The twelve months and the stars they are named for

Every one of the twelve Indian lunar month names is a star's name first. Chaitra, Vaishakha, Jyeshtha, Ashadha, Shravana, Bhadrapada, Ashwina, Kartika, Margashirsha, Pausha, Magha, Phalguna: each of those belonged to a patch of sky before it belonged to a month.

Say them out loud, in whatever shape your family says them. You have met that list on a wedding card. You have met it in the top corner of the panchang (pañcāṅga, पञ्चाङ्ग), the printed almanac sheet, and in the mouth of whoever at home announces that a thing will not be done in Pausha. Nobody sat you down and taught it to you. You absorbed it, the way you absorbed the names of relatives you have only seen twice. And like those relatives, each of the twelve has a history nobody thought to mention.

Here is the piece of sky the names come from. The moon goes once round the sky against the fixed stars in about 27 days, and the old reckoning cut its road into 27 stations, the nakṣatras (नक्षत्र), each one named for the star or the small cluster of stars sitting in it.

Once in each round the moon comes to the full. When it does, it is standing in one of those stations, and you can go outside and see which. That is the whole basis of the naming. The full-moon day on which the moon stood near Chitrā (चित्रा) was called Chaitri, and the month that held that full moon was called Chaitra. Robert Sewell and Sankara Balkrishna Dikshit set it out in one sentence in their 1896 handbook of the Indian calendar, and no part of it is complicated.

So the month is named after the star the moon was visiting on the one night of the month when everybody was looking up anyway.

Now the twelve, all of them, because the pattern only shows itself when you have the whole set in front of you.

Three months carry their star's name with almost nothing done to it. Jyeshtha the month comes from Jyeshtha the station. Shravana from Shravana, Magha from Magha. In roman letters they look identical. In Sanskrit the month's first vowel is drawn out a little, which is the ordinary way of turning a name into "the one belonging to that name", and English spelling loses the difference.

Six more do the same thing loudly enough to hear. Chitra gives Chaitra. Vishakha gives Vaishakha. Krittika gives Kartika. Mrigashira gives Margashirsha, which older books also call Agrahayana. Pushya gives Pausha. Ashwini gives Ashwina, written in some texts as Ashwayuja, after the station's older name.

The last three are named for a pair rather than a single star. Several stations come in twinned sets, an earlier and a later, pūrva (पूर्व) and uttara (उत्तर). Ashadha takes its name from the two Ashadhas, Bhadrapada from the two Bhadrapadas, and Phalguna from the two Phalgunis.

Three, six and three. That is the calendar, and there is not a single month in it whose name came from anywhere but the moon's road.

Two full moons in one year

A rule like that invites checking, so check it. Take the Delhi sheets for 2026 and look at two of the year's full moons.

Kartika's full moon falls on Tuesday 24 November. The pūrṇimā (पूर्णिमा), the full-moon lunar day, ends at 20:23 that evening, and the minute it ends is the minute the moon is exactly full. Krittika runs that day from 02:02 in the morning until 23:25 at night. So at the moment of the full moon, the moon was standing in Krittika, which is the star Kartika is named after. The appointment was kept.

Now April. Chaitra's full moon ends at 07:41 on the morning of Thursday 2 April. Chitra does not begin until 17:38 that evening. At the moment the moon came full it was still in Hasta, the station just before Chitra, about 10 hours short of the star the month is named after.

One year, one set of sheets, one hit and one miss.

[VERIFY: both full moons are read off the 2026 Delhi sheets; recheck the November figures if this post is dated after 24 November 2026]

Nine pairs and three threes

The miss is the normal case, and the almanac-makers knew it long before either of us did.

Three things work against a tidy answer. The twelve namesake stars are not spread evenly round the circle. The moon does not travel at a steady speed. And the month measured from one full moon to the next is not the same length as the moon's circuit of the stars. Put those together and the full moon simply cannot come back to the same station year after year.

So each month was given a band of stations instead of one. Sewell and Dikshit print the table. The 27 stations are handed out among the twelve months in order: nine months get two stations each, three months get three. The run begins at Krittika, which is where the oldest Indian star lists begin, and closes the circle at Bharani. Under that scheme, a Chaitra full moon in Chitra and a Chaitra full moon in Svati are both perfectly ordinary. April's, in Hasta, sits outside even that band, in the stretch the table hands to Phalguna.

Even the bands were not enough, and this is the part that changes how you should hold the names. The months stopped being named from full moons altogether. The sun does the naming now.

Every month the sun crosses from one sign of the zodiac into the next, and the crossing is called a saṅkrānti (सङ्क्रान्ति). The lunar month in which the sun crosses into Mesha (मेष), the first of the twelve signs, is Chaitra, and the other eleven follow it in order. That is the working rule on every sheet printed today. Where the moon stood at the full has nothing to do with it. Once in about three years a lunar month catches no crossing at all, and that month is handled by a separate convention.

Two smaller things, so they do not trip you later.

The star rule hangs on the full moon, and the full moon is the one night on which both ways of cutting the month agree. A purnima carries the same month name whether your household ends its month at the no-moon night or at the full one, which is why this naming worked across the whole country at once, and why amanta and purnimanta is a quarrel about dark fortnights and never about this.

And underneath all of it the frame itself is moving. The circle of stars and the circle of the seasons pull apart by roughly a degree every 72 years, so the months no longer sit in the year quite where the old texts left them; that slippage is precession, and the slow clock of precession measures it properly.

There is one more layer under the star names, and it is worth a minute because it tells you what these names replaced.

An older set of twelve is named in the Yajurveda: Madhu and Madhava for the spring, Shukra and Shuchi for the heat, Nabhas and Nabhasya for the rains, then Isha and Urja, Sahas and Sahasya, Tapas and Tapasya. Twelve names taken from weather and crops rather than from stars. Dikshit's note in the 1896 handbook points out that these season names are there in the Vedic hymn collections while the star names are not, and that the star names turn up later, and sparingly, in some of the Brahmanas. Exactly when the change happened is still argued over, and I would not hand you a date.

What is plain enough is the direction of it. The first set of month names looked at the fields. The second set looked up. The second set is the one that survived into your grandmother's mouth and onto your wedding cards, and it survived long after the rule that produced it had been quietly replaced by a sum about the sun.

Which is worth taking back outside once, to the place the names came from. Your next full moon is already printed on whatever sheet your house keeps, with a nakshatra sitting in the column beside it. If the moon is standing in the month's own star that night, you are watching the name being made. If it is one station short, you are watching the exact reason the almanac-makers stopped relying on it. Either way the month keeps its name, which is what becomes of a name once it has stopped being a measurement.