"When we say no we mean no": Interpreting negation in vision and language

This study looks into visual negation. It tests the assumption that visual negation operates along the same lines proposed for linguistic negation (Giora, 2006, 2007). Specifically, it assumes that, like linguistic negation, visually negated information is not unconditionally discarded. Instead, it is sensitive to discourse goals and requirements and will therefore allow information within its scope to remain accessible to comprehenders, should the circumstances require it. This must be true not only of highly restricting contexts that can tolerate no intricate inferencing (e.g. road signs) but also of contexts inviting complex inferential processes that could afford suppression and replacement with alternatives (e.g. works of art). On the basis of interpretations of straightforward and complex visual stimuli as well as empirical data collected from raters, we show that, as predicted, when communicators visually communicate "not X" interpreters often take them to mean "not X", retaining 'X' in memory rather than replacing it by an alternative opposite ('Y').