I'm currently researching my Evans and Partain ancestors of neighboring Hart and Elbert Counties of Georgia and Anderson County, South Carolina. My 4x Great-Grandparents, John H J and Lucy Maria Partian. I had recorded them in the 1850 census some time ago, but was reviewing their records in order to trace their children. While looking for their son Asa, I came across an alternative 1850 census listing for him. Click, Click. I was suddenly looking at a duplicate census listing for the entire family in 1850. I opened both up side by side:
Although the names are slightly different, it is the same family. The names of the family are the least of the inconsistencies however.
- John and Lucy Maria are suddenly seven years younger
- Benjamin and Sarah Adeline have switched orders in chronology
- They're both younger
- Asa and Lucy are younger
- John Hubbard is older
- Not shown, the family's property is worth $100 less
All of these changes - in only one day! The first image was record on October 8th and the second was recorded six pages later on October 9th. The census taker was the same man both times - William Steele. That I can tell by a quick look, none of the family's neighbors are repeated.
I have to wonder: why was the family recorded twice and why the differences? Why would the same census taker record the same family twice? At first I thought that perhaps the later entry was recorded as a correction to the previous one. But, when compared to the 1860 census, the first record seems more accurate. But how could the same man visit the same family in two days and not realize that he was recording them again? Dad says he must have been drunk. Another option is that he didn't actually go to the house to record the information. Maybe he came across someone on the property who gave the information, never reaching the house. Then, the other day he went to the house and didn't realize he'd already recorded the family. If he didn't see the home one day (or both) and saw different people when he questioned the family, he might not have realized what he'd done - especially given how different the information is on each listing.
Regardless of how it happened, this family stands as a good example of why census records cannot be taken as 100% proof of a family record.