My view in 2003 (unchanged now) was stated at the end of my post linked
above: "I think it's reasonable on the evidence to accept that
Charlemagne was born in 747, possibly on 2 April or after 15 August but
with little chance now of recovering any certainty about the date."
Einhard, who wrote Charlemagne's biography within a decade or so after
the emperor's death, could not find out where or exactly when he had
been born, so it is hardly likely that we can definitely work it out
1200 years later.
The following may repeat comments I have posted before:
The date 2 April is less reliable than has been credited by some
historians, nowadays usually following Matthias Becher's study of the
question in 1992. He preferred 748 over 747 on the basis that the birth
was reported after the departure of the child's paternal uncle from
Francia that evidently took place after mid-August in 747. However,
since the annals placing the events in this order were not compiled all
at once, it may be that the birth after the departure was entered
subsequently into a blank space when it had not been recorded at first.
The entry gives Charlemagne the title 'king', indicating that it was
written after 754 and probably after he had succeeded his father in 768,
but in any case before he became emperor in December 800. The entry for
Carloman's departure would have been old news by that time.
The calendar giving 2 April as the birthdate, with emperor as
Charlemagne's title, was written ca 840 at Prüm or possibly Metz. It is
in an imperfect copy of a lost original from Lorsch that was dated 789.
The birthdate may have been interpolated into this Lorsch exemplar
between December 800 and January 814, going by the imperial title. The
birthdate was not included in another copy made at Lorsch ca 810.
Arno Borst in 1996 considered that the detail might have been a verbal
tradition written down only in this single instance, while further
noting that the calendar entry under 2 April was presumably a direct
imitation of the official birthdays of consecrated emperors, 'natales
Caesarum', after the famous Roman chronography of 354 (that was copied
at least twice in the 9th century). These birthdays had been celebrated
in the ancient empire, and the selection of 'good' Caesars to be
remembered by posterity carried dynastic significance. An auspicious
birthdate was perhaps attributed to Charlemagne by analogy, chosen not
from factual knowledge but simply because the scribe knew that his birth
was recorded under 747 and checked that Easter had fallen on 2 April in
that year.
Obviously the silence of contemporaries about the emperor's birth
occurring on the most joyful day of the religious calendar makes it
somewhat implausible that he was really born on 2 April 747, yet that
this propitious coincidence was somehow forgotten in his own family and
at court only to be piously recalled later by a monk elsewhere.
Nonetheless, 2 April 748 was accepted as the birthdate and year by
Becher since he concluded that Charlemagne had been born after his uncle
Carloman departed for Rome. He proposed that the event was recorded
under 747 using Easter style, by which 748 did not start until 21 April.
This ingenuity relies on some flimsy pieces of evidence, the strongest
of which is that while almost all contemporary records place
Charlemagne's death on 28 January under 814, not 813 as would result
from using Easter style, the annals compiled at Prüm early in the 10th
century are an exception. But this is far from conclusive proof that the
same style was applied there ca 840, much less in 800/14 at Lorsch.
Peter Stewart