After an introduction to the general notion of relativism in philosophy, the chapter considers an approach to the study of scientific inquiry that is explicit in its commitment to relativism: the strong program in the sociology of knowledge. According to the strong program, which purports to give a social scientific account of science, scientific knowledge is not so much discovered as constructed by social dynamics that produce scientific consensus. The limits of such an account are explored by discussing both sociological and anthropological approaches. The social constructionist account has been applied to Robert Boyle’s experiments with an air-pump and the criticisms directed against them. Applying similar ideas to physicists’ attempts to detect gravity waves has led to the formulation of a problem known as the experimenters’ regress. Through such cases, the chapter sees how defenders of social constructionist accounts draw upon both history and social scientific investigations of current science. The chapter then surveys philosophical and historical criticisms of this approach.

In the first of two chapters on probability in scientific inquiry, the basic ideas of probability theory are introduced through examples involving games of chance. The chapter then focuses on the Bayesian approach to probability, which adopts the stance that probabilities should be understood as expressions about the degrees of belief. The Bayesian approach as a general framework for probability is explained through examples involving betting that extend beyond games of chance, which also allows the introduction of the idea of probabilistic coherence as a condition of rational partial belief. We are then finally ready for Bayes’s theorem, a theorem of the probability calculus that plays a central role in the Bayesian account of learning from evidence. That account is illustrated with a historically motivated example from the history of paleontology. The chapter considers objections to the Bayesian approach and the resources Bayesians may draw on for answering those objections.

Scientific realists defend the proposition that successful scientific theories in the mature sciences should be regarded as at least approximately true because that provides the best explanation of the fact that scientists use such theories successfully. Two important types of arguments against scientific realism are then considered. The historical argument appeals to the fact that seemingly successful theories have in the past turned out to be not even approximately true. The empiricist argument holds that because scientific realists believe claims about things that can never be observed, they violate the scientific commitment to subject claims to empirical assessment. Responses on behalf of scientific realism are considered. The chapter concludes by surveying engagements with realism in science that depart from the dialectic just sketched. These include considerations based on experimentation and experimental practice, varieties of structural realism, and perspectival realism.

This chapter examines how the history of science became a resource for the development and defense of important alternatives to logical empiricist views of scientific theory and the growth of scientific knowledge. The chapter also examines the different meanings attached to scientific paradigms in Thomas Kuhn’s account of scientific change and different notions of incommensurability implicated in his account. Like other postpositivist thinkers, Kuhn rejects the logical empiricist idea of separating observational and theoretical language, arguing instead that observation is theory-laden. The history of science plays an important but distinct role in Imre Lakatos’s methodology of scientific research programs, which aims to represent the rationality of scientific thought. The chapter concludes by examining Paul Feyerabend’s epistemological anarchism, which appears to cast doubt on the prospects for providing a systematic account of scientific rationality. Are Feyerabend’s views as extreme as their expression suggests, or is there another way to understand his provocations?

This chapter surveys some of the many types of models used in science, and some of the many ways scientists use models. Of particular interest for our purposes are the relationships between models and other aspects of scientific inquiry, such as data, experiments, and theories. Our discussion shows important ways in which modeling can be thought of as a distinct and autonomous scientific activity, but always models can be crucial for making use of data and theories and for performing experiments. The growing reliance on simulation models has raised new and important questions about the kind of knowledge gained by simulations and the relationship between simulation and experimentation. Is it important to distinguish between simulation and experimentation, and if so, why?

The concepts of inductive and deductive inference are introduced and contrasted. An artificial example is used to emphasize the logical structure of the problem of induction. To see how the problem of induction relates (and also does not relate) to a real episode of experimental inquiry, this chapter considers the case of Isaac Newton’s optical experiments using prisms to investigate the refraction of light. Although Newton did not concern himself with the problem of induction as philosophers now understand it, he used experimental strategies designed to address possible errors in the conclusions about light that he drew from his observations.

This chapter surveys influential ideas about scientific explanation. The idea that scientific explanation is a matter of logical deduction from scientific laws has played an important role both as the basis for positive accounts of scientific explanation and as a target of critical arguments spurring the investigation of alternative views. The chapter reviews some of the reasons in favor holding such a covering-law view of explanation and then turn to some alternatives. The chapter also considers a pragmatically oriented account of the act of explaining. Another alternative focuses on the idea that explanations unify phenomena, showing how seemingly different things are manifestations of a single truth about nature. Several approaches emphasize the way explanations indicate what causes something to happen, whether by reference to a process, a possible manipulation, or a mechanism.

The chapter reviews an approach to the development of a ‘scientific philosophy’ that developed in the early decades of the twentieth century in Central Europe. Logical empiricists combined an interest in using the resources of formal logic and an empiricist orientation to propose ways of distinguishing meaningful scientific discourse from what they regarded as cognitively meaningless metaphysical statements. In so doing, they articulated important and influential ideas about how to characterize the relationship between observations serving as evidence and the theories for which they are relevant. The chapter also examines their assumptions about the nature and structure of physical theories and how those shaped efforts such as Rudolf Carnap’s development of a theory purporting to quantify how much a particular body of evidence confirms a particular theory.

One philosophical approach that directly responds to the problem of induction is falsificationism, first proposed by Karl Popper. This chapter examines how falsificationists propose to account for the growth of scientific knowledge without appealing to inductive reasoning. Their approach relies on attempts to falsify general hypotheses through experiments and observations. Additional logical concepts are introduced in this chapter to facilitate the logical analysis of such falsification. The concept of corroboration, central to the falsificationist view, is introduced. The apparatus of falsificationism is applied to the example of Newton’s optical experiments introduced in Chapter 1. Finally, falsificationism is discussed in relation to conventionalism, a philosophical idea that in some ways falsificationism attacks and in other ways exemplifies.

In practice, much of statistical reasoning in science relies on probabilities subject to interpretation as relative frequencies. This chapter explains how probability can be understood in terms of relative frequencies and the uses scientists and philosophers have devised for frequentist probabilities. Particularly prominent in those uses are error probabilities associated with particular approaches to hypothesis testing. The approaches pioneered by Ronald Fisher and by Jerzy Neyman and Egon Pearson are outlined and explained through examples. The chapter then explores the error-statistical philosophy advocated by Deborah Mayo as a general framework for thinking about how we learn from empirical data. The error-statistical approach utilizes a frequentist framework for probabilities to articulate a view of severe testingof hypotheses as the means by which scientists increase experimental knowledge. Error statistics represents an important alternative to Bayesian approaches to scientific inquiry, and this chapter considers its prospects and challenges.