On the ability to inhibit complex movements: A stop-signal study of typewriting.

Experienced typists were instructed to stop typing when they detected a stop signal (a tone or a change in the display). The main question addressed in four experiments was whether typing responses were organized into words or into units smaller than the word (single letters or letter clusters). If subjects typed whole words after the stop signal, the probability of inhibiting the last letter of a word should be close to zero and should not be affected by word length. In fact, the probability was usually closer to one and was strongly affected by word length (see Experiments 1 and 2), suggesting that the unit is smaller than the word. If subjects typed whole words after the stop signal, they should type fewer letters after the signal the longer the stop-signal delay because fewer letters remain to be typed. For a given stop-signal delay, they should type more letters after the signal the longer the word because more letters remain to be typed. Neither of these predictions was confirmed whether subjects typed single words (see Experiments 1, 2, and 4) or sentences (see Experiment 3), again suggesting that the unit is smaller than the word. The only exception occurred in Experiment 3,