Understanding fluency: Beyond the automaticity of single word reading

Fluent reading is the hallmark of the skilled reader and a foundation for efficient comprehension. Yet we understand surprisingly little about the nature and development of reading fluency. Current theoretical models seem to assume that fluent reading is a corollary of automatized individual word recognition; however, the efficiency of word list reading is more closely associated with serial digit naming (RAN; a nonreading task) than with the speed of reading individual words. In this talk I will present a framework for conceptualizing reading of word sequences as a serial naming task by focusing on parallel processing due to temporally overlapping processing of successive items. This approach helps us understand differences between naming tasks, their relationship to reading fluency, and -- most importantly -- what are the additional prerequisites to reading fluency, beyond automaticity, which end up dominating the development of word-level reading skill past the beginner stages. The enriched view of fluency enabled by this approach can shed light on why fluency interventions are less efficient than interventions targeting earlier stages of reading development, as well as on why certain types of interventions have emerged as consistently more effective.