Fun with Merovingian Dates

I meant to post about this a while ago, after I received an email from Professor Apocalypse, Richard Landes. He noted a discrepancy between how I had dated a text in a manuscript in London (British Library, Cotton A ii) and how Wilhelm Levison had done it. The manuscript is a compilation of odds and ends put together by Sir Robert Cotton (d. 1631) which includes some leaves in the middle from the eighth century. Elias Lowe, who I mentioned in my last post, thought that it was Italian, and noted ‘mam’ as a typical Veronese abbreviation for ‘misericordiam’ amongst other features including a tall ‘e’ which indicate Beneventan influence. The text itself contains this crucial passage (translated loosely by me):

After the Passion [of Christ] there are 532 years by the cycle of Victorius. The completion of this year completes the 179th year [after that]. There are in total, from the beginning of the world up to the present year, 5928 and there remain 72 years of 6000 years. Subtracting 40, 32 remain.

Here are a cluster of things you need to know before we go any further:

1)      In the Merovingian period, few people used AD dates. Bede (d. 725) helped to popularise the use of the dates in historical works, but he was building on the widespread use of the Easter tables of Dionysius Exiguus in Ireland and areas of Northern Frankia. Both myself and Immo Warntjes at Queens Belfast are convinced that St Willibrord, an English saint with Irish training based at Echternach in modern-day Luxembourg, was crucial for promoting this on the continent.

2)      In Gaul, Germany and Northumbria, there were communities using the Easter table of Victorius of Aquitaine well into the eighth century. This table did not label the years according to AD dates, but rather according to the ‘Year of the Passion’ – i.e. they start in our AD28, when Victorius reckoned Christ was crucified. Victorius’s table lasted 532 years and started again, so someone referring to the year in AD 642 would say the year was ‘the 83rd year of Victorius’, not AP 615.

3)      Most people dated things according to regnal dates anyway. You might write that something happened in the ‘tenth year of King Sigibert’ (Fredegar, chapter 88), not that it happened in AD 642 or ‘the 83rd year of Victorius’.

4)      I suspect that the situation in Northern Italy was pretty much the same as in Merovingian Gaul. Not everyone agrees with me. I will write about this later in the Summer when I give my paper on it in Galway.

Anyway, that little text from the manuscript was quite confusing, wasn’t it? What did it mean? Well. The 179th year of the Victorian table is the equivalent of AD 738. The calculation that the world is 5928 years old is, according to the Merovingian version of the tradition popularised by St Jerome, the equivalent of AD 727. Some historians might wonder if that had been a slip, and the age of the world should be 5938, but since that doesn’t equate with the 179th year of the table of Victorius, this seems unlikely. (Later, some people would have tidied up the numbers, so 5938 would equal 738… but that’s definitely later). The next calculation is consistent with 5928, and that does leave 72 years until the 6000th year of the world, and forty years later that would leave 32 to go. The final statement would suggest that the year is 767.

Wilhelm Levison gives a rather different interpretation of this passage (1). Starting with the initial 179th year being equal to 738, he assumed that 5928 was a mistake for 5939, despite the consistency between the remaining numbers and the low likelihood of a mistake at the end between xxviii and xxxviiii. Then, noting the subtraction of 40 to give the final year, he calculated that this must make the year 778, because 738 + 40 = 778. There were, of course, then only 21 years to go to the 6000th year of the world, not 32, so Levison by implication condemned the accuracy of the note and the scribes ability to count.

But what if this were two statements, not one, copied by a relatively careless scribe? Scribes copying these kinds of numbers often noted down various things that they discovered. A single collection of computistical material, used for calculating Easter, might contain a number of indications of when the present year was, depending on which sources were being used. I have written about one such compilation in which the year is variously given as 743 and 751 (2). What we would then have with our note in the London manuscript is a record of two chronological notes, one jotted down from 738, and one jotted down in 767 based on a calculation initially made in 727. Later, perhaps not until 767, these were made into a single text.

This is all well and good, but what does it all mean? I think it indicates two things which are important to our understanding of the history of the period:

1)      The Victorian Easter table, not the Dionysiac one, was an important tool which people used to calculate the date well into the middle of the eighth century.

2)      People were also interested in how many years it was until the Year 6000. This might be because of a tradition that the world might end that year, and that is why Richard Landes was asking me. I think he might be right in this instance. But for the full context, you’ll have to wait for my book on the Apocalypse in the Early Middle Ages.



(1)   England and the Continent in the Eighth Century (Oxford, 1946), p. 304.

(2)   ‘Computus after the Paschal Controversy of 740’, in D. Ó Cróinín & I. Warntjes (ed.), The Easter Controversy of Late Antiquity and the Early Middle Ages (Turnhout, 2011).