No label defined (Q4236366)
From MaRDI portal
!
This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:
scientific article; zbMATH DE number 1269567
| Language | Label | Description | Also known as |
|---|---|---|---|
| default for all languages | No label defined |
||
| English | No label defined |
scientific article; zbMATH DE number 1269567 |
Statements
14 September 1999
0 references
conditional constructions
0 references
formal epistemology
0 references
Gärdenfors paradox
0 references
Ramsey test
0 references
rational acceptance of conditionals
0 references
monotonicity
0 references
Levi's non-propositional version
0 references
epistemic conditional
0 references
belief state
0 references
indexical approach
0 references
belief revision
0 references
0.9574091
0 references
0.92577195
0 references
0.90607196
0 references
0 references
0 references
0 references
0 references
0 references
The Ramsey test, in its standard modern formulation, gives the following intuitive criterion for the rational acceptance of conditionals. A conditional proposition ``If \(A\), then \(B\)'' is (rationally) accepted in a given state of belief \(G\) just in case \(B\) should be accepted if \(G\) were revised by \(A\) as a new piece of information. Using the notations ``\(A> B\)'' for ``If \(A\), then \(B\)'', and ``\(G*A\)'' for ``\(G\) revised by \(A\)'', the Ramsey Test (RT) can be written as: \[ A>B\text{ is accepted in \(G\) iff \(B\) is accepted in }G*A. \tag{RT} \] The aim of the paper is to investigate the Gärdenfors Paradox (GP) in the context of (RT). The authors show that (GP) is based on a number of questionable assumptions and discuss three approaches to solve (GP): (1) (RT) is divided into two logically independent conditions: Strict (RT) and Monotonicity (to remind that (RT) implies Monotonicity!). The proposal was to replace (RT) by Strict (RT), while abandoning Monotonicity. (GP) is avoided since it essentially involves the assumption of monotonicity. (2) The second approach, namely Levi's non-propositional version, involved also modifying (RT) such that an epistemic conditional \(A>B\) is accepted in a belief state \(G\) iff \(B\in G*A\). Since conditionals cannot themselws be members of belief states, Monotonicity is not derivable and so (GP) is avoided. (3) Finally, the indexical approach is considered, according to which epistemic conditionals were thought of as expressing genuine beliefs, but only in a context-sensitive manner, relative to a belief state. The paper presents a semantics that validates both the full (RT) at the level of acceptance sets and Gärdenfors' axioms for belief revision, provided these axioms are applied to descriptive belief sets and not to acceptance sets. This avoids (GP). The three different approaches to solving (GP) are not competing to each other, and may be viewed as corresponding to different uses of conditional constructions in epistemic contexts. For the entire collection see [Zbl 0908.90003].
0 references