<<

and the of Author(s): George Gaylord Simpson Source: Science, New Series, Vol. 139, No. 3550 (Jan. 11, 1963), pp. 81-88 Published by: American Association for the Advancement of Science Stable URL: http://www.jstor.org/stable/1710098 Accessed: 05-01-2016 23:56 UTC

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/ info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

American Association for the Advancement of Science is collaborating with JSTOR to digitize, preserve and extend access to Science. http://www.jstor.org

This content downloaded from 131.202.192.185 on Tue, 05 Jan 2016 23:56:53 UTC All use subject to JSTOR Terms and Conditions 11 January 1963, Volume 139, Number 3550 IM E ITC E

them-eventually abandoned the Greek way of deciding how things ought to be and gave us our way of observing how in fact things are. Definitions of science may differ in other respects, but to have Biology and the any validity they must include this point: the basis of science is observa- tion. This may be expressed and applied Nature of Science in different ways. Francis Bacon, a close contemporary of Kepler, had a concept Unification of the sciences can be most meaningfully of science as gathering all possible ob- servations and then deriving from them sought through study of the phenomena of . generalizations and laws by induction, in accordance with an elaborate system George Gaylord Simpson rooted in Scholastic logic. It has often been pointed out that Bacon's system does not really work and has not been followed by any successful (Bacon himself was not one). Neverthe- Day by day the researcher, the teach- expressed by Plato, for example, when less his respect for observation and his er, and the student think in terms of he said in The Republic, "We shall let operational approach to science are immediate goals and tasks: the project, the heavenly bodies alone, if it is our among the points in which his influence the lecture, the assignment. All of us design to become really acquainted has been profound and beneficial. should occasionally back off a bit and with ." In other words the ask some questions from a wider per- essence of things was believed to reside spective. What is science anyway? What in a philosophical ideal, and observation The Definition of Science do hope to accomplish? How of real phenomena was considered not does a particular science articulate with only unnecessary but also positively Thinking of science in terms of meth- science as a whole? wrong. Some five centuries after Plato, ods came more and more into vogue Ptolemy again formalized the Greek with the scientific triumphs of the 19th way and helped to embed it in Western century and science's great acceleration Escape from the Greeks thought for another 1500 years when into the 20th. Indeed one still, although he said that the goal of astronomy was now less frequently, hears of teaching There is a whole library of attempts "to demonstrate that all heavenly phe- the , as if science were to define science. The literature is so nomena are produced by uniform circu- a set routine applicable to any subject. prolix and in part so contradictory that lar motion." Now, that is not physically That tendency reached a climax with I cannot analyze it and should perhaps true, and Ptolemy knew that it was not. Karl Pearson and others a generation hesitate to add to it. The element of He was explicit that his intention was or so ago. The scientific method was confusion is well illustrated by a recent not to explore physical . The sometimes formalized as involving six statement that science is "thinking early astronomers' only gestures toward successive operations. about the world in the Greek way." reality were attempts to "save the ap- 1) A problem is stated. That is in fact an important thing that pearances," that is, to try to eliminate 2) Observations relevant to the prob- science is not. It has often been argued obvious contradictions without aban- lem are collected. that the Greek sense of order was a doning their a priori philosophical ideals, 3) A hypothetical solution of the necessary condition for the rise of sci- such as that of uniform circular motion. problem consistent with the observa- ence. Necessary, perhaps; sufficient, def- "Saving the appearances" was a euphe- tions is formulated. initely no. The actual origin of science mism for saving the philosophical postu- 4) Predictions of other observable in the modern sense involved a revolt lates. Facts were not to be explained, phenomena are deduced from the hy- against thinking in the Greek way. but to be explained away. pothesis. The Greek way, which became tradi- Science was born when a few think- 5) Occurrence or nonoccurrence of tional in medieval , was well ers decided that appearances were the predicted phenomena is observed. something not to be saved but to be 6) The hypothesis is accepted, modi- The author is Agassiz professor of , Museum of Comparative Zoology, respected. Those hardy souls-Coper- fied, or rejected in accordance with the Harvard College, Cambridge, Mass. nicus, Galileo, and Kepler were among degree of fulfillment of the predictions. 11 JANUARY 1963 81

This content downloaded from 131.202.192.185 on Tue, 05 Jan 2016 23:56:53 UTC All use subject to JSTOR Terms and Conditions There is no question that such a cut- fourth and fifth. The sixth step is not in fact form al deduction is niot invaria- and-dried method does work in partic- eliminated but is simply taken for grant- bly involved in scientific self-correction. ular instances, or that each of the six ed and not stated. On that more must be said, but here we operations is essential in various phases The main virtue of Conant's formu- have reached a point where another at- of scientific . Nevertheless, the lation is its recognition of the role of tempt to define science is in order. formulation fails as an overall charac- speculation, intuition, or just plain One way to approach definition is to terization of science. It is not a defini- hunch in finding a hypothesis. It ignores consider science as a process of ques- tion; it says nothing of the goals or the the fact that some observation inevita- tioning and answering. The questions nature of science. Its implication of a bly precedes the speculation, and both are, by definition, scientific if they are general routine that automatically solves formulations fail to note that observa- about relationships among observed any scientific problem is false. It quite tion (whether of nature or of an ex- phenomena. The proposed answers ignores the most difficult, most creative, periment) always is the first step in any must, again by definition, be in natural and most important elements of scien- scientific investigation. No one ever had terms and testable in some material tific endeavor. How does one discern a a hunch that was not about something way. On that basis, a definition of sci- problem, or decide what kinds of ques- -in the case of science, about possible ence as a whole would be: Science is an tions are to be asked? How does one relationships among facts already exploration of the material determine what observations are rele- known. that seeks natural, orderly relationships vant? And especially, what kind of Conant makes the essential point that among observed phenomena and that is hypothetical solutions are acceptable the aim of science is to seek and verify self-testing. We may well add, but not and where do they come from? Perhaps general ideas, relationships, and inter- as part of a definition, that the best the most cogent objection of all is that connections among phenomena. Obvi- answers are theories that apply to a important basic research has seldom ously science has nothing to do and wide range of phenomena, that are sub- really followed the "method" just as cannot exist if phenomena have not, in ject to extensive tests and that are sug- it is stated. fact, been observed, but there science gestive of further questions. It is such In our own days James Conant has begins, not ends. It follows that al- theories that contribute most to the on- strongly criticized that kind of formula- though the observation of facts or mem- going aspect of science so properly tion and has proposed a new definition orization of data is a necessary basis stressed by Conant. Nevertheless, most of science and another characterization and accomplishment of science, that in scientific endeavor has more limited ob- of its methods. His definition is: "An itself is not science. Science, truly to be jectives, and some endeavor, even interconnected series of concepts and such, must center not on descriptions though scientific by definition, has no conceptual schemes that have developed and names but on principles-that is, evident sequel. as a result of experimentation and ob- generalizations, theories, relationships, servation and are fruitful of further interconnections, explanations about experimentation and observation." He and among the facts. The Straying Physical Sciences characterizes scientific method as com- A second point often left implicit but prising: "(1) speculative general ideas, requiring meticulous attention is that It is noteworthy that almost all studies (2) deductive reasoning, and (3) experi- the materials of science are literally of the philosophy and methods of sci- mentation." Like all brief statements on material. The observations of science ence have referred primarily to the any subject, these are ambiguous and are of material, physically or objectively physical sciences. That is in part be- incomplete outside of the expanded observable phenomena. Its relationships cause the physical sciences do have a context given them by the author. The are material, natural relationships. This primacy-not, I insist, logically but his- definition, taken by itself, does not de- is not to say that science necessarily torically. The first sciences, as we now fine. If reread, it will be found to apply denies the existence of nonmaterial or strictly define science, were physical perfectly to the work of Picasso, for supernatural relationships, but only that, sciences. That was at a time when sci- example, and although Picasso's work whether or not they exist, they are not entists considered themselves to be also, is certainly creative and great it is with the business of science. This requires, or even primarily, philosophers, and in- equal certainty not science. Of course if you like, a measure of self-discipline deed "natural philosophy" was long Conant's point is to emphasize the among scientists, a recognition that their synonymous with "." The tradi- dynamic, ongoing nature of science. methods do not work properly in the tion has persisted. It has been reinforced That is a characteristic of the most im- absence of this restriction. by the reductionist half-truth (of which portant scientific investigations, but dy- The third feature that distinguishes more later) that all phenomena are ulti- namism is not confined to science and science from other fields of thought mately explicable in strictly physical does not characterize all of science. and of activity is that it is self-testing terms. Another factor has been the Conant may also be too hard on his by the same kinds of observations from prestige accruing from the thorough predecessors. His summary of scientific which it arises and to which it applies. and more obvious impingement of the method is freer, more impressionistic, It is, to use a currently popular but physical sciences on daily life through than the earlier formulation and to that perhaps overworked bit of jargon, a technology. It is also possible that more extent more nearly covers the varied cybernetic system with a feedback that of the most brilliant and thoughtful gambits of research. It is not, however, in spite of oscillations keeps its orienta- minds have gone into the physical sci- contradictory, and in some respects it is tion as nearly as may be toward reality. ences; I prefer not to think so, but I less complete or explicit. It starts with That is the point of the deductive phase suspect there is some truth in that. the formulation of the hypothesis, the stressed by both Pearson and Conant The point here is that most consider- third step of the earlier summary, and as well as by almost all other modern ations of the history, methods, and na- its other two steps are essentially the writers on scientific method, although tur~eofl science h~axebeen heavily biased

82 SCIlENC'E. SOL, 139

This content downloaded from 131.202.192.185 on Tue, 05 Jan 2016 23:56:53 UTC All use subject to JSTOR Terms and Conditions by concentration on physical science physicists have failed somewhere and on them is acceptable if there is con- and not on science as a whole. That has that there must be some rational way fidence that anyone in a position to been notably true of concepts of sci- to get over the difficulty. observe them would have observed entific laws, of predictability, of the The aspect that I spoke of as almost them. testing of hypotheses, and of causality. comic is this: well before the "revolu- In what used to be called the exact Francis Bacon warned, "Though there tion" life scientists had observed that sciences, which have turned out not are many things in nature which are laws, in the rigid 19th-century con- to be so exact, it was formerly as- singular and unmatched, yet it [the ception of physics, do not apply to sumed that uniform phenomena had human understanding] devises for them many phenomena in nature. Further, absolute constants measurable to any parallels and conjugates and relatives they knew that prediction (not the degree of accuracy. As a very simple which do not exist." Nineteenth-century only way of testing hypotheses) is com- example, the length and period of a physicists did not heed his warning. monly statistical and no less scientific pendulum were assumed to have an They refused to consider the unique ob- or confirmatory of a hypothesis for all infinitely exact and determinable value. ject or event and assumed that all that. They knew that this is no con- It now appears that this is not neces- phenomena could be reduced to sup- tradiction of the orderliness of nature, sarily true, and that is one of the dis- posedly invariable and universal laws and they discerned that only an un- coveries that so upset the physical sci- such as the gas laws or the law of gravi- necessarily restricted concept of causal- entists. But in the actual practice of tation. It followed that, once a law was ity is affected. The "revolution" was a observation it has always been evident known, its consequences could be fully revolution only for those who had in- that infinitely exact measurement is im- predicted. The consequences deduced sisted that everything must be explained possible. All that repetition and instru- from a hypothesis became predictions ultimately in terms of classical phys- mental refinement can do is to generate as to what would happen if an experi- ics-and where were there ever any a degree of confidence that a measure- ment were performed, and that is the real grounds for such a narrow view of ment (at any given time and under pertinent test embodied in Pearson's, science? It is true that understanding, given conditions) lies within a certain and still in Conant's (and many of statistical law and polymodal causal- range. Inference from the observation others'), descriptions of scientific meth- ity had crept over the life scientists takes into account the size of the range od. It further followed-or the physical gradually, so that the impact of these and the degree of confidence. The con- scientists thought it did-that when a concepts was not seen as revolutionary. clusion that even in principle the range law successfully predicted an event, the It is also true that not many cannot be infinitely small and confi- law explained the event as a result and are given to exploring the philosophical dence infinitely great makes no differ- specified its causes. implications of their science. There was ence operationally, at least. Here we in the 20th century have therefore little really clear discussion of It is further true that with many seen something curious and indeed al- causality in biology before that by phenomena the whole point of obser- most comic happen. Physicists have Ernst Mayr in 1961. vation is not an exact measurement or found that some, at least, of their laws determination of occurrence but estab- are not invariable; that their predictions lishment (again to some degree of con- are statistical and not precise; that some Self-Testing in Science fidence) of a probability. The classical observations cannot in fact be made; example is the tossing of a coin, and and that absolute confirmation by test- A fundamental, though not a suffi- here the biologists' point is that we do ing of a hypothesis therefore cannot cient, criterion of the self-testability of not expect the probability of throwing be obtained. Many have gone further science is repeatability. Norman Camp- heads to be exactly one-half. As mod- and concluded that causality is mean- bell's definition of science as "the study ern scientists and not ancient Greeks, we ingless and even that order in nature- of those judgments concerning which are examining real, objective coins and the last scientific relic of- our Greek universal agreement can be obtained" not the Platonic idea of a coin. By re- heritage-has disappeared. That is, of emphasizes this point. That is indeed peated observation of a real coin, we course, the so-called scientific revolu- not so much a definition of science as can establish a high degree of confidence tion wrought by quantum theory and of its field and its connection with re- that the probability is in a certain the principle of indeterminacy. The ality. Campbell's meaning is that the range. If the range is large, it is likely physicists' reactions to this (even in data of science are observations that to include the probability of one-half, my very limited knowledge of them) can be repeated by any normal per- but if the range is made small it is likely ran the gamut from reason to hysteria. son. That is as true of, say, the ob- to exclude that a priori ideal. Analogous Some-Bridgman is a sad example- servation of a fossil under a phenomena are very common in bi- found science coming apart in their microscope as it is of the height of ology. For example, we do not expect hands, further scientific knowledge im- in a tube in Torricelli's fa- an expanding population of flies to possible, and the universe and existence mous experiment, or of more recent ob- spread according to an exact law. We itself left wholly meaningless. Others, servations of separation by expect only to achieve confidence that such as Jeans, also accepted the whole chromatography and electrophoresis. the rate will be within a certain range of idea of orderlessness and acausality but, Illusion, even to the point of hallucina- probability, or to construct a frequency with almost mystical glee, likened the tion, is always a possibility, but it is distribution of rates. Discovery that release from physical law to release one that can be eliminated for all prac- Boyle's "law" has the same probabilistic from prison. Still others, with Schrdd- tical purposes by repetition of observa- nature neither surprises nor upsets us. inger, have had -what seems both the tions, especially by different observers We would expect it, because the mole- most mature and the most scientific re- and different methods. It is also true cules of gas, like the flies, are real in- action: they have concluded that the that unique events occur, but evidence dividuals which, however alike they are 11 JANUARY 1963 83

This content downloaded from 131.202.192.185 on Tue, 05 Jan 2016 23:56:53 UTC All use subject to JSTOR Terms and Conditions in other respects, have had different fidence in many things we call "facts" there is an answer. In fact there are histories. The Greeks could, but a sci- in daily life. That kind of nonpredictive several. Russell himself has provided entist cannot, be concerned with the testing most commonly occurs in fields one, apparently unwittingly, although ideal gas of classical physics. Perhaps that have a temporal or historical ele- it is dangerous to assume that he is ever the revolution in physics was only the ment, such as among the bi- unwitting. His example of what he final severing of the umbilical cord ological sciences or the time-linked calls the "known events" includes light from ancient . processes in geology among the physical from the bouncing off a man The most widespread and conclusive sciences. In fact a neglected historical named Jones and then entering the process of self-testing in science is test- component also affects many physical . "Jones himself" may still be ing by multiplication of relevant obser- laws, as in the example of the histories "wrapped in mystery," as Russell says, vations. In the natural sciences it is of the individual molecules in a gas. but evidently something happened out impossible to prove anything in the ab- there. The faith required is not that solute sense of, for example, a proof in "out there" exists, but that what hap- mathematics. Multiplication of obser- Science and Reality pens "in here" contains some informa- vations can only increase our confidence tion about it. Such an answer obviously within a narrowing range of probability. In discussing the nature and basic does not supply a philosophical abso- If confidence becomes sufficiently great procedures of science I have been quite lute, but it should satisfy a scientist's and the range is encompassed by the free in using such expressions as "re- more modest demand for reasonable hypothesis, we begin to call the hy- ality," "phenomena," "the material uni- confidence. pothesis a theory, and we accept it and verse," and so on. Philosophers have Norman Campbell has pointed out go on from there. The test, is of course, long since pointed out, and philo- that the fact that others demonstrably whether the range of probability is in sophical scientists are still worried by, receive the same sensations as we do fact within the scope of the hypoth- the fact that the very existence of such from the same stimuli is evidence that esis-in other words, whether the a thing as objective reality is uncertain. the outer world does exist. That is the observations are consistent with the I have already referred to Bridgman's basis for his remark, quoted earlier, on hypothesis. despairing conclusion that "the very the obtainability of universal agreement A key word in the expression "mul- concept of existence becomes meaning- in [observational aspects of] science. It tiplication of relevant observations" is less." In The Scientific Outlook Ber- is also evidence that the stimuli are relevant. The simplest definition is that trand Russell has discussed this matter structured-that is, do convey informa- relevant observations are those that more optimistically if equally incon- tion. Again a philosopher may quibble could disprove the hypothesis, for dis- clusively. He points out (in more and and say that the reactions of others proof is often possible even though different words) that what we call ob- have no bearing if the others are not absolute proof is not. The more obser- serving a phenomenon is in fact only really there, but a scientist will gain vations fail to disprove a hypothesis, sensing certain events that occur to and another degree of confidence. the greater the confidence in it. Predic- within ourselves. For example, when we Still another consideration seems to tion in the classical sense is a special think we have seen something, we know me the most interesting of all, and yet case of that general procedure. From only the event that light quanta of cer- I have never seen it clearly expressed the hypothesis consequences are de- tain energies and patterns impinged on elsewhere. It is, in a sense, a validation duced such that their failure to occur our retinas and produced other events of the " faith" given by Russell would disprove the hypothesis. Of in our nervous system. The object we (after Santayana) in the passage quoted course their occurrence would not think we saw "remains veiled in mys- earlier, as sole basis for assuming that prove anything; it would only increase tery." Russell asks finally, "Are circum- we really can obtain knowledge of the confidence. That this is in fact a spe- stances ever such as to enable us, from outer world. The fact is that man origi- cial case and not the touchstone of a set of known events [for example, nated by a slow process of evolution scientific theory is easy to demonstrate. those in our nervous system] to infer guided by . At every Again, examples are more familiar to that some other event [for example, the stage in this long progression our an- biological than to physical scientists, al- material existence of what we think we cestors necessarily had adaptive reac- though they occur in both fields. The see] has occurred, is occurring, or will tions to the world around them. As be- most striking example is the most im- occur?" He concludes, "I do not know havior and sense organs became more portant of all biological theories: that of any clear answer. . . . Until an an- complex, perception of sensations from of organic evolution. Although some swer is forthcoming, one way or those organs obviously maintained a quite limited predictions can be de- another, the question must remain an realistic relationship to the environ- duced from the theory, the theory was open one, and our faith in the external ment. To put it crudely but graphically, not in fact established by prediction world must be merely animal faith." the monkey who did not have a realistic and is not sufficiently tested by it. An Now, some feel that this is nonsense perception of the tree branch he jumped enormous number of observations enor- and that sensible people will not waste for was soon a dead monkey-and mously varied in kind are all consistent time on it. Whether or not there really therefore did not become one of our with this theory, and many of them are is an external world, we certainly have ancestors. Our perceptions do give true, consistent with no other theory that has to act as if there were, so we may as even though not complete, representa- been proposed. We therefore can and, well ignore the question. Indeed I shall tions of the outer world because that if we are rational, must have an ex- not here spend much time on it, but it was and is a biological necessity, built tremely high degree of confidence in has bothered many scientists, so it does into us by natural selection. If it were the theory--higher than legitimate con- seem worth while to point out that not so, we would not be here! We do 84 SCIENCE. VOL. 139

This content downloaded from 131.202.192.185 on Tue, 05 Jan 2016 23:56:53 UTC All use subject to JSTOR Terms and Conditions now reach perceptions for which our system observed includes the observer necessary casualty and that the con- ancestors had no need, for example of has quite a simple meaning. It merely cept of objectivity essential to science x-rays or electrical potentials, but we means that all systems in science have is saved by recognition that scientific do so by translating them into mo- a biological component. objectivity has a biological component. dalities that are evolution-tested. There is another, related sense in A related casualty that was almost in- which all science is partly biological. evitable in the struggle to develop mod- It is all carried on by human beings, ern science involves the concept of Biological Nature of Science a species of animal. It is in fact a teleology. part of animal behavior, and an in- The doctrine of final cause, of the That is one of the several senses in creasingly important part of the species- end's determining the means, was which science itself, as a whole, is fun- specific behavior of sapiens. another essential element in Greek damentally biological. A second sense From the functional point of view, thought, which was anthropomorphic in which that is true is involved in it is a means of adapting to the environ- in a truly primitive way. This doctrine another point that has lately been ment. It is now, especially through its was probably an inevitable outcome of bothering the physicists. The point is operating arm, technology, the princi- introspective and deductive philosophy. that whenever a scientist observes any- pal means of biological adaptation for Rational human actions are largely ex- thing he is himself part of the system civilized man. It is an evolutionary spe- plicable by their purpose, by the results in which the observing takes place. cialization that arose from more primi- they are expected to produce. It there- He therefore should not assume that tive, prescientific means of cultural fore seemed logical to conclude that what he observes would be exactly the adaptation, which in turn had arisen from the orderly intricacy of the world at same if he were not observing it. But still more primitive, prehuman behav- large was in a similar way purposeful he cannot very well observe what hap- ioral adaptation. I recently had occa- and governed by a foreseen end. Such pens when he is not observing! There- sion to point out to some ethnologists concepts were particularly important fore, the argument runs (but person- that culture in general is biological to Aristotle, and through his works ally I do not run with it), there is no adaptation and that they could resolve they came to be held as almost axio- such thing as objective knowledge, and some of their squabbles and find the matic in the western European milieu the goals of science are wholly de- common theoretical basis that eludes in which science finally arose. The lusive. Some atomic physicists say this them if they would just study culture broadly philosophical position was that does not matter as far as the man- from this point of view. The suggestion things exist, or events occur, as pre- sized world is concerned but matters was not well received, but it is true requisites of- their results, and that the only when you get down to their in- just the same. Some thought I was result, as final cause, is the real prin- visible, 'but all too obviously not imagi- being: a racist and some thought I ciple of explanation. In more popular nary, objects of study. Yet I really do was being a social Darwinian, both form, this view led to the belief that not see why' size matters in principle. quite rightly pejorative epithets in nature exists only for and in relation In either case the system actually ob- ethnological circles. Of course I was to man, considered as the ultimate pur- served contains something alive-to being neither one. I was just being a pose of creation or the overriding final wit (as a minimum), the observer. drawing attention to the really cause. Surely it would never occur to anyone quite obvious fact that culture is a bio- As physical science became more but an atomic physicist that because a logical phenomenon. That is true, in objective, it was apparent that tele- system includes something alive it heightened degree, of the special aspect ology, even if not rejected as a philos- cannot be properly studied! of culture we call science. ophy, had to be ignored as a means of To suppose that study, to be ob- scientific explanation. The scientist, as jective, should exclude the observer such, asked "What?" or "How?" about is as unrealistic as Plato. Science is Flight from Teleology phenomena such as gravity or gas pres- man's exploration of his universe, and sure, not "Why?" or "What for?" De- to exclude himself even in principle is As Gillispie has admirably shown in scription of how things fall, in terms certainly not objective realism-unless his book The Edge of Objectivity, the of masses, distances, and gravitational you insist that his inclusion is subjec- rise of science, in the strictest modern constants, is clearly scientific, but the tive by definition, but that would be sense of the word, centered around in- question, "What do things fall for?" merely playing with words. And to say creasing insistence on objectivity. It seems unscientific. It elicits no objec- that we cannot learn anything mate- now seems clear that in some instances tively testable answers. It was thus in- rially factual about a situation if we that insistence went too far. I have evitable that the strictest scientific atti- ourselves are in it is utter and non- noted that some scientists reached the tude should endeavor to exclude any sensical negation of the very meaning unnecessary and, in the last analysis, form of teleology, and in the physical of learning. The essential in objectivity absurd position that complete objectiv- sciences there seemed to be no great is not the pretense of eliminating our- ity would exclude the observer. Since difficulty in excluding it. One could, at selves from a situation in which we are exclusion of the observer is obviously least, readily evade teleology by ascrib- objectively present. It is that the sit- impossible in the practice of science, ing physical laws to a first rather than uation should not be interpreted in scientists who held that view, as we a final cause, although even here the terms of ourselves but that our roles have seen, tended either to fall into usual philosophical or theological belief should be interpreted realistically in despair or to revert to various more or continued to be that natural laws exist terms of the situation. To a biologist less covert forms of idealism. I have in order to make the world a suitable the discovery (to call it such) that every here maintained that this was an un- habitat for man. 11 JANUARY 1963 85

This content downloaded from 131.202.192.185 on Tue, 05 Jan 2016 23:56:53 UTC All use subject to JSTOR Terms and Conditions In the biological sciences the elimi- the second alternative. In The Origin tine effect. Many biologists threw out nation or even the brushing aside of of Species he made no entirely clear the baby with the bath water. In seek- crude teleology was incomparably more distinction between establishing the fact ing to get rid of nonscientific teleology difficult, and that is a principal reason that evolution has occurred and pro- they decided to throw out all the quite why a fully scientific biology lagged so posing a theory as to how natural proc- real and scientific problems that tele- far behind a scientific physics. It is not esses could produce organic adaptation. ology had attempted to solve. necessary or perhaps even possible to He has therefore been accused of un- That took several different forms. see any immediate, inherent purpose necessarily confusing two issues that One form in evolutionary studies was in a stone's falling, but it is quite in- should have been kept quite separate, the mutationist belief that do evitable that an animal's seeking its but that was not really the case. Evolu- not become adapted to a way of life food should be interpreted in terms of tion itself becomes a nonscientific issue but simply adopt the way of life that purpose or, at least, of an end served. if the explanation of adaptation in the their characteristics, originating at ran- All organisms are clearly adapted to course of evolution is left in the field dom, make possible. Another form was live where and how they in fact live, of metaphysics, philosophy, and theol- behaviorism, which also, in essence, and adapted in the most extraordinary, ogy. Darwin really went to the heart of sought to eliminate adaptation as a thoroughgoing, and complex ways. In the matter with unerring insight. Ex- scientific problem by refusing to con- fact they plainly have the adaptations planation of adaptation was the key sider behavior as motivated, as goal- in order to live as they do. The ques- point, and Darwin demonstrated, at directed, or even as serving needs (and tion, then, is how those key words in the very least, that a natural, objective hence in some sense having purpose) order to are to be interpreted. Until a explanation of adaptation is a rational in the as a whole. Behavior- century or so ago it occurred to very possibility and a legitimate scientific ism strove to be primarily descriptive, few naturalists to interpret them in any goal. That, at long last, made biology and what explanatory element was ad- but the classical teleological way. For a true and complete science. mitted was meant to be confined to example, to Cuvier, high priest of nat- Darwin fully respected the appear- consideration of the physiological sub- ural history in the early 19th century, ances and made no attempt to save strates and concomitants of the be- the validity of fully Aristotelian tele- them by explaining them away. The havior described. It is that latter as- ology seemed self-evident, and it was hand of man, for example, is made for pect that still influences a considerable the heart of his theoretical system. grasping. Darwin said so, and then pro- segment of opinion in biology, confining Cuvier went all the way to a man- vided a natural scientific explanation biological explanation to the physicist's centered teleological conception of the for the fact. He thus did not ignore question, "How?" and eschewing "What universe. He could think of no better the teleological aspects of nature but for?" This attitude, still strongly 'held reason for the existence of fishes- brought them into the domain of in some quarters, involves the idea that which he considered poor things, even science. Some of Darwin's contempo- scientific explanation must be reduction- to the watery, unromantic nature of raries and immediate successors recog- ist, reducing all phenomena ultimately their amours-than that they provide nized that fact by redefining teleology to the physical and the chemical. In food for man. That was also the period as the study of adaptation and by point- application to biology, that to the in of Paley's Natural Theology ing out that Darwin had substituted a quite extraordinary proposition that liv- and, later, of the Bridgewater Treatises scientific teleology for a philosophical ing organisms should be studied as test- "on the power, wisdom, and goodness or theological one. The redefinition did tube reactions and that their being alive of God, as manifested in the crea- -not take. The older meanings of the should enter into the matter as little as tion"-that is to say, on Christian tele- word teleology were ineradicable, and possible. As behaviorism omits the ology as a necessary and sufficient ex- they brought a certain scientific (al- psyche from psychology, so this form planation of nature, and most particu- though not necessarily philosophical) of reductionism omits the bios from larly of animate nature. disrepute to the whole subject. biology. The facts of adaptation are facts, and The physical scientists had earlier, the purposeful aspect of organisms is and more completely, evaded the issues incontrovertible. Even if the explana- of classical teleology. By the end of the Explanation in Biology tion offered by Aristotelian, and much 19th century, if not before, it had be- later by what was then orthodox Chris- come for them virtually a dogma that a Those tendencies were unquestion- tian, teleology were true, that would scientist simply must not ask, "What ably salutary in some respects. They definitely be an article of faith and not for?" Physical scientists considered the have helped to eliminate the last ves- of objectively testable science. Thus it question as applied to natural phe- tiges of pre-Darwinian teleology from was necessary either to conclude that nomena either completely meaningless biology. They have also helped to coun- there is no scientific explanation of or, at best, unanswerable in scientific teract vitalistic, metaphysical, and mys- organic adaptation or to provide an terms. Such was the priority and pri- tical ideas which, whatever one may acceptable, testable hypothesis that was macy of the physical sciences that this think of them in their own sphere, are scientific. Before Darwin most biologists position even came to be widely con- completely stultifying as principles of accepted the first alternative, which sidered a necessary qualification of truly scientific explanation. Here, however, (although few of them realized this scientific endeavor, part of the defini- the reductionist tendency has been two- fact) meant quite simply that there tion of science. That led in turn to a edged. By seeming to negate the very could be no such thing as a fully scien- very curious development that was at possibility of scientific explanation of tific biology. It was Darwin, more than its height in the 1920's and is still ex- purposive aspects of life, it has en- any other one person, who supplied erting a strong but now more clandes- couraged some biologists, who insist that

86 SCIENCE, VOL. 139

This content downloaded from 131.202.192.185 on Tue, 05 Jan 2016 23:56:53 UTC All use subject to JSTOR Terms and Conditions such aspects nevertheless exist, to seek tionist, is in terms of the adaptive use- nonliving systems and almost incom- explanations quite outside the legitimate fulness of structures and processes to parably more complicated. They are field of science. Naming of names is the whole organism and to the species not for that reason necessarily any less perhaps invidious, but to show that I of which it is a part, and still further, in material or less physical in nature. The am here setting up no straw man I will terms of ecological function in the com- point is that all known material proc- just mention Teilhard de Chardin in munities in which the species occurs. esses and explanatory principles apply Europe and Sinnott in the United It is still scientifically meaningful to to organisms, while only a limited num- States. say that, for instance, a has its ber of them apply to nonliving systems. The reaction went much too far. It thoroughgoing adaptations to predation And that leads to another point, my went so far as to falsify the very nature because they maintain the life of the final one. of biology and of science through lion, the continuity of its species, and supine acceptance of a dictum that all the economy of its communities. science is in essence physical science. Such statements exclude the grosser, Unity of the Sciences In fact, the life sciences are not only man-centered forms of teleology, but much more complicated than the physi- they still do not necessarily exclude a When science was arising, Francis cal sciences, they are also much broader more impersonal philosophical tele- Bacon insisted that all its branches in significance, and they penetrate much ology. A further question is necessary: should be incorporated into one body farther into the exploration of the uni- "How does the lion happen to have of fundamental knowledge. Bacon verse that is science than do the physi- these adaptive characteristics?" or, more placed this in an Aristotelian frame- cal sciences. They require and embrace generally and more colloquially, "How work really inappropriate for modern the data and all the explanatory princi- come?" This is another question that is science; he wrote it at a time when one ples of the physical sciences and then go usually inappropriate and does not mind could grasp the essentials, at least, far beyond that to embody many other necessarily elicit scientific answers as of all the sciences; and he was not him- data and additional explanatory prin- regards strictly physical phenomena. self a practicing scientist. Of course ciples that are no less-that are, in a In biology it is both appropriate and nowadays, as regards detailed knowl- sense, even more-scientific. necessary, and Darwin showed that it edge and adequate research ability, This can be expressed, as Mayr, Pit- can here elicit truly scientific answers, there is no such thing as a general sci- tendrigh, and others have expressed it, which embody those that go before. entist, a general biologist, or even a in terms of kinds of scientific explana- The fact that the lion's characteristics general entomologist. In the practice tions and kinds of questions that elicit are adaptive for has caused them and teaching of science, specialization them. "How?" is the typical question to be favored by natural selection, and and the accompanying fragmentation in the physical sciences. There it is often this in turn has caused them to be em- of the sciences have become absolutely the only meaningful or allowable one. bodied in the DNA code of lion hered- necessary. Yet this practical necessity It must also always be asked in biology, ity. That statement, which of course has not eliminated the force and value and the answers can often be put in summarizes a large body of more de- of the conception that the universe and terms of the physical sciences. That is tailed information and principle, com- all its individual phenomena form one one kind of scientific explanation, a re- bines answers to all three questions: grand unit and that there is such a ductionist one as applied to biological not only "How?" and "What for?" but thing as science, not just a great number problems: "How is heredity trans- also "How come?" Always in biology of special and separate sciences. mitted?" "How do muscles contract?" but not invariably in the physical sci- Bacon further maintained that the and so on through the whole enormous ences, a full explanation ultimately in- unity of nature would be demonstrated gamut of modern biophysics and bio- volves a historical-that is, an evolu- and the sciences would be incorporated chemistry. But biology can and must tionary-factor. into one general body by a fundamental go on from there. Here, "What for?" Here I should briefly clarify a point doctrine, a Prima Philosophia, uniting the dreadful teleological question-not of possible confusion. Insistence that the what is common to all the sciences. only is legitimate but also must even- study of organisms requires principles Despite the great change in philo- tually be asked about every vital phe- additional to those of the physical sci- sophical outlook, that has become a nomenon. In organisms, but not (in the ences does not imply a dualistic or traditional approach to the unification same sense) in any nonliving matter, vitalistic view of nature. Life, or the of the sciences. In our own days, Ein- adaptation does occur. Heredity and particular manifestation of it that we stein and others have sought unification do serve functions call mind, is not thereby necessarily of scientific concepts in the form of that are useful to organisms. They are considered as nonphysical or nonma- principles of increasing generality. The not explained, in this aspect, by such terial. It is just that living things have goal is a connected body of theory that answers to "How?"' as that heredity is been affected for upward of 2 billion might ultimately be completely general transmitted by DNA or that energy is years by historical processes that are in the sense of applying to all material released in the Krebs cycle. in themselves completely material but phenomena. In biology, then, a second kind of that do not affect nonliving matter, The goal is certainly a worthy one, explanation must be -added to the first or at least do not affect it in the same and the search for it has been fruitful. or reductionist explanation made in way. Matter that was affected-by these Nevertheless, the tendency to think of terms of physical, chemical, and me- processes became, for that reason, liv- it as the goal of science or the basis for chanical principles. This second form ing, and matter not so affected remained unification of-the sciences has been un- of explanation, which can bre called nonliving. The results of those processes fortunate. It is essentially a search for a compositionist in contrast with reduc- are systems different in kind from any least common denominator in science. 11 JANUARY 1963 - ~~~~~~~87

This content downloaded from 131.202.192.185 on Tue, 05 Jan 2016 23:56:53 UTC All use subject to JSTOR Terms and Conditions It necessarily and purposely omits much in quite the opposite direction, not Biology, then, is the science that the greatest part of science, hence can through principles that apply to all stands at the center of all science. It is only falsify the nature of science and phenomena but through phenomena the science most directly aimed at can hardly be the best basis for unify- to which all principles apply. Even in science's major goal and most definitive ing the sciences. I suggest that both the this necessarily summary discussion, I of that goal. And it is here, in the field characterization of science as a whole have, I believe, sufficiently indicated where all the principles of all the sci- and the unification of the various sci- what those latter phenomena are: they ences are embodied, that science can ences can be most meaningfully sought are the phenomena of life. truly become unified.

two groups were demographically simi- lar, and came from a single small homogeneous community, enhances the significance of our comparison but also limits the extent to which our findings Divergent Reactions to the can be considered representative of other groups with similar purposes. Threat of War Less than a month after they had formed, these two groups were sepa- rately approached by a member of our A peace and a shelter group were studied to examine research team and asked to participate their different responses to the Berlin crisis. in a research project. The six mem- bers of the research team had not worked together before, nor had any Paul Ekman, Lester Cohen, Rudolf Moos, Walter Raine, of us studied problems in the area of Mary Schlesinger, George Stone peace and war. We were, and remain, divided in our beliefs regarding civil defense and peace groups. These dif- ferences were purposely made explicit, and measurement techniques were ar- Different proposals for dealing with developer in a contemporary architec- rived at jointly in an attempt to coun- the threat of war had been offered and tural style. The first to form was the teract the influence of any one bias. discussed but generally aroused little Organization for Atomic Survival in It was not possible, however, to com- enthusiasm prior to the Berlin crisis. Suburbia (OASIS). Its members, who promise on the appropriate areas of With the intensification of international live fairly close to each other within inquiry. Instead, the domains of be- tension during the summer and early the community, planned to build a pri- havior sampled reflected our diverse fall of 1961 there was a rapid growth vate fallout shelter to accommodate a hypotheses, stemming from the differ- of interest in civil defense measures maximum of 100 people. A number of ing value orientations of the members and a proliferation of groups concerned them were also active in promoting a of the research team. The tests covered with peace. The desirability of fallout program for construction of commu- attitudes about war and peace, more shelters became a focus of conflict be- nity fallout shelters in the public schools. general opinions, personal character- tween proponents of these different ap- Members of the second group, People istics, background and life history, and proaches, and controversy was wide- for Peace, were originally brought to- game and risk-taking behavior. Most spread in Congress, among scientists, gether by their shared opposition to of the tests were specifically devised and at a community level. Within one community shelters, but they described for the study, although some parts were homogeneous community these diver- themselves as advocates of a "positive" borrowed from other studies (2). gent viewpoints were expressed in the program for peace, not just opponents A member of the research team ob- nearly simultaneous formation of two of shelters. served each meeting of the two groups groups, one organized to build a fall- People for Peace had 28 members from October 1961 to February 1962. out shelter, the other to oppose shelters. and OASIS had 26 at the time of the In the second week of January 1962, We studied these groups in order to un- study. A member was defined as any- derstand the factors which had led one who attended more than one meet- Paul Ekman is assistant professor of psy- them to adopt such different reactions ing. There were equal numbers of men chology, San Francisco State College, and a research fellow at the University of California to the threat of war. and women in OASIS; there were twice School of Medicine; Lester Cohen is a clinical The two groups that we studied were as many women as men in People for psychologist at the Langley Porter Neuro- psychiatric Institute, San Francisco; Rudolf formed within the same suburban up- Peace. Demographic data were similar Moos is assistant professor in the department of psychiatry, Stanford University Medical per middle-class community, about 20 for members of the two groups: most School; Walter Raine is a clinical psychologist miles from San Francisco (1). This is were in their mid-thirties, had more at the University of California (Los Angeles) Neuropsychiatric Institute; Mary Schlesinger is a community of about 8000 people than one child, had at least finished col- assistant professor of psychology, San Fran- who live in new, single-family dwell- lege, and were earning between $10,000 cisco State College; George Stone is a research psychologist at the Langley Porter Neuro- ings, most of them built by a single and $15,000 a year. The fact that the psychiatric Institute.

88 SCIENCE, VOL. 139

This content downloaded from 131.202.192.185 on Tue, 05 Jan 2016 23:56:53 UTC All use subject to JSTOR Terms and Conditions