Talking to Myself: How the Phonological Network Supports Inner Speech During Computer Code Comprehension

Abstract

Prior work has highlighted a connection between natural language skills and programming ability. In this dissertation, I investigate the role of a specific language-related process—phonological coding—and its potential role in aiding computer code comprehension. Using a neuroscientific individual differences approach I examine (1) whether the phonological system is actively engaged during Python code comprehension, (2) the mechanisms that might explain this relationship, and (3) how individual differences in behavioral factors indexing skill, capacity, and strategy modulate the involvement of the phonological system during code comprehension. Specifically, I investigate whether phonology's role in code comprehension is merely an epiphenomenon of accessing English word meanings, or if it serves a functional role in supporting comprehension. My results suggest that the phonological system is involved in code comprehension and that this involvement is modulated by individual differences in cognitive capacity and strategy. Together, this work suggests that phonological codes can be a functional building block for constructing an internal problem representation during computer programming tasks.