Dataset of 36 verbal images previously noted by critics and scholars. Each image is accompanied by contextual information and by a set of textual characteristics, from its size in words and characters to the semantic domains it is composed of, and the imageability ratings of its words.