You are viewing the SNL 2017 Archive Website. For the latest information, see the Current Website.

Poster A36, Wednesday, November 8, 10:30 – 11:45 am, Harborview and Loch Raven Ballrooms

The mental lexicon across the lifespan: Word associations from L1 and L2 speakers of Norwegian with and without dementia

Pernille Hansen1, Ingeborg Sophie Ribu1, Malene Bøyum1;1University of Oslo

Results from word association tasks may shed light on theories on the mental lexicon, on integration of L2 items into the lexicon, on changes in lexical access due to ageing and age-related cognitive decline, and on the influence of specific variables on lexical access. The current study touches on all of these aspects by investigating word associations in Norwegian from 213 participants across five groups: Young neurologically healthy L1 speakers of Norwegian (aged 20-30, n=122), elderly neurologically healthy L1 speakers of Norwegian (over 60 years old, n=51), elderly neurologically healthy L2 speakers of Norwegian (n=20), elderly L1 speakers of Norwegian diagnosed with dementia (n=10) and elderly L2 speakers of Norwegian diagnosed with dementia (10). A word association test with 100 cue words was developed, based on the methodology of Fitzpatrick et al. (2015). The cue words were nouns, verbs and adjectives randomly selected from the 2 k and 3 k bands of the NoWaC (Guevara, 2010), in order to control for the impact of word class and frequency. This list was administered on paper to the two cohorts of neurologically healthy L1 speakers of Norwegian, who were instructed to write the first word that came to mind for each of them. A short version of the test with 30 cue words was created based on the response patterns among these two groups, and administered both orally and on paper to the L2 speakers of Norwegian and the two groups of participants diagnosed with dementia. Response patterns were analysed within and across the five participant groups. Comparisons of the word associations show clear group differences both in the response frequency lists, and in the response category patterns. Differences were found between the two cohorts of neurologically healthy L1 speakers, as well as between the neurologically healthy elderly speakers and the participants diagnosed with dementia. Whereas the younger speakers mostly responded with single words, multi-word responses and blank responses were common across all four groups of elderly speakers. Responses were given on all levels of representation (i.e. related to meaning, position, and form), and there was much individual variation in the responses. These findings support a non-modular view of language, where linguistic knowledge is organised in a network. The findings also support the role of experience in usage-based theory. Further, significant differences were found between noun cues and verb cues, supporting the fundamental difference between nouns and verbs that is postulated in cognitive grammar. The results support previous findings of difficulties in the retrieval of and production of words as a result of age and age-related cognitive decline. References: Fitzpatrick, Playfoot, Wray & Wright (2015). Establishing the reliability of word association data for investigating individual and group differences. Applied Linguistics 36:1, 23-50. Guevara (2010). “NoWaC: a large web-based corpus for Norwegian”. Proceedings for the NAACL HLT 2010 Sixth Web as Corpus Workshop. Association for Computational Linguistics, 1-7.

Topic Area: Meaning: Lexical Semantics

Back to Poster Schedule