The Method of Multiple Working Hypotheses Author(s): David C. Raup and T. C. Chamberlin Source: The Journal of Geology, Vol. 103, No. 3 (May, 1995), pp. 349-354 Published by: The University of Chicago Press Stable URL: http://www.jstor.org/stable/30071227 . Accessed: 06/02/2011 11:25

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at . http://www.jstor.org/action/showPublisher?publisherCode=ucpress. .

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

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].

The University of Chicago Press is collaborating with JSTOR to digitize, preserve and extend access to The Journal of Geology.

http://www.jstor.org HISTORICAL ESSAY The Methodof MultipleWorking Hypotheses, by T. C. Chamberlin

Introduction by David C. Raup1 Professor Emeritus, Department of the Geophysical Sciences, The University of Chicago

INTRODUCTION

T. C. Chamberlin's classic essay on "The Method of Multiple Working Hypotheses" was originally published in Science in 1890 and has been cited by virtually all who have struggled to define the scientic method. The version of the essay reprinted here was revised and somewhat shortened by Chamberlin and first published in The Journal of Geology in 1897. Now, as a century ago, the essay is controversial. It is seen by some as the best prescription ever written for objectivity and impartiality in scientific research, with its eloquent arguments against being driven by preconceived ideas and its plea to shun any tendency to fall in love (Chamberlin'sword) with any one of several carefully constructed hypotheses. To other observers, however, Chamberlin's formula is both impossible to imple- ment (Nobody could be that objective!)and would not work anyway. A. F. Buddington,the great Princeton petrologist, was heard to say: "No progress without prejudice!"(A. T. Anderson, Jr.,pers. comm.). That Chamberlin's essay remains controversial today (see, for example, Johnson 1990; Locke 1990; Glen 1994) is testimony to our continued uncertainty about how best to do science. And thanks to Chamberlin's carefully crafted logic, with due notice of alternate views, the piece remains stimulating and highly relevant. To what extent should ad hoc theories, or just plain hunches, guide research?How can one divide one's affections evenly among competing hypotheses so that each hypothesis develops without special advantage over the others? As Chamberlain methodically (but charmingly) builds his thesis, he asserts several generalizations worthy of thought-then and now. For example, he urges all of us to be heartily skeptical of any explanation of a natural phenomenon that is simple. As one illustration, he takes an interesting swipe at the "Darwinian hypothesis" by suggesting that natural selection is fine as far as it goes, but that the Darwinian influence has caused "neglect of other lines equally important." In this criticism, Chamberlin is attacking not only the potential hegemony of a single, popular hypothesis, but is also raising concern over simple explanations for complex phenomena. In fact, he does not think most simple explanations are viable: "It is rare that [a geologist's] problem is a simple unitary phenomenon explicable by a single simple cause." No quick fixes or magic bullets allowed? I cannot help wondering how Chamberlin would react today to HIV as the sole cause of AIDS or an asteroid impact on Chicxulub as the trigger for the K-T mass extinction. Are we moving away from Chamberlin's multiple working hypotheses, and if so, should we? Because of the many continuing quandarieshe highlights, Chamberlin's essay is well worth reading again.

Glen, W., 1994, How science works in the mass-extinction debates, in Glen, W., ed., Mass-Extinction Debates: Stanford,Stanford University Press, p. 39-91. Johnson, J. G., 1990, Method of multiple working hypotheses: a chimera: Geology v. 18, p. 44-45. Locke, W. W., 1990, Comments and reply on "Method of multiple working hypotheses: a chimera": Geology, v. 18, p. 918.

Essay There are two fundamentalmodes of study. The tive nature. In the other mode the effort is to think one is an attemptto follow by close imitation the independently, or at least individually. It is pri- processesof previousthinkers and to acquirethe mary or creative study. The endeavor is to discover resultsof their investigationsby memorizing.It is new truth or to make a new compilation of truth study of a merely secondary,imitative, or acquisi- or at least to develop by one's own effort an indi- vidualized assemblage of truth. The endeavor is 1 Present address:R. R. 1, Box 168Y, Washington Island, WI to think for one's self, whether the thinking lies 54246. wholly in the fields of previous thought or not.

[The Journalof Geology, 1995, volume 103, p. 349-354] No copy claimed for this article-it is in the public domain. 0022-1376/95/10303-010$1.00

349 350 HISTORICAL ESSAY

It is not necessary to this mode of study that the tion is in its proper place, it is an almost certain subject-matter should be new. Old material may source of confusion and error when it runs before be reworked. But it is essential that the process of a serious inquiry into the phenomenon itself. A thought and its results be individual and indepen- strenuous endeavor to find out precisely what the dent, not the mere following of previous lines of phenomenon really is should take the lead and thought ending in predetermined results. The dem- crowd back the question, commendable at a later onstration of a problem in Euclid precisely as laid stage, "How came this so?" First the full facts, then down is an illustration of the former; the demon- the interpretation thereof, is the normal order. stration of the same proposition by a method of The habit of precipitate explanation leads rap- one's own or in a manner distinctively individual idly on to the birth of general theories.* When once is an illustration of the latter, both lying entirely an explanation or special theory has been offered within the realm of the known and old. for a given phenomenon, self-consistency prompts Creative study however finds its largest applica- to the offering of the same explanation or theory tion in those subjects in which, while much is for like phenomena when they present themselves known, more remains to be learned. The geological and there is soon developed a general theory ex- field is preeminently full of such subjects, indeed planatory of a large class of phenomena similar to it presents few of any other class. There is probably the original one. In support of the general theory no field of thought which is not sufficiently rich there may not be any further evidence for investi- in such subjects to give full play to investigative gation than was involved in the first hasty conclu- modes of study. sion. But the repetition of its application to new Three phases of mental procedure have been phenomena, though of the same kind, leads the prominent in the history of intellectual evolution mind insidiously into the delusion that the theory thus far. What additional phases may be in store has been strengthened by additional facts. A thou- for us in the evolutions of the future it may not be sand applications of the upposed principal of levity prudent to attempt to forecast. These three phases to the explanation of ascending bodies brought no may be styled the method of the ruling theory, the increase of evidence that it was the true theory of method of the working hypothesis, and the method the phenomena, but it doubtless created the im- of multiple working hypotheses. pression in the minds of ancient physical philoso- In the earlier days of intellectual development phers that it did, for so many additional facts the sphere of knowledge was limited and could be seemed to harmonize with it. brought much more nearly than now within the For a time these hastily born theories are likely compass of a single individual. As a natural result to be held in a tentative way with some measure those who then assumed to be wise men, or aspired of candor or at least some self-illusion of candor. to be thought so, felt the need of knowing, or at With this tentative spirit and measurable candor, least seeming to know, all that was known, as a the mind satisfies its moral sense and deceives it- justification of their claims. So also as a natural self with the thought that it is proceeding cau- counterpart there grew up an expectancy on the tiously and impartially toward the goal of ultimate part of the multitude that the wise and the learned truth. It fails to recognize that no amount of provi- would explain whatever new thing presented itself. sional holding of a theory, no amount of applica- Thus pride and ambition on the one side and ex- tion of the theory, so long as the study lacks in pectancy on the other joined hands in developing incisiveness and exhaustiveness, justifies an ulti- the putative all-wise man whose knowledge boxed mate conviction. It is not the slowness with which the compass and whose acumen found an explana- conclusions are arrivedat that should give satisfac- tion for every new puzzle which presented itself. tion to the moral sense, but the precision, the com- Although the pretended compassing of the entire pleteness and the impartiality of the investigation. horizon of knowledge has long since become an It is in the tentative state that the affections abandoned affectation, it has left its representa- enter with their blinding influence. Love was long tives in certain intellectual predilections. As in the since discerned to be blind and what is true in the earlier days, so still, it is a too frequent habit to * I use the term theory here instead of hypothesis because hastily conjure up an explanation for every new the latter is associated with a better controlled and more cir- phenomenon that presents itself. Interpretation cumspect habit of the mind. This restrained habit leads to the leaves its proper place at the end of the intellectual use of the less assertive term hypothesis, while the mind in the habit here sketched more often believes itself to have reached procession and rushes to the forefront. Too often a the higher ground of a theory and more often employs the term theory is promptly born and evidence hunted up to theory. Historically also I believe the word theory was the term fit in afterward. Laudable as the effort at explana- commonly used at the time this method was predominant. Journal of Geology HISTORICAL ESSAY 351 personal realm is measurably true in the intellec- As previously implied, the method of the ruling tual realm. Important as the intellectual affections theory occupied a chief place during the infancy of are as stimuli and as rewards, they are nevertheless investigation. It is an expression of a more or less dangerous factors in research. All too often they infantile condition of the mind. I believe it is an put under strain the integrity of the intellectual accepted generalization that in the earlier stages of processes. The moment one has offered an original development the feelings and impulses are rela- explanation for a phenomenon which seems satis- tively stronger than in later stages. factory, that moment affection for his intellectual Unfortunately the method did not wholly pass child springs into existence, and as the explanation away with the infancy of investigation. It has lin- grows into a definite theory his parental affecta- gered on, and reappears in not a few individual tions cluster about his offspring and it grows more instances at the present time. It finds illustration and more dear to him. While he persuades himself in quarters where its dominance is quite unsus- that he holds it still as tentative, it is none the less pected by those most concerned. lovingly tentative and not impartially and indiffer- The defects of the method are obvious and its ently tentative. So soon as this parental affection errors grave. If one were to name the central psy- takes possession of the mind, there is apt to be chological fault, it might be stated as the admis- a rapid passage to the unreserved adoption of the sion of intellectual affection to the place that theory. There is then imminent danger of an un- should be dominated by impartial, intellectual rec- conscious selection and of a magnifying of phe- titude alone. nomena that fall into harmony with the theory and So long as intellectual interest dealt chiefly with support it and an unconscious neglect of phenom- the intangible, so long it was possible for this habit ena that fail of coincidence. The mind lingers with of thought to survive and to maintain its domi- pleasure upon the facts that fall happily into the nance, because the phenomena themselves, being embrace of the theory, and feels a natural coldness largely subjective, were plastic in the hands of the toward those that assume a refractory attitude. In- ruling idea; but so soon as investigation turned stinctively there is a special searching-out of phe- itself earnestly to an inquiry into natural phe- nomena that support it, for the mind is led by its nomena whose manifestations are tangible, whose desires. There springs up also unwittingly a press- properties are inflexible, and whose laws are rigor- ing of the theory to make it fit the facts and a ous, the defects of the method became manifest pressing of the facts to make them fit the theory. and an effort at reformation ensued. The first great When these biasing tendencies set in, the mind endeavor was repressive. The advocates of reform rapidly degenerates into the partiality of paternal- insisted that theorizing should be restrained and ism. The search for facts, the observation of phe- the simple determination of facts should take its nomena and their interpretation are all dominated place. The effort was to make scientific study sta- by affection for the favored theory until it appears tistical instead of causal. Because theorizing in nar- to its author or its advocate to have been over- row lines had led to manifest evils theorizing was whelmingly established. The theory then rapidly to be condemned. The reformation urged was not rises to a position of control in the processes of the the proper control and utilization of the theoretical mind and observation, induction and interpreta- effort but its suppression. We do not need to go tion are guided by it. From an unduly favored child backward more than a very few decades to find it readily grows to be a master and leads its author ourselves in the midst of this attempted reforma- whithersoever it will. The subsequent history of tion. Its weakness lay in its narrowness and its re- that mind in respect to that theme is but the strictiveness. There is no nobler aspiration of the progressive dominance of a ruling idea. Briefly human intellect than the desire to compass the summed up, the evolution is this: a premature ex- causes of things. The disposition to find explana- planation passes first into a tentative theory, then tions and to develop theories is laudable in itself. into an adopted theory, and lastly into a ruling It is only its ill-placed use and its abuse that are theory. reprehensible. The vitality of study quickly disap- When this last stage has been reached, unless pears when the object sought is a mere collocation the theory happens perchance to be the true one, of unmeaning facts. all hope of the best results is gone. To be sure truth The inefficiency of the simply repressive re- may be brought forth by an investigator dominated formation becoming apparent, improvement was by a false ruling idea. His very errors may indeed sought in the method of the working hypothesis. stimulate investigation on the part of others. But This has been affirmed to be the scientific method. the condition is scarcely the less unfortunate. But it is rash to assume that any method is the 352 HISTORICAL ESSAY

method, at least that it is the ultimate method. thy and in parental relations (of adoption, if not The working hypothesis differs from the ruling of authorship) with every hypothesis that is at all theory in that it is used as a means of determining applicable to the case under investigation. Having facts rather than as a proposition to be established. thus neutralized so far as may be the partialities of It has for its chief function the suggestion and guid- his emotional nature, he proceeds with a certain ance of lines of inquiry; the inquiry being made, natural and enforced erectness of mental attitude not for the sake of the hypothesis, but for the sake to the inquiry, knowing well that some of his intel- of the facts and their elucidation. The hypothesis lectual children (by birth or adoption) must needs is a mode rather than an end. Under the ruling perish before maturity, but yet with the hope that theory, the stimulus is directed to the finding of several of them may survive the ordeal of crucial facts for the support of the theory. Under the work- research, since it often proves in the end that sev- ing hypothesis, the facts are sought for the purpose eral agencies were conjoined in the production of of ultimate induction and demonstration, the hy- the phenomena. Honors must often be divided be- pothesis being but a means for the more ready de- tween hypotheses. One of the superiorities of mul- velopment of facts and their relations. tiple hypotheses as a working mode lies just here. It will be observed that the distinction is not In following a single hypothesis the mind is biased such as to prevent a working hypothesis from glid- by the presumptions of its method toward a single ing with the utmost ease into a ruling theory. Af- explanatory conception. But an adequate explana- fection may as easily cling about a beloved intellec- tion often involves the coordination of several tual child when named an hypothesis as if named causes. This is especially true when the research a theory, and its establishment in the one guise deals with a class of complicated phenomena natu- may become a ruling passion very much as in the rally associated, but not necessarily of the same other. The historical antecedents and the moral at- origin and nature, as for example the Basement mosphere associated with the working hypothesis Complex or the Pleistocene drift. Several agencies lend some good influence however toward the may participate not only [one] but their propor- preservation of its integrity. tions and importance may vary from instance to Conscientiously followed, the method of the instance in the same field. The true explanation is working hypothesis is an incalculable advance therefore necessarily complex, and the elements of upon the method of the ruling theory; but it has the complex are constant varying. Such distribu- some serious defects. One of these takes concrete tive explanations of phenomena are especially con- form, as just noted, in the ease with which the templated and encouragedby the method of multi- hypothesis becomes a controlling idea. To avoid ple hypotheses and constitute one of its chief this grave danger, the method of multiple working merits. For many reasons we are prone to refer phe- hypotheses is urged. It differs from the simple nomena to a single cause. It naturally follows that working hypothesis in that it distributes the effort when we find an effective agency present, we are and divides the affections. It is thus in some mea- predisposed to be satisfied therewith. We are thus sure protected against the radical defect of the two easily led to stop short of full results, sometimes other methods. In developing the multiple hypoth- short of the chief factors. The factor we find may eses, the effort is to bring up into view every ratio- not even be the dominant one, much less the full nal explanation of the phenomenon in hand and complement of agencies engaged in the accom- to develop every tenable hypothesis relative to its plishment of the total phenomena under inquiry. nature, cause or origin, and to give to all of these The mooted question of the origin of the Great as impartially as possible a working form and a due Lake basins may serve as an illustration. Several place in the investigation. The investigator thus hypotheses have been urged by as many different becomes the parent of a family of hypotheses; and students of the problem as the cause of these great by his parental relations to all is morally forbidden excavations. All of these have been pressed with to fasten his affections unduly upon any one. In great force and with an admirable array of facts. the very nature of the case, the chief danger that Up to a certain point we are compelled to go with springs from affection is counteracted. Where each advocate. It is practically demonstrable that some of the hypotheses have been alreadyproposed these basins were river valleys antecedent to the and used, while others are the investigator's own glacial incursion. It is equally demonstrable that creation, a natural difficulty arises, but the right there was a blocking up of outlets. We must con- use of the method requires the impartial adoption clude then that the present basins owe their origin of all alike into the working family. The investiga- in part to the preexistence of river valleys and to tor thus at the onset puts himself in cordial sympa- the blocking up of their outlets by drift. That there Journalof Geology HISTORICAL ESSAY 353 is a temptation to rest here, the history of the ques- nary as its influence for good has been, as an illus- tion shows. But on the other hand it is demonstra- tration of this. While inquiry is thus promoted in ble that these basins were occupied by great lobes certain quarters,the lack of balance and complete- of ice and were important channels of glacial ness gives unsymmetrical and imperfect results. movement. The leeward drift shows much mate- But if on the contrary all rational hypotheses bear- rial derived from their bottoms. We cannot there- ing on a subject are worked codrdinately, thor- fore refuse assent to the doctrine that the basins oughness, equipoise, and symmetry are the pre- owe something to glacial excavation. Still again it sumptive results in the very nature of the case. has been urged that the earth's crust beneath these In the use of the multiple method, the reaction basins was flexed downward by the weight of the of one hypothesis upon another tends to amplify ice load and contracted by its low temperature and the recognized scope of each. Every hypothesis is that the basins owe some-thing to crustal deforma- quite sure to call forth into clear recognition new tion. This third cause tallies with certain features or neglected aspects of the phenomena in its own not readily explained by the others. And still it is interests, but ofttimes these are found to be impor- doubtful whether all these combined constitute an tant contributions to the full deployment of other adequate explanation of the phenomena. Certain it hypotheses. The eloquent expositions of "pro- is, at least, that the measure of participation of phetic" characters at the hands of Agassiz were each must be determined before a satisfactory elu- profoundly suggestive and helpful in the explica- cidation can be reached. The full solution therefore tion of "undifferentiated"types in the hand of the involves not only the recognition of multiple par- evolutionary theory. ticipation but an estimate of the measure and So also the mutual conflicts of hypotheses whet mode of each participation. For this the simultane- the discriminative edge of each. The keenness of ous use of a full staff of working hypotheses is de- the analytic process advocates the closeness of dif- manded. The method of the single working hy- ferentiating criteria, and the sharpness of discrimi- pothesis or the predominant working hypothesis is nation is promoted by the codrdinate working of incompetent. several competitive hypotheses. In practice it is not always possible to give all Fertility in processes is also a natural sequence. hypotheses like places nor does the method con- Each hypothesis suggests its own criteria, its own template precisely equable treatment. In forming means of proof, its own method of developing the specific plans for field, office or laboratorywork it truth; and if a group of hypotheses encompass the may often be necessary to follow the lines of in- subject on all sides, the total outcome of means quiry suggested by some one hypothesis, rather and of methods is full and rich. than those of another. The favored hypothesis may The loyal pursuit of the method for a period of derive some advantagetherefrom or go to an earlier years leads to certain distinctive habits of mind death as the case may be, but this is rather a matter which deserve more than the passing notice which of executive detail than of principle. alone can be given them here. As a factor in educa- A special merit of the use of a full staff of tion the disciplinary value of the method is one of hypotheses codrdinately is that in the very nature prime importance. When faithfully followed for a of the case it invites thoroughness. The value of a sufficient time, it develops a mode of thought of working hypothesis lies largely in the significance its own kind which may be designated the habit of it gives to phenomena which might otherwise be parallel thought, or of complex thought. It meaningless and in the new lines of inquiry which is contra-distinguished from the linear order of spring from the suggestions called forth by the sig- thought which is necessarily cultivated in lan- nificance thus disclosed. Facts that are trivial in guage and mathematics because their modes are themselves are brought forth into importance by linear and successive. The procedure is complex the revelation of their bearings upon the hypoth- and largely simultaneously complex. The mind ap- esis and the elucidation sought through the hy- pears to become possessed of the power of simulta- pothesis. The phenomenal influence which the neous vision from different points of view. The Darwinian hypothesis has exerted upon the inves- power of viewing phenomena analytically and syn- tigations of the past two decades is a monumental thetically at the same time appearsto be gained. It illustration. But while a single working hypothesis is not altogether unlike the intellectual procedure may lead investigation very effectively along a in the study of a landscape. From every quarter of given line, it may in that very fact invite the ne- the broad area of the landscape there come into glect of other lines equally important. Very many the mind myriads of lines of potential intelligence biologists would doubtless be disposed today to which are received and coardinated simultane- cite the hypothesis of natural selection, extraordi- 354 HISTORICAL ESSAY

ously producing a complex impression which is re- on the part of the practitioner of this method to cordedand studied directly in its complexity. If the taciturnity. The remedy obviously lies in coordi- landscape is to be delineated in language it must nate literary work. be taken part by part in linear succession. An infelicity also seems to attend the use of the Over against the great value of this power of method with young students. It is far easier, and thinking in complexes there is an unavoidable dis- apparentlyin general more interesting, for those of advantage.No good thing is without its drawbacks. limited training and maturity to accept a simple It is obvious upon studious consideration that a interpretation or a single theory and to give it wide complex or parallel method of thought cannot be application, than to recognize several concurrent renderedinto verbal expression directly and imme- factors and to evaluate these as the true elucida- diately as it takes place. We cannot put into words tion often requires. Recalling again for illustration more than a single line of thought at the same the problem of the Great Lake basins, it is more to time, and even in that the orderof expression must the immature taste to be taught that these were be conformed to the idiosyncrasies of the language. scooped out by the mighty power of the great gla- Moreover the rate must be incalculably slower ciers than to be urged to conceive of three or more than the mental process. When the habit of com- great agencies working successively in part and si- plex or parallel thought is not highly developed multaneously in part and to endeavor to estimate there is usually a leading line of thought to which the fraction of the total results which was accom- the others are subordinate. Following this leading plished by each of these agencies. The complex and line the difficulty of expression does not rise to the quantitative do not fascinate the young student serious proportions. But when the method of si- as they do the veteran investigator. multaneous mental action along different lines is The studies of the geologist are peculiarly com- so highly developed that the thoughts running in plex. It is rare that his problem is a simple unitary different channels are nearly equivalent, there is phenomenon explicable by a single simple cause. an obvious embarrassment in making a selection Even when it happens to be so in a given instance, for verbal expression and there arises a disincli- or at a given stage of work, the subject is quite nation to make the attempt. Furthermorethe im- sure, if pursued broadly,to gradeinto some compli- possibility of expressing the mental operation in cation or undergo some transition. He must there- words leads to their disuse in the silent processes fore ever be on the alert for mutations and for the of thought and hence words and thoughts lose that insidious entrance of new factors. If therefore there close association which they are accustomed to are any advantagesin any field in being armed with maintain with those whose silent as well as spo- a full panoply of working hypotheses and in habit- ken thoughts predominantly run in linear verbal ually employing them, it is doubtless the field of courses. There is therefore a certain predisposition the geologist.