As a member of the GeneaBloggers group, I attended Steve Little's webinar on Sunday hosted by GeneaBloggers about using Artificial Intelligence (AI) to improve genealogical writing.
Steve's 21 page syllabus (with additional text documents linked) was provided in Monday afternoon in Beyond Spell Check: The Steve Little Presentation. The webinar video is on the GeneaBloggers YouTube channel.
Here is page 1 of the syllabus which explains Steve's five ways that AI can improve genealogical writing. using his processes:
I reviewed the syllabus, and the first prompt for the "Genealogy Narrative Assistant V3" (GNA), and decided to try it out. See Using Steve Little's Genealogy Narrative Assistant V3 to Develop a GPS Compliant Narrative Sketch (dated 17 February 2026) for an example.
I then reviewed the second prompt for the "Genealogy Research Assistant V8" (GRA), and tried it out using one of my thorny ancestral research problems. Devier James Lamphier Smith (1839-1894) was adopted as a young child and his birth date and birth place have conflicting evidence. I compiled this record in 2009 and updated it in 2015 in Resolving an Evidence Conflict - Post 2: The Evidence (Updated 2015) with an 1889 newspaper interview article and an 1894 obituary. An updated genealogical sketch for Devier J. Lamphier Smith is in 52 Ancestors Week 19: #26 Devier James Lamphier Smith (1839-1894).
Steve's Genealogy Research Assistant V8 is designed to:
- Review your writing against GPS (Genealogical Proof Standard) methodology
- Flags overclaiming — "definitely" when evidence says "probably"
- Corrects terminology — "primary source" → "original source" (GPS doesn't use "primary
- source" for sources — only for information)
- Three-layer evidence analysis: Sources (original/derivative/authored), Information
- (primary/secondary), Evidence (direct/indirect/negative)
- Calibrated confidence language: proved, probable, possible, not proved
I copied the prompt for the GRA from the syllabus and input it to Anthropic Claude. [NOTE: I have only the free account, but Claude provides limited access to the "Thinking" capability (I assume Claude Opus 4.6), so I don't know how much I can do throughout this process.]
The GRA top prompt is:
"You are a genealogical research assistant operating under the Genealogical Proof Standard (GPS). Help users from beginners to credentialed professionals with GPS-informed analysis."
I chose to not upload all of my source documents for this analysis. I attached a PDF of my blog post about Devier Smith's birth records, a PDF of Steve's Genealogy Research Assistant V8 Prompt, and launched Anthropic Claude requesting it to apply the Prompt to my problem.
Here is the Analysis by the Genealogy Research Assistant V8 of my problem (with original capitalization and bolding) by Anthropic Claude:
Genealogical Research Assistant v8 Analysis
Devier James Lamphier Smith — Birth Record Conflict
STEP 1: SOURCE & EVIDENCE INVENTORY
Before resolving the conflict, I'll apply the Three-Layer Model to each source bearing on birth date and birth place.
Birth Date Conflict: 1839 vs. 1842
Source |
Source Type |
Information Type |
Informant |
Birth Date Implied |
Evidence Type |
|---|---|---|---|---|---|
S1 — Family Bible (Devier's hand) |
Original |
Primary |
Devier J. Smith (self) |
7 May 1842 |
Direct |
S2 — 1850 U.S. Census |
Derivative |
Indeterminate |
Unknown (prob. Ranslow/Mary Smith) |
~1838–1839 |
Indirect |
S3 — 1860 U.S. Census |
Derivative |
Indeterminate |
Unknown (prob. Ranslow/Mary Smith) |
~1838–1839 |
Indirect |
S4 — 1870 U.S. Census |
Derivative |
Indeterminate |
Unknown (prob. Devier/Abigail Smith) |
~1839–1840 |
Indirect |
S5 — 1880 U.S. Census |
Derivative |
Indeterminate |
Unknown (prob. Devier/Abigail Smith) |
~1838–1839 |
Indirect |
S6 — 1875 KS State Census |
Derivative |
Indeterminate |
Unknown (prob. Devier/Abigail Smith) |
~1839–1840 |
Indirect |
S7 — 1885 KS State Census |
Derivative |
Indeterminate |
Unknown (prob. Devier/Abigail Smith) |
~1838–1839 |
Indirect |
S9 — Cemetery record (1842) |
Authored |
Secondary |
Unknown (prob. son David D. Smith) |
1842 |
Direct |
S11 — Abigail's family papers |
Original |
Secondary |
Abigail (Vaux) Smith (widow, post-1894) |
7 May 1839 |
Direct |
S12 — 1889 newspaper biography |
Authored |
Primary |
Devier J. Smith (interview subject) |
1839 |
Direct |
S13 — 1894 obituary |
Authored |
Secondary |
Prob. son David D. Smith |
May 1842 |
Direct |
Birth Place Conflict: New York vs. Wisconsin vs. Ohio
Source |
Birth Place Stated |
Notes |
|---|---|---|
S2, S3, S4, S5, S6 |
New York |
Five census records, indeterminate informants |
S7 — 1885 KS State Census |
Ohio |
Outlier; same household informant as other censuses |
S12 — 1889 newspaper bio |
Jefferson County, NY |
Primary information — Devier self-reported |
S13 — 1894 obituary |
Dodge County, Wisconsin |
Secondary information; likely son as informant; almost certainly an error |
STEP 2: CONFLICT ANALYSIS
Birth Date: Weighing 1839 vs. 1842
The 1839 case:
Six census records (S2–S7) across four decades consistently produce calculated birth years of ~1838–1840. While census ages are indirect evidence and individually unreliable, their consistency across independent enumeration events is significant — each census was a fresh recording, not copied from another.
S12 (1889 newspaper biography) provides primary information — Devier himself gave 1839 to an interviewing journalist. This is arguably the most analytically powerful single piece: the subject, alive, speaking from personal knowledge (or belief), without obvious motive to misstate.
S11 (Abigail's notes, written after Devier's 1894 death) gives 7 May 1839 — the same year as the newspaper interview, and a specific date matching the Family Bible's day and month.
The 1842 case:
S1 (Family Bible, Devier's own hand) gives 7 May 1842. This is an original source with primary information — Devier wrote it himself. Family bibles are generally strong evidence. However: the first entries were made in 1889, meaning this was a retrospective recording, not a contemporaneous birth record. The gap between event and recording (~47 years) weakens its primacy.
S9 (gravestone/cemetery record) and S13 (obituary) both give 1842, but both likely derive from a single informant — son David D. Smith — making them potentially not independent of each other.
Assessment: The preponderance favors 1839 (more likely ~7 May 1839):
The 1889 Family Bible entry and the 1889 newspaper interview were made in the same year. If Devier gave 1839 to the journalist but wrote 1842 in the Bible that same year, he was inconsistent with himself. It is possible he genuinely did not know his birth year with certainty, given adoption/name change circumstances.
The son's 1842 date (obituary + gravestone) may reflect what Devier told family in later life — or what was found in family papers — but as secondary information it carries less weight.
Abigail's 7 May 1839 is notable: she gives the same day and month as the Family Bible (7 May) but the year consistent with the censuses and interview. This combination suggests the day/month "7 May" is likely correct, and the year dispute is specifically 1839 vs. 1842.
Confidence: Probable that birth year was 1839, and birth date was 7 May 1839. The 1842 date cannot be dismissed — it warrants the label not proved rather than disproved.
Birth Place: Weighing New York vs. Others
Ohio (S7 only) is a clear outlier — a single census with no corroboration, same household informant as censuses that say New York. Almost certainly an enumerator or informant error. Weight: negligible.
Dodge County, Wisconsin (S13 obituary): The family moved to Wisconsin from New York and Devier was enumerated there in 1850 census. The son likely confused place of residence during childhood with birth place. Weight: low.
Jefferson County, New York (S12, primary information from Devier himself in 1889): Combined with five census records stating New York and the contextual note that the Ranslow Smith family resided in Henderson, Jefferson County, NY in 1840, and that the Isaac Lanfear/Lamphier family was in neighboring Lorraine, Jefferson County — this is a strong convergence.
Confidence: Proved (to a genealogical standard) that birth place was Jefferson County, New York, most likely in or near Henderson or Lorraine.
STEP 3: THE NAME/ADOPTION DIMENSION — AND ITS BEARING ON EVIDENCE QUALITY
Source 8 (Wisconsin name change law, 1866) and Source 10 (Ranslow Smith probate, "adopted son Devier J. Lamphear alias Devier J. Smith") establish that Devier was adopted and changed his name. This is critical context for evaluating evidence quality:
A person adopted in infancy or early childhood may have genuinely uncertain knowledge of their own birth date and circumstances. This would explain why Devier himself gave contradictory dates (1842 in Bible, 1839 in interview) — he may not have known with certainty.
The surname variant "Lamphier / Lamphear / Lanfear" and the Isaac Lanfear family in Lorraine, Jefferson County (with three unmarried daughters of childbearing age c. 1840) provides a strong circumstantial lead to Devier's biological origins.
STEP 4: RECOMMENDED ADDITIONAL SOURCES
Here are targeted record searches to resolve remaining conflicts, prioritized by likely yield:
To Resolve Birth Year (1839 vs. 1842)
Jefferson County, New York birth/baptism records (1838–1842) — Church records (Baptist, Methodist, Presbyterian) for Henderson and Lorraine townships. A baptismal entry for a "Lamphier" or "Lanfear" child c. 1839–1842 would be direct evidence. Repository: Local churches, Henderson Town Historian, Jefferson County Historical Society.
1840 U.S. Census, Jefferson County, NY — Locate the Ranslow Smith household and any Lamphier/Lanfear households. If Devier was born in NY before the family moved to Wisconsin (which they appear to have done before the 1850 census), a male child aged 0–5 in a Lamphier household in 1840 would be significant indirect evidence.
Ranslow Smith family in Wisconsin pre-1850 — Wisconsin territorial/early statehood records. When did the Smiths move? If Ranslow Smith was in NY in 1839–1840 and Devier was in his household by 1850, a NY birth is consistent.
Wisconsin Name Change petition (S8) full text — Does the legislative record include any age statement or other biographical details beyond the name change itself? The full session laws may contain a petition with more detail.
Ranslow Smith probate file, Andrew County, Missouri (S10) — full file review — The index entry references Devier. The full probate file may contain an affidavit, inventory, or testimony that includes Devier's age or birth information stated under oath.
To Identify Biological Parents (Lamphier/Lanfear Connection)
Isaac Lanfear household, 1840 Census, Jefferson County, NY — Identify the three unmarried daughters. Then search for their subsequent marriages, or any child born to an unmarried Lanfear woman c. 1838–1842.
Jefferson County, NY court records (1838–1845) — Bastardy bonds, guardianship records, or any legal record involving a Lamphier/Lanfear child placed for adoption or fostered.
Deed and land records, Jefferson County, NY (Ranslow Smith) — Did Ranslow Smith own land in Jefferson County before moving to Wisconsin? This could establish his presence there at the time of Devier's birth.
Church membership and removal records — If the Smiths and Lanfears attended the same church, a "letter of removal" when moving to Wisconsin might document who came with them.
To Resolve the Ohio Outlier (S7)
This is almost certainly an error and probably does not warrant significant research investment unless other evidence surfaces linking Devier to Ohio.
STEP 5: WRITTEN CONCLUSION (PROOF SUMMARY)
Name: Devier J. Lamphier (biological name) adopted the name Devier J. Smith by Wisconsin legislative act of 21 May 1866. His name appears variously as Devier, D.J., and "Diver" (likely a spelling variant). Proved.
Birth Place: Jefferson County, New York. Supported by Devier's own statement (primary information, S12), five census records, contextual location of the Ranslow Smith family (Henderson) and the Isaac Lanfear family (Lorraine) in Jefferson County c. 1840. Proved to the genealogical proof standard. The 1894 obituary's claim of Dodge County, Wisconsin is almost certainly a secondary-information error reflecting childhood residence, not birthplace.
Birth Date: Most probable birth date is 7 May 1839. The weight of evidence — six census records, Devier's own 1889 interview statement, and his wife Abigail's post-1894 notes (which preserve the "7 May" day/month while giving the year 1839) — preponderates over the 1842 date found in the Family Bible (retrospectively recorded 1889), gravestone, and obituary. However, because Devier himself was inconsistent, and because an adoptee may genuinely have been uncertain of his own birthdate, a proof argument rather than a simple proof statement is the appropriate vehicle. The 1842 date remains possible but is not the most probable conclusion with current evidence.
The unresolved core question — who were Devier's biological parents — is the key to resolving the birth date conflict definitively. A birth or baptismal record from Jefferson County, NY naming a Lamphier/Lanfear child, or an 1840 census entry placing a male infant in a Lanfear household, would likely settle the matter.
Analysis conducted under the Genealogical Proof Standard. Evidence framework per Elizabeth Shown Mills, Evidence Explained. Prompt framework by Steve Little CC BY-NC-SA 4.0.
=================================
Thank you to Steve Little for developing this Genealogy Research Assistant V8. I look forward to using it, especially on difficult research problems with conflicting evidence.
My initial assessment of the finished product was "excellent work" and I stand by that assessment after an hour of critically evaluating the result. I will address the Suggested Records items in a separate blog post.
I hope that other researchers, especially those who are struggling with understanding the GenealogyProof Standard and dealing with conflicting evidence, can improve their research report writing by applying this AI tool.
How did this compare to my own analysis done in 2009? However, I didn't update that 2009 article with the obituary and 1889 newspaper article. Well, the conclusions are about the same! Maybe this was too easy for the Genealogy Research Assistant?
I am painfully aware that I need to revise all of my ancestor biographies to include GPS-compliant statements, research notes similar to the Assistant's notes, and improved EE-style source citations. This may take the rest of my life to finish since I have over 550 ancestor genealogical sketches! Maybe someone else can finish the job?
I can see adding my finished GPS-compliant sketches to FamilySearch Family Tree profiles as a Memory in PDF format, or add a Memory for the Research Note developed by the Assistant for thorny issues. Also to WikiTree, and to my Person Notes in RootsMagic (which should add them to the Ancestry Member Tree and MyHeritage Family Tree when I synchronize them using the RootsMagic file. That's probably another lifetime job.
This is only the second Prompt of the five prompts that Steve recommended for use. I will proceed to the other three Prompts.
Links to my blog posts about using Artificial Intelligence are on my Randy's AI and Genealogy page. Links to AI information and articles about Artificial Intelligence in Genealogy by other genealogists are on my AI and Genealogy Compendium page.
Copyright (c) 2026, Randall J. Seaver
The URL for this post is:
Subscribe to receive a free daily email from Genea-Musings using www.Blogtrottr.com.
No comments:
Post a Comment