The effect of argument structure on the scope of advanced planning

The current study investigates whether the scope of advanced planning, i.e., how many words speakers plan ahead during language production tasks, depends on the argument structure of the verb. Unaccusative and unergative verbs are often compared in this respect. Both only allow for one argument (i.e., the subject) but whereas in unaccusative sentences the subject is an internal argument, in unergative sentences, it is an external argument.bIt has been argued that the grammatical encoding of subjects in unaccusative sentences require the encoding of the verb while this is not the case for unergative sentences (e.g, Perlmutter, 1978).
Momma and Ferreira (2021) examined this issue in two picture-word interference experiments. Participants produced sentences of the type The + noun+ is + verb (“The ballerina is running) upon presentation of a picture, preceded by a written distractor verb. The distractor was either semantically related or unrelated to the target verb. In the second experiment, they introduced a stop signal task to add complexity. The authors modeled the time interval between the onset of picture presentation and the onset of the subject noun (Onset + “The”) as well as the acoustic duration of the subject + is (e.g. “ballerina is”) as a function of relatedness. These dependent measures were assumed to reflect advanced and incremental planning, respectively They reported semantic interference effects in the Onset + “The” for both types of verbs in the first experiment and also in the duration of the subject + is for unergative verbs in the second experiment. They took these results to suggest unaccusative verbs always need to be planned in advance whereas unergative verbs are more flexible and can also be planned incrementally if the task is more complex.
In the present study we replicate and extend these findings. We first re-analysed the data of Momma and Ferreira (2021) with two changes. First, assuming that the smallest unit of advanced planning is the phonological word, we considered the time interval between onset of picture presentation and the onset of the sentence as a measure of advanced planning and the duration of the subject (the + noun) naming latency as a measure of incremental planning. Second,we examined the impact of the distractor on both the mean of these variables and their variance. The hypothesis that the scope of advanced planning for unergative sentences is flexible predicts a larger variance for these sentences. We found interference effects for both unaccusative and unergative verbs for both dependent variables. Semantically-related distractors led to more variance in production latencies in experiment 1 and in the duration of “The ballerina” in experiment 2 for unergative sentences. We are currently collecting data for a conceptual replication in Dutch with 48 participants. Unlike in English, in Dutch, the definite determiner depends on whether the noun has a gender or is neuter and can therefore not be uttered before the noun has been accessed. The study is pre-registered on OSF. The data of 43 participants have been collected. The results will be presented at the conference.