The 1870 Project

A Genealogy Workflow for Record Dead Ends

July 30, 2026 · 6 min read

A missing ancestor is not a failed search. It is a research problem with a specific cause: the name changed, the county boundary moved, the record was indexed badly, or the person appears only through someone else. A genealogy workflow for record dead ends gives you a way to diagnose that problem before you spend another night opening the same census result, cemetery index, or family tree.

This matters most at the 1870 brick wall. For many African American families, 1870 is the first federal census after Emancipation that records formerly enslaved people by name. The next step backward is rarely a simple name search. It is a disciplined shift from searching for one person to reconstructing a household, a place, and the people who had legal or economic ties to that family.

Stop searching harder. Define the dead end.

A dead end needs a one-sentence research question. “Find my great-great-grandmother” is too broad. “Identify the parents of Sarah Johnson, age 22 in the 1870 census of Lowndes County, Alabama, who may have used a different surname before 1865” gives you something you can test.

Write down the last record you trust, not merely the last record you found. Capture the person’s name as written, estimated birth year, race or color designation if recorded, household members, neighbors, county, and source date. Then state what is unknown and what would count as proof.

For example, your goal might be to connect a 1870 household to a possible enslaver’s 1860 probate file. A useful result is not just a matching surname. It is evidence that links the same people, place, ages, relationships, or property history across the two periods.

This small reset prevents one of genealogy’s costliest habits: treating every search result as equally promising. They are not. A record that adds a new county, a spouse’s maiden name, or a child’s birthplace may be more valuable than another copy of the same census fact.

Build a case file before opening another database

Your research file should make contradictions visible. Keep it simple enough that you will actually update it. For each person and household, track these four categories:

  • Confirmed facts, with the record and date that supports each one
  • Open questions, stated as specific research problems
  • Name variations, including nicknames, spellings, initials, and possible prior surnames
  • Geographic clues, such as counties, districts, churches, plantations, military posts, or migration routes

Also create a short timeline. Census ages drift. Birthplaces change. A person can be listed as Black, mulatto, or white in different records. None of those details automatically disprove an identity. They tell you where to look for corroboration.

For formerly enslaved ancestors, pay close attention to the whole 1870 household and its neighbors. The older adults may be parents, siblings, former community members, or people who left the same plantation after the war. A neighbor’s marriage record, Freedmen’s Bureau labor contract, or death certificate can solve a question that your direct ancestor’s records cannot.

Use place as your primary search tool

When a name stops working, geography often keeps the case moving. County-level research is especially powerful because records were created locally even when the event affected a wider family network.

Start with the county in the earliest known postwar record. Then identify its parent county if boundaries changed. Check adjoining counties, because families crossed county lines for work, marriage, church, and safety without moving far from home. If your ancestor appears in a city, identify the ward, district, and nearby rural counties rather than treating the city name as the entire location.

For a pre-1870 African American research question, the county can point you toward probate courts, deed books, tax lists, plantation records, Freedmen’s Bureau field offices, Freedman’s Bank records, and federal microfilm collections. Those sources may not name every enslaved person. But they can expose the people and institutions around them.

This is why a vague instruction to “search more archives” is not enough. The useful next step is narrower: search a specific county’s 1864-1867 labor contracts, inspect a named estate’s probate inventory, or review a particular field office’s marriage registers. Precision creates momentum.

Separate identity clues from proof

A surname match is a clue. A shared county is a clue. A DNA match is a clue. A family story is a clue. Each can be meaningful, but none should carry more weight than it can support.

Build a cluster of evidence instead. If a possible enslaver lived near your ancestor in 1870, appears in the same county’s 1860 slave schedule, and has a probate record naming people with matching given names and ages, the case deserves serious attention. If DNA matches also descend from that family or from connected families in the same place, the direction becomes stronger.

Still, be careful with assumptions. Enslaved people did not always take the surname of a former enslaver. They could use a parent’s surname, a spouse’s surname, a name adopted before or after freedom, or different surnames across records. A shared surname can open a search path, but it cannot close the case by itself.

The same caution applies to online trees. Treat them as leads to verify, not evidence to copy. A confident-looking tree with no documents can send you into the wrong county for years.

Bring DNA into the workflow at the right moment

DNA is most useful when it answers a focused geographic or family-network question. Looking at hundreds of individual matches without a plan can create more noise than answers.

Begin with your strongest matches that have trees, shared matches, or known family locations. Group them by recurring surnames and counties. Ask: Which place appears across several match groups? Which family lines overlap in the county where my ancestor first appears? Are there clusters tied to a likely former enslaver, neighbor, or migration destination?

DNA cannot tell you an ancestor’s name from a match list alone. It can, however, identify the family network worth testing in the records. That distinction matters. Use DNA to prioritize an archive path, then use documents to establish the relationship.

The 1870 Project’s DNA Decoder is designed for this exact handoff. Rather than leaving you with a screen of centimorgans and surnames, it translates match patterns into concrete directions such as likely counties and the federal record sets to check. The goal is not more information. It is the next best record to search.

Expand sideways before going backward

If your direct ancestor disappears before 1870, search their children, spouse, siblings, and neighbors. This is not a detour. It is often the fastest route to an earlier generation.

A child’s death certificate may name parents. A sibling’s marriage license may reveal a maiden name. A widow’s pension application may preserve testimony about a marriage, military service, or place of birth. An adult child’s obituary may identify a county or church that never appears in the ancestor’s own records.

Interview living relatives early, especially the oldest relatives who can identify photographs, cemetery plots, nicknames, migrations, and family tensions that never reached official documents. Ask narrow questions: Who was raised by whom? Which relatives stayed behind when the family moved? What church did they attend? Did anyone say the family came from a particular plantation, town, or county?

Record the answer as a lead, including who shared it and when. Memories can be incomplete, but they often preserve the precise clue that gives a document search its direction.

Set a stopping rule for every search path

A workflow protects you from chasing a lead forever. Before you search, decide what you are testing and what would make you pause. For example: “I will review this estate’s probate packet for the surnames, given names, ages, and associates in the 1870 household. If none appear, I will move to the neighboring county and test the next landowner cluster.”

Log negative results. Knowing that a marriage register, tax list, or probate file did not contain your family keeps you from repeating work and helps you see which possibilities are losing support. Negative evidence is not empty space. It narrows the map.

Then choose one next action, not ten. Order a record, inspect a microfilm roll, map a neighbor cluster, contact a cousin, or test a DNA-linked county. Complete it, log what changed, and let that result determine the next move.

Your family was here. The record trail may be broken, misfiled, or waiting in someone else’s name, but it is not beyond research. Give each dead end a precise question, follow the evidence through people and places, and keep moving one defensible step at a time.

More in Archives and search skills