This is Part 1 of a critical examination of the MMEL hypothesis of the Star of Bethlehem. Go to the index here.
In order to make the theory presented in the Star of Bethlehem documentary work, a major premise concerns the date when Herod the Great, a king of Judea, died. This is because the Gospel of Matthew says that Jesus was borne during his reign. Moreover, the Magi are said to come to King Herod, and it is only after Herod’s death that the Holy Family returned from Egypt to Palestine. So the events with the Star have to take place before Herod’s death in order to be consistent with Matthew’s account.
Now, there is a significant problem that this creates for the MMEL hypothesis. The Jupiter-Venus conjunctions take places in 3 & 2 BCE, but most historians place the death of Herod before the Passover in 4 BCE. This provides one bit of impetus for the attempt to change the order of historical events. For this, we have to look into what evidence do historians rely on to get the 4 or 5 BCE date for Herod’s death, and what is the evidence that the MMEL hypothesis uses to argue that instead Herod died in 1 BCE.
When it comes to the life, deeds, and death of Herod, the major source we have is the account of the Jewish historian Josephus in his Antiquities of the Jews (JA from hereon in), along with some things in his other important work, the Jewish War (JW). Writing in the 90s CE, Josephus was writing a history of the Jews from Genesis up to the time of the major revolt against Rome in the 60s CE. The Bible is an obvious source Josephus uses when it comes to the Patriarchs, but for Herod he had access to some good sources. The two most noteworthy are the Histories of Nicholaus of Damascus and Herod the Great’s own memoirs (JA 16.183-7, 15.174, 14.9). Nicholaus was a close friend of Herod’s, and he survived Herod so he knew about his death rather well. Josephus had other sources as well, so when it comes to putting a date on the most important aspects of his life he was in a good position. While Herod’s memories won’t have much on his death and funeral, Herod certainly knew about the events of his life, so we have primary sources in the hands of Josephus when he composed his narrative.
It should be noted that Josephus is far from a perfect historian. We know he makes mistakes, and his use of calendars has caused headaches. However, we can confirm events he speaks of with outside witnesses and artifacts, This will be brought up for the best reconstruction for Herod’s beginning as king of the Jews and his death.
Now, the basic way we know when Herod died is from the start of his reign and how many years he reigned. Josephus goes through each year he ruled, and he gives us the start date. In fact, he gives us two starting dates: when he was declared king of the Jews by Rome and when he took the thrown after defeating the Parthian Empire that controlled the region. According to Josephus, Herod ruled for 37 years after he was declared king by Rome (JW 1.665, JA 17.191), and 34 years after he was properly crowned. Josephus tells us very particularly that Herod was made king by Rome in 40 BCE by citing the consuls of that year (JA 14.389), and the consular list is something independently established by classicist and is not in dispute. The beginning of Herod’s rule also given by Josephus using the consuls again in 37 BCE (JA 14.487). The date of Rome made Herod king is also independently given by the historian Appian who was writing about the Roman civil war. In his Civil War 5.75, he talks about various rulers being established by Rome, including Herod. The details, such as the death of Marc Anthony’s wife (Fulvia) and the start of Parthia’s siding with Cassius and Brutus, place the events in 40 BCE, lining up with Josephus’ record.
From either of those years, you do the computation and you get Herod dying in or around 4 BCE. In addition, Josephus mentions a lunar eclipse taking place not long before his death (JA 17.167); we may be worrisome that this was used for dramatic effect, but the eclipses in 5 and 4 BCE can be used to also confirm that this dating fits what Josephus said about when Herod died.
Another indication that Josephus’ version of events is chronologically on-track is his mention of an event in Herod’s seventeenth reigning year when Augustus made a state visit to him (JA 15.368-370, cf. 15.354, 380). This would, on the normal chronology, be 20 BCE, and again we have independent verification. Augustus himself notes the occasion (Res Gestae 11), and is further confirmed by the historian Cassius Dio (Roman History 54.7.4-6, 54.9.3).
We have still another correlation to consider from Josephus’ account. The Jewish historian says that Herod died before he reached the age of 70 (JA 17.148; JW 1.647), so if we knew something about his birth year we can limit when Herod would have died. Josephus tells us that Herod was 25 when he was appointed governor of Galilee in 47 BCE (JA 14.158; note that there is likely a textual corruption as it says Herod was 15) just after Julius Caesar had temporarily stayed in Syria, so he would probably have been born in about 73/74 BCE, and thus he died before the year 3 BCE. Again, the data fits with what is expected from the 5/4 BCE death year of Herod rather than in 1 BCE.
Josephus also provides evidence in the other direction based on the sons of Herod. After Herod died, his kingdom was divided up among his four sons. One of his sons, Archelaus, ruled into his tenth year (JA 17.324-4) which ended 37 years after the famous Battle of Actium in 31 BCE (JA 18.26). As such, Archelaus started to rule in 4 or 5 BCE and ended in 6/7 CE, an end date confirmed by Cassius Dio (Roman History 55.27.6); since sons tend to rule after their father dies (and Josephus discusses the process of how the kingdom was divided up), Herod the Great must have died in 5 or 4 BCE. We see the same with the other sons of Herod. Philip died in the 20th year of Tiberius (33/34 CE) after ruling for 37 years (JA 18:106), which would again place the date of Herod’s death to 5/4 BCE. And if we look to Antipas, he lost his reign in 38/39 AD and his coinage only goes to the 43rd year of his rule, at least so far as been uncovered; again, the simple back-calculation will give you 4 BCE as when Antipas began to rule, and hence Herod must have died in 5/4 BCE. So we have several lines of evidence from several angles all pointing to the same time for when Herod died.
Now, all of this should give us confidence that Josephus is not making an error any greater than a year in his chronology, let alone three in order to get a death year of 1 BCE (and 1 CE has been argued by some as well, using the same reasoning). However, there is one principle argument that is used to prefer a different chronology. As noted, Josephus talks of a lunar eclipse before Herod died, and the most commonly cited one is from March of 4 BCE, which was a partial rather than full one. The main argument against this is that there was not enough time between this eclipse and the Passover for all the events mentioned by Josephus to have transpired. This was argued by Ernest Martin, who took what he considered the shortest plausible lengths of time for events to happen, add them up, and see that they cannot fit.
Before I get into the details, let us consider a point of order. Suppose that indeed the events between the eclipse and Passover could not have happened as described, meaning Josephus is mistaken. the question then is this: should we consider it a mistake when or what eclipse Josephus was referring to (or even placing it there for dramatic purposes as was common in antiquity), or should we change the chronology by several years? Considering how good Josephus’ chronology has been up to this point, it seems it is the greater sin to fudge his account by three years rather than one detail.
However, Martin’s argument has two ways out: one is a major mistake in his reading of events, and another is which eclipse to consider. One of the points Martin makes is that the funeral procession that traveled from Jerusalem to Herod’s burial site at the Herodium only traveled 8 stadia (about a mile) a day for a distance of 200 stadia. This would then take around 25 days, absorbing almost all the time between the lunar eclipse (March 13) and the beginning of Passover (April 11). However, Josephus does not say the procession is going 8 stadia a day, but simply the procession went 8 stadia (JA 17.199), and after that it isn’t said. A likely proposal is that after the procession in the vicinity of Jerusalem was over, the body was then taking to the burial place soon after, perhaps the same day (Jewish custom required prompt burial after funerary rights were observed). When you take away the artificial obstacle Martin created for the standard chronology, things have a greater potential to fit.
But even ignoring this, we have another, more prominent eclipse of the moon to consider from September 15 in 5 BCE. Even Martin’s strange chronology can fit into the space between this date and the Passover. In fact, the late 5 BCE eclipse is a better fit to Josephus’ description, giving it some preference. One of the major experts of the Herodian dynasty, Nikos Kokkinos, prefers the 5 BCE year as that of Herod’s death (The Herodian Dynasty, pp. 372-3), and it does ensure that there is not too much of a time compression of events happening between the eclipse and the Passover.
In fact, another source that has been used to date the death of Herod is independent of the record in Josephus, what is called the Megillat Ta’anit or Scroll of Fasting. While most fasts are given a reason for why they are special, there are two dates given which are not explained, 7 Kislev and 2 Shevat; the latter date is considered by scholars to in fact be the day after Herod died in the Jewish calendar (and some also believing the same for 6 Kislev, see Vered Noam, “Megilatt Taanit—The Scroll of Fasting” in Safrai, Safrai, Schwartz, Tomson, The Literature of the Sages, pp. 339-62, esp. p. 370; to compute between the Gregorian or Julian calendar and the Jewish calendar, go here). On that line of evidence, both of these dates come after the September 15 eclipse in 5 BCE, but before the January eclipse in 1 BCE as the MMEL hypothesis believes. As such, the Scrolldates fit the 5 BCE eclipse best (even better than the March of 4 BCE eclipse), and it provides yet another line of evidence against the revised chronology.
Now, we see that all the evidence provided by Josephus fits best when Herod died in 5/4 BCE, and other sources independently fit this time frame as well, while the arguments of Martin et al. cannot stand on their own (and even if so, we would consider something in Josephus’ record to be slightly inaccurate rather than completely change his timeline). However, there is one other line of evidence to consider. Most all the Apostolic Fathers in the Catholic tradition gave Jesus a birth year in the area of 3/2 BCE as in the MMEL hypothesis, thus putting Herod’s death later than 4 BCE. Did they have some other line of evidence? Unfortunately, these sources do not say they know this based on some other record from the time of Jesus’ birth. In fact, we can see how they likely derived that date. According to Luke 3, Jesus was “about 30″ when he came to see John the Baptist whose ministry began in the 15th year of Tiberius (28/29 CE). If one believes Jesus came to John at the beginning of his ministry, then you can easy get a date of 3/2 BCE from this. As such, the Catholic tradition is not independent verification of when Herod died, but it is derivative of the Gospel material (which is itself in contradiction on important matters). It is also worth mentioning that while it is true Luke does talk about a king Herod when discussing the birth of John and Jesus, there is reason to believe he may be talking about Archelaus (who was also called “king” and “Herod”). But for the historian, the key thing is we cannot use late sources that are not dependent on primary sources for their calculations to deny something we find in a source (Josephus) who is using multiple primary sources (the writings of Nicolaus and Herod himself).
At this point, it should be clear that the evidence we have is strongly in favor of the 5/4 BCE date of Herod’s death, while there is nothing but several lines of special pleading for a different date. These are not the only arguments to consider, and there has been plenty written on the matter in biblical studies journals. The 1968 paper by Timothy Barnes is still a strong and worthy argument (“The Date of Herod’s Death”, Journal of Theological Studies 19), and modern attempts to go against it are require that most other historical sources from antiquity to be wrong (such as details about the Augustan court).
However, there is one last-ditch effort to get past this problem, based on textual witnesses to Josephus’ works. That will be considered in Part 2 of this critique.
