How to influence when [person] given name & pronouns get used in the Narrative Report?

In the custom Fact Type, the sentence template simply used [person], but in certain places the Narrative Report, the person’s name gets replaced with a pronoun making the sentence read awkwardly. Is there anyway to influence when [person] given name & pronouns get used in the Narrative Report?

In this example below her full name is given to begin with, followed next by just her given name, before finally using the pronoun she when her might have been more naturally used instead.

The default is that using just [Person] causes RM to try to use variations on name styles from sentence to sentence so the sentences don’t seem quite so stilted. If you take the default, you have no control over the variations used by RM.

You can use options such as [Person:HeShe] or [Person:HisHer] or [Person:Given] or [Person:Nickname] etc. You can see the list of all the options at Sentence Template Language

I have taken a totally different route. I use what are called point form sentences. I only print the name once per person. So my only fact that has the [Person] variable in the sentence is the Birth fact. My option for the Birth fact is [Person:Full]. My reports looks something like the following. On it’s face, it would appear that my point form sentences would be even more stilted that RM’s default. But my experience is that the point form sentences read rather well. That’s especially true when the point form sentences include a lot of notes.

John Andrew Doe
Birth: 12 Jan 1848, Tennessee
Died: 29 Dec 1918, Missouri
Buried: 2 Jan 1919, Shady Grove Cemetery, Some County, Missouri

You can see a real example of my point form sentences in action at Sample RM Report with Point Form Sentences

1 Like

Perfect, that did the trick - thank you Jerry!

Thank you for sharing the sample_rm_report. I’ve started looking at your real example more closely this morning and appreciate the detail that’s included.

I’m wondering how much of those details you had to manually enter into RM and how much was automatically scraped. For example, in your example, Census:, contains details of the census that don’t show up in what Ancestry scrapes and tags as Residence.

In part why I ask, the default RM Sentence template for Residence includes the awkward conditional insertion of Occupation when it is given in the Ancestry Description field; e.g., David lived Occupation: …, He lived Occupation: …, etc.

I hesitate to ask at this early stage of my RM education, but regarding point form sentences, is this something that awaits discovery in RM or is it a guru creation?

I don’t know about “guru”, but it is a creation. I completely replaced all of the sentence templates for all of the facts I use in RM. I became frustrated by how difficult it was for sentences created by RM really to flow. Sentences created by human beings can put more than one fact in one sentence or sometimes can gracefully split information for one fact across two sentences, etc. It is difficult for computer generated sentences to do that sort of thing. But even with human generated sentences, it’s hard to be too creative with basic birth, marriage, death, and burial information.

I was also influenced by seeing how sentences worked in TMG (The Master Genealogist). It is a now defunct product that was favored by many very expert users of genealogy software. I was never a TMG user, by the way. But I purchased a license after it became defunct. Many former TMG users have converted to RM. By learning about TMG even though it was defunct, I was able to help many TMG users with their conversion to RM. TMG provided much more flexibility than RM or any of RM’s current competitors for creating sentences. But even TMG sentences were not what I was really looking for.

I was also influenced by a little family history book called simply The Nichols Book (my father’s maternal grandmother was a Nichols). There is a little excerpt from the book at the end of this message. Sorry for it’s length, but I wanted you to get the flavor of what I was looking for. I wish I could put this kind of information into RM and have sentences that look like this. And actually I sort of can have sentences that look like this in RM. Except that they are called notes, and they are not generated by RM’s sentence templates.

So in answer to one of your questions, an awful lot of the nice narrative in my reports comes from notes, not from facts and and not from my point form sentence templates. To take census as an example, I transcribe complete census entries, storing the complete transcriptions on my Web page outside of RM. By being complete, the transcriptions include information such as a person’s occupation. Then I copy and paste the transcriptions into RM. I do lot of the same thing with other documents. My point form sentence templates then provide the framework for the report that is generated by RM, and the notes try to provide the warmth and humanity.

Here at last is the promised excerpt from The Nichols Book.

Ruby, fourth child of George and Loretta Nichols Cox was about one year younger than myself. She was a fine pianist and as I remember her when visiting there, especially when her brother and sisters, John, Minnie and Ethel had married. In the evening she played the piano and the guitar and we would sing, especially hymns. Her father, my Uncle George, always was delighted in hearing her and I sing and play. She taught school for a few years, then married a young man from Jefferson County by the name of Emmert Bryan. When World War I came along he and Ruby went to Virginia where he worked as a ship builder until the war was over. He later returned to Morristown, Tennessee where he worked as a cabinet maker. Ruby died in 1925 and is buried at Wesley’s Chapel in Jefferson County. Ruby left three small children, namely, Jack, Lena Maud and Robert.

1 Like

@thejerrybryan - My delay in typing a reply to your post, has largely been following through with what it inspired. Also fanning the flames was the speaker our Hancock County Genealogical Society had yesterday morning and his presentation; Intro to Artificial Intelligence for Genealogists by Stephen Pelsue

One of the interesting aspects of the first proof of concept deliverable from this morning’s ChatGPT 3-hour charette was how it reminded me a little bit of (your?) point form sentences … hmm, I thought I could attached a .docx or even a .pdf, but I’m blanking out on how to do that. Screenshots should suffice.


I like it. It’s not like mine in detail, but it’s very like mine in spirit.