The Magi and Mithras: what 561 stones naming the god show

The theory, put to the test

What would have to be true#

The theory has a simple logic: in the first two centuries a large part of the Roman world served Mithras, a god with a Persian name; the Magi are Persian priests; so someone put them into the birth story to make the new faith look acceptable to those who were “not yet convinced”.

Three things follow that can be checked. First: in the years the story was written, followers of Mithras had to be a crowd. Second: they had to be where the story was told. Third: the Magi had to look like priests of Mithras from the start.

At the bottom corner the curve shows how many of the 442 dated stones naming Mithras can exist by each year; the small stars mark the years the story of the Magi is on record in writing.

First test: the stones

No inscription naming Mithras in this database can be securely dated to AD 100 or earlier#

The Gospel of Matthew is usually dated between 70 and 100, most likely 80–95 (Davies and Allison), and Antioch in Syria is the city proposed most often, without excluding others. A count of stones shows what a cult left in writing; its population remains unknown. This dataset contains only Latin inscriptions, which limits it as well: the audience of a story written in Greek in Syria would more likely have left Greek inscriptions, which are not in the database.

The database (LIST v1.2, with the Latin inscriptions of EDCS and EDH) assigns four stones naming Mithras to the period up to the year 100. None provides a convincing basis for placing the stone securely by AD 100: for Aquileia and Tortona the dating has no verifiable basis; the Mainz altar is dated 201–250 by EDCS and 200–250 by Ubi Erat Lupa, without a stated basis, and 71–100 comes only from EDH, after a restoration of the name of the unit; here it is left undated; the Carnuntum altar is dated from the period the legion was stationed there, between 71 and 113. With the datings chosen here, no date range ends by 100; three have a range ending by 120: Hyginus in Rome (68–117), Carnuntum (71–113) and Alcimus in Rome (101–120). In all, 4 have a range that begins before 100, Hyginus and Carnuntum among them; 0 of them carry only the database’s default range “from the year 1 to 300”, which is a placeholder, not a date.

With LIST’s dates as they are, four ranges end by 100 and 16 begin by then; with EDCS’s own dates, three and 13. The table “The oldest stones” shows them one by one. The scholar who re-examined the evidence in 2016 writes that no monument of the Roman cult can be persuasively dated before 75, and that the evidence stays “relatively negligible” until the second half of the 2nd century.

The few literary references are roughly contemporary with the earliest inscriptions, or slightly earlier: Statius (written about 80–92) mentions Mithras in a cave he calls “Persean”, and Plutarch writes, about 100, that the rites of Mithras of the Cilician pirates “continue to the present time”. Cilicia is next to Syria; neither speaks of followers at Antioch.

A century later

By 160 Justin retells the story; the stones are still few#

One of the earliest authors to repeat the story of the Magi is Justin, before 161: they came “from Arabia”, he says (Dialogue with Trypho 77–78). Writing around 150, he also gives the earliest surviving Christian reference to Mithras, and calls the mysteries a demonic imitation of the Eucharist (First Apology 66).

By this point the map holds at least 12 stones naming the god; the estimate is 59 of the 442 dated ones, between 31 and 87 depending on how the datings are assumed to be spread inside their ranges: a convention of calculation, not an observed total. By 250 there will be between 223 and 402.

The theory therefore needs the Magi to have been written into a text that was already circulating in the 80s or 90s, for an audience that at the time had left almost no Latin inscriptions in this database.

When the stones multiply

The stones become numerous between 150 and 250#

In the second half of the 2nd century and the first of the next, the lights come on along the Danube and the Rhine, in Rome and in Ostia: by 150 there were between 9 and 96 stones, by 250 there are between 223 and 402 of the 442 dated ones. 69 of the 561 carry the database label “soldiers”, and many name freedmen, slaves and customs men.

This is the time when the story of the Magi is already quoted at Lyon, at Alexandria, at Carthage. The stones of Mithras become numerous after the story of the Magi was written.

Roman Dacia

In Dacia, the evidence for Mithras comes after the Gospel#

Dacia was conquered in 106; no Roman altar of Mithras is known there earlier. The database holds 54 stones naming the god in the province: 24 at Apulum and 10 at Sarmizegetusa. The earliest mithraeum in Dacia is of the time of Hadrian (117–138), according to Sicoe’s catalogue as cited by Szabó.

As a local illustration: in Dacia the evidence for the cult of Mithras is later than the Gospel, and the earliest known mithraeum comes more than twenty years after the story of the Magi was written.

Second test: place and text

What Matthew’s text says#

The Greek says “magoi from the east” (Matthew 2:1), who saw a star (2:2): their number is not given, they are not called kings and they have no names. The word comes from Persian, but in the Greek of the time it had broader senses: Persian priest, astrologer, interpreter of signs, sometimes sorcerer; the text does not say which country they come from.

A reader in the 80s could recall a recent event on hearing of “Magi from the East”: in AD 66 King Tiridates of Armenia had come to Rome with magi, and Pliny writes that he had initiated Nero in “magical suppers”, though Nero could not take the magi’s art from him (Natural History 30.16–17). Dio, writing later, has Tiridates tell Nero that he has come to worship him “as I do Mithras” (63.5.2). Another source close at hand: the story of Balaam, who sees a star come out of Jacob (Numbers 24:17).

As most readers take them, the Magi are the first people from outside the Jewish people to worship in the book (“they worshipped”, 2:11); at the end the disciples are sent to “all nations” (28:19). The Gospel therefore has a motive of its own for the Magi, without Mithras. Elsewhere the same Matthew speaks critically of the pagans (5:47; 6:7; 18:17) and tells the disciples earlier not to go “on the way of the Gentiles” (10:5): the Gospel’s attitude to them is not a simple one.

Third test: how the Magi look

How the Magi became three kings#

If the Magi had been put into the Gospel as priests of Mithras, they would have to be described so from the oldest texts on. The surviving texts show otherwise: the features are added one by one, in the centuries that follow, and they have separate histories: “three”, “kings” and the names of kings.

• c. 80–95Matthew: magoi from the east. No number, no king, no names.
• before 161Justin: the Magi come from Arabia; Damascus, he notes, lies in Arabia.
• c. 200Clement of Alexandria: “the Magi of the Persians”, the earliest mention found of the Magi as Persians.
• c. 207Tertullian: the East generally regarded magi as kings, he says; he also brings them from Arabia.
• c. 248In Against Celsus 1.60 Origen reads the three gifts as signs that Jesus is king, mortal and God, without stating how many Magi there were. The Biblical Archaeology Society reports that he also compares the Magi with Isaac’s three visitors in Genesis 26:26 (Abimelech, Ahuzzath and Phicol); this second point rests on its account alone.
• 3rd–4th c.The Priscilla catacomb, Rome: three figures in short tunics and pointed caps, one of the earliest representations of the Magi. The Yale ISM Review dates it, on stylistic grounds, to the late 3rd or early 4th century; the Biblical Archaeology Society to the mid-3rd.
• 5th c.Leo the Great explicitly describes, in sermons, three Magi and ties them to the three gifts; the tradition of three is older than his sermons.
• 5th–6th c.The oldest surviving names: Bithisarea, Melichior, Gathaspa (Excerpta Latina Barbari).

The costume in Christian painting, tunic, cloak and Phrygian cap, is the conventional dress of figures from the East in Roman art: Paris, Orpheus and captive barbarians wear it, and Mithras is often shown in it. The likeness is real, and it comes from the same conventions for depicting the East. The Magi are called Persian around 200, in the years when the stones of Mithras are at their most numerous.

What is left of the theory

Where the theory is right#

Three religious traditions have to be told apart. The old Persian religion, in which Mithra is a god and the magi are priests (Strabo lists among those the Persians worship the Sun, “whom they call Mithras”). The Parthian-Armenian royal religion. The Roman cult of Mithras, with initiation rites, grades and sanctuaries called mithraea. They are not the same people and the same institution. In 66 Rome saw a king with magi who, according to Dio, said he worshipped Nero as he did Mithras: the account therefore attests a link between magi and Mithras in the Parthian-Armenian royal tradition; it does not establish the same link for the Roman cult. Mithra was worshipped in the East before the Roman mysteries are securely attested: at Nemrut in Commagene a king of the 1st century BC placed on his funerary sanctuary a representation of “Apollo-Mithras-Helios-Hermes”. And at Caesarea Maritima in Judea there is a mithraeum whose date is disputed: from the end of the 1st century (the first proposal, from pottery) to the 3rd (coins), no earlier than the end of the 2nd century in the most recent analysis.

None of these traditions shows that the Magi were added for the followers of Mithras; the second shows only that a link between magi and Mithra existed in Parthian-Armenian kingship. The Commagene representation is of another kind than the Roman mysteries. The Caesarea mithraeum is a single one, in a provincial port. The word “magus” does not occur in any of the 561 included records; that lexical absence does not establish whether any follower had connections with magi.

Two Christian authors of the 2nd and 3rd centuries who describe the mysteries of Mithras, Justin (First Apology 66; Dialogue 70) and Tertullian (Prescription against Heretics 40), take them for devilish imitations of Christian rites; Origen reports Celsus’s account without that charge (Against Celsus 6.22). None speaks of the followers of Mithras as an audience to be won.