Does Genealogy Have Its Own Methodology?

Eteläpohjalaiset Juuret 3/2005, pp. 38–55. ISSN 1795-3189.

Genealogy as a hobby has grown explosively over the past two or three decades. More has been published than ever before. But one sector has been quiet. Interest in theoretical questions — and writing on them — has been limited. In the 1980s there were some discussions about numbering systems, but that momentum stalled. Researchers have stayed with the established systems for numbering ancestors and descendants, largely because genealogy software now handles numbering automatically.

With one exception I have felt like “a voice crying in the wilderness” when writing in Genos in the 1980s about genealogical methods, and more recently about the definition of genealogy. No responses or feedback have appeared. I have tried on suitable occasions to argue that, in order to raise genealogy from its subordinate position as an auxiliary discipline of history to a more independent and respected field, methodological questions need to be addressed. The goal should be to establish at least one university post whose holder’s duties include the development of genealogical theory and methods. This should remain a goal even though it is difficult and no money exists for it. History departments have enough posts that genealogical theory could be incorporated into one of them, if there is genuine conviction that genealogy belongs within the historical sciences.

Genealogical Method

When genealogy has been treated as belonging to the historical sciences, the methods of historical research have been considered sufficient. First and foremost this means source criticism. But something more must be required, because criticism alone does not produce results. In fact, with sufficiently strong source criticism most conclusions can be made impossible. This brings me to the one exception I mentioned.

The Basic Method

Let me stay with the early modern period — the centuries that most genealogists work in. A core axiom is that every genealogist uses some method. Usually the researcher is not conscious of using one. A method can be defined as a way of conducting research. The method normally used by genealogists, as described in guidebooks such as Sukututkimus askel askeleelta (“Genealogy step by step”), can be outlined briefly:

• Always start from initial data: typically a name, date and place of birth.
• Look in the register of births for the period and place given.
• Use those details to find the person in the communion register (parish record) for the corresponding period.
• Then continue along the chain of communion registers, with side excursions to church history books (historiakirjat) as needed.

Source criticism appears in guidebooks mainly as advice to verify data and check that no two people of the same name were born on the same day. Research is straightforward as long as the data is consistent and the parish records are sufficient. Let us call this the basic method — or, more colloquially, the “zigzag method”: alternating systematically between church history books and communion registers.

Like a Cat Around Hot Porridge

You face methodological questions in a completely new way when church records no longer suffice. The practice has been to look for new sources. Choosing which sources is itself a methodological question, depending on available time, expertise and the existence of the sources. For the same reason researchers make various delimitations that must be stated clearly. Often it is an impossible task to go through every possible source. A genealogist may have to rely on just one source series — the Suomen asutuksen yleisluettelo (General Register of Finnish Settlement) or, failing that, poll tax rolls — when church records run out. Good methodological practice is to divide the review of a large mass of sources into stages: start with the easiest, progress to the harder, until results are found.

In the most difficult cases you are forced into multi-source research, and you have to think in a new way about what can and cannot be concluded from each source.

I have recently had to work on genealogies that push the limits of inference, as in medieval genealogy. Methodologically I have approached these like a cat around hot porridge — cautiously, feeling out the possibilities:

1. Map the initial data as precisely as possible.
2. Consult a map of the area and estimate from general knowledge where the search should focus.
3. Determine the critical period for research and what sources are available.
4. Carry out the research proper using the method of necessary and sufficient conditions (which I presented in Genos).

A Worked Example: Jaakko Markunpoika

I was given as a starting point the published family history of the Kihniä family in Peräseinäjoki, Ilmajoki parish. The task: trace the parentage of Jaakko Markunpoika, the earliest known ancestor, as fully as the sources allow.

From the family history: Jaakko died on the first Sunday after Trinity, 1697, in Ilmajoki, aged 67, giving a birth year of around 1630. He had farmed the Kihniä holding at Peräseinäjoki from 1662 to 1690. His wife was Riitta Valpurintytär, who died around 1694.

Note on the wife’s name: In my own notes from the General Register she appears as Pirkko Vilpuntytär. Pirkko and Riitta are the same Finnish name, both derived from the Swedish Brita. But Valpurintytär and Vilpuntytär (Philippsdotter) are entirely different names — an important discrepancy to resolve.

Geographical framing: Peräseinäjoki in the mid-17th century was a sparsely inhabited frontier area of Ilmajoki parish. No roads existed; travel was by river. The Satakunta border was not far. The key question is whether Jaakko came from the established villages of Ilmajoki or from further away. Since he bears no Savonian surname and the Kihniä place name predates him, the working hypothesis is that he was Ostrobothnian and his home village lay relatively close.

Searching for Markku: Two to four farms were the typical range needed, and searching the settlement register for men named Markku in Ilmajoki 1630–1660 produced three candidates:

1. Markku Matinpoika at Koivo (Kouko), later Hirvijärvi village — farmer 1636–75.
2. Markus Juhonpoika at Vähä-Kurikka — farmer 1606–53.
3. Markku Paavonpoika at Piirtola no. 3, Jouppila village — farmer 1635–60.

Key court record: In the Ilmajoki court records for 10 March 1654, a Jaakko Markunpoika from Jalasjärvi applied for permission to found a new farm. The jury testified that the site had been abandoned since the Club War, so completely overgrown that it needed 8 years’ tax exemption to establish. Jalasjärvi village at that time was the location of candidate no. 1 (Markku Matinpoika, at what would later be called Hirvijärvi). The date calculation confirms: 1654 + 8 tax-free years = 1662, exactly when the Kihniä holding appears consistently in poll tax rolls. The conclusion that Markku Matinpoika of Jalasjärvi was Jaakko’s father is strongly supported.

Finding Riitta’s parents: Searching for men named Vilppu in the same area and period produced two candidates:

1. Vilppu Jaakonpoika, later Jokipii village (Yli-Jokipii farm), farmer 1636–c.1662.
2. Vilppu Laurinpoika at Rahkola no. 2, Jouppila village, farmer 1659–77.

The age relationship resolves this: the Jouppila Vilppu is clearly too young to be Riitta’s father (born too late to have a daughter old enough to marry Jaakko in the 1650s). This leaves Vilppu Jaakonpoika of Jalasjärvi as the most probable father.

The Criterion of Proximity

The working principle throughout the above is the criterion of proximity (närhetskriteriet in Swedish), applied both in time and in place — a concept also used by the Swedish historian Matts Hallenberg in tracing provincial governors from the Vasa period. Results obtained through this criterion are probable rather than certain. The degree of probability increases the more systematically the population group and its sources have been surveyed.

Even the Best Method Can Lead to Wrong Conclusions

From my own experience, and from observing others, methodologically correct research can still produce wrong results. One incorrect detail in the initial data is enough. Birth dates and places can change between sources; two people with the same name and the same birth date can appear in the same parish.

Example: Tracing Salomon (Finnish: Salu) Heikinpoika, born Virrat 30 May 1839. Searching a genealogy database produces no exact match — there is a Salomon born on that date, but his father’s name is recorded as Herman, not Heikki. Checking the original microfilmed birth register confirms the father’s name should be Heikki. The parents are found in the Virrat communion register 1834–43. The research looks straightforward.

But the starting data included another detail: Salu Heikinpoika had lived at the Ala-Pöyhölä croft of the Vermas farm in Virtainkylä. When this is checked against the communion register, the Salomon found there does not match — the Salu of Ala-Pöyhölä is a different person. The parish certificate, drawn correctly from the communion register, referred to the wrong Salomon. The whole ancestry being built up was for the wrong man.

The lesson is that even when initial data comes from an official parish certificate, it can lead to the wrong person when two individuals share the same name, birth date and parish.

Conclusion

Good genealogical method involves careful preparation: precise mapping of initial data, geographical orientation, identification of the critical time period, systematic source review, and the conscious application of the criterion of proximity with explicit probability reasoning. Results in difficult cases are probable rather than certain — but well-founded probability, clearly stated, is a legitimate scholarly product. The method should be approached cautiously, like a cat around hot porridge, while securing the flanks as the research proceeds.