Philosophy of Science
Philosophy of Science
The phrase "philosophy of science" can be used most broadly to describe two different, though related, sorts of inquiry. On the one hand it can be used to describe the philosophy of particular sciences, such as the philosophy of physics, biology, or economics. On the other hand, it can be used to describe the study of epistemological issues in science more generally. Although an increasing majority of work in the philosophy of science is being done in the philosophy of particular sciences, it is this latter construal of the philosophy of science that remains the heart of the field and is the focus of this entry.
In a tradition that can be traced back to John Stuart Mill (1806–1873) and Francis Bacon (1561–1626), many have taken the scientific method to be inductive. An inductive inference is ampliative (i.e., the content of the conclusion goes beyond the content of the premises) and nondemonstrative (i.e., all true premises do not guarantee a true conclusion; at best they render the conclusion more probable). For example, suppose that one has observed a large number of mammals and every kind of mammal that one has observed has teeth; from this evidence one might make the inductive generalization that all mammals have teeth. It is possible, however, that the next mammal one observes (say, an anteater) might turn out not to have teeth. The fallibility of inductive inferences is often referred to as Hume's problem of induction, after the philosopher David Hume (1711–1776).
Carl Hempel (1905–1997) argues that the scientific method begins not with observations but with hypotheses. According to this hypothetico-deductive method one deduces certain observational predictions from the hypothesis and then rigorously tests them through further observation and experimentation. If the predictions are borne out, then the hypothesis is confirmed. Thus Hempel's method is still broadly inductive. Although the conclusion of an inductive argument is not certain, one would like to determine quantitatively how probable the conclusion is, given its premises (the evidence). The logical positivist Rudolf Carnap (1891–1970) sought to develop such a logic of confirmation. Other models of confirmation, such as Bayesian and bootstrapping models, are reviewed in John Earman's Testing Scientific Theories (1983).
Karl Popper (1902–1994) insists that the scientific method is deductive, not inductive. Observation always requires a prior point of view or problem. Like Hempel, Popper believes science begins with a bold hypothesis or conjecture. The way in which the scientist comes to the hypothesis (context of discovery) is irrelevant (e.g., it could come to the scientist in a dream); all that matters is the way in which it is tested (context of justification). Unlike Hempel, Popper does not think that hypotheses can be confirmed. If the observational prediction is borne out, deductively the scientist is unable to conclude anything (to conclude that the hypothesis is confirmed is to commit the deductive fallacy of affirming the consequent). If, however, the predictions are falsified, then, by the valid deductive inference modus tollens (if p then q, not q, therefore not p ) one can conclude that the hypothesis is falsified. Hence, Popper's method is known as falsificationism. According to Popper, the scientist should not seek to confirm theories but rather, refute them. A theory that has survived repeated attempts of falsification—especially in those cases where it has made risky predictions—has been corroborated, though not confirmed. On this view, a theory is demarcated as scientific if there are observational conditions under which one would be willing to reject the theory as falsified.
As a matter of historical fact, however, scientists typically do not abandon their theories in the face of falsifying evidence. Furthermore, in many cases it turns out to be sound scientific judgment to continue developing and modifying a theory in the face of recalcitrant evidence. In response to these sorts of difficulties, Popper's student, Imre Lakatos (1922–1974), developed a sophisticated falsificationism known as the "methodology of scientific research programs." For Lakatos, instead of evaluating an individual theory or modification of a theory as scientific or ad hoc, one should evaluate a whole series of theories developed over time. This series, called a research program, consists of a hard core, which defines the research program and is taken to be irrefutable, and a protective belt, which consists of auxiliary hypotheses and background assumptions to be modified in the face of falsifying data, thereby protecting the hard core. According to Lakatos, a research program is demarcated as scientific if it is progressive—that is, it continues to make new predictions that become corroborated. Once a research program ceases to make new corroborated predictions it becomes degenerative and its hard core should be abandoned.
Paul Feyerabend (1924–1994) was a close friend of Lakatos and also a student of Popper's. In his book Against Method (1978) he denies that there is such thing as the scientific method. He writes, "the idea of a fixed method, or a fixed theory of rationality, rests on too naïve a view . . . there is only one principle that can be defended under all circumstances. . . . It is the principle: anything goes " (pp. 27–28). Feyerabend's view is known as epistemological anarchism.
Scientific rationality and theory change
Beginning in the early 1960s there was a shift away from concerns about scientific methodology towards concerns about scientific change. This shift was in large part due to the publication in 1962 of Thomas Kuhn's (1922–1996) The Structure of Scientific Revolutions. Kuhn argues that the philosophy of science ought to be the product of a careful examination of the history of science. This involves recognizing the integrity of the science within its own time and not simply viewing it in relation to one's contemporary perspective. This new historiographical approach leads Kuhn to reject much of traditional philosophy of science: the confirmationist and falsificationist accounts of theory evaluation, the view that science is cumulative, the distinction between context of discovery and context of justification, and the idea of a crucial experiment.
Kuhn argues that science is characterized by three sorts of phases: pre-paradigm science, normal science, and revolutionary science. Central to understanding these phases is his notion of a paradigm, which he uses in two primary ways. First, he means an exemplar, a concrete problem solution or scientific achievement that serves as a model for solving other scientific problems (e.g., the planetary dynamics laid out in Isaac Newton's Principia ). Second, and more broadly, he means by paradigm a disciplinary matrix, which includes not only exemplars, but laws, definitions, metaphysical assumptions, and values (e.g., Newton's dynamical laws, the definitions of mass and space, and the mechanical philosophy). The paradigm determines what is to count as an acceptable scientific problem and an acceptable scientific solution. In the process of normal science, anomalies emerge that resist solution within the framework of the paradigm; if these anomalies persist and proliferate, they can lead to a state of crisis. Revolutionary science is described as "those noncumulative developmental episodes in which an older paradigm is replaced in whole or part by an incompatible new one" (p. 92). Kuhn refers to the pre- and post-revolutionary periods of normal science as incommensurable, and says that there is a sense in which scientists from different paradigms work in different worlds. Kuhn polemically refers to the conversion from the old to the new paradigm as being analogous to a Gestalt switch or religious conversion. Ian Barbour draws analogies between Kuhnian paradigms and religious paradigms in Religion and Science (1997).
In the 1969 postscript to The Structure of Scientific Revolutions and in the article "Objectivity, Value Judgment, and Theory Choice" (1977) Kuhn responds to charges that his account of science makes science irrational and leads to relativism. Against the charge of irrationality, Kuhn notes that values (such as predictive accuracy, simplicity, internal consistency, and coherence with neighboring theories) provide scientists with a shared basis for theory choice. Against the charge of relativism, Kuhn notes that ultimately paradigms are to be evaluated by their ability to set up and solve "puzzles." In this sense Kuhn does believe that there is objective progress in science: Newton solves more puzzles than Aristotle, and Albert Einstein more puzzles than Newton. What Kuhn rejects is realism, which claims that there is a coherent direction of ontological development and that science is getting closer to the truth.
Subsequent philosophers of science influenced by Kuhn developed different strands of his thought in different directions. Feyerabend, who developed an incommensurability thesis around the same time as Kuhn, came to later embrace the label of relativist. Others, such as Larry Laudan, sought to preserve the rationality of science against the threat perceived in Kuhn's holist picture of scientific change. According to Laudan, a closer look at the history of science shows not a wholesale exchange of one paradigm for another, but rather the components of the disciplinary matrix (e.g., methods, values, and ontology) being negotiated individually. Regarding theory choice he writes, "there is enough common ground between the rivals to engender hope of finding an 'Archimedean standpoint' which can rationally mediate the choice" (1984, p. 75). He calls this alternative view the reticulated model of scientific change.
Scientific realism versus antirealism
The labels realism and antirealism are each used to cover a wide spectrum of views. The main positions can be roughly distinguished by their answers to three questions: (1) Is there a mind-independent world? (2) What is our epistemic access to that world? (3) What is the aim of science? Realists (along with many antirealists) accept the existence of a mind-independent world. Those antirealists who deny this advocate some form of idealism. While realists tend to be optimistic about epistemic access to the world, antirealists argue in various ways that this optimism is unwarranted. Realists typically see the aim of science being truth, whereas antirealists argue the aim is something less.
At one end of the realist spectrum is naïve realism—the view that science is a perfect, undistorted mirror of the mind-independent world and that scientific theories are literally true. More sophisticated versions of realism, such as the view of Ernan McMullin, hold that realism means the long-term success of a scientific theory gives reason to believe that something like the entities and structures postulated by the theory actually exist (p. 26). According to McMullin, an important part of the aim of science is the development of fruitful metaphors. Many have argued for realism on the grounds that it provides the best explanation for the success of science; the widespread success of science would be "miracle" if scientific theories were not at least approximately true (Boyd 1984, Putnam 1975). Others argue that the proper question for realism is not whether some theory is true or approximately true, but whether some entity exists. According to Ian Hacking's entity realism, one can conclude, for example, that electrons exist because researchers experimentally build devices that use electrons to investigate other parts of nature. Between theory realism and entity realism is another view known as structural realism. This view, which John Worrall attributes to Henri Poincaré (1854–1912), affirms a mind-independent world but takes epistemic access to that world to be limited to its structural features. Thus, there is a continuity of structure across theory change despite radical changes in ontology. Although what is meant by structure is not entirely clear, in the physical sciences it is typically taken to be the structures expressed in the mathematical formalism of the theory.
Challenges to realism come from many sources and have led to a variety of antirealist views. Both Kuhn (1962) and Laudan (1981) argue that the history of science undermines realism. Kuhn's view can be classified as a form of instrumentalism, according to which scientific theories are merely useful instruments for making predictions and solving problems. Other antirealist views, such as Bas van Fraassen's constructive empiricism, come out of the empiricist tradition. According to van Fraassen (1980), science only aims to give theories that are empirically adequate and a theory is empirically adequate if what it says about observable things is true, that is, if it saves the phenomena (p. 12). On this view, one is not compelled to accept the existence of unobservable entities, such as electrons.
A third strand of antirealism, known as social constructivism, comes from sociology. The social constructivist seeks to understand scientific practice in the laboratory in a manner similar to an anthropologist seeking to understand a foreign culture. Social constructivists, such as David Bloor of the Edinburgh School, reject the philosophical understanding of knowledge as justified true belief, and instead take knowledge to be whatever is collectively endorsed by a particular group of people at a particular time (p. 5). This makes social constructivism a form of relativism. It is called constructivism because it takes scientific knowledge and facts to be constructed rather than discovered. Stronger and weaker versions of this view are obtained depending on whether this process of constructing scientific knowledge is, or is not, taken to be purely social. Arthur Fine, who argues that social constructivism has important methodological lessons for the philosophy of science (1996), himself rejects both realism and antirealism. Instead Fine advocates a minimalist position he calls the natural ontological attitude, which prescribes accepting the claims of science in the same way that one accepts the evidence from one's senses, without adding any additional claims such as "and it is really true" or "and it is only a useful fiction" (1986, p. 127).
Feminist philosophies of science
Since the 1970s, many feminist philosophers, historians, and scientists have been asking why there have traditionally been so few women in science and whether certain sexist, racist, or nationalist biases have shaped the practice and content of science. Detailed case studies, a representative sample of which can be found in Janet Kourany's The Gender of Science (2001), reveal many ways in which such biases have affected science. In response to these findings, many feminists sought to develop a new epistemology or philosophy of science. Following Sandra Harding (1986), feminist philosophies of science can be roughly divided into three traditions: feminist empiricism, feminist standpoint theory, and feminist postmodernism.
Helen Longino, whose work falls largely within the feminist empiricist tradition, introduces a view known as contextual empiricism. Longino sees empirical data as constraining, but nonetheless underdetermining, theory choice. This gap between theory and evidence is bridged by value-laden background assumptions belonging to a particular context. These contextual assumptions are one way biases can enter science. Longino criticizes traditional portrayals of the scientific method as individualistic. Instead she sees the objectivity of science being secured by its social character (e.g., peer review, replication of experiments, and an openness and responsiveness to criticism). She argues that diversity in science is important for making these often invisible assumptions explicit and open to criticism.
Harding takes this point about diversity in science a step further in her feminist standpoint theory. In contrast to empiricism, standpoint theory argues that the legitimacy of the knowledge claim depends on the social identity of the knower. Harding writes, "women's subjugated position provides the possibility of more complete and less perverse understandings. Feminism . . . can transform the perspective of women into a 'standpoint'—a morally and scientifically preferable grounding for our interpretations and explanations of nature" (p. 26). One standard criticism that Harding considers is whether there is such a thing as a "feminist standpoint" that cuts across all classes, races, and cultures.
Donna Haraway's work Simians, Cyborgs, and Women (1991) exemplifies the feminist postmodernism tradition. Haraway rejects the idea of single feminist standpoint and instead argues that all knowledge is locally situated. Like Longino, Haraway offers an alternative account of objectivity. She writes, "Feminist objectivity is about limited location and situated knowledge, not about transcendence and splitting of subject and object. In this way we might become answerable for what we learn how to see" (p. 190). Although Haraway's view shares some affinities with social constructivism, she explicitly rejects the label "relativist." She explains, "the alternative to relativism is not totalization and single vision. . . . [Rather, it is] partial, locatable, critical knowledges. . . . Relativism is a way of being nowhere while claiming to be everywhere equally. The 'equality' of positioning is a denial of responsibility and critical enquiry" (p. 191). Underlying many of these feminist philosophies is a central concern for the social and ethical implications of science.
Scientific explanations and laws
The most influential account of scientific explanation is Hempel's covering law model. On this model, explanations are understood as arguments in which the explanandum (the event, feature, or law to be explained) appears as the conclusion of an argument. The premises of the argument must contain at least one universal or statistical law used essentially in the derivation, and empirically verifiable statements describing particular facts or initial conditions. If the argument is deductive and involves a universal law, it is called a deductivenomological explanation ; if the argument is inductive and involves a statistical law, then it is called an inductive-statistical explanation. For example, suppose one wants to understand why an ice skater's angular velocity increases as she draws her arms in during a spin. The explanation would show that this event can be logically deduced from premises involving the law of conservation of angular momentum and statements such as her initial angular momentum was nonzero and her moment of inertia was reduced by drawing her arms in.
Several philosophers and historians have objected that Hempel's conditions are neither necessary nor sufficient for a scientific explanation. The most famous counterexamples fall into the categories of either irrelevance (although the event follows from the premises, as a matter of fact those premises are irrelevant to the explanation of the event) or symmetry (if the law involves a biconditional or equation then one can switch one of the premises with the conclusion and "explain" things such as why a flagpole is a certain height in terms of the length of its shadow). These sorts of problems have led philosophers largely to abandon Hempel's model and propose new alternatives. To handle the problems of irrelevance and symmetry, Wesley Salmon (1925–2001) introduces a causal model explanation, whereby to explain an event is to identify the causes of that event. Alternatively, van Fraassen in his pragmatic account of explanation embraces the possibility that the length of a shadow may explain the height of a pole. For van Fraassen (1980), an explanation is always relative to a particular context. Yet another model of explanation is provided by Philip Kitcher (1981), who understands explanation to be a unification of diverse phenomena by means of a common underlying structure or small number of processes. He sees Charles Darwin's theory of evolution as illustrating this model of explanation. The link between explanation and unification is challenged by Margaret Morrison in her book Unifying Scientific Theories (2000).
From reductionism to theoretical pluralism
Reductionism can be construed as a thesis about ontologies, laws, theories, linguistic expressions, or some combination of these. Considered as a relationship between scientific theories, it can be taken as a synchronic relation between two concurrent theories belonging to different levels of description or a diachronic relation between a historical predecessor theory and its successor. The classic formulation of theory reduction is due to the logical empiricist Ernest Nagel (1901–1985), who takes it to involve the logical derivation of one theory from another. More specifically, "a reduction is effected when the experimental laws of the secondary science . . . are shown to be the logical consequences of the theoretical assumptions . . . of the primary science" (p. 352). The standard example is the reduction of thermodynamics (secondary science) to statistical mechanics (primary science). In the physical sciences, reductionism is more often taken to be a correspondence between two theories under certain conditions, typically characterized by the limit of some quantity. As Thomas Nickles notes, this view is "best described by 'inverting' the usual concept of reduction, so that successors are said to reduce to their predecessors . . . under limiting operations" (p. 181). For example, special relativity is said to reduce to Newtonian mechanics in the limit of small velocities.
Challenges to reductionism have come from detailed case studies of the relations between particular scientific theories. One recurring challenge is known as the problem of multiple realizability. For example, in reducing Mendelian genetics to molecular biology, as Alexander Rosenberg points out in his 1989 "From Reductionism to Instrumentalism?",one discovers that a single Mendelian trait can be realized by a variety of molecular mechanisms, and furthermore, the same molecular mechanism can produce different Mendelian characteristics. Another set of challenges arises when the reducing theory is statistical (such as statistical mechanics or quantum mechanics) and the reduced theory is not as Lawrence Sklar indicates in this 1999 essay, "The Reduction (?) of Thermal Dynamics to Statistical Mechanics." These sorts of difficulties have led many to reject reductionism and instead argue for theoretical pluralism, or the so-called disunity of science. According to pluralism, each scientific theory has its own proper domain of applicability. In her book, The Dappled World (1999), Nancy Cartwright raises the possibility that "nature is governed in different domains by different systems of laws not necessarily related to each other in any systematic or uniform way" (p. 31). This view has been criticized on the grounds that it forfeits the benefits that come from examining inter-theoretic relations. The question of the unity or disunity of science remains a controversial topic.
See also Explanation; Philosophy of Science, History of; Positivism, Logical
barbour, ian. religion and science: historical and contemporary issues. new york: harpercollins, 1997.
bloor, david. knowledge and social imagery, 2nd edition. chicago: university of chicago press, 1991.
boyd, richard. "the current status of scientific realism." in scientific realism, ed. jarrett leplin. berkeley: university of california press, 1984.
carnap, rudolf. "on inductive logic." philosophy of science 12, no. 2 (1945): 72–97.
cartwright, nancy. the dappled world: a study of the boundaries of science. cambridge: cambridge university press, 1999.
earman, john, ed. testing scientific theories. minneapolis: university of minnesota press, 1983.
feyerabend, paul. against method: outline of an anarchistic theory of knowledge. london: verso, 1978.
fine, arthur. the shaky game: einstein, realism, and the quantum theory. chicago: university of chicago press, 1986.
fine, arthur. "science made up: constructivist sociology of scientific knowledge." in the disunity of science: boundaries, contexts, and power, ed. peter galison and david stump. stanford, calif.: stanford university press, 1996.
hacking, ian. representing and intervening: introductory topics in the philosophy of natural science. cambridge, uk: cambridge university press, 1983.
haraway, donna. simians, cyborgs, and women: the reinvention of nature. new york: routledge, 1991.
harding, sandra. the science question in feminism. ithaca, n.y.: cornell university press, 1986.
hempel, carl. aspects of scientific explanation and other essays in the philosophy of science. new york: free press, 1965.
hempel, carl. philosophy of natural science. upper saddle river, n.j.: prentice-hall, 1966.
hume, david. an enquiry concerning human understanding (1748). a critical edition, ed. tom l. beauchamp. new york: oxford university press, 2001.
kitcher, philip. "explanatory unification." philosophy of science 48 (1981): 507–531.
kourany, janet. the gender of science. upper saddle river, n.j.: prentice hall, 2001.
kuhn, thomas s. "objectivity, value judgment, and theory choice." in the essential tension: selected studies in scientific tradition and change. chicago: university of chicago press, 1977.
kuhn, thomas s. the structure of scientific revolutions, 3rd edition. chicago: university of chicago press, 1996.
lakatos, imre. "falsification and the methodology of scientific research programmes." in criticism and the growth of knowledge, ed. imre lakatos and alan musgrave. cambridge, uk: cambridge university press, 1970.
laudan, larry. "a confutation of convergent realism." philosophy of science 48 (1981): 19–49.
laudan, larry. science and values: the aims of science and their role in scientific debate. berkeley: university of california press, 1984.
longino, helen. science as social knowledge: values and objectivity in scientific inquiry. princeton, n.j.: princeton university press, 1990.
mcmullin, ernan. "a case for scientific realism." in scientific realism, ed. jarrett leplin. berkeley: university of california press, 1984.
morrison, margaret. unifying scientific theories: physical concepts and mathematical structures. cambridge, uk: cambridge university press, 2000.
nagel, ernest. the structure of science: problems in the logic of scientific explanation (1961). indianapolis, ind.: hackett, 1979.
nickles, thomas. "two concepts of intertheoretic reduction." the journal of philosophy 70, no. 7 (1973): 181–201.
popper, karl. the logic of scientific discovery. london: routledge, 1959.
popper, karl. conjectures and refutations: the growth of scientific knowledge. london: routledge, 1963.
putnam, hillary. mathematics, matter, and method, vol. 1. cambridge, uk: cambridge university press, 1975.
rosenberg, alexander. "from reductionism to instrumentalism?" in what the philosophy of biology is: essays dedicated to david hull, ed. michael ruse. dordrecht, netherlands: kluwer academic, 1989.
salmon, wesley. the foundations of scientific inference. pittsburgh, pa.: university of pittsburgh press, 1967.
salmon, wesley. scientific explanation and the causal structure of the world. princeton, n.j.: princeton university press, 1984.
sklar, lawrence. "the reduction (?) of thermodynamics to statistical mechanics." philosophical studies 95 (1999): 187–202.
van fraassen, bas. the scientific image. oxford: clarendon press, 1980.
worrall, john. "structural realism: the best of both worlds?" dialectica 43, nos. 1–2 (1989): 99–124.
Philosophy of Science
Philosophy of Science
Everyone takes for granted scientific progress in the natural sciences, but the topic raises controversy about the social sciences. Philosophy of science attempts to answer questions like: What makes some practice scientific? Why is science so successful? Why is there unevenness in the success of science? But it also attempts to explain what these questions might mean. Because social scientists so often deal with difficult and controversial topics—the role of religion in human welfare, tolerance for cultural difference, psychological bias and stereotyping—they routinely turn to philosophy of science for guidance.
Three systematic theories in the philosophy of science emerged at different times in the twentieth century— logical empiricism, scientific realism, and social constructivism. Each position is designed to explain apparent scientific development or progress. Logical empiricism—in many ways the dominant outlook among methodologically reflective psychologists and social scientists in the twentieth century—states that scientific knowledge is restricted to observable phenomena, and that theories are mere instruments for generating hypotheses to be tested by observation (for the modern version, see van Fraassen 1980). Scientific Realism, as elaborated by J. J. C. Smart (1968), Hilary Putnam (1975), and Richard Boyd (1984), is the alternative view that the best scientific theories should be understood as approximately accurate descriptions of a mind-independent reality, in both its observable (empirical) and unobservable (theoretical) aspects. Social constructivism (Kuhn 1962) is the doctrine that scientific knowledge depends so deeply on theory that it is best understood as the result of a construction rather than a discovery procedure. There are also complicated hybrids of these three views, such as Arthur Fine’s natural ontological attitude (Fine  1996).
Many social scientists accept another philosophical doctrine: philosophical naturalism. Naturalism is the doctrine that philosophy is methodologically and perhaps (meta)physically continuous with the natural sciences. According to this view, a philosophical doctrine about the social sciences gets tested and confirmed or disconfirmed in the same way hypotheses in the natural sciences do— by prediction and measurement.
Traditional methods in the social sciences did not emphasize prediction and measurement; the social sciences were typified by narrative methods (Martin and McIntyre 1994). Narrative methods typically describe the course of behavior by appealing to actors’ intentions and construct a story of how an effect came about. Narration has been a central method in social sciences such as anthropology and history, where a chief goal is to describe the course of behavior in a particular cultural setting or epoch. The anthropological writings of Bronislaw Malinowski, such as Coral Gardens and their Magic (1935) or Argonauts of the Western Pacific (1922), and E. E. Evans-Pritchard’s Witchcraft, Oracles, and Magic Among the Azande (1937) are prime examples of the narrative method, as are Garrett Mattingly’s Armada (1959) and classics by Oswald Spengler (The Decline of the West ) and Edward Gibbon (The Decline and Fall of the Roman Empire [1776–1788]) in history. Most of the social sciences have developed clever methods to expand the methodological power of their fields (Joseph Henrich, Richard Thaler, Steven Leavitt), but much contemporary work in anthropology, history, and the clinical arm of psychology remains in the narrative tradition. Because such appeals would be wholly inappropriate for explanations in physics, chemistry, and biology, narrative methods are unique to the social sciences.
Measurement has always played an important role in the natural sciences, from Isaac Newton’s (1642–1727) laws of motion to Robert Boyle’s (1627–1691) law based on the measurement of heat, pressure, and the volume of an enclosed container. The social sciences make room for another method of inquiry that uses the tools of statistical measurement. Sociology uses such techniques as causal modeling. Economics also uses causal models, as well as a wide range of regression techniques. Psychology uses experimental design and the standard array of statistical tests: analysis of variance, t-tests, multivariate analysis, and so on.
So there are some differences between the methods of inquiry in the natural and social sciences, but there is also broad agreement about such topics as the importance of measurement (Trout 1998). Other shared norms place a premium on observability and testability, however unevenly applied. Testability—at least in the experimental sense—is not a practical requirement in evolutionary theory, for example. But the testability criterion does represent an ideal. Scientific assertions, and the theories that recommend them, are testable: There is a well-established generalization in psychology that words that are more frequent in a language are identified more quickly, and in economics that markets clear, to cite just two examples.
Early philosophy of science made testability, verifiability, and falsifiability a necessary condition for scientific status—and this was a mistake. The philosophy of science has a regrettable record when it dictates the standards of science; it performs much better when it extracts lessons from reliable scientific practice. So, while it makes sense to ask what the ideal norms are in the philosophy of science, the intellectual trend runs in the opposite direction: What features of the practice of science are so successful that they deserve to be elevated to a normative standard in the philosophy of science?
The logical positivists, following their interpretation of David Hume’s (1711–1776) dictum that all nontautological knowledge arises from experience, attempted to formulate a way of distinguishing claims that have empirical content—and thus could be adjudicated objectively by appeal to observable evidence—from those that do not have empirical content. Fortified by the meaning-atomism of the early Rudolph Carnap (1891–1970) and Ludwig Wittgenstein (1889–1951), the logical positivists hoped to catch and identify those claims that are metaphysical, that describe conditions that are beyond experience and thus empirically meaningless. The principle of verifiability was formulated in a variety of ways (for the scientific rationale that was in the air at the time, see Schlick [1932–1933] 1991). But ultimately, the principle of verifiability could not accommodate the actual, epistemically reliable practices of scientists; indeed, it contradicted those practices. Scientists made patently metaphysical assertions, and no amount of “rational reconstruction” could cleanse that vocabulary. Indeed, it is now an adage that, when reliable scientific practice meets exacting philosophical principle, philosophical principle either yields or faces embarrassment. (For an excellent introductory overview of twentieth-century philosophy of science, see Godfrey-Smith 2003.)
With similar concerns about identifying and diagnosing intellectually disreputable practices, Karl Popper (1902–1994) proposed the principle of falsifiability—that the hallmark of a scientific theory is that it is falsifiable under empirical testing. Put more formally: a theory is potentially scientific if and only if there are possible observations that would falsify or refute it. According to Popper’s vision, that is what scientists do—they make conjectures and attempt to refute them.
Although Popper’s proposal proved enormously influential, it turned out that neither verifiability nor falsifiability captured the practices of science. Scientists attempt to confirm and not refute their hypotheses. Moreover, their expressions are not meaningless without an experimental test to verify their hypotheses. In addition, Popper’s notion of corroboration—the tendency of a hypothesis to survive repeated attempts at refutation— was too weak to capture the scientific practice of confirmation.
Thomas Kuhn (1922–1996) was a classic social constructivist. He argued that science is not progressive or even continuous, but is typified by a series of “ruptures” in the normal process of science. These ruptures, produced by scientific revolutions, make theories and the items they refer to “incommensurable,” to use Kuhn’s term. Scientists separated by fundamentally different theories “see different worlds.”
Imre Lakatos’s (1922–1974) philosophy of science deemed Kuhn’s view too much of a concession to holism and relativism. Departing from Carnap’s and Popper’s focus on testing individual hypotheses, however, Lakatos took the entire theory as the unit of analysis. When confronted with falsifying evidence, scientists will protect the core of their favored theory from refutation by introducing auxiliary hypotheses. In response to Kuhn, Lakatos argued that a progressive (rather than stagnating) research program reflects growth, as well as findings that are novel and surprising.
Social scientists have long desired to extract information from catalogs of demographic data. But the reigning philosophical and social science standards in hypothesis testing had prohibited the search for correlations after the fact; all hypothesis tests have to be a priori. With new, powerful data-storing capacities, social scientists have developed sophisticated post hoc search and “data-mining” methods (Spirtes et al. 2001). These methods have allowed the social sciences to extract causal information from correlation data about important social issues, such as the efficacy of the Head Start program for low-income children.
Philosophy of science has built its theories—such as realism, empiricism, and constructivism—not just on methodological success but also on historical evidence of scientific progress. This evidence consists of case studies, often casually assembled and intuitively interpreted. But philosophers of science have been alert to the findings in the psychology of problem solving establishing the inaccuracy of intuitive judgments. Accordingly, philosophy of science has adopted a more quantitative focus, one that replaces narratives with models, and stories about episodes with data.
Earlier antimetaphysical philosophy of science was profoundly suspicious of causal notions. But with the demise of antimetaphysical philosophy of science, the causal properties of scientific items have received close attention. Moreover, individual causal properties—like being wealthy, a Democrat, or a member of a union—can be involved in different disciplines. These causal properties produce not only unified explanations of phenomena, but also a more unified image of the world. In a roundabout way, then, contemporary philosophy of science, by focusing on causal properties generally rather than observable properties specifically, completed the unity of science program begun by the logical positivists.
The logical positivist model was one of unification by reduction, of higher-level to lower-level sciences. According to this view, sociology and economics are reducible to psychology, psychology is reducible to biology, biology to chemistry, and chemistry to physics. But the course actually taken by the social sciences has not confirmed this logical positivist account. Instead, adjacent portions of the disciplines appear to get integrated rather than assimilated, reduced, or eliminated. If anything, the social sciences have undergone a proliferation of durable specialization, a model of unification confirmed by the increasing interdisciplinary activity in these disciplines.
SEE ALSO Revolutions, Scientific
Ayer, Alfred Jules. 1946. Language, Truth, and Logic. 2nd ed. London: Gollancz.
Boyd, Richard. 1984. On the Current Status of the Issue of Scientific Realism. Erkenntnis 19 (1–3): 45–90. Reprinted in Boyd, Gasper, and Trout (1991).
Boyd, Richard, Philip Gasper, and J. D. Trout, eds. 1991. The Philosophy of Science. Cambridge, MA: MIT Press.
Fine, Arthur.  1996. The Shaky Game: Einstein, Realism, and the Quantum Theory. 2nd ed. Chicago: University of Chicago Press.
Godfrey-Smith, Peter. 2003. Theory and Reality: An Introduction to the Philosophy of Science. Chicago: University of Chicago Press.
Henrich, Joseph, Richard McElreath, Abigail Barr, et al. 2006. Costly Punishment Across Human Societies. Science 312: 1767–1770.
Kuhn, Thomas. S. 1962. Structure of Scientific Revolutions. Chicago: University of Chicago Press.
Lakatos, Imre. 1976. Proofs and Refutations: The Logic of Mathematical Discovery. Cambridge, U.K.: Cambridge University Press.
Leavitt, Steven, and Stephen Dubner. 2005. Freakonomics: A Rogue Economist Explores the Hidden Side of Everything. New York: Morrow.
Martin, Michael, and Lee McIntyre, eds. 1994. Readings in the Philosophy of Social Science. Cambridge, MA: MIT Press.
Mattingly, Garrett. 1959. The Armada. Boston: Houghton Mifflin.
Popper, Karl. 1962. Conjectures and Refutations: The Growth of Scientific Knowledge. New York: Basic Books.
Putnam, Hilary. 1975. Explanation and Reference. In Philosophical Papers. Vol. 2: Mind, Language, and Reality, 196–214. New York: Cambridge University Press.
Schlick, Moritz. [1932–1933] 1991. Positivism and Realism. In The Philosophy of Science, eds. Richard Boyd, Philip Gasper, and J. D. Trout, 37–55. Cambridge, MA: MIT Press.
Smart, J. J. C. 1968. Between Science and Philosophy: An Introduction to the Philosophy of Science. New York: Random House.
Spengler, Oswald. 1928. The Decline of the West. Vol. 2. Trans. Charles Francis Atkinson. New York: Knopf.
Spirtes, Peter, Clark Glymour, and Richard Scheines. 2001. Causation, Prediction, and Search. 2nd ed. Cambridge, MA: MIT Press.
Thaler, Richard. 1991. The Winner’s Curse: Paradoxes and Anomalies of Economic Life. New York: Free Press.
Trout, J. D. 1998. Measuring the Intentional World: Realism, Naturalism, and Quantitative Methods in the Social and Behavioral Sciences. New York: Oxford University Press.
van Fraassen, Bas. 1980. The Scientific Image. Oxford: Clarendon.
J. D. Trout