The processing of lexical stress during visual word recognition: Typicality effects and orthographic correlates

Abstract

Many studies that have examined reading at the single-word level have been restricted to the processing of monosyllabic stimuli, and, as a result, lexical stress has not been widely investigated. In the experiments reported here, we used disyllabic words and nonwords to investigate the processing of lexical stress during visual word recognition. In Experiments 1 and 2, we found an effect of stress typicality in naming and lexical decision. Typically stressed words (trochaic nouns and iambic verbs) elicited fewer errors than atypically stressed words (iambic nouns and trochaic verbs). In Experiment 3, we carried out an analysis of 340 word endings and found clear orthographic correlates of both grammatical category and lexical stress in word endings. In Experiment 4, we demonstrated that readers are sensitive to these cues in their processing of nonwords during two tasks: sentence construction and stress assignment. We discuss the implications of these findings with regard to psycholinguistic models of single-word reading.

The research reported here was based on a doctoral dissertation completed by J. Arciuli under the supervision of L. Cupples in the Speech, Hearing, and Language Research Centre at Macquarie University. Portions of the research were presented at the Annual Australian Experimental Psychology Conferences in 2000 (Noosa) and 2002 (Adelaide).

Acknowledgments

We are grateful to Dr Verna Rieschild for her assistance in our reliability analysis of sentence classification procedures in Experiment 4. We also thank Elaine Funnell, Debra Jared, Kathy Rastle, and Ryan McKay for helpful comments on an earlier draft of the paper.

Notes

The research reported here was based on a doctoral dissertation completed by J. Arciuli under the supervision of L. Cupples in the Speech, Hearing, and Language Research Centre at Macquarie University. Portions of the research were presented at the Annual Australian Experimental Psychology Conferences in 2000 (Noosa) and 2002 (Adelaide).

¹ A small percentage of disyllabic words have even stress across both syllables.

² In general, stress errors are accompanied by segmental changes due to vowel reduction in stressed syllables.

³ We restricted our measure to second-syllable consistency to avoid problems with syllabification—for example, irrespective of how the word “oppose” is syllabified (o/ppose, op/pose, or opp/ose), the body of the second syllable is the same.

⁴ We acknowledge that this definition poses some problems for a small selection of words—for example, words with syllabified final consonants such as “meddle”, where the orthographic body of the final syllable does not contain a vowel.

⁵ Stress on both syllables could equate to (a) even stress on both syllables, (b) primary stress on the first syllable and secondary stress on the second syllable, or (c) secondary stress on the first syllable and primary stress on the second syllable.

⁶ Our original set of experimental items, which was based on an earlier dictionary analysis, contained equal numbers of nonwords with noun endings and verb endings (24 each). When we updated our dictionary analysis and searched the CELEX database some of our original stimuli had to be omitted because their endings were not strongly biased.

⁷ Items that were common to both tasks were entered into the analysis only once. Unless otherwise noted, separate analyses of the word sets used in each task revealed the same pattern of statistical results as did analysis of the combined item set.

⁸ This difference did not reach traditional levels of significance when data for naming and lexical decision stimuli were analysed separately, with t(16) = 1.81, p < .10, and t(28) = 1.17, p > .10, respectively.