Nelson Goodman (1906–1998) was a 20th-century American philosopher who radically reshaped the academic study of philosophy, forcing fundamental reconceptions of its problems, ends, and means. Goodman’s work cuts across the various branches of philosophy, revealing shared features and connecting links that more narrowly focused philosophers overlook.
Goodman was born in Somerville, Mass., on Aug. 7, 1906. He attended Harvard University, where he studied with C. I. Lewis, earning his B.S. in 1928 and Ph.D. in 1941. (See Lewis, C. I.) During graduate school he worked at a Boston art gallery. There he met his future wife, the artist Katharine Sturgis, and developed a lifelong passion for the arts. He taught at Tufts University (1944–1945), the University of Pennsylvania (1946–1964), Brandeis University (1964–1967), and Harvard University (1968–1977). Goodman founded Project Zero, a research program in arts education, and the Harvard Summer Dance Program. He served as president of the American Philosophical Association, Eastern Division, in 1967 and was a fellow of the American Academy of Arts and Sciences. Goodman died on Nov. 25, 1998, in Needham, Mass.
A concern with words and other symbols lies at the heart of Goodman’s philosophy. Western philosophy long assumed that meaning and fact are distinct. Goodman thought otherwise. His Structure of Appearance (1951; 3d ed., 1977) and Ways of Worldmaking (1978) argue that the identity of a fact depends on the words and other symbols used to represent it. By devising systems of classification, we determine what facts there are. Things can be organized in divergent but equally valuable ways, yielding equally viable world-versions. For example, light can be conceptualized as waves or particles; although waves are not particles, neither conception is deficient. “Light consists of waves” is correct relative to one acceptable world-version, incorrect relative to another. Absolutely and independently of versions we construct, it is neither correct nor incorrect. Goodman’s position is a relativism with rigorous restraints. Not every world-version is acceptable, nor does every sentence belong to an acceptable world-version. Acceptability requires consistency, coherence, suitability for a purpose, accord with past practice, and antecedent convictions. Sentences that do not belong to systems with these attributes are unacceptable. They create no facts.
Goodman’s “new riddle of induction” is a vivid example. Induction is generalizing from limited evidence to a wider class of cases. We need to know when a body of evidence affords a basis for induction and what class we should generalize to. Fact, Fiction, and Forecast (1955; 4th ed., 1983) shows that the problem is more difficult than it seems. Every emerald we have ever seen is green, and so we infer that all emeralds are green. But those emeralds also belong to a vast number of divergent classes. What favors “green” over the rest? Goodman highlights the problem by introducing a new term, grue: something is grue if and only if it is examined before future time t and is found to be green, or is not examined before t and is blue. Since t is in the future, every emerald we have ever seen is grue as well as green. Why not infer that all emeralds are grue? If we did, we would predict that some future emeralds are blue, not green. Goodman contends that grue is a well-defined term and that the emeralds we have seen are, in fact, grue. “Green” is preferable for induction, he believes, not because it cuts nature at the joints, nor because it is somehow more basic than grue, but because it has an established role in inductive practice. This does not ensure that newly discovered emeralds will be green. Induction affords no guarantees. If a nongreen emerald is found, then, despite its track record, “green” will lose its privileged status. But in the absence of disconfirming evidence, “green” is appropriate for induction and “grue” is not, because, unlike “grue,” “green” fits with inductive practices and precedents.
Most analytic philosophers privilege science. Goodman did not. He believed that art also advances understanding. Works of art are symbols, as he maintains in Languages of Art (1968; 2d ed., 1976). To understand them requires understanding the language or symbol system to which they belong. The aesthetic attitude is not passive contemplation of the sublime and the beautiful but active intellectual engagement with symbols having meanings that may be elusive. Goodman recognizes that art is dynamic; it breaks new ground and defies established conventions. This undermines the prospect of discovering a fixed and final “essence” of art but does not make the concept of art vacuous or merely subjective. Works of art are symbols with identifiable properties. They tend to be complex, multifaceted symbols that reward repeated attention and admit of multiple acceptable interpretations. They typically consist not of bald statements of established fact but of symbols that convey insights through metaphor, exemplification, and allusion. They are apt to suggest rather than assert. If we are receptive to an artwork’s suggestions, Goodman maintains, we can understand ourselves and our world in new ways. We discover previously unnoticed features of things and of our responses to those things. Both the world and our understanding of it are transformed.