How to answer an unanswerable question? Factors affecting correct “don't know” responding in memory tasks

When reporting from memory, people may often be asked unanswerable questions—questions for which the correct answer has never been encoded. These unanswerable questions should be met with an “I don't know” response. Previous research has shown that a manipulation commonly used to enhance memory at retrieval—context reinstatement—reduces appropriate “do not know” responding to unanswerable questions. Here we investigated whether this reduction is due to increased belief that a given question is answerable, or solely to increased confidence in specific responses for questions already believed to be answerable. In two experiments, we show that context reinstatement reduces “do not know” responding even when a “do not remember” option is available to express beliefs that particular questions are answerable. These results indicate that improved access to contextual information at retrieval creates an erroneous belief that unanswerable questions are in fact answerable.     

On Bellert's proposal concerning quantificational universals

The aim of the paper is to formalize I. Bellert's (McGill) proposal to characterize distinct nonequivalent readings of quantificationally ambiguous sentences with help of two features: absoluteness and distributiveness. The formalisation makes use of set theoretical and model theoretical standard notions. Foundamental rules, proposed by Bellert, govering the interpretation of cooccurring quantifiers are quoted and outlines of proofs of important derived rules are given.This research was sponsored by the grant CPBP 08.15-2/3, but some of results were obtained during the author's visit (May 86) in the Department of Philosophy of the Salzburg University: thanks are due to Prof. P. Weingartner for his hospitality.     

No non-trivial quasivariety of BCK-algebras has decidable first order theory

Using the semantic embedding technique the theorem announced by the title is proved.     

On comparison of theories by their contents

Popper's definition of verisimilitude was criticized for its paradoxical consequences in the case of false theories. The aim of this paper is to show that paradoxes disappear if the falsity content of a theory is defined with help of dCn or Cn ^–1.To the memory of Jerzy SupeckiI am grateful to David Pearce, Gerhard Schurz, Peter Simons, Maciej Spasowski and Jan Zygmunt for their helpful discussions on issues analysed in this paper. Particularly, Zygmunt's and Spasowski's comments enabled me to correct several mistakes of the earlier drafts.