28 August 1959, Volume 130, Number3374 WSZ 'I IF1 contribution was formulated by Wallace before Darwin published it. Admittedly, almost the same calamity befell New- ton, but only with respect to his gravi- tational work; his optics, dynamics, and mathematics are each enough to place Explanation and Prediction him in the front rank. Moreover, Dar- win's formulations were seriously faulty, in Evolutionary Theory and he appears to have believed in what many of his disciples regard as super- stition, the inheritance of acquired char- Satisfactory explanation of the past is possible acteristics and the benevolence of Nat- even when prediction of the future is impossible. ural Selection. Of course, Newton be- lieved that some orbital irregularities he could not explain were due to the inter- Michael Scriven ference of angels, but he did achieve a large number of mathematically precise and scientifically illuminating deductions from his theory, which is more than can be said of Darwin. Somehow, we feel The most important lesson to be consequence is the reassessment of Dar- that Darwin didn't quite have the class learned from evolutionary theory today win's own place in the history of science that Newton had. But I want to suggest is a negative one: the theory shows us relative to Newton's. that Darwin was operating in a field of what scientific explanations need not a wholly different kind and that he pos- ao. In particular it shows us that one sessed to a very high degree exactly cannot regard explanations as unsatis- Darwin's Importance those merits which can benefit such a factory when they do not contain laws, field. In place of the social scientists' or when they are not such as to enable We often confuse three criteria in esti- favorite Myth of the Second Coming the event in question to have been pre- mating the importance of the great fig- (of Newton), we should recognize the dicted. This conclusion, which is con- ures in the history of thought. The first Reality of the Already-Arrived (Dar- trary to the usual view of scientific ex- is the indispensability of what they wrote win); the paradigm of the explanatory planation (1; 2, pp. 319-352), has im- or said, regardless of its effect, judged but nonpredictive scientist. portant consequences for research in as a stage in the development of our Let us proceed by examining briefly those subjects in which serious errors present beliefs. From this point of view, the attempts by Darwin and others to are known to arise in the application to earn a place in history, a man need encapsulate the principles of evolution of the available regularities to individ- only be the first to discover the material in the form of universal laws and base ual cases. These subjects include a great or express the idea in question. The sec- predictions on them; and let us contrast part of biology, psychology, anthropol- ond criterion is their effect on other their lack of success in these endeavors ogy, history, cosmogony, engineering, thinkers, and nonthinkers, which, unlike with the tremendous efficacy of the ex- economics, and quantum physics. I shall the first, requires publication and recog- planations they produced. During this refer to such studies as "irregular sub- nition (or misinterpretation). The third comparison we shall try to extract the jects"; and the thesis of this article is criterion is the extent of their personal formal properties of the two key types that scientific explanation is perfectly indispensability. To judge this, we must of proposition that are associated with possible in the irregular subjects even make some estimate of the time that explanations in the irregular subjects: when prediction is precluded. One con- would have elapsed before the same one type of proposition is a weaker sequence of this view is that the impos- contribution would have been made by relative of predictions, and the other sibility of a Newtonian revolution in the others, had the individual under assess- type is a weaker relative of laws. social sciences, a position which I would ment never existed. If we introduce an maintain on other grounds, is not fatal index of "lucky fame" as the ratio of a to their status as sciences (3). Another man's importance on the second cri- Hypothetical Probability "Predictions" terion to his importance on the third, it seems very likely that Darwin has the The suggestion that in evolution we The author is on the staff of the department of highest index of philosophy of Swarthmore College, Swarthmore, lucky fame in history. see the "survival of the fittest" has some Pa. In fact, what is often regarded as his key well-known difficulties. In the first place, 28 AUGUST 1959 477 the definition of "the fittest" is difficult you $100," whereas an ordinary predic- continual creation, had the added ad- even when made relative to a particular tion is like a check for the sum. vantage that it spread the inexplicable environment. It is fairly obvious that no A second kind of difficulty with the element thin, thus making it scientifi- characteristics can be identified as con- "survival of the fittest" principle is that cally more palatable than a large lump tributing to "fitness"in all environments. many organisms are killed by factors at the beginning, whether the lump be Thus, strength may increase the chance wholly unconnected with any character- matter or numbers of species. Darwin's of fighting so much that it decreases the istics they possess-for example, they success lay in his empirical, case by case chance of survival, and intelligence may happen to be sitting where a tree or a by case, demonstration that recognizable be antiadaptive in anti-intellectual socie- bomb falls. Of course, this is sometimes fitness was very often associated with ties. Furthermore, maximum specializa- due to a habit or property they possess; survival, and that the small random vari- tion for a particular environment is in but that is not always true, since even ations could lead to the development of general incompatible, morphologically identical twins with identical habits do species. He did not discover an exact and genetically, with maximum flexi- not always die together. This really universal law but the utility of a par- bility to withstand sudden environmen- shows that (i) even at the limits of ticular indicator in looking for explana- tal changes (4). We are inclined to say stretching, "the fittest" refers to charac- tions. that the organisms adopting the former teristics of an organism, and spatiotem- line of development tend to be "fitter" poral location is not such a characteristic until the change occurs, and the latter (in physics, the study of the "properties Survival of a Species fitter when it occurs. Whatever we say, of matter" covers elasticity and molecu- it is quite clear that we cannot predict lar structure but not location), and (ii), In this Darwin was greatly assisted by which organisms will survive except in location sometimes determines survival. a feature of the data which constitutes so far as we can predict the environ- So it is simply false to suppose that "fit- a third difficulty in the attempt to sum mental changes. But we are very poorly ness" universally determines survival. up his account under the formula "sur- equipped to do this with much pre- Of course, one could go a step further vival of the fittest," no matter how fittest cision since variations in the sun's out- and define "the fittest" as "those which is defined. This is, of course, the fact put and even interstellar influences have survive"; this is not stretching but break- that our concern is with survival for substantial effects, quite apart from the ing the concept, and this step would be thousands of generations, not with sur- local irregularities of geology and cli- fatal to all the scientific claims of the vival to adulthood for one; and certain mate. However, these difficulties of pre- theory. We can get by with a tendency- factors enter into, or are absent from, diction do not mean that the idea of statement instead of an exact law, be- an explanation of the form of the ulti- fitness as a factor in survival loses all cause it justifies hypothetical and hence mate descendants of a certain popula- its explanatory power. It is not only true occasionally testable predictions and also tion, by comparison with an explanation but obvious that animals which happen explanations, but not with a tautology. of the form of the adults in the original to be able to swim are better fitted for H. Graham Cannon is thus entirely population. In particular, we must add surviving a sudden and unprecedented mistaken when he says: "So Darwin the variations in reproductive efficiency inundation of their arid habitat, and in pointed out that in the struggle for ex- (including mating efficiency, where sex- some such cases it is just this factor istence it will be those most fitted to sur- ual reproduction is involved) and in which explains their survival. Naturally vive who do in fact survive. What parental rearing efficiency, as well as ge- we could have said in advance that if a are the fittest? Simply those that sur- netic variability, and subtract some con- flood occurred, they would be likely to vive" (6). Darwin's discovery was that in siderations that affect postclimacteric survive; let us call this a hypothetical the world the way it is (and has been), survival (7). However, this transition probability prediction. But hypothetical the fitness of the organism, in a perfectly to what Simpson calls the "differential predictions do not have any value for recognizable but complex sense of "fit- reproduction of the best-adapted" does actual prediction except in so far as the ness" was very often the explanation of not eliminate the effect of the first and conditions mentioned in the hypothesis its survival.