Word meaning access: The one-to-many mapping from form to meaning

Abstract

This chapter focuses on the process by which stored knowledge about a word's form (orthographic or phonological) maps onto stored knowledge about its meaning. This mapping is made challenging by the ambiguity that is ubiquitous in natural language: most familiar words can refer to multiple different concepts. This one-to-many mapping from form to meaning within the lexicon is a core feature of word-meaning access. Fluent, accurate word-meaning access requires that comprehenders integrate multiple cues in order to determine which of a word's possible semantic features are relevant in the current context. Specifically, word-meaning access is guided by (i) distributional information about the a priori relative likelihoods of different word meanings and (ii) a wide range of contextual cues that indicate which meanings are most likely in the current context. Word Meaning Access 3 Words: The Gateway to MeaningThe mental lexicon contains stored knowledge about familiar words: information about their spoken and written forms, their meanings and their grammatical properties (see other chapters in this volume). A complete theoretical account of lexical processing must specify not only the nature of these different forms of stored, lexical knowledge, but must also describe the computational processes by which these different forms of knowledge interact. Here we focus on the process by which stored knowledge about a word's form (either orthographic or phonological) maps forward onto knowledge about its meaning.Wordforms are the gateway to meaning: once a reader (or listener) has identified the particular spoken or printed word that is present in their environment they are able to access a rich store of conceptual knowledge that may include perceptual, abstract and episodic features of the relevant concept. These individual word-sized units of stored knowledge can be combined together in a hierarchical manner within sentences to communicate complex, novel ideas.A common approach in models of lexical processing is to assume that the lexicon contains abstract representations of individual words. For example, in Morton's highly influential Logogen model (1969), any given logogen becomes activated by information (perceptual or contextual) that is consistent with the properties of that specific word. Once a single logogen's activation level rises to its threshold level, the properties of that word (including its meaning) becomes available to the comprehender. This idea that individual words are represented by abstract lexical units is retained in subsequent models such as the Cohort model of spoken word recognition (Marslen-Wilson & Welsh, 1978), the Interactive Activation and Competition model of visual word recognition (McClelland & Rumelhart, 1981), and the Dual Route Cascaded model of reading aloud (Coltheart, Rastle, Perry, Langdon, & Ziegler, 2001). Word Meaning Access 4Thus, according to these models, lexical representations provide an abstract level of lexical knowledge that bridges the gap between form-based (phonological...

Diversity of narrative context disrupts the early stage of learning the meanings of novel words

“…To investigate how situational (contextual) diversity affects the earliest stages of word learning via reading, in this preregistered experiment 100 adults learned pseudowords encountered within either a single coherent narrative context, or across several different narrative contexts. Unlike previous studies, the specific semantic features associated with the pseudowords were held constant across the diversity conditions, to ensure that any observed effects were not driven by differences in polysemy, which is well known to influence how easily words are learned and processed (Rodd, 2020(Rodd, , 2022. Contrary to our predictions, contextual diversity did not influence word-form learning as assessed by both recall and recognition measures; Bayes Factors indicated moderate evidence for these null effects.…”

“…Thus in natural language, words with high situational (contextual) variability are more likely to be heterogeneous/polysemous (i.e., have more than one sense; Cevoli et al, 2021;Hoffman et al, 2013;Hsiao & Nation, 2018). It is well established that ambiguity between different related word senses (i.e., polysemy) is ubiquitous in language and has pervasive influences on both the learning and processing of words (Rodd, 2020(Rodd, , 2022Rodd et al, 2002Rodd et al, , 2004 because of the challenge associated with the one-to-many mapping from form to meaning.…”

The Role of Sleep in Learning New Meanings for Familiar Words through Stories

“…This task was designed to be a semantic analogue of tasks such as pause detection, which has been used in word-form learning studies to measure the impact of newly acquired knowledge about a novel word form (e.g., "cathedruke") on online processing of a familiar word form ("cathedral") (Gaskell & Dumay, 2003;Lindsay & Gaskell, 2013). The logic underlying this task fits within the framework of a distributed connectionist account of ambiguous word processing (Rodd, 2022;Rodd et al, 2002Rodd et al, , 2004, whereby recognition of words with multiple unrelated meanings is delayed by the competition between the words' mutually incompatible semantic representations (Rodd et al, 2012).…”

Studying Individual Differences in Language Comprehension: The Challenges of Item-Level Variability and Well-Matched Control Conditions

“…Successful comprehension of written or spoken language, requires the meanings of individual words to be activated and integrated into the wider discourse context such that a coherent mental model of the input can be constructed (Graesser et al, 1994(Graesser et al, , 1997Van Dijk & Kintsch, 1983). Given that most English word forms are associated with multiple meanings (Rodd et al, 2002;Rodd, 2018Rodd, , 2022, cues from the surrounding context must be used to activate and integrate the context-relevant meaning representation of each word (Duffy et al, 1988;Eddington & Tokowicz, 2015;Rodd, 2020). For example, the word "positive" is often used to describe a desirable event or feeling, but when awaiting a clinical test result, the same word form can take on the opposite meaning.…”