The Mental Lexicon - Volume 18, Issue 1, 2023

Most of the empirical evidence that lays the ground for research on recognition of printed morphologically complex words comes from experimental paradigms employing morphological priming, e.g., exposure to morphologically related forms. Furthermore, most of these paradigms rely on context-less presentation of isolated words. We examined whether well-established morphological priming effects (i.e., faster recognition of a word preceded by a morphologically related word) are observable under more natural conditions of fluent text reading. Using the GECO database of eye-movements recorded during the reading of a novel, we examined the long-lag morphological and identity priming in one’s first language (L1, English and Dutch) or second language (L2, English). While the effects of identity priming were ubiquitous, no evidence of morphological priming was observed in the L1 or L2 eye-movement record. We discuss implications of these findings for ecological validity and generalizability of select current theories of morphological processing.

Sentences like (1) “The singer began the album” are ambiguous between an agentive reading (The singer began recording/playing/etc. the album) and a constitutive reading (The singer’s song was the first track). The ambiguity is rooted in the meaning specification of the aspectual-verb class, which demands its complement be construed as a structured individual along a dimension (e.g., spatial, informational, eventive). In (1), the complement can be construed as a set of eventualities (eventive) or musical content (informational). Processing aspectual-verb sentences is shown to involve (a) exhaustive lexical-function retrieval and (b) construal of multiple dimension-specific structured individuals, leading to multiple compositions with agentive and constitutive readings. The ultimate interpretation depends on the biased dimensions in context. Our eye-tracking study comparing sentences in different contexts (agentive vs. constitutive-biasing) shows not only the aspectual-verb composition effect, previously reported for the agentive readings, but also a comparable processing profile for the constitutive readings, a novel finding supporting the unified linguistic analysis and processing implementation of the two readings. Regardless of reading, the composition effect is observable even after the complement has been retrieved, indicating that the fundamental lexico-semantic compositional processes must take place before context can serve as a constraining force.

This study examines speakers’ intuitions about novel word senses created through regular polysemy patterns. We investigate the effect of scalar regularity and lexical figure (metaphor vs. metonymy) on the identification of novel word senses, based on a detection experiment. It is shown that the more regular a polysemy pattern is, the less salient are the novel senses it produces, and that metaphorical patterns derive more salient novel senses than metonymic patterns. These results provide insights into the processing of novel word senses and support a non-homogeneous mental representation of regular polysemous words.

Laboratory studies on word learning in a foreign language (L2) have identified several variables involved in the learning process, key amongst them the orthotactic probability and neighborhood density of new words relative to learners’ native (L1) lexicons. More recently, learners’ sensitivity to orthotactic probability and neighborhood density relative to their developing L2 lexicons has come into focus. Past studies on word learning have largely focused on early stages of learning, in controlled studies spanning hours or days. Few studies have considered large corpora of ‘real-life’ learning data, spanning several weeks. In this study, we validate past findings outside of controlled laboratory conditions, by analyzing a dataset collected from Duolingo (Settles et al., 2018), a popular language learning app. Effects of orthotactic probability and neighborhood density observed in controlled studies persist under uncontrolled, big-data conditions for learners of Spanish, but not French. As learning progresses, we observe a previously unreported reversal of the effects of L1 orthotactic probability and neighborhood density, challenging theoretical models of word learning. Finally, we confirm the importance of orthotactic probability and neighborhood density relative to learners’ developing L2 Spanish lexicons, lending support to theories which posit that the same processes underly both L1 and L2 acquisition.

Cognates, words that are similar in form and meaning across two languages, form compelling test cases for bilingual access and representation. Overwhelmingly, cognate pairs are subjectively selected in a categorical either- or manner, often with criteria and modality unspecified. Yet the few studies that take a more nuanced approach, selecting cognate pairs along a continuum of overlap, show interesting, albeit somewhat divergent results. This study compares three measures that quantify cognateness continuously to obtain modality-specific cognate scores for the same set of Norwegian-English word-translation pairs: (1) Researcher Intuitions – bilingual researchers rate the degree of overlap between the paired words, (2) Levenshtein Distance – an algorithm that computes overlap between word pairs, and (3) Translation Elicitation – English-speaking monolinguals guess what Norwegian words mean. Results demonstrate that cognateness can be ranked on a continuum and reveal measure and modality-specific effects. Orthographic presentation yields higher cognateness status than auditory presentation overall. Though all three measures intercorrelated moderately to highly, Researcher Intuitions demonstrated a bimodal distribution, yielding scores on the high and low end of the spectrum, consistent with the common categorical approach in the field. Levenshtein Distance would be preferred for fine-grained distinctions along the continuum of form overlap.