Naming the days of the week

Ever wondered where the weekday names come from? The answer may go back a bit further than you think. But it’s also incomplete.

The easy part of the answer is that the days are named after the seven planets. Not the modern solar system, though. The planets as they were known to ancient astronomers like Ptolemy. And in English, six of the seven planet names were swapped out, and Old English words or gods’ names put in. But the idea’s the same.

Sunday. Sun; relief of Sol (Museo Nazionale Romano, Rome)

Planet	Roman name	Old English	Modern English
Sun	dies Solis	Sunne	Sunday
Moon	dies Lunae	Mōna	Monday
Mars	dies Martis	Tiw (~Týr)	Tuesday
Mercury	dies Mercurii	Wōden (~Odin)	Wednesday
Jupiter	dies Iovis	Þonar (~Thor)	Thursday
Venus	dies Veneris	Frīg (~Frigg)	Friday
Saturn	dies Saturni	—	Saturday

In antiquity the Sun and Moon were frequently counted among the planets because, like the planets, they travel along the ecliptic relative to the fixed stars. Sometimes ancient writers talk of seven planets (including Sun and Moon), sometimes just five.

That’s the easy bit. I think the more interesting questions are:

1. When did Greco-Roman gods get attached to the seven day cycle?
2. Why are they in that order?

Monday. Moon (CambridgeInColour.com); Luna in chariot (Arch of Constantine, Rome)

Roman weekday names

The weekday names apparently arrived in Rome in the late 1st century BCE. The earliest reference to a day bearing one of the modern names is in the elegiac poet Tibullus, in a poem dating to the early 20s BCE.

aut ego sum causatus aves aut omina dira
      Saturnive sacram me tenuisse diem.

I used birds or bad omens as a pretext,
      or that the day sacred to Saturn detained me.

Tibullus 1.3.17–18

Petronius’ Satyrica alludes to all seven planets being associated with weekdays. However, the date of the Satyrica is in some doubt these days. It used to be dated to the reign of Nero (54–68 CE), but some scholars now think it’s a 2nd century novel.

... two timetables were posted, one on each doorpost. One of them, if I recall correctly, had this note: ‘third day and day before Kalends of January [30–31 December]: our C. dines out.’ The other depicted the course of the moon and paintings of the seven stars.

Petronius, Satyrica 30

The modern sequence of days appears in two graffiti no later than 79 CE, found at Pompeii. One is in Greek; the other, in Latin, omits Wednesday. Both graffiti start the week on Saturday.

θεων ημερας | κρονου | ηλιου | σεληνης | αρεως | ε[ρ]μου | διος | [αφρο]δειτης
Days of the gods: (day of) Kronos, Sun, Moon, Ares, Hermes, Zeus, Aphrodite

Mau 1881: 30

saturni | solis | lunae | martis | iovis | veneris
(day of) Saturn, Sol, Luna, Mars, Jupiter, Venus

Sogliano 1901: 330

Note. These and some further references are helpfully compiled by Schürer 1905: 25–34.

A graffito similar to the second has been found in Tunisia at Thuburbo Maius, not far from Carthage (IL Tun. 710 = Merlin 1944: 126), but that’s probably a couple of centuries later. It too starts the week on Saturday. The canonical sequence of seven appears in many later writers — including a short 4th century poem by Ausonius. In Ausonius, the week now begins on Sunday.

Some of these later writers were puzzled by the order. Plutarch, in the 2nd century, devoted a section of his Table talk to the question ‘Why they name the days after planets but number them differently from their sequence’ (Plutarch, Moralia 672c). Unfortunately that part of the Table talk is lost.

Tuesday. Mars (Fvalk.com); C. F. von Saltza, ‘Týr’ (F. Sander, Edda Sämund den vises, Stockholm, 1893, p. 78)

The order of the seven (or five) planets

The five planets, aside from the Sun and Moon, are the ones that are visible to the naked eye: Mercury, Venus, Mars, Jupiter, and Saturn. (Technically Uranus is too — occasionally, and with a keen eye. But only for a rather Simpsons-esque sense of ‘technically’.)

Those who are skilled in astronomy say that there are seven bands, on which the seven stars are carried. On the highest is carried the star of Kronos; on the one after that the star of Zeus; on the third the star of Ares; on the fourth the star of the Sun; on the fifth the star of Aphrodite; on the sixth the star of Hermes; and on the seventh the star of the Moon.

Achilles, Eisagoge 16 (Maas 1898: 42,25–30)

So the standard sequence, starting from the outermost planet, is: Saturn, Jupiter, Mars, Sun, Venus, Mercury, and Moon.

This arrangement isn’t really ‘Ptolemaic’, or only incidentally. Ptolemy himself allowed doubt over whether the inner planets are ‘beneath’ or ‘above’ the Sun, since the Sun’s brightness makes it impossible to observe whether Venus and Mercury pass in front or behind. And he’s explicit that he can’t measure the planets’ distance, since he can’t measure any parallax. He just accepts the conventional sequence as, well, a convention.

πιθανωτέρα μᾶλλον ἡ τῶν παλαιοτέρων τάξις καταφαίνεται ...

the order assumed by the older [astronomers] appears the more plausible ..

Ptolemy, Almagest 9.1 (ii.207 Heiberg; tr. Toomer)

Some other features of the ‘Ptolemaic’ system aren’t really Ptolemaic either. Ptolemy has no word for ‘deferent’, the circular orbit on which the epicycle is centred (though he does use the concept); he doesn’t give figures for the distances of the planets. That’s all mediaeval.

Wednesday. Mercury (Cronodon.com); Ian McShane as Wednesday (promotional poster for American gods, 2017–2021)

The standard order took time to become settled. Otto Neugebauer gives a handy run-down of the different sequences seen in different ancient cultures and different ancient authors (1975: 690–693). Some of them omit the Sun and Moon; some are reversed; some omit Jupiter and Saturn. His verdict is that no standard order existed prior to the time of Hipparchos, in the 2nd century BCE.

Egyptian	Jupiter, Saturn, Venus, Mercury, Mars
older Babylonian	Jupiter, Venus, Saturn, Mercury, Mars
Persian/Hellenistic era Babylonian	Jupiter, Venus, Mercury, Saturn, Mars
Archimedes according to ps-Hippolytos Refutatio 4.7–11	Moon, Sun, Venus, Mercury, Mars, Jupiter, Saturn, fixed stars
Plutarch On generation 1028b (Pythagorean)	Central fire, Counter-earth, Earth, Moon, Mercury, Venus, Sun
Plutarch On generation 1029b	Mercury, Venus, Mars, Jupiter, Saturn
‘some’ according to Achilles, apparently including Eratosthenes (Maass 1898: 42,30–43,2)	[Mars], Venus, Mercury, Sun, [Moon]
‘others’ according to Achilles (ibid.)	[Mars], Mercury, Sun, Venus, [Moon]
Eudoxos papyrus, inscription of Keskinto	Saturn, Jupiter, Mars, Mercury, Venus
Vitruvius Architecture 9.1.5	Moon, Mercury, Venus, Sun, Mars, Jupiter, Saturn
Ptolemy Almagest 9.1, Cicero On divination 2.91–92, Pliny Natural history 2.34–44	Saturn, Jupiter, Mars, Sun, Venus, Mercury, Moon
Cicero On the nature of the gods 2.52–53	Saturn, Jupiter, Mars, Venus, Mercury
(Indian) Varāhamihira Pancha-siddhantika 13.39 (i.121 Neugebauer-Pingree)	Moon, Mercury, Venus, Sun, Mars, Jupiter, Saturn, nakṣatras

(Notice, by the way, that Vitruvius and the 6th century Indian astronomer Varāhamihira have the same sequence as Ptolemy, but reversed; so do Cicero and Plutarch, reversing one another, but both omitting the Sun and Moon.)

In spite of the variation, and the fact that Ptolemy regarded the sequence as purely conventional, the ‘Ptolemaic’ sequence came to be universally regarded as the standard order. Ptolemy calls it ‘the order of previous astronomers’ (ἡ τῶν παλαιοτέρων τάξις), but as mentioned above, it’s inconsistent with pre-2nd century BCE sequences, Neugebauer suggests that no sequence existed until around the time of Hipparchos.

Thursday. Jupiter (Aol.com); Chris Hemsworth as Thor (promotional poster for Thor: the dark world, 2013)

The order of the weekdays

Here’s how the planets get reordered into weekdays, starting with Saturn, as in Ptolemy and the Pompeii graffiti.

We don’t actually have any good evidence on why this reordering happened. It clearly isn’t random, though: notice how for each weekday, you skip two planets — or conversely, for each planet, you skip four weekdays

One modern book on the history of the week, by Eviatar Zerubavel, favours a theory based on the premise that Egyptian astronomers assigned each of the planets to hours of the day (1985: 14–17); however, everything about this theory is hypothetical. It’s claimed by an ancient author, as we’ll see below, but not a very trustworthy author. It can’t be corroborated as anything more than, well, some ancient guy making guesses.

There are three ancient theories on record, including the one Zerubavel prefers. Personally I think all three are pretty tenuous. The lost essay by Plutarch that I mentioned above may have had a fourth explanation, but alas, we’ll never know what it was.

Friday. Venus (Deepsky2000.com); C. E. Doepler, ‘Frigg and her handmaidens’ (W. Wägner, Nordisch-germanische Götter- und Heldensagen, 3rd ed. 1882, p. 109)

The first theory comes from a 6th century Indian astronomer, Varāhamihira.

(Ascending) up from the Moon (each successive planet) is lord of the month, (descending) down from Saturn lord of the hour. (Ascending) up in order (every) fifth (planet) is lord of the day; the lords of the year are clear.

Varāhamihira, Pancha-siddhantika 13.42 (i.121 Neugebauer-Pingree)

This is simply a restatement of the pattern I mentioned: start from Moon = Monday, then for each successive day, move on five planets (counting inclusively; four, counting exclusively). Why anyone would do that, he doesn’t explain. So this is a pretty weak theory.

The second and third are found in Dion Cassius (3rd century CE). He attributes them both to ‘the Egyptians’, which is ... doubtful, to say the least. But let’s hear him out. Here’s theory number two:

For if you apply the so-called ‘principle of the tetrachord’ (which is believed to constitute the basis of music) to these stars, by which the whole universe of heaven is divided into regular intervals, in the order in which each of them revolves, and beginning at the outer orbit assigned to Saturn, then omitting the next two name the lord of the fourth, and after this passing over two others reach the seventh, and you then go back and repeat the process with the orbits and their presiding divinities in this same manner, assigning them to the several days, you will find all the days to be in a kind of musical connection with the arrangement of the heavens.

Dion Cassius 37.18 (tr. Cary)

The tetrachord was indeed the basic element of ancient Greek music: it was a sequence of four notes, spanning an interval of what we would call a perfect fourth (a frequency ratio of 4:3). That is, this theory is that you rotate through the planets in the same way that musical keys modulate through a cycle of fifths.

That isn’t as crazy as it might sound. These are all real things: the tetrachord, the Pythagorean preoccupation with perfect harmonic intervals, and the idea that the planets are tied up with musical theory in some way. Unfortunately we know basically nothing about Pythagoreanism in the 1st century BCE, so we have no way of deciding whether this theory is plausible, or completely daft.

Saturday. Saturn (UniverseToday.com); relief of Saturnus from altar of Malakbel (Palmyra, Syria), Musei Capitolini, Rome (Wikimedia.org)

Here’s theory number three, also from Dion Cassius:

If you begin at the first hour to count the hours of the day and of the night, assigning the first to Saturn, the next to Jupiter, the third to Mars, the fourth to the Sun, the fifth to Venus, the sixth to Mercury, and the seventh to the Moon, according to the order of the cycles which the Egyptians observe, and if you repeat the process, covering thus the whole twenty-four hours, you will find that the first hour of the following day comes to the Sun. And if you carry on the operation throughout the next twenty-four hours in the same manner as with the others, you will dedicate the first hour of the third day to the Moon, and if you proceed similarly through the rest, each day will receive its appropriate god.

Dion Cassius 37.19 (tr. Cary)

This one depends on the premise that each hour of the day is assigned to a planet, and that they’re assigned in the Ptolemaic order.

Day	Day begins at hour	Associated planet
1	1	Saturn
2	25	Sun
3	49	Moon
4	73	Mars
5	97	Mercury
6	121	Jupiter
7	145	Venus
8	169	Saturn

and so on. This sounds kind of plausible. It also has the advantage of predicting Saturday as the first day of the week, which is exactly what we see in early sources like the Pompeii graffiti.

There are problems, however. First, there’s no corroboration in anything we know about Egyptian astronomy for the idea of assigning planets to hours. It could easily be a figment of Dion’s imagination.

Second, Egyptian astronomy is characterised much more by a division of the day (and night) into 12 hours, not 24. According to Robert Hannah, the concept of the 12 hour day is precisely of Egyptian origin (2005: 87). As Herodotos puts it,

But as far as human affairs are concerned, [the priests in Egypt] agreed on this: that the Egyptians were the first to discover the year, and the division of it into twelve seasonal segments; and they discovered this from the stars, as they said. ... They also said the Egyptians were the first to refer to a canon of twelve gods, and that the Greeks adopted this from them ...

Herodotos 2.4 (my emphasis)

(A 12 hour cycle would produce the same result, if we start from the Moon and work our way out. But then we lose the advantage of matching early sources by outputting Saturday as the first day of the week.)

And third, what we do find in actual Egyptian astronomy is the idea of associating hours with specific stars or constellations, not planets.

The hours became associated with certain stars or star groups which rose heliacally at ten-day intervals through the year. Sirius was one of these, and it was joined by 35 other stars ... Collectively they are now known as the ‘decans’ ...

Hannah 2005: 87

This produced a system of ten-day weeks in a seven-week cycle, not a cycle of seven days. Theory number three gives every appearance of being a post hoc rationalisation of the weekday names, not a true explanation.

None of the three theories has any corroboration. Theory 1 is certainly the weakest. But the mismatch between theory 3 and what is actually known about Egyptian astronomy is so glaring that I think it has to be rejected almost as strongly.

The weird result is that the Pythagorean explanation — theory 2, rotating between the planets in musical tetrachords — is the strongest.

Not that it’s a good theory, mind. It sounds quite daft to me. It’s just that, as things stand, we don’t have anything to rule it out.

References

• Hannah, R. 2005. Greek & Roman calendars. Constructions of time in the classical world. London.
• Maass, E. 1898. Commentariorum in Aratum reliquiae. Berlin. [Internet Archive]
• Mau, A. 1881. ‘Scavi di Pompei.’ Bullettino dell’Instituto di corrispondenza archeologica 1881,i–ii: 22–32. [Internet Archive]
• Merlin, A. 1944. Inscriptions latines de la Tunisie. Paris.
• Neugebauer, O. 1975. A history of ancient mathematical astronomy. Berlin/Heidelberg.
• Schürer, E. 1905. ‘Die siebentägige Woche im Gebrauche der christlichen Kirche der ersten Jahrhunderte.’ Zeitschrift für die Neutestamentliche Wissenschaft 6: 1–66. [Zenodo]
• Sogliano, A. 1901. ‘Regione I (Latium et Campania).’ In: Notizie degli scavi di antichità comunicate alla R. Accademia dei Lincei, anno 1901, s.v. ‘Luglio 1901’. Rome. 329–333. [Internet Archive]
• Zerubavel, E. 1985. The seven day circle. The history and meaning of the week. Chicago.