It helps if you know what you are looking for. One Sallie Maria had gone before me and had [allegedly] successfully extracted the line of interest:
Jep (or Jeppe) Berthelsen married Mette Nielsdatter on June 20, 1732
And here is a facsimile of the original document.
Do you not see the marriage record for Berthelsen? Me neither. Even after a scrutinous look my human eye can’t even locate the individual(s). But, we have the notion that the record is there somewhere. Sallie had somehow transcribed the mess Danish penmanship.
Any genealogy explorer and interested hobbyist like myself may trust but should verify the investigations of others. A problem arises where if enough people take a single researched answer for granted as fact and (copy | paste) assimilate them as there own piece of the puzzle family tree then herd consensus becomes de-facto. Search indexing picks up on the trend, AI is trained on the indexed data and… history is printed.
Not immediately accepting the reading of this church record as gospel (see what I did there?) I perform my own diligence by looking harder and with AI as an assistant. AI is not [yet] a silver-bullet. As good or as weak as it is for the translation and transcription, it needs the person with the keyboard to steer it in the right direction or to keep it focused.
There is also an AI behavior known as confabulation. Beware that AI aims to please an can actually make up stuff thinking it’s fulfilling an expectation bias. It’s early days and not a sinister act. When called out on it, the AI apologizes and amusingly so, confesses the deed. (but you do have to expressly raise the doubt to sus it out)
Keeping this in mind it is prudent to distrust the bot. So, why bother? because the AI can spur you to look in certain direction and give insight into what was not obvious. Here’s what I mean: Following my Chat GPT ask for a full transcription, AI found 3 blocks that I had previously by eye, deemed to be incomprehensible and seemingly meaningless. e.g. the third one: “Exaudi (Sunday before Pentecost) – Third reading (banns) for Jep Berthelsen from Vinge Parish and Anne Nielsdatter from Svinninge Parish.“
Without my asking the AI proceeded to explain: “These are the banns (public announcements of an intended marriage) for Jep Berthelsen of Vinge Parish and Anne Nielsdatter of Svinninge Parish, read on three consecutive Sundays leading up to their marriage. The specific Sundays mentioned are part of the liturgical calendar: Cantate, Rogate, and Exaudi, which fall in the weeks before Pentecost. This format suggests they followed the traditional practice of announcing marriages three times in church before the wedding could proceed.” Very good! I had no previous experience of what a Bann was and now I know. Nice.
The AI still hadn’t confirmed a date for this marriage and the AI had wandered off into the weeds with a subsequent Bann (Dom:14) from another party of record and made an [incorrect] assumption: “The 14th Sunday after Trinity would be in late August or early September, depending on the liturgical calendar of that year. This suggests their marriage would have taken place around early September 1772.” You can see what is meant by keeping the AI on point.
To expedite, I queried point blank: Is there a date of 20 JUNE to be found anywhere in this ANNO 1772 record? Answer: Yes, in the initial marriage record image you provided, there is indeed a reference to 20 June:
Torsdagen den 20 Juni, viet Jep Berthelsen af Vinge Sogn og Morten Nielsdatter af Svinninge Sogn.
AI Translation:
Thursday, 20 June – Jep Berthelsen from Vinge Parish and Morten Nielsdatter from Svinninge Parish were married. So, it appears that Jep Berthelsen and Anne Nielsdatter were actually married on 20 June 1772. It looks like the earlier banns we discussed (from Cantate to Exaudi Sundays) were for this marriage.
Now you can see it. With some coaxing which the lens of AI pinpointed and summarized we have clarity.
Betrothal: May 7, 1732
Three banns read on consecutive Sundays (Cantate, Rogate, Exaudi)
Marriage ceremony: Sunday, June 20, 1732
Sally Maria had it right.