Thinking of Biology608 BioScience • July/August 2007 / Vol. 57 No. 7 www.biosciencemag.orgFor more than two decades, there has been sustainedcriticism of the appropriateness of using methods that relysolely on null-hypothesis testing for observational studies inscience (e.g., Carver 1978, McBride et al. 1993, Anderson etal. 2000, Wade 2000, Johnson 2002). The disciplines of psy-chology, wildlife biology, and statistics have been in the fore-front of this conflict between two qualitatively differentinferential paradigms: model-selection methods, based on in-formation theory, and null-hypothesis testing, based on afrequentist approach. But many other areas of biology andecology have been implicated, including molecular biology,systematics, physical geography, medicine, and epidemiology(Johnson and Omland 2004). Perhaps this is because all thesefields readily provide case studies in which multiple causativefactors lead to real-world complexity that is difficult to reduceto a single, isolated mechanism.Strong proponents of the model-selection paradigm havedecried the use of null-hypothesis testing as outdated, andsome have colorfully suggested that the practice of reportingP values should be “euthanized” on philosophical grounds(Anderson and Burnham 2002). Others have taken an equiv-ocal stance, suggesting that the two inferential paradigmsprovide complementary tools for the investigator, and that hy-pothesis testing should be retained for manipulative experi-mental design (e.g., Johnson and Omland 2004). Stephens andcolleagues (2005) proposed that it may be more profitable todistinguish between studies of univariate causality, in whichnull-hypothesis testing may be sufficient, and multivariatecausality, in which model selection offers clear advantages (butsee Lukacs et al. 2007).Here we attempt to clarify some of the philosophical ter-rain relevant to this debate by discussing one of the keyphilosophical underpinnings of model selection. This is theconcept of the method of multiple working hypotheses(MMWH), as described by the geologist T. C. Chamberlin in1890, and later referred to by Platt (1964) in his notion of“strong inference.” Although the term has become almostmainstream in ecology, we contend that the core meaning ofChamberlin’s conceptualization has often been forgotten ormisinterpreted over time, and that this needs rectification. Forinstance, a common mistake is to equate the MMWH withthe method of developing alternative hypotheses. Yet sys-tematic application of the latter method occurred at least asearly as Francis Bacon (1620), whereas the former is qualita-tively different in construction and was intended by Cham-berlin to serve as a complement to the formal, “pure,” orclassic analytic method. Here we first describe the MMWHin general terms. Then we discuss its applicability to method-ologies that not only allow (or require) the simultaneous appraisal of more than one hypothesis but explicitly accom-modate various situations in which several hypotheses are simultaneously true.Louis P. Elliott (e-mail: [email protected]) was at the School for Environmental Research at Charles Darwin University in Australia whenthis essay was prepared; he is now with the Northern Territory Departmentof Natural Resources, Environment, and the Arts in Australia. Barry W. Brook (e-mail: [email protected]) is the director of the Research Institutefor Climate Change and Sustainability, School of Earth and EnvironmentalSciences, University of Adelaide, South Australia 5005, Australia. © 2007American Institute of Biological Sciences.Revisiting Chamberlin: MultipleWorking Hypotheses for the21st CenturyLOUIS P. ELLIOTT AND BARRY W. BROOKThe method of multiple working hypotheses, developed by the 19th-century geologist T. C. Chamberlin, is an important philosophical contributionto the domain of hypothesis construction in science. Indeed, the concept is particularly pertinent to recent debate over the relative merits of two different statistical paradigms: null hypothesis testing and model selection. The theoretical foundations of model selection are often poorly understoodby practitioners of null hypothesis testing, and even many proponents of Chamberlin’s method may not fully appreciate its historical basis.We contend that the core of Chamberlin’s message, communicated over a century ago, has often been forgotten or misrepresented. Therefore, we revisit his ideas in light of modern developments. The original source has great value to contemporary ecology and many related disciplines, communicatingthoughtful consideration of both complexity and causality and providing hard-earned wisdom applicable to this new age of uncertainty.Keywords: Bayesian statistics, hypothesis testing, model selection, philosophy of science, statistical significancewww.biosciencemag.org July/August 2007 / Vol. 57 No. 7 • BioScience 609The method of multiple working hypothesesThe concept of the MMWH was advocated over a century agoby the geologist Thomas Chamberlin (1890) in a paper thatwas later reprinted in Science—a testament to the perceivedimportance of its content. “With this method,” Chamberlinwrote,“the dangers of parental affection for a favorite theorycan be circumvented”(Chamberlin 1890). Chamberlin’s con-cerns have a timeless quality that makes his prose lucid andrelevant even today. He contrasted the MMWH with themethods of the “ruling hypothesis” and the “single workinghypothesis,” and contended that the ruling hypothesis is theworse of the latter two. This is because investigators’ affectionor loyalty to a theory may lead them to collect evidence to support only the ruling theory, and not sufficiently consideralternative explanations. Chamberlin also criticized the single-working-hypothesis approach, said to be the methodof the day: “Under the working hypothesis, the facts aresought for the purpose of ultimate induction and demon-stration, the hypothesis being but a means for the moreready...arrangement and preservation of material for the final induction” (Chamberlin 1890).The amendment that Chamberlin advocated is one familiar to all practitioners of science. However, like muchcogent advice, it is easier to follow in theory than in prac-tice: “[What is required] is to bring up into view every ra-tional explanation of new phenomena, and to develop everytenable hypothesis respecting their cause and history”(Chamberlin 1890). This