Finding, sharing and losing words: understanding the mental lexicon

Words produced spontaneously in response to cues can reveal important differences in the way individuals process language. Judicious analysis of such "word association" data can account for ways in which language acquisition, use, and attrition (dementias, aphasias) affect connections between words in our minds. Research to date has been fragmentary, but hints at individual differences in associative behaviour, and some shared group characteristics (in the associations of young or older adults, or people with dementia, for example).

However, the evidence is frustratingly inconsistent, partly because of the many influences on any individual's association behaviour, but also because methodological approaches have been unsystematic and un-coordinated.

This project addresses those challenges by applying robust, theory-driven methodologies to datasets with distinct characteristics. This will enable us to establish the role of variables, including age, heredity, first language and cognitive and sensory impairment, in shaping lexical connections and, by extension, patterns of communicative behaviour.

The project will be conducted in five stages:

i) Evidence from word association (WA) work to date will be comprehensively evaluated. This will be the first time that studies from disparate populations (children and adults, in first and second languages, neurodivergent and neurotypical participants, etc) have been synthesised to identify common themes in both the findings and the methods used. These themes will be mapped onto theoretical models of the mental lexicon.

ii) Informed by the empirical evidence and theoretical models investigated in stage one, a set of systematic, conceptually robust methodologies will be developed for the analysis of WA responses.

iii) In the third stage of the project, the new methodologies will be trialled on sets of WA data with sufficiently strong and distinct characteristics to enable differentiated patterns of WA behaviour to be detected. These existing datasets (from previous projects) include responses from -Monozygotic and dizygotic (identical and non-identical) twins

-Non-sighted individuals (and a sighted matched control group) -Young (16-years old) and older (65+ years old) adults -Individuals with semantic dementia

-Bilingual individuals and language learners providing word associations in each of their two (Welsh and English) or three (English, Welsh and Spanish) languages -Individuals with specific expertise (oncology professionals) and those without

iv) The evidence from stage three will enable us to generate new models of the mental lexicon, and to identify any systematic group-wise variations. Data visualisations and methodical analysis will offer insights into the ways age, perspective, genetic/experiential commonalities, first and second language use, and visual or cognitive impairment, influence lexical connections and therefore impact communicative effectiveness.

v) Through partnerships with health professionals and with language educators, interventions to support communicative effectiveness in cancer consultations and in language learning contexts will be designed and trialled.

By optimising the effectiveness of WA methodologies, this project will enable a robust evaluation of the contribution of WA research to investigations of the mental lexicon and of communicative effectiveness. Outputs from the project, which include a book and an online resource bank of word association data sets and visualisations, will facilitate and support future research in the field.

Engagement throughout the project with researchers and practitioners in domains including healthcare, genetics, psychology, gerontology and language teaching, will inform new thinking on the role of the mental lexicon in communicative effectiveness, and will maximise the impact and legacy of the project.