Authors:
Venagli, Ilaria, ilaria.venagli@univr.it, Università di Verona
Nardon, Andrea, andrea.nardon@univr.it, Università di Verona
Dal Maso, Serena, serena.dalmaso@univr.it, Università di Verona
Melloni, Chiara, chiara.melloni@univr.it, Università di Verona
Piccinin, Sabrina, sabrina.piccinin@univr.it, Università di Verona
Vender, Maria, maria.vender@univr.it, Università di Verona
Keywords: event nominalizations, reading comprehension, eye-tracking
Abstract:
Nominalization is the linguistic process whereby a noun is derived from other syntactic categories, especially verbs. This transformation typically involves morphological changes through the addition of a derivational suffix or internal modifications of the word. Moreover, in the case of event nominalizations, the base verb arguments can be expressed as prepositional phrases or possessives (Grimshaw, 1990). Although the presence of event nominalizations in written texts is widespread, just a few studies have empirically assessed their impact on readers’ processing and text comprehension. Overall, these have found that texts in which event nominalizations are unpacked (transposed into corresponding verbal structures) are generally read faster, with lower total and first-pass reading times (Wolfer, 2016).
In this study, 40 monolingual Italian adults (age range: 19-35) were administered a reading comprehension task to identify potential differences in reading behaviour and comprehension depending on whether the text included a nominalization realized with full argument structure or its unpacked active and passive verbal forms. Participants were instructed to read the text carefully (Screen 1) before clicking the mouse to bring up the comprehension question and the three response options (Screen 2), while the text remained available. Eye-movements were recorded monocularly at 1000 Hz.
Behavioural data revealed that participants had a ceiling performance in answering comprehension questions, with no significant differences across conditions. However, the analysis of participants’ eye movements on texts consistently showed a significant interaction between Screen and Condition (first-pass fixation duration: ꭓ2 = 65.01, p < .001; total fixation duration: ꭓ2 = 137.44, p < .001; regressions: ꭓ2 = 157.91, p < .001), indicating that differences across conditions only became significant when participants were presented with the comprehension question and were thus engaged in a more demanding process that required integrating and retrieving information from the text. This suggests that answering a question on the passage containing a nominalization imposed higher processing costs, as reflected by longer fixations and more regressions.
These findings suggest that even for experienced readers, real-time processing appears to be facilitated when events are realized as (active or passive) verbal structures rather than as nominalizations with full argument structure.
References:
[1] Gaeta, L. (2002). Quando i verbi compaiono come nomi. Un saggio di Morfologia Naturale. Milano: Franco Angeli.
[2] Grimshaw, J. (1990). Argument Structure. Cambridge, Mass.: MIT Press.
[3] Wolfer, S. (2016). The impact of nominalisations on the reading process: A case-study using the Freiburg Legalese Reading Corpus. In S. Hansen-Schirra & S. Grucza (eds.), Eyetracking and Applied Linguistics (pp. 163– 186). Berlin: Language Science Press.