Dates are one of the key ways to characterize events. In fact, every Event needs to be associated with associated with some date. I consider that having some associated date is what distinguishes an Event from a Fact.
Gedcom standards identify several valid date forms. My discussion is only going to cover English language and English style dates as those are virtually the only ones that most genealogy programs are able to support.
My standard is that all dates will be represented as DD MMM YYYY, as in 24 June 2021 or 24 Jun 2021. As I have readers to my websites from both the US and UK, representing dates only numerically can result in errors. For instance, 10/12/2020 who be read in the US as 12th October, but in Britain as 10th December. By explicitly spelling out the month, less confusion should exist.
If the date you have is not recorded as such in a record but come from some record that allows you to approximate the date, use the About date construct, About 27 April 1834. If you are identifying the birth date based on a calculation, say of age at death, then use the Calculated date form, Cal. 27 April 1834. Make sure that you use the event note or a source citation to identify how you arrived at the date.
I NEVER publish a date that I am estimating. My genealogy software always we to make a date private. I add estimated birth dates to virtually everyone to help sort out the index of names in the family file. Having all of the John Lewis grouped by their approximate age in the index makes it convenient when trying to place a record. But I never wish to publish a date that is no more than a guess. I am always afraid that some newbie will take it as real. A date I publish needs to have some basis in a real record.
If you only know the date of an even to precision of Month, you would document it as March 1765; if the the year, then 1765. But use these only if you have a record that firmly places the event in either that month or that year. If you have a age from a census, assume it is approximate. If the 1851 census says a person is 34 years old, I will document their birth as “Abt. 1817”. If the 1900 US census says they were born in April 1864, I will record their birth as “Abt. Apr 1864”.
Range Dates
Gedcom supports two other date forms that I call generically “Range Dates”:
“Between 25 Dec 1880 and 14 Jan 1881” – this says that an event happened at some single date between 25 Dec 1880 and 14 Jan 1881. It could be a person’s death. You know they were alive on the first date because they signed a Will on that date and you know they were dead on the second date because their Will went to probate on that date.
“From 25 Dec 1880 to 14 Jan 1881” – this says that the event continued from the first date to the second date. This could have been a family reunion, and you have documentation that the reunion continued for the duration between those dates. This form is often used when documenting Residence.
In both of these range date forms, the date do not need to be precise to the day; however, Gedcom does not allow use of about or calculated dates.
Most programs allow users to enter dates like 1910-1920 or 1910/1920 to indicate a range date. These forms are not allowed by Gedcom, and I do not use them.
Double Dates
No discussion of Event Dates would be complete without discussing double dates which seem to be the bane of too many genealogists. And are incorrectly implemented, if at all, by some genealogy programs.
Let’s start with a bit of simplified history. In October 1582, Pope Gregory issued a Papal Bull introducing a new Calendar. This was to replace the Julian Calendar that had over time started to have some problems. Most of the Catholic countries introduced it very soon; however, the Protestant countries, including England, continued to use the Julian calendar. One aspect of the new Gregorian calendar is that the year now starts on 1st January, instead of Lady Day, 25th March, in the Julian Calendar. Scotland adopted 1 January as the start of the year in 1600.
The difference in start of the year meant that between 1st Jan and 25th Mar, the Julian world and the Gregorian calendar world were operating in different years. (It is actually much more complicated as we have to consider how leap years affect things and that year by year the two calendars got more out of sync. See Wikipedia if you need to see all the gory details.)
In the most of the English speaking world, most church and state records continued to use the Julian calendar, but the English world that had to deal with international commerce began to use the Gregorian calendars. As a result you started to see dates in records like 23 February 1723/24 or 23 February 1723 o.s. or 23 Feb 1724 n.s. All of those are the same date. Actually, what you are seeing are Julian Dates with an indication of what the Gregorian year would be.
The dates of the two calendars continued to move further out of sync so that it was decided that in 1752, the English world, including the American colonies, would adopt the Gregorian calendar. The year 1752 then began on 1 January. There remained the problem of aligning the calendar in use in England with that in use in Europe. It was necessary to correct it by 11 days: the ‘lost days’. It was decided that Wednesday 2nd September 1752 would be followed by Thursday 14th September 1752.
What this means for documenting your genealogy is that between 1582 and 1752, you have to carefully document what year dates between 1st Jan and 24th Mar represent. Sometimes the records are sufficiently clear if you can read the originals. Transcriptions can be more problematic. And dates in compiled genealogies are usually wrong, even if they have attempted to use double dating.
My standard is always to use double dating and trying to determine from original records what is the correct Julian year. I do make the assumption with a transcription or other record that does not make the year clear that I am looking at a Julian date. I usually make note if the year is estimated.
Double dates would generally be of the form 22 February 1722/23; however, you would also have 22 February 1699/00. Remember, that dates in the double date period are Julian calendar dates and in the period between 1st Jan and 24th March the double date just indicates what the Gregorian calendar year would be.
