TITLE PAGE & CONTENTS | PHL347 HOME | LH'S PHILOSOPHY HOMEPAGE 
Unauthorized
Online Supplement to
A Companion to the Philosophy of Mind
Samuel Guttenplan (ed.)
(Oxford: Basil Blackwell Ltd., 1995)
Unauthorized Online Supplement editor:  Larry Hauser 
Inference When I heard that Airforce One was at Detroit Metro Airport I inferred that Bill Clinton was in Detroit. Like such paradigmatic PROPOSITIONAL ATTITUDES as BELIEF and DESIRE, inferences clearly have propositional CONTENT; hence INTENTIONALITY: the content of my inference was the PROPOSITION President Clinton is in Detroit; it was about Clinton being in Detroit. Yet inferring also smacks of ACTION. Inference seems something one does. One simply has beliefs and desires, one draws inferences. Also, inferring, like many actions, has success conditions. Success conditions for inferences in fact, are (among other things) the subject matter of Logic. Attitudes, on the other hand, do not seem subject to success conditions in the same way. I may succeed or fail in obtaining the object of my desire; but its odd to speak of succeeding in desiring it. Similarly beliefs may be true or false; yet it's odd -- unless under special or peculiar circumstances -- to speak of succeeding in believing something.

While there is always a proposition inferred in inference, there is generally, if not always a second proposition (or set of propositions) involved in inferring. Inference seems to be essentially a three-place relation between the individual doing the inferring and two propositions (or a proposition and set thereof): a proposition inferred (the conclusion); and the proposition or propositions (the premises) from which the conclusion is inferred. Logicians here distinguish inductive from deductive inference. Deductive inferences establish or purport to establish the certainty of the conclusion given the premises. Inductive inferences establish (or purport to establish) the probability of the conclusion given the premises. If the truth of the conclusion really is certain given the truth of the premises, the premises are said to "imply" or "entail" the conclusion, and the corresponding inference (the actual drawing of the conclusion) is deemed "deductively valid". If the conclusion really is probable on the evidence cited (i.e., given the truth of the premises stating it) the corresponding inference is said to be "inductively strong".

The intentionality of inference, as pretheoretically understood, poses a special bootstrapping challenge for so called "inferential role semantics". Inferential role semantics hold that the meaning or semantic content of a term (or concept) is determined by its role in the various inferences it enters into. But if inferential processes are themselves intentional -- as they are on our everyday understanding -- whence do these inferences derive their content? More generally, any sort of CONCEPTUAL ROLE SEMANTICS -- inferential or otherwise -- faces a similar challenge. Conceptual roles, pretheoretically understood, being roles that CONCEPTS enter into; and concepts, pretheoretically understood, being already imbued with semantic content; this content already has to exist before there can even be conceptual roles as pretheoretically understood. Consequently, conceptual role theories must give "concept" and "conceptual role" technical senses that do not already presuppose SEMANTIC content: their theoretical characterization (under the conceptual role theory) must be purely SYNTACTIC.
 
See also NORMATIVE, RATIONALITY
 

LARRY HAUSER
1998
TITLE PAGE & CONTENTS | PHL347 HOME | LH'S PHILOSOPHY HOMEPAGE 
CROSS REFERENCE AND HYPERLINK CONVENTIONS
Plain SMALL CAPS refer to entries in Guttenplan's Companion to the Philosophy of Mind.  Underlined  SMALL CAPS hyperlink to Unauthorized Online Supplement entries.  Underlined ITALICS SMALL CAPS hyperlink externally to other resources online.   ITALICS SMALL CAPS without underlining refer to nonexistent would-be Supplement entries ("vapor ware").