Genealogical Proof Standard #1: How Much Research Is Enough?

If you missed my last post, Searching Isn't Researching, I talked about why finding records and doing genealogical research aren't quite the same thing. Searching finds records. Research begins with a question and uses those records to seek an answer..

But, how much research is enough?

The first component of the Genealogical Proof Standard calls for reasonably exhaustive research. Notice the word reasonably. We aren't expected to search every record ever created. But we also shouldn't stop with the first record that seems to answer our question.

Reasonably exhaustive research means looking beyond the obvious sources and considering what other records might provide relevant evidence. That means identifying the sources most likely to address our research question—whether census records, probate files, deeds, tax lists, church registers, newspapers, military records, court records, or records involving an ancestor's friends, associates, and neighbors (FAN).

Sometimes, however, the record we really want simply can't be found.

When the Record You Need Isn't There

One of my own research problems illustrates this perfectly.

For years, I couldn’t find the first marriage record for my second great‑grandmother, Melvina. I knew her first husband’s surname was Gaylor, and that missing record might have revealed the maiden name I’d been searching for.

But it simply wasn’t there.

So I asked myself: What other records might contain her maiden name?

Melvina died sometime between 1920 and 1930 in Texas, but despite searching, I haven’t located a death certificate—one that might have named her parents. Texas required death registration during this period, but that didn’t guarantee that every death was actually recorded or that the record survived.

I turned to the death certificates of four of Melvina's children. Some of her other children died very young, before the state required death certificates.

My great-grandfather and his sister both identified their mother's maiden name as Melvina Mitchell. A third sibling's death certificate listed her maiden name as unknown. But the fourth identified her as Melvina Nicholas¹.

Two records said Mitchell. One said Nicholas. One didn't know.

Because her first husband was a Gaylor, Nicholas couldn't be explained as a married name. I had conflicting information that I couldn't just dismiss.

It would have been easy to choose Mitchell because two of the four records agreed. Melvina also lived with my great-grandfather for more than a decade before her death, which gave me additional reason to think he might have known his mother's maiden name.

But I didn't want to stop there.

Family information provided two additional clues: Melvina had a half-sister named Elizabeth Scott, who later married George Hedrick, and a brother named Tully.

These clues provided me with names and relationships to search for.

I began looking for a Melvina Mitchell using the information I already had: she was born about 1842 in Missouri. In the 1860 census of Taney County, Missouri, I found the household of John Mitchell. Living in that household were Melvina Mitchell, age 18, and Tully Mitchell, age 11.²

That caught my attention. I now had a Melvina of the right age and birthplace living with a brother named Tully.

But Elizabeth Scott wasn't in the household.

So I looked back to 1850. In Taney County, I found the household headed by J. Mitchell. There I found M. Mitchell, female, age 9; T. S. Mitchell, male, age 1; and E. A. Scott, female, age 11³.

Next, I followed Elizabeth forward. In 1860, just a few pages from the Mitchell household, I found George Hedrick with Elizabeth, age 21—the right age for the E. A. Scott I had found ten years earlier⁴.

Now I had much more than two death certificates reporting the name Mitchell. I had:

• a Melvina of the right age and birthplace
• living with Tully, the brother named in family information
• appearing ten years earlier with Tully and Elizabeth Scott
• Elizabeth later living near the Mitchells with the man she was said to have married.

Family information hadn't proved these relationships, but it did tell me who to look for and where to keep searching.

Had I stopped with the first death certificate, I might have arrived at the correct answer. But I wouldn't have had nearly as much reason to trust it.

Looking Beyond the Easy Answer

Reasonably exhaustive research also means recognizing that our ancestors didn't always make themselves easy to find. Names were misspelled. People used nicknames or middle names. Indexes contain errors. County boundaries changed. Families moved. Some records were never created; others were lost or destroyed.

Sometimes we need to try multiple spellings, search neighboring counties, or go page by page through the original images. Sometimes we need to stop searching only for our ancestor and begin researching the people around them.

We also need to remember that not everything is online. A reasonably exhaustive effort may take us beyond the large genealogy websites to a courthouse, archive, library, historical society, cemetery, church, or another repository.

Even searches that don't produce a record can be important. Tracking where we've looked—and what we didn't find—keeps us from repeating the same work and helps us decide where we should look next.

The goal is to search broadly and thoughtfully enough that we aren't overlooking evidence that could change the answer to our research question.

My search for Melvina's maiden name didn't end when I found a record that said Mitchell. Staying anchored to my research question gave me a roadmap and helped me identify the sources most likely to provide an answer—including records created by or about members of the FAN club.

Confidence comes from knowing you didn't stop with the easy answer, but followed the evidence wherever it led.

----------------------------------------------------------

^1.“Texas, U.S., Death Certificates, 1903–1982,” database with images, Ancestry (https://www.ancestry.com : accessed 1 March 2010), state file no. 3641, entry for R. F. Wickliffe, died 22 January 1946; citing Texas Department of Health, Bureau of Vital Statistics, Austin. Also, “Texas, U.S., Death Certificates, 1903–1982,” database with images, Ancestry (https://www.ancestry.com : accessed 1 March 2010), state file no. 53972, entry for Florence Addielene Moffitt, died 23 November 1950; citing Texas Department of Health, Bureau of Vital Statistics, Austin. Also, “Texas, U.S., Death Certificates, 1903–1982,” database with images, Ancestry (https://www.ancestry.com : accessed 1 March 2010), registrar no. 70225, entry for John Benton Wycliff, died 22 January 1941; citing Texas Department of Health, Bureau of Vital Statistics, Austin. Also, “Arkansas, Death Certificates, 1914-1969,” database with images, Ancestry (https://www.ancestry.com : accessed 1 March 2010), state file no. 00402, entry for Clint N. Wickliffe, died 7 February 1962; citing Arkansas State Board of Health, Bureau of Vital Statistics, Little Rock. ↩

^2.1860 U.S. census, Taney County, Missouri, population schedule, Newton, p.7 (penned), p.861 (penned), dwelling 41, family 42, John Mitchell household; digital image, Ancestry (http://www.ancestry.com: accessed 07 January 2012); citing NARA microfilm publication M653, roll 658.↩

^3.1850 U.S. census, Taney County, Missouri, population schedule, Swan, p. 346 (stamped), p. 689 (penned), dwelling 88, family 88, J. Mitchell household; digital image, Ancestry (http://www.ancestry.com: accessed 08 January 2012); citing NARA microfilm publication M432, roll 420. ↩

^4.1860 U.S. census, Taney County, Missouri, population schedule, Newton, p.1 (penned), p. 855 (penned), dwelling 4, family 4, George Hedrick household; digital image, Ancestry (http://www.ancestry.com: accessed 08 January 2012); citing NARA microfilm publication M653, roll 658.↩