Festival calendars
A Socratic walk-through of festival calendars — reasoned out one step at a time, not lectured.
The question we started with
THE QUESTION #Why do a society's largest festivals keep falling at the same points every year?
A society's biggest gatherings are not scattered evenly through the year. They bunch. Midwinter carries several, late autumn carries more, high summer carries a few, and long stretches of the calendar carry almost nothing that anyone travels for.
The tempting explanation is religious: the dates are sacred, so they are where they are. But that just moves the question, since a religion could have sanctified any date at all. What decides which points a year's festivals settle on — and why do they stay there once they have?
Reasoning it through
REASONING #Ask what a large festival actually requires. Not devotion; devotion is available on a Tuesday in March. Three practical things, all at once.
It needs people simultaneously free. A multi-day gathering drawing a whole district is impossible when everyone is needed in a field. In a cereal-farming society labour demand is not flat — it is violently peaked at ploughing, sowing, haymaking and harvest, and slack in between. Those peaks are not a matter of preference: miss the harvest window and there is no year.
It needs a surplus to consume. Feasting is by definition eating more than the daily ration, and that is only possible when the stores are full or when something perishable must be eaten anyway. Grain is at its maximum just after harvest and at its minimum in late spring, before anything new is ready — the interval old English sources call the hungry gap.
And it needs a date everyone can arrive on. This is the constraint most easily forgotten, because we have printed calendars. Without one, a date must be independently checkable by every household: a solstice observed against a horizon marker, a first crescent anyone can see, the flooding of a river, the heliacal rising of a bright star, the onset of rains.
Now overlay those three. The intersection is narrow. After the harvest, all three coincide — labour has stopped, the stores are full, and the sun's position is near a marked point. Midwinter is the second obvious solution, once the ground is too hard to work, the animals that cannot be fed through winter have been slaughtered, and the solstice supplies an unmistakable anchor. That is not a claim that any particular festival was consciously placed; it is a claim about which placements were possible, and possible placements are the only ones that can persist.
Does that predict anything checkable? It predicts that societies on different production cycles should put their great days at different points, tracking their own cycle rather than converging on a shared date. That is what we find: Egypt's year was organised around the Nile's flood rather than the solstice; monsoon societies around the rains; pastoralists around the moves to and from summer pasture; fishing communities around a run that arrives when it arrives. The refuting observation would be pastoral, fishing and cereal societies in the same climate all clustering on the same dates — which would mean the cycle was never doing the work.
There is a second half to the question: why do they stay? Because a coordination point, once established, is expensive to move and cheap to keep. Everyone's plans — markets, fairs, hiring, courts, marriages, debts falling due — accrete onto the date, so each year's observance makes the next year's more certain. That is also why festivals outlive the conditions that placed them: a harvest festival survives in a society where almost nobody harvests.
The analogy
THE ANALOGY #Think of a large family fixing the date of its annual reunion. Nobody picks the date they like best. What actually decides it is the intersection of everyone's free weeks, whoever can host, and when there is money to do it — a narrow window, often just one. And once that window has been used twice, it stops being negotiated at all: it becomes the weekend, and people arrange their year around it rather than the reverse.
a family can phone round and agree, whereas a society without writing has no channel for agreeing anything — so its date cannot merely be convenient, it must be readable off the sky or the river by every household independently, which is why the anchors are astronomical rather than administrative.
Clarifying the model
THE MODEL #Two refinements hold this together. First, "the festival is where the labour trough is" is a statement about large, travelled-to, multi-day gatherings only. Small observances sit happily at labour peaks — blessings before sowing, a harvest-home eaten in the field — because they cost nobody a day. It is the expensive gatherings whose placement is forced. Second, the three requirements are ranked: the anchor is what fixes the date to a day, while labour and food fix it only to a season. That division explains a persistent oddity — festivals whose season is agricultural but whose exact date is astronomical.
Now where my account is weakest. Religions travel by conquest and conversion, carrying their dates intact into places where they fit nothing: a midwinter feast observed in the tropics is not an agricultural fact about the tropics. So some of the clustering we observe is diffusion, not local optimisation, and the two are hard to separate in any single case.
Several familiar supporting stories are also less secure than they sound. The reading of Lent as a ritual dignification of the hungry gap is a functionalist interpretation, appealing and contested, not an established origin. So is the claim that autumn feasts reflect a mass slaughter forced by an absence of winter fodder — historians now regard the scale of that practice as overstated. I have quoted no proportions or dates for these deliberately: the honest position is that the timing pattern is robust while the specific causal story behind any one festival usually is not.
A picture of it
THE PICTURE #How to readThe rim is the year running clockwise from January, and the two curves are the constraints a large gathering must satisfy — illustrative shapes for a northern cereal region, not measurements. A festival needs the labour curve pulled in toward the centre and the stores curve pushed out toward the rim at the same point. Where those overlap is the autumn sector after harvest and the midwinter sector, which is where the great feasts sit. Then look at late spring, where the curves cross the wrong way round — work rising, stores nearly empty — and note what occupies it instead: fasts rather than feasts.
What became clearer
WHAT CLEARED #Festival dates are not arbitrary sanctifications, and they are not really chosen either. A big gathering can only happen where free labour, spare food and a publicly checkable date coincide, and in an agricultural year that intersection occurs at only a couple of places. Those places got used, and once used they became coordination points that everything else was scheduled around — which is why the dates survive intact into societies whose fields, stores and skies no longer have anything to do with them.
Where to go next
ONWARD #- Why fasting and penitential seasons occupy the parts of the year feasts cannot, and whether they are the same mechanism seen from the other side.
- Calendar drift is the adjacent piece and asks a different question: this one asks why a given point in the year was chosen, that one asks why the calendar stops pointing at it.
Key terms
TERMS #| Term | What it means |
|---|---|
| Hungry gap | the late-spring interval when the previous harvest's stores are exhausted and nothing new is ready. |
| Heliacal rising | a star's first reappearance before dawn after a period of invisibility, used as an annual anchor in several ancient calendars. |
| Coordination point | a date whose value comes from everyone expecting everyone else to use it, so it becomes self-reinforcing. |
| Movable feast | an observance whose date is recomputed each year from an astronomical rule rather than fixed to a calendar day. |
Every term the collection defines is gathered in the glossary.