An example of the identity problem begins with a book transcribing funeral invitations including one for a Mrs. Nancy Eckles in 1839. The transcriber attributed it to the first wife of Charles Eckles of Georgetown, Scott County, Kentucky. (Marie Dickore, compiler, Copies of Names on Invitations to Funerals and Burials In Scott County and Fayette County, Kentucky, 1821 - 1898 (Cincinnati, Ohio: n.pub., 1942), 8.) Based on other evidence I came to believe his wife had died earlier and that the 1839 death was really his mother. On a recent trip to Kentucky I found an item seeming to confirm that Charles' wife died earlier (see below.)
John and Martha both signed with their mark. John Boyd had been the most likely candidate for her father, so it is especially rewarding to find that he apparently was the one. Information on nicknames (History and Genealogy Unit, “A Listing Of Some 18th and 19th Century American Nicknames,” Connecticut State Library (http://www.cslib.org/nickname.htm: accessed 31 August 2011).) claims that Nancy could be a nickname for Agnes. The consistent list of names over the three wills seems to agree with that. The mother's will uses a number of nicknames for her other children as well. Another document tying Nancy to John Boyd is a guardian report by Charles Eckles in Scott County, KY. (Charles Eckles guardian report (recorded 17 January 1831), Scott County Will Book E, Part 2: 161, Scott County Clerk's Office, Georgetown, KY.) He received payment from the estate of John Boyd on 27 January 1830 as guardian of Mary and Nancy Eckles. They were his daughters. This seems to imply that their mother, the named heir of John Boyd, was probably deceased by this time and her daughters were receiving her inheritance.
Next question: was her mother-in-law really Agnes too?