This talk addresses a well-known dichotomy of intonation research, compression versus truncation. The dichotomy describes two strategies which can be used to adjust an f0-pattern to a reduced scope of voicing: an intended pattern can either be realised in a shorter time span which leads to an increased velocity of f0-trajectories (compression), or it can be produced incompletely so that f0-targets are cut off in their frequency (truncation). I am going to argue that this dichotomy is insufficient to capture the whole range of systematic variation in such time pressure situations. Experimental data from German and Russian will be presented to substantiate this claim. Contrary to previous results (Grabe, 1998), German will be shown not to truncate a fall and not to compress a rise. Russian will be shown neither to use compression nor (the 'typical') truncation.