A Conflict of Visions: Ideological Origins

Thomas Sowell has achieved an enviable reputation in many different areas of economics. His many works on the economics of immigration, culminating in Ethnic America, have won him an eminent place in this field. His Knowledge and Decisions applies in a comprehensive way the insights of Friedrich A. Hayek to a vast number of social phenomena. As if this were not enough, Sowell has also published widely in the history of economic thought. It is with no little anticipation, then, that one turns to Sowell's venture into the history of ideas. Will the insight and imagination displayed in his previous work enable him to contribute a new way of looking a t intellectual history? Admittedly, perusal of his Marxism: Philosophy and Economics dims one's enthusiasm. That book offered little but a tired rehash of elementary Marxist economics, presented as a major piece of scholarship.' Further, except for in the final chapter Sowell manifested a surprising sympathy for Marxism. Sowell has made the work of analysis of his book as straightforward as possible, since he has carefully constructed it around a central thesis which the title adumbrates. What does Sowell mean by a vision? He informs us that "a vision is a sense of causation." I t is more like a hunch or a "gut feeling" than it is like an exercise in logical or factual verification. "These things come later, and feed on the raw material provided by the vision" (p. 16). The "hunches" that Sowell concerns himself with do not primarily involve moral judgments. "People with the same moral values readily reach different political conclusions. ... Labeling beliefs 'value premises' can readily become one more means by which conclusions insulate themselves from confrontation with evidence or logic" (p. 217). Before plunging into Sowell's distinction between "constrained" and "unconstrained" visions, the principal subject of the book, a pause over "vision" seems necessary. Sowell may mean by this an innocuous truism: theorists rarely arrive a t a total system a t once but rather extend and shape an initial conjecture as evidence turns up and as