In two experiments we study how redundant size modifiers influence the perceived size of objects. We show that when objects are referred to with overspecified descriptions (for example, using a description like “the large red chair” in a situation where all chairs are equally large but different in color), participants subsequently estimate the object to be larger than when objects are referred to using minimally distinguishing descriptions (e.g., “the red chair”). In Experiment 1, we show this effect with adult language users and different kinds of size modifiers. In Experiment 2, the same effect is shown for children of two different age groups (7- and 10-year olds), and for different kinds of visual size contrasts. Interestingly, we observe an inversely proportional relation between the age of our child participants and the difference in size estimates for minimal and overspecified descriptions, suggesting that language users gradually become better at avoiding false pragmatic inferences from redundant adjectives as they grow older.