## Philosophy Lexicon of Arguments | |||

Author | Item | Excerpt | Meta data |
---|---|---|---|

Lewis, David Books on Amazon |
Probability Conditional | V 133f Conditional/Probability/Lewis: thesis: the probability of conditionals is the conditional probability. - (VsStalnaker) - Logical form: P(A->C) and P(C I A) - But not for the truth-functional conditional >(horseshoe). - Because here they are only sometimes equal. - Therefore, the indicative conditional is not truth-functional. - We call it probability conditional. - Problem: then the probability of conditionals would hint at the relation of probability of non-conditionals. - That would be incorrect. - Solution: assertibility is not possible with not absolute probability in the case of the indicative conditionals. --- V 135 Conditionalisation/Lewis/(s): E.g. conditionalisation on B: P(A) becomes P(A I B) the probability A given B. --- V 135f Probability conditional/prob cond/Lewis: here the probability of the antecedent must be positive. - A probability conditional applies to a class of probability functions - universal probability conditional: applies to all probability functions (Vs). --- V 137 Right: C and ~ C can have both positive values: E.g. C: even number, A: 6 appears. - Then AC and A~C are both positive probabilities. - Important argument: A and C are independent of each other. - General: several assumptions can have any positive probability if they are incompatible in pairs. - The language must be strong enough to express this. - Otherwise it allows universal probability conditionals that are wrong. --- V 139 Indicative conditional/Probability/Conditional probability/Lewis: because some probability functions that represent possible belief systems are not trivial - (i.e. assigns positive probability values to more than two incompatible options). - The indicative conditional is not probability conditional for all possible subjective probability functions. - But that does not mean that there is a guaranteed conditionalised probability for all possible subjective probability functions. - I.e. the assertibility of the indicative conditional is not compatible with absolute probability. --- V 139 Assertibility is normally associated with probability, because speakers are usually sincere. - But not with indicative conditionals. - Indicative conditional: has no truth value at all, no truth conditions and therefore no probability for truth. --- V 144 Conditional/Probability/Lewis/(s): the probability of conditionals is measurable - antecedent and consequent must be probabilistically independent. - Then e.g. if each has 0.9, then the whole thing has 0.912. --- V 148 Probability/conditional/Lewis: a) picture: the picture is created by shifting the original probability of every world W to WA, the nearest possible worlds - (Picture here: sum of the worlds with A(= 1) or non-A(= 0) - This is the minimum revision (no unprovoked shift). - In contrast, the reverse: b) conditionalisation: it does not distort the profile of probability relations (equality and inequality of sentences that imply A). - Both methods should achieve the same. |
LW I D. Lewis Die Identität von Körper und Geist Frankfurt 1989 LW II D. Lewis Konventionen Berlin 1975 LW IV D. Lewis Philosophical Papers Bd I New York Oxford 1983 LW V D. Lewis Philosophical Papers Bd II New York Oxford 1986 LwCl I Cl. I. Lewis Mind and the World Order: Outline of a Theory of Knowledge (Dover Books on Western Philosophy) 1991 |

> Counter arguments against **Lewis**

back to list view | > Suggest your own contribution | > Suggest a correction

Ed. Martin Schulz, access date 2017-03-27