mglory@hutchtel.net wrote: > > > > Are these census errors and bad indexing in recent discussions > > > coming from errors in the original enumerations? Or, are some or > > > all being found only in incorrect transcriptions by ancestry.com? > > > Important distinction. > > > > > > John H Ballard <jballard@dslextreme.com> > > > > <snip> > > > > We also have to remember that all the indexers have to work with is > > microfilm - and a lot of that is not clearly legible. I sometimes > > look at a bunch of dim chicken scratches and marvel that the > > indexers got ANY of the letters right - and they had them all > > correct! > > > > "Richard A. Pence" <richardpence@pipeline.com> > > As someone who has indexed a state census for just one county I want > to say that at some point the indexer has to make a decision if it > is a "P" or "B" or it would never be done. The hand writing on just > the one county census ran the course from --shall we say > ornate--to--unreadable scratches. I just had to say that. > > mglory@hutchtel.net Amen. Even the 1930 isn't a model of legibility. :( And, messing around with some 1850 census today, I offer up the following METHODS-y tips: Wise may be indexed as Urise, because in at least ONE instance that's what it looks like. Now, anyone doing this for himself would *assume* that a Urise in the midst of 8 WISE would more likely be WISE than Urise, but paid indexers are paid to think, they're paid to type. Likewise a LONG family's 3rd child was surnamed VONY or possibly VONG. a BULLITT family is probably indexed as BULLELL because the bar on the T is in the next box. A GEORGE family has a child whose surname is LEORGE ... I have no idea what the name UUmrl should be, but I'll bet it's NOT Uumrl Cheryl singhals@erols.com