Saturday, May 28, 2011


© Kathy Duncan, 2011

Two researchers contacted me within the last week asking about the deaths of Martha (Sales) Duncan and her son William who apparently died on the same day. Their deaths on the same day suggests illness or an accident. Either would make for an interesting story. Maybe they died in a house fire or overturned wagon? Maybe they died in an epidemic? After sifting through my information, I’ve come to the conclusion that they did not die on the same day. No interesting story.

According to find-a-grave, both Martha and her son died on October 9th, Martha in 1869 and William in 1865. Both years are incorrect.

First, let’s look at the records of Martha (Sales) Duncan. Her tombstone in Coop Prairie Cemetery in Scott County, Arkansas was broken before this picture was taken in the late 1970s to early 1980s. The damage runs right through her year of death. In Cemeteries of Sebastian County, Arkansas, Wanda Gray transcribed Martha’s date of death as 1865. Larry Duncan who lived in Waldron and was a frequent visitor to Coop Prairie also transcribed Martha’s date of death as 1865. Since Martha’s husband Isaac Duncan married Susan P. (Reese) Hodges in 1867, Martha’s year of death would need to be well before 1869. Therefore, 1865 seems like an acceptable date of death for Martha. We can guess that the death years for Martha and her son William may have been transposed on find-a-grave.

Next, let’s consider what is known of William Duncan. His records are more problematic, and an easy conclusion can not be drawn at this time. Currently, find-a-grave links a tombstone for a William W. Duncan as the son of Martha (Sales) and Isaac Duncan. The entry states that William Duncan was born 6 Jan 1859 and died 9 Oct 1865. However the photograph of the tombstone clearly shows a death date of 2 Dec 1875. We could chalk that up to an recording error, but it is still not easy to link this William Duncan to Martha (Sales) and Isaac Duncan. The birth date on the tombstone is in conflict with two other records. The first conflict is with the birth date for William’s older sister Sarah Ann Duncan. Her tombstone states that she was born 7 Jan 1859. Were they twins born one day apart? No. Sarah’s birth date is consistent with the 1860 census which states that she is a one year old. This leads us to the second conflict. William’s age on the same census is 5/12 months; his age is off by one year from the birth date on the tombstone. In order for the tombstone to be William’s, the year on the stone should be 1860 not 1859. Granted, the tombstone could be William’s if the stone was carved incorrectly. Was Isaac Duncan too overcome with grief to remember the exact year of his 15 year old son’s birth? Did he write the date incorrectly or eligibly so that the tombstone mason refused to correct the error?

In order for the tombstone to belong to Martha’s William, we also need to be able to prove the he was still living in 1870. The 1870 Sebastian County, Arkansas census lists a ten year old William living with his father Isaac and step-mother Susan Duncan, so the stone could be William’s.

Note that neither census provides a middle initial for William Duncan while the tombstone provides a “W.” as a middle initial. If William were still living in his father’s household in 1880, it would be easy to decide if the tombstone belonged to him. Indeed, the 1880 census does not list a son William living with Isaac. That might suggest that William was deceased before 1880 except that there is 20 year old William Duncan, born in Arkansas, boarding with George Taylor in Palarma, Faulkner County, Arkansas. No birth place for parents is provided with this William Duncan’s entry, so it is impossible to rule him out as Martha (Sales) Duncan’s son.

We are left with questions. Is William Duncan’s tombstone in error? Did he die shortly before his 16th birthday? Or did he live well beyond 1875?