> The problem is that in most cases newspaper articles are researched to trace
> either trends or series of events...so that having every third (or whatever)
> day would be worse than useless in research! For example, suppose you were
> tracing the history of WWII, and your arbitrary selection left out June 6,
> 1944! Or stock market trends, and omitted "Black Thursday!"
> Or, worse yet, were culling an archive of the Chicago Tribune, and kept only
> a copy headlining "DEWEY WINS!"...
> Steven C. Barr
Then it wouldn't be a problem -- because SIGNIFICANT news stories have
follow-up stories. Thus, June 7, 8, 9... would cover the events of June
6th; "Black Thursday" would continue to be commented on; and it would
become clear from archival analyses that "Dewey Wins" was inaccurate. :)
Plus, you would have a cross-section of OTHER newspapers, where you
**did** have data for those specific dates -- just from different cities.
That being said: A professor here did a study on lynchings in the U.S.
South, and studied newspaper accounts to attempt a complete list of ALL
events. During certain periods, lynchings (unfortunately) were such
mundane events that they only received a two-paragraph write-up, with no
It depends on the specific research question being asked -- in which case,
those who would USE the archives -- historians, historical sociologists,
and the like -- would have VERY specific instructions on what degree of
archival retention would be in the "nice, but not necessary" realm, vs.