Witnesses, Not Fathers: A Note on a Misleading Transcription

In genealogical work, few sources are as valuable—or as frequently misunderstood—as parish records. In some cases, the records reveal Witnesses Not Fathers in critical roles that researchers can easily overlook. In recent years, large-scale transcription projects have made these materials more accessible than ever before. That is a welcome development. Yet with convenience comes a quiet risk: the temptation to trust the transcription without consulting the original.

The case discussed below serves as a reminder that even well-intentioned efforts can introduce serious errors. And when those errors concern something as fundamental as parentage, the consequences can ripple through an entire family tree.

I would like to draw attention to a transcription currently available at Digitalarkivet:
https://www.digitalarkivet.no/en/view/327/pv00000010031522

According to this transcription, Ole Hansen Fjeld and Peder Hagen are listed as the fathers of the married couple, Ole Olsen Fjeld and Sørinou Hansdatter Berg. This is incorrect.

The transcription is based on a church book revision from 1812. However, when we consult the original parish register, the relevant entry appears at the top of the left-hand page:

Østre Toten prestekontor, Parish register (official) no. 1 (1828–1839), p. 192
https://www.digitalarkivet.no/kb20070202620532

The key lies in the column heading: “Hvem forloverne” — “Who were the witnesses?”

This is crucial. The names recorded here are not the fathers of the bride and groom, but their forlovere—the men who served as witnesses to the marriage.

There is also an internal inconsistency that should immediately raise suspicion: if Peder Hagen were indeed Sørinou’s father, she would be named Pedersdatter, not Hansdatter.

More broadly, those listed as witnesses could be:

• fathers
• other relatives
• or simply friends, neighbors, or colleagues

In other words, their presence in this column does not establish parentage.

Transcriptions produced as part of the AMF collaboration—a joint initiative between Digitalarkivet, Ancestry, MyHeritage, and FamilySearch—appear to contain a disproportionately high frequency of errors compared to other transcription efforts.

As I have noted previously in my blog post “A Word of Warning,” the nature of these errors suggests that some of this work has been carried out with limited familiarity with local parish conditions, Norwegian naming conventions, language, and the structure of Norwegian church records.

Taken at face value, such transcriptions can lead to considerable frustration—and, more seriously, to family trees built on entirely false premises.

Conclusion

Transcriptions remain indispensable tools for locating and navigating historical records. But they must be used as they were always meant to be used: as guides, not authorities.

The older discipline still holds true. Read the original. Study the headings. Understand the naming customs. Take nothing for granted.

Only then can we be reasonably certain that the lines we draw between past and present rest on solid ground.