When to Stop Extending a Family Tree: Records, DNA, and the Limits of Evidence

Why careful genealogy sometimes means leaving branches unfinished

Purpose Statement

The purpose of this post is to explain a research dilemma I’ve encountered while studying my Peterson ancestry: how DNA evidence, early church records, and modern indexing can push family trees further back than the surviving documentation can reliably support, and why I’ve chosen to stop where the records stop.

When DNA Goes Further Than the Records

One of the hardest decisions in genealogy is not how to extend a family line, but when to stop extending it.

As I’ve continued working on my Peterson research, I’ve run into a dilemma that many family historians face today. DNA evidence suggests connections that stretch several generations beyond what the surviving records clearly support. Online trees reflect those connections with confidence, often presenting long ancestral lines that appear tidy and complete. Yet when I slow down and look for the documents that should anchor those relationships, the trail becomes much less certain.

What the Records Clearly Support

My direct ancestor in this line is Robert Peterson (1774–1834). The documentary evidence supporting Robert and his immediate family is solid enough to work with carefully. His father, John Peterson, who married Ruth Pyles, is supported by records that can be evaluated and cross-checked.

Beyond John, however, the evidence changes in character. Instead of clear statements of parentage, I encounter repeated names, church entries with descriptive terms, and DNA connections that point in a general direction but do not confirm a specific relationship.

When Repetition Starts to Look Like Proof

In several trees, John Peterson’s parents are confidently named, sometimes with detailed ancestral lines reaching back into Europe. These same names also appear across multiple DNA-connected trees, which can give the impression of confirmation.

But repetition is not proof. DNA can tell us that people are related; it does not, by itself, tell us exactly how they are related. Without records that explicitly connect a parent to a child, the relationship remains a hypothesis, no matter how often it appears online.

A Delaware Baptism Record as an Example

A Delaware baptism record illustrates how this problem can take shape. Indexed versions of the record list:

Peter Peterson Cauponi
Spouse: Magdalena
Child: Johan Peterson Cauponi

At first glance, it is easy to read this as a family with a compound surname, or to assume that “Cauponi” is part of the Peterson name. Some researchers have carried this forward by inserting “Caupany” or similar spellings into modern family trees as a middle name or additional surname.

The record itself does not support that reading.

What “Cauponi” Likely Means

The term cauponi comes from Latin and is associated with caupo, meaning an innkeeper, tavern keeper, or merchant. In early church records, especially those written or indexed using Latin conventions, it was common to include descriptive terms alongside a person’s name. These descriptors helped distinguish individuals with common names, much like an occupation or place might be used today.

In this case, the record identifies a man named Peter Peterson, his wife Magdalena, and their child Johan. The word cauponi appears to function as a descriptor attached to the name, not as a hereditary surname. It describes something about Peter, not who his ancestors were.

The safest way to read this record is not “Peter Peterson Cauponi” as a full name, but rather “Peter Peterson, described as cauponi.”

How Descriptors Turn Into Ancestors

Problems arise when descriptive terms like cauponi are lifted out of their original context and placed into modern name fields. Once entered as a surname or middle name, the descriptor begins to look like evidence of lineage. From there, it becomes easy to extend a tree backward another generation, even though the original record never intended to make that connection.

Over time, repetition across multiple trees can give the appearance that a conclusion is well established, when in fact it rests on a misunderstanding of how the record functions.

Why I Chose to Stop

This is the point where I have chosen to stop extending my own tree. Where records clearly state relationships, I use them. Where records introduce descriptive language that later becomes treated as a surname or proof of ancestry, I document the record itself but stop short of turning that descriptor into an ancestor.

This means my tree looks less complete than many others. It has fewer names filled in and more notes explaining uncertainty. That is intentional. I would rather leave a branch unfinished than attach it to the wrong family.

Genealogy is not just about how far back a tree can go. It is about understanding what records actually say, how they were created, and where their limits lie. Knowing when to stop is not a failure of research. It is part of doing the work responsibly.

Sources

1. “Delaware, U.S., Baptisms, 1697–1886,” database, Ancestry.com (https://www.ancestry.com : accessed 13 January 2026); original data from FamilySearch, “Delaware Baptisms, 1697–1886,” Salt Lake City, Utah.

2. FamilySearch Wiki, “Latin Genealogical Word List,” explaining Latin usage and descriptive terms in church records,
   https://www.familysearch.org/en/wiki/Latin_Genealogical_Word_List
   (accessed 13 January 2026).

3. Steve Peterson, profile for John Peterson (father of Robert Peterson), “Peterson and Hay Ancestry,” Ancestry.com,
   https://www.ancestry.com/family-tree/person/tree/10236187/person/-674227071/facts
   (accessed 13 January 2026; subscription required).