The Realism-Antirealism Debate

18

The Realism-Antirealism Debate

If one wants a slogan: realism is the truth and temperate rationalism the way.

W. H. Newton-Smith, The Rationality of Science

There is, I think, no theory-independent way to reconstruct phrases like “really there”; the notion of a match between the ontology of a theory and its “real” counterpart in nature now seems to me illusive in principle. Besides, as a historian, I am impressed with the implausibility of the view.

Thomas Kuhn, The Structure of Scientific Revolutions

1—Introduction

Our scientific discourse, as well as our everyday language, is permeated with terms for theoretical entities in science: quarks, quantum vacuums, electrons, DNA molecules, continental plates, and a host of others. Moreover, most people today are inclined to think that our current theories about these entities—for example, our current views about DNA molecules—are reasonably accurate, approximately true depictions of the real, theory-independent world. However, in spite of these widely held opinions, it is still legitimate, and it is part of the business of philosophy, to question these views by raising this question: Given that a scientific theory is “successful” in some sense of that term (e.g., it explains facts, makes good predictions), are we then entitled to believe in the reality of the theoretical entities postulated by that theory and in the approximate truth of our descriptions of those entities?

Ever since the ancient Greeks began to investigate nature, there has been a debate about this and related questions. What is the purpose of scientific theories? What does it mean to say a scientific theory is successful? Are successful scientific theories merely useful fictions that explain empirical observations and generate accurate predictions, or do their theoretical terms actually refer to real entities, and does a successful theory provide fairly accurate descriptions of those entities? This debate is called the realist-antirealist debate, and it is still being hotly contested among scientists and philosophers. Very roughly, scientific realism is the view that science progressively secures true, or approximately true, theories about the real, theory-independent world “out there” and does so in a rationally justifiable way. In contrast, antirealism, which comes in a number of different forms, denies realist interpretations of science in favor of alternative interpretations. Thus it is important to keep in mind that a defense of the existence of the theoretical entities of science is not ultimately a question of whether some scientific theory is “successful.” More fundamentally, it is a question about which kind of philosophy of science—scientific realism or some form of antirealism—should be preferred as a characterization of the success of scientific theories.

Why should Christians be concerned with this issue? Apart from the intellectual value of the debate, considered in and of itself, another aspect of this discussion is of interest to Christians: one’s views of the realism-antirealism controversy should be factored into one’s understanding of the integration of science and theology. If scientific realism is accepted as the correct view of a scientific theory—for example, the idea that by employing the notion of imaginary time as something real, one can avoid postulating that the universe had a beginning—then if that theory seems to run counter to some theological affirmation, say, that the universe had a beginning, then Christians will either have to refute that scientific theory, adjust their understanding of the theological affirmation, or adopt a different strategy. Thus much depends on what it means for a theory to be well established or successful. This, in turn, is related to the debate about realism and antirealism.

However, if antirealism is adopted for and limited to scientific theories, then one would not take a well-established scientific theory to be true or approximately true—perhaps the theory is just a useful fiction—and there will be no pressure to adjust the truth of the theological affirmation. For example, if a theologian believes that all physical events have causes, and if quantum physics seems to deny this, then if quantum theory is taken in an antirealist way, there would be no need to adjust one’s view of causation. On the other hand, there may be dangers in adopting antirealism for scientific theories because, once this move is made, it may be difficult to limit antirealism to scientific theories alone. For example, if one’s antirealism affects one’s theological assertions, then claims about God, life after death, and so forth could be interpreted in ways that cash out those claims without taking them to be referring to actual entities (God, the afterlife) in the real world. Figure 18.1 depicts some of the major options in the realist-antirealist dialogue.

Figure 18.1. Options in the realist-antirealist dialogue

In what follows, we will state and evaluate scientific realism, look at some of its anti­realist rivals, and offer some final thoughts about this debate relevant to the task of integrating science and theology.

2—Scientific Realism

2.1 Overview of Scientific Realism

Scientific realism was a minority view in the first half of the twentieth century, at least among the more vocal philosophers, but is now the majority position among current philosophers of science. Prominent scientific realists have been Ernan McMullin, Richard Boyd, W. H. Newton-Smith, Karl Popper, and Rom Harré. There are several different varieties of scientific realism, but its core tenets are these:

SR1: Scientific theories (in mature, developed sciences) are true or approximately true.

SR2: The central observational and theoretical terms of a mature scientific theory genuinely refer to entities in the world. These terms make existence claims.

SR3: Given two rival theories, it is in principle possible to have good reasons for thinking which is more likely to be true or approximately true. Rationality is an objective notion and conceptual relativism (what is rational for one person or group should not necessarily be so for another person or group since rationality itself is relative to a person, scientific community, or theory) is false.

SR4: A scientific theory will embody certain epistemic virtues (simplicity, clarity, an absence of internal and external conceptual problems, predictive success, empirical accuracy, scope of relevance, fruitfulness in guiding new research, utilization of appropriate ways of explaining things—e.g., prefer efficient to final causes) if and only if it is true or approximately true.

SR5: The aim of science is a literally true conception of the theory-independent external world. Scientific progress is a fact, and science tends to converge on truer and truer conceptions of the world, with later theories usually refining and preserving the best parts of earlier theories and coming closer to the truth than earlier theories in the same domain of discourse (i.e., where the earlier and later theories are talking about the same things and are thus comparable to each other).

2.2 Explanation of Scientific Realism

SR1 states that science involves some form of the correspondence theory of truth (see chap. 6): a theory is true if and only if what it says about the world is in fact the way the world is. The theory-independent world is what makes a theory true. SR1 also assumes that the notion of approximate truth is a coherent one. Some theories can better approximate the truth than others. Finally, SR1 contains the notion of a “mature, developed science,” which, in turn, has two different aspects to it. For one thing, an immature area of science can become mature if it has time to develop, test, and “vindicate” one of its major theories. The idea here is that a mature, developed science is one that has been around long enough to have proven itself.

There is another aspect to the notion of a mature, developed science: some sciences are considered mature in the sense of being the ideal, archetypal sciences. In this sense, physics and, perhaps, chemistry are considered mature in that they are the ideal sciences. Other areas of science are mature to the degree that they are reducible to or approximate the methods, theoretical concepts, and explanatory devices used in physics and chemistry. There is an implicit reductionism in this notion of a mature science: there is a unity to all sciences and, in some sense, other sciences are reducible to or replaceable by physics and chemistry. For example, some reductionists believe that the laws and theories of biology will someday be reduced to the laws and theories of chemistry and physics.

SR1 expresses what is sometimes called alethic realism or truth realism, according to which the theoretical entities postulated in mature, developed sciences (e.g., electrons) are real, and our descriptions of them (e.g., electrons have negative charge and a rest mass of such and such) are at least approximately true. By contrast, entity realism holds that ordinary observable entities (e.g., Mars), the microentities that could be observed if we were small enough (e.g., electrons), and entities that are in principle unobservable (e.g., quarks, magnetic fields) are real, but our descriptions of these entities are neither true nor approximately true. Entity realism is currently a minority position among realists, and it will be set aside for the purposes of this and the next chapter.

SR2 is a semantic thesis, that is, a thesis about the nature of language, about how terms in language succeed in referring to things in the world, and about the nature of linguistic meaning. Scientific theories include observational terms and theoretical terms. There is a debate about whether the distinction between these two types of terms is absolute (i.e., a term functions one way or the other, but not both ways at the same time in the same theory) or whether the distinction is best viewed as a continuum with observational terms being colored by theory and theoretical terms always having at least some observational content. Either way, the distinction is an important one for understanding scientific realism. Table 18.1 gives some examples of the two types of terms.

Table 18.1. Theoretical versus observational terms

Theoretical Term

Observational Term

electron mass

mass

red

floats

electric field

kinetic energy

longer than

wood

atom

temperature

left of

water

molecule

gene

hard

weight

virus

charge

volume

iron

The important thing about SR2 is the claim that theoretical terms are referring terms. To understand what this means, consider the following two sentences:

In sentence (1), the term Fido is a referring term. It refers to the dog Fido, an “extralinguistic entity” in the world, and says of him that he has the property of brownness. If sentence (1) is true, then because Fido is a referring term, sentence (1) has implications for reality: the entity referred to by the term Fido actually exists.

Now consider sentence (2). It contains a definite description, a phrase that begins with the and a description of the item that satisfies that description and, therefore, to which the description refers. (2) is grammatically similar to (1) in that the definite description “the average family” appears to function as a referring term like the proper name Fido. Both terms occupy the subject position in what appear to be subject-predicate sentences. However, on closer inspection, we realize that “the average family” is not a referring term at all. No one would try to locate the average family and count its children to see if, in fact, it has 2.5 children. Rather, “the average family” is a shorthand term for a set of mathematical operations. This can be made clear by replacing (2) with (2’ ):

2’. Add the number of children and divide by the number of families and you get 2.5.

(2’ ) may have some existence implications (there are families and children) but is basically a recipe for a mathematical operation, and “the average family” does not designate an entity in a straightforward manner as does Fido in sentence (1).

Now consider this sentence:

3. Protons have positive charge.

SR2 says that the theoretical term proton is a referring term, that is, it functions like Fido and not like “the average family.” In general, the theoretical terms in our well-established scientific theories actually refer to entities, most of them unobservable in practice or in principle (e.g., magnetic fields), but that genuinely exist.

What does it mean to say that an observational or theoretical term is central to a theory? Suppose we were trying to identify some substance, X, that we think is copper. Given our current theories about metals in general, and copper in particular, a number of theoretical properties or observational consequences are relevant for classifying X as copper: being reddish, having the melting point of 1083º C, and having the atomic number 29. Some of these are more relevant and, therefore, more conclusive than others. With respect to two theoretical or observational properties (and the terms that refer to them), one is more central to classifying X than another if and only if the possession (or lack) of that property tends to count more in favor of (or against) classifying X as copper than does the possession (or lack) of the other property.

With regard to copper, having the atomic number 29 is more important and central than the melting point, which in turn is more important than being reddish. The more central a property is, the more weight it has in settling a dispute about classification. The intuitive idea here is that scientific theories have several theoretical and observational terms in them, some of them are more important than others, and those judged more important are the ones that are more central for the purposes of classification. The more central a term is to a well-established theory, the more a scientific realist is committed to the existence claims that the term implies.

SR3 asserts that science is an objectively rational discipline, not in the sense that scientists have no biases, but in the sense that one can, at least in principle, have objective reasons or warrant for accepting or rejecting a given scientific theory. Rationality is a normative, objective notion, not a relative one. Moreover, when two rival theories are being judged, there are, in principle, considerations that can be used to decide in favor of one rival and against the other. Rival theories are commensurable: they can be compared vis-à-vis each other against some common ground—the data, the various epistemic virtues each embodies, and so forth.

SR4 claims that if a scientific theory has certain epistemic virtues, then it is objectively rational to believe that it is true or approximately true because it has these virtues. If T₁ is simpler or explains more data than T₂, gives more accurate predictions, and so on, then one has more reason to believe that T₁ is true or more approximately true than T₂. Furthermore, if a theory is true and its rivals are false, or if a theory is truer than its rivals, then one should expect it to be better eventually at predicting data, containing more clarified terms, or embodying other epistemic virtues. SR4 implies that rationality is connected to truth. As a theory grows in its rationality (as measured by the presence of epistemic virtues), our confidence in the truth or approximate truth of the theory grows as well.

Finally, SR5 is a thesis about the history and aims of science. It states that the aim of science is to give us not just theories that work (e.g., that help us control nature, predict phenomena, harmonize with our observations), but also theories that are true. Science tries to tell us the way the world really is, especially the way the unobserved theoretical entities and processes of the world really are.

Moreover, more recent scientific theories are more accurate pictures of the world than are their predecessors. Science progresses over time toward a true picture of the world. Scientific progress is to be understood as progress toward truth. There are two kinds of theory changes in science: replacement and refinement. In replacement, a latter theory replaces a former one and the earlier theory is abandoned altogether. On a realist view, the abandoned theory is judged to be false. The change from phlogiston to oxygen chemistry is an example. In refinement, when a latter theory refines a former one, the supposed truth-preserving components of the earlier theory are preserved in the later theory and made more precise and accurate. According to some, the change from Newtonian to Einsteinian views of space, time, and mass is an example of theory refinement. SR5 implies that the history of science, while surely containing examples of theory replacement, includes as well a number of cases of theory refinement. Either way, in theory replacement or refinement, newer theories get closer to the truth than older ones.

2.3 Evaluation of Scientific Realism

2.3.1 Positive Support

This, then, is a statement of scientific realism as a philosophy of science. How should we evaluate this position? There have been at least four basic kinds of arguments offered in favor of scientific realism. Due to space considerations, we can only state them briefly. Here is the first argument: Given that one must either embrace scientific realism or some form of antirealism, and given the fact that there are serious problems with each of the main forms of antirealism, then scientific realism, on balance, is to be preferred. Obviously, the strength of this argument turns on the case for the inadequacies of the various forms of antirealism. Later on in this chapter we will present the main antirealist position. At that time, you can get a feel for the various views. Hopefully, this will help you to begin to formulate your own evaluation of their strengths and weaknesses.

Second, a scientific realist can offer arguments in favor of SR1-SR5. For example, the scientific realist could offer arguments in favor of the correspondence theory of truth, in favor of the idea that rationality is intimately connected to truth, and against various forms of conceptual relativism. Similarly, the realist could also try to justify a reading of the history of science as a history of clear progress through the refinement and replacement of theories toward a better and better depiction of the way the world is. The first two issues (the correspondence theory of truth and the connection between rationality and truth) have already been discussed in chapters six and three, respectively. The third topic (conceptual relativism) was the subject of much of chapters four and six. The fourth factor would involve a detailed investigation of the history of science, among other things.

We cannot go into these debates here; however, one important point should not be missed. All of these topics are deeply philosophical in nature. Thus the defense or rebuttal of scientific realism illustrates the fact that the philosophy of science is presuppositional to science itself. That is, the question of how we should understand the existence claims of a given scientific theory will be answered, in part, by one’s attitude toward scientific realism, and one’s attitude about this will, in turn, be justified largely in philosophical terms. Thus the realism-antirealism debate serves as another illustration of the truth of an external philosophy of science that was discussed at the beginning of chapter seventeen.

The third argument for scientific realism is, perhaps, the main one. According to this argument, scientific realism is the best explanation for the fact that (1) our theories actually work (i.e., embody various epistemic virtues); (2) science makes progress in solving its problems; (3) often, a scientific theory will have a host of independent, empirical confirmations for it that converge together to support the theory, even if some of those empirical confirmations were not originally conceived as part of the domain for which the theory was thought to be responsible. Scientific realists claim that the best explanation for these three facts is that our theories succeed in laying hold of reality and giving at least approximately true descriptions of what really exists. For the scientific realist, it is because our theories capture the way the world is that those theories embody epistemic virtues, allow us to solve problems, obtain empirical confirmation, and can be extended into new, previously unthought of domains of investigation. If we abandon scientific realism, say its advocates, these facts about science can only be regarded as fortuitous miracles.

This argument places a burden of proof on antirealists; that is, realism should be assumed to be correct unless antirealists can show that their views are to be preferred to scientific realism. Among other things, this means that antirealists are required either to offer alternative accounts for these successful aspects of science or to deny that these successes really exist. As we will see later, Larry Laudan and Bas C. van Fraassen take the former approach and Thomas Kuhn takes the latter.

One final argument has been offered in favor of scientific realism, and it is theological in nature. According to some Christian philosophers of science, God has created us to know his creation and think his thoughts after him, given us cognitive and sensory faculties suited to be obtainers of truth for the world in which he placed us, and given us the desire to know his world. And these theological facts lend support to scientific realism. Other Christian philosophers of science do not agree. According to them, these points about God, our faculties and desires, and the world only support the existence and availability of knowledge that comes from theology, philosophy, and common sense. But they do not give support to such a technical enterprise as science, which, they argue, is better understood in antirealist ways.

2.3.2 Objections to Scientific Realism

What are some of the main arguments against scientific realism? A number of them have been raised. First, antirealists have criticized the realist notion of truth in two different ways. To begin with, antirealists have criticized the correspondence theory of truth and/or the notion of a theory-independent external world to which we have access. Sometimes this latter point is supported by the claim that all observations are theory laden, there is no direct, theory-independent access to the external world, and thus the real­ist notion of such a world is either unintelligible or irrelevant. These criticisms have been the focus of much of chapters five and six, so we will not rehearse those issues here. But you should bear in mind that the epistemological and metaphysical topics mentioned in those chapters have a direct bearing on what view of science one will take.

A different way that antirealists have criticized realist notions of truth centers on the concept of approximate truth. Few would argue that our current theories are true in any complete or final way. Rather, scientific realists claim that our current well-established theories are approximately true to one degree or another. Another word for “approximate truth” is verisimilitude. Verisimilitude should not be confused with the epistemological notion that a given belief or set of beliefs has a high probability of being true, that is, that our certainty regarding the belief or set of beliefs is probable, but not the belief itself, which is completely true if true at all. Verisimilitude means that a belief or set of beliefs is itself approximately true. Some critics of scientific realism point out that no one has been able to clarify sufficiently what approximate truth means, especially if a correspondence theory of truth is in view. Something is either true or false, and truth does not come in degrees. A false proposition may serve certain purposes better than another false proposition, but it is still false. If a lunch box is on the kitchen floor, then the proposition that it is on the kitchen table may help you find the lunch box more than the proposition that it is in the den, but strictly speaking both propositions are false. Similarly, if we accept Einstein’s laws of motion, space, time, and mass, then Newton’s laws of motion are literally false, even though at low velocities they are useful for a number of purposes (e.g., predicting where a projectile will land, flying an airplane). Thus antirealists claim that the realist notion of verisimilitude is an inadequate one.

A second antirealist criticism claims that theoretical terms in scientific laws or theories can be eliminated without loss of empirical content and we need not take those terms to be genuine referring terms. Remember, realists interpret the role of theoretical terms as referring to theoretical entities that actually exist. But if we can eliminate those terms without losing our ability to predict observations accurately and fit our theories and laws with the empirical world, then we can regard those terms as unnecessary.

An example may help clarify this antirealist argument. Some empirical generalizations are limited in their range of application and have exceptions. Consider the following:

Some antirealists argue that most or perhaps all of the theoretical terms supposedly referring to unseen entities can be similarly eliminated without loss of empirical content or predictive ability for our laws and theories.

Third, antirealists claim that successful theories can turn out to be false such that the truth or approximate truth of a theory is not a necessary condition for the theory’s being successful (its embodying various epistemic virtues), as scientific realists claim in SR4. In the history of science, many theories have explained phenomena, generated fruitful research and accurate predictions, yet were later abandoned as false. For example, Augustin-Jean Fresnel (1788–1827) used optical aether theory to predict a novel test result: a small beam of light directed at a circular disk will undergo diffraction (bending around the edges of the object) in such a way that a bright spot at the center of the shadow of a circular disc should appear. When tested, the prediction proved correct. But we no longer believe in optical aether theory.

Fourth, antirealists claim that throughout the history of science, theories that would later be judged to have been approximately true based on a realist understanding of contemporary science (e.g., early atomic theory like that advocated by William Prout [1785–1850]) have been unsuccessful for long periods of time compared to rival theories that we now consider false. Thus the truth or approximate truth of a theory is not a sufficient condition for its success. Arguments three and four directly attack SR4 by showing that the truth of a scientific theory is neither necessary nor sufficient for the theory to be successful.

A fifth and final objection to scientific realism claims that a proper understanding of the history of science shows that, surface appearances notwithstanding, science does not make progress toward truer and truer theories. Some antirealists like Thomas Kuhn support this contention by arguing that there is no progress whatever in science. We will look at Kuhn later in this chapter. But, briefly stated, his point is that the history of science is really a history of later theories completely replacing earlier theories and not refining them. This is clearly true in some cases, like theories that placed the earth at the center of the universe, but, Kuhn argues, it is even true when our current theories still use the language of earlier theories. For example, the shift from Newton’s to Einstein’s theory about space, time, and mass represents a change in which the new theory still uses many of the same terms as did the older theory (mass, space, time are still used in Einstein’s theory), but the entities represented by these terms are depicted as so radically different in the new theory compared to the old that advocates of the two theories use the terms equivocally (both use the same terms but with very different meanings) and we no longer believe that Newtonian mass, space, and time exist. Thus the history of science is a history of discarded theories, suggesting that our current theories will some day be discarded as well. Thus there is no reason to think that we now have or will have true theories.

Other antirealists, like Larry Laudan (also discussed later in this chapter), believe that science does, in fact, make progress through history, but that progress has nothing whatever to do with the notion that later theories are more true than earlier ones. Rather, our current theories represent progress compared to earlier ones in that our current theories are more pragmatically useful—they make better predictions, they help us explain more things, they allow us to generate better technology. But pragmatic usefulness should not be confused with truth, say some anti­realists, and scientific progress, though real, is a function of the pragmatic usefulness of theories and not of their truthfulness.

Other objections have been raised against scientific realism, but we have described the main ones. It is time now to take a look at antirealist alternatives to scientific realism. In what follows we will look at four rational nonrealist (i.e., instrumentalist) views in the order presented in figure 18.1 and then look at one nonrational nonrealist position.

3—Antirealism

3.1 Rational Nonrealism (Instrumentalism)

Rational nonrealism, also called instrumentalism, is a family of views that agree on two main points: (1) Science is an objectively rational enterprise, and conceptual relativism is false as a way of understanding the ration­ality of science. (2) Scientific theories do not give us true or approximately true pictures of the unobservable, theoretical entities and processes that are causally responsible for what we empirically observe, nor do the theoretical terms in those theories refer to actual entities in the theory-independent world (and even if, by accident, they do refer to existent entities, this would be utterly irrelevant for science). Rather, scientific theories are useful tools or instruments (thus the name instrumentalism) that help us accomplish certain things. But what, then, are the things that scientific theories help us to accomplish? Here, instrumentalists differ, and at least four answers have been given to this question.

1. Phenomenalism. Among other things, phenomenalism is an epistemological doctrine that includes a view about the nature of perception. It was discussed in chapter eight, and thus it will only be briefly treated here with special focus on the way phenomenalism relates to the nature of science. Phenomenalism is a view that was more popular in the previous century. Major proponents of one form or another of phenomenalism have been Benjamin Brodie, Ernst Mach, and A. S. Eddington. Essentially, phenomenalism is a radical empiricist theory of epistemology to the effect that all our knowledge is derived from and is about immediate sensory experiences. Applied to science, the view implies that scientific knowledge is about what can be directly observed. Any thing or process that cannot be perceived cannot be supposed to exist for science.

Second, the meaning of expressions that appear to be referring to unseen theoretical entities (e.g., electron, atom) must be taken as referring to sensory experience. Since only propositions about what is directly observable are meaningful or have the status of knowledge, claims about unobservable theoretical entities are really disguised statements about sensory experiences and must be translated or reduced to such. Science merely tries to save the phenomena, that is, to facilitate our ability to describe the successive sensory experiences we have of the world or to anticipate and predict future sensory experiences. As Mach put it, “It is the object of science to replace, or save, experiences, by the reproduction and anticipation of facts in thought.” Theoretical terms are merely summaries of data, labor-saving devices for classifying observations. A theory is a mnemonic device by which sensory facts are stored and recollected; it is an elliptical formula of the relations of dependence that regularly occur between observable events and properties. Thus theoretical terms can be translated without remainder into a set of statements about a potentially infinite set of actual or possible sensory experiences.

Furthermore, phenomenalists hold that experience is nothing more than a succession of immediately given, conscious states known by direct inspection or acquaintance with the flow of one’s own consciousness. These conscious states are variously called ideas, sense data, or impressions.

Some examples may help to clarify what all this means. According to phenomenalism, the word hydrogen refers not to an atom that exists but to a set of laboratory observations of “hydrogenated space” (i.e., a set of laboratory observations of colorlessness, weight, and volume that are experienced after a series of prior sensory experiences occur that involve ways of preparing the “hydrogenated space”). The statement “There is presently an electric current in the wire on the table” does not refer to the theoretical entity, an electric current, but rather the statement can be translated into a set of statements like this: “If the galvanometer on the desk there were introduced into the circuit, the pointer would be deflected from its present position.” This statement can, in turn, be translated into a statement about a set of sensory experiences (“If one orients one’s head in a certain direction at the right time, one will have a pointer type sense datum with a certain shape”). The law 2H₂ + O₂ = 2H₂O merely asserts that, in the past and most likely in the future, two units of a certain observed volume, weight, and odorlessness produced by certain techniques have been added to one unit of volume, weight, and odorlessness produced by certain other techniques, and a wet, colorless substance of two units has always followed. Note that the essential features of the law are all capable of being analyzed in terms of sensory experiences (e.g., observed volume, weight, and odorlessness).

We cannot undertake a criticism of phenomenalism here, but one point should be made. There has always been a tension between the empirical aspects of science and the belief that science gives us the truth about the theoretical, unobservable entities and processes in the world. To the degree that one emphasizes the empirical aspects of science, it is difficult to sustain scientific realism. On the other hand, to the degree that one embraces scientific realism, one must allow that science goes beyond and involves itself in factors other than those that can be directly tested through immediate observation.

2. Operationalism. Operationalism, occasionally called operationism, is an approach to science very similar to phenomenalism. Its major proponent has been P. W. Bridgman (1882–1962). Whereas phenomenalism links scientific terms, laws, and theories to actual or possible sensory experiences, operationalism links them to actual or possible laboratory operations. For the phenomenalist, scientific laws and theoretical terms really refer not to mind-independent entities and events but to mind-dependent sensations. For the operationalist, scientific laws and theoretical terms really refer to experimental activities and operations. Theoretical terms are like the term “the average family” that we discussed in conjunction with SR2 above.

Bridgman stated the central tenet of operationalism in this way: “In general, we mean by any concept nothing more than a set of operations; the concept is synonymous with the corresponding set of operations. . . . The proper definition of a concept is not in terms of its properties but in terms of actual operations.” For example, length is not an attribute that a physical body has; rather, it is defined as a set of operations of sliding rulers, marking coincidences, and counting how many operations one has made. The “length” of a room is nothing but the operation of taking a stick with thirty-six markings (call them inch markings), starting at one end of the room, and laying the stick end after end until one crosses the room, and recording all one’s activities. Similarly, the statement that x amount of some acid neutralizes y of some base means that if one pours a certain measurement of one liquid (which, in turn, is defined in terms of certain operations used to obtain it) into a beaker and adds a certain measurement of another liquid, then certain measurements follow (e.g., the combined liquids change color when the measurements are done in a certain way).

Before we move on to the next instrumentalist view, a final observation about operationalism should be made. Operationalism is reductionistic in spirit and thus is an inadequate theory. Length is a property of bodies that may be measured by a set of measuring operations, but which is not identical to or the same thing as those measurements. Thus we should be careful to distinguish operationalism as a philosophy of science from the legitimate use of operational definitions. It is quite proper to come up with some operational definition for something—say, intelligence, depression, or length—in which a set of measurable, empirically testable operations are formulated as a test for the thing being studied. It is one thing to say that measurements x, y, and z are good tests for the presence of some phenomenon P, which is what the legitimate use of operational definitions seeks to do. However, it is another thing altogether to claim, as operationalists do, that phenomenon P is identical to, is nothing other than, measurements x, y, and z. Thus one can use operational definitions without being a full-blown operationalist.

3. Pragmatism. A third instrumentalist view is called pragmatism, and the main contemporary advocate of this view is philosopher Larry Laudan. According to Laudan, any philosophy of science must take into account three features of the history of science: (1) At each period of science, one can find examples of agreement (where scientists agree about the adequacy of some theory) and disagreement (where various theories are competing for allegiance). (2) Science is an objectively rational, progressive discipline, but scientific progress has little or nothing to do with truth. We must therefore understand scientific progress in ways other than the scientific realist proposal expressed in proposition SR5. (3) Science tries to solve empirical and conceptual problems (see chap. 17). A solution to a problem need not be true; it must only remove our puzzlement about some phenomenon and allow us to make useful predictions about the natural world and harmonize our theories with what we observe.

Laudan’s philosophy of science is an attempt to explain these three features of the history of science. For Laudan, then, the main goal of science is to solve problems. The solved problem is the basic unit of scientific progress. For example, if our theory of gravity helps us explain and predict the movement of the tides, then we have used our theory to remove puzzlement about this problem, and our solution is a good one because it works for us, irrespective of whether our solution is true. Thus scientific progress should be defined in terms of the relative rate, number, and importance of solved problems accomplished by a theory or research program, not in terms of convergence toward the truth (a research program is, roughly, a group of theories focusing on the same issues, e.g., all models of atomic theory through history would be part of the same research program). If theory A solves important problems more quickly than theory B, then it is more rational to accept theory A than B. The possession of various epistemic virtues (simplicity, predictive success, etc.) by a theory tells us merely that the theory is successful, that it works well in solving its problems, not that the theory is approximately true.

In sum, Laudan claims that problem solving is the fundamental feature of science, that scientific progress should be defined, not in terms of truth, but in terms of the rate and success of a theory’s problem-solving track record, and that scientific rationality should, in turn, be defined in terms of the progress of a theory—the more progressive a theory is the more rational it is to accept it. According to Laudan, the realist concept of truth is not relevant to any of these notions.

But the scientific realist will want to know how science can make progress in solving its problems if its theories are not at least approximately true. If the truth of our theories is not what is responsible for making them work, what is? Laudan’s answer involves a coherence theory of epistemic justification (see chap. 5). The point of our theories is for them to work, and when a theory has problems we will continue to make adjustments in various ways. We may deny that the problematic data are really accurate (perhaps our instruments are not to be trusted), we may change the aims and methods we appropriate (e.g., start preferring theoretical simplicity to empirical accuracy, try different test methods that harmonize with our theory and abandon test methods that do not), or we may adjust part of our theory itself. The point is, we will find some way to reestablish coherence among our theories, the data, and our aims and methods. But once we have obtained such coherence, that does not mean the theory is approximately true; it just means that scientists have become good at fitting things together until they work. For Laudan, this is all it takes to explain how science makes progress in solving its problems.

4. Constructive empiricism. Princeton philosopher Bas C. van Fraassen holds a view about science that he calls constructive empiricism. According to van Fraassen, scientific realism is inadequate; the point of science is to “save the phenomena,” that is, to develop theories that are consistent with empirical observations and predictions. Science aims at giving us theories that are empirically adequate, and acceptance of a theory only involves the belief that it is empirically adequate.

Van Fraassen calls his philosophy constructive because he sees the aim of science to be the construction of theoretical models, not the discovery of truth concerning unobservables. It is called empiricism because, as already mentioned, the main goal of science is to develop empirically adequate, predictively successful theories. Van Fraassen makes a distinction between believing and accepting a theory. Accepting a theory merely involves belief that the theory is empirically adequate, that it is true regarding what it says about what we can observe. It involves no commitment to the truth of the theoretical, unobservable entities and processes included in the theory, but merely the pragmatic commitment to act as if such entities and processes are real if such commitment leads us to better and better empirical success. Believing a theory, on the other hand, involves a commitment to the belief that a good theory truly describes both the observational and unobservable world. Scientific realists believe good theories; constructive empiricists merely accept them.

If scientific realism is false, then how is it that theories make progress in “saving the phenomena”? Van Fraassen’s answer is similar to the one offered by Laudan: since the point of a theory is empirical adequacy in the first place, scientists will keep adjusting things until our models obtain such adequacy. Scientific theory competition is like natural selection: only the strong (i.e., the empirically adequate) survive.

This completes our brief overview of four instrumentalist views of science. All of them agree that science is an objectively rational set of disciplines. Nonrational nonrealists reject this notion. They agree with instrumentalists that truth about real, theoretical entities and processes is not the goal of science. But they also reject the notion that there is some objective sense of rationality that science embodies.

3.2 Nonrational Nonrealism

Thomas Kuhn is the best-known advocate of nonrational nonrealism, and we will look at his philosophy of science as a representative of the view. Kuhn’s view of science has two key aspects: the history of science and the epistemology of science. For Kuhn, the epistemic authority of science does not reside in the fact that scientists use something called the scientific method (he denies that there is such a thing) but rather in the scientific community itself, seen as a group of practitioners of a craft working within a shared paradigm.

The notion of a paradigm is central for understanding Kuhn’s history and epistemology of science. What is a paradigm? Some have claimed that Kuhn has used as many as twenty-two different definitions of the term, but here are some of the main ones: the paradigm is an entire worldview, a specific scientific theory (like Newton’s views of space, time, motion, and mass), a pedagogical tool scientists use to indoctrinate new practitioners into a field (like the ideal gas equation PV = nRT, which teaches students how to view gases), a way of seeing the world, and something that constitutes nature itself by telling us what the world in our theory amounts to (there are atomic corpuscles in some paradigms, waves in another, both in still different paradigms).

Regarding the history of science, there are periods in which science can be conceived of as normal science. These are times when a field is dominated by one, universally accepted paradigm. During this time, the paradigm is not questioned and scientific activity focuses on extending the range of application for the paradigm, making it more precise and so forth. However, if the paradigm develops a number of anomalies that it cannot adequately handle, then a period of crisis emerges when rival paradigms are formed. Eventually, a scientific revolution occurs by which the old paradigm is completely abandoned and a new paradigm is accepted, ushering in a new period of normal science. Accordingly, Kuhn rejects any notion of progress throughout the history of science or of theory refinement as a type of theory change. The history of science is a history of theory change and replacement through cycles of normal science, crisis, revolution, and normal science.

Regarding the epistemology of science, Kuhn holds that there are no such things as neutral facts or data. Observation is theory laden and there are no data or epistemic virtues that could be used to justify a paradigm that are not themselves already dependent on that paradigm. Thus the notions of truth as correspondence and of an external, theory-independent world are bogus and irrelevant for science. Practitioners of two “rival” paradigms literally see a different world. Rival paradigms are incommensurable; that is, they cannot even be compared with each other to see which is more rational or closer to the facts because there is no common ground between them and nothing outside the paradigm that could serve, even in principle, as a basis of such a comparison. Different paradigms literally describe different worlds.

This means that the shift from one paradigm to another is like a conversion experience. It represents a different way of seeing things. Such paradigm shifts are not matters of gaining a truer or rational theory in favor of a discarded one. The rational-irrational distinction is itself relative to different paradigms. What is rational for one community of scientific practitioners defined by one paradigm will not necessarily be rational for another. Thus science is not objectively rational because such a notion is not applicable to science. The rational authority of science is primarily a cultural matter and relies on various types of persuasion; it is not a matter of objective rationality. Our current theories are not more rational than those of five hundred years ago, they are just more rational to those of us who accept them. It should be obvious that Kuhn’s view of science has a postmodernist flavor to it in its rejection of truth as correspondence, of the notion of an objective theory-independent world, and of the idea of objective rationality. It should also be obvious that it suffers from some of the problems that accrue to postmodernism in general (see chap. 6).

4—Integration and the Realist-Antirealist Debate

Christians have the desire, the responsibility, and the privilege of integrating their theological beliefs with justified beliefs from sources outside theology. We will take up the topic of the integration of science and theology in the next chapter, but for now, let us ask how the realist-antirealist debate informs the task of integration. Should Christians be realists or antirealists in light of their joint commitment to truth wherever it is found and to Christian teaching? In the introduction to this chapter we touched on the importance of this question. It is important for its own sake, since as Christians and simply as responsible human beings we want to know the truth about things, the realist-antirealist debate included.

This question is also important for the task of integration. For example, if some well-established scientific theory seems to be contrary to some well-established part of our Christian worldview, then one way to resolve this dilemma would be to take the scientific theory in question in antirealist terms. We could then accept that the theory is successful, but not take that to mean that the theory is true. Thus it would not constitute a direct challenge to the truth of the theological belief in question. Of course, such a strategy could easily be abused or charged with being a convenient way out of intellectual difficulties, so anyone who adopts such a strategy would need to use it with integrity. Moreover, such a recourse assumes that antirealism is an option, at least some of the time.

Should Christians adopt realism or antirealism as a philosophy of science? This is no easy question, and responsible Christian philosophers are divided on the issue, as we indicated earlier in this chapter. It would seem that, insofar as Christian doctrine is concerned, each position is a live and legitimate option. However, we need not leave the matter here because we think that there are some arguments against scientific realism that are not live options for Christians, while others are live options for Christians.

In our view, common sense, solid philosophical argument, and Christian theology unite to justify belief in three things relevant to the realist-antirealist dialogue: (1) The notion of truth as correspondence is at least part of an adequate philosophical account of truth and is implicitly taught in and presupposed by the Bible, as is the notion of a theory- or (finite) mind-independent external world. (2) Our natural sensory and cognitive faculties, though fallen, are still adequate to give us justified true beliefs about the world, which is itself the sort of world suited for being known by creatures with the types of faculties we possess. Moreover, any attempt to limit human knowledge within the confines of strong forms of empiricism (as in phenomenalism or operationalism) is mistaken. (3) Rationality is intimately connected with truth in such a way that the main value or point of behaving rationally and of obtaining rational beliefs is that such behavior and beliefs put us in a better position to live and to have the truth. Thus antirealist arguments that run contrary to these three notions are to be rejected on rational and theological grounds.

On the other hand, three antirealist arguments do seem to carry weight: (1) The history of science clearly offers a number of cases where past theories now judged to have been false in light of realist interpretations of current science embodied several epistemic virtues, and theories judged as approximately true in light of realist interpretations of current theories failed to embody various epistemic virtues for some time compared to their rival theories. (2) Due to our finite and fallen condition, various psychological and sociological nonrational factors, while not making knowledge impossible, do, nevertheless, affect the processes by which we form and justify our beliefs. This is especially true in areas of life that have a high degree of social prestige, as does science. Thus the widespread acceptability of some theory by the scientific community (e.g., evolutionary theory) may well be due to various nonrational factors and not to the fact that the theory is a well-justified, approximately true view of the world. (3) Much of the point of some areas of science, especially those that are highly mathematical (like theoretical physics and speculative cosmology), is to “save the phenomena” and not much is lost if the more metaphysical aspects of some theories in these areas are rejected.

Where, then, does this leave us? In our view, we are not forced to choose between realism or antirealism. Rather, we are free to adopt one or the other on a case-by-case basis. We should start with an examination of specific, actual scientific theories, assess the strengths and weaknesses of realist and antirealist interpretations of those theories in light of the comments just made, and make our choice accordingly. Note that one should start with particular cases in science and not with general criteria for choosing between realism and antirealism. General, though not universal, criteria for such choices emerge only after careful consideration of particular cases.

Are there any criteria that can be used to help us decide between a realist or antirealist view of some specific theory? We know of no complete set of adequate criteria, but the following are some criteria that, while admitting of exceptions, are still valuable. On the strength of the three points just mentioned in favor of scientific realism (truth as correspondence and the reality of the external world; the existence of knowledge beyond the empirical; the importance of truth for rationality), let us assume scientific realism and place the burden of proof on the antirealist. Then one should interpret a scientific theory in realist terms unless one or more of the following obtain:

These criteria should be understood as mere guides and not as absolute rules. So understood, they can help inform any attempt to understand science and can be useful in the attempt to integrate science with theology. Principles (2) through (5) require little comment. (2) focuses on the realist emphasis on the progressive, refinement view of the history of science and asserts that, when absent or scarce in an area of science, this counts against realism. (3) captures the insight that if a theory’s acceptance can be adequately explained by nonrational (e.g., sociological) factors, then we need not appeal to the approximate truth of the theory to explain why it is viewed as a success. (4) brings out the idea that in areas of science (perhaps some areas of mechanics, theoretical physics and speculative cosmology) where the epistemic virtue of empirical adequacy carries most of the weight for a theory (often these will be areas that are heavily mathematical or that depict a phenomenon in ways that are vastly different from the world of common sense, e.g., the quantum world), the more metaphysical aspects of the theory are less important for the rational acceptability of that theory. Finally, ad hoc adjustments of a theory are those solely designed to save it from being falsified by an anomaly and that accomplish little or nothing else. While it is not easy to tell when an ad hoc adjustment of a theory is appropriate or inappropriate, nevertheless, to the degree that a theory has been maintained through inappropriate ad hoc readjustments, it loses its claim to understand its success as a sign of its approximate truth.

That leaves us with (1), which may be the most important of the five for the task of integration. (1) tells us that if a theory, realistically interpreted, has problems of vagueness, circular definition, or lack of clarity among its concepts, then if an antirealist understanding of the theory removes these problems, that counts in favor of the antirealist understanding. Moreover, if a realist understanding of a theory conflicts with rationally justified philosophical or theological beliefs whereas an antirealist view does not, this too counts in favor of the antirealist position. For example, if we have good philosophical or theological reasons for believing that the space-time universe had a beginning a finite time ago, then if a “successful” scientific model runs counter to this belief, it may be best to interpret the model in antirealist terms.

Principle (1), as well as principles (2) through (5) and indeed the realist-antirealist debate in general, are important to keep in mind when we try to integrate our theological and scientific beliefs. Can more be said about how philosophy can be an aid to the task of integration? Most definitely, as the next chapter will illustrate.

Chapter Summary

Science regularly uses terms that stand for unobservable, theoretical entities (e.g., electrons). Deciding whether to believe in these entities involves taking a position on the scientific realism-antirealism debate. Moreover, understanding this debate can also inform the task of integrating science and theology.

Scientific realism is the view that science is an objectively rational set of disciplines that progressively obtain truer and truer theories, whose central observational and theoretical terms genuinely refer to real entities and processes and whose theories embody epistemic virtues if and only if they are approximately true. Scientific realists argue that their view is superior to antirealist positions, that SR1-SR5 are rational to believe, that scientific realism explains the otherwise miraculous fact that scientific theories work, make progress, and can be extended into new domains, and that scientific realism fits with certain theological beliefs. Antirealists reject these claims and argue that the notion of truth as correspondence and approximate truth are problematic, that theoretical terms can be eliminated from science without loss of empirical content, that successful theories can be false and theories taken as true by realists can be unsuccessful for a long time, and that science either does not make progress or that its progress is unrelated to truth.

Antirealist alternatives come in two main groups. First, there is rational nonrealism (instrumentalism). Instrumentalists agree that science is objectively rational, but they reject the realist claim about truth regarding theoretical entities. Rather, theories are useful instruments for summarizing sensations (phenomenalism) or laboratory operations (operationalism), for pragmatically solving problems (Laudan), or for constructing empirically adequate pictures of the world (van Fraassen). Second, there is nonrational nonrealism, represented by Thomas Kuhn, which rejects both the truthfulness and referential aspects of science, as well as the notion that science is objectively rational.

Regarding the task of integration, Christians are divided about this debate. Moreover, there are good and bad arguments against scientific realism. In light of this, an eclectic model may be the best. This model implies taking a realist or antirealist view on a case-by-case basis, guided by some useful principles to aid in such decision making.

Checklist of Basic Terms and Concepts

ad hoc

alethic or truth realism

antirealism

commensurable

constructive empiricism

entity realism

epistemic virtues

incommensurable

nonrational nonrealism

observational terms

operational definition

operationalism (operationism)

paradigm

phenomenalism

pragmatism

rational nonrealism or instrumentalism

reductionism

refinement

replacement

save the phenomena

scientific realism

theoretical terms

verisimilitude

↑ Back to top