Authors:

TAYLOR, J.S.H., Plunkett, K. & Nation, K.

Reference:

Year of publication:

2011

CBU number:

7170

Abstract:

Two experiments explored learning, generalization and the influence of semantics on orthographic processing in an artificial language. In Experiment 1, 16 adults learned to read 36 novel words written in novel characters. Post-training, participants discriminated trained from untrained items and generalized to novel items, demonstrating extraction of individual character-sounds. Frequency and consistency effects in learning and generalization showed that participants were sensitive to the statistics of their learning environment. In Experiment 2, 32 participants were pre-exposed to the sounds of all items (lexical phonology) and to novel definitions for half of these items (semantics). Pre-exposure to either lexical phonology or semantics boosted the early stages of orthographic learning relative to Experiment 1. By the end of training, facilitation was restricted to the semantic condition and to items containing low frequency inconsistent vowels. Pre-exposure reduced generalization suggesting that enhanced item-specific learning was achieved at the expense of character-sound abstraction. The authors' novel paradigm provides a new tool to explore orthographic learning. Although the present findings support the idea that semantic knowledge supports word reading processes, they also suggest that item-specific phonological knowledge is important in the early stages of learning to read. This article is not Open Access Compliant