II: Insights in the Analysis of Intra-Speaker (In)Stability
Individual and Group Trajectories Across Adulthood in a Sample of Utah English Speakers
Despite several years of research into the issue, we still cannot describe with certainty the ways people behave linguistically across the lifespan - that is, the degree to which individuals change (or vary) across adulthood, and the range of variation within which individuals operate. Since these questions deal with the behavior of individuals, they can best be investigated by the use of panels, a set of individuals tracked across real-time, thus allowing intraindividual change and/or variation to be investigated.
It is unclear, however, to what extent the linguistic behavior of individuals as tracked across their lives can tell us about the behavior of linguistic variation and change in communities, even in the same time- frame. This is in great part because of the widespread assumption (dating back at least to Lenneberg 1967) that once past adolescence, an individual’s linguistic production remains essentially stable (assuming, of course, no particular shock affecting the linguistic system).1 If this is the case, then any given individual can only give us reliable insight into the state of a local variety as spoken in the past, as the individual was acquiring their linguistic system, and not an ongoing present (see Bailey et al. 1991; Tillery and Bailey 2003; Bailey et al. 2016 for more on this).
In investigating the reality of this assumption, then, there are two issues that deserve close attention: whether speakers actually do exhibit change (or at least variation) in their linguistic systems at anything beyond a marginal level once adolescence is past, and, if such changes do occur, whether those changes occur in the same direction as the changes being exhibited in the wider community. The first of these issues questions the assumption that adults are linguistically stable, which is an issue questioned by the emerging literature on variation and change in the linguistic behavior of adult speakers (see Bowie 2005; Sankoff and Blondeau 2007; Wagner 2008; Bowie and Yaeger-Dror 2015; and others); given the findings of these studies, one would presumably expect to find variation of some sort throughout adulthood, at least for some sizable number of speakers. The second issue, however, is ar present murkier because research findings have been mixed (see also the chapters in this volume). To take just a few examples: Sankoff and Blondeau (2007) found that while most speakers in their panel stayed stable in production across adulthood, a sizable minority did not moved in the same direction as the community’s change; Nahkola and Saanilahti (2004) found that their panel generally exhibited changes across adulthood in the same direction as the community, but not for all variables; and Bowie (2011, 2015) found that individuals in his panels did not necessarily show a consistent trend (whether of stability or change) throughout adulthood and that any discernable trends did not always match what was happening in the community.