Homophonic and heterophonic homograph processing in a semantic suppression task : an ERP study (Hebrew)

Homographs are words that share the same written form but convey different meanings and sometimes different pronunciations. For example, the word bark can refer either to the sound made by a dog or to the outer layer of a tree. In Hebrew, two main types of homographs are distinguished: homophonic homographs, which are spelled and pronounced identically (e.g., mesaper – “storyteller” vs. “haircutter”), and heterophonic homographs, which share the same spelling but differ in pronunciation (e.g., zar – “a stranger” vs. “a bouquet”).

The present study compared homophonic and heterophonic homographs in Hebrew and examined how the brain processes their multiple meanings. Specifically, it explored how readers suppress the irrelevant (dominant) meaning and select the contextually appropriate (subordinate) meaning when the word appears in a sentence that biases interpretation toward the subordinate sense.

An Event-Related Potential (ERP) paradigm was used to measure the N400 component, a neural index of semantic integration. Participants performed a semantic judgment task in which sentences containing homophonic or heterophonic homographs were followed by target words. Data were analyzed using Linear Mixed Models (LMM), focusing on reaction times, accuracy, and electrophysiological responses.

Behavioral results revealed that ambiguous words elicited slower reaction times compared to unambiguous words, and that heterophonic homographs caused greater response delays than homophonic ones. Ambiguity also reduced response accuracy. At the neural level, a significant N400 effect was found for target words, with stronger amplitudes for heterophones, particularly over left anterior and central scalp regions. Ambiguous words elicited more negative N400 responses than unambiguous words, and target words evoked stronger N400 activity than the homographs themselves.

These findings suggest that semantic and phonological suppression mechanisms differ between the two types of homographs, with heterophanes requiring dual suppression – both semantic and phonological. The results highlight the important role of the left inferior frontal gyrus (LIFG) in these suppression processes. Overall, the study provides real-time neurophysiological evidence for homograph processing in Hebrew and sheds light on the cognitive demands involved in reading ambiguous words. The findings also hold potential implications for educational and clinical interventions aimed at enhancing reading comprehension and language processing, including the use of targeted brain stimulation techniques (tDCS).