August birthdays: facts, records and changes for 2027

Ask what somebody's birthday is and you have already assumed which calendar the answer is in. For a large part of the world that assumption is wrong at least some of the time, because more than one calendar is in daily use, and a birth recorded in one of them does not have a single fixed Gregorian date.

What to take away

• Several calendars run alongside the Gregorian one, and a birthday kept in some of them moves against the Gregorian year.
• The year, the month and the day in one record can each belong to a different system.
• An entry should carry the calendar its date was recorded in. Converting without saying so destroys information that cannot be recovered.

Calendars a record can be in

Type	How it counts	What happens to the birthday
Solar, Gregorian months	Months fixed to the solar year	Recurs on the same date every year
Solar, different year count	Gregorian-like months with another starting point for the year number	Day and month convert cleanly; the year needs an offset
Solar, different months	A different division of the year, such as twelve months of thirty days plus a short thirteenth	Day and month both need a conversion table
Lunisolar	Months follow the moon; an extra month is inserted in some years to stay with the seasons	The observance lands on a different Gregorian date each year, within about a month
Purely lunar	Twelve lunar months with no correction to the seasons	The observance moves earlier through the Gregorian year by roughly eleven days a year

Working out which day a date fell on has always taken tables and rules rather than arithmetic alone.

Why a lunisolar birthday moves

A lunisolar calendar keeps its months aligned with the moon and its years aligned, roughly, with the sun. Twelve lunar months fall short of a solar year, so a thirteenth month is inserted in some years to catch up. Anyone whose birthday is observed by such a calendar has an observance that falls on a different Gregorian date each year. There is no single Gregorian birthday to record. There is the Gregorian date of the original birth and the moving date of the observance, and they are different facts.

A purely lunar calendar drifts without correction, so the observance walks all the way round the seasons over about a third of a century. A birth month in that calendar can fall in a Gregorian summer in one decade and a Gregorian winter in the next.

Year numbers are their own problem

Two records can agree on the day and month and disagree on the year because they count years from different starting points. Under Japanese era names a year is numbered from the start of a reign, so the same Gregorian year is written differently depending on the era it falls in, and an era can change partway through a year. Other systems count from a religious epoch, so the year is offset from the Gregorian one by a fixed amount. A year of birth copied from such a record into a Gregorian list without the offset is wrong by centuries, and the mistake looks like a typo rather than what it is.

What an entry should record

• The date as written, in the script and system of the source.
• The calendar it belongs to, by name.
• The converted Gregorian date, if one is given, with the source of the conversion.
• For a lunisolar or lunar date, whether the entry describes the original birth or the recurring observance.
• Whether the year, the month and the day all came from the same record.

The January index requires a date type on every entry. Calendar is the companion field, and the March page explains why even the Gregorian one is an approximation with a history. The day offset between Julian and Gregorian dates in older European records is a separate matter, covered on the site's historical pages.

Reading a converted date

When a public list gives one Gregorian date for a person whose records are kept in another system, ask three things. Who converted it. Whether the year was converted as well as the day. Whether the date is a birth or an observance. A page that cannot answer has published a guess with a confident face, which is the failure the May page and the June page each describe from a different side.

Common questions

Is a converted date less reliable than a native one?

It is one step further from the record, and every conversion is a place for an error. A conversion that names its method and source is nearly as good as the original. One that does not is a claim about a claim.

Should a calendar site list a lunar birthday under this year's Gregorian date?

Only if it says that is what it is doing, and updates it every year. Fixing the observance to one Gregorian date is wrong in most years.

Do these questions arise for living people?

Constantly. Many people observe a birthday in one calendar and hold documents in another, and both dates in their own life are correct. The July page covers the arithmetic that assumes a single Gregorian date and what that assumption silently requires.