A response to “Darwinism: Philosophical Preference, Scientific Inference, and Good Research Strategy” by Michael Ruse, Darwinism: Science or Philosophy, Chapter 2, Proceeding of symposium entitled Darwinism: Scientific Inference or Philosophical Preference (March 26-28, 1992).
I appreciate very much the opportunity to respond to Professor Ruse. Though it is in the nature of a response to disagree, I must say that I always appreciate his work in philosophy of biology. His arguments are presented forcefully and cogently. Because of this they have always served to clarify my thinking even when I find myself on the opposite side of a particular philosophical issue.
As a philosopher of science, I also appreciate the title of Professor Ruse’s paper; I too believe that an idea can be both a philosophical preference and a scientific inference. Recent work in philosophy of science on something called the “demarcation problem” suggests that it may be difficult to separate philosophical and scientific considerations in part because both science and philosophy share a common concern with explanation. Establishing a rigid line of demarcation between science and philosophy is especially difficult in the vexing world of origins research. So I appreciate Professor Ruse’s drawing our attention to what is perhaps a false dichotomy in the title of this conference, the one between “scientific inference” and “philosophical preference.”
Nevertheless, my philosophical preferences are somewhat different from Professor Ruse’s. It will be part of the purpose of this response to suggest that inferences with decidedly theistic implications may also be considered properly scientific. In fact, I would like to suggest that although Professor Ruse’s definition of science certainly serves certain philosophical preferences, it does not always promote theoretical openness, nor as a result, intellectual rigor.
In this response, I want to challenge two assertions that Professor Ruse has made. The first concerns his definition of the scientific attitude. Second, I want to challenge his claim that evolution, defined as common descent,1 is a fact.
Challenge to Ruse’s Definition of Science
Professor Ruse has suggested that to adopt the scientific outlook one must accept that the universe is subject to natural law,2 and that further, one must never appeal to (an intervening) agency as an explanation for events. Instead we must always look to what he calls “unbroken law” if we wish to explain things as scientists.
There are several problems with this assertion and with the so-called “covering law” conception of science that underlies it and to which Professor Ruse has appealed.3 Indeed, unsolved problems with the covering law idea of Science are legion.4 It can be no purpose of mine, however, to recall or explain all of them. Nevertheless, one of the more salient difficulties with this philosophy of science-i.e., this theory about what constitutes a proper scientific theory-is relevant to my critique of Professor Ruse’s suggestion that science is primarily concerned with explanation via natural law.
This difficulty is as follows: the covering law model incorrectly conflates scientific laws and explanations. There are two sides to this difficulty.
- In the first place, many laws are descriptive and not explanatory. Many laws describe regularities, but do not explain why the events they describe occur. A good example of this from the history of science was Newton’s Universal Law of Gravitation, which Newton himself freely admitted did not explain, but instead merely described, gravitational motion. As he put it in the “General Scholium” of the second edition of the Principia: “Hypothesis non fingo” (i.e., “I do not feign hypotheses”) 5 To assert that science must explain by reference to “natural law” would necessarily eliminate from the domain of the properly scientific all fundamental laws of physics that describe mathematically, but do not explain, the phenomena they “cover.”
- Laws cannot be equated with causes or explanations for a second reason. Many scientific explanations do not depend, either principally, or at all, upon scientific laws. Many scientific explanations depend primarily upon antecedent causal conditions and events, not laws, to do what I have called the “primary explanatory work.”6 That is to say, citing past causal events often does more to explain a particular phenomenon than citing the existence of a regularity in nature. This is, in part, because many things do not come into existence via a series of events that regularly reoccur. For example, if a historical geologist seeks to explain the unusual height of the Himalayas, he or she will cite particular antecedent factors that were present in the case of the Himalayan orogeny but were not present in other mountain-building episodes. Knowing the laws of geophysics that describe orogeny generally (if there even are such things) will aid the geologist very little in accounting for the contrast between the Himalayan and other orogenies. What the geologist needs in this situation for an explanation is not knowledge of a general law, but evidence of a particularly distinctive set of past conditions.7
The situation is similar to the situation faced by historians generally. Historical explanations of why World War I began-whether it was the ambition of the Kaiser’s generals, the Franco-Russian defense pact, or the assassination of Archduke Franz Ferdinand-invariably and primarily involve the citation of events, conditions, or actions (and not laws) that are believed to be causally antecedent to the explanandum. As Michael Scriven has shown, we can often know what caused something (both in history and in disciplines like historical geology) even if we cannot relate causes and effects to each other as formal nomological statements.8 Similarly, William Alston has shown that laws alone often do not explain particular events even when we have them.9 Thus he concludes that to equate a law with an explanation or cause “is to commit a ‘category mistake’ of the most flagrant sort.”10
Perhaps another example will help. If I wish to explain why human beings were able to fly to the moon, when apples usually fall to the earth, I will not primarily cite the law of gravity. Such a law is far too general to be primarily relevant to explanation in this context, because the law allows for a vast array of possible outcomes depending on initial and boundary conditions. The law stating that all matter gravitates according to an inverse-square law is consistent both with an apple falling to the earth and with an astronaut flying to the moon. Explaining why the astronaut flew, when apples routinely fall, therefore, requires more than citing the law, because the law is presumed operative in both situations. Accounting for the differing outcomes (i.e.. between the apple and the astronaut) will require references to the antecedent (prior) conditions and events, which differed in the two situations. In other words, explanation in this case requires an accounting of the way in which engineers have altered the initial and boundary conditions affecting the astronauts to allow them to overcome the constraints ordinarily imposed by gravity on all earthbound objects.
Confusion about the role of antecedent conditions and laws in scientific explanation has led many to create a false dichotomy between “unbroken law” and the action of agency. In Professor Ruse’s case this dichotomy is manifest in his assertion that invoking the action of a divine agent constitutes a departure from a commitment to natural law. I disagree. Posing the action of agents against the laws of nature creates an unnecessary dichotomy. The reason for this is simple. Agents can change initial and boundary conditions, yet in so doing they do not violate laws. Most scientific laws have the form, “If A then B will follow, given conditions X.” If X are altered, or if A did not obtain, then it constitutes no violation of the laws of nature to say that B did not occur, even if we expected it to. Agents may alter the course of events, or produce novel events that violate our expectations, without violating the laws of nature. To assert otherwise is to misunderstand the distinction between antecedent conditions and laws.
The tendency to conflate laws with causes, on the one hand, and to treat natural laws and agency as mutually exclusive ontologies, on the other, has produced a confused set of expectations about what scientifically acceptable origins-theories must look like. This confusion has been heightened by the positivist tendency to see all scientific practice as identical and by talk about the scientific method.
In my own research, I have argued that a clear and logical distinction exists between questions that motivate historical and nonhistorical (or what might be called “inductive” or “nomological”) science. Whereas nomological or inductive science addresses questions of the form, “How does nature normally operate or function?” historical science addresses questions of the form, “How did natural feature X arise?”11
This distinction has important implications for evaluating the scientific status of theories that invoke an antecedent cognitive act as a scientific explanation. I personally think that it suggests the legitimacy of such postulations if they also possess features such as wide explanatory power, internal consistency, and coherence. Let me explain.
When a research program concentrates on questions about how nature normally (i.e., unassisted by agency) operates, any reference to agency (whether divine or human) becomes inappropriate because it fails to address the question of interest. As I have argued elsewhere,12 much nonhistorical scientific endeavor typically seeks to infer or explain nomological relations (i.e., scientific laws), whereas historical sciences typically seek to infer past causal events. To propose a divine act (construed as an event in space and time) where a nomological relation or law is required is to misunderstand the context and character of the relevant inquiry. Neither divine nor human action qualifies as a law. To offer either when a law is sought is clearly inappropriate. On this I believe, both theists (such as myself) and others (such as Professor Ruse) can agree.
It is not at all clear, however, that references to agency are similarly inappropriate when reconstructing a causal history — i.e., when attempting to answer questions about how a particular feature in the natural world (or the natural world itself) arose. In the first place, many fields of inquiry routinely invoke the action of agents to account for the origin of features or events within the natural world, Forensic science, history, and archaeology, for example, all sometimes postulate the past activity of human agents to account for the emergence of particular objects or events. Several such fields suggest that a clear precedent exists for inferring the past causal activity of intelligent agents as part of historical inquiry (Imagine the absurdity of someone claiming that scientific method had been violated by the archaeologist who first inferred that French cave paintings had been produced by human beings rather than by natural forces such as wind and erosion.)
There is a second reason that postulating the past action of agency may be appropriate in the historical sciences. That has to do with the nature of historical explanations. Historical explanations require the postulation of antecedent causal events; they do not seek to infer laws (though they may use laws to make retrodictive inferences or to enhance the plausibility of a postulated causal history).13 To offer past agency as part of an origins scenario or explanation is therefore (at least) logically appropriate, because the type of theoretical entity provided corresponds to the type required by historical explanations. Simply put, past agency is a causal event. 14 Agency, therefore, whether seen or unseen, may serve as a valid theoretical entity in a historical theory, even if it could not do so in a nomological or inductive one. Mental action may be a cause, even if it is certainly not a law.
I would like to press my case against Professor Ruse’s prohibition against agency in science even further. I would like to argue that to exclude intelligent design a priori as a working hypothesis in, for example, historical biology is both gratuitous and anti-intellectual. Unlike Darwin, modern Darwinists can scarcely bring themselves to Consider the possibility of intelligent design, let alone actually argue against it as he did. Professor Ruse, who to his credit has spent many hours directly confronting various creationist heresies, fails in this paper to mention intelligent design on his list of scientific possibilities. Yet it must be mentioned that this is precisely the theory that Darwin himself spent most of his time arguing against.
Indeed, it must be acknowledged that it is at least logically possible that a personal agent existed before the appearance of the first life on earth. It is therefore at least logically possible that such an agent (whether visible or invisible) designed or influenced the origin of life on earth. Moreover, as Bill Dembski will argue, we do live in the sort of world where knowledge of such an agent could in principle be accessible empirically. This suggests that it is logically and empirically possible that such an agent (whether divine or otherwise) designed or influenced the origin of life on earth. To insist that postulations of past agency are inherently unscientific in the historical sciences (where the express purpose of such inquiry is to determine what happened in the past) suggests we know that no personal agency could have existed prior to man. Not only is such an assumption intrinsically unprovable, it seems entirely gratuitous in the absence of some noncircular account as to why science should presuppose metaphysical naturalism.
Moreover, to exclude by assumption a logically possible answer to the question motivating historical science seems anti-intellectual and theoretically limiting, especially since no equivalent prohibition exists on the possible nomological relationships that scientists may postulate in nonhistorical sciences. The (historical) question that must be asked about biological origins is not “Which materialistic scenario will prove adequate?” but “How did life as we know it actually arise on earth?” Since one of the logically appropriate answers to this latter question is that “Life was designed by an intelligent agent that existed before the advent of humans,” I believe it is anti-intellectual to exclude the “design hypothesis” without consideration of all the evidence, including the most current evidence, that might support it.
There is one final reason that a priori exclusions of design are anti-intellectual, indeed, even unscientific. Recent nonpositivistic accounts of scientific rationality suggest that scientific theory evaluation is an inherently comparative enterprise. Notions such as consilience 15 (which Professor Ruse mentions) and Peter Lipton’s Inference to the Best Explanation 16 (IBE) imply the need to compare the explanatory power of competing hypotheses and/or theories. If this process is subverted by metaphysical gerrymandering, the rationality of scientific practice is vitiated. Theories that gain acceptance in artificially constrained competitions can claim to be neither “most probably true” nor “most empirically adequate.” Instead such theories can be considered only “most probable or adequate among an artificially limited set of options ” Moreover, where origins are concerned, only limited numbers of basic research programs are logically possible, as Professor Ruse mentions. (Either brute matter has the capability to arrange itself into higher levels of complexity or it does not, and if it does not, then either some external agency has assisted the arrangement of matter or matter has always possessed its present arrangement.)
The exclusion of one of the logically possible programs of origins-research by assumption, therefore, seriously diminishes the significance of any claim to theoretical superiority by advocates of a remaining program. Professor Ruse’s prohibitions notwithstanding, an openness to empirical arguments for design is a necessary condition of a fully rational historical biology. In my opinion, a rational historical biology therefore must address not only the question, “Which materialistic evolutionary scenario provides the most adequate explanation of biological complexity?” but also the question, “Does a strictly materialistic evolutionary scenario, or one involving intelligent agency, or some other, best explain the origin of biological complexity, given all relevant evidence?” To insist otherwise is to insist that materialism holds a metaphysically privileged position, Since I see no reason to concede that assumption, I see no reason to concede Professor Ruse’s conception of science.
The Fact of Evolution
For me, two things follow from the inadequacy of Professor Ruse’s definition of science. First, because I reject Professor Ruse’s view of science, I am unmoved by other similar philosophical arguments (especially from scientists) against the appropriateness of design theories in general. Indeed, almost all philosophical objections to the scientific status of intelligent design are predicated upon some untenable neo-positivist criterion of proper scientific practice. Many are predicated upon precisely the same “covering law” view of science that Professor Ruse has espoused.17 Given recent work in philosophy of science by Laudan and others,18 I doubt that Professor Ruse can offer a credible and metaphysically neutral demarcation criterion that succeeds in defining science narrowly enough to exclude the possibility of a scientific theory of design without also excluding evolutionary theories such as common descent.
Second, because I reject Professor Ruse’s view of science, I am also unconvinced by his assurances that common descent is a fact, or as he once put it, “a fact, fact, FACT!”19 I say this with no particular glee or malice, since I personally could quite easily accommodate common descent to my own belief that life owes its origin in some measure to intelligent design. I am simply unconvinced by the arguments for descent and by the philosophy of science that Professor Ruse and others invoke to make their case for it.
It might seem that Professor Ruse’s philosophy of science and his arguments for common descent are unrelated. In fact, they are not. He acknowledged as much in his paper when he stated that “If you think in terms of unbroken law, then evolution makes the most sense.” What if you don’t think in terms of unbroken law — is common descent still a fact? Or rather, what if you reject the covering law model that leads Professor Ruse to speak of “unbroken law” — does common descent remain a fact then? Is it even still the best explanation? The fact is that common descent is not a fact, and Professor Ruse is letting his philosophical predilections about the nature of acceptable science drive his conclusions about biological history.
Strictly speaking, common descent is an abductive or historical inference,20 as Professor Ruse himself acknowledges when he speaks more accurately of “inferring historical phylogenies.” As defined by C. S. Peirce, abductive inferences attempt to establish past causes by viewing present effects. 21 Hence it is more accurate to refer to common descent as a theory about facts, i.e., a theory about what in fact happened in the past. Unfortunately, such theories, and the inferences used to construct them, can be notoriously underdetermined.22 As Elliot Sober points out, many possible pasts often correspond to any given present state. Establishing the past with certainty, or even beyond reasonable doubt, can therefore, be very difficult especially when the past in question occurred billions of years ago. In my opinion, none of Darwin’s five main arguments for descent-neither fossil progression, biogeographical distribution, homology, embryological similarity, nor the existence of rudimentary organs 23 -establish common descent beyond reasonable doubt, though I admit that some of those arguments do strongly suggest the common ancestry of many disparate organisms within limited groups.
I also admit that the theory of common descent produces an admirable consilience. But that is just the point. Theories have the property of consilience; facts do not. In any case, consilience is a comparative notion, and to my mind the question of whether or not a monophyletic view of biological history can achieve a greater consilence than a polyphyletic view has not yet been settled. Indeed, even supposedly invincible arguments from molecular homologies depend for their efficacy upon a priori certainty that similarity cannot be the product of common principles of design. Such certainty in my experience often seems to have been acquired on the basis of rather naive dismissals of the metaphysics of others it also seems to me to have been acquired without adequate reflection upon the implications of the molecular biological revolution which is now again suggesting to many of us the possibility of design.
- Thomson (1982), pp. 529-531.
- In addition to his conference paper, see Ruse (1982b), pp. 72-78.
- Ruse (1986), pp. 68-73, especially footnote #9, p. 73. Ruse (1988), p. 301. Hempel, (1942), pp 35-48. Hempel (1962), pp. 9-33.
- See for example Lipton (1991), pp. 43-46. Meyer (1990), pp. 39-76. Graham (1983), pp. 16-41. Scriven (1966), pp. 238-264. Mandelbaum (1961), pp. 229-242. Scriven (1959a), pp. 477-482. Scriven (1959b), pp. 448-451.
- Newton (1958), p. 302.
- Meyer (1990), pp. 47-75.
- 1bid., pp. 51-56. Scriven (1975), p. 14. Lipton (1991), pp. 47-81.
- Scriven (1959b), pp. 446-463.
- Alston (1971), pp. 17-24.
- Ibid., p. 17.
- For a thorough exposition of this, see Meyer (1990), pp. 1-136.
- Indeed, none of the above denies that laws or process theories may play necessary roles in support of causal explanation, as even opponents of the covering-law model (such as Scriven) admit. Scriven notes that laws (or other types of general process theories) may play an important role in justifying the causal status of an explanatory antecedent and may provide the means of inferring plausible causal antecedents from observed consequents. Scriven (1959b), pp. 448-449; (1959a), p. 480; (1966), pp. 249-250. See also Meyer (1990), pp. 18-24, 36-72, 84-92.
- For a more complete discussion of the prevailing neo-positivistic confusion of laws and causes, and the subsidiary role that nomological understanding does play in historical science, again, see Meyer (1990), pp. 36-76.
- Thagard (1978), p. 79. Whewell (1840), vol. 2:242. Gould (1986), p. 65. Laudan (1971), pp. 371-378.
- Lipton (1991), pp. 82ff.
- Ruse (1986), p. 73, especially footnote #9.
- See also Gillespie (1979), pp. 1-18, 41-66, 146-156. Saunders and Ho (1982), pp. 179-196. Quinn (1984), pp. 32-53. Laudan (1988), pp. 337-350. Meyer (1990), pp. 111-136. Lipton (1991). The untenable nature of Ruse’s position is manifest in his own admission that modern evolutionary theory does not meet the demarcation standards that he promulgates elsewhere as normative for his opponents. See, for example, his discussion of population genetics in Darwinism Defended [Ruse (1982a), p. 86] where he acknowledges that “it is probably a mistake to think of modern evolutionists as seeking universal laws, at work in every situation.”
- Ruse (1982a), p. 58.
- Meyer (1990), pp. 112-130. Gould (1986), pp. 60-69.
- Meyer (1990), pp. 24-34. Fann (1970), p. 33. Peirce (1931), vol. 2:375. Peirce (1956), pp. 150-156.
- Sober (1988), pp. 1 – 17.
- Ho (1965), pp. 8-20. Darwin (1859), pp. 331-434.
Alston, W. P. (1971) “The place of the explanation of particular facts in science.” Philosophy of Science 38:13-34.
Darwin, C. (1859) The Origin of Species by Means of Natural Selection. London. All quotes taken from version reprinted by Penguin in Harmondsworth, England, 1984.
Fann, K. T. (1970) Peirce’s Theory of Abduction. The Hague: Martinus Nijhoff.
Gillespie, N. C. (1979) Charles Darwin and the Problem of Creation. Chicago: University of Chicago Press.
Gould. S. J. (1986) “Evolution and the triumph of homology, or why history matters.” American Scientist 74:60-69.
Graham, G. (1983) Historical Explanation Reconsidered. Aberdeen: Aberdeen University Press.
Hempel. C. (1942) “The function of general laws in history.” Journal of Philosophy 39:35-48.
Hempel, C. (1962) “Explanation in science and in history.” In: Frontiers of Science and Philosophy, R. Colodny (ed.), pp. 933. Pittsburgh.
Ho, Wing Meng (1965) Methodological Issues in Evolutionary Theory. D.Phil. thesis, Oxford University.
Laudan, L, (1971) “William Whewell on the consilience of inductions.” The Monist 55:368-39 1.
Laudan, L. (1988) “The demise of the demarcation problem.” In: But Is It Science? Ruse, M. (ed.), pp. 337-350. Buffalo: Prometheus Books.
Lipton, P. (1991) Inference to the Best Explanation. London: Routledge.
Mandelbaum M. (1961) “Historical explanation: the problem of covering laws.” History and Theory 1:229-242,
Martin, R. (1972) “Singular Causal Explanation.” Theory and Decision 2:221 -237.
Meyer, S. C. (1990) “Of clues and causes: a methodological interpretation of origin of life studies.” Cambridge University Ph.D. thesis.
Newton, Isaac (1958) Isaac Newton’s Papers and Letters an Natural Philosophy. I. Bernard Cohen (ed.), Cambridge: Harvard University Press.
Peirce. C. S. (1931) Collected Papers, vols. 1-6. C. Hanshorne and P. Weiss (eds.). Cambridge: Harvard University Press.
Peirce, C. S. (1956) “Abduction and Induction.” In: J. Buchler (ed.) The Philosophy of Peirce, pp. 150-156. London: Routledge.
Ruse, M. (1973) The Philosophy of Biology. London.
Ruse, M. (1982a) Darwinism Defended: A Guide to the Evolution Controversies. London: Addison-Wesley.
Ruse, M. (1982b) “Creation science is not science.” Science, Technology and Human Values. vol. 7, no. 40. pp. 72-78.
Ruse, M. (1986) “Commentary: the academic as expert witness.” Science, Technology and Human Values, vol. 11, no. 2, pp. 66-73.
Ruse, M. (1988) “Witness testimony sheet: McLean vs. Arkansas.” In: But is it Science? Ruse, M. (ed.), pp. 301-306. Buffalo: Prometheus Books.
Saunders, P. T., and Ho, M. W. (1982) “Is Neo-darwinism falsifiable? – and does it matter?” Nature and System 4:179- 196.
Scriven, M. (1959a) “Explanation and prediction in evolutionary theory.” Science 130:477-482.
Scriven, M. (1959b) “Truisms as the ground for historical explanations.” In: Theories of History, P. Gardiner (ed.), pp. 443-475. Glencoe: The Free Press.
Scriven, M. (1966) “Causes, connections and conditions in history.” In: Philosophical Analysis and History, W. Dray (ed.), pp. 238-264. New York.
Scriven, M (1975) “Causation as explanation.” Nous 9:3-15.
Sober, E. (1988) Reconstructing the Past. Cambridge: M.I.T. Press.
Thomson, K. S. (1982) “The meanings of evolution.” American Scientist 70:529-531.
Whewell, W. (1840) The Philosophy of the inductive Sciences, 2 vols. London.