Growing old in Sheffield: Insights from longitudinal semantic data
The study of language use across the lifespan allows answering a number of intriguing questions regarding trajectories of language change, or sociolinguistics of aging processes. Longitudinal studies so far demonstrate that: (1) individuals may display patterns of stability; (2) individuals may change in later life in the direction of a community-level change with older speakers adapting to patterns typical of younger speech; or (3) speakers may display retrograde change in later life, with older speakers reverting to earlier community patterns as they age (Sankoff 2013).
Most of these insights are based on a modest, but growing body of longitudinal research in grammar or phonology. When it comes to the lifespan use of lexis, linguists occasionally refer to anecdotal evidence, such as that older speakers keep on using older expressions, but such lexical data is not systematically interrogated. In this context, my talk will show how the understanding of processes of lifespan change can be furthered by considering the evidence from the use of words.
In my recent research I have traced the use of fifteen polysemous adjectives, such as awesome, skinny, and gay in Sheffield between 2005 and 2015. The results provide support to the overall stability of semantic variables across the life of individuals, with indicators, such as skinny, being the most stable. Data on variables, such as wicked or cool indicate that speakers are slightly changing towards the usage of younger generations. The data also shows that speakers’ awareness increases over the course of change and this is particularly seen in cases of awesome and also, to some extent, gay. My study adds to previous observations in that speakers can change in later life by rejecting the use of a given adjective with all its senses, and not necessarily by reverting to previous ‘pre-change’ usage (3). This happens in a situation when the community-level change increases the use of a meaning variant that individuals do not accept. The analysis of individual variables is supplemented with insights of the histories of individual speakers. For example, I show that speakers with child-rearing responsibilities change their semantic usage by participating in community-wide change more often than speakers without child-rearing responsibilities.
A number of observations based on this study allow me to comment on methodological intricacies of socio-semantic research and to conclude by mapping out the most fruitful lines of enquiry for future investigations of lifespan change.