works
D. M. Armstrong The identification problem and the inference problem article The Identification and Inference problems challenge the ‘relations of universals’ theory of laws by questioning the nature of the law-making relation and how it entails empirical regularities. These issues are resolved by grounding the theory in the perception of singular causation. Because token-token causal sequences are epistemically primitive and frequently exhibit patterns, such regularities justify the postulation of universals. The Identification Problem is addressed by identifying the law-making relation between universals as the causal relation itself—the same relation experienced in singular interactions, but hypothesized to hold between types rather than tokens. This identification facilitates a solution to the Inference Problem: if a causal relation holds between universals, it is conceptually or analytically necessary that tokens of those universals will instantiate the corresponding causal effects. Under this framework, laws of nature are higher-order atomic facts that explain the uniformity of the physical world. This account demonstrates how laws as relations of universals can both entail and explain regularities, providing a mechanism that covers causal laws and potentially extends to probabilistic ones, though non-causal laws remain a subject for further inquiry. – AI-generated abstract.

The identification problem and the inference problem

D. M. Armstrong

Philosophy and Phenomenological Research, vol. 53, no. 2, 1993, pp. 421

Abstract

The Identification and Inference problems challenge the ‘relations of universals’ theory of laws by questioning the nature of the law-making relation and how it entails empirical regularities. These issues are resolved by grounding the theory in the perception of singular causation. Because token-token causal sequences are epistemically primitive and frequently exhibit patterns, such regularities justify the postulation of universals. The Identification Problem is addressed by identifying the law-making relation between universals as the causal relation itself—the same relation experienced in singular interactions, but hypothesized to hold between types rather than tokens. This identification facilitates a solution to the Inference Problem: if a causal relation holds between universals, it is conceptually or analytically necessary that tokens of those universals will instantiate the corresponding causal effects. Under this framework, laws of nature are higher-order atomic facts that explain the uniformity of the physical world. This account demonstrates how laws as relations of universals can both entail and explain regularities, providing a mechanism that covers causal laws and potentially extends to probabilistic ones, though non-causal laws remain a subject for further inquiry. – AI-generated abstract.

PDF

First page of PDF