This surprises us today but it followed established precedent and should have been what was expected at that time. What must have been surprising, actually, was the announcement (whenever it came) that Rome's Emperor had split the kingdom and made Galilee independent.
Augustus Caesar was a genius, a political progressive and a natural innovator who left his creative and re-organizational mark on virtually every aspect of Roman government, culture and life, including overseas policy. The provincial government was reorganized, taxation was reformed, the army was permanentized and creatively financed, colonization and road building was expanded, and client kings were aggressively courted and groomed, their potential heirs being shipped into Italy for formal education in the ways of Imperial control. In settling dynastic disputes overseas, Augustus was basically ad hoc, showing no fear of imposing whatever situation seemed most workable from his vantage point, whether precedented or not.
In contrast to the Emperor, while Herod the Great was progressive in foreign relations and ambitious about economic development, the King of the Jews had remained conservative domestically, where he nurtured his constituencies among very traditional people. Politically, Herod made it quite clear for decades that he intended to establish his own dynasty in some form or another, following Hellenistic and Hasmonean precedent, a thoroughly conservative ambition.
It should stand without question that Herod intended his kingdom to go on with his descendants in power because it was not until late in the game that he became so famously volatile in favoring or disfavoring his various sons as chosen heir, and disowning or executing the most recent offender/s. These sins were anomalous from the larger continuity. As late as 6 BC, Herod was sharing rule with family members in subordinate positions, his eldest son Antipater and his brother Pheroras. Further, each late dynastic crisis was immediately patched with a new heir, and there were both early and later versions of Herod's will inclusive enough to name subordinate heirs, while the revision of 12 BC called for joint rule among three, an anomaly towards the other extreme, which was probably misguided. In all this, by and large, the family succession remained an unchanging assumption, with absolutely no mention or the faintest hint of a notion about dividing the kingdom politically.
The radical adjustment in 4 BC was entirely due to Augustus. The circumstances which compelled him to it probably involved the increased difficulty of governing greater Judea after such chaos and rebellion had erupted, and certainly had to do with the increasingly bitter contentions among the royal family as they waited for Caesar's decision. One also might speculate that Augustus also had one eye looking forward to claiming Judea directly, but the bottom line was probably that Antipas and Archelaus were fighting so venomously in Rome. How could the Emperor deliberately confirm one as subordinate governor under his rival? But whichever aspects of the situation were most responsible for inspiring Augustus decision, the critical point to observe is that Caesar innovated this solution autocratically, and that it was unprecedented.
The narrative of Josephus offers two details which indicate this with particular clarity.
The first point he puts as a thought into the head of P.Q.Varus, Governor of Syria during the conflict that year. Undoubtedly following the account left by Nicolas of Damascus at this point, Josephus has to explain how and why the youngest Herodian prince, Philip, winds up so suddenly in Italy for the judgment, late in sailing season, when Philip had been left in Judea by the family so many months earlier, before the fighting broke out. Whatever caused Philip's travel in fact, Josephus puts it down to the advice of his new friend, Governor Varus, who supposedly "saw a partition coming" (Loeb, AJ 17.303, Cf. BJ 2.83).
Now, while we obviously don't know what Varus may have begun to foresee, the more helpful perspective is that Josephus himself has shared this conception with us, and it expresses the author's account of those days somewhat directly. In using Varus' impressive and unique foresight to get Philip to Rome, Josephus takes no pains to defend or explain the content of that contrived vision, the not-yet-in-effect and the uncommonly-foreseeable nature of the "coming" division. Rather, the narrative completely assumes what ought to be evident by now, that in fact the kingdom had not yet been divided at that time. Therefore, since Josephus' narration shows that Herod's will had not effected such a division, even tentatively, it seems impossible to understand how Josephus' earlier description of Herod's will could be taken as evidence that Herod had been the one to stipulate such a division.
The second and more practical set of details involves finance. There is a point at which it becomes indisputably clear that Augustus' problem was not merely something political, like the assigning of jurisdictions to a proconsul or any other official. No, the real world implications of Josephus' narrative swing dramatically at the point when Augustus transfers control of the real estate, by permanently redistributing the direct receipt of all property-based revenues. Again, note that Josephus introduces this as the personal decision of Augustus, which of course it could only have been, but the key point is to recognize that these revenues must have been assigned differently before Augustus' surprising solution. In other words, the most dramatic change that the Emperor caused was not for Archelaus to lose a mere title or prestige, but for Archelaus to lose control of the revenues which accrued directly from half of his previous territory, which meant that Archelaus lost control of those territories. This, in turn, implies that Archelaus had possessed them all previously, which obviously requires that Antipas' original position in Galilee, as inherited, must have been a merely subordinate tetrarchy.
From first analysis, it should have been difficult to think the Great King had proposed an independent Galilee for the sixteen year old Antipas, even during his swelling psychopathy in those final days. For all of these reasons, Antipas' original position must be understood as a subordinate tetrarchy, as indeed the Galilean tetrarchy had always been previously, and certainly as it had been, both exclusively and explicitly, in Josephus' Antiquities, prior to King Herod's death.
The Augustan settlement therefore must have come as a surprise to everyone in Judea and Galilee, whenever it was finally announced, late in 4 BC or perhaps early in 3 BC. In the final experience of those people, it may not have been Archelaus' demotion to "ethnarch" that was seen as the most humbling consequence of his earliest mistakes as a ruler, but it was probably Archelaus' loss of Galilee - the shocking, near-unthinkable dissolution of the Kingdom, no less - by which later posterity would come to associate most strongly with its cultural memory of the Herodian princes' return from their voyage to Rome.
Briefly, now. Why does any of this matter?
If Matthew's Jewish readership in the mid to late first century brought this contextual understanding to its reading of the background referenced in verse 2:22, then the literary effect of that verse becomes much more significant. Recognizing that Galilee's independence from Judea was neither immediate nor immediately known to the general public, the text now appears to glorify God in unstated retrospect, for his divine foreknowledge, as it therefore also lifts up Joseph's bravery in following God's strange advice. In the world of the story, with chronological precision, God is telling Joseph to choose Galilee at a time before Galilee became independent of the infamously horrifying Archelaus, whom Joseph purportedly worked to avoid.
The implications for reader knowledge will be discussed at a future time. After reconstructing history from Josephus' narrative, we can further reconstruct some details about what the lived experience must have been like, for the common people. Finally, it is from that second reconstruction that we may observe some connection between which experiences were most likely to have lingered in the cultural memory, and thus informed Matthew's readership. At least, if Josephus' Italian readership in the late first century was aware enough to appreciate this distinction without too much authorial assistance, then how much more easily might a Judean readership have remembered it, possibly in the mid-first century?
Although it may be anticipated that this work will be able to find (or easily be tempted to "produce") a substantial connection between the textual phrasing with and whatever posterity is supposed to have developed (perhaps, skeptically, however much this literary reading needs the text to evoke?!), it could, nevertheless, be profoundly interesting to see how plausibly such a case can be made, and how likely it could seem - in the final analysis - that Matthew's original readers may have actually known (or rather, remembered!) this particular context, which, I very firmly believe, Matthew 2:22 was originally composed to rely upon.
In the end, it may not be the reconstructed reader memory that supports said reading of the text. It may be that a compelling reading of the text can transform into equally compelling evidence for particular reader knowledge, at least, that assumed by the author. Maybe! ; - )