Who is dominating the Dutch neighbourhood? On the role of subsyllabic units in Dutch nonword reading

Abstract

To assess the role of the subsyllabic units onset–nucleus (ON; spa rk) and rime (sp ark ) in Dutch visual word recognition, we compared lexical decisions to four groups of nonwords in which the existence of ONs and rimes was orthogonally manipulated. Nonwords with existent ONs and/or rimes were rejected more slowly and less accurately. ON and rime neighbours thus influence Dutch nonword reading to the same extent. Simulations with the interactive activation model (McClelland & Rumelhart, 1981) revealed that this model with left-to-right coded representations could not replicate the effects found in the lexical decision data whereas an adapted version with representations of onset, nucleus, and coda could. Effects of the larger units ON and rime emerged from activation patterns created by the smaller units onset, nucleus, and coda.

Keywords:

• Subsyllabic units
• Rime
• Orthographic neighbourhood
• Lexical decision
• Interactive activation model
• Simulations

Acknowledgments

This research was made possible by a grant of the Flemish Fund for Scientific Research (FWO) to Dominiek Sandra. The item by item data are available upon request from the first author.

Notes

1 Word frequencies were taken from the list of Dutch word forms in CELEX (Baayen et al., 1995).

2 As a consequence of the orthogonal design, each ON and rime unit was presented twice. An ANOVA revealed that the repetition factor did not interact with either the ON effect or the rime effect (all ps  >  .05).

3 The correlations in the ONC coding condition remain very similar for different parameter settings—for example, latencies, r = .538 with a lower word-to-ONC excitation (0.15); latencies, r = .420 with a lower ONC-to-word excitation (0.03) and inhibition (0.03); latencies, r = .519 with a higher ONC-to-word excitation (0.06) and inhibition (0.06).

4 A Levene's statistic showed that the variance between the four conditions was heterogeneous, F(3, 74) = 6.6, p < .001. Therefore, we also performed a nonparametric Mann–Whitney U test on the ON and rime existence factors. These tests also show a clear effect of both ON existence (U = 340.5, p < .001) and rime existence (U = 295.0, p < .001).

5 Because of the heterogeneity of variance between the conditions (see Footnote 4), we performed a nonparametric alternative analysis for the ANCOVA. Mann–Whitney U tests on the residuals of a regression with neighbourhood density and frequency revealed a clear effect of both ON existence (U = 408.5, p < .001) and rime existence (U = 334.5, p < .001).