Author: David Paterson

Occasionally, a historical researcher may wish to reconstruct families, but explicit documentary statements of kinship are lacking. We can sometimes elicit unstated information from court documents, such as probate records, if we:

(1) understand the customary ways in which the data in these documents were compiled.

(2) compare related records, rather than treat each record in isolation.

Case Study Summary. We examine an estate that begins with 33 slaves (19 of them children) and increases over five years to 45 slaves. During these five years, the 19 children increase to 33 children (not including an estimated 3 deaths). Only two of the original children (and three subsequent unnamed infants) are ever specifically identified in the records with a particular mother; however, by comparing and analyzing the estate's legal records, it is possible to demonstrate, with virtual certainty, the maternity of 22 more of these children, and the most likely maternity of the remaining 6 children.

Sources. The primary records for the Alpheus Beall estate (Upson County, probated 1848-1853) are in Upson County Record of Accounts Book C, pp. 40, 44, 167, 288-9, 486, 592-3. Additional, valuable details are found in the Record of Vouchers Book A, pp. _____________. A lawsuit contains some useful data (William & Elisha H. Beall vs. William J. Starling, administrator of Alpheus Beall, May Term 1853, Upson Superior Court Writ Book H, pp, 247-276). The division and allotment of slaves resulting from the suit is recorded in Upson Superior Court Minute Book B, pp. 529-31 (May Term 1853, Decree in Equity in the case of William & Elisha H. Beall vs. William J. Starling, administrator of Alpheus Beall).

Method. An effective way to study multiple slave listings from the same owner or estate is to arrange the data in tables so that information about each individual can be easily compared. Arranging in tables also helps to resolve ambiguities in identity when more than one slave has the same name, or when slaves' names are recorded in different forms. I have placed data from the Alpheus Beall records in three tables. I have ignored minor variations in name spellings when the identity was obvious.

Note for Table 1. Only column 1 of this table lists persons in the same order as the source document. The data in columns 2 through 6, originally recorded in various orders, is here conformed to the order of persons listed in column 1 for purposes of comparison.

Column 1. The "Schedule of Property in the Hands of the Temporary Executrix" (widow, Mary C. Beall), made on July 10, 1848, lists all 33 slaves by name, but does not suggest any family relationships. The list is reproduced in column 1 in the same order as recorded. Note that the people are divided into three categories: men, women, and children. It was most common in such lists for the adults to be listed in age order (for example, Sukey is presumably the oldest woman). Children might be listed in descending age order, or by family groups in descending age order � but we cannot be sure what arrangement was used until we compare this list with other data.

This list differentiates two women named Eliza as "old Eliza" and "young Eliza," but most later documents call them "Big Eliza" and "Little Eliza." The return for 1852 differentiates "Old Eliza" from "woman Eliza [with] four children," allowing us to determine that: Big Eliza = old Eliza, and Little Eliza = young Eliza.

Table 1, Column 2. Three and a half months later, on October 25, 1848, the Inventory & Appraisement (I&A) was made. The format of this list is different from the earlier Schedule. Women are listed with the number (no names) of their youngest children. No ages are given, but based on common practice, youngest children listed with their mothers might include those from infancy up to 6 years or ten years old. The average maximum age of those listed with their mothers was commonly about 8 years old. The entry for Becky is typical: "Woman Beccy & two children $950." I have summarized such entries in column 2 with a "plus" sign and the number of unnamed children followed by the valuation, so that Becky is shown as "+2 $950".

Appraised values are included in the table because they often indicate children's relative ages (the older, the more valuable, with boys generally more valuable than girls of the same age), and thus can sometimes suggest family groupings.

Notice that Julius, Calvin, and all children following Henry are not appraised individually. This means that they are among the unnamed children numbered with the women (their mothers). Notice that there are twelve named children in the earlier Schedule of July 10, but a total of thirteen unnamed children in the I&A on October 1848. This suggests that one additional child was born between July and October. For the moment, we do not know whose child this might be, but see the later discussion of Little Eliza's children.

Table 1, Columns 3 through 6. Each year, most slaves of the estate were hired out. Their earnings (or expenses) were included in an annual return. Although the amount of hire for each slave can often provide valuable clues to a slave's age or physical condition, to simplify the present analysis, the amounts of hire are omitted from this table. The number of children hired with each woman is indicated in the same manner as in column 2, along with any additional information given in the returns concerning pregnancy, births, or deaths. An "x" in the appropriate column marks persons individually named in each annual return. The report to Superior Court (May 1853) mentions a return for 1853, but that return was not recorded.

Note that girl Emily's hire is listed each year except 1849, and that, coincidentally, Big Eliza is listed with 2 children in 1849, but in 1848 and 1850 she has only one. Might Emily have been hired with Big Eliza in 1849, and therefore be her child? We will return to this possibility later.

Harriett's absence from the list of hires for 1851 is explained in the Record of Vouchers, Book A (p. 88), where she is claimed as an expense to the estate due to a long sickness that year.

There are two pieces of data for which I do not have satisfactory explanation: (1) Becky and her children were omitted from the return for 1849, and (2) the return for 1852 lists Big Eliza without any children. Absence of a person, or reduction of number of children listed with a mother, often indicates death � but this explanation obviously does not apply to Becky or Harriet, and is not supported for Big Eliza by analysis of subsequent records from Superior Court (see Table 2). The number of children listed with a mother can also decrease when a child becomes old enough to hire individually � but, in this case, the return for 1852 names no new person who might correspond to Big Eliza's child. These discrepancies might have been mistakes in the original returns, or transcription errors by the clerk who copied them into the record book.

In response to a lengthy dispute over the estate (William & Elisha H. Beall vs. William J. Starling, administrator of Alpheus Beall, Upson Superior Court Writ Book H, pp, 247-276), widow Mary C. Beall filed a statement for the April 1850 Term of Superior Court, saying, in part, that since the inventory & Appraisement, "five negro children have been born and are now living: Billy, Mary, Willmond, Martha's infant and Amanda's infant." This establishes birthdates for these children between October 25, 1848, and April 1850. Comparison with the Return for 1849 narrows the birth of Billy [Becky's William], Mary, and Martha's infant, to about November-December 1848.

[BUT, DESCRIPTION OF MARTHA'S "INFANT" IS MORE CONSISTENT WITH AN UNNAMED CHILD OF RECENT BIRTH (SAY, 1849-50). DID SHE HAVE AN EARLIER ONE THAT DIED? THIS IS CONSISTENT WITH THE PHRASE "BORN AND NOW LIVING."]

The Division. Table 2 shows the "equitable division" of the estate's slaves between three legatees, made on December 28, 1853. The persons in each lot are listed here in exactly the same order as shown in the Superior Court record. Names of women known from records to be mothers are displayed in bold letters for easier reference in the following analysis. We find that all the names of children last seen on the list of 7/10/1848 are here, except that Pathena is abbreviated to Thena, and Fredonia is called Doney. Bill, named in the 1848 list, may correspond to one of the two Williams in the 1853 division, but can we tell which one?

Let us imagine the proceedings of the two men (Thomas W. Goode and Obadiah C. Gibson) tasked by Upson Superior Court with making an "equitable distribution" of Alpheus Beall's slaves.

• Their primary consideration, dictated by law, was to ensure that the lots were as equal in value as possible.
• A secondary goal, encouraged by compassion and practical considerations, was to keep the mothers and youngest children in "family" units (When economically convenient, the term "families" might expand to include husbands or older children).

The larger the number of slaves, and the fewer the heirs, the easier it was maintain the integrity of families while dividing the estate. The common method of making up lots was to distribute the prime men first, then the women with children, then the other children, and lastly the elderly or crippled slaves.

Because Alpheus Beall had only three men, that part of the distribution was easy. It is natural to wonder whether each man was husband to one of the women in the same lot, but the records are completely silent on marriage relationships � except that Mary C. Beall (under her remarried name, Mary C. Stallings) signed a will on 3/24/1854 that suggests Amanda and Martha had husbands belonging to other owners.

What of the distribution of women and children? Do the lots made by Goode and Gibson reveal family structures?

Whose child is whose? The three lots in Table 2 list certain children's names immediately under a particular woman's name. Does this arrangement reflect a deliberate grouping of mothers and their children? Let's bring in our other evidence.

Let's look first at Amanda in Lot #2. Three children (Ellen, Peter and Jefferson) are listed directly under her. Ellen and Peter were listed sequentially in the first list of 7/10/1848 (Table 1, column 1), but Jefferson was not listed, presumably because he was not yet born. Amanda is shown to have two children in October 1848, and was hired with two children for 1849. Ellen and Peter are surely these two children. Jefferson would have been the third child, born in 1849, because Amanda is hired for 1850 through 1852 with three children. Thus, the evidence supports the following family structure:

(mother) Amanda

(children) Ellen

Peter

Jefferson born 1849

The records for Little Eliza and her children are more challenging. Look at Table 2, Lot #1, "Little Eliza and infant," and notice that there are four children listed directly under her: Sarah, Frances, Jones and Louisa. Of these, the first three were listed, in exactly the same order, in July 1848 (Table 1, column 1).

• The I&A of October 1848 lists Little Eliza and 4 children. Sarah, Francis and Jones could be three of them, and a fourth could have been born between July and October.

• Little Eliza is hired in 1849 with 4 children � let us say those same 4 children.

• Little Eliza is hired in 1850 and 1851 with 5 children. Women usually gave birth at two year intervals, so it would be unusual, but possible, for Eliza to have had a fifth child in late 1849.

• Little Eliza is hired in 1852 with only 4 children, which suggests one child had died during 1851. She bore another child "in summer or early in fall" of 1852.

Thus, by the time we get to December 1853, we are looking for Eliza with 4 older children and an infant. That is exactly what we find suggested by the list in Lot #1. Dollar values given to the children suggest their relative ages.

(mother) Little Eliza

(children) Sarah

Francis

Jones

[child] born either 1848 or 1849, died 1851

Louisa born either 1848 or 1849

(infant) born 1851

Similarly, we can show that the four children listed under Nancy are indeed Nancy's children. Bill (William) and Fredonia (Doney) from the first list (7/10/1848) are probably the two children counted with Nancy in October 1848. William's relatively low valuation in 1853 probably reflects the crippling effects of burns he suffered in 1851 (Record of Vouchers Book A, p. 86??????). Two more births are suggested by the annual returns (see Table 1, columns 4 through 6):

(mother) Nancy

(children) William

Fredonia (Doney)

Wilmond probably born 1849

Ariann probably born 1851

Martha's three children, with whom she was hired in 1851-2, appear to be Laura, Elizabeth and Nancy. A fourth, Henry, is a young infant, as shown by the fact that he and his mother are valued as a unit in the December 1853 distribution.

Likewise, it can easily be demonstrated that Dallas is woman Mary's older child, and that Mary, her younger child, was probably born in 1848 (a third was born and died in 1850).

Listed immediately following Big Eliza in Lot #1 are Emily and Chloe. Chloe was a young child inventoried in 7/10/1848, and is probably the child listed with Big Eliza in October 1848, and hired with her through 1851. We noticed that Emily was not hired independently in 1849 � the same year that Big Eliza is shown hired with two children. Emily's place in Lot #1 confirms our earlier suspicion that she, too, is Big Eliza's child.

Woman Elizabeth, who bore a child in April 1851, is distributed without any children, suggesting that her baby had died during 1852 or 1853.

Susan, listed among the older children in 1848, is distributed to Lot #3 with her infant child, probably born in 1852-3.

Ann also had her first-born child about the same time as Susan had hers (1852-3).

Let us now consider how Goode and Gibson may have finished making up the three lots. After each lot receives two mothers with their young children, we are two-thirds of the way down the list for each of the 3 lots. Goode and Gibson still have sixteen people left to distribute, including Becky and her three children, and 60-year-old Sukey, whose useful working life is almost finished. They distribute the older teenagers and young adults, including childless women and women with infants. These include Thena, Harriett, Susan and child, Elizabeth, John, Henry, and Ann and child. They are distributed according to their appraised value, regardless of who their mother might be. Now, only woman Becky, three boys, and old Sukey remain.

The three boys (Julius, Calvin, and William) must be Becky's children. Becky had two children in 1848 and a third child by 1850. Two of the older boys named among the children in 1848, Julius and Calvin, are never hired by themselves, and cannot be accounted for in any other way unless they are Becky's younger sons. Although we cannot be sure of their ages, Julius and Calvin were (by 1853) considered old enough to separate from their mother � and so, Becky's "family" of younger children was broken up to make the three lots even. Her youngest son, only about 4 years old (see Table 1), would have been kept with his mother; thus, we may safely assume that William, listed immediately above Becky in lot #2, is the third and youngest son.

Another Pattern Appears. As we determine the probable maternity of each child named in the Schedule of Property from 7/10/1848, we can see a pattern emerge. Table 3 reproduces the names of the children in the order in which they were listed on 7/10/1848. The dollar valuation after seven of the names comes from the Inventory & Appraisement in October 1848 (I will later explain what this valuation tells us). The second column shows the children's mothers: those who have been determined from our previous analysis, and those (distinguished by parentheses and italics) who will be suggested in the following paragraphs.

The order and grouping of children's names in the Schedule strongly suggest (1) that the children of each mother were listed together, (2) that the list started with children of the most senior women, then continued in descending age to the younger mothers, and (3) that the children of each mother were listed in descending age order.

The first two children on the list, Harriett and Pathena, were given an appraised value in October 1848, a sure signal that they were considered old enough to sell or distribute apart from their mother. If they were Becky's older children, then they would customarily have been listed first � just as it appears they were, ahead of their younger brothers, Julius and Calvin. Previous discussion has demonstrated that Julius and Calvin are Becky's children. In 1848, Calvin would have been Becky's youngest child, and therefore the last-listed member of her family group.

The next child on the list, Susan, was given an appraised value in October 1848, suggesting a somewhat advanced age. If our assumption about the Schedule of Property is correct, she was the oldest child in the next family group. I suggest that Susan, Josephine, Emily, John, Henry, and Chloe are Big Eliza's children, based on their grouping together, and because we have already demonstrated that Emily and Chloe are surely hers. It is reasonable that an older woman like Big Eliza (also known as "old Eliza") probably had older children (teenagers) who would have been appraised and hired separately from her, as five of these children were. Note that boys, John and Henry, are assigned greater value than the girls who are probably their older sisters.

This pattern confirms our earlier identification of child Bill. Because he is listed after Amanda's Ellen and Peter, and before Nancy's Fredonia, absent other evidence, he could have been either Amanda's youngest child, or Nancy's older child; but the 1853 allotment of "William, Nancy's son" (see Table 2, Lot #3) confirms his maternity.

Conclusion. Our first piece of evidence, the Schedule of Property (7/10/1848), gave us a list of names, but (by itself) did not tell us who was related to whom. The Inventory and Appraisement, and the subsequent four annual returns (1848-1852), defined the numbers of young children associated with each mother, but did not tell us their names. The Division of Negroes (12/28/1853) gave us another list of names, organized differently from the first list. By comparing the numbers of children expected to be with each mother, we were able to determine that the children named in the Division were listed under the names of their mothers. Then, by comparing the children's names in the Division with the earlier names on the Schedule of Property, we could see that the Schedule was arranged to group children in "families," and in approximate age order. Thus, the older children who were not kept with their mothers in the Division, could still be matched to their most likely mother. Throughout the analysis, we used appraised values of slaves to guess their approximate relative ages, and used evidence in the annual returns to estimate birth dates.

Upson County Page last updated: Sunday, 24-Feb-2002 18:41:33 MST