Blog Menu

Did Epiphany really happen?

Three_wise_men_6th_Century_Roman_MosaicThe Feast of the Epiphany in the church’s liturgical calendar is based on the events of Matt 2.1–12, the visit of the ‘wise men’ from the East to the infant Jesus. There are plenty of things about the story which might make us instinctively treat it as just another part of the constellation of Christmas traditions, which does not have very much connection with reality.

The first is the sparseness of the story. As with other parts of the gospels, the details are given to us in bare outline compared with what we are used to in modern literature. We are told little of the historical reality that might interest us, and the temptation is to fill in details for ourselves. This leads to the second issue—the development of sometimes quite elaborate traditions which do the work of filling in for us. So these ‘magoi’ (which gives us our word ‘magic’) became ‘three’ (because of the number of their gifts), then ‘wise men’ and then ‘kings’ (probably under the influence of Ps 72.10. By the time of this Roman mosaic from the church in Ravenna built in 547, they have even acquired names. Christopher Howse comments:

[T]hink how deeply these three men have entered our imagination as part of the Christmas story. “A cold coming they had of it at this time of the year, just the worst time of the year to take a journey, and specially a long journey, in. The ways deep, the weather sharp, the days short, the sun farthest off, in solstitio brumali, the very dead of winter.”

Those words, in a tremendous sermon by Lancelot Andrewes that King James I heard on Christmas Day 1622, were brilliantly stolen by TS Eliot and incorporated into his poem The Journey of the Magi. And we can see it all: the camels’ breath steaming in the night air as the kings, in their gorgeous robes of silk and cloth-of-gold and clutching their precious gifts, kneel to adore the baby in the manger.

Yet, that is not entirely what the Gospel says…


In response to this, critical scholarship has moved in the other direction, and by and large has pulled apart Matthew’s story and confidently decided it that none of it actually happened. Instead, it was constructed by Matthew out of a series of OT texts in order to tell us the real significance of Jesus. So Marcus Borg and Dominic Crossan, in The First Christmas: what the gospels really teach about Jesus’ birth, come to this conclusion:

In our judgement, there was no special star, no wise men and no plot by Herod to kill Jesus. So is the story factually true? No. But as a parable, is it true? For us as Christians, the answer is a robust affirmative. Is Jesus light shining in the darkness? Yes. Do the Herods of this world seek to extinguish the light? Yes, Does Jesus still shine in the darkness? Yes. (p 184).

The approach presents problems of its own. For one, the stories are not presented as parables, but in continuity with the events Matthew relates in Jesus’ life later in the gospel. For another, if God in Jesus did not outwit Herod, on what grounds might we think he can outwit ‘the Herods of this world’? More fundamentally, Matthew and his first readers appeared to believe that the claims about Jesus were ‘parabolically true’ because these things actually happened. If none of them did, what grounds do we now have? Even if the events we read about are heavily interpreted, there is an irreducible facticity in testimony; if this has gone, we ought to question the value of the testimony itself.


A good working example of this approach is found in Paul Davidson’s blog. Davidson is a professional translator, rather than a biblical studies academic, but he offers a good outline of what critical scholarship has to say about Matthew’s nativity.

His basic assumption is that Matthew is a ‘multi-layered’ document—Matthew is writing from the basis of other, differing sources. He takes over large parts of Mark’s gospel, as does Luke, and Matthew and Luke never agree in contradiction to Mark, a key piece of the argument of ‘Marcan priority’, that Mark was earlier than either of the other two. Whether or not you believe in the existence of the so-called Q, another early written source (and with Mark Goodacre, I don’t), Matthew is clearly dealing with some pre-existing material, oral or written. It is striking, for example, that Joseph is a central character in Matthew’s account before and after the story of the magi, and is the key actor in contrast to Luke’s nativity, where the women are central. Yet in this section (Matt 2.1–12) the focus is on ‘the child’ or ‘the child and his mother Mary’ (Matt 2.9, 2.11; see also Matt 2.14, 20 and 21). Some scholars therefore argue that this story comes from a different source, and so might be unhistorical.

This is where we need to start being critical of criticism. Handling texts in this way requires the making of some bold assumptions, not least that of author invariants. If a change of style indicates a change of source, then this can only be seen if the writer is absolutely consistent in his (or her) own writing, and fails to make the source material his or her own. In other words, we (at 20 centuries distant) need to be a lot smarter than the writer him- or herself. Even a basic appreciation of writing suggests that authors are just not that consistent.

Davidson goes on in his exploration to explain the story of the star in terms of OT source texts.

The basis for the star and the magi comes from Numbers 22–24, a story in which Balaam, a soothsayer from the east (and a magus in Jewish tradition) foretells the coming of a great ruler “out of Jacob”. Significantly, the Greek version of this passage has messianic overtones, as it replaces “sceptre” in 24:17 with “man.”

He is quite right to identify the connections here; any good commentary will point out these allusions, and it would be surprising if Matthew, writing what most would regard as a ‘Jewish’ gospel, was not aware of this. But if he is using these texts as a ‘source’, he is not doing a very good job. The star points to Jesus, but Jesus is not described as a ‘star’, and no gospels make use of this as a title. In fact, this is the only place where the word ‘star’ occurs in the gospel. (It does occur as a title in Rev 22.16, and possibly in 2 Peter 1.19, but neither make any connection with this passage.)


Next, Davidson looks at the citation in Matt 2.5–6, which for many critical scholars provides the rationale for a passage explaining that Jesus was born in Bethlehem when he is otherwise universally known as ‘Jesus of Nazareth’ (19 times in all four gospels and Acts). But, as Davidson points out, Matthew has to work hard to get these texts to help him. For one, he has to bolt together two texts which are otherwise completely unconnected, from Micah 5.2 and 2 Sam 5.2. Secondly, he has to change the text of Micah 5.2 so that:

  • Bethlehem, the ‘least’ of the cities of Judah, now becomes ‘by no means the least’;
  • the well-known epithet ‘Ephrathah’ becomes ‘Judah’ to make the geography clear; and
  • the ‘clans’ becomes ‘clan leader’ i.e. ‘ruler’ to make the text relevant.

Moreover, Matthew is making use of a text which was not known as ‘messianic’; in the first century, the idea that messiah had to come from Bethlehem as a son of David was known but not very widespread.

All this is rather bad news for those who would argue that Jesus’ birth was carefully planned to be a literal fulfilment of OT prophecy. But it is equally bad news for those who argue that Matthew made the story up to fit such texts, and for exactly the same reason. Of course, Matthew is working in a context where midrashic reading of texts means that they are a good deal more flexible than we would consider them. But he is needing to make maximum use of this flexibility, and the logical conclusion of this would be that he was constrained by the other sources he is using—by the account he has of what actually happened.


St Denis 2012 - 26 - Version 2Davidson now turns to consider the magi and the star. He notes a certain coherence up to the point where the magi arrive in Jerusalem.

So far, the story makes logical sense despite its theological problems (e.g. the fact that it encourages people to believe in the “deceptive science of astrology”, as Strauss noted). The star is just that: a star.

Then everything changes. The star is transformed into an atmospheric light that guides the magi right from Jerusalem to Bethlehem, where it hovers over a single house—the one where the child is. We are no longer dealing with a distant celestial body, but something else entirely, like a pixie or will-o’-the-wisp.

Here again critical assumptions need some critical reflection. Matthew’s inclusion of magi is theologically very problematic indeed. Simon Magus and Elymas (Acts 8.9, 13.8) hardly get a good press, not surprising in light of OT prohibitions on sorcery, magic and astrology. Western romanticism has embraced the Epiphany as a suggestive mystery, but earlier readings (like that of Irenaeus) saw the point as the humiliation of paganism; the giving of the gifts was an act of submission and capitulation to a greater power. For Matthew the Jew, they are an unlikely and risky feature to include, especially when Jesus is clear he has come to the ‘lost sheep of the house of Israel’ (Matt 10.6, 15.24).

There have been many attempts to explain the appearance of the star scientifically. The best contenders are a comet (for which there is no independence evidence), a supernova (observed by the Chinese in 4 BC) or the conjunction of Jupiter with Saturn in the constellation Pisces. I think the latter is the best candidate; Jupiter signified ‘leader’, Saturn denoted ‘the Westland’, and Pisces stood for ‘the end of the age’. So this conjunction would communicate to astrologers ‘A leader in the Westland [Palestine] in the end days.’ This highlights a key problem with Davidson’s criticism; the issue is not whether a star could in fact indicate a particular house in our, modern scientific terms. This is clearly impossible. The real issue is whether Matthew thought it could—or even whether Matthew thought the magi thought it could. As Dick France highlights in his NICNT commentary, this was actually a common understanding for which we have documentary evidence. And any naturalistic explanations miss Matthew’s central point: this was something miraculous provided by God. If you don’t think the miraculous is possible, you are bound to disbelieve Matthew’s story—but on the basis of your own assumptions, not on any criteria of historical reliability or the nature of Matthew’s text.

Davidson cites the 19th-century rationalist critic David Friedrich Strauss in his objection to the plausibility of Herod’s action:

With regard to Herod’s instructions to report back to him, Strauss notes that surely the magi would have seen through his plan at once. There were also less clumsy methods Herod might have used to find out where the child was; why did he not, for example, send companions along with the magi to Bethlehem?

In fact, we know from Josephus that Herod had a fondness for using secret spies. And in terms of the story, the magi are unaware of Herod’s motives; we are deploying our prior knowledge of the outcome to decide what we think Herod ought to have done, which is hardly a good basis for questioning Matthew’s credibility.


botticelli-c-1475-adoration-of-the-magiFinally, we come to the arrival of the magi at the home of the family. Interestingly, Matthew talks of their ‘house’ (Matt 2.11) which supports the idea that Jesus was not born in a stable—though from the age of children Herod has executed (less than two years) we should think of the magi arriving some time after the birth. No shepherds and magi together here!

Davidson again sees (with critical scholars) this event constructed from OT texts:

According to Brown, Goulder (2004), and others, the Old Testament provided the inspiration for the gifts of the magi. This passage is an implicit citation of Isaiah 60.3, 6 and Psalm 72.10, 15, which describe the bringing of gifts in homage to the king, God’s royal son.

But again, the problem here is that Matthew’s account just doesn’t fit very well. Given that these OT texts uniformly mention kings, not magi, if Matthew was constructing his account from these, why choose the embarrassing astrologers? And why three gifts rather than two? Where has the myrrh come from? Again, it is Irenaeus who first interprets the gifts as indicators of kingship, priesthood and sacrificial death respectively, but Matthew does not appear to do so. In the narrative, they are simply extravagant gifts fit for the true ‘king of the Jews’. Subsequent tradition has to do the work that Matthew has here failed to do, and make the story fit the prophecies rather better than Matthew has managed to.

Davidson closes his analysis of this section with a final observation from Strauss:

If the magi can receive divine guidance in dreams, why are they not told in a dream to avoid Jerusalem and go straight to Bethlehem in the first place? Many innocent lives would have been saved that way.

Clearly, God could have done a much better job of the whole business. But it rather appears as though Matthew felt unable to improve on what happened by fitting it either to the OT texts or his sense of what ought to have happened.

The modern reader might struggle with aspects of Matthew’s story. But it seems to me you can only dismiss it by making a large number of other, unwarranted assumptions.

(Reposted from Epiphany 2015)


You can listen to my sermon from Epiphany 2014 here, in which I explore historical, narrative and theological approach to reading this text in order to understand what we might learn from it.


Follow me on Twitter @psephizo


Much of my work is done on a freelance basis. If you have valued this post, would you consider donating £1.20 a month to support the production of this blog?

, , , , ,

8 Responses to Did Epiphany really happen?

  1. Phill January 6, 2016 at 10:37 am #

    Thanks for this Ian – very helpful.

    I think it takes far *more* faith to believe in critical scholarship than simply believing what the gospels actually say. As you point out several times, second-temple Judaism did not interpret texts in the way that Matthew and the evangelists did. There was simply no precedent for seeing Jesus as the kind of Messiah that he was, not to mention the fact that no-one believed in resurrection in the way that Jesus rose (Tom Wright does a good job of outlining this in The Resurrection of the Son of God).

    So in order to get round the fact that the events that Matthew outlines are unlikely, you have to invent an even more unlikely (i.e. wildly improbable) sequence of events. All this while affirming that Jesus is apparently ‘a light shining in the darkness’. It’s incredible.

  2. Paul January 6, 2016 at 10:15 pm #

    Sorry, there are just way too many logical problems and contradictions between Matthew and Luke to think the birth stories are historical.

    For one thing, how do you get around the problem of where Mary and Joseph are from? In Matthew, they live in Bethlehem and move to Nazareth to avoid Herod. In Luke, they are natives of Nazareth and go to Bethlehem for the most implausible census in the history of mankind that was not recorded by any other historian.

    How did the Jesus family avoid detection? They had angels at the birth, went to the temple and Jesus was publicly declared special by a prophet, and they had visits from foreign magistrates of some sort who gave them expensive gifts. Yet Herod’s people were unable to find him, nobody ratted on the Jesus family to save their own children from execution?

    And what did they do with the gifts of the magi? Certainly they were expensive gifts, the family would have been extremely wealthy with a haul of gold and expensive goods. Why was the family then a typical poor peasant family after that?

    Why didn’t Mary remember the Angelic visits later in life? There is no sign that Mary encouraged Jesus’ ministry. But if she really had visits from angels, would she not have been his biggest supporter?

    I could go on all day like this, but you get the drift.

  3. John Oliver January 6, 2016 at 11:44 pm #

    Is there any suggeston that Magi may include wise women, seers and similar?

  4. Peter Llewellyn January 7, 2016 at 6:29 am #

    As requested Ian:

    Dialogue from Ian’s Facebook post yesterday:

    PL: I am in no doubt that Matthew believed that what he wrote really happened; he no doubt faithfully transmitted the tradition he received. However I think traditions from a whole generation prior to the bulk of his Jesus materials would have come to him pretty well worked over, though not necessarily from the historicist perspective you ?Ian Paul offer. Your analysis of the text as it stands seems to me to be sound and persuasive, but your arguments do not simply transfer mutatis mutandis to the earlier transmission of the tradition. In short, the story as it came to Matthew would be pretty close to what we now have; but how it got to him is another matter.

    IP replied: Yes, ?Peter I would agree with this analysis. This implies that scepticism about any events behind this relies on all of
    a. a time lag between events and any recording of them
    b. the generation of story through (possibly oral) transmission
    c. a lack of interest in facticity in the gospel writers.

    And IP added: For me, research over the last 30 years and more on the dating of the gospels, the study of oral tradition, the importance of eyewitness testimony, and the practices of writing in first century culture all put big question marks against some or all of these three.

    PL replied: I largely agree with this – it is right to “put big question marks” against all three aspects. But I think we can also offer some responses to the question marks.

    When I first was asked to lecture on the Gospels (Australian College of Theology Bachelor degree studies) in 1982 there was thought to be a consensus in favour of the four document source hypothesis, a tendency to support the Markan hypothesis (Mark is simpler than the other Synoptics, therefore earlier, and thus closer to Jesus and therefore more reliable), and more consensus stuff from the Form and Redaction critics. All this pushed, in various ways, against the perception of the historical reliability of the Gospels. But then the Scandinavian school (Gerhardsson & co.) argued powerfully for the reliability of oral transmission, in the process also reducing the significance of any time lag. Studies in the Jewish context of the Gospel writers (which was my personal field of research at the time) also pushed in the same direction. Luke explicitly affirms an interest in “accuracy” (???????. Luke 1:3) but whether that amounted to facticity is another question (see Loveday Alexander’s excellent analysis of this issue in her The Preface to Luke’s Gospel: Literary convention and social context in Luke 1.1-4 and Acts 1.1, Cambridge: CUP, 1993). I agree with Gregory Sterling (Historiography and Self-Definition: Josephos, Luke-Acts and Apologetic Historiography, Leiden: E.J. Brill & Co., 1992) that Luke counts as a good, generally reliable Hellenistic apologetic historian, but I don’t think the same can be so confidently said of Matthew. The upshot is that Luke’s Jewish apologetic is clearer and more obviously on the surface than Matthew’s.

    I think the displacement of the Matthean community in the 80s CE, out of the Synagogue to form their own new culture, had two effects on his composition of Matthew: first, it made everything urgent – witness Matthew 18 where church discipline has become a key question; and second, everything was skewed away from the Judaism they knew. None of this counts directly either for or against the historicity of the Epiphany story, but indirectly I think we can discern a tendency in Matthew to diminish the Judaism of the Synagogue and to favour a more inclusive Judaism (i.e. like Matthew’s community) – and this fits the Epiphany story.

    Of course there are questions about every aspect of this brief outline of a reconstruction, but if it is correct then we can see a movement away from concern about facticity in a story such as the Epiphany story – which then takes a more constructed role in the Matthean agenda. In that case, a time lag of more than 80 years between the supposed events and the writing of Matthew may permit more variation than a purer oral transmission would; and it may also invite more creativity in framing the story.

    There are three contentious historical claims in the Matthew Epiphany story, as others have noted: the Magi, the star, and Herod. Ignoring the factual question of whether Herod was still alive when Jesus was born, I don’t think it is difficult to see some creativity in constructing the story in all three elements, so that Matthew receives a well-framed story which he doesn’t have to tweak very much to suit his apologetic purpose, viz. to establish a new Torah with Jesus as the prophet like Moses.

    I know all this is dreadfully shorthand and skips over a lot! – I hope it isn’t too incoherent. In saying all this I want to affirm the integrity and unity of Matthew, whether it is dependent on Mark or not, and to put the author in the same general category as the author of Luke-Acts – as having an apologetic purpose towards the survivors of Judaism, but of course in quite a different context.

  5. John Grayston January 7, 2016 at 9:06 am #

    Ian, thanks as ever.

    A small point, and I am not wanting for one moment to put magi and shepherds together in the house (which I believe to be right), stable or cave, but can we infer anything about the time of the visit of the magi from the age of the children killed by Herod? I know there is a common understanding here but for me there are too many unknowns. Herod only enquires about the time the star appeared – does this coincide with the birth? Is it prior to the birth? What about preparations for the journey and journey time, real or presumed by Herod? Would Herod, from what we know of him, have been too scrupulous about only killing those within the strict parameters determined by (the magi’s interpretation of) the star? How long, without the arrival of the magi and the instructions to go to Egypt, would Mary and Joseph have stayed in Bethlehem? If, as is often argued, Jesus was about 18 months by this time, that seems a long time to be away.from Nazareth. (I know I am conflating Matthew and Luke at this point and that Matthew seems to know nothing of Nazareth until the return from Egypt, but I think that is more logically explained by Matthew either having no knowledge of it or omitting it than by Luke inventing it, and that is therefore a reasonable question to ask.)

  6. Penelope Wallace January 7, 2016 at 9:17 am #

    Thank you, Ian, for your helpful analysis. As a non-scholar, my own difficulty with Matthew 2 has always been the problem of fittijng it together with Luke 2. It requires a huge struggle to make the two accounts of where Mary and Joseph lived before the birth, what they did after the birth, and why Jesus was brought up in Nazareth match, and if Jesus had spent months as an infant in Egypt (visiting Jerusalem, surely the most dangerous of places for him according to Matthew, before or afterwards) why does Luke not mention it?

  7. Dave Burke January 11, 2016 at 4:54 pm #

    I came here following Steve Walton’s link to your article on the Economist piece on the C of E, and very glad I stayed to read this article too. Helpful and useful piece of work – thanks!

Leave a Reply