolvability" is part of the definition of IC or not. If "unevolvable" is part of the definition, then saying "IC systems cannot/are very unlikely to evolve" is nothing but a tautology and proves nothing. Common assertions, e.g. that IC is a subset of SC, support the notion that evolvability is part of the definition.
OTOH, I think that the most fair reading of Behe (particularly his piece in Philosophy and Biology, which has some clarifications) is that IC is defined based on present characteristics. Then, an argument is made about the evolvability of independently-defined IC systems.
C4: There remains ambiguity even within the "present characteristics" group of IC definitions. To me, it appears that "multiple parts required for identified function" is the only essential characteristic required to meet the definition. This is the only characteristic universally invoked by IDists, and it's the only characteristic that bears weight in the typical IC-->ID argument. To wit, "the system must have all of the parts before it will be functional" or versions thereof. Multiple-parts-required is also easily operationalized and unambiguously identified, e.g. by knockout experiments. And further, it is easily applied to mousetraps as well as things like computer programs.
So, IMO the way to fix C3 and C4, and probably C1 and C2, is simply to identify IC with "multiple-parts-required" once-and-for-all.
C5 is a lesser issue, but: what constitutes a "part"? In DBB, Behe tends to emphasize what I would call "logical parts" -- a propulsion motor needs a filament, stator, and rotor, etc. However, the current standard is usually to identify proteins as parts -- thus the flagellar stator is made up of 2 protein types, the filament of many, etc. I say this is a lesser issue as usually protein parts can be hierarchically fit into different logical parts, and as proteins are a fairly natural unit it seems fair to use them.
OTOH, evolutionarily speaking it may be the "logical parts" which are more important to consider. Consider the filament: as far as we know (I think), all bacterial flagella make use of a hook protein. But is a hook protein absolutely necessary from a logical standpoint? For example, if one had a spherical cell with one flagellum, couldn't it get by with just a rigid helical filament attached directly to the rod? No u-joint (Universal joint) is logically needed if the filament can function perpendicular to the cell membrane. A related concern comes from the archaeal flagella: they have similar logical parts and the same function, but all of the proteins are completely different. So are proteins the right thing to call "parts"? Still not a huge issue, but worth considering.
Where I get concerned however is when parts-counts are artifically inflated, e.g. when partial chunks of proteins start getting called separate "parts".
IP: Logged
|
|
yersinia
Member
Member # 324
|
posted 30. May 2003 00:39
C7: IC is, properly speaking, a scale-invariant concept. This is shown by Behe's usage of molecular (flagella), microscopic (cilia; although some Drosophila sperm cilia are several centimeters long), and macroscopic (mousetraps, bicycles, rube goldberg machines, snares made out of vines, etc.) examples.
However, when it comes to biology, a "molecular systems only" rule is sometimes invoked (not always; plenty of pro-ID people call hearts, kneecaps, etc. IC). This has the effect of ruling out most of the evolutionist's long-studied, best examples (middle ear bones, feathers, eyes, etc.), and therefore has an obvious advantage in debates, but I don't think it is supportable. Consider the following paradox:
A standard snap mousetrap is IC, because Behe says so if nothing else.
Let's say that we replace all of the parts of the mousetrap with parts made of cells (the wooden base is already made of cells anyhow, as are the vines in Behe's snares in DBB). It has a superimposable shape and functions in the same fashion. Is this made-of-cells mousetrap suddenly not IC? This is what the molecular-systems-only rule would imply, even though the made-of-cells mousetrap works or doesn't work, based on removal of parts, in exactly the same fashion as the regular mousetrap.
IP: Logged
|
|
Alix Nenuphar
Member
Member # 686
|
posted 30. May 2003 01:35
C8: IC as currently defined appears to lack application.
IC is defined it terms of the current function of the biological structure. It is unclear whether any information regarding current function and component dependency may legitimately be extrapolated to precursor structures.
IP: Logged
|
|
Rex Kerr
Member
Member # 632
|
posted 30. May 2003 05:54
This complaint is relevant only to the "Behe original" and "Dembski enhanced" versions of IC; the evolutionary IC is a markedly different proposition.
C9: The utility of IC is hampered by the asymmetry of evolutionary processes. In particular, gain of function and loss of function are not symmetric; gain is much rarer than loss. This is critical because IC systems are those which are "well-matched" in the sense that all pieces are required to function. Yet an obvious way to generate a well-matched system is to start with a poorly-matched system and throw things out until everything that remains is essential--this is loss, or degeneration, from a prior reducible state, and because of the asymmetry of evolution, we expect this to be "easy" (whereas regenerating this prior system would require gain of function and therefore be "hard"). Hence, identifying a system as IC leads us to conclude either that it was very improbable for it to spring forth all at once by chance; or that there was a prior poorly matched system that performed the same function. Since IC contains no tools for either judging the probability of generating the poorly matched system, or for judging the probability of generating a well-matched system from a poorly-matched one, it is of very limited utility.
IP: Logged
|
|
RBH
Member
Member # 380
|
posted 30. May 2003 08:37
For reference, given Rex's (appropriate) differentation of the several definitions of IC, here are the entries from the ISCID Beta version of its Encyclopedia: quote:
Irreducible Complexity
Michael Behe's Original Definition:
A single system composed of several well-matched, interacting parts that contribute to the basic function of the system, wherein the removal of any one of the parts causes the system to effectively cease functioning. (Darwin's Black Box, 39)
William Dembski's Enhanced Definition:
A system performing a given basic function is irreducibly complex if it includes a set of well-matched, mutually interacting, nonarbitrarily individuated parts such that each part in the set is indispensable to maintaining the system's basic, and therefore original, function. The set of these indispensable parts is known as the irreducible core of the system. (No Free Lunch, 285)
Michael Behe's "Evolutionary" Definition
An irreducibly complex evolutionary pathway is one that contains one or more unselected steps (that is, one or more necessary-but-unselected mutations). The degree of irreducible complexity is the number of unselected steps in the pathway.
C10 (Dembski's Enhanced Definition) "the system's basic, and therefore original, function" is sometimes interpreted historically ('original' = whatever the system did when it [and/or its component subsystems?] first appeared) and sometimes ahistorically ('basic' = the system's [primary?] current function), and there is no clear guidance on which is intended.
C11 "Well-matched" is a subjective aesthetic judgement. It is easy to conceive of examples in which the parts are ill-matched by some aesthetic criteria yet are IC by the knockout criterion.
RBH
[ 30. May 2003, 08:40: Message edited by: RBH ]
IP: Logged
|
|
|
Printer-friendly view of this topic