As we approach the end of October, and more and more shops are trying to sell us sweets to hand out to strange children as part of some American end of October custom called “trick or treat”, children are deciding what grisly thing they will dress up as for their halloween party.
Halloween is the eve of All Saints day - celebrated since the ninth century, and much of the superstition surrounding the day developed from superstitions surrounding this date. But the date was not idly chosen. Like many Christian festivals, the date was chosen because of related pagan festivals at or about the same time.
Now which pagan festival was originally on the night before November 1st? Many people will tell you this is Samhain or the Celtic New Year - one of the Celtic fire festivals.
I disagree.
I don’t think the historical Samhain is the 1st November, which would be an illogical choice adapted only to suit the Roman system of counting of the months.
Before I go on, I make the point that November 1st may now be celebrated as the Celtic New Year, in much the same way that April 6th is celebrated as the start of the fiscal year. Not because April 6th is special, but because the start of the year has become disconnected with the reasons for it.
I make this point, because it makes no difference on what day we celebrate such things - but when people make spurious special claims for a certain night of the year as being somehow distinct from other nights, based on Celtic history - then it is worth noting that the night they celebrate is not the night the ancient Celts probably celebrated.
My argument is that prior to the spread of Roman culture throughout the Celtic world, and the Roman calendar with it, the more likely date for Samhain would have been the day that we now call November 5th (although 6th November is also a candidate).
There is good evidence that the Celts of northern Europe kept an accounting of the passing of time by use of standing stones and such like, so whilst we do not need to delve into too much fantasy over how good astronomers they were, there is clear evidence that they would have known the dates of key events such as the solstices and equinoxes. We also know that they held the Beltain and Samhain festivals as especially significant, but prior to the Roman accounting of time there would be no way that we could get an exact record of the date of these festivals except by the astronomical method and counting.
Now consider the calendars I have reproduced below. The Autumnal equinox on 21st September, and winter solistice on 21st December are two well known significant astronomical dates. Recent evidence suggests that the winter solstice in particular was celebrated at Stone Henge (not the summer solstice, as many suppose).
Whilst Stone Henge pre-existed the Celts in Britain, we know that the current arrangement of the stones was carried out by the pre-Roman Celts. Thus it is important information that these Celts held 21st December to be a key date.
In the calendar below I have marked 45 dates starting on the autumnal equinox in orange. I have marked in green the 45 days prior to the winter solstice. Notice that leaves one day unmarked between the two dates - November 5th.
This date, equidistant between equinox and solstice, is - I suggest - the date on which pre-Romanised Celts would celebrate their fire festival, and begin their year. This would be the historical date of Samhain.
| September |
October |
November |
December |
Su Mo Tu We Th Fr Sa
                1 2
 3 4 5 6 7 8 9
10 11 12 13 14 15 16
17 18 19 20 21 22 23
24 25 26 27 28 29 30
|
Su Mo Tu We Th Fr Sa
 1 2 3 4 5 6 7
 8 9 10 11 12 13 14
15 16 17 18 19 20 21
22 23 24 25 26 27 28
29 30 31
|
Su Mo Tu We Th Fr Sa
          1 2 3 4
 5 6 7 8 9 10 11
12 13 14 15 16 17 18
19 20 21 22 23 24 25
26 27 28 29 30
|
Su Mo Tu We Th Fr Sa
                1 2
 3 4 5 6 7 8 9
10 11 12 13 14 15 16
17 18 19 20 21 22 23
24 25 26 27 28 29 30
31
|
The Romans had their own festival on November 1st - the feast of Pomona, the goddess of the autumnal harvest and orchards. This festival may have been sufficiently similar to Samhain that the two became conjoined on 1st November in the Roman calendar (a calendar which, in any case, knew considerable drift over time, so that 1st November at this point might conceivably have coincided with the observed and derived date for Samhain).
Sir James Frazer notes that every European fire festival other than the Celtic fire festivals was linked to a significant astronomical date, and it is he who (in chapter 62 of the Golden Bough) suggests that the Celtic dates are more siginificant to the herdsman, although he credits another with the revelation.
It seems that Frazer was not aware of the significance of the date of November 5th though, and his survey of customs that were still extant puts most of them on the night of 31st October. That does not, however, mean that throughout history the date was always 31st October, but only that this has become the significant date in modern consciousness, and that - I suggest - is owing to the feast of Pomona, and the spread of Roman culture.
We note that this could also explain the date of Christmas being 25th December - this being also the date of the Roman Saturnalia feast, which again may have drifted from a solstice date on 21st December pnce the Romans introduced their calendar.
Someone may object that we have a Celtic calendar - the Coligny calendar from 1st Century Gaul. Their year was shorter than ours, and they used intercalary days and months on a regular basis to maintain a balance between the lunar and solar year.
Festival days drifted during the course of a calendrical cycle, and could vary as much as month earlier or later in relation to where they fell during the first year of the cycle.
What we don’t know is whether this calendar with its festival drift was representative of an earlier Celtic tradition, or something that arose only after the Roman influence caused these Gauls to start accounting time with a calendar.
If we suppose that the Coligny calendar is representative of original Celtic tradition then the dates of these festivals would drift. Further it would suggest that a full moon was indeed chosen to recognise the date of Samhain.
If this is so then my thesis that November 1st is not Samhain, but the nearest Roman festival to Samhain would remain, but my thesis that the correct night is November 5th would be wrong - it would rather be a full moon close to 5th November (perhaps the first full moon thereafter). We cannot know for sure - and I think that is a good thing.
Whilst the calendar is not itself in any respect Roman, it is certainly quite possible that the whole concept of a calendar (and thus intercalary months) was derived from the Roman practice, and that an older celtic tradition would not necessarily have placed the bonfire nights within specific months at all. I admit that beyond the recognition that Halloween does not seem to historically fit exactly with Samhain, all else is speculation based on whatever pieces of archeological evidence one wishes to cite in support of ones thesis.
However, if a pagan says to you (as one said to me, some years ago) “Halloween is Samhain, and that festival is one of ours”, politely point out to them that Halloween is not the real date of Samhain, and that All Saints day is very much a Church festival, and has been for nearly 1200 years.