Maximal Specificity and Lawlikeness in Probabilistic Explanation

Dados Bibliográficos

ID10704283
AutoresCarl G Hempel (Princeton University, autor correspondente)
Ano1968
Volume35
Fascículo2
Páginas116-133
Data de publicação1968-06-01
Peer ReviewedSim
Open AccessSim
TipoARTICLE
PeriódicoPhilosophy of Science (JOURNAL)
Identificadores do periódicoISSN: 0031-8248 • E-ISSN: 1539-767X
EditoraCambridge University Press (CUP) (PUBLISHER)
DOI10.1086/288197
OpenAlexW2059634082
IdiomaEN
Citações recebidas33
Referências citadas5

The article is a reappraisal of the requirement of maximal specificity ( RMS ) proposed by the author as a means of avoiding “ambiguity” in probabilistic explanation. The author argues that RMS is not, as he had held in one earlier publication, a rough substitute for the requirement of total evidence, but is independent of it and has quite a different rationale. A group of recent objections to RMS is answered by stressing that the statistical generalizations invoked in probabilistic explanations must be lawlike, and by arguing that predicates fit for occurrence in lawlike statistical probability statements must meet two conditions, at least one of which is violated in each of the counterexamples adduced in the objections. These considerations suggest the conception that probabilistic-statistical laws concern the long-run frequency of some characteristic within a reference class as characterized by some particular “description” or predicate expression, and that replacement of such a description by a coextensive one may turn a statement that is lawlike into another that is not. Finally, to repair a defect noted by Grandy, the author's earlier formulation of RMS is replaced by a modified version

Ambiguity · Class (philosophy) · Counterexample · Discrete mathematics · Epistemology · Linguistics · Mathematical economics · Predicate (mathematical logic) · Probabilistic logic · Statement (logic) · Bayesian Modeling and Causal Inference · Computer Science · Mathematics · Philosophy