Science in Christian Perspective



Science and the Concept of Miracle
Department of Philosophy John Wesley College Owosso, Michigan
Department of Philosophy Tabor College Hillshoro, Kansas

From: JASA 30 (December 1978): 164-168.

Does there exist an unavoidable conflict between empirical science and religious faith in relation to the concept of miracle? Historically, such a conflict was perceived as very real and the issues often contested bitterly. To a certain extent this is still true. It is the intent of this discussion to elucidate the exact nature of this purported conflict and, thereby, hopefully motivate theists (and non-theists) to rethink in a more rigorous manner (1) exactly what it means to say that an observable phenomenon is a "miracle" and (2) exactly how such "miraculous events" function within the theistic belief system.

One of the basic assumptions of empirical science is that all observable phenomena can, in principle, he explained in terms of natural laws. Yet one of the basic tenets in most religious systems is that some observable events occur for which no totally adequate natural explanation is possible (i.e., that miracles occur). Does there not then exist an unavoidable conflict between empirical science and religious faith in relation to the concept of miracle?

Historically, such a conflict was perceived as very real and the issues often contested bitterly. To a certain extent this still true. It is the intent of this discussion to elucidate the exact nature of this purported conflict and, thereby, hopefully to motivate theists (and isoiitheists) to rethink in a more rigorous manner (1) exactly what it means to say that an observable phenomenon is a 'miracle" and (2) exactly how such "miraculous events function within the theistic worldview (i.e., exactly what their purpose and value is).

Meaning of Natural Law

Prior to the Renaissance, natural laws were generally considered prescriptive. That is, it was generally held that, just as God had prescribed certain moral laws (principles) to promote proper personal interaction among humans, he had also established (prescribed) certain binding cause/effect patterns (natural laws) to insure proper (uniform) interaction among the physical elements in his universe. Man was considered capable of "discovering" and utilizing some of these "laws" for his own benefit. But man was in no sense considered responsible for the nature of such "laws" or the fact that they existed.1 Accordingly, it seemed quite reasonable to define a miracle as "a violation of a natural law by God." Since God was thought to have established the natural laws in question, it did not seem problematic to hold that he could "violate" (bypass or modify) these normally binding interactive patterns to display his power, verify his identity or bring about a desired state-of-affairs.

This concept of natural law, however, became increasingly less acceptable during the Renaissance. First, as it became increasingly popular for the humanist to openly challenge the existence of God (at least as lie had been traditionally conceived), the concept of a lawgiver (prescriber) who had established certain cause/effect patterns could no longer simply be accepted as fact. Second, dependence upon "supernatural" explanations for observable phenomena greatly lessened as the scientific method (based on observation and experimentation) became an increasingly successful means of relating to the physical environment. For these and other reasons, it became philosophically fashionable to simply set aside the whole question of divine causation and contend that the only significant statements concerning the physical universe with which man ought to concern himself were those based on (concerned with) sense experience.2 Not surprisingly, this necessitated a new conceptualization of "natural law." No longer were natural laws conceived of in ontological terms as divinely prescribed causal patterns or forces which men discover. Rather, natural laws (scientific laws) came to he viewed as linguistic conventions. That is, natural laws came to be understood as general descriptive statements which summarize (are formulated to explain) our knowledge concerning observed regularities in the physical universe.3 No longer was it claimed, for example, that fire "produced" heat4 because God had so decreed that such a causal pattern would be binding. Rather, it was argued (in philosophical terms) that, since fire had "produced" heat in all observed occasions in the past, it was justifiable to affirm as a generalization (hypothesis) that "fire always produces heat" and on the basis of this confirmed "lawlike statement" to predict that fire would "produce" heat in each relevant instance in the future.5

Given this reading of "natural law," to define a miracle as "a violation of a natural law" became (and has remained) extremely problematic. Strictly speaking, it makes sense to claim that an event is a violation of a natural law only if (1) there exist identifiable, consistently binding, cause/effect patterns which are objective (exist and function whether perceived or not) and (2) we know (have good reasons for believing) that these patterns were not operative in relation to the occurrence under consideration. Thus, it did make sense to talk about divine violations of (suspensions or modifications of) prescriptive natural laws which God had created and man discovered. But such is not the ease when discussing descriptive natural laws. Such "laws" we have seen, do not denote objective cause/ effect patterns which operate apart from man's experience and must be discovered; they are descriptive statements which purport only to summarize (purport only to be hypothetical generalizations set forth to explain) man's past experience with various sorts of phenomenal regularities. There is no attached claim that such descriptions necessarily tell us anything about the ultimate nature of reality (i.e., there is no claim that such "laws" inform us about the nature of possible states-of-affairs-including causal states-of-affairs-which may exist apart from man's experience). Accordingly, even if an event occurs which can be proven to be a valid eounterinstanee to a present descriptive law, what necessarily follows is only that we presently possess no descriptive generalization (working hypothesis) to "explain" the occurrence. It would make no sense to claim that the occurrence was a violation of the relevant natural laws as such "laws" are not viewed as describing (do not purport to denote) the sort of objective (discoverable) cause/effect patterns which can be violated.

What does it mean to say that an observable phenomenon is a 'miracle"? Exactly how do such "miraculous events" function within the theistic worldview?

Redefinition of Miracle

To avoid such semantical confusion, many philosophers of religion and philosophical theologians have begun to characterize the miraculous, not as a violation of a natural law, hut as a permanently inexplicable event. That is, the claim has become, not that a miracle is a circumvention or modification of a law of nature, but that it is an observable occurrence in relation to which empirical science will never be able to generate descriptive covering laws (i.e., an occurrence permanently unexplainable by the empirical scientist) .6

This move certainly does circumvent the semantical problem related to the "violation model." But the characterization of the miraculous as a permanently inexplicable event has also come under attack. Anthony Flew states the basic objection quite clearly;

1. Scientific laws are like a geographer's map. Just as the geographer uses his map to describe the actual landscape, the scientist uses scientific laws to describe what actually occurs in our experience.
2. Hence, just as a discrepancy between the actual landscape and a map necessitates a change in the map, an unusual event which is not presently subsumable under a scientific law demonstrates only that the relevant laws are inadequate and in need of revision or extension.
3. This is not to say that all such revisions will be immediately forthcoming. Some observable occurrences might remain in "explanatory limbo" for lengthy periods of time.
4. However, due to the descriptive nature of the scientific enterprise, even the most recalcitrant of events must he seen as, in principle, subsumable under scientific laws.
5. This in turn means that every event-no matter how unusual or bizzare-must be seen as, in principle, explicable scientifically.7

But philosophers such as Margaret Boden and R. G. Swinhurne disagree. Boden grants that observable phenomena cannot normally be dismissed as lying forever outside the range of science but is not convinced this would always be the case. For example, she argues, let us take the logically possible case of a leper whose missing fingers reappear instantly tinder the most stringent fraud-detecting conditions (e.g., in the presence of doctors, TV. cameras, etc.). Such an event, we are told, would conflict with so many well established scientific facts that any attempt at revising our present scientific laws in such a way as to accommodate it would so weaken the predictive power of such laws that they would no longer be of practical value. Accordingly, she concludes, if such an event were actually to occur, the scientist, of necessity, would be forced to identify it as a permanently inexplicable phenomenon.8

Ssvinhurue uses a somewhat more sophisticated line of reasoning to arrive at the same conclusion;

I. Counterinstances to scientific laws are of two types: repeatable and nonrepeatable.
2. The occurrence of a repeatable counterinstance demonstrates that the relevant law is inadequate and must be revised. But the occurrence of a nonrepeatable counterinstance does not affect the explanatory adequacy of the relevant law. It only demonstrates that the event in question is a circumvention or suspension of it.
3. The test for deciding whether a counterinstance is repeatable or not is the following:
   a. An event E is a repeatable counterinstance if a new law L' accommodating F can be devised which is simple, coherent and yields new and more correct predictions than the current law L to which E is a counterinstance.
   b. But an event E is a nonrepeatable counterinstance if there can be no new law U devised which accommodates F and is simpler, coherent and better able to yield successful predictions than the current law L to which F is a counterinstanee.9
4. It is true that, based on such a test, the labeling of any given counterinstance as "repeatable" is a corrigible matter. Hence, it might be doubted that we could ever decide with certainty that a eounterinstance actually was non repeatable. But all claims to knowledge about matters of fact are corrigible, and we must reach provisional conclusions about them on the evidence available to us.
5. We have to some extent good evidence about what are the "laws of nature," and some of them are so well established and account for so many data that any modification of them which we would suggest to account for the odd counterinstance would be so clumsy and ad hoe that they would upset the whole structure of science. For example, let us imagine we experience the "resurrection from the dead . . . a man whose heart has not been heating for twenty-four hours and who was dead by other currently used criteria" or "water turning into wine without the assistance of chemical apparatus or catalysts." In such eases, it would be most reasonable for the scientist to label such phenomena "permanently inexplicable events."10

Such reasoning is confused. First, it sets up a false dilemma for the scientist. Boden and Swinburne would have us believe that, when faced with an extremely unusual occurrence which is not presently subsumable tinder known scientific laws, the scientsist must either immediately accommodate such an event within a new law or consider it "permanently inexplicable." h reality, many other options are open. The scientist could, as Flew seems to be suggesting, continue indefinitely to conduct tests, hoping thereby to gain new information which would allow him to subsume the occurrence in question under a new or present law of some sort. Or if the occurrence were singular, the scientist could simply label it as a "freak" event and await the occurrence of similar phenomena before seriously investigating further. In other words, it is simply not the ease that the scientist must make a present, conclusive decision concerning the explicability status of each unusual occurrence he encounters.

Moreover, even if the scientist were forced to make an immediate judgment on the "explicability status" of a given event, he could never justifiably decide to label an event "permanently inexplicable." Since it is logically possible that even the most bizarre event E could have necessary and sufficient empirical antecedent causal conditions, it is possible with respect to any, such event that the scientist could identify such causal factors and thereby place "E-type" events tinder a scien tific covering law of some sort. Or stated differently, since the empirical scientist could not know a priori that regularity patterns between ally given event-type E and a given set of empirical causal conditions will never he found, he could never justifiably claim-as Bodess and Swinhurne would have us believe-that E could never he subsumed under a scientific law of some sort. It may be that some occurrences will, as a matter of fact, never be explained. But this in no way entails that the scientist can justifiably attempt to identify such events.

Our criticism of the Boden-Swinburne position, however, does not mean that we are in total agreement with Flew's line of reasoning. Flew's contention is not only that the empirical scientist could never justifiably claim that an observable event is permanently inexplicable. Flew (and others) also hold that, since a valid counterinstance to an existing scientific law (set of laws) demonstrates only that the relevant law is inadequate and, accordingly, in need of revision or cxtention, there could he, in principle, no permanently inexplicable event. That is, they claim that the idea of a permanently inexplicable event is unintelligible (i.e., conceptual nonsense) 11 This is surely too strong a contention. The claim that all observable phenomena can, in principle, be subsumed under true scientific laws is plausible only if it is implicitly assumed that all observable phenomena actually do have a set of necessary and sufficient empirical antecedent causal conditions (even if science has not or will not discover them). But such an assumption cannot be granted, as it is logically possible that some observable occurrences have no sufficient set of empirical antecedent causal conditions or no empirical antecedent causal conditions at all (i.e., it is logically possible that some observable occurrences have solely-or at least some-nonempirical antecedent causal conditions)

For example, imagine the following presently inexplicable occurrence: a leper who has lost his fingers suddenly discovers that they have regrown. One possible explanation would be that this occurrence is solely the result of some rare, hitherto undetected, natural chemical process. If such were the ease, it would be within the power of empirical science, in principle at least, to discover the set of empirical antecedent causal factors involved and place such an occurrence under a scientific law of some sort. But to claim, a' la Flew, that such an occurrence would have to be, in principle, subsumable under some scientific law is unwarranted, for it would also be logically possible that the regrown fingers were totally (or partially) the result of a set of essential, uonempirical antecedent causal conditions (e.g., the action of a god). Accordingly, it would be possible that empirical science could never formulate even a small scale, true scientific law under which the occurrence (i.e., event-type) in question could be subsumed.

This fact, however, does not lessen the genuine tension that exists between empirical science and the concept of miracle if the latter is defined as a permanently inexplicable event. True, it is logically possible that miracles in this sense do (could) occur. But it is impossible under this reading for the Christian (or any theist) to contend justifiably that any given occurrence is in fact miraculous, as it is always possible that rele vant scientific laws will he developed sometime in the future.12

Significance for Apologetics

Some theists will sorely argue that this conclusion is relatively unimportant, since as they see it, the true significance of the miraculous is related not to its cxplicability status as an observable event but to the fact that it is a 'sign" from God. In other words, some theists will be quite willing to drop "permanently inexplicable" as a defining characteristic of the miraculous, and argue rather that the identifiability of a miraculous occurrence must be based upon the fact that it is an awe-producing (unique) 'act of God' which is brought about to demonstrate his presence, goodness or approval.

This stance obviously does circumvent the identification problem posed above and, therefore, may appear appealing. But there is an "apologetical price" to pay. When the miraculous is defined as a permanently inexplicable event, the explicahility status of such an occurrence becomes an objective identification criterion open to both the theist and nontheist (i.e., the explieability status becomes a"common ground" for othrworldview discussions concerning the identification of purportedly miraculous events). Accordingly, if in event could he identified as "permanently inexplicable," it would tend to make belief in God's existence (i.e., belief in the existence of a supernatural being who intervenes in earthly affairs) more plausible as it would demonstrate that empirical causation alone is not sufficient to explain all types of observable phenomena. Or, stated in even weaker terms, if the miraculous is conceived of as a "permanently inexplicable" event, the more convincingly it can be argued that science cannot (will not) be able to explain an observable occurrence (e.g., a resurrection), the more plausible it becomes to affirm a nonnatural (supernatural) explanation. In short, if miracle is defined as a "permanently inexplicable" event, the concept possesses interworldview apologetical value.

But once it is acknowledged that a miraculous event need not be permanently inexplicable ('.e., once it is acknowledged that empirical science can, in principle, explain (or recreate) any "miraculous" occurrence)
the explieability status of an occurrence can, of course, no longer function as an objective "common ground" for interworidview identification purposes. Rather, the miraculous becomes a solely religions concept identifiable only by those already possessing a theistic perspective (i.e., identifiable only by those who already affirm the possibility of divine intervention). Under this reading, the miraculous loses its inter-worldview apologetical value as the fact that an event is miraculous now becomes a consequent of, not support for, the fact that it is an "act of God."
Some theists may find this apologetieal implication distasteful, but others will not as within some theological perspectives it has long been held that miracles are " open only to the eyes of faith" (i.e., that the miraculous is a religious concept which has its meaning and value solely from within a religions perspective). But the theist's feelings on the matter are irrelevant. If a miracle is, ill principle, a scientifically explicable event, its traditional interworldview apologetical function disappears.

The most viable alternative is to define "miracle" as a religious concept (an act of God) which derives its uniqueness not from its explicahility status, but from the fact that it is part of an unusual event sequence.

Distinguishing A Miracle

Moreover, once the miraculous becomes a solely intratheistic concept, a new identification issue arises: How is the theist to identify an awe-producing "act of God" (a miracle)? Or, stated differently, since from a theistic perspective all events are in some sense "acts of God," the question becomes: Flow is the theist to distinguish awe-producing "acts of God" (miracles) from their nonawe-producing counterparts?

This is a complex question, but one obvious point must be reemphasized. Theists who conceive of the miraculous as an intratheistic concept cannot define "awesomeness" in terms of scientific explicahility since (for whatever reason-e.g., to avoid the identification problem posed above) they have already acknowledged that a miraculous event need not be a scientifically inexplicable occurrence. Given this fact, it seems most reasonable to assume that for such theists "awesomeness" is in some sense tied to the timing or sequencing of a given occurrence. That is, it seems most reasonable to assume that for such theists the identifiability of an "awesome" act of God (a miracle) is in some sense tied to the fact that it is an observable event which most rational individuals would not normally have expected to occur within the given event sequence of which it is a part. Consider, for example, the following situation.

June, a Bible College student, will soon he dismissed from school if she fails to pay a $500 debt. She feels very strongly, however, that God wishes her to remain in school arid, therefore, asks God to verify her "call" by providing her with the necessary funds. A few days later, June receives a letter from a distant aunt with whom she has not corresponded in years. The aunt writes that, while praying a few days earlier, she had suddenly received the feeling that June might be in need of financial assistance. Moreover, the letter continues, she feels led to send June exactly $500.

The event sequence described includes no observable event-token for which the scientist could not offer an empirical (natural) explanation. It is quite normal for college students to have financial needs and for relatives to offer assistance (even unsolicited assistance). But the fact that June's aunt sent exactly the right amount of money at exactly the right time, even though she had no empirical information that June needed help, makes the event sequence quite aweproducing as we would not normally expect June's need to be met ill a manner such as this. In fact, this event sequencing seems so extraordinary that, when coupled with the fact that there appears to have been divine intervention, it is easy to imagine some theists labeling Juries procurement of funds "miraculous."
But whether an event sequence is awe-producing is a relative issue-relative to the psychological perspective of each individual or group of individuals making the judgment. In other words, once the timing or sequencing of events becomes the intratheistic criterion for the identification of a miraculous event, it seems that the theist no longer has even an objective intratheistie basis upon which such an identification can be made. He seems rather to have only a subjective, psychological criterion which allows him to affirm only' that a given event is "miraculous" to him.

The Theist's Dilemma

We have been discussing the relationship between science and the concept of miracle. What has emerged is that the theist seems to he in somewhat of a dilemma, it seems that, all things being equal, it would he most appealing for the theist to define the miraculous as a permanently inexplicable event, as he would then possess an objective interworldview criterion which would allow the miraculous to function as a useful apologetical tool. however, we have seen that a miraculous event, if defined in this matter, is not identifiable. The most viable alternative, it seems, is to define "miracle" as a religious concept (an act of God) which derives its uniqueness not from its explicablity status, but from the fact that it is part of an unusual event sequence. But we have seen that under this reading the identifiability of a miraculous event becomes a subjective issue which destroys its interworldview apologetical value and greatly weakens its objective status from an intratheistie perspective.

Some theists will gladly accept (have accepted) this "weaker" concept of miracle. For those theists who are uncomfortable with it, the challenge is clear: they must generate an objective identification criterion that is built neither on the explicability status nor the timing or sequencing of the event in question. We personalh are very doubtful such a criterion can be formulated. Nor do we believe such a criterion is necessary as we feel that the "weaker" concept of miracle is sufficient for an intellectually defensible and experientially satisfying theistic belief system.


1See, for example, Augustine, City of God, XXII and Aquinas, Sunsmo Contra Gentiles, XCVIIICIII.
2See, for example, David Hume, Enquiries Concerning the Human Understanding, Section X.
3We are not talking here about simple "induction by enumeration," in relation to it would be said that "laws" are simply generalizations from past experiences (past instances). We are basically talking about the "hypotheticodeductive method" of scientific reasoning, in relation to which particular predictive "lawlike" statements are hypothesized as accurate generalizations about some aspect of the physical universe and then confirmed or disconfirmed by observation and experimentation.
4David Home, and others, have argued that experience gives us no knowledge concerning true causality; that we can at best speak of the "constant conjunction" of various phenomena in our experience. For these philosophers, the claim that any "X" produces (is the cause of) any "Y" must always remain a philosophically unverifiable (but psychologically unavoidable) hypothesis.
5To affirm a descriptive understanding of natural law, one need not deny that prescriptive laws "exist," but only deny (or doubt) that such laws can be identified (or discovered).
6See, for example, R. C. Swinburne, The Concept of Miracle (New York: MacMillian and Company, 1970), and Margaret Boden, "Miracles and Scientific Explanation," Ratio (December, 1969), pp. 137-41.
7Anthony Flew, God and Philosophy (New York: Dell Publishing Co., 1966), p. 150.
8Bodcn, pp. 137-41.
9It is of course true that every event-token is nonrepeatable (i.e., each event happens, and will only happen, at only one specific time.) Scientific laws, accordingly, describing regular (repeatability) patterns between certain types of phenomena and sets of antecedent causal conditions.
10Swinbnrne, pp. 29-32.
115ee for example, Ahstair McKiunon, " 'Miracle' and 'Paradox' ", American Philosophical Quarterly (October, 1967), pp. 305-312.
12We are talking here about the identification of a miraculous event on the basis of experience (on the basis of observable characteristics related to the event). Thus, the contention that biblical miracles can be known to have occurred because they were identified as such in the biblical record and the biblical record is a trustworthy source, is not, strictly speaking, subject to this epistemological qualifier. But the fact that the Bible identifies certain occurrences as miraculous is helpful in the present context only if it can be demonstrated that the Bible characterizes the miraculous as a permanently inexplicable event. It is not at all clear that this is the case (e.g., it is not at all clear that the parting of the Red Sea by a great wind is an inexplicable state-of-affairs).