QuickLesson 27: Verifying Historical 'Facts'—A Blueprint

 

We hear the caution everywhere: Don’t just accept what you see in print. Confirm it. Verify it! How, exactly, do we do that?

Verification is essential to prevent the spread of false or misleading information. It is critical to the identification of people and relationships between them. It matters for students and scholars who analyze the actions of participants in historic events and turn to genealogy websites for easy “identification” of those people. It matters for family historians who try to use genetic evidence, DNA shared-matches, and the online “trees” of those matches to break through ancestral stalemates.

The challenge for genealogists goes beyond ensuring that each ancestor is identified correctly and that one’s own personal lines are accurately assembled.  We must also invest considerable time into the “confirmation” or “verification” of every relevant fact and relationship on trees posted by every DNA match we attempt to work with.  A failure to prove every relevant assertion and generational link in each and every match’s tree will invalidate the conclusions we reach from genetic evidence.

That last bears repeating:   A failure to prove each and every relevant assertion and generational link in each and every match’s tree will invalidate the conclusions we reach from genetic evidence. Period.

Thus, two key questions:

• What constitutes confirmation or verification?
• Exactly what do we do to confirm or verify a relationship or an identity?

Whatever our research role, we all face one basic issue. The words confirm and verify are subjective. Most individuals use these two words interchangeably, and different individuals define the concept in different ways. They apply different expectations and varying levels of acceptability.

As with the injunction that reliable conclusions must be based on reasonably exhaustive research,¹ there is no quantifiable or measurable standard for confirmation or verification. At present, each of us must decide for ourselves whether a point argued by someone else has been reliably confirmed or verified. If we are the researcher, then any time we assert that we have “confirmed” or “verified” a point, we owe it to those who use our work to explain how and why we consider the point confirmed or verified. This Quicklesson will provide a blueprint to assist our verification efforts as researchers and consumers of the work of others.

Basic definitions:

Merriam-Webster supplies these generic definitions:

Confirm: to give new assurance of the validity of [something]; to remove doubt about, by authoritative act or indisputable fact

Verify:  to establish the truth, accuracy, or reality of [something]; to confirm or substantiate in law or by oath

Problems with these definitions:

Confirm. This definition of confirm is obviously problematic for historians. Few historical facts are “indisputable.” In most cases, we can only establish likelihood, based on available evidence. Similarly, the concept of removing doubt “by authoritative act” implies that a sole authority already exists for every point we investigate. 

This Authority Standard might work, or it might fail us, when we seek answers to historical questions. For example:

• Confirming an assertion that the U.S. Congressional Pension Act of 1814 authorized cash bounties for volunteers to serve in the military conflict with Great Britain is possible by locating the published act and studying its text.² (The Authority Standard works.)
• Confirming an assertion that the War of 1812 marked “the last time that Indians played a major role in determining the future of the continent” is not possible because various authorities disagree on that interpretation.³ (The Authority Standard does not work.)

 

Verify. The definition of verify is equally problematic. How exactly does one establish “the truth, accuracy, or reality” of an assertion, whether it be “fact” or opinion?

Synthesizing the collective wisdom of the Internet, the new Artificial Intelligence tool ChatGPT outlines six “key points” we can apply for verification. Chat’s advice is basic to any research process and it is sound, insofar as it goes. But it does not define the parameters we need for a Truth-Accuracy-Reality Standard and it merges two separate issues:

• our evaluation of the source (a container of information); and
• the accuracy of any one information statement made within the source (i.e., the contents of the container).

Chat’s six points are outlined below, with EE’s parenthetical observations in italics:⁴

1. CHAT: Identify the source, whether primary or secondary.⁵

EE: This is a necessary step in the research process but an identification of the source does not constitute verification of a fact. Likewise, categorizing a source as “generally good” (primary) as opposed to “more-questionable”(secondary) does not verify any fact asserted therein. Any source type or person can err. Any source can be right on one assertion and wrong on another.

2. CHAT: Evaluate the reliability of the source. taking note of bias or motives.

EE: As with Point 1, this assessment of the reliability of the source is essential but a favorable judgment of the source overall does not verify a specific piece of information.

3. CHAT: “Cross-check” with other sources “to confirm the accuracy of the information.”

EE: This recommendation implies that all facts are carved immutably in stone, that any given question has only one correct answer, and that all sources are created equal. To the contrary, if five sources all copy one mistaken author or errant record, the fact that they all agree does not verify the assertion.⁶

4. CHAT: Check for inconsistencies by comparing different sources.

EE: Again, “checking for inconsistencies” is an evaluation of the likelihood that an assertion is correct; but it does not verify the assertion.  A lack of obvious conflict does not mean an assertion is correct.

5. CHAT: Consider broader historical context.

EE: As with Points 1–4, this is a basic step in the research process, rather than the verification process.Reasonably exhaustive research is a basic requirement for any credible conclusion and that reasonably exhaustive research includes the study of broader historical context—not just a look-up of a name or fact.⁷

6. CHAT: Consult with experts.

EE: This sends us back to an observation made earlier: “Authorities” do disagree on matters of history; they contest many “facts” and passionately differ on interpretations. Finding one authority that agrees with another does not verify or confirm the assertion. It only tells us that more than one person believe that point.

So, without a Truth-Accuracy-Reality Standard, what exactly can we do to confirm or verify a “fact” (i.e., an assertion)? 

 

— A BLUEPRINT FOR VERIFYING HISTORICAL ASSERTIONS—

 

Step One: Identify the Specific Assertion To Be Verified

Before any verification effort begins, we must define precisely what we are trying to confirm or verify. Five questions are basic, starting with

1. Does the assertion we seek to verify consist of just one discrete “fact” or does it contain multiple “facts” that must be separated and verified individually?

For each fact we define, we then ask:

2. Is this an immutable fact for which only one correct answer should exist?
3. Is this an interpretation of an event or situation for which diverse opinions may validly exist?

Every assertion we attempt to verify will do one of two things: It will identify its source or it will neglect to identify its source. If a source is identified, then as a beginning point toward verification we must ask:

4. Does the source cited by our source actually exist?
5. Does the cited source actually say what our source attributes to it?

 

Step Two: Research

Following the injunction for reasonably exhaustive research, we identify and examine all possible sources that might speak to whatever we are attempting to verify. This is the point at which verification actually begins.

In this research process, we might achieve verification through one of five approaches:

1. If no source is cited, then our first challenge is to find another source that asserts the same “fact,” as a launch point for deeper research.
2. If a source is cited, then we access it and appraise its likely reliability. ChatAI’s Points 1 and 2 are helpful here. EE 1.30–1.41 provides more explicit guidance for this texual analysis. Once this analysis is made, however, we still have not confirmed or verified any assertion within the source. We’ve merely made a preliminary and superficial  judgment of the source’s validity.
3. If the cited source is a derivative source (an abstract, extract, index entry, transcription or translation, etc.), then we seek a more-authoritative source: the original record or the most-original version that survives.
4. If a source is cited and our textual evaluation suggests that it is likely reliable, then we seek other quality sources that report the same “fact.” However, each other source must be independently created from the one we’ve first used and each must be created independently from each other.
5. If no other source is found that reports the same exact “fact,” we may still be able to build a case using indirect evidence.⁸

Any conclusion we reach on the basis of these five approaches must also meet three other conditions rooted in the proof standard adopted by genealogists to ensure accurate identification of people and relationships:⁹

• We should seek and use all sources potentially relevant to that particular “fact.”
• We should resolve any conflict between those sources.
• We should create a written summation of the evidence found and explain why we believe it sufficiently supports the “fact” we are attempting to verify.

 

Step 3: Proving Identity

When we reach the point of concluding that an assertion is indeed “verified,” then one further assessment is critical:

• What proves that the person named in the source is our person-of-interest and not some other same-named individual?

This final point is the one at which many verification efforts fail and a point on which we can be easily bamboozled—especially with family trees online. Technology makes it easy to attach documents to individual profiles or “facts,” which lend an air of authority that can be both impressive and deceptive. The attachment of a document to a person or to an assertion does not constitute verification. We must still prove that the individual discussed in that document is (or is not) one and the same as our person of interest.

 

—APPLYING THE BLUE PRINT —

Case One: Verification of Identity

A distinguished historian—the Chief Historian of the U.S. National Park Service at one point in his career—edited, annotated, and published a diary maintained by a Confederate military officer.  There, we find the following:¹⁰

“Monday Nov. 16 [1863]. At Dr. Scrugg’s earnest & repeated invitation my uncle & I rode over & took dinner with him. ... At Sunset we attended the funeral of Mrs. Eliza Rachal.⁵¹”

^{51 “}Elizabeth Rachal had been born in Texas in 1838. In 1860 she made her home with the Henry Yeeters. Ibid. [The previous source to which ibid. refers is “Eighth Census, Natchitoches Parish, State of Louisiana, National Archives.”]

 

Step One: Identifying the Issue To Be Verified

This simple statement and its citation present four separate issues that need confirmation or verification:

1. Statement that “Mrs. Eliza Rachal” was buried on 16 November 1863.
2. Statement that “Elizabeth Rachal” had been born in Texas in 1838.
3. Statement that “Elizabeth Rachal” in 1860 resided in the Henry Yeeters household.
4. Assertion that, “Mrs. Eliza Rachal,” buried in 1863, was indeed the woman born in Texas in 1838, living 1860 in the Henry Yeeters household.

This historian’s academic and professional credentials are impressive. On their face, they generate trust. He was trained in (and had significant experience with) reading manuscript materials such as this diary. His attached citation informs us that he did outside research to identify the person being discussed. Our next step is to examine his cited source:

 

Step Two: Research

When we consult the source cited by editor Bearss in his reference note, we find this:

1860 U.S. census, Natchitoches Parish, La.

Page 1: Town of Natchitoches, 25 July 1860

No. 5/5 Henry Jeeter, 40, male, overseer, $450 personal property, born Miss.

                Susan Jeeter, 27, female, born Texas

                Cornelia Jeeter, 8, female, born La.

                Charles E. Jeter, 4, male, born La.

                Virginia Jeeter 1, female, b. La.

                Elizabeth Rachal, 22, female, b. Texas

 

Assessment: Census data

This census entry supports Bearss’ assertions that one Elizabeth Rachal appears on the 1860 census as a female, born about 1828 in Texas, living in the household of Henry and Susan Jeter. We can accurately say that we have “confirmed” those points made by Bearss in his editorial note. However, the cited census does not verify or confirm the “facts” that this Elizabeth was called Eliza and that she was buried three years later with Lt. Poché in attendance. Conversely, it does not verify or confirm Bearss’ assertion that the woman whose burial Poché attended on 25 November 1863 was the woman of the Jeter household who was born in Texas in 1838.

Analysis of both the census data and the diary entry against context from elsewhere (ChatBPI’s Point 5) raises three flags that also cast doubt upon the asserted identity of the buried woman:

• This 1860 Yeeters household was in the town  of Natchitoches, the parish seat. According to Poché’s diary entry, he “took dinner” with Dr. Scruggs and then “at Sunset” went to Mrs. Rachal’s funeral.

Dr. Samuel E. Scruggs was a resident of the village of Cloutierville, about 25 miles downriver from the town of Natchitoches.¹¹ Between “dinner” (typically the noon meal in this society) and “sunset” on that November day (4:51 pm¹²) the 1863 diarist could not have traversed those 25 miles between Natchitoches and Cloutierville.

• Elizabeth Rachal is presented here as a single woman or possibly a young widow without children, while Poché’s diary (as transcribed by Bearss) states that the deceased was “Mrs. Eliza Rachal.”  

Eliza is a well known English-language dimunitive for Elizabeth. However, the use of the term “Mrs.” is incongruent with Elizabeth’s identity as a 22-year old Texan with no children, living in the household of a man from Mississippi whose wife was also born in Texas and was close in age to Elizabeth. The 1860 census does not specify marital status; but, in both Louisiana and Texas culture where single females were expected to live with relatives, Elizabeth was more likely to be an unmarried relative of Mrs. Henry Jeter.¹³

• Poché, a French Créole writing in French, would have written the name as Mme. E____ Rachal (as opposed to Mrs. E____ Rachal). By the custom of the time and place, the word “Mme.” (Madame) would not have been attached to Mme. Rachal’s given name. The given name to which it would be attached would be that of her husband. The word Mme. did not mean “a married woman named E____ Rachal.” To the contrary, it meant “the wife of  E_____ Rachal.”

Poché’s diary, in its manuscript form, would be an original document. In its published form, it is a derivative work that reflects the editor’s reading and interpretation of the manuscript.  The verification process now calls for us to seek and examine two identifiable original documents:

• The original diary
• A church record of the burial (or, as a next-best source, a newspaper obituary for the deceased)¹⁴

 

Assessment: Original diary

Poché’s diary is held today by Prescott Memorial Library, Louisiana Tech University.¹⁵  An examination of the diary and a comparison with the published edition revealed that his editor, overall, did a worthy job of reading the penmanship and translating the text. However, as when most diaries are translated and published many decades after the fact by someone from a different region and culture, the editor’s paleographic skills were challenged by personal names unfamiliar to him.  In this case, the original text writes the name as Mme. Elizée Rachal rather than Mme. Eliza Rachal. Elizée, variously spelled Elisée, was a male saint’s name within Louisiana’s French Creole (and Catholic) population.

 

Assessment: Burial record

Mme. Rachal was indeed buried from the parish church in the village of Cloutierville. The burial act recorded in the parish register verifies that the deceased indeed Mme. Elizée Rachal, rather than "Mme. Eliza Rachal."  The entry—which was recorded a week later by the parish priest, J. M. Beaulieu, who was by nature lackadaisical with his record keeping—places her in the following sequence:

1863. “Le vingt huit October. J’ai enterré Austin Rachal, agé d’un an. [28 October. I have buried Austin Rachal, aged one year.]

1863. “Le vingt cinq November. J’ai enterré Madame Elisée Rachal, agée du quarante ans. [25 November. I have buried Madame Elisée Rachal, aged forty years.]

1864. “Le huit Janvier. J’ai donné la sépulture ecclésiastique au corps de Benjamin Rachal, agé vingt quatre ans. [8  January. I have given an ecclesiastical burial to Benjamin Rachal, aged twenty-four.]¹⁶

 

Step 3: Identification

The deceased woman to whom diarist Poché referred was neither “Elizabeth Rachal” nor “Mrs. Eliza Rachal.” Further research establishes the fact that Elisée Rachal, on 9 February 1836, wed his cousin Marie Clementia Rachal, daughter of Manuel Rachal and Marie Lise LaCour.¹⁷ By the custom of her era, Marie Clementia was buried in November 1863 under the name of her husband, with the title “Mme.” used to distinguish her from him.¹⁸

 

Conclusion: The attempted verification not only failed but also disproved the assertion. The published diary’s entry for Mme. Rachal’s burial, together with the identification added by the editor, created four facts in need of verification. The facts attributed to the census record are correctly stated, more or less; we can confirm that the census enumerates Elizabeth Rachal in the Jeter household, stating a Texas birthplace for her and an age that would place her birth in 1827 or 1828.¹⁹ However, editor Bearss’ identification of the buried woman has failed verification, despite the fact that it was attributed to two original records: the diary itself and the federal census.

 

Case Two: Verification of Relationship

A respected, peer-reviewed, history journal recently published the following statement:²⁰

“Two Chitimacha sisters, Jean and Marie Terese de la Grande Terre, married François Derbenne and Jacques Guedon, and their métis children became prominent persons in the colonial community at Natchitoches.⁵”

 

Step One: Identify the Issue To Be Verified

This one sentence carries two separate assertions that need verification:

• The “fact” that Jean[ne] de la Grande Terre and Marie T[h]erese de la Grande Terre were sisters.
• The opinion that their métis children became prominent persons at Natchitoches.

For the purpose of this example, we will focus on the stated relationship: a “fact” for which there can be only one correct answer: either they were sisters or they were not.

 

Step Two: Research

The first issue to be addressed is whether our source has cited its own source(s) for the assertion we are attempting to verify. The author has followed the standard practice used by historians (but not genealogists²¹): Individual statements of “fact” are not keyed to specific sources.  Rather, a cluster of related statements are keyed to one reference note in which multiple sources are cited. In the transcription below, coloration separates each individual source, for clarity:

  5. Gregory A. Waselkov and Bonnie L. Gums, eds., Plantation Archaeology at Rivière aux Chiens, ca. 1725–1848 (Mobile, AL, 2000), 35; H. Sophie Burton and F. Todd Smith, Colonial Natchitoches: A Creole Community on the Louisiana-Texas Frontier (College Station, TX, 2008), 30, 39–41, 55; Dayna Bowker Lee, “From Captives to Kin: Indian Slavery and Chancing Social Identities on the Louisiana Colonial Frontier,” Native American Adoption, Captivity, and Slavery in Changing Context, eds. Max Carocci and Stephanie Pratt (New York, 2012), 79–96.

Verification now calls for an examination of all three sources to determine which one was the author’s “authority” for the assertion that Jeanne and Marie Thérèse were sisters. That examination reveals the following:

• Waselkov and Gums: No reference appears to either Jeanne or Marie Thérèse. {Line space}
• Burton and Todd: The five pages cited from this source do discuss these two women but only one page is relevant to the asserted relationship. A second relevant passage is found on a sixth page not cited by Usner. Those two passages tell us this:

"Two of these men wed women from Gulf Coast tribes. François Derbanne, a French Canadian trader who had accompanied St. Denis to Mexico in 1717, and Jacques Guedon, a settler from Nantes, married the Chitimacha Indian sisters, Jeanne and Marie Anne Therese de la Grande Terre. (p. 31)

"15. The De la Grande Terre sister's [sic] were probably among the women captured and sold into slavery from this raid [on the Acolapissa Tribe at Lake Pontchartrain]. (p. 176)

• Bowker Lee: This source offers one relevant passage:

“When the French established a post at the Natchitoches Caddo village on Red River in 1714 at least two men were accompanied by their formerly enslaved Chitimacha wives: Jeanne de la Grande Terre, wife of François Guyon des Prés Derbanne, keeper of the king’s warehouse, and Marie Thérèse de la Grande Terre, wife of merchant Jacques Guedon (E. Mills 1981; Higginbotham 1977, pp. 444-5)."²⁰

 

Assessment:

Usner’s citation to Waselkov and Gums has no bearing on the assertion we are attempting to verify. Waselkov and Gums were cited to support a related but different statement within Usner’s paragraph.

Usner’s citation to Burton and Smith is clearly the source of his assertion that Jeanne and Marie Thérèse were sisters. When Burton and Smith made that assertion, they inserted a reference note where readers might expect to find a citation to the evidence for the asserted kinship. Burton and Smith’s appended note cites no source. It only adds an “aside” that expresses an additional opinion for which no evidence is offered.

Usner’s citation to Bowker Lee is relevant to the two women; but it does not support the “fact” we seek to verify—i.e., the sibling relationship. When we consult the sources that Bowker Lee herself cites²³ (which are now two sources removed from our starting point and are also derivative), those sources do discuss both women, and the sources cite original historical documents. However, neither make any statement of kinship between Jeanne and Marie Thèrese. When we consult the original documents cited by Bowker Lee’s sources, those original records also fail to show any kinship or other association between the two women.

 

Conclusion: While Usner’s article, in a highly respected academic journal, does cite three sources for the paragraph in which his assertion was made, those sources either failed to provide evidence or else they tracked back to original documents that do not at all support the assertion. The asserted kinship remains unverified, with no evidence to support it.

 

BOTTOM LINE:

As researchers, we expect our derivative sources to identity their own sources, but the mere presence of citations does not constitute verification. For the unwary, citations can be an impressive smoke screen. Whether or not the cited material verifies an author’s statement can only be determined by pushing through that screen, and others behind it, until we arrive at the original source for the published assertions.

Verification and confirmation, as dictionary-defined, are subjective concepts. They become concrete—embodied with both a specific process and measurable standards—when we apply the Evidence-Explained Blueprint for Verifying Historical Assertions, as demonstrated in this QuickLesson.