Morphological facilitation and semantic interference in compound production: An ERP study

On the lexical representation(s) of compounds: A continuous picture naming study.

“…Other studies have reported results that are best explained by the single-lemma account (Levelt et al, 1999; e.g., Lorenz, Mädebach et al, 2018; Lorenz, Zwitserlood, et al, 2021; Lüttmann et al, 2011), or have only found weak evidence for multiple lemmas in their naming error analysis (e.g., Lorenz, Regel, et al, 2018). As discussed above, we believe that the discrepancy in results between those previous studies and the present one is due to the different experimental paradigms used.…”

“…Semantic interference was obtained for distractors related to the compound, not for distractors merely related to one of its constituents. If, during compound production, the compounds’ constituent lemmas were indeed activated in addition to the compound lemma, the categorical relationship of the distractor and the compound’s first constituent should cause (at least some) interference (see also Lorenz, Regel, et al, 2018; Lorenz, Zwitserlood, et al, 2021). Further evidence comes from grammatical gender effects in compound production.…”

“…Many studies on compound production point toward morphemesized representations at the word-form level instead of compoundsized word forms (e.g., Bien et al, 2005;Lorenz, Regel, et al, 2018;Lorenz & Zwitserlood, 2016;Lüttmann et al, 2011;Roelofs, 1996; but see Janssen et al, 2008;Janssen et al, 2014). Research on the lemma representation of compounds, however, is sparse and results are partially inconsistent (e.g., Lorenz & Zwitserlood, 2016;Lorenz, Mädebach, et al, 2018;Lorenz, Zwitserlood, et al, 2021;Mondini et al, 2004;Semenza et al, 1997). The current study provides new experimental data from speech production on the lexical representation of compounds at the lemma level and aims to assumptions: (a) lexical-semantic encoding includes at least two levels, a conceptual and a lexical level, and (b) conceptually related lexical entries are initially coactivated before the target entry is selected from among the activated items (e.g., Caramazza, 1997;Dell et al, 1997;Levelt et al, 1999;Mahon et al, 2007;Roelofs, 2018;Schade, 1999).…”

“…First, one could argue that our participants were particularly aware of the morphological structure of our items, which might have resulted in a different way of processing the compounds. In our blocked LEXICAL REPRESENTATION(S) OF COMPOUNDS design, participants had to produce only simple nouns in the first and only compounds in the second half of the experiment (or vice versa), whereas participants only produced compound targets in previous PWI studies (e.g., Lorenz, Zwitserlood, et al, 2021). Our participants might therefore have noticed the morphological structure of the compounds, particularly when presented after the simple words, and might have processed their constituents more consciously.…”

Encoding category-level and context-specific phonological information at different stages: An EEG study of Mandarin third-tone sandhi word production

“…Zhang and Damian, 2019), the target pictures were presented with all three types of primes for each participant. Moreover, each condition was also repeated to increase the experimental power (see Lorenz et al, 2021 for a similar practice). Thus, each target word was repeated 6 times (3 prime conditions × 2 repetitions), giving rise to a total of 96 trials for each prime type.…”