Experiment Icelandic 2 (ICE2) was run by Dietmar Roehm in Iceland in 2005 as part of research conducted by the Max Planck Research Group Neurotypology (Max Planck Institute for Human Cognitive and Brain Sciences, Leipzig; head: Ina Bornkessel-Schlesewsky). The data were originally analysed using "traditional" ERP analysis methods, i.e. single-subject averaging per condition, time window and electrode for statistical analysis via ANOVAs; grand average ERPs for visualisation. Results showed variability across conditions (verb classes ACT, ALT, DAT; see paper for details), which was deemed of potential interest for the relationship between language processing – and specifically electrophysiological measures in relation to behavioural acceptability – and language change. The findings were submitted following completion of data acquisition and analysis but never published. The new analysis of the data (2019), which is presented here, aimed to take into account trial-by-trial variability by subject and item in order to obtain a more fine-grained perspective on the data. See the README file for a detailed description of the processing pipeline.