Proceedings of the Annual Meeting of the Cognitive Science Society

The Stroop effect is a well-known phenomenon where participants take longer when they name the color of letters that spell the name of a different color. This effect has been extensively studied, with various proposed theories and computational models. However, most models do not integrate with existing frameworks for color processing or reading, limiting their generalizability. We introduce the BRAID-Color probabilistic model, an extension of BRAID, an existing model of reading and reading acquisition. BRAID-Color consists in adding to an orthographic information processing branch, able to recognize words spelled by a sequence of letters, a color processing branch, able to name colored visual stimuli. Information fusion at the lexical level combines both to simulate color naming with colored letters. With default parameters, simulations in the congruent, neutral and incongruent conditions illustrate that the model successfully and robustly accounts for the Stroop effect.