Authors: Sato, Y., Wajima, Y., Ueda, K.

Title: Strategy analysis of non-consequence inference with Euler diagrams

Journal(書誌情報): Journal of Logic, Language, and Information

doi: 10.1007/s10849-017-9259-x


Abstract: How can Euler diagrams support non-consequence inferences?
Although an inference to non-consequence, in which people are asked to judge
whether no valid conclusion can be drawn from the given premises (e.g., All
B are A; No C are B), is one of the two sides of logical inference, it has
received remarkably little attention in research on human diagrammatic
reasoning; how diagrams are really manipulated for such inferences remains
unclear. We hypothesized that people naturally make these inferences by
enumerating possible diagrams, based on the logical notion of
self-consistency, in which every (simple) Euler diagram is true
(satisfiable) in a set-theoretical interpretation. The work is divided into
three parts, each exploring a particular condition or scenario. In condition
1, we asked participants to directly manipulate diagrams with size-fixed
circles as they solved syllogistic tasks, with the result that more
reasoners used the enumeration strategy. In condition 2, another type of
size-fixed diagram was used. The diagram layout change interfered with
accurate task performances and with the use of the enumeration strategy;
however, the enumeration strategy was still dominant for those who could
correctly perform the tasks. In condition 3, we used size-scalable diagrams
(with the default size as in condition 2), which reduced the interfering
effect of diagram layout and enhanced participants’ selection of the
enumeration strategy. These results provide evidence that non-consequence
inferences can be achieved by diagram enumeration, exploiting the
self-consistency of Euler diagrams. An alternate strategy based on
counter-example construction with Euler diagrams, as well as effects of
diagram layout in inferential processes, are also discussed.

著者Contact先の email: ueda[at] (K. Ueda) ([at]を@