Written production of German compounds: Effects of lexical frequency and semantic transparency

In this study, we present an experiment in which we examined the time course of typing German compounds. The compounds varied according to three criteria: (1) whole word frequency (high vs. low), (2) head frequency (high vs. low) and (3) semantic transparency (transparent vs. opaque). In this experiment, we recorded the interkey intervals (IKIs) and concentrated on the IKI measurements found at the boundary of the two immediate constituents in compounds. We refer to this boundary type as an SM-boundary because (S)yllable and (M)orpheme boundaries coincide at this word position. As we found effects of lexical frequency for SM-IKIs in a series of previous studies, we argue that possible differences in SM-IKIs found for compounds of different frequency classes and of different degrees of semantic transparency can give an insight into the processes involved in the written production of German compounds: whole word procedures and/or compositional procedures. Our findings show that SM-IKIs are affected by compound frequency, head frequency and semantic transparency. We therefore argue that both whole word procedures and compositional procedures are involved in the written production of German compounds. These findings are in line with those versions of dual-route models which postulate that the two routes run in parallel and interact.