When was the Viking first day of summer? Winter?

Question

I am writing a historical Viking novel taking place in the early 11th century. I have done extensive research about basically everything, including the Viking calendar. Several places mention that the Vikings only had two seasons, summer and winter, of 26 weeks each, with an occasional extra summer 'leap' week to accommodate that the year is more than 364 days long. At the first day of summer, everybody became one winter older, which is how the Vikings measured age.

What I haven't been able to find, however, is exactly when this first day of summer was. At this time, as far as I can tell, the Vikings still used lunar calendars, with an occasional leap month, so I doubt it was at the start of any given month. Nor does it seem to be at the vernal equinox.

I did find one source which states that the start of summer at April 14th - but the same source uses fixed rather than lunisolar months and fixes the start of the first summer month, Harpa, at April 14th, which wouldn't be the case with a lunisolar calendar.

My best guess, based on which months were considered summer months (though different sources disagree on this), is that the first week of summer starts on the fourth monday after the spring equinox (including the day of the equinox, if that is a Monday). The first day of winter then presumably starts on the fourth Monday after the autumnal equinox, not counting an equinox Monday (thus making the summer sometimes a week longer than winter).

Does anybody have a source that either confirms or disproves this? Or just a better informed guess than mine?

(Though it isn't important for the story, I like for such details to be accurate. I have even found what days of the year the moon was new; i.e. when a new month starts; by looking at the dates for historical solar eclipses.)

Answer 1

Every reference I've been able to find for Old Norse or Viking calendars ended up pointing me to information from the Old Icelandic calendar. That's not that weird, as much of our knowledge of pre-Christian Germanic culture in other areas also is typically inferred from Icelandic sources.

There's a good English-language paper^(PDF) on the subject from Svante Janson of Uppsala University I'll reference here.

Yes, it looked like they had only 2 seasons, sumar and vetur. Their years all had exactly 52 weeks, with an extra leap week added when necessary to keep the seasons from drifting from the solstices/equinoxes. This had the interesting mathematical side benefit of causing the first day of each season to be the same day of the week every year.

They still used 12 months (because they were using the same moon as everyone else), but since 364 isn't evenly divisible by 12, the extra 4 days were put a the end of sumar. It also seems likely they didn't really care about months much until they needed to start reconciling their calendars with the Christian ones in the 12th century, which is why we don't know a lot of their Old Icelandic month names (perhaps they didn't actually name them). Their dating system appears to have been more apt to use the week numbers.

After the 12th Century, pains were taken to keep this somewhat aligned with the varioius Christian calendars, which produced the relation shown in Janson's table below:

So it looks like from the 12th Century on the first day of Sumar would have been in their reckoning Thursday, Sumar week 1, somewhere in the 9-15th of April Julian, and somewhere in the 19th-25th of April Gregorian. Their first day of Vetur would have been Saturday of Vetur week 1, which would have been somewhere in the 11th-18th of October Julian, or 21st-28th October Gregorian. However, there's apparently some scholarly debate on if Vetur had always started on Saturday, or instead moved there from Monday when the aforementioned 12th Century calendar reforms happened.

Another interesting thing I noticed in there is that there's no proscribed order to these seasons. They alternate of course, but the Norse didn't number their years, so they had no concept of a specific day that a year started in. Ages of people and things were reckoned in Winters, but sometimes Summers (just to keep us on our toes, I guess).

Also, since neither season was favored as "first", and the two season's first weeks started on different days of the week, it appears there may not have been a concept of a "first day of the week". Of course we modern folk today argue over what that day is, so perhaps lack of consensus among the Old Norse on the matter should not seem so strange.

Answer 2

Summer (probably) started 10 weeks before midsummer

According to O.S. Reuter's Skylore of the North, in Iceland in 999, the start of summer was set to be 10 weeks before the Althing, which was held at midsummer. This was probably based on traditional usage.

Also, according to Reuter, Scandinavians used both solar and lunisolar calendars at the same time, with the latter used for festivals, which tended to be held at the full moon. The year's first sacrificial feast, Dis, was hence held at the full moon following the first new moon after midwinter. An 8-year cycle was followed of 13, 12, 13, 12, 12,13, 12, 12 months.

A solar calendar of 12 months of 30 days, plus a 5-day winter festival, seems to have been in use in Norway and Iceland, while a calendar of 13 months of 28 days was followed in Scandinavia, with a leap week every 5 or 6 years added to the midsummer month. Hence, all months started on Thursdays. This schedule was apparently also followed in Iceland at some point (they must have been very confused).

Thanks to everybody who spent time trying to find an answer! Your efforts sparked further efforts by myself and made me find Reuter's article, which likely is the best answer available, giving the dearth of contemporary sources.