Books on Amazon
Definition induction/Goodman: requires that some classes are seen as relevant classes by excluding others.
The sharpest criticism VsHume/Goodman: his analysis relates at best to the origin of predictions, not to their entitlement.
Deduction/Goodman: accordance with accepted practice.
The problem of induction is not a problem of prove, but a problem of definition of the difference between justified and unjustified predictions.
Mutual adjustment between definition and language use.
Grue/Goodman: Problem: same data support contrasting predictions - question: in what essential property hypotheses must be the same> law: not in connection with e.g. an object in my pocket. - "Grue" does not work as conventional non-law-like hypotheses (limited in space or time) - one can reverse the derivation: red and green from gred and reen.
Law-like or resumable hypotheses are not to characterize purely syntactically.
What confirm certain data, is not what is obtained by generalization of separate individual cases, but that which is obtained by generalization of the entire body of data material.
Weisen der Welterzeugung Frankfurt 1984
Tatsache Fiktion Voraussage Frankfurt 1988
Sprachen der Kunst Frankfurt 1997
N. Goodman/K. Elgin
Revisionen Frankfurt 1989