Accentuate the positive: semantic access in english compounds.

Victor Kuperman

doi:10.3389/fpsyg.2013.00203

Abstract

The present study supplements research on semantic effects in word processing by focusing on the role that meanings of morphemes play in recognition of complex words. We present an overview of behavioral effects of six semantic properties characterizing the emotional and sensory connotations of English compounds and their morphemes, as well as their semantic richness. Semantics of compounds affected latencies to those compounds, and semantics of morphemes affected latencies to those morphemes presented as isolated words. Yet semantics of morphemes had little bearing on recognition of compounds, with the exception of longer recognition times for compounds with emotionally negative morphemes (e.g., seasick). We interpret the data as evidence against obligatory decomposition and dual-route accounts of morphological processing and in favor of the naive discriminative learning account that posits independent, morphologically unmediated, and simultaneous access to all meanings activated by orthographic cues in the visual input. We discuss selectivity and division of attention as driving forces in complex word recognition.

Highlights

What role do morphemes play in visual recognition of complex words? This question has been on the forefront of research on morphological processing for about four decades and gave rise to a spectrum of processing theories
Six candidate properties were selected on the basis of their availability for a large number of English compounds and their constituents: emotional valence and arousal, imageability, concreteness, body-object interaction rating, and sensory experience rating
The values were further correlated with lexical decision latencies to compound words and their constituents presented in isolation

Summary

Introduction

What role do morphemes play in visual recognition of complex words? This question has been on the forefront of research on morphological processing for about four decades and gave rise to a spectrum of processing theories (for recent reviews, see Amenta and Crepaldi, 2012; Diependaele et al, 2012). Dual- and multipleroute models propose that the meanings of both the complex word and its morphemes can be activated simultaneously: the processing preference for either the morphemic or the whole-word route is not categorical and can be biased by the formal properties of the complex word (e.g., Caramazza et al, 1988; Schreuder and Baayen, 1995; Kuperman et al, 2008). A radically different class of models does away with morphemes as an independent level of representation and argues for a learned mapping between co-activating formal and semantic units that is either direct (Baayen et al, 2011) or mediated by a layer of hidden units (Seidenberg and Gonnerman, 2000). We critically review the current practices of identifying semantic access and propose complementary diagnostic measures

Objectives

Methods

Results

Conclusion