ABSTRACTThe present approach estimates the strength of intensifiers in Dutch by computing their information values in a language corpus, that is, contextual information content (Cohen Priva, 2008; Piantadosi, Tily, & Gibson, 2011) and Shannon Information (Shannon & Weaver, 1948), to respectively explain the use value and the expressive value of intensifiers when they intensify a predicative adjective. Conflicting strength values help in understanding the high number of intensifiers commonly available in particular languages and the constant need for adding new ones. Our approach underlines the relevance of two measures of information content (IC) for ranking intensifiers: (i) IC in context: the more combinatorial or transitional options an intensifier has, the higher its contextual information content and consequently its use value; and (ii) IC in relation to all alternative intensifiers: the higher the surprisal value that the occurrence of an intensifier evokes, the higher its expressive value. We shall investigate the validity of these two measures by researching a large corpus of Dutch tweets and shall test whether the values of these two measures can predict the stacking order in sequences of intensifiers.