Pathways to Knowledge is a collection of recent articles, all but one recently published. Gathered together, they show the fecundity of Alvin Goldman's reliabilist epistemology, addressing topics ranging from a priori knowledge to the epistemic status of introspection to the current state of social epistemology. This review will focus on those ideas of most interest to philosophers of science.
Goldman espouses an “externalist” reliabilism that easily coheres with both naturalistic epistemology and evolutionary psychology. Consistent and careful application of reliabilism shows that naturalism is not inconsistent with a priori warrant since (some or all) intramental cognitive processes may be reliable. Thus construed, a priori knowledge is neither certain nor unrevisable. So a priori knowledge does not have privileged epistemic status, as was assumed from Descartes on. To the extent that it is reliable, we may thank the forces of evolution on cognition.
Goldman's reliabilism entails that the ultimate epistemic virtue is truth (or at least accuracy). This view is called veritism. Goldman does not simply stipulate this, but attempts to give evidence for it by considering other suggested epistemic virtues such as coherence, and showing that veritism underlies these virtues. Goldman has some particularly intriguing things to say about epistemic virtues in science, such as generality and simplicity. He does not think that they “reduce” to truth but claims that they specify the kinds of truth that scientists are interested in. In this sense, he allows that science is responsive to interests. Moreover, Goldman distinguishes science from everyday knowledge and other kinds of knowledge, claiming that the epistemic virtues of science, other than truth, do not hold sway elsewhere. And Goldman takes on some recent work in philosophy of science (notably Nancy Cartwright's), which denies that true generalities are the goals of science. Goldman maintains that although true generalizations are not always achieved, they are the goal of scientists.
One of the disputed questions in philosophy of psychology is the epistemic status of introspection. Goldman argues that if, as he thinks, reliability of evidence is the overriding goal, then intersubjective access of evidence is to be valued only insofar as it contributes to reliability. In the case of introspection, intersubjective access is not possible but reliability is likely, at least over a modest range of introspective reports (not over the wide range that, for example, Wilhelm Wundt expected). In particular, Goldman thinks that reports of introspection are reliable in the kinds of contexts in which they are currently used in experimental psychology and neuroscience, for example, to report awareness or unawareness of objects in one's visual field.
Goldman's reliabilism does not assume—as do many traditional epistemologists—that all rational belief-forming processes are “cold” cognitive ones, that is, free of emotion or motivation. Goldman claims that beliefs can be contagious when there is a prior bond of affection between individuals. (He distinguishes this from the currently popular “meme” theory, in which ideas are communicated by indiscriminate imitation. The latter process, he argues, lacks much power.) The reliability of BABA (the bonding approach to belief acceptability) is understood through evolutionary psychology: those with affective bonds will have less incentive to deceive one another than those without. BABA (as it sounds) is, in particular, the foundation of parent-child educational interactions.
“Externalist” reliabilism makes the shift from individual epistemology to social epistemology straightforward: in addition to assessing the reliability of individual processes, Goldman examines the reliability of social processes. In particular, Goldman looks at the kinds of social interactions that are conducive to true and justified beliefs in a community of individuals. Individual justification is in part due to favorable social circumstances. Goldman's social epistemology is developed at length in his Knowledge in a Social World (Reference Goldman1999); Part III of Pathways to Knowledge should be read as supplementary material. Goldman's social epistemology is similar to those of Philip Kitcher (Reference Kitcher1993), Paul Thagard (Reference Thagard1993), and others. They differ in degree of sociality from my work in the social epistemology of science (Solomon Reference Solomon2001), in which I argue that, even at times of consensus, normative assessment is of social states and individual scientists need not be justified in their beliefs. That is, I see no need for individual justification in science. All differ in kind from Helen Longino's social epistemology (Reference Longino1990), which assesses social epistemic states procedurally rather than instrumentally.
Goldman is conservative in his analysis of the goals of science. Veritism allows for preferential interests in some kinds of truth, and indeed Goldman argues that scientists have a special interest in general and simple truths about nature. Goldman could argue—as have a number of recent philosophers of science in the tradition of “Minnesota Pluralism” (Ronald Giere, Helen Longino, Kenneth Waters)—that scientific goals typically include personal and political goals, which influence scientists to produce partial but still representational accounts of the world.
Pathways to Knowledge is satisfying reading, as are Goldman's previous books. Goldman's “externalist” reliabilism is a simple but powerful idea, which he applies carefully and thoroughly. It is important reading for epistemologists of science.
