Suppressing contextually irrelevant meanings of homophonic versus heterophonic homographs: A tDCS study targeting LIFG.

Readers frequently encounter homographs (e.g., bank) whose resolution requires selection-suppression processes: selecting the contextually relevant meaning, while suppressing the irrelevant one. In two experiments, we investigated how these processes are modulated by the phonological status of the homograph (homographs with one vs. two possible pronunciations); and what is the involvement of the left inferior frontal gyrus (LIFG, including Broca's area) in these processes. To these ends, Experiment 1 utilized the context verification task with two types of Hebrew homographs: homophonic (e.g., bank) and heterophonic (e.g., tear). In the task, participants read sentences ending either with a homograph (e.g., bank) or an unambiguous word (e.g., shore). The sentences were biased towards the homograph's subordinate meaning (e.g., The fisherman sat on the bank/shore), and were followed by a target word related to the homograph's dominant meaning (e.g., MONEY). The participants were asked to judge whether the target was related to the overall meaning of the sentence. An ambiguity effect was observed for both types of homographs, reflecting interference from the irrelevant dominant meaning. However, this ambiguity effect was larger for heterophonic than for homophonic homographs, indicating that dominant meanings of heterophonic homographs are more difficult to suppress. Experiment 2 was identical, except that the procedure was coupled with transcranial direct current stimulation (tDCS) over the LIFG (including Broca's area). We found that stimulating the LIFG abolished the ambiguity effect, but only in the case of heterophonic homographs. Together, these findings highlight the distinction between phonological and semantic levels of selection-suppression processes, and the involvement of the LIFG in the phonological level of these processes.