Prof. Massimo Pigliucci
SUNY- Stony Brook – Stony Brook, NY
A new brand of creationism has appeared on the scene in the last few
years. The so-called neocreationists largely do not believe in a young Earth or
in a too literal interpretation of the Bible. While still mostly propelled by a
religious agenda and financed by mainly Christian sources such as the Templeton
Foundation and the Discovery Institute, the intellectual challenge posed by
neocreationism is sophisticated enough to require detailed consideration (see
Edis 2001; Roche 2001).
Among the chief exponents of Intelligent Design (ID) theory, as this new
brand of creationism is called, is William Dembski, a mathematical philosopher
and author of The Design Inference (1998a)
. In that book he attempts to show that there must be an intelligent
designer behind natural phenomena such as evolution and the very origin of the
universe (see Pigliucci 2000 for a
detailed critique) . Dembki’s (1998b)
argument is that modern
science ever since Francis Bacon has illicitly dropped two of Aristotle’s
famous four types of causes from consideration altogether, thereby unnecessarily
restricting its own explanatory power. Science is thus incomplete, and
intelligent design theory will rectify this sorry state of affairs, if only
close-minded evolutionists would allow Dembski and co. to do the job.
Aristotle’s four causes in science
Aristotle identified material causes, what something is made of; formal
causes, the structure of the thing or phenomenon; efficient causes, the
immediate activity producing a phenomenon or object; and final causes,
the purpose of whatever object we are investigating. For example, let’s say we
want to investigate the “causes” of the Brooklyn Bridge. Its material cause
would be encompassed by a description of the physical materials that went into
its construction. The formal cause is the fact that it is a bridge across a
stretch of water, and not either a random assembly of pieces or another kind of
orderly structure (such as a skyscraper). The efficient causes were the
blueprints drawn by engineers and the labor of men and machines that actually
assembled the physical materials and put them into place. The final cause of the
Brooklyn Bridge was the necessity for people to walk and ride between two
landmasses while avoiding to get wet.
Dembski maintains that Bacon and his followers did away with both formal
and final causes (the so-called teleonomic causes, because they answer the
question of why something is) in order to free science from philosophical
speculation and ground it firmly into empirically verifiable statements. That
may be so, but things certainly changed with the work of Charles Darwin (1859)
. Darwin was addressing a complex scientific question in an unprecedented
fashion: he recognized that living organisms are clearly designed in order to
survive and reproduce in the world they inhabit; yet, as a scientist, he worked
within the framework of naturalistic explanations of such design. Darwin found
the answer in his well-known theory of natural selection. Natural selection,
combined with the basic process of mutation, makes design possible in nature
without recourse to a supernatural explanation because selection is definitely
non-random, and therefore has “creative” (albeit non-conscious) power.
Creationists usually do not understand this point and think that selection can
only eliminate the less fit; but Darwin’s powerful insight was that selection
is also a cumulative process – analogous to a ratchet – which can build
things over time, as long as the intermediate steps are also advantageous.
Darwin made it possible to put all four Aristotelian causes into science.
For example, if we were to ask what are the causes of a tiger’s teeth within a
Darwinian framework, we would answer in the following manner. The material cause
is provided by the biological materials that make up the teeth; the formal cause
is the genetic and developmental machinery that distinguishes a tiger’s teeth
from any other kind of biological structure; the efficient cause is natural
selection promoting some genetic variants of the tiger’s ancestor over their
competitors; and the final cause is provided by the fact that having teeth
structured in a certain way makes it easier for a tiger to procure its prey and
therefore to survive and reproduce – the only “goals” of every living
Therefore, design is very much a part of modern science, at least whenever there is a need to explain an apparently designed structure (such as a living organism). All four Aristotelian causes are fully reinstated within the realm of scientific investigation, and science is not maimed by the disregard of some of the causes acting in the world. What then is left of the argument of Dembski and of other proponents of ID? They, like William Paley (1831) well before them, make the mistake of confusing natural design and intelligent design by rejecting the possibility of the former and concluding that any design must by definition be intelligent.
One is left with the lingering feeling that Dembski is being disingenuous about ancient philosophy. It is quite clear, for example, that Aristotle himself never meant his teleonomic causes to imply intelligent design in nature (Cohen 2000) . His mentor, Plato (in Timaeus), had already concluded that the designer of the universe could not be an omnipotent god, but at most what he called a Demiurge, a lesser god who evidently messes around with the universe with mixed results. Aristotle believed that the scope of god was even more limited, essentially to the role of prime mover of the universe, with no additional direct interaction with his creation (i.e., he was one of the first deists). In Physics, where he discusses the four causes, Aristotle treats nature itself as a craftsman, but clearly devoid of forethought and intelligence. A tiger develops into a tiger because it is in its nature to do so, and this nature is due to some physical essence given to it by its father (we would call it DNA) which starts the process out. Aristotle makes clear this rejection of god as a final cause (Cohen 2000) when he says that causes are not external to the organism (such as a designer would be) but internal to it (as modern developmental biology clearly shows). In other words, the final cause of a living being is not a plan, intention, or purpose, but simply intrinsic in the developmental changes of that organism. Which means that Aristotle identified final causes with formal causes as far as living organisms are concerned. He rejected chance and randomness (as do modern biologists) but did not invoke an intelligent designer in its place, contra Dembski. We had to wait until Darwin for a further advance on Aristotle’s conception of the final cause of living organisms and for modern molecular biology to achieve an understanding of their formal cause.
There are two additional arguments proposed by ID theorists to
demonstrate intelligent design in the universe: the concept of “irreducible
complexity” and the “complexity-specification” criterion. Irreducible
complexity is a term introduced in this context by molecular biologist Michael
Behe in his book Darwin’s Black Box (1996)
. The idea is that the difference between a natural phenomenon and an
intelligent designer is that a designed object is planned in advance, with
forethought. While an intelligent agent is not constrained by a step-by-step
evolutionary process, the latter is the only way nature itself can proceed given
that it has no planning capacity (this may be referred to as incremental
complexity). Irreducible complexity then arises whenever all the parts of a
structure have to be present and functional simultaneously for it to work,
indicating – according to Behe – that the structure was designed and could
not possibly have been gradually built by natural selection.
Behe’s example of an irreducibly complex object is a mousetrap. If you
take away any of the minimal elements that make the trap work it will loose its
function; on the other hand, there is no way to assemble a mousetrap gradually
for a natural phenomenon, because it won’t work until the last piece is
assembled. Forethought, and therefore intelligent design, is necessary. Of
course it is. After all, mousetraps as purchased in hardware stores are indeed
human products; we know that they are intelligently designed. But what of
biological structures? Behe claims that, while evolution can explain a lot of
the visible diversity among living organisms, it is not enough when we come to
the molecular level. The cell and several of its fundamental components and
biochemical pathways are, according to him, irreducibly complex.
problem with this statement is that it is contradicted by the available
literature on comparative studies in microbiology and molecular biology, which
Behe conveniently ignores (Miller
1996) . For example, geneticists are continuously showing that
biochemical pathways are partly redundant. Redundancy is a common feature of
living organisms where different genes are involved in the same or in partially
overlapping functions. While this may seem a waste, mathematical models show
that evolution by natural selection has to produce molecular redundancy because
when a new function is necessary it cannot be carried out by a gene that is
already doing something else, without compromising the original function. On the
other hand, if the gene gets duplicated (by mutation), one copy is freed from
immediate constraints and can slowly diverge in structure from the original,
eventually taking over new functions. This process leads to the formation of
gene “families,” groups of genes clearly originated from a single ancestral
DNA sequence, and that now are diversified and perform a variety of functions
(e.g., the globins, which vary from proteins allowing muscle contraction to
those involved in the exchange of oxygen and carbon dioxide in the blood). As a
result of redundancy, mutations can knock down individual components of
biochemical pathways without compromising the overall function – contrary to
the expectations of irreducible complexity. (Notice
that creationists, never ones to loose a bit, have also tried to claim that
redundancy is yet another evidence of intelligent design, because an engineer
would produce backup systems to minimize catastrophic failures should the
primary components stop functioning. While very clever, this argument once again
ignores the biology: the majority of duplicated genes end up as pseudogenes,
literally pieces of molecular junk that are eventually lost forever to any
biological utility: Max 1986)
be sure, there are several cases in which biologists do not know enough about
the fundamental constituents of the cell to be able to hypothesize or
demonstrate their gradual evolution. But this is rather an argument from
ignorance, not positive evidence of irreducible complexity. William Paley
advanced exactly the same argument to claim that it is impossible to explain the
appearance of the eye by natural means. Yet, today biologists know of several
examples of intermediate forms of the eye, and there is evidence that this
structure evolved several times independently during the history of life on
earth (Gehring and Ikeo 1999) .
The answer to the classical creationist question, “what good is half an
eye?” is “much better than no eye at all”!
Behe does have a point concerning irreducible complexity. It is true that some
structures simply cannot be explained by slow and cumulative processes of
natural selection. From his mousetrap to Paley’s watch to the Brooklyn Bridge,
irreducible complexity is indeed associated with intelligent design. The problem
for ID theory is that there is no evidence so far of irreducible complexity in
Dembski uses an approach similar to Behe to back up creationist claims, in that
he also wants to demonstrate that intelligent design is necessary to explain the
complexity of nature. His proposal, however, is both more general and more
deeply flawed. In his book The Design Inference (Dembski
1998a) he claims that there are three essential types of phenomena in
nature: “regular”, random, and designed (which he assumes to be
intelligent). A regular phenomenon would be a simple repetition explainable by
the fundamental laws of physics, for example the rotation of the earth around
the sun. Random phenomena are exemplified by the tossing of a coin. Design
enters any time that two criteria are satisfied: complexity and specification (Dembski
are several problems with this neat scenario. First of all, leaving aside design
for a moment, the remaining choices are not limited to regularity and
randomness. Chaos and complexity theory have established the existence of
self-organizing phenomena (Kauffman
1993; Shanks and Joplin 1999) , situations in which order spontaneously
appears as an emergent property of complex interactions among the parts of a
system. And this class of phenomena, far from being only a figment of
mathematical imagination as Behe maintains, are real. For example, certain
meteorological phenomena such as tornados are neither regular nor random but are
the result of self-organizing processes.
But let us go back to complexity-specification and take a closer look at
these two fundamental criteria, allegedly capable of establishing intelligent
agency in nature. Following one of Dembski’s examples, if SETI (Search for
Extra Terrestrial Intelligence) researchers received a very short signal that
may be interpreted as encoding the first three prime numbers, they would
probably not rush to publish their findings. This is because even though such
signal could be construed as due to some kind of intelligence, it is so
short that its occurrence can just as easily be explained by chance. Given the
choice, a sensible scientist would follow Ockham’s razor and conclude that the
signal does not constitute enough evidence for ET. However, also according to
Dembski, if the signal were long enough to encode all the prime numbers between
2 and 101, the SETI people would open the champagne and celebrate all night.
Why? Because such signal would be both too complex to be explained by chance and
would be specifiable, meaning that it is not just a random sequence of numbers,
it is an intelligible message.
specification criterion needs to be added because complexity by itself is a
necessary but not sufficient condition for design (Roche 2001). To see this,
imagine that the SETI staff receives a long but random sequence of signals. That
sequence would be very complex, meaning that it would take a lot of information
to actually archive or repeat the sequence (you have to know where all the 0s
and 1s are), but it would not be specifiable because the sequence would be
Dembski is absolutely correct that plenty of human activities, such as
SETI, investigations into plagiarism, or encryption, depend on the ability to
detect intelligent agency. Where he is wrong is in assuming only one kind of
design: for him design equals intelligence and, even though he admitted that
such an intelligence may be an advanced extraterrestrial civilization, his
preference is for a god, possibly of the Christian variety.
The problem is that natural selection, a natural process, also fulfills
the complexity-specification criterion, thereby demonstrating that it is
possible to have unintelligent design in nature. Living organisms are indeed
complex. They are also specifiable, meaning that they are not random assemblages
of organic compounds, but are clearly formed in a way that enhances their
chances of surviving and reproducing in a changing and complex environment.
What, then distinguishes organisms from the Brooklyn Bridge? Both meet
Dembski’s complexity-specification criterion, but only the bridge is
irreducibly complex. This has important implications for design.
In response to some of his critics, Dembski (2000)
claimed that intelligent design does not mean optimal design. The
criticism of suboptimal design has often been advanced by evolutionists who ask
why God would do such a sloppy job with creation that even a mere human engineer
can easily determine where the flaws are. For example, why is it that human
beings have hemorrhoids, varicose veins, backaches, and foot aches? If you
assume that we were “intelligently” designed, the answer must be that the
designer was rather incompetent – something that would hardly please a
creationist. Instead, evolutionary theory has a single answer to all these
questions: humans evolved bipedalism (walking with an erect posture) only very
recently, and natural selection has not yet fully adapted our body to the new
condition (Olshansky et al 2001). Our closest primate relatives, chimps,
gorillas and the like, are better adapted to their way of life, and therefore
are less “imperfect” than ourselves!
Dembski is of course correct in saying that intelligent design does not
mean optimal design. As much as the Brooklyn Bridge is a marvel of engineering,
it is not perfect, meaning that it had to be constructed within the constraints
and limitations of the available materials and technology, and it still is
subject to natural laws and decay. The bridge’s vulnerability to high winds
and earthquakes, and its inadequacy to bear a volume of traffic for which it was
not built can be seen as similar to the back pain caused by our recent
evolutionary history. However, the imperfection of living organisms, already
pointed out by Darwin, does do away with the idea that they were created by an
omnipotent and omnibenevolent creator, who surely would not be limited by laws
of physics that he himself made up from scratch.
The four fundamental types of design and how to recognize them
Figure 1: the four kinds of design in nature and
how to distinguish them.
The first kind of design is non-intelligent-natural, and it is
exemplified by natural selection within earth’s biosphere (and possibly
elsewhere in the universe). The results of this design, such as all living
organisms on earth, are not irreducibly complex, meaning that they can be
produced by incremental, continuous (though not necessarily gradual) changes
over time. These objects can be clearly attributed to natural processes also
because of two other reasons: they are never optimal (in an engineering sense)
and they are clearly the result of historical processes. For example, they are
full of junk, non utilized or under-utilized parts, and they resemble similar
objects occurring simultaneously or previously in time (see, for example, the
fossil record). Notice that some scientists and philosophers of science feel
uncomfortable in considering this “design” because they equate the term with
intelligence. But I do not see any reason to embrace such limitation. If
something is shaped over time – by whatever means – such that it fulfills a
certain function, then it is designed and the question is simply of how such
design happened to materialize. The teeth of a tiger are clearly designed to
efficiently cut into the flesh of its prey and therefore to promote survival and
reproduction of tigers bearing such teeth.
The second type of design is intelligent-natural. These artifacts
are usually irreducibly complex, such as a watch designed by a human. They are
also not optimal, meaning that they clearly compromise between solutions to
different problems (trade-offs) and they are subject to the constraints of
physical laws, available materials, expertise of the designer, etc. Humans may
not be the only ones to generate these objects, as the artifacts of any
extraterrestrial civilization would fall into the same broad category.
The third kind of design, which is difficult, if not impossible, to
distinguish from the second, is what I term intelligent-supernatural-sloppy.
Objects created in this way are essentially indistinguishable from human or ET
artifacts, except that they would be the result of what the Greeks called a
Demiurge, a minor god with limited powers. Alternatively, they could be due to
an evil omnipotent god that just amuses himself with suboptimal products. The
reason intelligent-supernatural-sloppy design is not distinguishable from
some instances (but by all means not all) of intelligent-natural design
is Arthur C. Clark’s famous third law: from the point of view of a
technologically less advanced civilization, the technology of a very advanced
civilization is essentially indistinguishable from magic (such as the monolith
in his 2001: a Space Odyssey). I would be very interested if someone
could suggest a way around Clark’s law.
Finally, we have intelligent-supernatural-perfect design, which is
the result of the activity of an omnipotent and omnibenevolent god. These
artifacts would be both irreducibly complex and optimal. They would not be
constrained by either trade-offs or physical laws (after all, god created the
laws themselves). While this is the kind of god many Christian fundamentalists
believe in (though some do away with the omnibenevolent part), it’s quite
clear from the existence of human evil as well as of natural catastrophes and
diseases, that such god does not exist. Dembski recognizes this difficulty and,
as I pointed out above, admits that his intelligent design could even be due to
a very advanced extraterrestrial civilization, and not to a supernatural entity
at all (Dembski 2000) .
In summary, it seems to me that the major arguments of Intelligent Design theorists are neither new nor compelling. a) It is simply not true that science does not address all Aristotelian causes, whenever design needs to be explained. b) While irreducible complexity is indeed a valid criterion to distinguish between intelligent and non-intelligent design, these are not the only two possibilities, and living organisms are not irreducibly complex (e.g., see Shanks and Joplin 1999) . c) The complexity-specification criterion is actually met by natural selection, and cannot therefore provide a way to distinguish intelligent from non-intelligent design. d) If supernatural design exists at all (but where is the evidence or compelling logic?), this is certainly not of the kind that most religionists would likely subscribe to, and it is indistinguishable from the technology of a very advanced civilization.
Therefore, Behe’s, Dembski’s, and other creationists’ (e.g., Johnson 1997) claims that science should be opened to supernatural explanations and that these should be allowed in academic as well as public school curricula is unfounded and based on a misunderstanding of both design in nature and of what the neo-Darwinian theory of evolution (Mayr and Provine 1980) is all about.
Acknowledgments. I would like to thank Melissa Brenneman, Will Provine and Niall Shanks for insightful comments on earlier versions of this article, as well as Michael Behe, William Dembski, Ken Miller and Barry Palevitz for indulging in correspondence and discussions with me over these matters.
M.J. (1996) Darwin's black box. The
biochemical challenge to evolution. Free Press, New York.
S.M. (2000) The four causes. Accessed on 5/16/00, web page,
C. (1859) The origin of species by means
of natural selection: or, the preservation of favored races in the struggle for
life. A.L. Burt (1910), New York, NY.
W.A. (1998a) The design inference.
Cambridge University Press, Cambridge, UK.
W.A. (1998b) Reinstating design within science. Rhetoric & Public Affairs
Dembski, W.A. (2000) Intelligent design is not optimal design. Accessed on 2/3/00, web page, www.meta-list.org.
T. (2001) Darwin in mind: Intelligent Design meets artificial intelligence.
Skeptical Inquirer 25(2): 35-39
W.J. and Ikeo, K. (1999) Pax 6,
mastering eye morphogenesis and eye evolution. Trends in Genetics 15:371-377.
P. (1997) Defeating Darwinism by opening
minds. InterVarsity Press, Downers Grove, IL.
S.A. (1993) The origins of order.
Oxford University Press, New York.
E.E. (1986) Plagiarized errors and molecular genetics: another argument in the
evolution-creation controversy. Creation/Evolution 9:34-46.
E. and Provine, W.B. (1980) The
evolutionary synthesis. Perspectives on the unification of biology. Harvard
University Press, Cambridge, MA.
K.R. (1996) The biochemical challenge to evolution. Accessed on 10/30/99, web
S.J., Bruce A.C. and Butler, R.N. (2001) If Humans Were Built to Last.
Scientific American, March.
W. (1831) Natural theology: or, Evidences
of the existence and attributes of the Deity, collected from the appearances of
nature. Gould, Kendall, and Lincoln, Boston, MA.
M. (2000) Chance, necessity, and the new holy war against science. A review of
W.A. Dembski's The Design Inference.
BioScience 50 1, January, pp. 79-81.
D. (2001) A bit confused: creationism and information theory. Skeptical Inquirer
N. and Joplin, K.H. (1999) Redundant complexity: a critical analysis of
intelligent design in biochemistry. Philosophy of Science 66:268-282.