Naturalism and design WILLIAM DEMBSKI
66; J. P. Moreland, Stephen C. Meyer, and Richard H. Bube, “Conceptual Problems and the Scientific Status of Creation Science: a Discussion,” Perspectives on Science and Christian Faith
46 (1994): 2–25.
102 Smith, “Uncaused Beginning of the Universe,” 135. Elsewhere he has written,
[This world] exists nonnecessarily, improbably, and causelessly. It exists for absolutely no reason at all. . . . The impact of this captivated realization upon me is overwhelming. I am completely stunned. I take a few dazed steps in the dark meadow, and fall among the flowers. I lie stupefied, whirling without comprehension in this world through numberless worlds other than this one” (Quentin Smith, The Felt Meanings of the World [Lafayette, Ind.: Purdue University Press, 1986], 300–1).
In Theism, Atheism, and Big Bang Cosmology, Smith claimed that the universe came into being uncaused out of nothing at the Planck time; but he has since recanted that position under the realization that the whole field of quantum cosmology is then studying a complete fiction!
103 Smith, “Uncaused Beginning of the Universe,” 135.
104 Peter W. Atkins, Creation Revisited (New York: W. H. Freeman, 1992), 129, 149– 51.
105 John Leslie asks incredulously, “How could such nonsense have been churned out by the author of Physical Chemistry, a superb textbook?” (John Leslie, “Is It All Quite Simple?,” Times Literary Supplement, 29 January, 1993, 3). For a good critique of Atkins, see Keith Ward, God, Chance, and Necessity (Oxford: One World, 1996), chap. 1.
106 John Gribbin, In Search of the Big Bang (New York: Bantam Books, 1986), 374.
107 Isham, “Quantum Cosmology and the Origin of the Universe,” 8.
108 See Robert Weingard, “Do Virtual Particles Exist?” in Proceedings of the Philosophy of Science Association, 2 vols., ed. Peter Asquith and Thomas Nichols (East Lansing, Mich.: Philosophy of Science Association, 1982), 1: 235–42.
109 Kanitscheider, “Physical Cosmology,” 346–7.
110 John Post, Metaphysics: a Contemporary Introduction (New York: Paragon House, 1991), 85.
111 Ibid., 87.
112 Smith, “Uncaused Beginning of the Universe,” 120.
113 Adolf Grünbaum, “The Pseudo-Problem of Creation in Physical Cosmology,” Philosophy of Science 56 (1989): 373–94. For a response, see William Lane Craig, “The Origin and Creation of the Universe: a reply to Adolf Grünbaum,” British Journal for the Philosophy of Science 43
(1992): 233–40.
114 Adolf Grünbaum, “Creation as a Pseudo-Explanation in Current Physical Cosmology,”
Erkenntnis 35 (1991): 233–54. For a response, see William Lane Craig, “Prof. Grünbaum on Creation,” Erkenntnis 40 (1994): 325–41.
115 Brian Leftow puts this nicely when he writes,
If God existed in time once time existed and time had a first moment, then God would have a first moment of existence: there would be a moment before which He did not exist, because there was no ‘before’ that moment. . . . Yet even if He . . . had a first moment of existence, one could still call God’s existence unlimited were it understood that He would have existed even if time did not. For as long as this is true, we cannot infer from God’s having had a first moment of existence that God came into existence or would not have existed save if time did (Brian Leftow, Time and Eternity, Cornell Studies in Philosophy of Religion [Ithaca, N.Y.: Cornell University Press, 1991], 269; cf. 201).
Senor has dubbed such a model of divine eternity “accidental temporalism” (Thomas D.
Senor, “Divine Temporality and Creation ex nihilo,” Faith and Philosophy 10 [1993]: 88). See
further William Lane Craig, “Timelessness and Creation,” Australasian Journal of Philosophy 74
(1996): 646–56.
116 Adolf Grünbaum, “Some Comments on William Craig’s ‘Creation and Big Bang Cosmology’,” Philosophia naturalis 31 (1994): 225–36. For a response, see William Lane Craig, “A Response to Grünbaum on Creation and Big Bang Cosmology,” Philosophia naturalis 31 (1994): 237–49.
117 See discussion and references in William Lane Craig, “Divine Timelessness and Personhood,” International Journal for Philosophy of Religion 43 (1998): 109–24.
118 John C. Yates, The Timelessness of God (Lanham, Md.: University Press of America, 1990), 173.
119 Quentin Smith, “Causation and the Logical Impossibility of a Divine Cause,” Philosophical Topics 24 (1996): 169–70.
120 In Quentin Smith, “The Concept of a Cause of the Universe,” Canadian Journal of Philosophy 23 (1993): 1–24, Smith actually arrives at this conclusion himself. He states, extant definitions of causality are incorrect since they do not cohere in the proper way with the concept of a cause of the universe . . . .
This entails that either there is some other (as yet undiscovered) definition of a cause that is correct or that a cause is indefinable. In the latter case, the concept of a cause would be primitive and the causal relation a simple relation known only by ostension (as is arguably the case with such relations as being in contact with or being earlier than). I know of no means of discovering or formulating a correct definition of a cause and know of no reason to think that there is such a definition. Accordingly, I think it is reasonable to conclude that the causal relation is indefinable.
One way to avoid this conclusion would be to reject the assumption that the various examples of causes of the big bang . . . are genuine examples of causes. . . . I would say that claims that God’s creation of the big bang singularity and other examples given . . . are not cases of possible causation are counterintuitive and are ad hoc attempts to retain a counterexampled theory. It is more plausible to think that a cause cannot be defined than to think that a mind’s creation of a big bang singularity could not be a causal act (Ibid., 1, 24).
Smith came to think God’s relation to the Big Bang is not causal because no cause is logically sufficient for its effect. But Smith does not justify why the actions of an omnipotent being would not be exceptions to this rule.
121 See J. P. Moreland, “Libertarian Agency and the Craig/Grünbaum Debate about Theistic Explanation of the Initial Singularity,” American Catholic Philosophical Quarterly 81 (1998): 539–54. I am indebted to my colleague for several interesting discussions pertinent to agency and creation.
122 Smith, “Logical Impossibility of a Divine Cause,” 176.
123 Sydney Shoemaker, “Time without Change,” Journal of Philosophy 66 (1969): 363–81.
124 David Hume to John Stewart, February, 1754, in The Letters of David Hume, 2 vols., ed. J. Y. T.
Greig (Oxford: Clarendon Press, 1932), 1: 187.
125 Thomas Aquinas Summa theologiae 1a.2.3.
William Dembski
1. Introduction
Naturalism has become the default position for all serious inquiry. From the humanities to the natural sciences to biblical studies to the arts, inquiry is allowed to proceed only under the supposition of naturalism, the view that nature is self-sufficient. Naturalism is not atheism. To affirm that nature is self-sufficient is not to deny God’s existence. God could, after all, have created the world to be self-sufficient.
Nonetheless, in the interest of keeping human inquiry productive, naturalism insists that the existence and activity of God be bracketed. Naturalism affirms not so much that God does not exist as that God need not exist. And because God need not exist, intellectual honesty demands that we get about the serious work of life without invoking him. This is the received wisdom.
Naturalism comes in a variety of forms. Its principal form in Western culture is known as scientific naturalism. Scientific naturalism locates the self-sufficiency of nature in the undirected natural causes studied by science. Accordingly, scientific naturalism would have us understand the universe entirely in terms of such causes. In particular, since human beings are a part of the universe, who we are and what we do must ultimately be understood in naturalistic terms. This is not to deny our humanity.
But it is to interpret our humanity as the consequence of brute material processes that were not consciously aiming at us. Nor is this to deny God. But it is to affirm that if God exists, he was marvelously adept at covering his tracks and giving no empirical evidence of ever interacting with the world. To be sure, it remains logically permissible for the scientific naturalist to affirm God’s existence, but only by making God a superfluous rider on top a self-contained account of the world.
If naturalism is false, how could we know it? In the last thirty years, evidence from cosmology has suggested that a designing intelligence ordered the universe. With the rise of Big Bang cosmology, the discovery of anthropic coincidences, and the proliferation of fine-tuning arguments, the existence of a designing intelligence has gained plausibility. This same evidence, however, has been interpreted naturalistically. Stephen Hawking, for instance, seeks a theory of everything that omits
all contingency and thereby eliminates any need for fine-tuning.1 Alan Guth’s inflationary cosmology attempts to wash out the need for fine-tuning by giving the early universe a brief but explosive period of exponential growth.2 John Barrow and Frank Tipler show that by positing an ensemble of universes one can explain the design of the universe as a selection effect.3 Because the ensemble contains so many universes, some are bound to be interesting and include rational agents like ourselves.
That we exist and live in an apparently designed universe may surprise us, but is fully to be expected simply by chance given a large enough ensemble of universes.
Deciding whether the universe as a whole is designed always requires additional metaphysical assumptions. Thus, to reject design, Hawking needs to assume an underlying necessity to the laws of nature, Guth needs to assume “eternal inflation”
(making inflation a metaphysical principle for universe construction), and Barrow and Tipler need to assume an ensemble of universes that are causally isolated from our own.4 On the other hand, to accept design, theists like Hugh Ross begin with a prior belief in God and then formulate arguments to show how the fine-tuning of the universe actually does support design.5 Now I don’t mean to suggest there is no principled way to distinguish among metaphysical assumptions that alternately support or undercut the design of the universe. William Lane Craig and Richard Swinburne, for instance, have argued convincingly for the strength of theism as a metaphysical position in relation to cosmology.6
My interest in this paper, however, is different. Certainly I want to answer the question, If naturalism is false, how could we know it? But in answering this question I will consider not whether the universe as a whole is designed, but rather whether we are able to detect design within an already given universe. The universe provides a well-defined causal backdrop (many physicists think of it as a field characterized by field equations). Although one can ask whether that causal backdrop is itself designed, one can as well ask whether events and objects occurring within that backdrop exhibit features that convincingly implicate design. At issue here are two types of design: (1) the design of the universe as a whole and (2) instances of design within the universe.
An analogy illustrates the difference. Consider an oil painting. An oil painting is typically painted on a canvas. One can therefore ask whether the canvas is designed.
Alternatively, one can ask whether some configuration of paint on the canvas is designed. The design of the canvas corresponds to the design of the universe as a whole. The design of some configuration of paint corresponds to an instance of design within the universe.
Though not perfect, this analogy is useful. The universe is a canvas on which is depicted natural history. One can ask whether that canvas itself is designed. To decide this question, however, presupposes a serious investment in metaphysics. On the other hand, one can ask whether features of natural history depicted on that canvas are designed. In biology, for instance, one can ask whether Michael Behe’s irreducibly
complex biochemical machines are designed.7 Although metaphysical assumptions are not irrelevant here, they play a much weaker role than they do in cosmology. This is only to be expected. The origin and early evolution of the universe remain highly speculative. On the other hand, the causal backdrop for life is well-defined. The nuts and bolts of life occur at the level of biochemistry. Biochemistry provides the causal backdrop against which design in biology must be decided. That backdrop is well-defined and subject to far less speculation than the origin and early evolution of the universe.
The aim of this chapter, then, is to show how detecting design within the universe, and especially within biology, challenges naturalism. To do this, I shall review my own work on detecting design. I shall show that detecting design within the universe follows a well-defined methodology. Moreover, when applied to the irreducibly complex biochemical systems of Michael Behe, this methodology convincingly implicates design. I conclude that design within the universe renders naturalism untenable.
2. Admitting design into science
When the physics of Galileo and Newton displaced the physics of Aristotle, deterministic natural laws became the preferred mode of scientific explanation.
When the physics of Bohr and Heisenberg in turn displaced the physics of Galileo and Newton, the preferred mode of scientific explanation came to include both deterministic natural laws and chance processes. Chance and necessity, to use a phrase made famous by Jacques Monod, henceforth set the boundary for scientific explanation.
That boundary needs now to be redrawn again. Chance and necessity have proven too thin an explanatory soup on which to nourish a robust science. Without invoking the rightly discarded teleologies, entelechies, and vitalisms of the past, one can still see that chance and necessity together do not span the full range of scientific explanation.
A third mode of explanation is required, namely, design. Chance, necessity and design
– these three modes of explanation are needed to properly span the full range of scientific explanation.
Many scientists tacitly admit this point. For instance, Richard Dawkins begins his book The Blind Watchmaker by stating, “Biology is the study of complicated things that give the appearance of having been designed for a purpose.”8 Statements like this echo throughout the biological literature. In What Mad Pursuit, Francis Crick writes,
“Biologists must constantly keep in mind that what they see was not designed, but rather evolved.”9 Granted, the biological community thinks it has accounted for the apparent design in nature apart from any actual design (typically through the Darwinian mechanism of mutation and selection). The point to appreciate, however,
is that in accounting for the apparent design in nature, biologists regard themselves as having made a successful scientific argument against actual design.
Scientific refutation is always a double-edged sword. Claims that are refuted scientifically may be wrong, but they are not necessarily wrong. Alternatively, for a claim to be scientifically falsifiable, it must have the possibility of being true. To see this, consider what would happen if microscopic examination revealed that every cell was inscribed with the phrase “Made by Yahweh.” Of course cells do not have “Made by Yahweh” inscribed on them, but that is not the point. The point is that we would not know this unless we actually looked at cells under the microscope.
Design always remains a live option in biology. A priori prohibitions against design are easily countered, especially in an age of diversity and multiculturalism where it is all too easy to ask, Who sets the rules for science anyway? Nonetheless, once we admit that design cannot be excluded from science on first principles, a weightier question remains: Why should we want to admit design into science? To answer this question, let us turn it around and ask instead: Why shouldn’t we want to admit design into science? What’s wrong with explaining something as designed by an intelligent agent?
Certainly there are many everyday occurrences which we explain by appealing to design. Moreover, in our workaday lives it is absolutely crucial to distinguish accident from design. We demand answers to such questions as: Did she fall or was she pushed?
Did someone die accidentally or commit suicide? Was this song conceived independently or was it plagiarized? Did someone just get lucky on the stock market or was there insider trading?
Not only do we demand answers to such questions, but entire industries are devoted to drawing the distinction between accident and design. Here we can include forensic science, intellectual property law, insurance claims investigation, cryptography, and random number generation – to name but a few. Science itself needs to draw this distinction to keep itself honest. In a recent issue of Science, a Medline websearch uncovered a “paper published in Zentralblatt für Gynäkologie in 1991 [containing] text that is almost identical to text from a paper published in 1979
in the Journal of Maxillofacial Surgery. ”10 Plagiarism and data falsification are far more common in science than we would like to admit. What keeps these abuses in check is our ability to detect them.
If design is so readily detectable outside science, and if its detectability is one of the key factors keeping scientists honest, why should design be barred from the actual content of science? With reference to biology, why should we have to constantly remind ourselves that biology studies things that only appear to be designed, but that in fact are not designed? Isn’t it at least conceivable that there could be good positive reasons for thinking biological systems are in fact designed?
The biological community’s response to these questions has been to resist design resolutely. The worry is that for natural objects (unlike human artifacts), the distinction between design and non-design cannot be reliably drawn. Consider, for
instance, the following remark by Darwin in the concluding chapter of his Origin of Species:
Several eminent naturalists have of late published their belief that a multitude of reputed species in each genus are not real species; but that other species are real, that is, have been independently created. . . . Nevertheless they do not pretend that they can define, or even conjecture, which are the created forms of life, and which are those produced by secondary laws. They admit variation as a vera causa in one case, they arbitrarily reject it in another, without assigning any distinction in the two cases.”11
It is this worry of falsely attributing something to design (here identified with creation) only to have it overturned later that has prevented design from entering science proper.
This worry, though perhaps justified in the past, is no longer tenable. There does in fact exist a rigorous criterion for distinguishing intelligently caused objects from unintelligently caused ones. Many special sciences already use this criterion, though in a pre-theoretic form (e.g. forensic science, artificial intelligence, cryptography, archeology, and the Search for Extra-Terrestrial Intelligence). The great breakthrough of the intelligent design movement has been to isolate and make precise this criterion. Michael Behe’s criterion of irreducible complexity for establishing the design of biochemical systems is a special case of this general criterion for detecting design.12
Before examining this criterion, I want to offer a brief clarification about the word
“design.” I use “design” in three distinct senses. First, I use it to denote the scientific theory that distinguishes intelligent agency from natural causes, a theory that increasingly is being referred to as “design theory” or “intelligent design theory”
(IDT).13 Second, I use “design” to denote what it is about intelligently produced objects that enables us to tell that they are intelligently produced and not simply the result of natural causes. When intelligent agents act, they leave behind a characteristic trademark or signature. The scholastics used to refer to the “vestiges of creation.” The Latin vestigium means footprint. It was thought that God, though not physically present, left his footprints throughout creation. Hugh Ross has similarly referred to the “fingerprint of God.” It is “design” in this sense – as a trademark, signature, vestige, or fingerprint – that this criterion for discriminating intelligently from unintelligently caused objects is meant to identify.14 Lastly, I use “design” to denote intelligent agency itself. Thus, to say that something is designed is to say that an intelligent agent caused it. But note, to say that an intelligent agent caused something is not to prescribe how an intelligent agent caused it. In particular, design does not presuppose miracles.15
3. The complexity-specification criterion
What does the criterion for distinguishing intelligently from unintelligently caused objects look like? Although a detailed explanation and justification of this criterion is fairly technical,16 the basic idea is straightforward and easily illustrated. Consider how the radio astronomers in the movie Contact detected an extra-terrestrial intelligence.
This movie, based on a novel by Carl Sagan, was an enjoyable piece of propaganda for the SETI research program – the Search for Extra-Terrestrial Intelligence. To make the movie interesting, the SETI researchers actually had to find an extra-terrestrial intelligence (the non-fictional SETI program has yet to be so fortunate).
How, then, did the SETI researchers in Contact find an extra-terrestrial intelligence? To increase their chances of finding an extra-terrestrial intelligence, SETI researchers monitor millions of radio signals from outer space. Many natural objects in space produce radio waves (e.g., pulsars). Looking for signs of design among all these naturally produced radio signals is like looking for a needle in a haystack. To sift through the haystack, SETI researchers run the signals they monitor through computers programmed with pattern-matchers. So long as a signal does not match one of the pre-set patterns, it will pass through the pattern-matching sieve (and that even if it has an intelligent source). If, on the other hand, it does match one of these patterns, then, depending on the pattern matched, the SETI researchers may have cause for celebration.
The SETI researchers in Contact did find a signal worthy of celebration, namely the following:
110111011111011111110111111111110111111111111101111111111
111111101111111111111111111011111111111111111111111011111
111111111111111111111111011111111111111111111111111111110
111111111111111111111111111111111111101111111111111111111
111111111111111111111101111111111111111111111111111111111
111111111011111111111111111111111111111111111111111111111
011111111111111111111111111111111111111111111111111111011
111111111111111111111111111111111111111111111111111111111
110111111111111111111111111111111111111111111111111111111
111111111111101111111111111111111111111111111111111111111
111111111111111111111111111101111111111111111111111111111
111111111111111111111111111111111111111111111011111111111
111111111111111111111111111111111111111111111111111111111
111111111110111111111111111111111111111111111111111111111
111111111111111111111111111111111111110111111111111111111
111111111111111111111111111111111111111111111111111111111
111111111111110111111111111111111111111111111111111111111
111111111111111111111111111111111111111111111111111111101
111111111111111111111111111111111111111111111111111111111
1111111111111111111111111111111111111111111
In this sequence of 1126 bits, 1’s correspond to beats and 0’s to pauses. This sequence represents the prime numbers from 2 to 101, where a given prime number is represented by the corresponding number of beats (i.e. 1’s), and the individual prime numbers are separated by pauses (i.e. 0’s).
The SETI researchers in Contact took this signal as decisive confirmation of an extra-terrestrial intelligence. What is it about this signal that decisively implicates design? Whenever we infer design, we must establish two things – complexity and specification. Complexity ensures that the object in question is not so simple that it can readily be explained by chance. Specification ensures that this object exhibits the type of pattern that is the trademark of intelligence.
To see why complexity is crucial for inferring design, consider the following sequence of bits:
110111011111
These are the first twelve bits in the previous sequence representing the prime numbers 2, 3 and 5 respectively. Now it is a sure bet that no SETI researcher, if confronted with this twelve-bit sequence, is going to contact the science editor at the New York Times, hold a press conference, and announce that an extra-terrestrial intelligence has been discovered. No headline is going to read, “Aliens Master the First Three Prime Numbers!”
The problem is that this sequence is much too short (i.e. has too little complexity) to establish that an extra-terrestrial intelligence with knowledge of prime numbers produced it. A randomly beating radio source might by chance just happen to output the sequence “110111011111.” A sequence of 1126 bits representing the prime numbers from 2 to 101, however, is a different story. Here the sequence is sufficiently long (i.e. has enough complexity) to confirm that an extra-terrestrial intelligence could have produced it.
Even so, complexity by itself is not enough to eliminate chance and implicate design. If I flip a coin 1000 times, I will participate in a highly complex (or what amounts to the same thing, highly improbable) event. Indeed, the sequence I end up flipping will be one in a trillion trillion trillion . . ., where the ellipsis needs twenty-two more “trillions.” This sequence of coin tosses will not, however, trigger a design inference. Though complex, this sequence will not exhibit a suitable pattern. Contrast this with the previous sequence representing the prime numbers from 2 to 101. Not only is this sequence complex, but it also embodies a suitable pattern. The SETI
researcher who in the movie Contact discovered this sequence put it this way: “This isn’t noise, this has structure.”
What is a suitable pattern for inferring design? Not just any pattern will do. Some patterns can legitimately be employed to infer design whereas others cannot. The basic intuition underlying the distinction between patterns that alternately succeed or fail to implicate design is, however, easily motivated. Consider the case of an archer.
Suppose an archer stands 50 meters from a large wall with bow and arrow in hand. The wall, let us say, is sufficiently large that the archer cannot help but hit it. Now suppose each time the archer shoots an arrow at the wall, the archer paints a target around the arrow so that the arrow sits squarely in the bull’s-eye. What can be concluded from this scenario? Absolutely nothing about the archer’s ability as an archer. Yes, a pattern is being matched; but it is a pattern fixed only after the arrow has been shot. The pattern is thus purely ad hoc.
But suppose instead the archer paints a fixed target on the wall and then shoots at it. Suppose the archer shoots a hundred arrows, and each time hits a perfect bull’s-eye.
What can be concluded from this second scenario? Confronted with this second scenario we are obligated to infer that here is a world-class archer, one whose shots cannot legitimately be referred to luck, but rather must be referred to the archer’s skill and mastery. Skill and mastery are of course instances of design.
The type of pattern where the archer fixes a target first and then shoots at it is common to statistics, where it is known as setting a rejection region prior to an experiment. In statistics, if the outcome of an experiment falls within a rejection region, the chance hypothesis supposedly responsible for the outcome is rejected.
Now a little reflection makes clear that a pattern need not be given prior to an event to eliminate chance and implicate design. Consider the following cipher text: nfuijolt ju jt mjlf b xfbtfm
Initially this looks like a random sequence of letters and spaces – initially you lack any pattern for rejecting chance and inferring design.
But suppose next that someone comes along and tells you to treat this sequence as a Caesar cipher, moving each letter one notch down the alphabet. Behold, the sequence now reads,
methinks it is like a weasel
Even though the pattern is now given after the fact, it still is the right sort of pattern for eliminating chance and inferring design. In contrast to statistics, which always tries to identify its patterns before an experiment is performed, cryptanalysis must discover its patterns after the fact. In both instances, however, the patterns are suitable for
inferring design.
Patterns divide into two types, those that in the presence of complexity warrant a design inference and those that despite the presence of complexity do not warrant a design inference. The first type of pattern is called a specification, the second a fabrication. Specifications are the non- ad hoc patterns that can legitimately be used to eliminate chance and warrant a design inference. In contrast, fabrications are the ad hoc patterns that cannot legitimately be used to warrant a design inference. This distinction between specifications and fabrications can be made with full statistical rigor.17
4. Specification
A rigorous elucidation of the complexity-specification criterion requires explicating the dual notions of specification and complexity. We treat specification first.
Specification is what must be added to an otherwise highly improbable chance event to render chance an unacceptable explanation of that event. Left to themselves highly improbable events do not warrant the elimination of chance. The precise sequence of heads and tails in a long sequence of coin tosses, the precise configuration of darts gotten by throwing darts at a dart board, and the precise seating arrangement of people at a cinema, if explained by chance, are all highly improbable events. It is only when the precise sequence of heads and tails has been recorded in advance, when the precise configuration of the darts locates all the darts at the center of the target, and when the precise seating arrangement at the cinema corresponds to seats people have been assigned on their ticket stubs – that is, when what at first blush appears to be a highly improbable event is also specified – that we are warranted in eliminating chance.
What exactly then are specifications? In the first place specifications are patterns.
As I stressed earlier, what is needed to eliminate chance in explaining an event is that the event under consideration conform to a pattern. Specifications, however, are not just any old patterns, for it is always too easy to commit a post hoc fallacy, concocting patterns after the fact so that the only reason an event conforms to a pattern is because the pattern was read off the event, thereby leaving open the possibility that the event was due to chance after all.
Patterns thus come in two varieties, specifications and fabrications. For our purposes specifications are the good patterns, the ones that legitimately warrant eliminating chance, whereas fabrications are the bad patterns, the ones that are ad hoc and thus give us no warrant for eliminating chance. What then marks the difference between specifications and fabrications? An example will help answer this question. Consider the following event E, an event that to all appearances was obtained by flipping a fair coin 100 times:
THTTTHHTHHTTTTTHTHTTHHHTTHTHHHTHHHTTTTTT
THTTHTTTHHTHTTTHTHTHHTTHHHHTTTHTTHHTHTHT
HHHHTTHHTHHHHTHHHHTT E
Is E the product of chance or not? A standard trick of statistics professors with an introductory statistics class is to divide the class in two, having students in one half of the class each flip a coin 100 times, writing down the sequence of heads and tails on a slip of paper, and having students in the other half each generate purely with their minds a “random looking” string of coin tosses that mimics the tossing of a coin 100
times, also writing down the sequence of heads and tails on a slip of paper. When the students then hand in their slips of paper, it is the professor’s job to sort the papers into two piles, those generated by flipping a fair coin, and those concocted in the students’
heads. To the amazement of the students, the statistics professor is typically able to sort the papers with 100 per cent accuracy.
There’s no mystery here. The statistics professor simply looks for a repetition of six or seven heads or tails in a row to distinguish the truly random from the pseudorandom sequences (the truly random sequences being those gotten by flipping a coin and the pseudo-random sequences being those concocted in the students’ heads). In a hundred coin flips one is quite likely to see six or seven such repetitions. On the other hand, people concocting pseudo-random sequences with their minds tend to alternate between heads and tails too frequently. Whereas with a truly random sequence of coin tosses there is a 50 percent chance that one toss will differ from the next, as a matter of human psychology people expect that one toss will differ from the next around 70 percent of the time.
How then will our statistics professor fare when confronted with E? Will E be attributed to chance or to the musings of someone trying to mimic chance? According to the professor’s crude randomness checker, E would be assigned to the pile of sequences presumed to be truly random, for E contains a repetition of seven tails in a row. Everything that at first blush would lead us to regard E as truly random checks out.
There are exactly 50 alternations between heads and tails (as opposed to the 70 that would be expected from humans trying to mimic chance). What is more, the relative frequencies of heads and tails check out: there were 49 heads and 51 tails. Thus it is not as though the coin supposedly responsible for generating E was heavily biased in favor of one side versus the other.
Suppose, however, that our statistics professor suspects she is not up against a neophyte statistics student, but instead a fellow statistician who is trying to put one over on her. To help organize her problem, study it more carefully, and enter it into a computer, she will find it convenient to let strings of 0’s and 1’s represent the outcomes of coin flips, with 1 corresponding to heads and 0 to tails. In this case the following pattern D will correspond to the event E:
01000110110000010100111001011101110000000100100011
01000101011001111000100110101011110011011110111100 D
Now the mere fact that the event E conforms to the pattern D is no reason to think that E did not occur by chance, since, as things stand now, the pattern D has simply been read off of the event E.
We are back to the question of how to explain an arrow stuck in a bull’s eye.
Whether we are impressed with an arrow sticking in a bull’s eye depends on whether the bull’s eye was in place before the arrow was shot, or whether it was painted around the arrow only after the arrow landed. Where the arrow landed (= event) certainly conforms to the pattern set by the bull’s eye. But if the bull’s eye was painted around the arrow only after the arrow landed, there is no reason to think that where the arrow landed was dictated by something other than chance. Only if the position of the target (= pattern) is in some sense independent of whatever chance process might be responsible for the arrow’s flight can this pattern preclude chance in case the arrow actually happens to be sticking in the bull’s eye.
In what sense then, if any, is D independent of E? More to the point, what have we failed to consider so far that would warrant our rejecting E as the product of chance?
To answer this question let us rewrite D as follows:
0
1
00
01
10
11
000
001
010
011
100
101
110
111
0000
0001
0010
0011
0100
0101
0110
0111
1000
1001
1010
1011
1100
1101
1110
1111
00
D
By viewing D in this way, anyone with the least exposure to binary arithmetic immediately recognizes that D was constructed simply by writing binary numbers in ascending order, starting with the one-digit binary numbers (i.e. 0 and 1), proceeding then to the two-digit binary numbers (i.e. 00, 01, 10 and 11), and continuing on until 100 digits were recorded. It is therefore intuitively clear that D does not describe a truly random event (i.e. an event gotten by tossing a fair coin), but rather a pseudorandom event, concocted by doing a little binary arithmetic.
Although it is now intuitively clear why E cannot properly be explained by reference to chance, we need to consider more closely why this mode of explanation fails here. We started with a putative chance event E, supposedly gotten by flipping a fair coin 100 times. Since heads and tails each have probability 1/2, and since this probability gets multiplied for each flip of the coin, it follows that the probability of E
is 2−100, or approximately 10−30. In addition, we constructed a pattern D to which E
conforms. Initially D proved insufficient to eliminate chance as the explanation of E
since in its construction D was simply read off of E. Rather, to eliminate chance we had also to recognize that D could have been constructed quite easily by performing some simple arithmetic operations with binary numbers. Thus to eliminate chance we needed to employ some additional side information, which in this case consisted of our knowledge of binary arithmetic.
Two aspects of this side information are crucial to the elimination of chance in this example. The first is that the side information provides no information about the event E in question. Probabilistically this means that the probability of E does not change once the side information becomes available to us (a property that probabilists refer to as “conditional independence”). This is certainly the case here since our knowledge of binary arithmetic does not affect the probabilities we assign to coin tosses. The second crucial aspect of the side information is that it enables us to construct the pattern D to which E conforms. This is also the case here, where our knowledge of binary arithmetic enables us to arrange binary numbers in ascending order, and thereby construct the pattern D. The side information that enables us to eliminate chance must enable us to construct the pattern to which an event conforms, yet without recourse to the actual event. This is the crucial insight. Because the side
information provides no information about the event E (i.e. is conditionally independent of E), any pattern constructed from this side information is obtained without recourse to the event E. In this way any pattern like D that is constructed from the side information avoids the charge of being ad hoc.
We are now in a position to define what it means for a pattern D to be a specification of an event E, and thus for D to be capable of eliminating chance in explaining E: given a pattern D to which the event E conforms, D is a specification of E provided side information K has been identified satisfying the following two conditions CINDE
Under the assumption that E occurred by chance, K provides no information about E (i.e. K is conditionally independent of E).
TRACT
The task of constructing D from K is doable (or as complexity theorists would say, “tractable”).
CINDE and TRACT are respectively the conditional independence and tractability conditions for side information K. In case both the conditions CINDE and TRACT
obtain, we say that the pattern D specifies the event E, and that the pattern D is a specification. These conditions characterize the type of side information needed to eliminate chance. In the previous example, K consisted of our knowledge of binary arithmetic. Our knowledge of binary arithmetic clearly does not affect the probabilities we assign to coin tosses (hence CINDE was satisfied), but does allow us to construct the pattern D representing an ascending sequence of binary numbers (hence TRACT was also satisfied). Thus it became obvious that the event E could not reasonably be attributed to chance.
As they stand, CINDE and TRACT require considerable technical elaboration.
Unfortunately, I can offer only the briefest indication of what is at stake with these conditions here, and must refer the reader elsewhere for a thorough technical treatment.18 CINDE is a probabilistic condition, stating what probabilistic properties K must possess if K is to be suitable for eliminating chance in explaining E. TRACT is a complexity-theoretic condition, stating what complexity-theoretic properties K must possess if K is to be suitable for eliminating chance in explaining E. Probabilistically, K must leave whatever probabilities we assign to the event E untouched. Complexity-theoretically, K must render the task of constructing the pattern D tractable or doable.
The tasks we find doable depend on the resources we are given. Thus given current technology we find sending a person to the moon doable, but sending a person to the nearest galaxy non-doable. The doable tasks are by definition the tractable ones, the non-doable tasks the intractable ones. Complexity theory, as a generalization of computational complexity theory, ranks tasks according to difficulty, and then determines which tasks are sufficiently easy to be tractable.19 Thus, according to TRACT the task to constructing D from K has to be sufficiently easy to be tractable.
5. Complexity as probability
The complexity part of the complexity-specification criterion is not the same complexity that appears in the tractability condition used to define specification. The complexity of the complexity-specification criterion is essentially a probability.20 To see the connection between complexity and probability, consider a combination lock.
The more possible combinations of the lock, the more complex the mechanism and correspondingly the more improbable that the mechanism can be opened by chance.
Complexity and probability therefore vary inversely: the greater the complexity, the smaller the probability. Thus, to determine whether the complexity-specification criterion is satisfied, it is necessary to determine not only whether an object or event is specified, but also whether it has sufficiently small probability (i.e. sufficiently high complexity).
What makes a probability a small probability? To answer this question we shall need the concept of a probabilistic resource. 21 A probability is never a small probability in isolation, but only in relation to a set of probabilistic resources that describe the number of relevant ways an event might occur. There are two types of probabilistic resources, replicational and specificational. To understand the first of these, imagine that a massive revision of the criminal justice system has just taken place. Henceforth, a convicted criminal is sentenced to serve time in prison until he flips k heads in a row, where k is selected according to the severity of the offense (we assume that all coin flips are fair and duly recorded – no cheating is possible).
Thus for a 10-year prison sentence, if we assume the prisoner can flip a coin once every five seconds (this seems reasonable), eight hours a day, six days a week, and given that the average streak of heads is 2 (=Σ1≤i≤ ∞i2−i), the prisoner will on average attempt to get a string of k heads once every 10 seconds, or 6 attempts a minute, or 360 attempts an hour, or 2880 attempts in an eight-hour work day, or 901440 attempts a year (assuming a six-day work week), or approximately 9 million attempts in 10 years. 9
million is approximately 223. Thus if we required a prisoner to flip 23 heads in a row before being released, we would on average expect to see him out within approximately 10 years. Of course specific instances will vary – some prisoners being released after only a short stay, others never recording the elusive 23 heads!
Now consider a prisoner’s reaction when after approximately 10 years he finally flips 23 heads in a row. Is he shocked? Does he think a miracle has occurred? Certainly not. His reaction is more likely to be, It’s about time! Given the number of opportunities he had for observing 23 heads in a row, he has about an even chance of getting out of prison within 10 years. There is thus nothing improbable about his getting out of prison within this span of time. It is improbable that on any given occasion when he flips his coin he will flip 23 heads in a row. But when all these occasions are considered jointly, it becomes quite probable that he will be out of prison within 10 years.
Given the event E of flipping 23 heads in a row and given n opportunities for E to occur, the question is not whether E taken by itself is improbable, but whether E
remains improbable once all the opportunities for E to occur have been factored in.
If we let E(n) denote the event that E occurs in at least one of these n trials, the question then is whether E(n) is improbable. Thus in the case of our coin-flipping prisoner, even though the probability of E is 2−23, and therefore quite small, it is the probability of E(9,000,000)), which is approximately 1/2, and which clearly is not small, that gives the prisoner hope of actually getting out of prison within his lifetime.
If the prisoner’s life expectancy is better than 10 years, he stands a reasonably good chance of getting out of prison.
The following two quantities therefore offset each other: the probabilistic resources n available to the prisoner, on the one hand, and the number of heads k that he must flip in a row to secure his release, on the other. If the prisoner’s probabilistic resources (=n) are big enough, it does not matter how many heads in a row (=k) he must obtain to get out of prison – his probabilistic resources will then be adequate to get him out of prison. If, on the other hand, the number of heads (=k) that must be flipped in a row is exorbitant, the prisoner’s probabilistic resources (=n) will stand little chance of getting him out of prison.
Consider, for instance, a prisoner condemned to flip 100 heads in a row (=k). The probability of getting 100 heads in a row on a given trial is so small that the prisoner has no practical hope of getting out of prison, even if his life expectancy and coin tossing ability were dramatically increased. If he could, for instance, make 10 billion attempts each year to obtain 100 heads in a row (this is coin-flipping at a rate of over 500 coin flips per second, 24 hours a day, seven days a week, for a full year), then he stands only an even chance of getting out of prison in 1020 years. His probabilistic resources are so inadequate for obtaining the desired 100 heads that it is pointless for him to entertain hopes of freedom.
Let us now turn to the other type of probabilistic resource. In the preceding example probabilistic resources consisted of the number of opportunities for a certain event to occur. We will call this type of probabilistic resource a replicational resource (i.e., the number of trials or replications for a given event to occur).
Replicational resources are not the only type of probabilistic resource, however.
Probabilistic resources can also assume another form in which the question is not how many opportunities there are for a given event to occur, but rather, how many opportunities there are to specify an as-yet undetermined event. We wi’ll call this other type of probabilistic resource a specificational resource. Because lotteries provide the perfect illustration of specificational resources, we consider next a lottery example.
In the interest of eliminating the federal deficit, the federal government decides to hold a national lottery in which the grand prize is the nation’s gold reserves at Fort Knox. The federal government has so constructed the lottery that the probability of
any one ticket winning is 1 in 2100, or approximately 1 in 1030. Specifically, the lottery has been constructed so that to buy a ticket, the lottery player pays a fixed price, in this case $10, and then records a 0–1 string of length 100 – whichever string she chooses so long as it does not match a string that has already been chosen. She is permitted to purchase as many tickets as she wishes, subject only to her financial resources and the time it takes to record the 0–1 strings of length 100. The lottery is to be drawn at a special meeting of the United States Senate: in alphabetical order each senator is to flip a single coin once and record the resulting coin toss.
Suppose now that the fateful day has arrived. A trillion tickets have been sold at ten dollars a piece. To prevent cheating, Congress has enlisted the National Academy of Science. In accord with the NAS’s recommendation, each ticket holder’s name is duly entered onto a secure database, together with the tickets purchased and the ticket numbers (i.e. the bit strings relevant to deciding the winner). All this information is now in place. After much fanfare the senators start flipping their coins, beginning with Senator Amy Aaron and ending with Senator Zig Zygmund. As soon as Senator Zygmund has announced his toss, the database is consulted to determine whether the lottery has a winner. Lo and behold, the lottery does indeed have a winner, whose name is . . . .
From a probabilist’s perspective there is one overriding implausibility to this example. The implausibility rests not with the federal government sponsoring a lottery to eliminate the national debt, nor with the choice of prize, nor with the way the lottery is decided at a special meeting of the Senate, nor even with the fantastically poor odds of anyone winning the lottery. The implausibility rests with the lottery having a winner. By all means the federal government should institute such a lottery if it seemed likely to redress the national debt, for it is obvious that if the lottery is run fairly, there will be no winner – the odds are simply too much against it. Suppose, for instance, that a trillion tickets are sold at 10 dollars apiece (this would cover the deficit as it stands at the time of this writing). An elementary calculation shows that the probability that one of these tickets (qua specifications) will match the winning string of 0’s and 1’s drawn by the senate cannot exceed 1 in 1018. Even if we increase the number of lottery tickets sold by several orders of magnitude, there still will not be sufficiently many tickets sold for the lottery to stand a reasonable chance of having a winner.
Sometimes it is necessary to consider both types of probabilistic resources in tandem, those depending on the number of opportunities for an event to occur (i.e.
replicational) as well as those depending on the number of opportunities to specify a given event (i.e. specificational). Suppose, for instance, that in the preceding lottery the Senate will hold up to a thousand drawings to determine a winner. Thus instead of having Senators Aaron through Zygmund flip their pennies in succession just once, we have them repeat this process up to a thousand times, stopping short of the thousand repetitions in case there happens to be a winner. If we now assume as before
that a trillion tickets have been sold, then for this probabilistic set-up the probabilistic resources include both a trillion specifications and a thousand possible replications.
An elementary calculation now shows that the probability of this modified lottery having a winner is no greater than 1 in 1015. It therefore remains highly unlikely that this modified lottery, despite the increase in probabilistic resources, will have a winner.
Probabilistic resources comprise the relevant ways an event can occur (replicational resources) and be specified (specificational resources). The important question therefore is not what is the probability of the event in question but rather what does its probability become after all the relevant probabilistic resources have been factored in? Probabilities can never be considered in isolation, but must always be referred to a relevant class of probabilistic resources. A seemingly improbable event can become quite probable when placed against a sufficiently large set of probabilistic resources. The anthropic principles that look to multiple universes bank on precisely this point: while the emergence of intelligent life in our universe is vastly improbable when considered strictly with respect to our universe, when we factor in all the possible universes that might have given rise to us, the emergence of intelligent life is rendered a virtual certainty. Whereas the status of possible universes other than our own is a matter of controversy, their role as probabilistic resources is evident.22
Let us now return to our original question: Given a probability p, what makes p a small probability? The primary intuition underlying small probabilities is that an event has small probability only if its probability remains small after all the probabilistic resources relevant to its occurrence have been factored in. Thus a prisoner sentenced to flip 23 heads in a row to be released from prison does not regard this event as having small probability because this event is reasonably likely given 10
years of coin flipping. A prisoner sentenced to flip 100 heads in a row to get out of prison, however, is right to regard this event as having small probability since it remains improbable even after all conceivable probabilistic resources have been factored in.
Once one has fixed all the probabilistic resources relevant to the event under consideration, how does one assess whether the probability of this event is indeed small? There seems to be an infinite regress here, for given what has been said, the probability of an event is small just in case its probability remains small after the relevant probabilistic resources have been factored in. Recall the prisoner example.
A prisoner sentenced to flip k heads in a row regards this event as having small probability just in case after factoring in all his probabilistic resources for attaining this event, the recomputed probability that factors in all his probabilistic resources remains small. Determining what it is for a probability to be small is thus transformed into the problem of determining what it is for a recomputed probability to be small.
As it turns out, however, this is progress, for it is much easier to say what it is for a recomputed probability to be small (i.e. a probability that has been recomputed on the basis of relevant probabilistic resources), than what it is for an arbitrary probability
to be small (i.e. a probability that is given without reference to relevant probabilistic resources).
Indeed, things could not be simpler: a probability recomputed on the basis of probabilistic resources is small just in case it is less than 1/2. The number 1/2 marks an important probabilistic boundary. Events of probability strictly less than 1/2 are less likely than not, whereas those of probability strictly greater than 1/2 are more likely than not. For an arbitrary event E of probability p, whether p is greater or less than 1/2 does not determine whether p has small probability since the probabilistic resources relevant to E’s occurrence have yet to be factored in. But once those probabilistic resources are factored in, whether the probability recomputed on the basis of those probabilistic resources is greater or less than 1/2 becomes crucially important.
A prisoner sentenced to flip k heads in a row in order to be released from prison has nothing to fear from any inherent improbability of this sequence, but everything to fear from not having enough coin flips (i.e. probabilistic resources) to render this sequence more likely than not. If all the coin flips potentially available to the prisoner cannot render this sequence more likely than not, the prisoner’s probability of getting out of prison will remain less than an even chance, and the prisoner should expect to remain in prison. An event whose probability remains less than 1/2, even after all the relevant probabilistic resources have been factored in, is an event which all the relevant probabilistic resources cannot conspire to render more likely than not.
To recap, a probability is never small in isolation, but only in relation to a relevant set of probabilistic resources. Once the relevant set of probabilistic resources has been prescribed, a probability is considered small just in case the probability recomputed on the basis of those resources is strictly less than 1/2. Thus for the complexity-specification criterion to be satisfied, the complexity of the event or object in question must correspond to a probability which, when recomputed against the relevant probabilistic resources, is less than 1/2.
6. Why the criterion works
Why does the complexity-specification criterion reliably detect design? To see why this criterion is exactly the right instrument for detecting design, we need to understand what it is about intelligent agents that makes them detectable in the first place. The principal characteristic of intelligent agency is choice. Whenever an intelligent agent acts, it chooses from a range of competing possibilities.
This is true not just of humans, but of animals as well as of extra-terrestrial intelligences. A rat navigating a maze must choose whether to go right or left at various points in the maze. When SETI researchers attempt to discover intelligence in the extra-terrestrial radio transmissions they are monitoring, they assume an extra-
terrestrial intelligence could have chosen any number of possible radio transmissions, and then attempt to match the transmissions they observe with certain patterns as opposed to others. Whenever a human being utters meaningful speech, a choice is made from a range of possible sound-combinations that might have been uttered.
Intelligent agency always entails discrimination, choosing certain things, ruling out others.
Given this characterization of intelligent agency, the crucial question is how to recognize it. Intelligent agents act by making a choice. How then do we recognize that an intelligent agent has made a choice? A bottle of ink spills accidentally onto a sheet of paper; someone takes a fountain pen and writes a message on a sheet of paper. In both instances ink is applied to paper. In both instances one among an almost infinite set of possibilities is realized. In both instances a contingency is actualized and others are ruled out. Yet in one instance we ascribe agency, in the other chance.
What is the relevant difference? Not only do we need to observe that a contingency was actualized, but we ourselves need also to be able to specify that contingency. The contingency must conform to an independently given pattern, and we must be able independently to formulate that pattern. A random ink blot is unspecifiable; a message written with ink on paper is specifiable. Ludwig Wittgenstein in Culture and Value made the same point: “We tend to take the speech of a Chinese for inarticulate gurgling. Someone who understands Chinese will recognize language in what he hears. Similarly I often cannot discern the humanity in man.”23
In hearing a Chinese utterance, someone who understands Chinese not only recognizes that one from a range of all possible utterances was actualized, but is also able to specify the utterance as coherent Chinese speech. Contrast this with someone who does not understand Chinese. In hearing a Chinese utterance, someone who does not understand Chinese also recognizes that one from a range of possible utterances was actualized, but this time, because lacking the ability to understand Chinese, is unable to specify the utterance as coherent speech.
To someone who does not understand Chinese, the utterance will appear gibberish. Gibberish – the utterance of nonsense syllables uninterpretable within any natural language – always actualizes one utterance from the range of possible utterances. Nevertheless, gibberish, by corresponding to nothing we can understand in any language, also cannot be specified. As a result, gibberish is never taken for intelligent communication, but always for what Wittgenstein calls “inarticulate gurgling.”
This actualizing of one among several competing possibilities, ruling out the rest, and specifying the one that was actualized encapsulates how we recognize intelligent agency, or equivalently, how we detect design. Experimental psychologists who study animal learning and behavior have known this all along. To learn a task an animal must acquire the ability to actualize behaviors suitable for the task as well as the ability to rule out behaviors unsuitable for the task. Moreover, for a psychologist to recognize
that an animal has learned a task, it is necessary not only to observe the animal making the appropriate discrimination, but also to specify this discrimination.
Thus to recognize whether a rat has successfully learned how to traverse a maze, a psychologist must first specify which sequence of right and left turns conducts the rat out of the maze. No doubt, a rat randomly wandering a maze also discriminates a sequence of right and left turns. But by randomly wandering the maze, the rat gives no indication that it can discriminate the appropriate sequence of right and left turns for exiting the maze. Consequently, the psychologist studying the rat will have no reason to think the rat has learned how to traverse the maze.
Only if the rat executes the sequence of right and left turns specified by the psychologist will the psychologist recognize that the rat has learned how to traverse the maze. Now it is precisely the learned behaviors we regard as intelligent in animals.
Hence it is no surprise that the same scheme for recognizing animal learning recurs for recognizing intelligent agency generally, to wit: actualizing one among several competing possibilities, ruling out the others, and specifying the one actualized.
Note that complexity is implicit here as well. To see this, consider again a rat traversing a maze, but now take a very simple maze in which two right turns conduct the rat out of the maze. How will a psychologist studying the rat determine whether it has learned to exit the maze? Just putting the rat in the maze will not be enough.
Because the maze is so simple, the rat could by chance just happen to take two right turns, and thereby exit the maze. The psychologist will therefore be uncertain whether the rat actually learned to exit this maze, or whether the rat just got lucky.
But contrast this now with a complicated maze in which a rat must take just the right sequence of left and right turns to exit the maze. Suppose the rat must take one hundred appropriate right and left turns, and that any mistake will prevent the rat from exiting the maze. A psychologist who sees the rat take no erroneous turns and in short order exit the maze will be convinced that the rat has indeed learned how to exit the maze, and that this was not dumb luck.
This general scheme for recognizing intelligent agency is but a thinly disguised form of the complexity-specification criterion. In general, to recognize intelligent agency we must observe an actualization of one among several competing possibilities, note which possibilities were ruled out, and then be able to specify the possibility that was actualized. What is more, the competing possibilities that were ruled out must be live possibilities, and sufficiently numerous so that specifying the possibility that was actualized cannot be attributed to chance. In terms of complexity, this is just another way of saying that the range of possibilities is complex. In terms of probability, this is just another way of saying that the possibility that was actualized has small probability.
All the elements in this general scheme for recognizing intelligent agency (i.e.
actualizing, ruling out, and specifying) find their counterpart in the complexity-specification criterion. It follows that this criterion formalizes what we have been
doing right along when we recognize intelligent agency. The complexity-specification criterion pinpoints what we need to be looking for when we detect design.
As a postscript it is worth pondering the etymology of the word “intelligent.” The word “intelligent” derives from two Latin words, the preposition inter, meaning between, and the verb lego, meaning to choose or select. Thus according to its etymology, intelligence consists in choosing between. It follows that the etymology of the word “intelligent” parallels the formal analysis of intelligent agency inherent in the complexity-specification criterion.
7. Irreducible complexity
Perhaps the most compelling evidence for design in the universe comes from biology, and specifically from biochemistry. In a recent issue of Cell, Bruce Alberts, president of the National Academy of Sciences, remarked, “The entire cell can be viewed as a factory that contains an elaborate network of interlocking assembly lines, each of which is composed of large protein machines. . . . Why do we call the large protein assemblies that underlie cell function machines? Precisely because, like the machines invented by humans to deal efficiently with the macroscopic world, these protein assemblies contain highly coordinated moving parts.”24
Nonetheless, Alberts sides with the majority of biologists in regarding the cell’s marvelous complexity as only apparently designed. The Lehigh University biochemist Michael Behe disagrees. In Darwin’s Black Box Behe presents a powerful argument for actual design in the cell. Central to his argument is his notion of irreducible complexity.
A system is irreducibly complex if it consists of several interrelated parts so that removing even one part completely destroys the system’s function. As an example of irreducible complexity Behe offers the mousetrap. A mousetrap consists of a platform, a hammer, a spring, a catch, and a holding bar. Remove any one of these five components, and it is impossible to construct a functional mousetrap.25
Irreducible complexity needs to be contrasted with cumulative complexity. A system is cumulatively complex if the components of the system can be arranged sequentially so that the successive removal of components never leads to the complete loss of function. An example of a cumulatively complex system is a city. It is possible successively to remove people and services from a city until one is down to a tiny village
– all without losing the sense of community, which in this case constitutes function.
From this characterization of cumulative complexity, it is clear that the Darwinian mechanism of selection and mutation can readily account for cumulative complexity.
Indeed, the gradual accrual of complexity via selection mirrors the retention of function as components are successively removed from a cumulatively complex system.
But what about irreducible complexity? Can the Darwinian mechanism account
for irreducible complexity? Certainly, if selection acts with reference to a goal, it can produce irreducible complexity. Take Behe’s mousetrap. Given the goal of constructing a mousetrap, one can specify a goal-directed selection process that in turn selects a platform, a hammer, a spring, a catch, and a holding bar, and at the end puts all these components together to form a functional mousetrap. Given a pre-specified goal, selection has no difficulty producing irreducibly complex systems.
But the selection operating in biology is Darwinian natural selection. And this form of selection operates without goals, has neither plan nor purpose, and is wholly undirected. The great appeal of Darwin’s selection mechanism was, after all, that it would eliminate teleology from biology. Yet by making selection an undirected process, Darwin drastically abridged the type of complexity biological systems could manifest. Henceforth biological systems could manifest only cumulative complexity, not irreducible complexity.
Why is this? As Behe explains in Darwin’s Black Box: An irreducibly complex system cannot be produced . . . 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. . . . Since natural selection can only choose systems that are already working, then if a biological system cannot be produced gradually it would have to arise as an integrated unit, in one fell swoop, for natural selection to have anything to act on.26
For an irreducibly complex system, function is attained only when all components of the system are in place simultaneously. It follows that natural selection, if it is going to produce an irreducibly complex system, has to produce it all at once or not at all. This would not be a problem if the systems in question were simple. But they’re not. The irreducibly complex biochemical systems Behe considers are protein machines consisting of numerous distinct proteins, each indispensable for function, and together beyond what natural selection can muster in a single generation.
One such irreducibly complex biochemical system that Behe considers is the bacterial flagellum. The flagellum is a whip-like rotary motor that enables a bacterium to navigate through its environment. The flagellum includes an acid powered rotary engine, a stator, O-rings, bushings, and a drive shaft. The intricate machinery of this molecular motor requires approximately fifty proteins. Yet the absence of any one of these proteins results in the complete loss of motor function.27
The irreducible complexity of such biochemical systems counts powerfully against the Darwinian mechanism, and indeed against any naturalistic evolutionary mechanism proposed to date. Moreover, because irreducible complexity occurs at the biochemical level, there is no more fundamental level of biological analysis to which
the irreducible complexity of biochemical systems can be referred, and at which a Darwinian analysis in terms of selection and mutation can still hope for success.
Undergirding biochemistry is ordinary chemistry and physics, neither of which can account for biological information. Also, whether a biochemical system is irreducibly complex is a fully empirical question: individually knock out each protein constituting a biochemical system to determine whether function is lost. If so, we are dealing with an irreducibly complex system. Mutagenesis experiments of this sort are routine in biology.28
The connection between Behe’s notion of irreducible complexity and my complexity-specification criterion is now straightforward. The irreducibly complex systems Behe considers require numerous components specifically adapted to each other and each necessary for function. On any formal complexity-theoretic analysis, they are complex in the sense required by the complexity-specification criterion.
Moreover, in virtue of their function, these systems embody patterns independent of the actual living systems (cf. section 4). Hence these systems are also specified in the sense required by the complexity-specification criterion.
Biological specification always denotes function. An organism is a functional system comprising many functional subsystems. The functionality of organisms can be cashed out in any number of ways. Arno Wouters cashes it out globally in terms of the viability of whole organisms.29 Michael Behe cashes it out in terms of the minimal function of biochemical systems.30 Even the staunch Darwinist Richard Dawkins will admit that life is specified functionally, cashing out functionality in terms of the reproduction of genes. Thus in The Blind Watchmaker Dawkins will write, “Complicated things have some quality, specifiable in advance, that is highly unlikely to have been acquired by random chance alone. In the case of living things, the quality that is specified in advance is . . . the ability to propagate genes in reproduction.”31
8. So what?
There exists a reliable criterion for detecting design. This criterion detects design strictly from observational features of the world. Moreover, it belongs to probability and complexity theory, not to metaphysics and theology. And although it cannot achieve logical demonstration, it does achieve statistical justification so compelling as to demand assent. This criterion is relevant to biology. When applied to the complex, information-rich structures of biology, it detects design. In particular, the complexity-specification criterion shows that Michael Behe’s irreducibly complex biochemical systems are designed.
What are we to make of these developments? Many scientists remain unconvinced.
So what if we have a reliable criterion for detecting design and so what if that criterion tells us that biological systems are designed? How is looking at a biological system and
inferring it is designed any better than shrugging our shoulders and saying God did it? The fear is that design cannot help but stifle scientific inquiry.
Design is not a science-stopper. Indeed, design can foster inquiry where traditional evolutionary approaches obstruct it. Consider the term “junk DNA.” Implicit in this term is the view that because the genome of an organism has been cobbled together through a long, undirected evolutionary process, the genome is a patchwork of which only limited portions are essential to the organism. Thus on an evolutionary view we expect a lot of useless DNA. If, on the other hand, organisms are designed, we expect DNA, as much as possible, to exhibit function. And indeed, the most recent findings suggest that designating DNA as “junk” merely cloaks our current lack of knowledge about function. For instance, in a recent issue of the Journal of Theoretical Biology, John Bodnar describes how “non-coding DNA in eukaryotic genomes encodes a language which programs organismal growth and development.”32 Design encourages scientists to look for function where evolution discourages it.
Or consider vestigial organs that later are found to have a function after all.
Evolutionary biology texts often cite the human coccyx as a “vestigial structure” that hearkens back to vertebrate ancestors with tails. Yet if one looks at a recent edition of Gray’s Anatomy, one finds that the coccyx is a crucial point of contact with muscles that attach to the pelvic floor. Now anatomy is nothing else than an exercise in design, studying the large-scale design plans/blueprints for bodies. Thus here again we find design encouraging scientists to look for function where evolution discourages it.
Examples where the phrase “vestigial structure” merely cloaks our current lack of knowledge about function can be multiplied. The human appendix, formerly thought to be vestigial, is now known to be a functioning component of the immune system.33
Admitting design into science can only enrich science. All the tried and true tools of science will remain intact. But design also adds a new tool to the scientist’s explanatory tool chest, namely, the complexity-specification criterion. Moreover, design raises a whole new set of research questions. Once we know that something is designed, we will want to know:
1.
How was it constructed (cf. reverse engineering)?
2.
What is its function/purpose?34
3.
How has noise/age/friction/mutation obscured the original design plan?
4.
Can we reconstruct the original design plan (cf. art restoration)?
5.
What are the constraints within which the object functions well and outside of which it malfunctions or breaks?35
6.
In what sense is the design optimal?36
Even so, design does much more than enrich and transform science. Indeed, design has profound metaphysical and world-view implications. The discovery of design in
the world, and especially in biology, enables us to answer the question with which we started this essay: If naturalism is false, how could we know it? Naturalism, by limiting the fundamental causation of nature to chance and necessity, must reduce all appearance of design in nature to just that – mere appearance. What the complexity-specification criterion shows, however, is that design cannot be reduced to chance and necessity. Moreover, when applied to the complex, information-rich structures of biology, like Michael Behe’s irreducibly complex biochemical systems, this criterion shows that design actually is present in nature.
Naturalism, therefore, fails on its own terms. The great hope of scientific naturalism was, after all, that naturalism could be redeemed in the coin of science.
That hope is now effectively dashed. There is no principled way of excluding design from the causal structure of the world. Moreover, design is, via the complexity-specification criterion, fully amenable to scientific investigation. Naturalism is therefore seen to be false on strictly scientific grounds. The logic of this conclusion is straightforward: naturalism allows only certain sorts of fundamental causes (chance and necessity). Those causes are (demonstrably) incapable of generating specified complexity. But nature exhibits specified complexity, especially in biology. Therefore naturalism is false.
If naturalism is false, how can we know it? This is how.
Notes
1 Stephen Hawking, A Brief History of Time: From the Big Bang to Black Holes (New York: Bantam, 1988).
2 Alan Guth, The Inflationary Universe: The Quest for a New Theory of Cosmic Origins (Reading, Mass.: Addison-Wesley, 1997).
3 John Barrow and Frank Tipler, The Anthropic Cosmological Principle (Oxford: Oxford University Press, 1986).
4 See respectively Hawking, Brief History, 171–5; Guth, Inflationary Universe, 247 (“eternal inflation”); Barrow and Tipler, Anthropic Cosmological Principle, 22–3.
5 Hugh Ross, The Fingerprint of God, 2nd ed. (Orange, Calif.: Promise, 1991), 132–8.
6 See William Lane Craig, “Design and the Cosmological Argument,” in Mere Creation: Science, Faith and Intelligent Design, ed. William A. Dembski (Downers Grove, Ill.: InterVarsity, 1998), 335–7; Richard Swinburne, The Existence of God (Oxford: Oxford University Press, 1979), 107–15.
7 See Michael Behe, Darwin’s Black Box (New York: Free Press, 1996).
8 Richard Dawkins, The Blind Watchmaker (New York: Norton, 1986), 1.
9 Francis Crick, What Mad Pursuit (New York: Basic Books, 1988), 138.
10 Eliot Marshall, “Medline Searches Turn Up Cases of Suspected Plagiarism,” Science 279 (23
Jan 1998): 473–4.
11 Charles Darwin, On the Origin of Species, facsimile 1st ed. (C ambridge, Mass.: Harvard University Press, 1964 [1859]), 482.
12 Cf. Behe, Darwin’s Black Box, 39–45.
13 Cf. Dembski (ed.), Mere Creation.
14 The relevant theory is developed in William A. Dembski, The Design Inference (Cambridge: Cambridge University Press, 1998).
15 For a non-naturalistic account of intelligent agency see Edmund Runggaldier, Was Sind Handlungen? Eine Philosophische Auseinandersetzung mit dem Naturalismus, in Münchener Philosophische Studien, vol. 12 (Stuttgart: Verlag W Kohlhammer, 1996).
16 For a full account see my book The Design Inference.
17 Ibid., ch. 5.
18 Ibid.
19 For an overview of complexity theory, see Dembski, The Design Inference, ch. 4.
20 For probability as a special case of complexity, see Dembski, The Design Inference, sec. 4.6.
21 For a full account of probabilistic resources, see Dembski, The Design Inference, ch. 6.
22 John Leslie analyzes this line of reasoning masterfully in his book Universes (London: Routledge, 1989).
23 Ludwig Wittgenstein, Culture and Value, ed. G. H. von Wright, trans. P. Winch (Chicago: University of Chicago Press, 1980), 1e.
24 Bruce Alberts, “The Cell as a Collection of Protein Machines: Preparing the Next Generation of Molecular Biologists,” Cell 92 (8 Feb 1998): 291.
25 Behe, Darwin’s Black Box, 39–45.
26 Ibid., 39.
27 Ibid., 69–72.
28 See, for example, Nicholas Gaiano, Adam Amsterdam, Koichi Kawakami, Migeul Allende, Thomas Becker, and Nancy Hopkins, “Insertional Mutagenesis and Rapid Cloning of Essential Genes in Zebrafish,” Nature 383 (1996): 829–32; Carolyn K. Suzuki, Kitaru Suda, Nan Wang, and Gottfried Schatz, “Requirement for the Yeast Gene LON in Intramitochondrial Proteolysis and Maintenance of Respiration,” Science 264 (1994): 273–6; Qun-Yong Zhou, Carol J. Qualfe, and Richard D. Palmiter, “Targeted Disruption of the Tyrosine Hydroxylase Gene Reveals that Catecholamines are Required for Mouse Fetal Development,” Nature 374 (1995): 640–3.
29 Arno Wouters, “Viability Explanation,” Biology and Philosophy 10 (1995): 435–57.
30 Behe, Darwin’s Black Box, 45–6.
31 Dawkins, Blind Watchmaker, 9.
32 John W. Bodnar, Jeffrey Killian, Michael Nagle, and Suneil Ramchandani, “Deciphering the Language of the Genome,” Journal of Theoretical Biology 189 (1997): 183.
33 Percival Davis and Dean Kenyon, Of Pandas and People, 2nd ed. (Dallas, Tex.: Haughton, 1993), 128.
34 Note that we can detect design without knowing what something was designed for. There is a room at the Smithsonian filled with obviously designed objects for which no one has a clue
about their purpose. See Del Ratzsch’s article “Design, Chance, and Theistic Evolution,” in Dembski, ed., Mere Creation, 294.
35 Design implies constraints. An object that is designed functions within certain design constraints. Transgress those constraints and the object functions poorly or breaks.
Moreover, we can discover those constraints empirically by seeing what does and doesn’t work. This simple insight has tremendous implications not just for science but also for ethics.
If humans are in fact designed, then we can expect psychosocial constraints to be hardwired into us. Transgress those constraints, and we personally as well as our society will suffer.
There’s plenty of empirical evidence to suggest that many of the attitudes and behaviors our society promotes undermine human flourishing. Design promises to reinvigorate that ethical stream running from Aristotle through Aquinas known as natural law. See J.
Budziszewski, Written on the Heart: The Case for Natural Law (Downers Grove, Ill.: InterVarsity, 1997).
36 Optimality of design is a separate question from design. Something can be designed without being optimally designed. What’s more, optimality of design must not be confused with perfection of design. Indeed, there is no such thing as “perfect design.” As Henry Petroski aptly notes, “All design involves conflicting objectives and hence compromise, and the best designs will always be those that come up with the best compromise.” See Henry Petroski, Invention by Design: How Engineers Get from Thought to Thing (Cambridge, Mass.: Harvard University Press, 1997), 30.
Aaron, Senator Amy 268
Borde, Arvind 224, 225, 231
‘abstract particulars’ 70
Boyle, Joseph 197
Akaike criterion 59, 60, 61
Bradley, F.H. 97, 103
Alberts, Bruce 273
Braine, David 173
A- and B-theories of time 232–4
Brandenberger, Robert 224
anti-foundationalism 42
Brentano, F. 40
anti-naturalism 43
Bunge, Mario 235
antirealism 13
Aquinas, Thomas 93, 244
Campbell, C.A. 143
Arguments-on-Paper Thesis, the 42
Campbell, Donald 199–200
Aristotle 3, 120, 193, 215, 231, 255
Campbell, Keith 69, 77, 100, 102, 103
Armstrong, D.M. 69, 77, 78–9, 83, 84,
early version nominalism 81–5
96–104, 156–7
revised version of nominalism 86–96
‘aspects’ 70
Cartesian dualism 5, 79
Atkins, Peter 236
‘cases’ 70
atomic theory of matter 76
causal theories of action 159–61
attribute-agreement 70
Chalmers, David 157, 175, 177, 178,
Aufbau projects 43
180
Augustine, St 194
chaotic inflationary model 224–5
Aune, Bruce 77
Chomsky, Noam 147
Austin, J.L. 153
Chu Hsi 193
Churchland, Paul 8, 50, 133, 135–8,
Barrow, John 217, 226, 228, 254
141–4, 146–7
Bedau, Mark 125
Clarke, Randolph 164, 165
Behe, Michael 255, 257, 275, 277
Coherentism 43
Darwin’s Black Box 273, 274
complexity as probability 265–70
Bekenstein–Hawking formula 229
complexity-specification criterion
belief 32–3
258–61
Bergmann, Gustav 83
detection of design 270–3
Bhaskar, Roy 73
conceptualism 113
bio-social optimism 199
Council, Richard 101
Block, Ned 140, 141
consciousness 38, 148–9
Bodnar, John 276
conservatism, epistemological principle
Bohr, Niels 255
of 20–1
Bondi, Hermann 219
Contact (film) 258–9
BonJour, Laurence 79
contemporary materialism 156
contingent identity 6
eliminative ontological naturalism 8
conventionalism 113
emergentism 180
Copernicus 51
empirical sciences 5, 8, 10, 11
Copleston, Frederick 215
Epicurus 62, 235
core methodological naturalism 10
epistemic warrant 21, 22
core ontological naturalism 10
epistemological naturalism 20
Core Scientism 10–14, 16–17, 19, 22
ethical intuitionism 27
correspondence 35–6
evolution, theory of 76
naivety of 36–9
evolutionary biology 76
Craig, William Lane 254
Ewing, A.C. 143
Crick, Francis: What Mad Pursuit 255
externalism 44, 74, 75
critical thinking 192
extreme nominalism 70
cumulative complexity 273
extreme particularism 197–8
curved space, theory of 51
falsity 35
Darwin, Charles 76, 77, 274
Fichte, I.H. von 37
Origin of Species 257
Field, Hartry 68
Davidson, Donald 8, 164
Finnis, John 197
Davies, P.C.W. 217, 231
first philosophy 3–4, 69
Dawkins, Richard: Blind Watchmaker, The
Fodor, Jerry 9, 54
125, 255, 275
Forster, Malcolm
De Vito, Scott 60
on simplicity 58–61
Democritus 62
Foster, John 141
Dennett, Daniel 8, 40–1, 50, 77, 134,
foundationalism 43, 92
147, 149
Frankfurt, Harry 159, 160
Derrida, Jacques 36, 38
Frege, Gottlob 25, 41
Descartes, René 5, 6, 24, 25, 173, 174,
Friedman, Alexander 216
204
Friedman–Lemaître (Big Bang model)
Discourse on Method 43
216–18
descriptivism 27
full-orbed naturalism 62
design theory 257
Dewey, John 24, 27, 30
Galileo 255
Dirac, Paul 51
Gamow, George 220
divine attributes 150–1
Gauss, J.K.F. 51
Doe, Jane 138
General Theory of Relativity 51, 216,
Dostoyevsky, F.
231
Brothers Karamazov 208
Gibbard, Allan 200
Dretscke 53
Gilligan, Carol 197
dualism 134, 135, 157
Ginet, Carl 158, 165, 166–7
Ducasse, C,J. 143
Goetz, Stewart 143
Gold, Thomas 219
Eddington, Sir Arthur 217
Goldman 30
Einstein, Albert 44, 216, 227
Gott, J. Richard 231, 232
theory of gravitation 51
Grand Story 75–9
electron, theory of (Dirac) 51
gravitation, theory of 51
electroweak theory 51
Green, T.H. 24
eliminative materialism 135–8
Gribbin, John 235
eliminative methodological
Griffin, James 196
naturalism 9
Grisez, Germain 197
Grossmann, Reinhardt 68, 70
K-containing regions 116
group theory 51
Kenny, Anthony 152
Grünbaum, Adolf 238–9
Kicher, Patricia 30
Guth, Alan 254
Kieseppä, I.A. 60
Kim, Jaegwon 26, 77, 87, 134, 170–4,
Habermas, J. 191
177–9
Hacker, P.N.S. 152
kind concepts 115–16
Hansen, Norwood 42
kind-constituting matter 116
Hare, R.M. 191, 192, 193
knowledge, definition 31–4
Hart, W.D. 143
Kornblith 42
Hartle, James 225
Kukla, Andre 60
Hartle–Hawking models 226
Hawking, Stephen 220, 225, 227, 228,
Leibniz, G.W. 142, 215, 231
253
Lemaître, Georges 216
Hawking–Penrose Singularity
Lewis, David 56, 240, 241–2
Theorems 220
Lewis, H.D. 143
Hegel, G.W.F. 36, 37, 198
libertarian freedom 161
Heil, John 157, 171, 175
Linde, Andrei 224, 225
Heisenberg, W.K. 255
linguistic rule theories 44
heliocentric model (Copernicus) 51
Li-Xin Li 231, 232
Hicks, G. Dawes 90
local naturalists 78
Hook, Sidney 24, 30
Locke, John 36, 37
Horkheimer 42
logical grounding 33
Hoyle, Sir Fred 217, 218
logical psychologism 43
Hubble, Edwin 216
logical relations 41–2
Hull, David 77
Lotze, R.H. 25
Hume, David 32, 37, 81, 133, 215,
Lucretius 62, 235
230, 244
Luther, Martin 194, 202
Husserl, Edmund 24, 25, 41, 43,
Lycan, William 141, 144, 147
71, 98
Conscious and Experience 134
identity materialism 139–47
Mackie, J.L. 133, 135, 152
integrative dualism 152–3
Miracle of Theism, The 133
intelligent design theory (IDT) 257
Maddy, Penelope 68–9
intentionality 50
Malcolm, Norman 215
internal symmetry, theory of 51
Marx, K. 198
internalism 44, 74–5
material constitution, problem of 114
intuitive thinking 192
Inwagen, Peter van 116
materialism 4–5, 6, 29
irreducible complexity 273–5
Mavrodes, George 135
Isham, Christopher 222, 223, 236
Mayr, E. 76, 77
McGinn, Colin 68, 139, 140, 147, 148,
Kagan, Shelly, 195
150, 151, 153, 176
Limits of Morality, The 194
mental causation, nature of 151–2
Kane, Robert 162–4
metaphilosophical naturalism 62
Kanitscheider, Bernulf 235, 237
metaphysical cosmology 225
Kant, I. 24, 25, 33, 36, 37
metaphysics and epistemology
on morality 189–90, 191–2, 193,
143–5
197, 198, 201–8
methodological monism 28
methodological naturalism 9, 14–17
nominalism 67, 70
conceptual necessity 17–18
non-eliminative non-reductive
metaphysical necessity 18–19
methodological naturalism 9
Midas touch picture of consciousness 38
non-eliminative non-reductive
Mijuskovic, B. 151
ontological naturalism 8
Mill, John Stuart 191, 192, 194
non-eliminative reductive ontological
Millikan, Ruth Garrett 53, 54, 56, 124,
naturalism 8
125
non-eliminative reductive
Milton, John: Paradise Lost 35
methodological naturalism 9
mind–body identity 5–6, 134, 151,
nonreductive materialism
169
147–53
mind–brain relation 176–7
normativity 26–7
mind-dependent material objects
Novikov, I.D. 221
118–19
modal antirealism 113–14
O’Connor, Timothy 165, 179–80
‘moments’ 70
O’Shaughnessy, Brian 150
monism 25
ontological dualism 5
Monod, Jacques 255
ontological naturalism 4–6
Moore, G.E. 27
definition 50
moral demand 190–1
ontological pluralism 5
moral faith 201–8
oscillating models 220–2
providence 203–6
overfitting 59
pure rationalism 201–3
experience of evil 206–8
Papineau, David 8, 53, 56, 74, 78,
moral gap 189
156–7
Murphy, A.E. 30
Parfit, Derek 215
Penrose, Roger 220, 229
Nagel 140, 173
‘perfect particulars’ 70
National Academy of Science 268
phenomenalism 37
National Endowment for the
philosophical monism 3
Humanities 10
philosophical naturalism 3
National Science Foundation 10
philosophical Puritanism 30
natural capacities 191–3
physicalism 24, 29–30, 34, 68
natural selection 274
physics-ism 29–30
naturalism
Plantinga, Alvin 43, 122, 122
definition 14–17, 25–7
Plato 3, 24, 25, 67, 99
logical status of 14–19
Pluralism 30
syntactic tautology 14
Poland, Jeffrey 78
traditional realist properties and
Popper, Karl 42
79–80
positivism 37
naturalist ontology 78
possible holy being 193–4
naturalistic monism 30
Post, John 237
naturalized epistemology 26
pragmatism 13, 61
Neander, Karen 124–7
Principle of In toto Location (PIL)
Newton, I 255
97–8
theory of gravitation 51
Principle of Instantiation 99
Nietzsche, F.W. 193, 201, 207
probabilistic resource 266
Noddings, Nel 197, 198
proper-function-phenomena 119–20,
noetic unity 34, 42–4
122
properties 67–104
Search for Extra-Terrestrial Intelligence
attribute agreement 70
(SETI) response 258–9, 270
nominalism 70
Searle, John 28, 23, 73, 76, 78, 147,
realist view 70–2
169–75, 177, 179–81
‘property-instances’ 70
Seibt, Johanna 72, 102
Ptolemaic system 51
Sellars, Roy Wood 32
Putnam, Hilary 24, 26, 36, 38–40
Sellars, Wilfrid 68, 69, 73, 78, 110
Putnam–Kripke–Marcus theory of
Silk, Joseph 221
linguistic reference to natural
simplicity 50–3, 58–61
kinds 19
Singularity Theorems 220
Slipher, Vesto 216
qualia 145–6
Smart, J.J.C. 8, 9
quantum gravity models 225–30
Smith, Quentin 218, 235, 238, 240,
quasi-tropes 86, 90
241–3
Quine, W.V. 6–7, 8, 9, 17, 25–7, 30, 39,
Sober, Elliott on simplicity 58–61
42, 44, 62, 68
social constructionism 43–4
epistemiological naturalism of 19–22
Socrates 112
Sophocles 206, 207
Ramsey, Frank 56
space (space–time) trope 91
Rawls, John 191, 193
spatial extension 4
realism 70–1
specification 261–5
red-shift 216
specificational resource 266
referential opacity 141–2
Spener, Philipp Jakob 202
Reichenbach, Hans 61
statistical normality (SN) accounts of
Relativity, General Theory of 51, 216,
proper functions 122, 123–4
231
Stebbing, L.S. 43
reliability 53–5, 61
stereotopic fallacy 144
replicational resource 266, 267
Stich, Stephen 30, 50
representation, theories of 36
Stout, G.F. 90
representational naturalism 50
Strawson, Galen 140, 147
centrality of reliability to 53–5
strong naturalists 78
Rey, Georges 141, 147, 149
Stubenberg, Leopold 140, 147
Richardson, R.C. 151
Suarez, Francis 86, 92
Riemann, G.F.B. 51
supervenience naturalism 168, 170, 180
Robinson, Howard 67
Swinburne, Richard 143, 151, 152, 234,
Rorty 38, 39
254
Ross, Hugh 254, 257
Russell, Bertrand 215
Tarski, Alfred 21
Taylor, Richard 162
Sagan, Carl 258
teleological properties 53
Santayan, George 24
theism 148–50
scientific naturalism, contemporary
Thomists 197
72–9
Thomson, Judith Jarvis 196
naturalist epistemic attitude 73–5
Thucydides 206, 207
naturalist grand story 75–9
Tipler, Frank 217, 254
scientific realism 56–8, 74
‘tropes’ 70, 81–5
definition 49–50
truth 33, 34–6, 39–41
scientism 29–30
Tryon, Edward 222
Scotus, Duns 86, 96
Turney, Peter 61
‘unit properties’ 70
vitalism 145
unity of apperception 44
universe, origin of, theories 215–30
Wagner, Steven 73
chaotic inflationary model 224–5
Warner, Richard 73
oscillating models 220–2
weak naturalists 78
quantum gravity models 225–30
Weinberg, Steven 51, 52
standard Big Bang model
Wesel, Elie 208
(Friedman–Lemaître model)
Weyl Curvature Hypothesis 229
216–18
Wheeler, John 216
steady state model 218–31
Williams, Bernard 206, 207
vacuum fluctuation models 222–3,
Williams, D.C. 83, 95, 96
237
Wittgenstein, Ludwig 215
utilitarianism 191
Culture and Value 271
Wundt, Wilhelm 43
vacuum fluctuation models 222–3,
237
Yates, John 240
verificationism 43
Vilenkin, Alexander 224, 225, 227, 231
Zeldovich, Ya. B. 221
Vilenkin models 226
Zygmund, Senator Zig 268