Philosophical Foundations of the Social Sciences: Analyzing Controversies in Social Research.
Philosophers of science today are less reproving than their teachers were; they say more about what natural science is or has been than what it ought to be. The story is different with the social sciences; here philosophers continue to say what ought to be done and to prescribe as well as describe scientific practice. Kincaid's book is prescriptively rich; where the methods of confirmation or explanation used in the social sciences are not those of the natural sciences, Kincaid argues that they ought to be.
The prescription, Kincaid explains in his introduction, is practical, for without scientific knowledge of society "government intervention in social and economic affairs would be inane". Why inane? Because judgments of what ought to be need to be informed by knowledge of what is and, in particular, knowledge of what the causes are. So, according to Kincaid, unless the social sciences discover the causes and effects of human behavior, social policy is groundless.
A central aim of the book is to show how the social sciences can best discover causes and so offer a ground for policy. The search for causes must be an objective, impartial pursuit of the truth and so, on Kincaid's view, the social sciences should be free of all non-epistemic values, for knowledge of causes is a matter of what is, not what ought to be. Interpretivists are no help here, according to Kincaid, for they eschew value-neutrality and deny that human behavior can be understood in terms of causes and so have nothing to contribute to social policy.
Individualists do better, according to Kincaid, but not well enough, for though they search for causes, they assume that only the properties of individuals can be causes and so lend credence to voluntaristic, merit based social policies and leave more collectivist policies groundless. Kincaid argues that social scientists are often better able to satisfy the standards of good science by offering macro rather than micro accounts of their data; holistic or macrosocial theories, hypotheses about social causes, are often better confirmed and display more of the basic scientific virtues than individualistic ones. Large scale social and economic events, Kincaid argues, are often better explained in terms of social conditions or forces than individual expectations or preferences.
The book Contains eight chapters. Chapter one introduces the naturalist and holist positions Kincaid wishes to defend. Chapter two rehearses and opposes arguments against scientific rationality and defends what Kincaid calls the evidential and explanatory virtues of good science. Chapter three shows that the social sciences can provide well-confirmed causal explanations and laws by countering conceptual arguments that social phenomena do not have causes or are not law-like and by offering an example of a theory in the social sciences that does provide them. Many macrosocial explanations are functional, and chapter four surveys arguments that functional explanations cannot be tested and confirmed and shows how they can be.
Chapter five makes the case for holism. Many good social theories are not reducible to individualistic accounts, Kincaid explains, because the social properties are multiply realizable, and so there are no bridge laws between the theory and an individualist treatment of the phenomena. But reduction, he argues, is not the only route to good science and, as long as the social theory is well-confirmed and fits with other well-confirmed theories, the science is good. Chapter six considers whether the social sciences can be interpretive and also naturalistic; they can, Kincaid argues, for interpretation is slight as long as the theory is macrosocial, and there is nothing about the practice of interpretation that falls outside the methods of the natural sciences.
Chapter seven is about economics and is one of the book's best. According to conventional wisdom, neither naturalism nor holism can be applied to economics, for the core of economics, the theory of the rational economic man, is ruthlessly individualistic, and few if any economic theories are well-confirmed or provide causal explanations of economic events. Kincaid takes a different view; philosophers of economics, he argues, have mistakenly identified economics with theories of rational choice or equilibrium theories, but the best economics is in supply and demand theories of aggregate market behavior. The laws of these theories, he shows, can be confirmed independently of theories of individual consumption or production decisions. Chapter eight offers a recipe for how a social scientist can win virtue, how she can leaven her theories and explanations with the features of good science, and speculation about why so few theories in the social sciences have the "symptoms of good science".
Kincaid's book is rich in argument; every favored position is clearly supported, every objection is fairly considered and carefully assessed, and no line of reasoning is overlooked or left unattended. His arguments for holism are especially compelling, for they show that some of the best theories in the social sciences are at the aggregate rather than individual level. His case for naturalism, however, is less successful; he focuses too much on the explanation and too little on the individuation of facts in the social sciences, for often the subjects' way of sorting the facts of social life is what identifies or individuates them.
Interpretivists, as Kincaid explains in chapter six, offer the following argument against naturalism: (1) facts in the social sciences are meaningful; (2) because they are meaningful, they must be interpreted to be understood; (3) interpretation is not amenable to naturalistic methods; therefore, (4) no naturalistic science of society is possible. Kincaid believes that the argument displays an individualist bias and questions each of the premises. He allows that individual actions are meaningful but thinks that when facts about individuals are aggregated, the meanings are unimportant and the macrosocial theories about these aggregates are not committed to them. "When social science is about the large scale", he writes, "it can abstract from individual detail and thus from explanations in terms of meaning" (p. 194). While facts about individual actions often must be explained using the subjects' concepts or categories, these can be identified using naturalistic methods, and, in any case, when the facts are about aggregate objects or events, the concepts or categories of the subjects no longer matter.
Kincaid is mistaken on two counts. First, in order to identify the subjects' own categories, we need to interpret their speech and actions, but there are constraints on the practice of interpretation that have no counterpart in the natural sciences, viz. charity, for we cannot understand too much of our subjects' thought or talk to be different from our own without doubting our beliefs about the content or meaning of what they think or say. Second, even if the subjects' concepts or categories don't explain the aggregate data, they shape their description. Consider, for example, the number of crimes in the US and how crimes are counted. An act counts as a crime in the US only if the subjects make, through legislation, the act a crime. The legislation enters into the individuation of the crime even if not into an explanation of changes in the crime rate. What distinguishes the social from the natural sciences is not how data are explained but how they are counted.
The claim that non-epistemic values can be removed from the practice of social science is also mistaken, for the discovery of causes often depends on such values. One aim in developmental psychology, for example, is to discover why children differ in some measure of intellectual competence, for example, reading or writing ability. A cause of the differences might explain why, but how the differences are identified depends on non-epistemic values, viz. on what counts as a competence, and these values are passed on to our identification of the causes. The situation is much like that of discovering the cause of a disease. That a condition is disease, that is, pathological, or even counted as a kind of condition at all depends on non-epistemic values, but so too do judgments about its etiology.
Ought one group of sciences (social) emulate another (the natural sciences)? Throughout this book, Kincaid maintains that broad conceptual considerations cannot tell us how sciences should be done, and he shows that many conceptual arguments against naturalism or holism in the social sciences are faulty. However, Kincaid must show more than that naturalism and holism in the social sciences are possible; he must show that they are desirable. In claiming that the social sciences ought to be more scientific, Kincaid turns away from naturalism and value-neutrality and to questions which, on Kincaid's view, science cannot answer, for example, why ought the social sciences aim to discover the causes of a crime rather than the values in virtue of which a community makes the act a crime, or why can these sciences describe but not oppose any of these values? Kincaid is eager to tell social scientists what to do but says too little about why they have more reason to listen to him than natural scientists had for listening to his teachers.
Department of Philosophy University of Minnesota 224 Church Street S.E. Minneapolis, MN 55455 USA