Most of my prior genealogy research was based on family records. Very little of it came from hard sources like census records. Some folks would claim that what I have is not genealogy, but folklore. So I've recently been spending considerable time looking for "official" documents, like county marriage records, census records, land registries, etc.

What I have found does not help much in coming up with concrete "facts".

Take my grandfather as an example:

  • The family folklore/history says he was born in Brooklyn Iowa in 1860.
  • The US Census records of 1890 and 1910 say he was born in Illinois.
  • One Census record shows the family name as spelled Ferrell (two E's).
  • Another record shows him as being born in St Louis, Missouri.

I know that the family spells our name Farrell. So the Census record is wrong. Hey, typos happen. I am sure that he was born near 1860, maybe 90% sure it was plus or minus one year of 1860, but it could be more. And I could believe that he was born somewhere within 100 or so miles of the Mississippi river, starting at St Louis, and going up to say Dubuque, Iowa. But the location of his birth is at best a fuzzy fact.

How do serious folk record this uncertainty?

  • 1
    Welcome to G&FH.SE! I've done some light editing on your question and title to make it a bit easier to read. I hope you enjoy exploring the site.
    – Jan Murphy
    Jan 3, 2016 at 3:16

4 Answers 4


If you went around to researchers and asked them this question, I suspect you'd get a different answer for each one. Here are some of your options for recording genealogical information:

  • paper systems designed for genealogy research, such as the forms in the workbook accompanying Emily Anne Croom's book Unpuzzling Your Past
  • software made specifically for 'doing genealogy' (more on this below)
  • keeping your findings in narrative format in a research journal and your search results in software made for general office work (Word / Excel / editor of choice), notekeeping (OneNote/Evernote) or writing (Scrivener)

One problem with using software designed for "doing genealogy" is that most people are only familiar with the lineage-linked genealogy software which based on GEDCOM, a standard created by the LDS Church to communicate data about the lineages we think we know.

In my opinion, lineage-linked software is not well-suited to the task of managing information while we are still in the process of figuring stuff out -- it is focused on keeping track of people, when what the researcher is doing is collecting and analyzing records about people. It is good for recording 'conclusions' -- or if you don't like that term, you can substitute 'theory' or 'your best guess about what's true' -- but not necessarily for Managing the Multiple Maybes.

See What tools exist for collecting and managing evidence? and Transitioning from person-based genealogy to record-based genealogy? for previous questions and answers that touch on this problem.

As people become more familiar with modern standards for doing genealogical research such as the Genealogical Proof Standard, more source-centric software and evidence-management tools are becoming available to us, such as:

  • Lineascope, an online browser-based application that aids the user in analyzing information in sources, and preserving the analysis for use in a proof statement
  • Evidentia, software that supports the user as they go through the analysis of a source (walking you through the process used in the Genealogical Proof Standard)
  • Clooz, software which is designed to help the user inventory and analyze documents and other source material
  • Custodian, designed for people doing large one-name and one-place studies -- it helps "store, index and organize the information" collected for the study

But the best way to answer the question of "what do serious folk do" is to look at the actual work product of accredited and certified genealogists. If you are at RootsTech, NGS, or other large genealogy conferences, look for booths run by the Board for Certification of Genealogists (BCG) or the Association of Professional Genealogists (APG) to see if they have sample portfolios you can examine.

On the web, the BCG offers work samples on their website. Elizabeth Shown Mills, the author of Evidence Explained, offers a series of QuickLessons including QuickLesson 20: Research Reports for Research Success -- the notes in that quicklesson have links to her website Historic Pathways where she generously shares copies of publications and work samples.

Another thing that helps us get better at analyzing records is to learn more about how the records were created by reading journal articles like Claire Prechtel-Kluskens' Who Talked to the Census Taker?, the instructions given out by the agencies at the time the records were created, or the reference information papers and finding aids created by the archivists who have custody of the records now.

Taking your example:

the family folklore/history says he was born in Brooklyn Iowa in 1860. The US Census records of 1890 and 1910 say he was born in Illinois. One Census record shows the family name as spelled Ferrell (two E's). Another record shows him as being born in St Louis MO.

What we all need to do, and often fail at, is to write down why it is that we thought these records belonged to our person, despite the fact that they don't agree with our family stories.

Whatever method you choose to keep track of your research process, until you are comfortable with the concepts in the The Evidence Analysis Process Map, the best thing to do is write your analysis out explicitly, so you can go back later and see what you've done. Even if we don't want to create formal research reports and proof statements for publication, it's still helpful to write these things out for ourselves -- it makes it easier to go back to a problem later and pick up where we left off.

Whether you keep the information in a separate program like Word of Scrivener, or write research notes that get attached to specific people in your lineage-linked software, or something else, write everything out explicitly so it will be easier to review later.

A narrative in a plain text file has the advantage of being easier to access by multiple programs and operating systems. I use Scrivener for a lot of my research notes because the files are stored in Rich Text Format rather than a proprietary format, and can be read independently outside of Scrivener's environment.

  • 2
    Thanks, good information. I agree, the linkage-oriented software packages have an orthogonal design to what is needed for fuzzy facts. I chased through several of your links, and found this discussion to be enlightening. familysearch.org/learningcenter/lesson/… I would never have thought of using Word for any of this. IMHO its too complex, the file format is terrible (Can't store it useful in git) and 99% of the features don't help. Jan 3, 2016 at 4:47
  • 2
    I keep most of my research notes in Scrivener or in Family Historian (my main lineage-linked software). The most promising software approach I've seen to date is STEMMA by @ACProctor -- see his blog: parallax-viewpoint.blogspot.com/2015/11/stemma-v40.html
    – Jan Murphy
    Jan 3, 2016 at 5:23

@JanMurphy in her answer has nicely described the usefulness of source-centric software in a case like this. However, as the most widely used software is lineage-linked, you raise an important question.

There is no right or wrong way to record uncertain information, but I think some approaches are better than others. In my lineage-linked database, my goal is to enter the "fuzzy" information in a way that will not be misinterpreted by myself or others in the future.

For locations, this means defaulting to the level of geography in which I have reasonable confidence. For example, in your case you have evidence that an individual was born in Iowa, Illinois, and Missouri. Obviously all of these cannot be correct, but if I have no indication that any is more likely than the other, then in my database I would not specify a city or state. I would simply enter the birth place as "United States." I would then attach the relevant source material, and explain in the notes why there is uncertainty, what the possibilities are, and what sources you should examine to find more evidence.

"Fuzzy" dates are generally more straightforward as most programs support the circa (c.) or about (abt.) notation before a date, or a date range. Again, I only enter a date if I am somewhat confident in an approximate date. In your example, it would be reasonable to enter the birth date "c.1860". However, if I had no evidence for when an individual was born, or highly conflicting evidence such as birth years over a decade apart, I would rather leave the birth date blank than make a wild guess.

I find "fuzzy" name spellings most problematic of all. Especially in the eighteenth century and earlier, it is not uncommon to find a surname spelled differently even within the same document. I think you have to settle on a convention you want to use and be consistent with it. My approach is to use the name spelling as it was at birth as the primary spelling in my database. By this I do not necessarily mean the spelling that is on the birth certificate or baptism record, rather the spelling that I think was most consistently used by the individual's family around that time. Other name spellings can be entered in notes or as Alternate Name facts.

Obviously there is always a degree of subjectivity when entering uncertain information into a database. Whatever approach you settle on, my advice is to be consistent, and (as Jan mentioned) write good notes explaining your uncertainty - as much for your own benefit as others.


This is actually a very interesting question, and I think there is a reasonably simple and acceptable way of handling fuzzy facts.

My own genealogy adventure started many years ago with a similar set of stories from great-aunts and uncles about my ancestors. There was a tree which was handwritten by my father's aunt. And then there were many family tales about my great-grandmother's brother who came over for a few years from England during the first World War, and after the war left for the United States with his wife and was never heard from again, who I am still looking for.

The simple answer is to record the best information you have. Feel free to put a big asterisk beside it to state what is behind this information.

Put down that your grandfather was born either in Brooklyn Iowa or in Illinois and give the two references to your sources (family folklore, and US Census records). You could add in your belief that he was born within 100 miles of the Mississippi River, but if you do so, you must provide the reason for your belief (it doesn't come out of thin air - did you mother tell you that, or what?)

What has happened here is you simply have not found evidence you trust enough to come to a conclusion about where your grandfather was born. So don't conclude anything yet. Just provide all the information about all the possibilities.

In genealogy software, if I don't conclude the birthplace, then I leave it blank, and I add all my theories as a note, with sources attached to the note.

One day, you'll find some evidence that will give you substantial reason to believe a birthplace. At that point, you can record the birthplace as your conclusion along with the sources that support it. You should then attach the note about your fuzzy birthplace to it, and update the note to reflect your new conclusion and to indicate your earlier beliefs in the light of the new conclusion. One of those earlier beliefs may in the future still prove to be the correct birthplace, so you don't want to lose those. Be sure to also keep the sources that were attached to the note.

So it's really just a matter of recording your conclusions. If you don't have enough information to conclude, then record the possibilities as a note. And always attach your sources.

Same holds true with regards to birthdate, spelling of surname, and everything else.

  • Thanks for your answer. I agree with everything you said, except that its easy. One nit, you said "if I don't conclude the [elided], then I leave it blank, and I add all my theories as a note, with sources attached to the note." I haven't found this to be a good approach. Its handy to have approximate or guessed value, but you need query and reporting tools that deal with the fuzzy values. On my grandfather in the OP, you can say that he was not born in Ireland or even New Jersey Jan 5, 2016 at 6:18
  • Pat: I wouldn't try to fudge my information just for query and reporting. If you don't know, I wouldn't fill in the field, else the fuzzy data will spread to online databases. For reporting, a note is good. For query, your program hopefully has a full text search and you can find Mississippi again.
    – lkessler
    Jan 5, 2016 at 12:48
  • 1
    I think we are agreeing, lkessler. the needs of querying and reporting are orthogonal to assessing facts, fuzzy or not, and coming to conclusions. This is stuff serious historians have been doing since the beginning of written history. Jan 6, 2016 at 3:11

FamilySearch is a free tool that lets you save multiple dates, even for the same event. You can also mark the date you think should be displayed in a person's Summary view.

  • 2
    Hi, welcome to G&FH.SE. I suspect your answer has been downvoted because FamilySearch's Family Tree is a collaborative environment and not really suitable for storing the only copy of one's own research. Many lineage-linked software packages also allow you to save multiple dates for the same event, but there are disadvantages to doing it that way.
    – Jan Murphy
    Jan 3, 2016 at 18:50

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.