What are those criteria? We've been looking for them, so far without notable success. What's shaped our inquiry so far is the hope that there is a simple, clear, sharply defined line between starters and nonstarters in the race for truth. It's high time that we reconsidered our hope since, it seems, it's gotten us nowhere worth staying.
Our working model of successful theories has led some to embrace falsifiability as the essential mark of the scientific. If a theory has any observational consequences at all, then it should be possible to determine whether it is false by making the appropriate observations. If, on the other hand, a theory has no observational consequences at all, it will be impossible to test observationally, and thus incapable of predictive success or predictive failure.
Creationists have sometimes charged that, unlike the rest of science, evolutionary theory is unfalsifiable and so not science. At least one historian has argued that Creationists operate with an outdated conception of science, one that was current during the nineteenth century, when their position was first enunciated in this country. Whether this claim is historically correct, it is clear now - and has been clear to philosophers of science for at least thirty years, and perhaps three hundred years - that the naive view of falsifiability and the 'predictive success' model of theory assessment on which it is based are both fundamentally wrong. They do not give a correct idea of how science has ever worked. This can be shown by considering one of the greatest genuine successes of Newtonian mechanics, from which we can also see how that insight comes into play.
Newtonian celestial mechanics is one of the star achievements in the history of science. It was used to compute the planets' orbits to extraordinary accuracy, and it is still used by NASA in determining the trajectories of rockets. As predictive successes go, it was a great one. By 1830, it was quite clear that the observations of Uranus's path diverged from the one predicted by Newtonian mechanics - it wobbled. So something had to give. Astronomers were faced with three main options: (a) give up on Newtonian mechanics altogether; (b) suppose that there were special nongravitational forces operating on Uranus alone; (c) suppose that there was an unobserved mass exerting a perturbing gravitational force on Uranus. No one seriously considered taking the first option - but if predictive success were really all that mattered, then we'd have expected (a) to be actively considered. The fact that it was not is a clear signal to us that Newtonian mechanics had clearly perceived and well-appreciated virtues other than predictive success. The third option was independently suggested in 1845 by two astronomers (Adams and Leverrier). They actually computed what the orbit of a perturbing planet would have to be to explain the deviations of Uranus from the predicted path. Astronomers looked for the perturbing planet, and, in 1846, Neptune was discovered. (A similar procedure was followed later and led to the discovery, in 1930, of Pluto.)
Although scientists saw that Newtonian celestial mechanics was worth saving, it was not just because they thought it was a predictively successful theory. It was because it was part of a larger theory that gave a highly unified explanation of much of what goes on in the physical world. There are two aspects of this moral to the story. The first aspect has to do with unification: Newtonian physics had a single small set of principles and strategies to use in solving a very broad range of problems and answering a very broad range of questions about the physical world; it had the power of generality that we have mentioned before (in our discussion of phrenology's alleged lack of systematicity). The second aspect of this moral has to do with explanation. Although we do want predictive success from scientific theories, we want more; that is, we want to know more than what will happen, we also want to know why it will happen (or why it has happened, or why it is happening - explanation is not restricted to the future; neither, by the way, is prediction). There have been many times in the history of science when perfectly regular connections have been observed and summarized in what are called "phenomenological descriptions"; but we are almost never satisfied by mere descriptions of a regularity or pattern. We also want to know what the underlying reason is for the occurrence of that particular pattern. To take one example from the recent history of science: atomic spectra were observed around the turn of the century. When atoms were bombarded with electrons or photons of certain energies, they would emit radiation in highly regular series, or spectra, and, contrary to all contemporary expectations, the spectra were full of gaps instead of being continuous. But being able to predict that they would emit certain patterns of radiation was not enough: physicists wanted to know why there were gaps, what the reason was for their occurrence. That was what drove the search for a theory of atomic spectra. To be sure, having that theory allows additional predictions to be made, which is a desirable consequence; but the desire for explanation remains. From current physics, we have another example: there are just four known forces in nature, gravitational, electromagnetic, strong nuclear, and weak nuclear. Ever since it's been thought that these are all the forces, physicists have been hard at work looking for a deeper explanation of how they are related to one another, how they might be different manifestations of a single, underlying reality. Grand Unified Theories (GUTs) promise a unified understanding of all but the gravitational force. Totally Unified Theories (TUTs) promise such understanding of all four forces in one unified scheme. One of the hottest contenders for a TUT - Superstring theory - says that space has nine dimensions, six of which you haven't noticed because they are curled up like tiny six-dimensional spheres attached to each point in three-dimensional walking-around space. That shows you how far scientists will go to get a wholly unified explanation.
The best scientific theories suggest new avenues of research, new questions to ask, and new ways to look for answers. They have what Kitcher calls "fecundity." Whatever its flaws may have been - and there were plenty - phrenology did force neuropsychologists to give serious consideration to a new hypothesis: localization of brain function. With the traditional Cartesian view in place, investigation of 'higher' mental functions seemed pointless. It is largely thanks to phrenology that modern neuropsychology enjoys great fecundity. This is not to say that any theory that causes a lot of activity and investigation is a good one. An awful theory can do that. The point is that scientific theories, even very good ones, can be incomplete, and that that can be a good sign, or not necessarily a bad one. (This issue comes up in evaluating evolutionary theory when it is criticized for not giving a complete account of the origin of life on Earth. What drives the criticism is a misguided metaphor: the theory is seen as a skyscraper which must have a firm foundation. But a better metaphor is that of a vast, intricate, but incomplete mosaic or jig-saw puzzle, with many pieces filled in, and known constraints on how the rest will look. Part of the picture that's missing some pieces is the part that tells us how self-replicating molecules were formed in the first place.)
In the discussion of phrenology, we examined the idea that the most systematic theories are the best ones. The six qualities that make an explanation systematic come in degrees, so theories can be more or less good depending on how much of each they have. If we wish, we may reserve the label "pseudoscientific" for those theories that are almost wholly lacking in all of these qualities. But there's an important point about Kitcher's us of the term "pseudoscientific": the difference between science and pseudoscience is not a sharp difference - as between starters and nonstarters - but a difference in degree. Theories can be more or less pseudoscientific, and there can be borderline cases. So if we employ Kitcher's way of drawing the distinction, we should drop the race comparison, and talk simply of good, mediocre, and bad theories. Our concern will then be with placing theories in a (multidimensional) continuum, and not with deciding who shouldn't be allowed to start the race. Evolutionary theory belongs in the best part of this spectrum. There is no concise, handy principle with which to dismiss theories that deserve to be discarded: there is no substitute for the hard work of seeing whether a theory has the three complex and subtle characteristics in question. Kitcher's own practice is good example: his book's very title makes his attitude toward Scientific Creationism quite clear, but the only way to justify that attitude is to go through a book-length examination of the Creationist criticisms and proposals.
With all that in mind, we can evaluate Scientific Creationism fairly.
[I've often been asked if I 'practice what I preach' - always taking time to do the hard work of evaluating every idea that comes my way. The answer is, of course, "No," and that's not what I meant to be preaching. I do what most people do: use a mish-mash of principles and desires to decide what to investigate thoroughly. Like most people, I have some fairly well-defined professional and family obligations that help narrow the field. Otherwise, I'm driven by curiosity, when I can afford that luxury. Among the principles I use are those that most of us use to decide when (not) to trust a source of information or assertion. Does she have relevant professional credentials? Has she been a reliable, reasonable source in the past, and if so, is there anything special about the present case that should make me skeptical - for example, she hates the theorist whose criticism she's criticizing, she hasn't had a decent night's sleep in six weeks, etc. So I am not counseling the abandonment of common sense.]
It's worth pausing to look at a Scientific Creationist criticism of evolutionary theory that threatens to gut the latter if it succeeds. In raising some issues about the nature of order and disorder and how they can be explained, it will help us to evaluate Scientific Creationism's use of the notion of design.
Everyone agrees that living things, taken either individually or in populations, are highly - even hierarchically - ordered. A, perhaps the, main goal of biological theories - Scientific Creationism and evolutionary theory among them - is to give an account of this Big Fact. Evolutionary theorists delight in surprising people by saying that the main driving force behind evolution is a random mechanism, namely, mutation, which produces new bits of genetic information (alleles). This sounds counterintuitive because, when we think of randomizing, we think of disorder. Scientific Creationists urge us to hold tightly to that contrary intuition; it spells doom for evolutionary theory. A quick trip to the dictionary instructs us that random mechanisms operate purely by chance. When events occur by chance, they follow no predictable patterns and so cannot be described by the highly regular patterns mandated by natural laws. Instead, random mechanisms, by definition, produce disorder - as anyone who's shuffled a deck of cards surely knows. It's just not plausible to suppose that the random mechanism of mutation could produce the very high degree of order that it is the goal of evolutionary theory to explain. And dragging in natural selection as a constraint on the disorder is hardly enough to yield the extraordinarily high degree of order that living things typify; constrained disorder is still disorder. So, evolutionary theory deploys precisely the wrong sort of conceptual resources to achieve its explanatory goal. The theory is rotten at its conceptual core. To summarize:
Random Argument
If evolutionary theory is true, then the order of living things is produced by a random mechanism (i.e., mutation).
Random mechanisms operate purely through chance, so are not governed by law and produce disorder.
Living things are highly ordered.
Therefore, evolutionary theory is not true
The first and less serious flaw in this argument is in the first premise. Although it is generally agreed that mutation is the more fundamental of two mechanisms for producing variation, there is also recombination of existing bits of genetic information. The first premise focuses on mutation to the exclusion of recombination.
But this is a relatively minor flaw compared to the gross error in the second premise. There is a distinction - not peculiar to evolutionary theory, but familiar to all of us - between apparent randomness and real randomness. Many complex processes are wholly deterministic in the sense that if one knew all the relevant values of variables, one could in principle predict the outcome. But many such processes are so complex that it's unlikely that we will in fact know all the relevant values of the variables. Although the motion of every piece of matter involved may be rigidly predetermined by law-governed physical forces, it all seems 'random' to us because of our partial understanding of how the forces are acting on each piece at each moment. Suppose, for example, that I hurl an open box of Ping-Pong balls into a room full of people. There'll be Ping-Pong balls everywhere, bouncing off walls, furniture, ceiling, and the room's occupants. Eventually, things will settle down and there'll be a final distribution of Ping-Pong ball resting places, aptly describable as a chaotic mess. Although I don't recommend it as a pastime, we could determine this final distribution on a ball-by-ball basis, calculating the trajectory of each through the obstacle course. Because there is no theoretical or practical purpose served by doing so, we are content to rest instead with incomplete, statistical descriptions of the final distribution. In this sense, apparent randomness is an indicator of acceptable levels of ignorance about complex systems.
Much of what evolutionary theorists talk about as random is just this apparent randomness, and a lot of progress has been made in molecular biology and mathematics toward fuller understanding of the processes involved. Another kind of randomness may play a significant role in mutation, operating at the molecular or atomic level. If what a system will do next is a really random event, then it is in principle unpredictable, even on the basis of all the information there is about the system's present state; there is simply nothing in the system now that determines exactly what it will do next. According to what most physicists think about certain atomic processes, atomic events are really random. But even in this case, there are also extremely well-verified laws governing groups of such events (the theory that explains almost all chemical interaction, quantum electrodynamics, has been verified to over ten decimal places), the probabilistic laws of quantum mechanics, which give numerically and geometrically precise descriptions of how these random events array themselves in groups. Ultimately, these laws generate the highly ordered Periodic Table of the Elements (see below). The charge that no patterns are to be expected there is thus wildly off the mark.
The time has now come for us to assess Scientific Creationism as its proponents insist that it be assessed: as a scientific theory. In our discussion, we must keep firmly in mind that it is the theory, not its proponents or its origins, that is to be assessed. Its proponents may themselves be good, bad, or anything in between: that's irrelevant. They may have a hidden political agenda: that's irrelevant. The view may be religiously inspired: that's irrelevant, too. So it is no criticism of the theory in this context to say that its proponents want to change the educational and political system by imposing their interpretation of the Bible and what they believe follows from that interpretation. Even if all of those charges are correct, they detract not one whit from Scientific Creationism itself.
This stricture cuts both ways, of course. In this context, it is no defense of the theory to observe that it is religiously motivated, that it accords well with one interpretation of the Bible, or that its defenders may be well-intentioned clergy, theists, and parents who want the best for their children. (And if Scientific Creationism's proponents do wish to claim that its fitting their interpretation of Genesis is evidence for the theory, then they open themselves up to challenges for justification and must admit at least the possibility of counter-evidence as well. Would that be a welcome result?)
So let's take a look at Scientific Creationism. When we try to do so, we encounter a rather odd problem. Evolutionary, right or wrong, is a complex, well-articulated system of hypotheses and explanations. It's been under development for over a century by lots of smart people, and it takes many hundreds of textbook pages just to summarize its current state. But it is very difficult to find any theory at all in the writings of Scientific Creationists. Once one gets beyond their negative remarks about or criticisms of evolutionary theory, there is little in the way of a positive view to be found. A brief look at the best of what they have to say turns out to be enough.
There is no disputing the fact that the fossil record is ordered, indeed stratified. Scientific Creationism claims that the record was not produced by evolutionary mechanisms but by a Great Flood that rearranged the remains of organisms made in the initial, simultaneous creation. There is a good deal of vagueness in what they have to say, and a lot of backtracking and sheer avoidance of difficult questions. But the most serious challenge to this claim involves the matter of differential evidence: we have no better evidence for the claim that aliens have periodically tampered with Earth's biosphere, than we have for believing the Great Flood story (Kitcher, 132). It is true that the Great Flood story may accord better with what's suggested by the Bible, but why should that count as evidence here? As long as we are to understand Scientific Creationists to be proposing a scientific theory, as they insist we do, we are entitled, even obligated, to ask them: What independent evidence is there for the truth of this Biblical account, as you interpret it? And they have nothing persuasive to offer.
At this point, and at many others, the proponents of Scientific Creationism take refuge in ignorance and mystery. They say that the detailed mechanisms for ordering the fossil record are forever beyond the power of human beings to know, that the ways of God are many and mysterious, and that it all goes according to God's plan for the world, the Great Design of the Greatest Designer.
What exactly is wrong with this? There are two possible criticisms here, and we must not confuse them. Perhaps you expect me to object at this point that mentioning God makes their reply unscientific. Surely, you might say, if a theory makes reference to God, then it's got to be religion masquerading as science:
Scientific Creationism is pseudoscience because it makes reference to God, the Great Designer of the Universe.
But I do not want to object to Scientific Creationism in this way, and neither do any other clear-headed critics. For example, Kitcher explicitly declines this option, remarking that, " ... postulating an unobserved Creator need be no more unscientific than postulating unobserved [sub-atomic] particles." (125) And we have already considered and rejected Gilkey's criterion, which is in the same family as this one. So reference to God is not the issue. (I'll return to this point in discussing Separation by Reference.) To see what the issue really is, let's first look at another claim made by Scientific Creationism.
One of the central tenets of Scientific Creationism is that all the basic kinds of organisms were created simultaneously; it allows, however, that since the Great Flood, some change has taken place in the forms of some organisms, but not in the basic kinds. So the rule for organisms and their ancestors seems to be: once of a basic kind, always of that basic kind. This allows change of only a rather modest sort, since too much change would be incompatible with the preservation of the structure of basic kinds.
One of the most striking recent successes of evolutionary theory is in accounting for the newly discovered biochemical similarities among organisms. Humans and chimpanzees not only share a great deal of genetic material, but are genetically similar, in precisely specified ways, to horses and carp. So, according to evolutionary theory, animals that share a relatively recent ancestor have proteins with similar structures. Can Scientific Creationism account for this sort of similarity? It is sometimes claimed by proponents that these biochemical similarities track with the more traditional similarities among basic kinds, but this is not true in general. For example, humans and chimpanzees are alleged to be of different basic kinds, despite the virtual identity of their cells' biochemical structures. We won't go too far wrong if we think of each basic kind of organism as the object of a separate decision and creative act by God: He saw that the plan required beetles, so He created beetles; He saw that the plan required humans, so He created humans; etc. Given God's nature, however, it was within His power to have fashioned each kind of organism with wholly disparate genetic codes or even from entirely different materials, one carbon-based, another silicon-based, a third from some other material stuff. There is nothing in Scientific Creationism that would lead us to expect the actual overlap among organisms of different kinds.
When pressed, proponents of Scientific Creationism once again claim that the structural similarities and differences among organisms of different basic kinds are all part of God's Great Design: organisms have the features that they have because they need them to serve their part - their purpose - in the divine plan. What is wrong with this reference to design is not the invocation of a deity, but the fact that it explains nothing. It's not that the notion of a Great Design is used that's objectionable; it's how it is used in this context.
In this context, I find it very helpful to look at creation as did the seventeenth century philosopher, mathematician, legal theorist and theologian, Gottfried Wilhelm von Leibniz. Leibniz expended an enormous amount of his considerable intellectual resources on trying to understand creation and he devised an excellent model. According to this model, the universe might have been quite different from the way it actually is - there are infinitely many possible universe designs. Some differ little from the actual one (you and I might be missing from the plan); others differ radically in omitting living things altogether and in having entirely different natural laws (inverse cube, say, instead of inverse square). As an omniscient being, God knows every detail about every possible design for a universe. As an omnipotent being, it is within God's power to implement any one of these designs. But as a supremely good, morally perfect being, it is in God's nature to choose the very best design and to create a universe in accord with it. So, Leibniz argued, this, the actual universe, must be the best of all possible worlds. And since God would not have populated this world with sentient beings wholly unable to appreciate any of the details of His design, Leibniz held, we are endowed with a divinely given cognitive capacity to discern these details and to appreciate them through at least a partial understanding of their mutual fit. Science, on this view, may be thought of as the attempt to discover the details and to answer the question, Why this design above all others? There are more than a few scientists who view themselves as professional students of God's Great Design. Some are even evolutionary biologists, who study (what they take to be) the divinely initiated production process that resulted in us and the world as we find it.
Once we accept the Scientific Creationists' invitation to assess their theory on its merits, we find that their vague gesture in the general direction of a Great Design gives no explanation. If anything is clear about the organisms on Earth, it is that their needs are in large part determined by their 'design'. If fish didn't need to eat, they wouldn't have mouths, digestive tracts, etc. Until we are shown, at least in outline, what the Great Design is and how, in detail, it is a good - indeed, the best possible - design, we have been given no explanation of how its parts allegedly fit together. So if the reference to Great Design is to offer any hope of explanation, it must be specified in much greater detail than Scientific Creationism tries to do. After all, lots of designs were possible: why this particular one?
If, at this point, Scientific Creationism retreats into mystery and ignorance, then it really has withdrawn from the competition with evolutionary theory. It has conceded that explanations of the order of living things are impossible.
But the picture is even worse for Scientific Creationism because, in the end, it is not just one theory in one corner of biology that its proponents urge us to flush; geology, chemistry, and physics will be forced down the drain, too. To see why, let's look at what Scientific Creationism has to say about the age of Earth.
One of the most widely-known tenets of Scientific Creationism is its claim that Earth is not over 4 billion years old, as most of science would have it, but just a few thousand years old. The details of the procedures used by most scientists to date the rocks on Earth and Earth itself are complex and rooted in extremely well-confirmed, thoroughly tested theories in physics, chemistry, and geology - not to mention evolutionary theory itself. In radiometric dating of rocks, one measures the relative proportions of certain radioactive elements in rocks and then computes the rock's approximate age using a known rate of radioactive decay. Since the way in which the procedures are applied is genuinely complex, it is open to misunderstanding and challenge at a number of points. Kitcher discusses several challenges, responds to them, and tells you where to find additional information. I will mention just one challenge made by Scientific Creationism for the sake of illustration.
In determining the age of rocks by radiometric dating, one assumes that there is a constant rate of decay for each element under consideration. Some Scientific Creationists challenge the assumption of constancy of the rate. That assumption, however, is not a physicist's fantasy but a consequence of the way that nuclear forces operate. Such forces are well understood by physicists, and the particular claim of constancy has been subjected to numerous tests and has passed with flying colors. What we are then faced with here is a particularly stark choice: Scientific Creationism or modern physics.
Notice, as well, that even if Scientific Creationists were to succeed in raising well-taken doubts about dating the age of Earth, they would not thereby establish that its age is but a few thousand years. That requires additional justification - and, as I've emphasized before, according with an interpretation of the Bible is not by itself sufficient justification.
Similar conflicts arise between Scientific Creationism and other well established chunks of contemporary science. In the end, we must reject either Scientific Creationism or much of what is best in science, including its standards and methods.
Go to next section on Angelic Science
