Editor's note: In celebration of the release of the 1940 census, the Heritage Hunting column over the next few weeks will concentrate on the genealogical values and pitfalls of using all census records.
It wasn't until 1870 that the government asked the marital status of each person in a household. But beginning in 1850, information is presented in such a way that we can form theories for our genealogical research.
We usually see an adult male followed by an adult female of similar age, and then a list of children. Responsible researchers won't draw a conclusion that this man and woman are married to each other, but we can form some theories that we'll use as a basis to seek more evidence to support our thoughts.
Let's explore how to build a theory so we'll know when and where to begin looking for a marriage record.
Prior to the mid-20th century, "birth control" was haphazard and often not even understood. So researchers can count on finding large families, usually with two years between each child.
This cultural knowledge helps in many cases to determine when a couple married. Look at the age of the oldest child and compute his approximate date of birth. For example: In an 1860 Alabama census, a couple (the man age 28 and the woman age 26) has four children, ages 1, 3, 5 and 7 (born about 1859, 1857, 1855, and 1853). A woman usually had her first child within the first two years of marriage, so we can estimate that the couple married about 1851 (this also would explain why we didn't find them on the 1850 census).
The 1860 census included a place of birth for each person. Suppose the two youngest children were born in Alabama. The 5-year-old was born in Georgia and the 7-year-old was born in South Carolina. Now we can estimate not only that this couple married about 1851, but that it was in South Carolina. Even though we probably won't find them as a married couple on the 1850 census, we might find the man as an 18-year old in someone else's (his parents perhaps) household. This could give us a county (and we should consider surrounding counties) in which to start our search for a marriage record.
Later censuses take some of the guesswork out of the marriage puzzle. The 1900 and 1910 censuses asked individuals to give the length of their current marriage. In 1910, the government also told census takers to indicate an M1 or M2 in the marital status column. The M1 indicated the individual was in a first marriage. The M2 meant it was something other than a first marriage. It could be a second, third, or more (so you won't find an M3 or M4, etc.).
In 1930, the government asked the age of each person at his/her first marriage.
It is important to be responsible in using marriage data from the census. If you are writing a narrative on your ancestor and the only marriage information you have is based on census research, your report should read that a couple "probably was married about [insert year]." After you've pursued the census leads to more definitive records, you can be more precise.
Net week, we'll explore how to determine relationships using clues in the census.