Recently, a colleague of mine received a reference request that highlights a common misperception some people have regarding original records. The request involved a researcher who complained that their family name had been misspelled on the 1930 census and they wanted the National Archives to “fix” the mistake. Now, most expert researchers know that original records—and especially census records—are not perfect. They always need to be evaluated for reliability (sometimes the author’s perspective can be suspect), but many original records also contain all kinds of inherent or contextual flaws, including misspelled names and misidentified people. (On the 1910 census, for example, my own grandmother appears as a 2-year-old toddler named “Pearl” even though her proper name was Cora Marie!)
As primary source material, however, original records are still considered sacrosanct because they provide the foundation—the essential evidence—upon which all historical and genealogical research is based. From an archival standpoint, therefore, we never, under any circumstances, alter the original! To use the vernacular, the original record “is what it is.” Tampering with it in any way would compromise the very integrity and meaning of the record. In the case of the 1930 census, it would also be a physical impossibility to correct a misspelled name because the Census Bureau destroyed the originals after microfilming the schedules. (For practical reasons, we also can’t go back and change the microfilm.) Either way, the National Archives cannot “fix” mistakes on existing records. Our first responsibility is to preserve the records “as is” so they will be available to those who come after us, to be used and understood in the same context. The best we can do now is alert researchers to the possibility that mistakes exist and encourage everyone to conduct research with an open mind.