Group Thinking: Cluster Research

By Emily Anne Croom Premium

Sign up for the Family Tree Newsletter Plus, you’ll receive our 10 Essential Genealogy Research Forms PDF as a special thank you!

Get Your Free Genealogy Forms

"*" indicates required fields

This field is for validation purposes and should be left unchanged.

You’ve heard the expression, “We’re all in this together.” Try thinking of your ancestors the same way. Just like you, your ancestors were not isolated individuals:

  • They were part of a family, with siblings and cousins, aunts and uncles, parents and grandparents. Groups of relatives often lived near each other, worshiped together, witnessed each other’s documents and were buried in the same cemetery.
  • Like you, ancestors had friends and neighbors. Often their spouses came from neighboring families or were the siblings of classmates and military buddies. They and their friends were lodge brothers or officers in the ladies’ literary society.
  • Your ancestors had business partners and co-workers, were clients of local doctors and lawyers and bought dry goods from the local mercantile store. They were known in their communities and occasionally got their names in the local newspapers.
  • Sometimes, even small communities had several families by the same surname. When you start researching, you may not know whether these same-name families were related to yours. They may have been, and it could be to your benefit to find out.

Studying your ancestors in the context of this community of relatives, friends, neighbors, associates and same-name families is practicing “cluster genealogy.”

Why study the cluster?

Studying this ancestral community — the cluster — often helps you learn more about individual ancestors. For example, if the mercantile store owner’s descendants placed his store ledgers in a local archive, you may learn what your ancestors purchased and when, whether they paid their bills on time, and whether they bought ready-made clothes or fabric for making clothes at home. Each time an ancestor shopped, you learn that he or she was alive and in that place at that time.

Or consider families of the same surname. If census records show three Smitherman families in the same community, and one was yours, you need to learn whether the others were related to yours. The Smitherman men may have been brothers, or father and sons, or cousins, or not related at all. Studying such families along with yours could help you learn such things as a female ancestor’s maiden name or where your family lived before moving to that community. They may be the key to working back to previous generations.

Sometimes studying members of the ancestral cluster is the only way to prove generational links. The process is like a journey. If you’re trying to travel from A to B and the road is washed out, you may have to travel from A to C to reach B.

On the genealogical journey, for example, you may want to identify Ann’s parents, but no birth certificate exists for Ann. However, her obituary names her brother, Cal. If you study Cal, you may be able to prove that Cal was a child of Bill and Betty and then show convincingly that Ann was also a child of those parents — going from A to B by way of C.

When is it cluster time?

Look at the cluster idea from another perspective. Successful genealogists use more than a one-person-one-name-only approach to research. If Ann’s birth, along with her parents’ names, was not registered in public records, you have to look for her parents’ names in other documents. If you look only for records that say “Ann’s parents were X and Y,” you may never find your answer. The question may require a more creative approach — branching out into Ann’s cluster to identify and study her relatives and, if necessary, her friends, neighbors and associates. You can read case studies illustrating this process in The Sleuth Book for Genealogists (Betterway Books) and the National Genealogical Society Quarterly (included with $40 annual NGS membership, 800-473-0060).

You collect part of the cluster all the time — siblings. After all, one principle of successful genealogy is to study each generation of siblings. Because you work backward in time from yourself, you gather information first on yourself and your siblings. Then, you study your parents and their siblings, your grandparents and their siblings, and so forth. Remember, genealogy is the study of lineages — links between children and their parents — one generation at a time.

At some point in your genealogical journey, the road forks, and finding information on a certain ancestor is not as straightforward as it was for more recent generations. This situation is not necessarily a “dead end” or “brick wall,” but it is time to study the cluster.

Every ancestor was unique, so the research needed to answer your questions will be specific to that ancestor. In general, however, the research questions for which the cluster is a good tool often involve an ancestor’s

  • birth or death date and place
  • spouse’s name and origin
  • children’s names
  • siblings’ names or
  • parents’ names.

Reconstructing the family cluster

To study the cluster, you need to reconstruct it — identify its members. Start with family members. Each cluster was unique, beginning with the ancestor, spouse and their children — a nuclear family. So the first step in reconstructing an ancestor’s cluster is updating a family group sheet or outline on this nuclear family, including each person’s name, vital statistics (birth, marriage and death dates and places) and spouse’s name. Expand the J charts to include the children and grandchildren in their nuclear families. This process should include documentation of every piece of information on your group sheets — list the specific record that J gave you each detail.

Family group sheets such as those in Unpuzzling Your Past, 4th edition (Better-way Books, $18.99), and The Unpuzzling Your Past Workbook (Betterway Books, $15.99) provide space to record each family’s data and the source information for every fact. Because these charts show your proof with your data, they’re great tools for your research and appropriate ways to share your data with others.

Remember, however, that family databases on the Internet, on CD-ROM or in published family histories aren’t always reliable. They provide clues to investigate, but they may or may not contain accurate details about your ancestors. Because you’ll need to build on the information you have in order to reconstruct the cluster or solve genealogical problems, you’ll need to know that you have correct information to start with. That knowledge comes through research in family and public records.

When you tackle the unknowns in genealogy, you can’t work effectively on all your ancestors at once. To study details and research more thoroughly, it helps to limit the scope of your efforts to one “focus ancestor” at a time. This narrowed framework helps you concentrate on that ancestor’s life and cluster.

If you’ve identified your focus ancestor’s siblings, make family group sheets or outlines for them as parents and for the childhood family in which they grew up, even if you don’t know the names of their parents. If families of the same surname lived in the ancestor’s county or neighborhood, include them in your search as well, for they may be relatives.

Research your known relatives and the same-surname families in records contemporary with their lives — such as newspapers, cemeteries, censuses, church records, vital records, tax rolls and land and probate records.

Sometimes this study gives you the information you seek on the focus ancestor and lets you proceed to the next research question.

For example, a study of same-surname records identified the siblings and parents of one Ferdinand Coleman. The process began with the study of probate records in Cumberland County, Va., for anybody named Coleman, whose relationship, if any, to each other and to Ferdinand was unknown. Of the records found, three wills indeed revealed his childhood family. It’s important to note that he didn’t create any of these records but was named in the records of others — his family cluster.

Ferdinand’s father’s will in 1818 named five sons, including Ferdinand, and implied a sixth. In a September 1829 will, Newton H. Coleman added two middle names and another brother. He named in his will his sisters Martha and Susan and six brothers — John Henry, William Pride, Elliott R., Ferdinand, Archer A. and Creed. Newton also mentioned his mother but didn’t name her. His mentioning her did indicate, however, that she was alive at the time, as were his siblings. The mother’s will in 1850 named her four deceased sons, three living sons, one living daughter and two grandchildren. Her will also provided her name, Elizabeth.

In addition to the three wills, a deed (land) record of John A. Allen indicated a third Coleman daughter, John’s wife, who died before 1825. So, together, the wills and the deed record confirmed the seven sons and three daughters of Elliott G. and Elizabeth W. Coleman, including Ferdinand. One record alone would have given only a partial picture of the family.

Beyond the family cluster

Once you’ve studied your focus ancestor’s family cluster, you’ll know whether you also need to reconstruct the cluster of friends, neighbors and associates. At this point, if your research question is still unanswered, you should use the same kinds of records mentioned above, and any others available, to identify and study the broader cluster.

As you reconstruct the cluster, you aren’t trying to study an entire county’s population. Rather, you’re looking for people who can help you get the answers you seek. You’re looking for at least two kinds of “suspects.” One group is people who may be relatives of your focus ancestor; the other group is people who lived near, interacted with or had something in common with your ancestor. At first, you may not know who is in which group. In either case, however, your task probably involves a process of elimination as you study and narrow down the list to those who seem to be important in your particular quest.

For example, in census records, neighbors are usually those in the same “enumeration district” (1880 and after) or listed on the two to 10 pages on either side of your family. Sometimes, the closest neighbors are separated by several pages in the census because of the route the census taker used in visiting the families. As you read these pages, note especially other families of your ancestor’s surname or families with similar birthplace patterns or other details in common. These families may need to be on your cluster list.

If, for instance, your family lived in rural Tennessee in 1860 and all the neighbors were born in Tennessee, you may not learn much from this record alone, but you can begin getting acquainted with the neighborhood. If, on the other hand, several neighboring families, including yours, reported the parents born in Virginia and children born successively in Virginia, North Carolina and Tennessee, make note of their entries. You need to study them: Did they migrate together? Were they related? Remember, the cluster may contain people of any surname, not just your focus ancestor’s.

Among the sources that can help you reconstruct and study this cluster of friends and associates are tax, land, probate, marriage, vital and court records of ancestral counties, along with federal, land and military records. For example, some tax records were compiled alphabetically, but others give clues to neighborhoods — and thus, neighbors. Tax records may show a person who paid taxes for another as his “agent.” Such an agent was often a relative or close friend. So the roll provides a valuable clue to investigate in other records.

Many marriage records show the names of witnesses or sureties; they were often related to the bride or the groom, even if the surname was different. Witnesses to land or probate records may or may not have been relatives but become part of the cluster to study. Executors of wills often were relatives or close friends. Court records may show ancestors joined with others as defendants or plaintiffs in lawsuits; these people are also part of the cluster. Names of cluster members come from many kinds of records — you never know which ones until you look.

The cluster may be large or small, depending on the research question and the records from which you draw the names. -One census cluster I studied began with 274 same-surname families. The list narrowed as I checked off those who didn’t match the criteria of the family I was trying to find. Another cluster began with 37 men who interacted with an ancestral family in county records. A cluster could be as few as two or three couples you suspect of being an ancestor’s parents.

More clues from clusters

You can also use the cluster approach when you’re researching an ancestral surname, especially a common one, and you discover several people by the same given name. If you have several people named Richard Williams, for example, how can you sort them out to determine which is your ancestor? One way is to study each Richard’s cluster in whatever records are available:

  • Research the children and siblings connected with each man.
  • Study the wives and their families.
  • Identify their neighbors. Sometimes, repeated contact with certain neighbors and friends, as shown in the records, helps you sort out the Richards.
  • Study census, tax, land, probate and other records for all the Richard Williamses in the location where your ancestor lived.

In addition, signatures and marks individuals made on records sometimes help separate people of the same name. Although county record books rarely show the actual signatures of ancestors, they indicate whether the participants signed their names or made their marks. While one Richard Williams may have signed his name consistently, another may have marked his documents with a capital R. This difference could help separate the two men in the records.

Still another application of the cluster principle helped me solve a census problem — finding my grandmother as a child in the 1910 and 1900 censuses. I knew she lived with her widowed grandmother in a small Texas town, but I didn’t find a 1910 Soundex entry for her grandmother, Susan Mood. (Soundex is an index based on the sounds in the surname. See Unpuzzling Your Past for more information.)

Since the two females had different surnames, I looked next for the child’s name in the 1910 Soundex. It took longer than expected because the Soundex, based on the census, listed her by her middle name, not her given name. I hadn’t expected her to be listed that way and didn’t think to look for the middle name the first time I tried.

Because she had a different surname from the head of her household, her grandmother, her Soundex card was an “individual card” rather than a “family card.”

It showed she was living with her grandmother, Susan Wood. (No wonder I didn’t find her in the Soundex.) With the information from the card, I located the two females in the census, with Grandma Susan named correctly as Mood. I also noted the other families on their street and remembered my grandmother talking about these neighbors.

Next, for the 1900 census, I tried the Soundex and found neither my grandmother nor Susan Mood listed. Rather than use one of several other time-consuming options available to me, I decided to try finding them through their cluster of neighbors. I chose to look for the Booty family — a less common name meant a shorter Soundex search — and found their entry on the census schedule. Two houses away lived my grandmother and Susan Mood.

This time, Susan Mood was enumerated (and again Soundexed, I later discovered) under the surname Wood. Furthermore, my grandmother, whose surname was McKennon, was listed as McKenna. The Soundex indexes names by reducing them to a code of a letter plus three numbers. Although McKennon has the Soundex code M255, the name McKenna has the code M250, which, for Texas, was on a different roll of microfilm. I hadn’t thought of that particular spelling variation — indeed a different name — when I first looked for her and therefore hadn’t found her. It was through the neighbors that I found my two ancestors.

Seeing the two ancestral surnames in the record as completely different names taught me two valuable lessons — first, think of how the I initial letter might have been written and then copied, as in W for M, and, second, pronounce the name with the regional accent to imagine what the census taker may have heard, as in McKenna for McKennon. Even more important, this experience reinforced the value and efficiency of using the cluster in research.

Give the cluster approach a try in your own family history. You may discover that, at least in genealogy, there’s no such thing as too much togetherness.

Emily Anne Croom is the author of four books on genealogy, including the best-selling Unpuzzling Your Past (Betterway Books), now in an enlarged, updated and revised fourth edition.

A version of this article appeared in the December 2001 issue of Family Tree Magazine is a participant in the Associates Program, an affiliate advertising program designed to provide a means for site to earn advertising fees by advertising and linking to affiliated websites.