Thursday, December 13, 2012

Watch Out for Early Dates in Ancestry's "Massachusetts Town and Vital Records, 1620-1988" Collection

One of the very best record collections on is the "Massachusetts Town and Vital Records, 1620-1988" set of town record books.  These are, usually, the original source for most of the vital records that appear in the published town vital record books and in the International Genealogical Index database.

The database was indexed by, which was a monumental task because of the really poor handwriting, or page blotting and damage, on many of the pages from these record books.

There is another significant problem with some of the indexed dates:

Most genealogists are aware of the change from the Julian calendar to the Gregorian calendar that occurred in September 1752 (see for a detailed explanation).  Many Massachusetts town clerks did not use month names in the 17th century, but rather wrote dates like "12 (1) 1647" denoting the 12th day of the first month of 1647.  The first month was March, and the second month was April, etc., down to the 11th month being January and the 12th month being February.

Consequently, we see entries in the records like this:

My target here is Sarah Larkin, who is the 4th person from the bottom of the left-hand page, with a birth date of 12 (1) 1647.  That should be 12 March 1647.  As you can see, there are entries in early 1647 for 11th month (meaning January) and 12th month (meaning February) above Sarah's entry.  The clerk who transcribed this record book helpfully added "/47" to the year for those entries.

How did the Ancestry indexers index the date?  You guessed it - Sarah's is indexed as 12 January 1647 instead of 12 March 1647, as you can see in the screen below;

So, it appears that some of the early colonial vital record entries in the Ancestry index are incorrect.  It is impossible to know, of course, without looking at the actual records to determine if the month was represented by a number or written out.

Consequently, many persons who blindly attach these records (which are almost the only actual "historical records" available on Ancestry for Massachusetts colonials) will get the dates wrong for these persons.  And proliferate them in their Ancestry Member Trees.

My guess is that there is absolutely no way to correct the indexes at this point in time.  Each image with records before 1752 would have to be reviewed and the index corrected.

The URL of this post is:

Copyright (c) 2012, Randall J. Seaver


Daniel Dillman said...

Frankly, judging by the speed and success of the 1940 US Census indexing project recently completed, I've been thinking it would be a great idea to start re-indexing some other records as well, perhaps other censuses, as many of the indexes I've come across in my searching have been really poor. Sometimes this is due to poor handwriting or spelling on the part of the original census taker. Sometimes it's due to lousy transcribing on the part of whoever did the original indexing. Either way, I think the way they did it with the 1940 Census, using a pair of indexers and an arbiter on each record, would yield much improved indexes for other records as well.

Marian Pierre-Louis said...

Very good catch, Randy! Thanks for posting this.

Dona said...

Randy, I'm sure you'll point this out to Ancestry next time they ask, or sooner, please! I totally agree, Daniel, that the model using two indexers and one arbiter is the standard to use. And it's always helpful to use indexers who are familiar with the area and the surnames there when indexing. It certanly leads to more reasonable results.

Unknown said...

You can always make correction on incorrectly indexed records on I always try to do so as I come across them in my own reserarch. You can explain your reasons for the change. I agree they should be re-indexed, but, in the meantime, some of the records will reflect accurate information. Melissa

Heather Rojo said...

This has been a pet peeve of mine for New England and Quaker genealogy research for many, many years. Thanks for blogging about it!

M Murphy said...

Ancestry should make it easier to edit incorrect entries

Geolover said...

Thanks for this very dandy Randy-post.

However, those who wish for re-indexing by should be very careful what they wish for.

Since June they have been reworking the Drouin Collection indexes, in the process mangling place/parish names, putting places in the wrong groups (the rare and valuable 1750s Fort DuQuesne records now placed in "Acadia" where it decidedly was not), and not co-ordinating this reworking at all with the drop-down menu of place-names in the search form. Records are now much, much harder to find.

A much more minor recent reworking has deleted a lot of information from the extracts from the wrongly titled (in the original) _Calendar of Sussex County, Delaware Wills_. This volume abstracts loose papers from estate files as separated and refiled by the Delaware Archives. They include Administrators' Bonds, Inventories, Accounts and other items in addition to wills. But the reworking now calls each document a will, which is very confusing in addition to just plain wrong in the majority of entries.

Barbara J. Mathews, CG said...

My pet peeve is that most towns have several volumes of records, but that Ancestry's presentation doesn't permit the user to figure out which volume the image is in.

This was based on a microfiche collection. I've used the fiche which clearly are marked and organized by volume. When I access them on Ancestry, I get "image 2431of 5861" or similar. I can still find the early images with the overview of the volumes in each town. I just can't figure out where volumes start and end on Ancestry.

Where this gets irritating is in towns that have transcriptions of earlier volumes. I found a town in which the modern transcripts were indexed but I couldn't find the corresponding "hit" in the original volume.

Accessing them on Ancestry is easier than driving to Boston Public Library's Microtext Department. I just find my hands tied in that part of my job which is to know what I'm using.