Academic journal article Psychonomic Bulletin & Review

Form and Meaning in Early Morphological Processing: Comment on Feldman, O'Connor, and Moscoso del Prado Martín (2009)

Academic journal article Psychonomic Bulletin & Review

Form and Meaning in Early Morphological Processing: Comment on Feldman, O'Connor, and Moscoso del Prado Martín (2009)

Article excerpt

Feldman, O'Connor, and Moscoso del Prado Martín (2009) reported evidence for differential priming of semantically transparent (talker-TALK) and semantically opaque (corner-CORN) morphological pairs under masked presentation conditions. The present commentary argues that these data should not call into question the theory that morphologically structured words undergo a segmentation process based solely on form, because (1) these results do not contradict existing evidence for morpho-orthographic segmentation, (2) funnel plots suggest that the lack of priming observed for semantically opaque items in this study is inconsistent with findings in the existing literature, and (3) orthographic characteristics of the semantically opaque pairs in this study (rather than semantic factors) are the most likely explanation for these discrepant results.

Behavioral evidence from repetition priming (Marslen- Wilson, Tyler, Waksler, & Older, 1994), frequency effects (Schreuder & Baayen, 1997), and eyetracking (Hyönä & Pollatsek, 1998) has converged in showing that morphologically complex words such as departure are recognized in terms of their constituent morphemes (i.e., {depart}1{ure}). Despite widespread agreement that word recognition involves the analysis of morphemic elements, it remains contentious whether morphemes contribute to lexical processing by virtue of their role in conveying the meanings of words (e.g., Marslen-Wilson et al., 1994; hereafter, morpho-semantic decomposition) or whether morphemic elements also have a privileged status at earlier levels of lexical processing involved in recognizing orthographic form (e.g., Taft, 1994; morphoorthographic decomposition, after Rastle & Davis, 2008). Critical to this debate is whether morphemic analysis is confined to items (such as departure) in which the meaning of the whole form can be transparently derived from the combination of its constituent parts or also extends to semantically opaque morphemic items (such as department) in which there is no semantic relationship between the meaning of the whole word and the combined meanings of the constituent morphemes {depart}+{ment}.

The article by Feldman, O'Connor, and Moscoso del Prado Martín (2009; hereafter, FOM) presents evidence from masked priming that semantically opaque complex words both fail to prime their stems and elicit significantly less priming than do semantically transparent items. Similar results have been reported in a number of priming paradigms, including delayed repetition priming (Marslen-Wilson & Zhou, 1999), cross-modal priming (Marslen-Wilson et al., 1994), and long stimulus onset asynchrony (SOA) paired priming (Rastle, Davis, Marslen-Wilson, & Tyler, 2000). However, whereas these previous studies were characterized by overt presentation of complex words (and thus, adequate processing time for participants to access the meanings of prime words), the priming effects obtained by FOM were obtained using a masked visual-priming method usually thought to reflect the earliest form-based processing of written words (Forster & Davis, 1984). The semantic influences observed by FOM therefore

call into question the autonomy of morpho- orthographic from morpho-semantic processing and the universality of the form-then-meaning assumption within models of word recognition. (p. 688)

This conclusion, in particular, runs contra to perhaps all accounts of the cognitive and neural stages involved in word identification, which propose that the initial stages of word recognition probed by visual masked priming are largely independent of word meaning (see, e.g., Dehaene, Cohen, Sigman, & Vinckier, 2005; Norris & Kinoshita, 2008). Must the form of written words be processed prior to the meanings of those words being accessed? Or is the meaning of written words processed concurrently with their orthographic form?

In this article, we consider whether the data presented by FOM merit rejecting form-then-meaning accounts of morphological processing (in particular) and word recognition (in general) by (1) reviewing prior empirical evidence that led to the proposal of form-based morphoorthographic segmentation; (2) assessing whether the effects reported by FOM are consistent with this prior literature, using funnel plots (graphs of sample size against effect size) derived from a meta-analysis of masked morphological priming studies (Rastle & Davis, 2008); and (3) considering whether methodological aspects of FOM can explain their results. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.