David Corfield inferences

Idea

Give examples of inferences in natural language where the grammar allows regular valid steps, but it’s not an overt logical truth.

When is it a question of rewriting, sugaring, and when is it part of Sellars-Brandom material inference?

E.g.,

1. It is raining and today is Tuesday. Therefore, today is Tuesday. Clearly logical.

2. The dog is brown and quick. Therefore the dog is quick. Surely logical.

3. That man lives in the house. Therefore the house is occupied. Material? And yet it’s just a rewriting of forming the bracket type.

Examples

• X had already A-ed when Y B-ed. Therefore, X A-ed.

x: Person, A(x): Achievement, y: Person, B(y): Achievement, a:A(x), b:B(y), t(a) < t(b) < n, then x: Person, A(x): Achievement, a:A(x), t(a) < n.

• There is a person in the house. Therefore, the house is occupied.; This book has been read by someone, it is second-hand.

A: Type, a: A, then $\vert a \vert: \vert \vert A \vert \vert$, or $\vert \vert A \vert \vert$ is true.

• Caesar was killed by his friend Brutus. Therefore, Caesar was killed, Brutus killed someone. Caesar had not been killed earlier. Latter depends on specificity of killing marking the end of a life, so can only cause to die once.