Why Easter Moves

Christmas is on the same day every year. Easter isn’t, and the range is wider than most people realize: it can land anywhere from 22 March to 25 April in the Western calendar, a spread of thirty-five days.

The reason is that Easter is calculated from a lunar rule laid over a solar calendar, and the two don’t divide into each other. That mismatch produced one of the longest-running technical problems in European history, occupied serious mathematicians for over a thousand years, and still generates a different answer in Eastern and Western churches most years.

The more I read about the computus, the more I enjoy it. Monks kept those tables for centuries so that a village with no astronomer could still land on the right Sunday, and I think that’s a heroic piece of work that almost nobody thanks them for. It also creates a practical problem for any working writer who plans content, publishing or events on an annual cycle.

What is the rule that sets the date of Easter?

The short version is the first Sunday after the first full moon on or after the spring equinox. Every part of that sentence has been adjusted.

The equinox is fixed at 21 March by convention, regardless of when the actual equinox occurs, which in some years it doesn’t. The full moon isn’t the real moon either. The calculation uses what’s called the ecclesiastical moon, a tabular approximation designed to be computable centuries in advance without observation. In some years the ecclesiastical full moon falls a day or two away from the astronomical one, and the date of Easter follows the table instead of the sky.

So the working rule is the first Sunday after the first ecclesiastical full moon on or after a fixed nominal equinox. It’s a rule about a model of the sky, not the sky.

That’s a deliberate engineering choice and a good one. An observed rule would require astronomers in every location and would produce different answers in different places. A tabular rule can be printed in a book and used identically everywhere.

I’m firmly on the table’s side here. A rule everybody can compute identically beats a more accurate rule that produces a different answer in every town, and I’d make the same trade in any system I had to run.

What is the computus?

The body of calculation developed to work out the date, and for centuries it was among the most demanding arithmetic ordinary institutions performed.

The problem is that lunar months of roughly 29.5 days don’t fit neatly into a 365-day year. The approximation used relies on a cycle of nineteen years, after which the phases of the moon fall on approximately the same calendar dates. That cycle, and the corrections layered on top of it, are what the computus manages.

Working it out required tables, rules for the leap year interaction, and a way of finding which day of the week a date falls on. Monasteries maintained the tables, disputes about method were serious matters, and the ability to compute the date correctly was a real technical skill. Some historians argue that this had real consequences for European mathematics, since it kept a demanding computational problem alive in institutions during periods when little else required that kind of work. Whether or not the strong version of that claim holds, the labor was real.

Gauss later produced an algorithm that computes the date directly. That’s what most modern implementations descend from.

Why do Orthodox and Western Easter differ?

Because they use different calendars, and both apply the same rule to them.

The Western churches use the Gregorian calendar, introduced in the sixteenth century to correct a drift in the older Julian calendar. Most Orthodox churches continue to calculate Easter using the Julian calendar. That has now drifted thirteen days from the Gregorian.

Applying the same rule to two calendars that disagree about what day it’s produces two dates. They sometimes coincide, and usually don’t, and the Orthodox date is always the same as or later than the Western one.

There have been repeated attempts to agree a common date, including proposals to use an astronomically observed equinox and full moon. That would produce a single answer everywhere. None has been adopted, and the reasons are largely not technical.

For anyone who plans around the season, the gap is a nuisance I’d happily see fixed. The arithmetic was never the hard part, and I don’t expect to see a single date in my lifetime.

For anybody publishing internationally, the practical consequence is that Easter is at least two, and the gap between them can be more than a month.

Which holidays move with Easter?

A major run of the spring calendar. That’s why a single moving date shifts so much.

Everything in the season before Easter is counted backward from it. Ash Wednesday and the whole of Lent, roughly six weeks, is positioned relative to the date. Palm Sunday, Maundy Thursday and Good Friday are the week immediately before it.

Everything after is counted forward. Easter Monday, and then Ascension and Pentecost at fixed intervals afterward.

Mardi Gras and the carnival season sit immediately before Lent begins. That means the whole of that festival calendar moves too, including celebrations in cities where almost nobody is tracking the religious reason for the date.

So a single lunar calculation determines the position of roughly three months of the spring calendar across a large part of the world, and it moves every year.

When is Easter in the next few years?

The spread is the thing to look at, because it shows the problem better than any explanation does.

Across any run of consecutive years the Western date jumps around by weeks, moving late one year and early the next with no pattern a person can hold in their head.

It’s not drifting steadily in one direction. That would at least be predictable. It’s oscillating inside its thirty-five day window according to where the ecclesiastical moon falls.

The extremes are rare. The earliest possible date, 22 March, and the latest, 25 April, both occur very infrequently, and most years cluster somewhere in the first three weeks of April. That clustering is why the problem catches people out. Easter is usually near enough to last year’s date that a reused plan looks approximately right, until the year it’s three weeks out and everything tied to it falls in the wrong month. If you need the actual dates, take them from a computed calendar instead of a list somebody typed. Published lists of Easter dates contain errors more often than you’d expect, because the calculation is unintuitive enough that mistakes are hard to spot.

Typed-up lists of Easter dates make me nervous for exactly that reason. One wrong digit can sit in a list like that for years because nobody’s instinct flags it, and I’ll trust a formula over a stranger’s table every time.

Where did the spring festival come from?

An older layer sits underneath, and The details are contested. The English name is unusual. Most European languages use a word derived from the Hebrew Pesach. It’s where the festival’s own calendar anchor comes from. English and German use something else entirely, and the standard explanation traces it to a pre-Christian spring goddess named by Bede in the eighth century.

That explanation rests largely on a single passage in one source, and scholars differ on how much weight it carries. Some treat it as reliable evidence of an older festival; others think Bede may have inferred the goddess from the month name instead of the other way round.

What’s not in doubt is that northern Europe had spring observances long before the Christian calendar arrived, that the same agricultural reality drives the seasonal timing everywhere, and that customs attached to the season have accumulated from several directions.

The eggs and the hare belong to that accumulated layer and their documented history is much shorter than the folklore claims. Both are attested in early modern Germany and traveled from there. That’s recent by the standards of the arguments usually made about them.

This is a useful case for anyone writing about tradition. The strong origin story is frequently the least documented part, and the real version is more interesting than the confident one. This is the section I enjoy most. People repeat the confident version, with the ancient goddess and the ancient eggs, every spring with total certainty, and the certainty is the part I distrust. I’ll take a messy, documented history over a tidy legend any day.

Why does the date of Easter matter to publishers?

Because most planning tools assume dates repeat, and this one doesn’t.

Anybody building an annual content calendar, a publication schedule or a seasonal campaign works from a template they reuse. Fixed holidays copy forward correctly. Easter doesn’t, and every year somebody discovers that their spring plan is anchored to last year’s date.

The practical failure modes are consistent. Content scheduled relative to Easter drifts a month out of position. Retail and gift-related material misses its window. School holidays, which in many countries move with Easter, shift the audience’s availability without warning.

The fix is to treat Easter as a computed value instead of a stored one. Any calendar system worth using can calculate it, and the rule should never be typed in by hand from memory.

The wider lesson applies past Easter. A great many things a planner treats as annual constants aren’t, including school terms, tax dates and any holiday defined as the nth weekday of a month.

What did the dispute over the date cost?

More than a technical disagreement should. That’s the interesting part.

The question of how to calculate Easter was contested seriously for centuries. Different regions used different methods and arrived at different dates. That meant neighboring communities could be fasting and feasting in opposite weeks.

The English resolution came at a synod in the seventh century, where the competing traditions were argued and one method was adopted. Accounts of it survive and read as a political settlement as much as a technical one, because the choice of method carried an implication about which authority you followed.

That’s worth holding onto as a piece of general history. A dispute about arithmetic became a proxy for a dispute about allegiance, and the arithmetic was settled by deciding the allegiance.

Anybody who has watched an argument about a standard, a format or a style guide will recognize the shape of it exactly.

That’s the part of the story I find most human. People fought over a date for centuries because the date stood for whose side they were on, and arguments over far smaller standards still run on exactly that fuel.

How do you find the date without doing the arithmetic?

Use a computed source, and never a remembered one.

Every serious calendar library implements the calculation, most spreadsheet packages have a documented formula for it, and any reference that lists the date forward for decades is deriving it instead of recording it. If you want to understand the mechanism, Gauss’s algorithm is short enough to work through by hand once, and doing so makes the nineteen-year cycle and the leap year corrections concrete in a way reading about them doesn’t.

I’d put Easter in front of anyone who thinks the calendar is simple. It’s a thousand-year-old algorithm hiding inside a holiday, and every spring it punishes the plan that assumed otherwise.

For the rest of the seasonal calendar on this site, the full index is at literature related holidays, the personal version of the season is at Easter, and the narrative structure underneath the season is covered in resurrection as a narrative structure.

Frequently Asked Questions