This paper introduces a technique for viewing text objects under changing size constraints in 2D environments. Our approach automatically combines font size reduction and content reduction to preserve legibility of key words. Unlike traditional semantic zooming, our approach creates intermediate representations and transitions automatically. The main benefit is that it provides more meaningful views for different object sizes without additional authoring effort. Keywords Text reduction, semantic zooming, spatial hypertext.