Goodman's problem and scientific methodology

Journal of Philosophy

Volumn 100, Issue 11, 2003, Pages 573-590

[No Author Info available]

Author keywords

[No Author keywords available]

Indexed keywords

EID: 60949715291     PISSN: 0022362X     EISSN: 19398549     Source Type: Journal    
DOI: None     Document Type: Article

References (13)

1. Goodman's most famous discussion of his problem is in his Fact, Fiction, and Forecast (Cambridge: Harvard, 1955).
2. For a collection surveying solutions proposed in the last fifty years, see Douglas Stalker, ed., Grue! (Chicago: Open Court, 1994).
3. John Pollock has expressed the alleged need for this supplement to standard statistical theory explicitly: "It seems likely that... [a] strong projectibility constraint should be imposed upon all familiar patterns of statistical inference. The need for such a constraint seems to have been ignored in statistics" - "The Projectibility Constraint," in Stalker, ed., p. 141.
4. See, for example, Colin Howson and Peter Urbach, Scientific Reasoning: The Bayesian Approach (Chicago: Open Court, 1989), chapter 4;
5. and Elliott Sober, "No Model, No Inference: A Bayesian Primer on the Grue Problem," in Stalker, ed., pp. 225-40.
6. See Jackson, "Grue," this JOURNAL, LXXII, 5 (March 13, 1975): 113-31 (reprinted in Stalker, ed.);
7. and Jackson and Pargetter, "Confirmation and the Nomological," Canadian Journal of Philosophy, X (1980): 415-28.
8. The original study (which is now quite controversial) is discussed in E. G. Mayo, The Human Problems of an Industrial Civilization (New York: Macmillan, 1933). Mayo's discussion tends not to distinguish between the effects of special kinds of work supervision and the effects of being studied per se. I am grateful to Art Owen for bringing the Hawthorne effect to my attention.
9. See also J. Moreland's On Projecting Grue" for an earlier attempt to deal with the grue problem (though not emerose problems) using the concept of randomness, applied within a Carnapian framework - Philosophy of Science, XLIII, 3 (1976): 363-77. Note also that while I use the term "bias" in a specific way here, the term can also be used more generally to refer to any (nonchance) source of spurious association.
10. The debate between John Dewey and Hans Reichenbach in Dewey's "Schilpp volume" includes an interesting exchange on this issue. Reichenbach modeled all nondeductive inference on statistical estimation, and hence saw the role of sample size in generating convergence on a true value as crucial. Dewey had a different model of nondeductive inference, in which all hangs on the ability to find an individual that is representative of its kind; if we can do this, then one individual is enough. I am suggesting that both Dewey and Reichenbach were on the right track with respect to understanding some inferences in science, but both were too inclined to generalize. See P.A. Schilpp and L.E. Hahn, eds., The Philosophy of John Dewey (Library of Living Philosophers, La Salle, IL: Open Court, 1939).
11. For the bearing of natural kinds on induction, see especially W.V. Quine, "Natural Kinds" (in Ontological Relativity and Other Essays (New York: Columbia, 1969), reprinted in Stalker, ed.);
12. and also Hilary Kornblith, Inductive Inference and Its Natural Ground (Cambridge: MIT, 1993).
13. In biology", the much-debated statistical concept of heritability is an example (see R.C. Lewontin, "The Analysis of Variance and the Analysis of Cause," American Journal of Human Genetics, XXVI (1974): 400-11). Heritability is slowly becoming less important to genetics, as particular causal pathways linking genes and phenotypic traits are uncovered in more detail.