Panel studies provide a clear overall result: people as they age register lesser differences from their earlier selves than does the community over the same time interval,
as measured by a trend study. This means first, that it must be younger speakers who are
in the vanguard of change. Those adult speakers who change are (a) in the minority; (b)
concentrated in the younger age cohorts of adults and (c) make less signficant advances
than the community as a whole. The second implication of this result is that apparent
time must consistently underestimate the rate of change. I will exemplify this last point
by considering the real time trajectories from [r] -> [R] for 31 Montreal French speakers
between 1971 and 1984. Though a majority were stable, a minority of 11, mostly
younger speakers, changed over this period in the direction of the change. If the 1971
data were not available and we interpreted the 1984 values according to apparent time,
we would assume that, for example, those speakers registering 90% - 100% of the
innovative variant had begun their lives as children with those same values. However,
we know that approximately 1/3 of those speakers had significantly lower values thirteen
years earlier, and thus they actually changed over the period.