When I was in graduate school I took a very interesting seminar taught by James Van Cleve (who was on loan to Duke from Brown University at the time--now he's at USC) called "Realism and Anti-Realism". Virtually all of the readings were anti-realist in their orientation, and one of the students asked why the syllabus had such an anti-realist slant to it. Van Cleve answered "I just assumed that most people were probably already realists, and would be less familiar with the anti-realist literature." That turned out to be an accurate assessment, and the course was one of the most interesting I took at Duke principally because I found a great many of my own a priori metaphysical assumptions being challenged on every front.
That was fifteen years ago, at a time when many of my deeply held convictions were coming under heavy fire not only in my coursework, but from within my own conscience as well. In particular, I was struggling with the task of making sense out of my religious faith in the face of what Bas Van Fraassen has called "the scientific image" of man. As it happens, it was an encounter with Bas Van Fraassen's anti-realist philosophy of science that helped me through this rather difficult period. It occurred to me that a believing Christian (as opposed to the mind-numbingly banal "Christian" Owen Flanagan recommends being) could not possibly accept scientific realism, and Van Fraassen's work introduced me to a rather large body of literature that demonstrated that going anti-realist is a good idea for many other reasons as well.
We may define scientific realism in roughly this way: the task of science is to describe the operations of nature as they really are, on the assumption that representations of nature in the form of scientific hypotheses, formulae, theorems, etc., are literally true or false to the extent that they accurately representing nature. Anti-realism about science can take many forms, but the form that I defend is not particularly complex, it is merely the denial of this form of scientific realism. In particular, the anti-realism that I defend is simply the denial that scientific hypotheses, formulae, theorems, etc., can be literally true or false. Truth and falsity are, of course, rather important norms, but for the anti-realism that I defend a more important norm is usefulness.
To see what I mean, consider three different maps. Map (A) is a satellite photograph of the earth's surface such as you might find using Google Earth, only without any street overlays, place names, etc.: just the photograph of the ground. Map (B) is a road map of the sort you might purchase at the local gas station. Map (C) is a line drawing that I make to show you how to get from uptown Athens to my house. Now, in terms of "accuracy of representation" I think that it is arguable that there is a continuum here from "most accurate" to "least accurate", with map (A) being the "most accurate" to map (C) being the "least accurate". On a "realist" account of mapping, the satellite photograph will be the best, since it is the "truest" in the sense of being the most accurate representation of what's on the ground. But if you've ever used Google Earth to get somewhere without turning on the street overlays, you know that it's almost useless. You need to know street names, directions, distances, etc. So map (B) is far more useful than map (A). And if you have a specific sort of trip in mind, such as trying to get to my house from uptown Athens, then map (C) will be even more useful, since it has been designed to do precisely what you need to do, i.e., get yourself to my house from uptown Athens. So, which map do you want? Well, it sort of depends on what you want the map for. If, like me, you sort of enjoy looking at satellite photographs of the earth, I suppose map (A) will suit you, but most people want maps to get places, and so maps (B) and (C) are superior to map (A), even though they are less true in the sense of being less directly representational. They are, of course, representations, but they are not representations in the sense of being "literally true" in the realist's sense.
So the anti-realist favors usefulness over truth, but what could this mean in the domain of science? Isn't a scientific theory useful precisely to the extent that it's true? This is, in fact, a version of an argument that realists make all the time. They say that scientific realism must be true, because science is so successful. If science did not accurately represent reality, they say, there would be no way for science to be as successful as it is, because the only way for science to be successful is for it to "get things right" about the external world. Hence, to the extent that "science works", it is also literally true.
Now, it seems to me that even a moron could see that this is a circular argument, and yet realists make it all the time. Indeed, it seems as though the anti-realist would have a perfectly simple rejoinder to this kind of argument: if scientific realism is true, then the history of science along with the hypothetico-deductive model of theory formation teach us that all of our scientific theories are always false, regardless of how useful we may find them to be at one time or another. I don't actually want to get into that particular problem, though, because for the Christian that whole issue is academic. Scientific realism also presupposes both global materialism and global empiricism, and the Christian must reject both. Global materialism is false for the Christian because the Christian believes that there is at least one non-material being, namely God. Global empiricism is false because the Christian believes that it is possible to have knowledge of a non-perceptible entity, namely God. Christians, of course, can be scientists, since they may quite easily adopt both local materialism and local empiricism (that is, materialism and empiricism when doing science) without violating the principles of their Christian belief.
So the Christian is interested only in scientific theories that are useful, and is not interested in whether or not they are true (mostly they will have to be regarded as false for one reason or another anyway, so why even get into that whole problem?). Consider the difference between the heliocentric and the geocentric models of the solar system. Both of them have been "useful" at one time or another, though at present we believe that one of them is false. Some would like to believe that the other one is true, but of course that presupposes that because the other one works better that it is also truer. It may turn out to be the case that the heliocentric model will need to be modified as more data are collected, thus showing the present version of it to be false. This is perhaps unlikely, but a genuine scientist cannot rule out the possibility that his present hypothesis is false, otherwise science could never move forward, it could never take into account new data. Rather than say, "this model is false" the Christian says "whether or not this theory is true or false, it's the one that I can use right now to make these predictions or to accomplish these goals." One of the things that led to the abandonment of the geocentric model in favor of the heliocentric was precisely this: one model was more useful than the other. It's not that there were new data available--there weren't. Both models can make a certain amount of sense of the observational data. What's more important than that is that the heliocentric model enables us to do more given the other theories that we also accept. Hence the criterion is not truth (since on that criterion we could never finally accept any theory), but usefulness.
Looking at science in this way enables the Christian to accept a theory like evolution even when many defenders of evolution claim that accepting it means rejecting religious belief. That would only be the case if two conditions were met: (a) evolutionary theory logically excludes the truth of any religion and (b) scientific realism is true. Since both (a) and (b) are false, and since evolutionary theory is very useful in explaining the origins and development of life, the Christian not only can but ought to accept evolutionary theory.