Same Name, Same Person?
Updated
IntermediateWhat you'll learn
~55 min- Ask an AI tool, in plan mode, for a comparison page that sets two records side by side and keeps your decision on each pair
- Check the plan for how it will compare records before anything is built
- Read an identifier (a service or registration number) character by character, with the system that issued it and how sure its reading is, before you use it to link two records
- Tell two men of one name apart, and link one man who carries two different numbers
- Follow one family through three household examination rolls, the parish's lists of its residents household by household, past drifting birth years, a changed farm number and a volume nobody has seen
- Record Match, Needs review or No match yourself, with a reason a relative could check
The story, and what you will end up with
You search an online enlistment index for Walter Engstrom and get two hits. One is ENGSTROM WALTER O, enlisted 15 January 1942, living in Tamarack County, Wisconsin, born 1919. The other is ENGSTROM WALTER E, enlisted 2 March 1942, living in Pell County, Minnesota, born 1921. Both have the family surname and the right first name. If you attach the wrong one to Great-uncle Walt, every record you find for him afterwards goes on the wrong man.
The family line has the same problem a generation further back. The Tallmora baptism register in Sweden, whose page 47 you checked line by line in lesson 03, has a boy named Olof born to Erik Olofsson and Brita Persdotter in May 1856, and another boy named Olof born to the same parents in March 1858. An emigrant index has an Olof Ersson, aged 25, leaving Gråbacka village in April 1883 for Kettle Ford, Wisconsin. Both boys have the same name, the same parents and the same farm, so the name alone cannot say which one left. Ersson is not a family name passed down: it means Erik’s son, made from the father’s first name. The family in Wisconsin is called Engstrom; no record in this file shows the emigrant taking that name, and linking him to the Engstroms is a later step.
By the end of this lesson you will have a small comparison page on your own computer that sets two records side by side, shows where they agree and where they differ, and keeps your own decision on each pair, with your reason. You build it by asking your AI tool in plain English, and then you do the deciding.
Before you start
First, if your AI tool is still open, leave it: type /exit and press Enter (if you use Antigravity, press Ctrl+D with nothing typed instead), and wait until you see the $ or % prompt. The next line goes there, in the terminal, not in the AI tool. It makes a folder for this lesson and goes into it:
mkdir -p ~/archive-playbook/06-same-person && cd ~/archive-playbook/06-same-personStart your AI tool with the start line you kept from the setup lesson, the one with approval prompts off. Then ask it to set the folder up:
Set this folder up so we can undo any change you make, download https://learn.intellimetrics.net/starters/history/06/same-name.json into a data folder, and save a checkpoint.Look for it saying three things: the folder is set up with Git (the program that saves snapshots, from setup), the file (a text file of data) is saved as data/same-name.json, and a first snapshot, your checkpoint, is saved. If it does not mention one of them, ask: “Did you save a checkpoint with the file in it?” If the download does not work, download same-name.json in your browser, check where it went (the browser’s list of downloads shows the folder), and tell your AI tool, for example: “same-name.json is in my Downloads folder. Copy that one file into a data folder here, and save a checkpoint.” In Windows PowerShell or a desktop app, “my Downloads folder” works as is; only say “my Windows Downloads folder” if you use Ubuntu (WSL) on Windows, which keeps its own Downloads folder apart from Windows.
Everything in the file is fictional, made up for this course.
Two kinds of number appear in it. An army serial is the number the Army gave a soldier, the kind an enlistment index and a casualty file carry. A draft serial is the number on a draft registration card, a record of the Selective Service System’s registration. In this case they come from two separate numbering systems, so a number can be compared only with another of its own kind.
Open this box once, because it says what each code used further down stands for.
ℹWhat is in same-name.jsonclick to expand
Ten records, seven pairs of them to compare, and four household examination rolls: fourteen codes in all, each in the tables below. Nine records have an entry of their own; the tenth, R-02, is carried inside R-01’s entry. The codes number every record in this course’s whole case, which is why they skip: this lesson uses some of them.
Each record has its code, its title, whether it is an Original (the record in its first form, or a good image of it) or a Derivative (a later version made from it, such as a typed index), its words exactly as written, and the details picked out of them: name, identifier and the kind of number it is, birth, place, county or parish, and the people named with the person. A reading note says how sure the reading of the identifier is. In a reading, [?] stands for one character that cannot be read.
| Code | What it is |
|---|---|
| R-01 | Baptism of the first Olof, Tallmora, 2 May 1856 |
| R-02 | Death and burial entry of 1857, carried inside R-01’s entry |
| R-03 | Baptism of the second Olof, Tallmora, 14 March 1858 (the page from lesson 03) |
| R-09 | Emigrant index entry: Olof Ersson, 25, leaving Gråbacka on 20 April 1883 |
| R-12 | Walter Olof Engstrom’s draft registration card, 1940, with a draft serial number |
| R-13 | Enlistment index entry, ENGSTROM WALTER O, with an army serial number |
| R-14 | Enlistment index entry, ENGSTROM WALTER E, with an army serial number |
| R-18 | Morning report of 23 March 1945, a unit’s daily report of changes in each soldier’s status; a microfilmed carbon copy |
| R-20 | Loss report of the aircraft Walt was aboard, prepared 5 April 1945: his line from the crew list |
| R-26 | Tamarack County honor roll, an online index typed in 2009 that does not say what it copied |
The pairs are M-1 to M-7:
| Pair | Compares |
|---|---|
| M-1 | R-13 and R-18 |
| M-2 | R-13 and R-14 |
| M-3 | R-12 and R-13 |
| M-4 | R-13 and R-26 |
| M-5 | R-01 and R-09 |
| M-6 | R-03 and R-09 |
| M-7 | R-20 and R-13 |
The four rolls are R-05 (1855 to 1860), R-06 (1861 to 1871, not seen: you would have to request it), R-07 (1872 to 1877) and R-08 (1878 to 1883). Each gives its pages, the farm heading at the top of each household, and every line as written; one line in R-05 is marked struck through. N:o is the register’s way of writing “number”, so Gråbacka N:o 3 is farm number 3 in Gråbacka village.
{ "record": "R-18", "title": "Morning report, 23 March 1945 (microfilm of a carbon)", "kind": "Original", "identifier": "TRAINING-SN-20458[?]", "identifier_type": "army serial", "reading_note": "last character faint on the carbon: 3 or 8"}That is part of one record; the rest of its fields are left out here.
Switch to plan mode
Your AI tool is already open. Switch to plan mode now: Shift+Tab until it says plan, or /plan (the keys for each tool). In plan mode it usually reads your file, asks its questions and shows you a plan before it builds anything; the order can vary.
The conversation
Each message below is an example. Say it your own way.
Say what you want
Say what is in the file, and what you want to do with it:
The records in data/same-name.json include two Walter Engstroms and two baby boys named Olof. Help me build a small tool to compare each pair and record my own decision. Ask me what you need first.Your AI tool reads the file and comes back with questions. It may notice on its own that the file holds seven pairs, more than the two you mentioned.
Answer its questions
Answer each in a sentence. These are questions AI tools often ask, with one way to answer each. Yours may ask them in a different order or show a list of choices:
| If it asks | You might say |
|---|---|
| Only the pairs you named, or all seven in the file? | All seven pairs. |
| A page in your browser, or a program in the terminal? | A page in my browser is fine. |
| What choices do you want: same person, different people, can’t tell? | Match, Needs review or No match. Make me write a reason for each one. |
| Should it stay neutral, or point out what differs? | Show what agrees and what differs, but never suggest a decision or a score. |
| Should the Olof pairs also show the death record and the household rolls? | Yes, show R-02 and the household rolls. |
| Where should your decisions be saved? | Save my decisions to a separate file with the date; leave my data file alone. |
| Save a checkpoint after every decision? | Checkpoints only when you change the page, or when I ask. |
If it asks several questions at once, answer the ones that matter to your research in a sentence, and for the rest say that its suggestions are fine. If it asks something not listed here, answer it the same way, in a sentence.
Read the plan
Next it shows you a plan: what it will build and how it will check it. Read it for six things:
- Both records of a pair are shown in full, as written, side by side, with what agrees and what differs between them.
- Identifiers are compared character by character, and only when both numbers come from the same system: an army serial against an army serial, never against a draft serial.
- A character marked
[?]stays unread. It is never filled in as a 3 or an 8. - The household rolls sit side by side in order, the roll nobody has seen shown as not seen, with what changes from one roll to the next pointed out.
- Nothing suggests a decision: no score, no verdict. Your decision needs a reason, and no record is changed or merged into another.
- Your file is read and left as it is.
A first plan often leaves out how identifiers are compared, or what changes between the rolls. Say so in a sentence, in your own words, for example:
Compare the numbers one character at a time, and only numbers of the same kind. Also show me what changes from one household roll to the next.When the plan has all six, approve it: if your AI tool offers a choice to go ahead, pick it; otherwise say “Yes, go ahead and build it.” In Codex, press Shift+Tab first to leave plan mode. In Copilot, saying it brings up a choice to go ahead; pick that. Let it build. When it has finished, ask it: “Did you install anything, and where?” Nothing, or something inside this lesson’s folder, is fine for these lessons. If it names anywhere else on your computer, ask it what that is for, and to remove it if you do not want it there.
Look at what it built
Ask it to open the page:
Start the comparison page and open it in my browser.Your AI tool may open the page straight from its file, or start a small program on this computer that serves the page at a local address, one ending in a number, such as http://127.0.0.1:8000/ (127.0.0.1 is an address that means this computer); either is fine. If it gives you an address without opening it, copy the address into your browser’s address bar and press Enter. If it says the address is already in use, say: “Something else is using that address. Don’t stop it. Use a different address for my page.” Leave the program running while you work; when you are done, ask your AI tool to stop it. If your AI tool does not take a message after opening the page, wait a minute. If it still does not, close the terminal window, open a new one, type this lesson’s folder line again, start your tool, and say: “Start the page again so it keeps running while we talk.” The page and your saved work are files in the folder, so nothing is lost.
Before you decide anything, find these on the page. If one is not there, tell your AI tool what you see and what you expected, for example: “M-1 shows the last character of R-18’s number as 3. It should stay unread.”
- All seven pairs, each with no decision recorded yet, and nothing anywhere that suggests one.
- M-7: the two numbers are the same, from the same system (both army serials), with no unreadable character.
- M-1: the last character of R-18’s number shown as unreadable, not as a 3 or an 8, and not counted as a difference.
- M-4: one character differs, an 8 against a 3, the fifth of the six digits.
- M-2: the numbers, the birth years (1919 and 1921) and the counties differ.
- M-3: the two numbers are from different systems (draft serial and army serial) and are not compared with each other.
- M-5: R-02, the death and burial of the first Olof in January 1857, shown with R-01. You may have to open a section of the page to see it.
- The household rolls: Brita Persdotter’s birth year changing 1827, 1826, 1828; Olof’s 1858, 1857, 1858; the farm heading changing from Gråbacka N:o 3 to Gråbacka N:o 5; R-06 shown as not seen; and Myrbo N:o 1, another farm, on the page next to the family’s in all three rolls (compare the page numbers).
Decide each pair yourself
A decision is yours: Match, Needs review or No match, with a reason. A good reason says what agrees, what differs, what could not be compared, and which records you relied on, so that a relative can check it without asking you. If your page has no place for something, such as the records you relied on, put it in the reason. Work through the pairs in this order.
- M-7, the easy one, done carefully. The loss report’s crew line and the enlistment entry give the same army serial, and neither has an unreadable character. The name agrees too. The loss report gives no birth or place to compare, so the link rests on the number and the name, and your reason should say so. The records settle this one: same number, same system, same name and initial, nothing unread. That supports a Match, with a reason such as: “Same army serial, no unreadable character in either; same name and initial; R-20 gives no birth or place to compare.” Before you save it, try saving with the reason left empty, and check that the page refuses. If saving downloads a file, tell your AI tool its name and that it is in your Downloads folder, and ask it to move that one file, for example: “I saved decisions.json; it’s in my Downloads folder. Move that one file into this folder’s data folder.” Use the name your page gave the file.
- M-2, two men of one name. The numbers differ, and so do the birth years and the counties. A differing number alone would not prove two men; with the other two differences the records settle it. That supports No match, with a reason that gives all three: “Different army serials, birth years 1919 and 1921, counties Tamarack and Pell; R-13 and R-14.”
- M-3, one man, two numbers. The draft card’s number and the enlistment number share nothing, but they come from two systems, so their difference says nothing either way. Compare the rest: name and middle name (Olof, against the initial O), birth year, county. Decide it, and say in your reason that the numbers were not compared. Match and Needs review are both honest here, if the reason says what the decision rests on.
- M-1, a number you cannot fully read. One hasty reader sees “204583” and links; another sees “204588” and rejects. Neither reading is on the page. Another record in your file does give his number in full beside the same names: the loss report, R-20, prepared later, on 5 April 1945. Its crew list has Walter with the same rank and the same four crewmates as R-18 (Prell, Sorley, Reese and Dahlke), and its number is the one you matched to R-13 in M-7. That is a strong reason to think R-18’s Walter is the same man, but it comes from R-20 and his crewmates, not from R-18’s number, which no other record can read for you. This pair is open on its own two records. A finished decision for it is Needs review, with a reason that names the unread last character, what agrees, the support from R-20 through M-7, and what would settle the number itself: a clearer copy of the same morning report. Written that way, Needs review is a complete record of where the pair stands.
- M-4, a number that differs by one character. You have just met a 3 and an 8 that could not be told apart, and R-26 is a typed index that does not say what it copied. Neither correct the number in your head nor reject the entry because of it. This pair is open too: the records in your file cannot say whether the index typed a 3 for an 8 or copied a different man. A finished decision is Needs review, with a reason that names the differing character, the index’s missing source, and what would settle it: the record the index copied.
- M-5 and M-6, two babies called Olof. Because R-09 indexes the departure under
Ersson, searching the emigrant index for “Olof Engström” found nothing. (In the family-line search log, a different log from the one in lesson 01, that empty search is S-03, and S-04 is the search for “Olof Ersson” that found R-09. Lesson 01’s report listed both numbers as gaps because they were not in its log.) Now read R-02 beside R-01 in M-5:1857. Jan. 9 död, 13 begr. Olof, s. af Bd. Erik Olofsson i Gråbacka. 8 mån., that is, died 9 January, buried 13 January 1857, Olof, son of the farmer Erik Olofsson of Gråbacka, 8 months. The first Olof died as a baby, eight months old, so he cannot be the man of 25 who left in 1883. The records settle this one, and they support No match, with a reason such as: “R-02 records the death of R-01’s Olof in January 1857; R-09’s emigrant was alive in 1883.” Record M-5 now. For M-6, the birth and the parish fit, but a baby and a man leaving 25 years later are not yet linked. Open the household rolls before you decide it. - The household across three rolls. Read what changed. The farm is N:o 3 in R-05 and N:o 5 in R-07 and R-08: a hasty reader decides this is a different household and loses the family. Brita’s birth year is 1827, 1826 and 1828, always on 20 January; no one roll is “the” answer, and none is a reason to drop her. Olof’s drifts the same way. In R-05 the first Olof’s line is struck through, ending
† 9/1 57, died 9 January 1857, which agrees with R-02. In R-08, Olof’s line endsUtfl. 12/4 1883 t. N. Amerika: moved out on 12 April 1883 to North America; leaving the parish and leaving the country, on 20 April, are two events. Then the neighbours. Myrbo N:o 1 is on the page next to the family in every roll. Hans Jonsson and Kerstin Olsdotter were the two baptism witnesses on R-03. In R-05, Hans is a farmhand at Myrbo. In R-05, Kerstin is a maid in the family’s own household at Gråbacka. By R-07, Hans is the farmer at Myrbo, and Kerstin is his wife. Same parents, same children with the same days and months, the same farm next door, roll after roll. - Back to M-6. The household chain settles the link between the 1858 baby and the emigrant, and it supports Match. Your reason names the chain and the rolls it runs through, and says what it does not settle, for example: “Same parents, same children with the same days and months, Myrbo N:o 1 next door in R-05, R-07 and R-08; R-08 has Olof moving out to North America on 12 April 1883; R-09’s age 25 fits 1858. R-06 not seen.”
What the chain does not settle: whether Gråbacka N:o 5 is N:o 3 under a new number, or the family moved within the village between 1861 and 1871. R-06 is the volume for those years. Nor does M-6 link the emigrant to the Olof Engstrom of Kettle Ford; that is a separate comparison, against records this file does not hold.
Check it for these
- After you recorded M-5 and M-6, every record is still there exactly as it was, side by side; nothing was merged.
- After you reload the page, M-5 and M-6 still show your decisions, each with its reason and the date.
- When you ask “Has my data file changed since the checkpoint?”, your AI tool compares it with the checkpoint and says no.
If something on the list is not true, tell your AI tool in a sentence what you see and what you expected.
Notice what one unreadable character does
Ask:
In this folder, make a copy of my data file where R-18's number ends in 8 instead of [?]. What does M-1 show on the copy?Look for it saying that on the copy M-1 shows a difference at the last character, 3 against 8, where your file showed one character that could not be read. If it changed your own file instead, say: “Put my data file back the way it was at the checkpoint.”
One guessed character turns “cannot be compared” into “different”, and a guess of 3 would have turned it into “the same”. The number on the carbon has not changed; only the reading has. Then M-3 showed the opposite case: two numbers that share nothing, about one man, because they were issued by two different systems.
A matching identifier is a strong comparison clue. Check its reading, issuing context and other evidence before linking records.
Now remove the copy:
Delete the copy now, and save a checkpoint.One real run
Here is part of what one comparison page showed for M-1, built with Claude Code on 10 November 2026. Yours will be laid out differently; check it against the lists above, not against this.
Numbers are compared character by character only when both records use the same numbering system.
Prefix: same text; no differing positions; uncertain reading at position 6.
Position 1 2 3 4 5 6R-13 2 0 4 5 8 3R-18 2 0 4 5 8 [?] same same same same same uncertainFictional training example.
The same page said of the household rolls that the way it lined up lines from roll to roll was “only how the lines are arranged, not a finding that they are the same person”.
How real archives differ
The people, farms, numbers and registers in this case are invented. The record types are real, and their real rules are worth knowing:
- The Swedish National Archives (Riksarkivet) describes the household examination rolls (husförhörslängder): “Församlingens invånare redovisas hushåll för hushåll, ordnade efter bostadsort” (the parish’s residents are listed household by household, arranged by place of residence), with contents including “alla familjemedlemmar, föräldrarnas födelsedata … eventuellt inflyttningsår, eventuellt utflyttningsår” (all family members, the parents’ birth data … any year of moving in, any year of moving out) (Riksarkivet, kyrkoarkiv). Riksarkivet explains these only in Swedish; the English in brackets is a translation for you.
- On names it says: “Idag följer ofta samma efternamn från generation till generation, men detta är en 1900-talsföreteelse. Det vanliga var att faderns förnamn bildade barnens efternamn” (today the same surname often follows from generation to generation, but this is a 1900s phenomenon; the usual practice was that the father’s first name formed the children’s surname), and that women kept that name after they married (Riksarkivet, släkt- och personforskning). That is why Brita Persdotter does not take Erik’s name. No record in this case shows when Olof took the name Engström.
- The registers in this case use abbreviations made up for this course, with these meanings:
Bd.farmer,h.wife,s.andd.son and daughter,pig.maid,dr.farmhand,Enkanthe widow,†died,d:oditto,Utfl.moved out,t.to,begr.buried,mån.months. - The US National Archives says its series of World War II Army Enlistment Records “does not have records for Army officers” (NARA, Army serial number FAQ). An enlistment index will not find everyone who served. If someone may appear under more than one number, look for each.
Keep going
Small changes do not need plan mode. Say each in a sentence:
- “When I change a decision, keep the earlier one with its date.” Your page may do this already; if so, your AI tool will say so.
- “Give me a way to add a pair of my own.” Before you rely on it, try it with one more made-up pair.
- “When I get the 1861 to 1871 roll, let me enter its lines, and mark them as entered by me.”
- “Save a checkpoint, including my decisions file.” Your AI tool saves a snapshot you can return to (people who use Git call this a commit).
When you have your own records to compare, compare them in the finished page. It runs on your real records by itself, with your AI tool closed. Do these steps in order; the first two happen now, while this folder still holds only practice data:
- Go through the data check, the questions in setup about what the material is, who decides how it is used, and which AI tool would see it. Go on only if it allows this use.
- While this folder still holds only the practice files, ask your AI tool: “How do I open the comparison page myself on a file in another folder, and save my decisions there too?”
- Write down what it tells you.
- Ask your AI tool to stop the page’s program, then leave the tool: type
/exitand press Enter (if you use Antigravity, press Ctrl+D with nothing typed instead). - Keep your real files in a folder of their own, outside this course folder, and do what you wrote down. Your decisions file is saved beside your real files, not in the course folder.
Try another case
Before the quiz, write two or three sentences for yourself. A cousin finds two John Karlsons in one county census, ten years apart: one aged 30, one aged 43, both farmers, both married to an Anna. She wants to merge them. What would you compare before deciding, and what would you write if you could not settle it?
A good answer compares more than the name and wife’s name (the children and their ages, the neighbours, the farm), notices that 30 and 43 are thirteen years apart in a ten-year gap, keeps both entries separate, and records Needs review with a reason that names what would settle it.
Key takeaways
- Ask briefly, and let your AI tool ask back. Say what is in the file and what you want to decide; its questions fill in the rest.
- Check the plan for how it compares, not how it is coded.
- A differing identifier does not prove two people, and a matching one does not prove one: M-3 links across two numbers, and one character that differs in an index is a question, not an answer.
- Follow the household, not the name: through drifting birth years, a changed farm number and a volume nobody has seen, the parents, children and neighbours carry the family from roll to roll.
- Link, never merge, and write the reason down: a Needs review that names what is open and what would settle it is a complete decision.
You are tracing a Nils Berg. A lodge membership card and a county tax roll both show the number 4471 beside his name. The card gives his age as 34; the tax roll gives no age, and lists him in a different township. What do you record?
What’s next
You now have a Walter whose records you can link with reasons: the draft card, the enlistment entry and the loss report. Next you put them in order. The Life and Service Timeline keeps the date of an event, the date a record was written and the date of an order apart, separates where his unit was from where he was, and flags the gaps no record covers.