UC Irvine

Among many things, Frank Ramsey is famous for his invention of a logical device called the Ramsey sentence. The Ramsey sentence of a scientific theory is a conjunction of the theory’s propositions where the theoretical vocabulary is existentially quantified out. The sentence appears in an obscure paper titled “Theories”. This paper is extremely difficult because Ramsey likely did not finish it before his death and it mostly consists of a mathematical model of scientific theories and a labyrinthine toy example. Consequently, philosophers have proposed many radically different and incoherent interpretations of the Ramsey sentence and Ramsey’s philosophy of science. An important fact interpreters have ignored is that Ramsey had intended “Theories” to be a chapter in a book centered on a revised version of his formal theory for decision-making under uncertainty. The principal accomplishment of my dissertation is to leverage this fact to provide the first complete and philosophically satisfying interpretation of “Theories”, including Ramsey’s toy example and the Ramsey sentence.

Ramsey’s theory for decision-making is incompatible with his philosophical commitments in “Theories” and contemporary papers. In my dissertation, I revise that theory to make it consistent with those commitments. I combine this with the formal model he describes in “Theories” to provide a unified reconstruction of Ramsey’s intended model for scientific theories. I use the reconstructed model to explain Ramsey’s toy example, and I furnish a complete interpretation of Ramsey’s arguments in “Theories”. With these parts, I develop a detailed account of the laws and chances derived from scientific theories and a novel interpretation of the Ramsey sentence. This has several important philosophical consequences.

First, Ramsey did not use the Ramsey sentence in the ways philosophers have previously proposed. Philosophers view the Ramsey sentence either as a tool for eliminating theoretical terms or for asserting the existence of theoretical properties. Neither of these claims is correct. Instead, I argue that the Ramsey sentence is a communication and deliberation device aimed at representing the propositions an individual believes with her behavioral dispositions. Informally, it expresses as new propositions the mixture of rules a person uses to guide her behavior so that she may deliberate and communicate those rules with herself and other people. This suggests the Ramsey sentence can be used to read off the content of an agent’s belief from their behavior.

Second, Ramsey formulated a novel and interesting type of non-reductionistic scientific anti-realism. That is, for Ramsey theoretical propositions are fictitious and yet not eliminable for observational propositions. Theoretical propositions are fictitious because their meaning depends on theoretical laws, which, being universally quantified, are also fictitious. But theoretical propositions cannot be eliminated because they entail conceptual possibilities that are richer than those given by the observation language.

Third, Ramsey held laws and chances to be rules of deference for degrees of belief. His account of laws and chances foreshadows pragmatic, subjectivist accounts of laws and chances from later in the twentieth century. Degrees of belief are representations of preferences over gambles involving observable propositions. A chance then is a rule for preferring some gambles over others, and a law is a chance that assigns unity to its degrees of beliefs. This prefigures the views that chances and subjunctive conditionals are reducible to properties of degrees of belief.