Proceedings of the Annual Meeting of the Cognitive Science Society

Does phonological onset overlap in a prime-target pair (espada–esparadrapo / sword- adhesive tape) facilitate target recognition as strongly as morphological relatedness (globo–globaliza–åão / globe-globalization)? Brazilian Portuguese speakers perform an auditory–visual priming lexical-decision task while EEG is recorded and analyzed in the time–frequency domain. Prime–target pairs manipulate (i) relation type: onset phonological overlap vs shared morphological base, and (ii) target length (7, 9, 12 graphemes). Within each length, targets vary in morphological depth: simplex [root+category] (e.g., carangueijo / crab) versus derived forms with two to four additional layers (e.g., penaliza–åão / penalization). We test whether delta-band (–ï1–4 Hz) power indexes decompositional structure building: we predict greater delta power for morphologically complex targets, scaling with derivational depth, and weaker or absent scaling for purely phonological overlap. Data collection is ongoing; full analyses and results will be reported at the meeting.