Science in Christian Perspective




The Scientific Status of Theology:
Imre Lakatos, Method and Demarcation

Greg Peterson*

Thiel College
Greenville PA 16125

From: Perspectives on Science and Christian Faith 50 (March 1998): 22-31.  Response: Siemens

Is theology a science? This claim would normally receive considerable skepticism, not only from scientists but from many philosophers and theologians as well. To the contrary, I use the philosophy of science as developed by Thomas Kuhn and Imre Lakatos to argue that the scientific method applies across many different disciplines, including theology. The result is that there is no firm demarcation between science and non-science. One must judge, instead, between good and bad science, as well as between progressive and degenerative research programs. In this light, theology can and should be considered a science, but with significant limitations and qualifications.

Is theology a science? This claim would normally receive considerable skepticism, not only from scientists but from many philosophers and theologians as well. Is not the story of science the story of its liberation from the theological dogmatism of the Middle Ages? Are not religion and science at odds or, at the very least, dealing with quite separate realms (the subjective and objective)? To the contrary, I argue that the scientific method applies and is used across many different disciplines, including theology, and not simply among those natural sciences which usually claim exclusive right to the title of science. This argument is based upon the analysis of scientific method by Imre Lakatos, a well-known philosopher of science, and the further application of his analysis beyond the natural sciences. Lakatos' account, while not perfect, accurately describes the practice of science and reveals that there is no firm demarcation between science and nonscience. The judgment, rather, must be between good and bad science and between progressive and degenerative research programs; between good and bad theology and between progressive and degenerative theological research programs.

What Is Science?  A Lakatosian Approach

While there is some agreement on very basic structures and characteristics of science, there is currently no consensus view on a normative or even simply descriptive account of scientific methodology among philosophers of science. This may strike some (especially scientists) as surprising, since science goes about its business daily with quite a bit of success. Certainly, some features of scientific method are used regularly and are not in dispute. Science frequently relies on quantitative analysis, and those disciplines which are considered "most scientific" (physics, mathematics, chemistry) usually are the ones that rely most heavily on quantitative measurement. Science also involves hypothesis testing. Across disciplines it is generally considered that a hypothesis must be testable to be scientific, though some fields, such as cosmology, stretch this criterion considerably. But other issues, such as whether science is progressive, whether scientific theories represent reality or are merely instrumental, or whether science proceeds in terms of evolution or revolution, remain subject to intense debate.

One real problem lies in delineating what is and is not science. This is the well-known problem of demarcation. Physics is a science. So are astronomy, computer science, and biology. But is economics a science? Or psychology? Or history? All these disciplines can include some level of quantitative analysis and some level of hypothesis-testing. The latter three, however, are usually considered more qualitative and, thus, more subjective. Can a firm demarcation be made between science and nonscience?

Karl Popper thought so, and drew the line between those ideas which were subject to falsification and those which were not.1 Physics, because it could provide hypotheses which could testably be falsified, was a science. Freudian psychology, because its hypotheses were so all encompassing as to be untestable and irrefutable, was not. This strict demarcationism was prevalent throughout much of twentieth century philosophy of science and is still prevalent in some circles today.

The historicist revolution, led by Thomas Kuhn, Paul Feyerabend, and others, significantly changed the landscape in the philosophy of science when their work first began to appear and be recognized in the 1960s and 1970s.2 Among other things, Kuhn argued that the history of science revealed that normal science is remarkably immune to falsification, through the development of ad hoc hypotheses, the conviction that somehow the measurements made were imprecise, or the faith that scientists had not yet considered undiscovered phenomenon. The Newtonian paradigm, for instance, could never accurately predict Mercury's orbit, which can only be satisfactorily described in terms of relativity theory. This anomaly, however, was never considered a serious threat to the paradigm, since elsewhere Newtonian mechanics was so powerfully predictive. It was only toward the end of the nineteenth century, when a range of increasingly disturbing anomalies began to appear that Newtonian mechanics was seen to be in crisis. Kuhn's interpretation, in particular, was seen as a challenge, if not a single knockdown blow, to both Popper's criterion of falsification and nearly all positivist accounts of science.

It is not, however, Kuhn's philosophy of science on which I wish to dwell, but rather that of Imre Lakatos. This is for several reasons. Kuhn and Lakatos together provide two of the most prevalent approaches to the philosophy of science which also take into account the data that the history of science has to offer. Lakatos, however, was more directly concerned than Kuhn with the problems of demarcation and of scientific progress, so that his account is somewhat fuller and more relevant to our purposes than that of Kuhn.

Among Kuhn's difficulties was describing the rationality of scientific revolutions, the switch by scientists from one scientific theory to another. While Kuhn did seem to believe that science is progressive, he nevertheless seemed to think that the switch between paradigms could not be rationally described and even referred to this switch as a conversion experience.3 It is this seeming irrationality at the basis of science that has sparked much of the debate over Kuhn's theory. Lakatos himself derisively referred to this portion of Kuhn's thesis as a matter of mob psychology.4 But Kuhn's historicist approach certainly impacted Lakatos and it is in his response and rejection of Kuhn that one finds also his departure from Popper. Lakatos, it would appear, was trying to find a third way that salvaged the rationality of science as depicted in the Popperian tradition while adopting the more historicist posture of Kuhn and others.

At the center of Lakatos' theory lies the research program which he describes as follows:

The basic unit of appraisal must be not an isolated theory or conjunction of theories but rather a "research programme," with a conventionally accepted (and thus by provisional decision "irrefutable") "hard core" and with a "positive heuristic" which defines problems, outlines the construction of a belt of auxiliary hypotheses, foresees anomalies and turns them victoriously into examples, all according to a preconceived plan.5

Like Kuhn's paradigm, a research program possesses a set of unquestioned assumptions called its hard core. The hard core is the explanatory center of the research program and it is to the core's defense and propagation that the program is devoted. A negative and a positive heuristic then protect the hard core.

The negative heuristic protects the hard core by deflecting criticism from it to the auxiliary (ad hoc) hypotheses stated to be the subjects of examination and test or by simply denying the significance of anomalies with the faithful proviso that eventually scientists will explain them. Darwin, for instance, never had an adequate account of how humans inherited individual traits. It was not until the much later recognition of first Mendelian genetics and then the discovery of DNA that scientists discovered a satisfactory account of inheritance compatible with natural selection.

The negative heuristic protects the hard core by deflecting criticism
 from it to the auxiliary (ad hoc) hypotheses stated to be the subjects of examination
and test or by simply denying the significance of anomalies...

The positive heuristic, on the other hand, serves to strengthen the research program by the discovery of novel facts that the hard core can explain. These novel facts can be either new discoveries or anomalies in older research programs that the scientist's own research program unexpectedly confirms. A second form that the positive heuristic takes is to develop increasingly sophisticated models. While Lakatos' formulation of what counts as a model is unclear (it seems to me that the word "theory" conveys better Lakatos' intention), it seems that a model is a refutable variant of the hard core used to develop plausibility. Apparently, a model is a way of systematizing the hard coreóif one way of systematization fails, it is replaced with another. In this vein Lakatos cites the development of Newton's theory.6 Newton's hard core was gravity as the source of planetary motion. In the course of his work, he went through several different formulations (models) of how this might be explained. First he assumed that planets did not revolve around the sun but revolved around each other. Then he revised this by assuming that the planets were not point objects but had extension, and so on. Each model becomes more elaborate and increases in explanatory power.

In turn, it is this description of positive and negative heuristics, Lakatos believed, that allowed for an evaluation of competing research programs. We can now describe research programs as either progressive or degenerating. First, is one program more consistent than another? Secondly, is the program producing novel facts and how fast is it producing them? Thirdly, is the program plagued by continual modification and addition of ad hoc hypotheses? If the success of its positive heuristic characterizes the program more than the failures of its negative heuristic, it is progressive. If, however, novel facts are becoming scarce or nonexistent while the negative heuristic is having problems and consistent models are becoming more difficult to formulate, the program is degenerating. Lakatos states:

A research programme is said to be progressing as long as its theoretical growth anticipates its empirical growth, that is, as long as it keeps predicting novel facts with some success ("progressive problemshift"); it is stagnating if its theoretical growth lags behind its empirical growth, that is, as long as it gives only post hoc explanations either of chance discoveries or of facts anticipated by, and discovered in, a rival programme ("degenerating problemshift").7

Lakatos asserts, however, that we must give a certain tolerance to degenerative research programs, especially young ones. In contrast to naive falsificationism, there can be no one crucial experiment (here like Kuhn) upon which the research program rises or falls. For a research program to be dealt a decisive death blow, it must have suffered a string of obvious setbacks without any prospects of novel facts expected. There is, in Lakatos' words, no such thing as "instant rationality" by which we can by an immediate appraisal determine which theory may be better.8 It is only by evaluating programs in a historical context that we can objectively evaluate them. With younger programs, this may include having to wait until the program matures. A temporarily degenerative program may rebound. It is only in the face of obvious, continued failure of a program competing with one or more, more successful programs that we are required to abandon one line of research for another.

Several criticisms have been lodged at the Lakatosian approach over the years, but the more severe ones are largely without merit. It has been asserted, for instance, that Lakatos' methodology does not tally itself with the history of science. Lakatos himself described his task as primarily a normative one and not a descriptive one. That is, he was concerned primarily with how science should operate and only secondarily with how science has operated, though, as Kuhn pointed out, that is of considerable relevance too. A second criticism, made by Feyerabend, is that Lakatos' methodology is so general that "anything goes."9 This comment, unfortunately, is somewhat misleading. Anything can go, but it can go for only a limited amount of time. The prescription against instant rationality means that an evaluation of a research program must take place over a range of time. However, that discipline must first be structured as a research program to be identified as scientific (or possibly scientific). Also scientists will hardly take the hard core of a theory seriously if it quickly degenerates without any promise. One thinks here of phrenology in the nineteenth century, which advanced itself as a science but became thoroughly discredited. No one considers phrenology a science today, and one would be hard pressed to find any current champions. Lakatos' account, however, does leave much to be filled in. A comparison with the work of Kuhn is enlightening, for while these two philosophies are usually considered antagonistic to each other, there are significant elements which are actually complementary.

The positive heuristic, on the other hand, serves to strengthen the research program
by the discovery of novel facts that the hard core can explain.

Kuhn, it may be recalled, challenged the older evolutionary approach of the positivists and neo-positivists and claimed that science is characterized by long periods of normal science interrupted by periodic revolutions. At the basis of normal science lies the much popularized notion of a paradigm. Roughly speaking, the paradigm is the core set of beliefs, models, and exemplars that provide the basis for the conduct of normal science. Standard problem sets in textbooks, foundational experiments such as the Michelson-Morley experiment, and particular formulas such as F = ma or e = mc2 can all serve as part of the functional paradigm. Problem-solving, the sometimes tedious everyday type of research that can serve to expand and concretize the paradigm, characterizes normal science, secure within the knowledge of its established paradigm.

As the paradigm develops, however, new anomalies appear and some old anomalies persist. At some point, these anomalies accumulate to such a point as to threaten the veracity of the paradigm itself. At this point, normal science gives way to revolutionary science. Theories compete with each other for acceptance as each tries to account for new evidence that sparked the collapse of the old paradigm. When a satisfactory position emerges, it quickly gains momentum. Scientists relatively swiftly abandon the old paradigm and the foundation of the new theory becomes the new paradigm and new basis for normal science. This switch from one paradigm to another is, according to Kuhn, not fully explainable and he likens it to a conversion experience. At the basis of science, for Kuhn, lies things not fully explainable in rational terms.

In comparing Lakatos to Kuhn, one may observe at the outset that Lakatos relates his notion of research programs to that of Kuhn's paradigms.10 This is, however, incorrect. The point of comparison in Kuhn to Lakatos' research programme is the idea of normal science. In this regard, Lakatos' notion of research programme is clearly an improvement. It possesses greater detail than that of normal science and better explains the competition among programs as opposed to the monolithic unity of normal science. Kuhn, on the other hand, better shows the underlying assumptions shared among research programs. For a field to be even recognized as a scientific discipline there must be some basic level of agreement between research programs for them even to compete. We could say that Newton's laws functioned in this way in the nineteenth century, while Einstein's relativity theory functions this way in the twentieth century. Competing research programs in (say) stellar evolution still have some basic agreements among them.

Secondly, the proper point of comparison of Kuhn's paradigm (as he has further developed it) is Lakatos' hard core. Here it seems that Kuhn is more specific than Lakatos (we may recall the difficulty above of relating the notion of a hard core to that of a model). Arguably, it is this extended notion of a paradigm that best describes what the hard core is, namely the shared matrix of values, beliefs, symbolic expressions, and exemplars that form the basis of the research program. It is important to enunciate, as Kuhn has done, that the hard core often contains different types of statements. Evolutionary biology contains computer models of ecosystems, law-like generalizations, famous examples (e.g., the panda's "thumb"), and so forth. All these are used to varying degrees in the scientist's work and belie the notion that scientific work is simply contained in mathematical operations.

Thirdly, I would argue that Lakatos undervalues the role of puzzle solving in Kuhn's work. While the notion of a positive heuristic is useful, something remains to be said for the more nondescript and often seemingly irrelevant research that many scientists engage in but that sometimes produces interesting and novel (in the full Lakatosian sense) results. This observation refers back to our first: there are different levels of agreement and disagreement. It is the puzzle solving (particularly as manifested in the various engineering professions) that is the predominant scientific work. Puzzle solving, while displaying the disagreements typical of Lakatosian research programs, nevertheless requires some significant agreement (an underlying paradigm).

In spite of this, Lakatos displays a methodology that is both reasonably true to the actual practice of science, while it also answers the questions of scientific progress and demarcation. It is this latter concept, however, that plays a particularly important role for our purposes here. The idea of demarcation determines how we define theology and in what sense we may consider it a science.

Science and Demarcation

Like Popper, Lakatos was concerned with the problem of demarcation: what does and does not count as a science? For Popper, the line that demarcated science from nonscience was the line of falsificationóthose disciplines that could produce falsifiable statements could be construed as a science. The solution that Lakatos proposed, however, was significantly different. Research programs, some of which are progressive and some of which are degenerative, characterize science. When degenerative programs reach a certain nadir, we no longer can consider them science. Lakatos himself refused to say exactly when this takes place and limits himself to stating that we should give younger programs more leniency than older ones. Superstring theory is given a longer leash than steady state cosmology. But how do we know? This is left to the judgment of the scientific community, which presumably operates rationally.

This lack of some sharp criteria of demarcation will trouble many, no doubt, but it does accurately describe the current state of the sciences. No one, for instance, questions whether physics is a science, although there are many who question whether physics can or should operate as the primary model of the sciences as it has done in the past. Physics is largely quantitative and is composed of well-defined research programs whose hard cores share many similarities and are intensely competitive. We demarcate research programs as progressive or degenerative with relative ease in a relatively short amount of time. Quark theory is now firmly established, but bootstrap theory is nearly forgotten.

What of paleontology, a discipline that is usually classed among the sciences although it is not very quantifiable or experimental (two typical criterions of being a science)? Paleontology, from a Lakatosian viewpoint, is a science because it has been and continues to be characterized by progressive research programs, while degenerative programs that do occur are replaced by progressive ones. One may take as an example the fairly recent dinosaur "revolution," championed by Robert Bakker and others.11 The picture of dinosaurs as cold-blooded, lumbering, thoughtless brutes has been significantly challenged by a body of research that suggests that dinosaurs were nimble, warm-blooded animals who cared for their young. Sometimes bitter competition has characterized the two approaches, and while the newer warm-blooded picture has considerable momentum, dissension remains. Both programs are, currently, scientific, although the day will come when one no longer is.

Lakatos displays a methodology that is both reasonably true to the actual practice of science,
while it also answers the questions of scientific progress and demarcation.

Unlike paleontology, psychology has inhabited a sort of scientific netherworld, wanting to be a science but never quite gaining acceptance. The behaviorism of B. F. Skinner succeeded only at the high price of sacrificing nearly everything that psychology was supposed to be about (namely, the psyche). Some current branches of psychology, such as cognitive psychology, are now largely considered within the fold of science, while Freudianism (typically) only gets marginal respect. Was Freudian psychology ever scientific? Under a Lakatosian approach, one may reply yes and say that it was even progressive for a while. But its often sloppy methodology and its increasing difficulty in producing novel facts or reconciling anomalies (such as the superiority of biochemical explanations and treatments of some forms of mental illness to psychoanalytic ones) has resulted in a serious (if tenacious) degeneration of the program. In the twentieth-first century, Freudianism may be treated much as phrenology is now.

One might move, however, from these increasingly borderline cases to one that is clearly seen to be not a science: metaphysics. Why is metaphysics not currently regarded as a science? One might observe that metaphysics, as it is currently practiced, does not identify a single subject matter. Therefore, we cannot organize it into competing research programs, though specific topics within metaphysics may take on the character of a research program and thus be considered as scientific. But how scientific are such programs? As long as they are progressive, they are scientific. If their limited scope inhibits the discovery of confirming data or novel facts, this progress might be limited to such an extent that it would get only marginal recognition or may eventually be seen as degenerating. If neither realism nor idealism has any measurable consequences or tie-in with other theories, then their scientific status is endangered.

Finally, let us take a bogus case: astrology. Can astrology be a science? Presumably, one could organize astrology in a way that follows a Lakatosian research program. There would be a hard core of shared beliefs and exemplars (a theory of how the planets affect human lives, anecdotal accounts of past successes, etc.) protected by auxiliary hypotheses (astrology only predicts general patterns, not specific events) that produce a series of papers (the positive heuristic) predicting several consequences of astrology, reconciling discrepancies, demonstrating successes, and so forth. Of course, one wonders how long astrologers could carry it off, and the suspicion of most of us is: not very long at all. In fact, one may observe that they have already tried this. In the Middle Ages and through the Renaissance period, astronomy was tied to astrology. But as the former gained in power and scope the other quickly began to degenerate, until only vestiges remain today. Astrology is a failed research program.

Is Theology a Science?

Is theology a science, or can it be? Surprisingly, several theologians have, in pondering the status of their discipline in relation to the rest of academia, considered this question. In part, this is because the concept of science as found in continental thought is broader, where the distinction is not between science and the humanities (nonscience) as between the natural sciences (naturwissenschaft) and the human sciences (geisteswissenschaft), the former often characterized as mere "technical" reason, the latter as dialectical reason. In this vein, Thomas Torrance writes of a theological science based upon the revelation of God, practicing a unique method distinct from that of the physical sciences.12 Wolfhart Pannenberg, by contrast, argues that theology is the science of God and that its method is largely the same as that of the physical sciences, though some differences remain.13 Since, in Pannenberg's theology, God is the all-determining reality, all of reality is necessarily the object of theology. Theology differs from the physical sciences, however, in the degree of incompleteness and in its emphasis on the future. Ian Barbour, while not claiming a scientific status for theology, does observe that certain broad similarities exist between theories in the sciences and in religion.14 Barbour relies heavily on the concepts of model and paradigm, suggesting that the roles these concepts can and do play in religious life are very analogous to the ways that they are used in the sciences.

Theology differs from the physical sciences...
in the degree of incompleteness and in its emphasis on the future.

Equally interesting is previous research suggesting that theology can or does act as a research program in the Lakatosian sense. William Austin is, to my knowledge, the earliest advocate of this approach, but Philip Hefner and Nancey Murphy have suggested it as well.15 Of these scholars, Murphy has developed this approach at greatest length. She argues that the Lakatosian method solves the methodological crises that have faced academic theology in recent decades and that if theology follows a Lakatosian approach, then theology can be a science. The concern here is a normative one. In the past, Murphy argues, theology has not been a science, but it can be. Unlike Pannenberg, Murphy does not take the time to state what theology is a science of (namely God), although one might presume this from the tenor of her works. Murphy is concerned, however, with what counts as data and devotes considerable attention to this. Data, for Murphy, comes primarily from religious experience, whether past or present, as "Christian discernment," meaning the relevant Christian community, evaluates it. In Murphy's conception, Christian discernment becomes a theory of instrumentation, a device for evaluating religious experiences as either true or false, much in the way (presumably) a theory of optics allows us to interpret freely the objects under a microscope.16 Similarly, Scripture, which is the result of the religious experiences of others, provides another data set, with theological hermeneutics playing the role of a theory of instrumentation in this case. Murphy is not altogether convincing in these passages, and for the idea of theories of instrumentation to work much more needs to be said. What, after all, is it about religious experiences that count as data? Not the experiences themselves, since they are private. Presumably, we must rely on the verbal or written reports of these experiences and what they purportedly disclose. These statements, presumably, then form the basis of testable hypotheses about God and the world, from which we can build a theological theory.

Murphy argues that the Lakatosian method solves the methodological crises
that have faced academic theology in recent decades and that if theology
follows a Lakatosian approach, then theology can be a science.

More important is Murphy's claim that theology has not been a science, but that it can be. Murphy herself is not totally consistent on this point, as she draws (in my mind not altogether successfully) on the Catholic Modernist movement of the turn of the century as an example of a briefly progressive theological research program. It seems to me, however, that the history of theology reveals numerous theological programs which, in their day were seen as progressive or degenerating. The highly technical scholastic theology of the Middle Ages or the theological programs of Martin Luther and John Calvin and, even later, of Friedrich Schleiermacher and Karl Barth can be seen in terms of a scientific research program. The question remains, however, as to how scientific these programs were. The above discussion about which disciplines we may correctly see as a science now becomes relevant. There it seemed that while the physical sciences possess the clearest claim to the status of science, theirs is not an exclusive one. Psychology and metaphysics, if properly approached, can be sciences. Even astrology can take a scientific approach, although its lack of scientific claim should soon become clear. No discipline is barred, a priori, from the scientific approach, although the attempt to scientize a subject may make inherent theoretical weaknesses apparent.

The great difficulty lies in the fact that while the natural sciences seem obviously progressive, theology does not. One can historically trace the development of Newton's theory from its initial appearance to its near unanimous acceptance within the physics community. One can further trace the demise of that theory in its absolute form with its replacement by Einstein's relativity physics, which again eventually (and relatively speedily) gained near unanimous acceptance within the physics community. Theology, it is argued, only rarely, if ever, achieves this level of consensus, and it is not nearly as obviously progressive as the physical sciences are, if it can claim to be progressive at all.

This dipolar characterization, however, is misleading, as the previous discussion on demarcation should indicate. While the physical sciences do have a broad consensus within specific fields on fundamental theories and goals, disagreement also characterizes them, as the debates regarding cosmology in astrophysics or regarding the extinction of the dinosaurs in paleontology should indicate. Disagreement and conflict are a normal part of science, and without it scientific disciplines would not thrive but ossify. The question is, how much disagreement is permissible among research programs within a given field without calling into question its scientific integrity.

From a Lakatosian perspective, as long as competing research programs characterize a field, it is scientific. The scientific status of any one research program within the field will depend on whether it is progressive or degenerative in comparison to the other programs it is competing with. The real distinction, however, seems to be between good science and bad or marginal science. Recall here our previous thumbnail characterization of several fields from physics to astrology. Most would agree that physics is a good science. Most would agree that biology is a good science as well, though it is not nearly as quantitative as physics is. Economics, by some lights, is on its way to being a good science. Psychology, however, is still in muddy waters in the eyes of many.

 The great difficulty lies in the fact that while the natural sciences seem obviously progressive,
 theology does not.

A distinction, then, needs to be made between the research programs themselves, and the scientific character of the field as a whole, with the latter being dependent on the health and coherence of the competition among individual research programs. From this perspective, theology qualifies as a science in the Lakatosian sense of the term, since it is characterized by competing research programs that may be variously described as progressive or degenerative. But theology as a field does not represent a good or strongly rooted science in its present state, because of the level of disagreement, not only on particulars but on basic methodological questions. One might observe that while we can characterize some research programs in theology as scientific (with a hard core, auxiliary hypotheses and instances of confirmation/disconfirmation), others may not be.

From a Lakatosian perspective, as long as competing research 
programs characterize a field, it is scientific.

One might even observe that this is an occupational hazard of "higher level" disciplines, that the greater the level of generalization required the poorer the grade of science that can be achieved. Physics, by its very nature, will always be more precise (and in some ways more limited) than chemistry, which is more precise than paleontology, which is (possibly) more precise than psychology. Each scientific discipline has its own character and its own strengths and weaknesses. Theology, arguably the most general of all disciplines, would naturally suffer from this reality as well.

Is theology, then, a science? One can answer yes, but with a few qualifications. We can broadly construe certain theological research programs as scientific in character. We might fairly say that neo-orthodox theology was a progressive research program in its heyday. It possessed a hard core based upon the self-revelation of God and was arguably progressive in relation to its competitors on many fronts. Work done in the analytic philosophy of religion on very specific issues (such as arguments over the existence of God and over God's attributes of omniscience and omnipotence) has also reached a high level of technical formulation and expression and possibly represents another progressive research effort.

But while theology may be considered scientific, currently we cannot say that it is, on the whole, a good science. The level and kind of theological disagreement are remarkably high when compared to well-formulated sciences. When one considers that Christian theology is in competition with Islamic and Jewish theology, and all three are in competition with nontheistic programs (such as Buddhism), the level of disagreement is even higher.

The real question, then, is whether theology can be a good science, or at least a better one. Murphy, in essence, argues that theology can be such. The possibility is worth considering, although the difficulties are rather steep, as many important changes would have to occur. Theological research programs are, historically speaking, much more loosely knit than scientific ones. Both Kuhn and Lakatos emphasize that much of the core of a research program is unspecified, but I am inclined to believe that it is nevertheless better articulated than many theological programs. To be a good science, theology would have to be composed of well-formed, competing research programs operating within a coherent framework. Furthermore, some of these programs would have to be progressive, discovering novel facts or resolving old anomalies. The category of novel fact is probably the most troublesome and its applicability will depend mostly on the theology at the base. One can, of course, imagine what might count as confirming facts: answer to prayers, changes in attitudes/lifestyles among influenced groups, achievement of social justice, developments in cosmological theory and so on, depending on the program in question. But for novel facts to be found would mean that theologians would have to look for them. Do they? I would argue that in, some sense, many theologians do, by appeal to better interpretations of scripture, accord with the latest physics or other criteria. But this is not done in a rigorous and premeditated fashion. Can theology do this? Probably, but it would require a significant reorientation for many theologians.

While theology may be considered scientific, 
currently we cannot say that it is, on the whole, a good science.

These would be the minimum requirements for such an approach. Whether it would succeed or fail is beyond my guess (success being relative in this context anyway) and would be as dependent on the personalities and the theories employed as the method itself. It may be, too, that the nature of the data, the higher order character of theology, ultimately prevents theology from being a good science, much as the higher order character of the social sciences force similar problems on them. But if God is a true God, then God is made manifest in the world, whether through nature, Scripture, or religious experience. To the extent that God is faithful, and rational, we can put forth a scientific theology which relates the activity of God to the world.

Some Considerations and Conclusions

The idea of theology as a science will strike many as difficult to accept. We have often seen religion and science as traditionally opposed, and likewise faith and reason. From a Lakatosian (as well as Kuhnian) perspective, this dichotomy is poorly constructed. One should instead speak of a faithful reason. The prohibition against "instant rationality" means that there is a faith component in one's choice between scientific programs. At any given time, then, the choice between research programs is governed not simply by rational considerations, at least in the usual sense. Faith (of a limited sort) or a hunch also governs it. The relationship between faith and reason are more complex than the usual antagonistic faith-science dichotomy would indicate.

Even so, we may object that theology seems to lack several features typical of the sciences. Science depends on the strict demarcation (at least in theory) between theory and data. Science requires testable hypotheses. Can theological hypotheses be testable? The answer here seems to be yes, although it would take considerable effort to indicate in what way theology may be testable. Murphy, in her discussions of religious experience and Scripture, has taken one approach to testability and the data/theory distinction. Even if one disagrees with her specific proposals, her work can nevertheless serve as a guide to further progress. One should observe that there are different grades of testability, and this has been an insight of the historicist approach to the philosophy of science. The falsificationist program inspired by Popper saw science in terms of bold theories that thrived or failed in the face of critical experiments. The more historicist approach taken by Lakatos recognizes that such events are rare and predominantly recognized well after the fact, such as with the Michelson-Morley experiments that were only much later seen as the critical test for the existence of ether. One may observe as well that there are grades of confirmation/disconfirmation, and that this is especially true of "higher level" sciences. To refer to paleontology again, the discovery of a layer of iridium and of a massive crater in the Caribbean basin certainly establishes that an asteroid did indeed hit Earth at the end of the Cretaceous period, but it has proven much more contentious to link that directly to the extinction of the dinosaurs.17 One should only expect the same level of ambiguity, if not more, in theology.

Finally, in recognizing theology to be a science, are we putting science on a pedestal, as the only valid source of knowledge and thereby denigrating other ways of knowing? Such a question assumes that there is a hard and fast distinction between scientific knowledge and nonscientific knowledge, whereas part of the task here has been to argue that there is no easy distinction. Even so, science, as presented here, does constitute a certain form of activity. Science is both a social and critical activity and we need both aspects for a research program to be possible at all. Theology has historically been both social and critical when it is at its best, so the inclusion of theology among the sciences has some merit to it. Certainly, there are other forms of knowledge that are generally considered nonscientific, although I have not taken the time to elucidate them here. The everyday knowledge of my work telephone number, how to do the dishes, and how to write a good letter are generally not considered forms of scientific knowledge. Nevertheless, these things still constitute knowledge of a kind.

The challenging task will be to see to what extent theology can be construed as a good science
 and what level of theological consensus and progress can be achieved in a world 
that seems to be increasingly global and increasingly fractured at the same time.

One might further observe that there remains a distinction between theology and religion. Religion is the living of the faith; theology represents a second order reflection on that faith. The two are ultimately interconnected. Theology, at its best, clarifies and elucidates faith. Faith, in its day-to-day living, poses many problems, at least on a personal level, that theology has to deal with.

The more difficult task, however, still awaits us. A Lakatosian approach aids us in seeing how we can say that theology can be a science. But currently, theology exists in a type of scientific netherworld. While we can say that some programs have a broadly scientific approach, the sheer diversity of approaches tends to cancel the effectiveness of theology as a field. The challenging task will be to see to what extent theology can be construed as a good science and what level of theological consensus and progress can be achieved in a world that seems to be increasingly global and increasingly fractured at the same time.



1Karl Popper, The Logic of Scientific Discovery (London: Hutchinson, 1959).

2Thomas Kuhn, The Structure of Scientific Revolutions (Chicago: University of Chicago, 1970); Paul Feyerabend, Against Method: Outline of an Anarchistic Theory of Knowledge (London: NLB, 1975).

3The Structure of Scientific Revolutions, 158. It should be pointed out, however, that Kuhn does not see progress in terms of increasing verisimilitude, but more instrumentally in terms of the ability to solve a greater range of puzzles. See Thomas Kuhn, "Afterwords," in World Changes, edited by Paul Horwich (Cambridge, MA: Bradford Books, 1993).

4Imre Lakatos, "Falsification and the Methodology of Scientific Research Programmes" in The Methodology of Scientific Research Programmes, edited by John Worrall and Gregory Currie (Cambridge: Cambridge University Press, 1978), 91.

5Imre Lakatos, "History of Science and its Rational Reconstructions," in The Methodology of Scientific Research Programmes," op. cit., 112.

6Lakatos, "Falsification and Methodology," 50ñ1.

7Lakatos, "History," 112.

8Lakatos, "Falsification," 87.

9Paul Feyerabend, in Imre Lakatos and Alan Musgrave, Criticism and the Growth of Knowledge (Cambridge: Cambridge University Press, 1970).

10Lakatos, 91, n. 2.

11Robert T. Bakker, The Dinosaur Heresies: New Theories Unlocking the Mystery of the Dinosaurs and Their Extinction (New York: William Morrow, 1986).

12Thomas Torrance, Theological Science (Oxford: Oxford University Press, 1969).

13Wolfhart Pannenberg, Theology and the Philosophy of Science, translated by Francis Mcdonagh (Philadelphia: Westminster Press, 1976).

14Ian Barbour, Myths, Models and Paradigms (New York: Harper & Row, 1974); Religion in an Age of Science (San Fransisco: HarperCollins, 1990).

15William H. Austin, "Religious Commitment and the Logical Status of Doctrines," Religious Studies 9 (1973): 39ñ48; The Relevance of Natural Science to Theology (London: Unwin Brothers, 1976); Philip Hefner, The Human Factor: Evolution, Culture and Religion (Minneapolis: Fortress Press, 1993); Nancey Murphy, Theology in the Age of Scientific Reasoning (Ithaca, NY: Cornell University Press, 1990).

16The application of theories of instrumentation to theology is elaborated in Nancey Murphy, "What Has Theology to Learn From Scientific Methodology," in Science and Theology: Questions at the Interface edited by Murray Rae, Hilary Regan, and John Stenhouse (Grand Rapids, MI: Eerdmans, 1995).

17See William Glen, ed., The Mass-Extinction Debates: How Science Works in a Crisis (Stanford: Stanford University Press, 1994).