Estimating dates and ages in genealogical writing

Beginner Tips
A sundial shows the time.

Dates and ages are fundamental elements of genealogy. As many researchers know, it's not uncommon to arrive at a date or age in our family history that we know isn't exact - but when you do, what's the proper way to document your estimate and describe it in your writing? And how do you come up with the estimation in the first place? 

The reality is that estimated dates and ages are acceptable and necessary when documenting your family history - nobody is able to confidently confirm all dates and ages they encounter. The key is making sure the estimation is accurate and communicated properly to your audience in your writing and documentation. 

The following article appeared in the Spring 2010 issue of the New York Researcher, and was written by Patricia Law Hatcher, FASG, - the former editor of The New York Genealogical and Biographical Record. She is the author of over two hundred published articles on genealogical subjects. We hope your enjoy the article:


Dates are fundamental to genealogy — and it is fundamental that genealogists estimate them properly, and use them correctly in their writing.

Expressing Dates

Except when quoting from an original source, dates should be expressed in genealogical format 2 March 2010 (no commas).

If the original does not give a standard date with month, day, and year, the original text should be preserved, with the interpretation provided in editorial square brackets so that the reader can be assured that the date is correct: 1st day 3rd month [May] 1730, 1st day 3rd month [March] 1830, 9 7ber [September] 1800, 12th next [April] 2010, first Monday in Easter [5 April] 2010.

Bracketing Events

Very often we do not have an exact date for a vital event. When possible, it is best to define the vital date in terms of dated events that we do have.

For example, instead of saying John Smith died in 1708, it is genealogically better to say he died between 9 June 1708 (will) and 10 August 1708 (probate), because in this way it is clear that he could have been the man who sold land in February 1708, but could not have been the man who sold land in December 1708.

When bracketing death dates, think beyond events directly related to death: he died between 30 June 1708 (deed) and 30 July 1708 (wife called widow in church membership).

Using About/Circa and Say

In scholarly publications we use specific terminology to describe how we have determined a non-exact date.

About or circa (Latin for about) are used to indicate a date calculated from an inexact age, as on a census, deposition, or tombstone: born about 1847 (1850 census), born about 1798 (age at death on gravestone), born about 1740 (deposed about 45 years in 1785).

Say indicates a date estimated from life events (see below): born say 1790 (first land purchase in 1815), married by say 1702 (child baptized in 1703), born by say 1677 (estimating marriage at 25). It is based on statistically common patterns and therefore could be off by a number of years.

Tips for Estimating Dates

Estimation of birth years based on life events, together with the construction of family groups, is an essential genealogical technique. In the introductory material to each volume in the Great Migration project, Robert Charles Anderson, FASG, discusses its value:

Criteria for approximating dates: When we do not have an exact date for a vital event, such as a birth, baptismal, marriage or death date, we will in all instances create an approximated date for that event. We do this for a number of reasons. Sometimes this type of chronological analysis will reveal an unsuspected contradiction in previous treatments of a family, indicating perhaps that not all the children of a man could have been born to his only known wife. …

Ordering families: We wish to fit the children with well-estimated years of birth. … This process of determining the order of birth of the children in a family, although timeconsuming, frequently provides some of the best new data on that family. . . . the value of this process in pointing out conflicts and contradictions and in directing further research is worth the effort.

Over and over again, I too have found this process to be valuable. For The Record I require a fully documented Genealogical Summary, even for simple articles — often generating surprising new conclusions. When exact dates are lacking, I go through this age-estimation and child-ordering process.

The pay-off is amply shown, for example, in the greater understanding of the Moore, Osborne, and Titus families. An unexpected finding was the likelihood that at least some — probably all — of the eldest sons of John Moore were not by his wife Margaret Howell (with her prestigious ancestry) as has long been stated.

There are several key life events on which we rely to establish ages: earliest appearance in records, appearance in positions of authority, first evidence of land ownership, and first marriage. For women we look at span of childbearing, especially the birth of the last child.

For this methodology to be useful, it is necessary to focus equally on all members of a family, sometimes even following their children in order to determine likely ages and hence birth years to understand potential placement within a family group.

Cautious researchers will realize that there is a meaningful difference between the age at which someone could do something and the age at which they typically did that something. A case in point is being a witness:

The age of majority was twenty-one. Minority was divided in three sections of seven years each. Between fourteen and twenty, a child could be responsible for his or her own actions. Therefore, a fourteen-year-old could legally witness a document. However, since everyone knew that person was a minor, they were not likely to ask them to be a witness if it could be avoided. A minor would have been a witness only if there were no adult available, perhaps when death was unexpectedly imminent.

Men did not usually achieve positions of importance in a community until their thirties (late twenties at earliest). Men from prominent families might well be given such positions earlier than lowly farmers. Not all town offices or military roles indicate importance; some were routinely given to men in their twenties. The researcher will need to understand the duties involved in the position in question.

Estimating Marriages and Births

Family researchers tend to believe that first marriages occurred much earlier than was the norm. Although in the early colonial period marriages in England were permitted at a very young age (fourteen for girls, sixteen for boys), for practical reasons they usually did not occur until the husband was in a position to support a family. Most men simply were not established enough to support a wife and family until they were in their mid-twenties.

In England an apprentice could not marry. Apprenticeships typically ran until age twenty-one, but in cities that age was twenty-five. A man generally married a woman near his own age whom he had known much of his life. It was not uncommon for the wife to be a year or two older. This understanding of the appropriate ages for marriage was carried to early America, but with a bit more flexibility.

A couple’s first child was typically born a year after the marriage, with subsequent children arriving every two years. An exception, but one rarely seen in America, was among very well-to-do families in which a wet nurse was employed, in which case the children might be more closely spaced.

This was the norm, but exceptions did, of course, occur (in statistics these are called outliers). Early pregnancy was a common reason. Men who came from well-to-do families did not necessarily have the impediment of needing to earn the capital to establish a home or farm. On the other hand, if they sought higher education, they often married much later than the norm.

Older men, such as widowers, often seemed to marry much younger women. An exception that has bearing on the Moore-Osborne article is that children, particularly daughters, living in a household with a stepparent, were apt to marry at an age much younger than typical, especially if the situation was unpleasant or if both biological parents had died.

First marriage for men near the age of twenty-five and women at twenty or twenty-one holds up well, statistically, over a broad range of time and place. It is inappropriate for genealogists to grasp at an unlikely birth date or marriage age because it is convenient or gives a preferred ancestry (as with Margaret Howell).

It is vitally important that for estimated dates we stick to the estimating protocol throughout the article or book. We do not adjust dates here and there to make family groups look neater. For example, if three sisters all married in 1745, each of their entries in the Genealogical Summary would say born say 1725 (estimating marriage at 20). We do not have a crystal ball to tell us their birth order.

Responsible practices in expressing dates and ages increases both genealogical accuracy and the potential for future research.


by Patricia Law Hatcher, FASG

Originally published in The New York Researcher, Spring 2010

© 2011 The New York Genealogical and Biographical Society

All rights reserved.