Researchers selected this weird and wonderful assortment from tens of millions of possibilities because they have common attributes as well as unique characteristics. They're practical: A model must be cheap and plentiful; be inexpensive to house; be straightforward to propagate; have short gestation periods that produce large numbers of offspring; be easy to manipulate in the lab; and boast a fairly small and (relatively) uncomplicated genome. This type of tractability is a feature of all well-used models.At one level, the reliance of biomedical researchers on standardized creatures for experimentation is mere practical necessity. Biological materials are, by their nature, variable and complex; life scientists have sought to control the variability they face by selecting out and standardizing particular experimental subjects. Yet these organisms, no matter how standardized they become as laboratory instruments, maintain an independent existence in a contingent world. They are not models in the traditional sense -- they are not smaller versions of humans, and they do not exactly replicate our experiences or diseases. Unlike the idealized representational models characteristically featured in the history of the exact sciences, in which the model (e.g., the Bohr atom) has been supposed to mirror a natural system (hydrogen) by embodying the mathematical laws and structure from which the behavior of the system can be deduced, model systems maintain their own autonomy and specificity. That is, model systems do not directly represent humans as models of them. Rather, they serve as exemplars or analogues that are probed and manipulated in the search for generic (and genetic) relationships. They serve as models for human attributes. The use of standardized organisms in biomedicine is part of a broader model-systems approach in the life sciences that includes the investigation of a far wider range of entities, from specific proteins (e.g., hemoglobin) to particular lakes (e.g., Linsley Pond in Connecticut), and whose utility in producing general knowledge relies on the routine use of analogies to other examples and entities.These distinctions between representational and representative functions, between models of and models for, have proven quite useful in discussing the characteristics of model systems. We suggest that insofar as similar objects inhabit spaces far beyond biology laboratories, the same distinctions extend to other areas, areas where relations of similarity rather than deduction have grounded claims to generality and where specificity has been a resource rather than a problem. Many fields have developed canonical examples that have played something like the role of model systems, which serve not only as points of reference and as illustrations of general principles or values but also as sites of continued investigation and reinterpretation. What we here call model objects of this sort in this volume include Athenian democracy in political theory, the ritual in anthropology, and the so-called Prisoner's Dilemma in game theory. Through what processes do particular organisms, cases, materials, or texts become foundational to their fields? How do they serve a classificatory function for the organization of knowledge, whether it is in a biology laboratory or an art museum? When does the specificity or idiosyncrasy of an example threaten its utility?Examining the pursuit of knowledge organized around exemplars rather than around fundamental laws, we aim to reopen the old question of the relation between the human sciences and the natural sciences. In the nineteenth century, the question was cast in terms of the relation between the generalizing lawlike sciences (nomothetic, in the canonical formulation of Wilhelm Windelband) and the particularizing sciences (idiographic), where lawlike referred to the universal laws of physics as the ideal of science. It is no longer the case, however, that universal laws either do or can serve as a model for all science, even natural science. This has become most apparent with the emergence of biology in the past thirty years as the so-called science of the future. It is not clear that there are any high-level laws in biology, in the sense of predictive laws that determine the future behavior of a biological system (except perhaps in evolutionary theory); we will not be concerned with whether such laws may emerge. Instead, we want to show how the model-systems approach so pervasive in biology compares with the use of cases, exemplars, and related methods in other fields. Interestingly, it appears that many of these approaches grew up in response to the challenge of producing something like lawlike knowledge in disciplines in which laws seemed incapable of capturing the specificity and complexity of organisms, geological processes, or human productions. If the result has not been laws, it has nevertheless been reliable systematic knowledge. Thus, our title: Science without Laws.A model system in biology refers to an organism, object, or process selected for intensive research as an exemplar of a widely observed feature of life (or disease). The traditional contrast between laboratory physiology and natural history within nineteenth-century biology provides some background to the emergence of model systems in twentieth-century laboratories. Of the two traditions, it was experimental physiology that most closely emulated the ideals of the physical sciences. Nineteenth-century physiologists used animal models such as frogs because of their accessibility for experimental manipulation, and they aimed to produce universalistic knowledge through reductionist (and often instrument-based) approaches. This mechanistic approach treated the organism itself as a workshop, but there was considerable debate as to whether vital phenomena could be completely reduced to physico-chemical principles or laws. Hermann Helmholtz, Emil Du Bois-Reymond, Carl Ludwig, and others strove to reduce physiology to the physics of atoms and forces in their attempts to show how laws (such as the conservation of energy) might account for processes in both living and non-living materials. Not all physiologists agreed with the aims of this "organic physics," however. Claude Bernard cautioned that biology might borrow methods and instruments, but not theory, from physics and chemistry. In fact, radical reduction soon went bankrupt and even its most committed adherents had to redirect their energies. Ludwig, who in the early 1850s had asserted that "physiology is nothing other than applied physics," had also at the same time expressed his "hope someday to work with a capable clinician or pathological anatomist . . . and together with him experimentally reproduce the conditions of disease. . . . It ought to be possible to generate innumerable illnesses similar to those found in man." This is the expansive program he took up with Karl Wunderlich when he moved to Leipzig in 1871 to establish the new physiology laboratory there. It is also the program that turned increasingly to animal models rather than physical laws for insight into biological processes. The turn to model systems in the twentieth century resulted from the conjunction of this experimental tradition with a new industrial infrastructure. Accompanying the increasing mass-production of scientific materials and equipment was a narrowing of the number of organisms intensively studied and also the commodification of many of the laboratory's experimental inhabitants. The ascendance of the fruit fly in genetics research grew out of T. H. Morgan's laboratory in the 1910s and 1920s; during the same time-period maize became a dominant organism for plant genetics, particularly cytogenetics. By mid-century, inbred strains of mice were widely used not only to understand mammalian genetics, but also for investigations of cancer and other human afflictions. And the molecular emphasis of postwar biology grew out of the intensive study of a handful of microbes, especially yeast (Saccharomyces cerevissiae), bacteria (most notably E. coli) and viruses (particularly the bacteriophages and tobacco mosaic virus). Since the 1960s, scientists have domesticated new animals for research, such as the nematode worm and the zebrafish. Each of these model organisms has a unique history in the laboratory, with particular physical features and experimental advantages. Yet each of these model systems has also gained ground by virtue of its historically acquired prominence within a field of study. Indeed, model systems exhibit a self-reinforcing quality: the more the model system is studied, and the greater the number of perspectives from which it is understood, the more it becomes established as a model system. Even for the many biologists who do not study one of the canonical model organisms, these systems tend to serve as benchmarks and methodological guides when they turn to other organisms and objects as researchers.Beginning with model systems as understood in recent biology, this book compares the scope and function of model objects in domains as diverse as geology and history, attending to differences between fields as well as to epistemological commonalities. What distinguishes this collection of essays from other studies of models in science is both its attention to model systems as concrete autonomous objects and the breadth of disciplines it addresses. Traditional approaches to scientific models derive from the philosophy of science, especially the philosophy of theoretical physics, in which the model is presented as a precise, usually mathematical, representation of the phenomenon in question. But this picture of mathematical models, valorized by physicists and philosophers, is highly idealized. As Nancy Cartwright has argued, the models mobilized by physical scientists to illustrate their theories are far from real-world situations; they are, rather, "nomological machines" that manufacture universal laws of nature by providing the kind of simplified mechanical or mathematical evidence that could not be found in "nature." Not only the character of the model but also the presumption that the exact sciences should provide the basis for understanding general scientific method or rationality has recently come under question. To Ian Hacking's enumeration of six "styles of reasoning" that characterize the sciences, for example, John Forrester has proposed "reasoning by cases" as a seventh scientific method, widely used not only in the human (and biological) sciences, but also in law, medicine, and ethics. Case-based reasoning relies on relations of similarity rather than on conventional reductionism and treats specificity as a resource, not a problem. The essays in this book attend to case based modes of inquiry usually neglected by historians and philosophers of science, demonstrating that their epistemological practices and patterns extend far beyond the boundaries of science.Angela N. H. Creager, Elizabeth Lunbeck, and M. Norton Wise
mail@the-scientist.comThis excerpt is from Science Without Laws: Model Systems, Cases, Exemplary Narratives, Duke University Press (2007), edited by Angela N. H. Creager, Elizabeth Lunbeck, and M. Norton Wise.Links within this article:Model organisms for biomedical research http://www.nih.gov/science/models/Model Organisms, The Scientsist http://www.the-scientist.com/supplement/2003-6-2/Evelyn Fox Keller, "Models of and Models for: Theory and Practice in Contemporary Biology," Philosophy of Science 67 (2000): S72?S86. http://links.jstor.org/A. Gawrylewski, "The trouble with animal models," The Scientist, July 2007. http://www.the-scientist.com/article/display/53306/Wilhelm Windelband, "Geschichte und Naturwissenschaft," in Präludien: Aufsätze und Reden zur Philosophie und ihrer Geschichte (Tübingen, Germany: Mohr, 1915)Timothy Lenoir, "Science for the Clinic: Science Policy and the Formation of Carl Ludwig's Institute in Leipzig," in Instituting Science: The Cultural Production of Scientific Disciplines (Stanford: Stanford University Press, 1997), 113, 127. http://www.sup.org/book.cgi?book_id=2642%202925Robert E. Kohler, Lords of the Fly: Drosophila Genetics and the Experimental Life (Chicago: University of Chicago Press, 1994). http://www.press.uchicago.edu/cgi-bin/hfs.cgi/00/12521.ctlMary S. Morgan and Margaret Morrison, eds., Models as Mediators: Perspectives on Natural and Social Science (Cambridge: Cambridge University Press, 1999). http://tinyurl.com/24jgfhNancy Cartwright, "Nomological Machines and the Laws They Produce," in The Dappled World: A Study of the Boundaries of Science (Cambridge: Cambridge University Press, 1999), 49-74. http://www.cup.cam.ac.uk/John Forrester, "If p, Then What? Thinking in Cases," History of the Human Sciences 9 (1996). http://hhs.sagepub.com/cgi/content/citation/9/3/1Science Without Laws: Model Systems, Cases, Exemplary Narratives http://www.dukeupress.edu/
