“…Assuming that both listeners and speakers should converge on a shared discriminative code, and that both incrementally process the signal, also listeners should benefit from a smooth signal. Evidence showing that listeners use language to predict not only states of the world but also parts of the upcoming signal already corroborates this idea (e.g., as previously mentioned, DeLong et al, 2005;Lau et al, 2016;Rohde & Rubio-Fernandez, 2021;Tribushinina & Mak, 2016;van Berkum et al, 2005;Wicha et al, 2004;Willems et al, 2016). If, in addition, it could be confirmed in future research that listeners benefit from longer but smoother noun phrases, this would further suggest that a smooth signal benefits the communication process as a whole, including speakers and listeners.…”