Segalowitz and Segalowitz distinguish between "speedup" (mean reaction time [RT] and mean standard deviation of responses in an RT task decrease to the same degree) and "automatization" (mean standard deviation decreases more than mean RT). The coefficient of variation, which is the standard deviation divided by the mean RT, decreases in the case of automatization while remaining unchanged in the case of speedup. We present data that are collected in two studies. The first one is a longitudinal Study spanning 2 years and comprising four RT tasks, both in second language (L2) English and first language Dutch (N > 200). The second study is an English L2 word training study. Students (N = 41) performed a lexical decision task before and after training. Convincing evidence for automatization was not found in either Study. The main problems in testing the Segalowitz and Segalowitz hypothesis is that gains in knowledge itself and gains in processing it cannot be adequately disentangled in the RT tasks currently used, characterized by a speed-accuracy trade-off. Although conceptually skill acquisition can be distinguished from knowledge accumulation, in reality, knowledge accumulation forms part of skill acquisition because, in real L2 learning, exposure to new words goes hand in hand with exposure to words encountered previously.
