Let's suppose that I am researching my great-grandfather's life. I find him and his family mentioned in each of the U.S. Federal Censuses from 1900 to 1930 living in the same small town. Can I conclude that from 1900 to 1930 that he lived in the same small town? I would think that based on our own experience and the assumption that the four U.S. Census records show continuity, that we would most probably conclude that he did not move from the town. But what if I find a record of a child apparently born to the family not just in another town but in another state? Let's further suppose that the information on the birth of the child is complete enough to convince me that the child must be from my great-grandfather's family.
Now, I have two seemingly conflicting pieces of information. My conclusion that the the family lived in the same place for thirty years is now challenged by information that may indicate that they moved for some period of time. Let's further suppose that this child does not appear on any of the four U.S. Census records. At this point, I have four possible conclusions that account for all of the information I have discovered to this point:
- I am mistaken about the parentage of the child despite being convinced of the relationship
- The mother moved to another state to have the baby and then returned to the original town
- The child was born in the other state and then died therefore the family moved between the Census years and then returned to the original town
- There is some other explanation of the information but I lack enough information to draw a conclusion
What additional information would I need to know to support a decision to adopt any one or more of the four possible conclusions? U.S. Censuses are taken every ten years. Many children are born and die between Censuses and therefore not recorded. I could look at the U.S. Census records more closely and see if the responses from this family answer any of the questions. For example, in the Census years 1900 and 1910, the Census questions asked how many children the mother had and how many of those were living at the time of the Census. Depending on when the child was born, a missing child can show up as a difference between these two numbers. These specific questions were discontinued in the 1920 and 1930 Censuses so whether or not they might help depend on the time period involved with the child's birth.
Of course there are a lot of other records or documents that I might want to find about the family that could persuade me one way or another on this question. But this is the type of issue that frequently arises in genealogical research. At this point, I could well be convinced that I had found an additional child for the family and move on to other issues. Would I place the birth of the child into my "uncertain" category or would I decide that my conclusion was well founded and move on?
Genealogical research is highly personal by its nature. We start by basing our conclusions on what we already believe about our family and only modify our conclusions, if we modify them at all, only when presented with persuasive information that differs from our own beliefs. Many genealogists support their beliefs by using terms such as "evidence" and "proof." I could have substituted those terms in my hypothetical but the substitution of the words would not have changed the process or further validated my conclusions. If I were to say, for example, that I had "proof" that the family had a child in another state, would that really add any more certainty to the situation? When all is said and done with genealogical research, we derive on our own conclusions from whatever information we tend to believe. To a great extent, any progress we make in genealogical research is dependent on the degree to which we can question our previous conclusions and modify those conclusions as we obtain additional information. From this standpoint, genealogical and hence, all historical, investigations are tentative and open to revision.
My hypothetical situation is based on a series of actual occurrences. My Great-grandfather did live in a small town in Arizona and from time-to-time other researchers try to add children to the family. In some cases, these children have even appeared in the U.S. Census records as children in the family. Because records about the family are kept on the FamilySearch.org Family Tree and because anyone who is a registered user can add in a child or otherwise modify the information in the Family Tree, these added children can appear. How are these proposed additional children evaluated?
First of all, in any situation where there is a degree of uncertainty, it is important to look at all the facts. My investigations have indicated that my Great-grandmother had eleven children. All of the children were born between the years of 1878 and 1899 when she was 42 years old. In both the 1900 and the 1910 U.S. Census records she is listed as having 11 children with 11 living. To add another child to the family, there would have to be information as to why my Great-grandmother only listed 11 children in two different U.S. Census years. If another child was proposed born after 1910, then the information would also have to indicate how she could have had a child when she was over 52 years of age. In fact, the additional children are usually stuck into the family in between existing children.
The issue of an uncertain connection can arise in many different contexts. One common issue occurs in the context of research back in time and attempting to determine which of a number of possible candidates are a direct line ancestor. This occurs when research discloses two or more people with similar names that reside in a relatively small area. The issue usually arises in the context of a time period when there are fewer records at some distant point in the past. Resolution of this multiple individuals issue depends on the degree that the investigator uses a broader base of information.
Because genealogical research is dependent on historical records, we may come to a hiatus where there appear to be no records available. As a matter of fact, as we research further back in time, ultimately any further determination of ancestry will end because of a lack of records. With a foundling or in an adoption situation, this might occur in one or two generations but even when some records are available, ultimately, moving into the dim past, the records all disappear.
There is really no strategy to move beyond a total lack of records. But my personal experience shows that most dead-end situations are not due to a lack of records, but rather to a lack of knowledge on the part of the researcher as to additional records that may be available. Let's return to the situation where we have multiple choices for a possible ancestor. This situation usually arises due to a lack of information about the children of the people with the same names. Rather than focusing on the possible parents, the solution will usually lie in addressing the information available about the children. It is time for another hypothetical situation.
Let's suppose that I have been researching my family back into the 1700s in the Colonial America. I have traced my ancestral line to an ancestor who was born in what is now Massachusetts in 1700. I am having difficulty finding the parents of my ancestor and discover two or more possible candidates for a father. The tendency is to examine records about the potential fathers and try to distinguish them from each other sufficiently to determine which of them, if any, is the father of my ancestor. At this point, what I usually find is that I really do not have enough information about the "known" ancestor to even begin to research his or her parents. In fact, I may have made several unsupported conclusions before I got back as far as my "known" ancestor. My strategy is to "back off" of the next generation and do additional research on the last known connection. The key here is the word "known." In this context, I mean an connection that can be adequately documented from sources that were created during the time the events supposedly occurred. This principle is usually referred to as "beginning with known facts and moving to the unknown." In other words, focusing on the multiple person situation assumes that we have already obtained sufficient information about the existing generation to move on to a previous one when it is most likely that we really need more information before moving back in time.
The principle here is that in any situation where we are not sure of a connection between two or more people, the information we lack usually lies with the one more closely related to us in time. For example, in order to determine if the information about a parent is correct, I will usually begin by determining whether or not the birth information about the children is consistent and complete.
To assure ourselves that we have made supportable conclusions, we need to consult a broad spectrum of records. In my example above, I focused on U.S. Census Records. But in addition, I might point out that I found six linear feet of documents about my Great-grandfather's family in a university special collections library. To date, there are 80 sources list for my Great-grandfather on the FamilySearch.org Family Tree. Of course, if we were to move back in time, the number of potential sources would decrease and, as I pointed out already, there comes a time when there is insufficient information to form a conclusion about the next possible generation.