Using AI to Build a Norwegian Genealogy Research Strategy

Here is what ChatGPT tells me when I ask for a research strategy for a named individual in Norway.

This is a follow-up to my article Using AI in Norwegian Genealogy Research.

Imagine you are researching a Norwegian ancestor and have only a few starting facts:

Name: Ole Rasmussen Skavnes
Approximate birth: 1850s
Known place: Skavnes, Romsdal, Norway

You could start searching for his name immediately. But that is not necessarily the best approach. A better use of AI is to ask it to help you build a research strategy: what records are likely to exist, where those records are held, what naming variations to expect, and which sources should be searched first.

The aim is not to have AI “find your ancestor.” It is to use AI to decide where to look next and why.

Start by asking AI to map the problem

A useful prompt might be:

I’m researching a Norwegian ancestor and want help building a search strategy — not an answer about who they were, just a plan for where to look next.

What I know:

Name: Ole Rasmussen Skavnes
Approximate birth year: 1850s
Home parish/farm: Skavnes, Romsdal, Norway

Based on this, help me think through:

• What record types would likely exist for this person, in this place, during this period?
• What jurisdiction — prestegjeld, sorenskriveri, fogderi, etc. — would the farm or parish likely have fallen under?
• What alternate spellings or naming patterns should I search under, given Norwegian patronymic and farm-name conventions?
• What neighbouring parishes or municipalities are worth checking if the home parish comes up empty?
• What would be a sensible order to search these records in, and why?

This is a much better question than simply asking, “Who was Ole Rasmussen Skavnes?”

The latter encourages an answer. The former produces a research plan.

1. Identify the historical jurisdiction

The first problem is geographical.

A Norwegian farm name is not enough by itself. Administrative boundaries have changed, and the municipality shown on a modern map may not be the jurisdiction under which the records were created.

For a nineteenth-century ancestor, it is useful to think of the hierarchy roughly as:

Farm → sokn → prestegjeld → fogderi → sorenskriveri → amt

Different records can therefore be found under different jurisdictions.

The prestegjeld is particularly important for parish registers. The sorenskriveri becomes important when looking for probate and legal records, while the fogderi appears in various administrative and taxation records.

This is one area where AI can be useful before you open a single record. Ask it to establish the historical jurisdiction of the farm, but treat the answer as a starting point to verify.

2. Start with the census

For someone born in the 1850s, the 1865 census is an obvious early target.

If Ole was born between 1850 and 1859, he would have been roughly six to fifteen years old in 1865.

The objective is not simply to find an Ole Rasmussen.

You want to identify the entire household.

Look for:

• Ole’s age
• birthplace
• farm or residence
• father’s name
• mother’s name
• siblings
• father’s occupation
• household position
• farm or holding number, if available

Finding a household such as:

Rasmus [surname], farmer or tenant
Wife: [name] Ole Rasmussen, age 10
Several siblings

is much more useful than finding an isolated name in a database.

It gives you a family structure that can be tested against other records.

3. Find the baptism record

Once an approximate birth year has been established, move to the parish register.

Search the Fødte og døpte records for the relevant parish, probably covering several years rather than only one.

The baptism record should be treated as a key piece of evidence.

Record the entire entry, including:

• date of birth
• baptism date
• father’s name
• mother’s name
• residence
• occupation or social status
• godparents
• marginal notes

The godparents are worth recording even if they appear irrelevant.

They were often relatives, neighbours or members of the family’s wider social network. Later, they may help confirm relationships.

4. Remember how Norwegian names worked

This is where a modern search engine can easily lead a genealogist astray.

Rasmussen is normally a patronymic. It means, essentially, “son of Rasmus.”

It does not necessarily represent a hereditary surname.

Skavnes, meanwhile, is a farm or residence name.

So the same man might appear in different records as:

• Ole Rasmussen
• Ole Rasmussen Skavnes
• Ole Rasmussøn
• Ole Rasmusson
• Ole Olsen, if a transcription is incorrect
• Ole [another farm name]

The farm name may also change during his lifetime if he moves.

This means that a search for the exact phrase “Ole Rasmussen Skavnes” can be far too restrictive.

A better strategy is to search the components separately and then compare the evidence.

Once you identify Ole’s father, the next generation becomes much easier to investigate. If his father was Rasmus Olsen, for example, you would search for Rasmus Olsen, not for another generation of “Rasmussen.”

5. Find the confirmation

Confirmation records are particularly useful for sorting out people with common names.

For an Ole born in the 1850s, look roughly fifteen to eighteen years later.

The confirmation record may provide:

• age
• residence
• birthplace
• parents
• vaccination information
• assessment of knowledge or conduct

More importantly, it gives you another independent appearance of the same person.

If the baptism says:

Ole Rasmussen, Skavnes

and the confirmation several years later says:

Ole Rasmussen, Skavnes

you have another piece of evidence connecting the individual to the farm.

6. Work backwards through the father

Once Ole’s parents have been established, stop concentrating on Ole for a moment.

Research his father.

Ask:

• Who was Rasmus?
• Where was he born?
• Who were his parents?
• Who did he marry?
• When did he appear at Skavnes?
• Was he a farmer, tenant farmer, husmann or servant?
• Did he inherit or acquire the holding?
• Had he moved from another farm?

This is often where a genealogy begins to become interesting.

The question changes from:

“Who was Ole Rasmussen?”

to:

“How did this family come to live at Skavnes?”

That can lead to an entirely different set of records.

7. Research the farm as well as the person

A farm history or bygdebok can be extremely useful.

Look up Skavnes and examine the families associated with the farm during the relevant period.

Record:

• farm owners
• tenants
• husmenn
• families
• marriages
• children
• transfers of property
• people arriving from other farms
• people leaving for other farms

The purpose is not to accept the bygdebok as proof of every relationship.

Instead, use it as a map of the family and the farm.

When the bygdebok says that a particular person was the father of another, check the original parish register where possible.

This distinction matters:

Secondary sources help you find the evidence. Primary sources help you prove the evidence.

8. Expand to neighboring parishes only when necessary

If the expected parish register produces nothing convincing, don’t immediately search the whole of Norway.

Expand geographically in stages.

For someone associated with the Romsdal/Vestnes area, neighbouring historical parishes might include places such as:

• Tresfjord/Sylte
• Tomrefjord
• Vestnes
• Ørskog
• Vågstranda
• Veøy
• Rødven

The exact list depends on which historical Skavnes is involved and the period being investigated.

This is another useful task for AI: ask it to identify historical neighboring parishes, rather than simply asking for modern neighboring municipalities.

9. Add military records

For a Norwegian man born in the 1850s, military records may eventually provide another way of identifying him.

Depending on the locality and period, look for:

• lægdsruller
• militærruller
• stamruller
• other military rolls

These can sometimes contain information such as:

• birth date
• residence
• farm
• physical description
• height
• military service
• movements

They are particularly useful when several men have the same name.

I would normally use them after establishing the basic family through church records and censuses, rather than making them the first port of call.

10. Leave probate and property records until the family is established

Probate records, or skifte, can be exceptionally useful, but they are generally easier to use once you know approximately when and where someone died.

A probate record may identify:

• widow
• sons
• daughters
• spouses of children
• property
• debts
• possessions
• heirs living elsewhere

Land and property records can similarly reveal whether the family actually owned the farm or occupied it as tenants.

Useful record types include:

• pantebøker
• tingbøker
• skjøter
• skylddelingsforretninger
• matrikler
• odels records

These sources can answer questions that parish registers cannot.

For example:

Was Rasmus actually the owner of the Skavnes holding, or was he a tenant?

That may completely change how you understand the family’s economic position.

11. Only investigate emigration once there is evidence of it

If Ole disappears from Norwegian records, don’t immediately assume he emigrated.

First check:

• neighbouring parishes
• marriage records
• censuses
• migration records
• military records
• death records

If there is evidence that he went abroad, then search the Norwegian emigration protocols and, if appropriate, American or other destination records.

The farm name can be especially valuable here because emigrants were often recorded with both a patronymic and a farm or residence name.

A sensible research order

A practical sequence for this particular example would be:

1. Establish the historical location of Skavnes

↓

2. Search the 1865 census

Identify Ole’s household.

↓

3. Find Ole’s baptism

Confirm his parents.

↓

4. Search the 1855 census

Look for the parents and family before Ole’s birth.

↓

5. Find the parents’ marriage

Establish the previous generation.

↓

6. Find Ole’s confirmation

Confirm his identity and residence.

↓

7. Research the Skavnes farm

Understand the household and property structure.

↓

8. Search the 1875 census

See where Ole went as a young man.

↓

9. Find his marriage

Establish his adult household.

↓

10. Follow him through later censuses

Look at 1891, 1900 and later records as appropriate.

↓

11. Check military records

Use them as an additional identity check.

↓

12. Search death and probate records

Confirm the final family structure.

↓

13. Search land and property records

Investigate ownership and inheritance.

↓

14. Investigate emigration

Only if the evidence suggests that he left Norway.

Keep an evidence table

One of the simplest ways AI can help is to suggest a structure for recording the research.

For example:

Record	Name	Age/Birth	Farm	Father	Mother	Confidence
1865 census	Ole Rasmussen		Skavnes
Baptism	Ole Rasmussen		Skavnes
Confirmation	Ole Rasmussen
1875 census	Ole Rasmussen
Marriage	Ole Rasmussen

This does something surprisingly important: it separates evidence from assumptions.

If two different Ole Rasmussen candidates appear, you can compare them rather than unconsciously merging them.

And that is probably the most useful role for AI in this kind of research.

AI can help you identify what to search, where to search, what variations to expect and what evidence would be useful next.

It should not quietly turn a plausible guess into a genealogical fact.

Go to the Download section and find The Genealogist Prompt Library which contain a number of prompts to get you started using AI for genealogy.