A Comparative Study in Science and Religion by Ian Barbour
Difficulties in religious language have been described by many authors in recent years. In Germany, Rudolf Bultmann has said that modem man can no longer speak of a God who acts in nature and history and has proposed a ‘demythologized’ version of the gospel. In England, Bishop Robinson’s Honest to God became a best seller, partly because of his frankness in expressing doubts about traditional ways of speaking of God. In the United States, three theologians who found themselves unable to accept theistic assertions were presented in the popular press as the ‘Death of God’ movement. These men are symptomatic of a widespread questioning of classical formulations.
There are many reasons for current debates about religious language. Biblical statements, if taken literally, are not credible to modern man. The God ‘up there’ is incompatible with our understanding of the universe. Classical discussions of the symbolic and analogical character of religious language were dependent on the metaphysical assumptions of Platonism or scholasticism, which can no longer be presupposed; more recent interpretations often hold that religious images are only symbols of man’s subjective life. The possibility of meaningful language about God is widely disputed today. Theological doctrines, on the other hand, seem to be divorced from human experience. Religious ideas without an experiential basis appear abstract and irrelevant.
For other persons, the encounter of world religions has led to the adoption of a total relativism in place of exclusive claims for a particular tradition. The confidence of the Catholic community in the authority of the church and the conviction of Protestant neoorthodoxy concerning the exclusiveness of revelation have been weakened by the new awareness of religious pluralism. The diversity of religious rituals and beliefs has been taken as support for historical and cultural relativism. Whereas teaching in theological seminaries had assumed the truth of one tradition, the growing study of religion in secular universities has been concerned about its functions in human life -- without reference to the question of its truth or falsity. Often this has ended in the reductionist view that religion is entirely the product of psychological and sociological forces.
One might also point to the secularization of contemporary society, which itself has many facets: the separation of political and educational institutions from the church, the autonomy of the intellectual disciplines, the dominance of this-worldly over otherworldly interests, the confidence in man’s ability to control his own destiny without divine assistance. But the present volume is concerned with the basic conceptual and methodological problems of religious language, and here the most significant influence has undoubtedly been science.
In past centuries, particular scientific theories have had a major impact on religious thought. In the eighteenth century, Newtonian mechanics led to a mechanistic view of the world and a deistic understanding of God the cosmic clockmaker. In the nineteenth century, Darwin’s theory of evolution encouraged new interpretations of divine immanence in the cosmic process, as well as naturalistic philosophies of man’s place in a world of law and chance. But in the twentieth century, the main influences of science on religion have come less from specific theories -- such as quantum physics, relativity, astronomy, or molecular biology -- than from views of science as a method.
Science seems to yield indubitable knowledge on which all men can agree. Its apparent objectivity contrasts with the subjectivity of religion. According to the popular stereotype, the scientist makes precise observations and then employs logical reasoning; if such a procedure is to be adopted in all fields of enquiry, should not religion be dismissed as prescientific superstition? And does not the scientist assume that nature is a self-contained order in which there is no place for God’s action?
It has been largely through the work of philosophers that thought about the methods of science has affected religious thought in recent decades. Specifically, writings in the philosophy of science have had major repercussions in the philosophy of religion. During the 1930’S and 1940’s, the positivists had taken science as the norm for all meaningful discourse. Religious language was considered neither true nor false but meaningless. The positivists had proclaimed the famous Verification Principle, which states that, apart from tautologies and definitions, statements are meaningful only if they can be verified by sense data. Accepting an oversimplified view of science as the prototype for all genuine knowledge, they dismissed religion as ‘purely emotive’.
During the 1950’s positivism came under increasing attack, but many of its assumptions were perpetuated in the empiricism which came to replace it as the dominant interpretation of science. Among the empiricist claims were the following. (1) Science starts from publicly observable data which can be described in a pure observation-language independent of any theoretical assumptions. (2) Theories can be verified or falsified by comparison with this fixed experimental data. (3) The choice between theories is rational, objective and in accordance with specifiable criteria. Philosophers under the sway of such empiricism continued to say that religion can legitimately make no cognitive claims. We will look particularly at the protracted debate concerning the falsifiability of religious beliefs which has occurred since I955, when Antony Flew issued his challenge to the theist: What would have to occur to constitute a disproof of the existence of God? Flew held that religious statements are not genuine assertions because the observable conditions which would falsify them cannot be specified.
But during the 1960’s, the empiricist assertions listed above were vigorously criticized. It is the thesis of this volume that recent work in the philosophy of science has important implications for the philosophy of religion and for theology. Three new viewpoints concerning science, and their consequences for the critique of religion, are the central themes of the book.
The first theme, the diverse functions of language, reflects a change in outlook among philosophers which was already under way in the 1950’s. It is well enough known that it need only be summarized here. The positivist principle that statements are meaningful only if they can be verified by sense data turned out to be too strict to satisfy even in science. The principle would have excluded scientific theories which can never be conclusively verified or proved to be immune to modification. Weaker versions were attempted, for example: a statement is meaningful only if some possible sense data are relevant to the probability of its truth or falsity. But it was extraordinarily difficult to specify at what point the ‘relevance of data’ was to be considered too indirect to qualify under this more generous charter.
Increasingly, philosophers came to acknowledge that language has many forms serving varied functions; science was no longer taken as the norm for all discourse. Linguistic analysis, the most prominent school of contemporary philosophy, asks how men use different types of language. Each field -- science, art, ethics, religion, and so forth -- has a different task, and its approach must be judged by its usefulness in accomplishing its own particular functions. The value of a statement depends on what one wants to do with it; every type of language has its own logic, appropriate to its specific purposes.
The linguistic analysts have described various functions of religious language. Sometimes it evokes and expresses self-commitment. At other times it recommends a way of life, declares an intention to act in a particular way and endorses a set of moral principles. Or again, it may propose a distinctive self-understanding and engender characteristic attitudes towards human existence. Many philosophers stress these non-cognitive functions; they insist that these tasks are valuable and legitimate but are very different from the tasks of scientific language. This is an attractive solution to issues between science and religion; the two fields cannot possibly conflict if they serve totally different functions. The function of scientific language is the prediction and control of nature; that of religious language is the expression of self-commitment, ethical dedication, and existential life-orientation. But the price of this division of labour is that religion would have to give up any claims to truth, at least with respect to any facts external to ones own commitment. Religious beliefs would be useful fictions which fulfil important functions in human life but are not entitled to make any assertions. Throughout this volume a ‘useful fiction’ is to be regarded not as false (as in the popular usage of ‘fictional’), but as neither true nor false.
The diversity of functions of religious language has also been presented in the writings of anthropologists about myths. Myths are stories which are taken to manifest some aspect of the cosmic order. They provide a community with ways of structuring experience in the present. They inform man about his self-identity and the framework of significance in which he participates. Archetypal events in primordial or historical time offer patterns for human actions today. Myths are re-enacted in rituals which integrate the community around common memories and common goals. According to many interpreters, myths are neither true nor false; they are useful fictions which fulfil these important social functions.
However, I would want to join those philosophers who also defend cognitive functions of religious language. For religion does claim to be in some sense true as well as useful. Beliefs about the nature of reality are presupposed in all the other varied uses of religious language. We can at least say that religion specifies a perspective on the world and an interpretation of history and human experience. It directs attention to particular patterns in events. It makes assertions about what is the case.
I will thus be mentioning both similarities and differences between science and religion. Existentialism and positivism, while disagreeing violently in their estimation of subjectivity, agreed completely in portraying a sharp contrast between the objectivity of science and the subjectivity of religion. I will try to show that science is not as objective, nor religion as subjective, as these two opposing schools of thought both assumed. Despite the presence of distinctive functions and attitudes in religion which have no parallels in science, there are also functions and attitudes in common, wherein I see differences of degree rather than an absolute dichotomy. Some of these comparisons are spelled out in the discussion of models and paradigms.
The second theme of the book is the role of models. In the last decade there has been considerable interest in model-building within many intellectual disciplines. Broadly speaking, a model is a symbolic representation of selected aspects of the behaviour of a complex system for particular purposes. It is an imaginative tool for ordering experience, rather than a description of the world. There are, of course, some objects of which actual physical replicas can be built -- such as a ‘scale model’ of a ship or a ‘working model’ of a locomotive. We will be concerned, however, with mental models of systems which for various reasons cannot be represented by replicas, such as the economy of a nation, the electrons in an atom or the biblical God.
There are many types of models serving a diversity of functions. In the social sciences, models of economic development or of population growth allow quantitative predictions of a few variables to be studied under a set of simplifying assumptions. With computer models one can carry out calculations concerning the complex interaction of many variables, among which specified relationships are assumed. The simulation of the behaviour of military, industrial and urban systems is carried out in the new fields of ‘operations research’ and ‘systems analysis’. Models of the political behaviour of an electorate are used to project election returns. Engineering models are used to solve practical problems when it is difficult to experiment on the original system.
I will deal especially with theoretical models in science, which are mental constructs devised to account for observed phenomena in the natural world. They originate in a combination of analogy to the familiar and creative imagination in the invention of the new. I will argue that theoretical models, such as the ‘billard ball model’ of a gas, are not merely convenient calculating devices or temporary psychological aids in the formulation of theories; they have an important continuing role in suggesting both modifications in existing theories and the discovery of new phenomena. I will try to show that such models are taken seriously but not literally. They are neither literal pictures of reality nor ‘useful fictions’, but partial and provisional ways of imagining what is not observable; they are symbolic representation of aspects of the world which are not directly accessible to us.
Models in religion are also analogical. They are organizing images used to order and interpret patterns of experience in human life. Like scientific models, they are neither literal pictures of reality nor useful fictions. One of the main functions of religious models is the interpretation of distinctive types of experience: awe and reverence, moral obligation, reorientation and reconciliation, interpersonal relationships, key historical events, and order and creativity in the world. I will delineate some parallels between the use of scientific models in the interpretation of observations and the use of religious models in the int&pretation of experience. Ultimate models -- whether of a personal God or an impersonal cosmic process -- direct attention to particular patterns in events and restructure the way one sees the world.
Other functions of religious models have no parallel in science. Models in religion express and evoke distinctive attitudes. They encourage allegiance to a way of life and adherence to policies of action; their vivid imagery elicits self-commitment and ethical dedication. Religion demands existential involvement of the whole person; it asks about the ultimate objects of man’s trust and loyalty. Its language expresses gratitude, dependence and worship. This self-involving and evaluational character of religion contrasts with the more detached and neutral character of science.
A separate chapter is devoted to the question of ‘complementary models’. The term originates in modern physics, where both wave and particle models are used for electrons, photons, and other inhabitants of the atomic world. No single model is adequate for the interpretation of experiments in micro-physics, though the probability of the occurrence of particular observations can be predicted from a unified mathematical formalism. I will argue that there is some parallel in the complementarity among diverse models within religious language. I do not believe, however, that the term should be extended to call science and religion ‘complementary’, since they are not talking about the same phenomena and their models are of differing logical types serving differing functions.
I will suggest that the recognition that models are not pictures of reality can contribute to tolerance between religious communities. In a day when the religions of the world confront each other, the view proposed here might engender humility and tentativeness in the claims made on behalf of any one model. In place of the absolutism of exclusive claims of finality, an ecumenical spirit would acknowledge a plurality of significant religious models without lapsing into a complete relativism which would undercut all concern for truth. Analysis of models provides a path between literalism and fictionalism in religion also.
Both the cognitive claims of religion and its living practice must be grounded in experience. If inherited religious symbols are for many people today almost totally detached from human experience, a return to the experiential basis of religion is important for its renewed vitality in practice, as well as for a sound epistemology in theory. Implicit in this position, of course, is a rejection of the positivists’ restriction of attention to sense-experience; all symbol-systems are selective, ordering those aspects of experience which men consider most significant.
The third theme of this volume is the role of paradigms. The term has received wide currency through Thomas Kuhn’s influential book, The Structure of Scientific Revolutions (1962). Kuhn maintained that the thought and activity of a given scientific community are dominated by its paradigms, which he described as ‘standard examples of scientific work that embody a set of conceptual, methodological and metaphysical assumptions’. Newton’s work in mechanics, for instance, was the central paradigm of the community of physicists for two centuries. In the second edition (1970) of Kuhn’s book and in subsequent essays, he distinguished several features which he had previously lumped together: a research tradition, the key historical examples (‘exemplars’) through which the tradition is transmitted, and the set of metaphysical assumptions implicit in its fundamental conceptual categories. Adopting these distinctions, I will use the term paradigm to refer to a tradition transmitted through historical exemplars. The concept of paradigm is thus defined sociologically and historically, and its implications for epistemology (the structure and character of knowledge) must be explored. Let me summarize three issues in this discussion and then indicate their implications for religion:
1. The influence of theory on observation. The empiricists of the 1950’s had claimed that science starts from publicly observable data which can be described in a pure observation-language independent of any theoretical assumptions. By the early 1960’s this claim had been challenged by a number of authors who tried to show that there is no neutral observation-language; both the procedures for making observations, and the language in which data are reported, were shown to be ‘theory-laden’. Kuhn’s volume gave historical illustrations of the paradigm-dependence of observations. He concluded that rival paradigms are ‘incommensurable’. I will maintain that even though data are indeed theory-laden, it is possible to make pragmatic distinctions between more theoretical and more observational terms in any particular context. Rival theories are not incommensurable if their protagonists can find an overlapping core of observation-statements on which they can concur.
2. The falsifiability of theories. The empiricists had claimed that even though a theory cannot be verified by its agreement with data, it can be falsified by disagreement with data. But critics showed that discordant data alone have seldom been taken to falisfy an accepted theory in the absence of an alternative theory; instead, auxiliary assumptions have been modified, or the discrepancies have been set aside as anomalies. I will suggest that comprehensive theories are indeed resistant to falsification, but that observation does exert some control over theory; an accumulation of anomalies cannot be ignored indefinitely. A paradigm tradition, then, is not simply falsified by discordant data, but is replaced by a promising alternative. Commitment to a tradition and tenacity in exploring its potentialities are scientifically fruitful; but the eventual decision to abandon it is not arbitrary or irrational.
3. The choice between rival paradigms. The empiricists had portrayed all scientific choices as rational, objective and in accordance with specifiable criteria. Kuhn replied that criteria for judging theories are themselves paradigm-dependent. He described the change of paradigms during a ‘scientific revolution’ as a matter not of logical argument but of persuasion and ‘conversion’. I will argue that there are criteria of assessment independent of particular paradigms. But in the early stages, when a new contender first challenges an accepted paradigm, the criteria do not yield an unambiguous verdict; the experimental evidence and the relative weights assigned to diverse criteria are debatable and subject to individual judgment. Yet because there are accepted criteria common to all scientists, the decision can be discussed, reasons can be set forth, and an eventual consensus can be expected.
Corresponding to these three issues arising from the discussion of paradigms in science are three similar issues in religion:
1. The influence of interpretation on experience in religion is more problematical than the influence of theory on observation in science. There is no uninterpreted experience; but descriptions of religious experience can be given which are relatively free from doctrinal interpretation. To be sure, any set of basic beliefs tends to produce experiences which can be cited in support of those beliefs, and agreement on the data of religion seems to be exceedingly difficult to achieve. Yet because members of different religious traditions can appeal to areas of shared experience, communication is possible.
2. Flew’s demand that the theist should specify falsifying conditions for religious beliefs seems unreasonable if such falsifying conditions cannot even be specified for comprehensive scientific theories. I will submit that though no decisive falsification is possible, the cumulative weight of evidence does count for or against religious beliefs, but with greater ambiguity than in science. Religious paradigms, like scientific ones, are not falsified by discordant data but replaced by promising alternatives. Commitment to a paradigm (understood, again, as a tradition transmitted through historical exemplars) allows its potentialities to be systematically explored.
3. There are no rules for choice between religious paradigms, but there are criteria of assessment. The application of such criteria is even more subject to individual judgment in religion than in the controversies between competing paradigms during a ‘scientific revolution’. Moreover religious faith includes personal trust and loyalty; it is more totally self-involving than commitment to a scientific paradigm. Nevertheless the existence of criteria means that religious traditions can be analysed and discussed. Religious commitment is not incompatible with critical reflection. It is my hope that the new views of science described here can offer some encouragement to such a combination of commitment and enquiry in religion.
These three themes -- the diverse functions of language, the role of models and the role of paradigms -- combine to support the position of critical realism which I will defend in both science and religion. Such a position recognizes the distinctive non-cognitive functions of religious language, but it also upholds its cognitive functions. Critical realism avoids naive realism, on the one hand, and instrumentalism, which abandons all concern for truth, on the other. Naive realism is untenable if models are not literal pictures of reality and if the history of science is characterized by major paradigm shifts rather than by simple cumulation or convergence. But the inadequacies of naive realism need not lead us to a fictionalist account of models, or to a total relativism concerning truth, if there are indeed data and criteria of judgment which are not totally paradigm-dependent. In the concluding chapter I will suggest some implications of critical realism for the academic study of religion and for the encounter of world religions, as well as for personal religious faith.