legacy materials

[William J Poser]

Dan Harvey wrote:
>I disagree that analysis be done later...

I agree that analysis cannot be separated from data collection.
When I said that analysis of legacy materials can be done later,
I was referring only to circumstances in which live data is
available, the point being not only that we will end up with
more data in toto but that an interaction between data gathering
and analysis is only possible when working with living speakers.
When I said that the analysis can be done later, I meant only
that since the legacy data is already "dead", someone in the future
can do as good a job of studying it as I can, whereas I can do
better working with living speakers than someone in the future will
be able to for the simple reason that there probably won't be any
in the future.

The idea that one can simply gather an unanalyzed corpus
and store it away, which some people are promoting, is
I think quite fallacious.  It encourages people to bypass
the interactive data gathering and analysis that is likely
to produce the greatest insight, and all too often it seems to
be associated with projects that expend an awful lot of time and
money to obtain a very small amount of data.

Bill