This article demonstrates the potential of sound iconicity for automatic text analysis. This study claims that--at least in poetic language--the ratio of plosive versus nasal sounds in a text predicts its emotional tone as it is perceived by readers; that is, poems that have a relatively high frequency of plosive sounds are more likely to express a pleasant mood with high activation, whereas a relatively high frequency of nasal sounds indicates an unpleasant mood with low activation. Moreover, these findings are universal (i.e., they are independent of specific languages or language families). This article presents the results of an intercultural study testing the previous hypothesis. (Contains 4 tables, 2 figures, and 3 footnotes.)