Seasonal Reckonings

When do the seasons start? In England, the common opinion reflected in calendars and the media is that they start with the equinoxes and solstices. But this is neither spatially nor temporally invariant. The cross-quarter days, for example, celebrated in Irish Gaelic as Samhain (starting Winter), Imbolc, Beltane, and Lughnasadh, have also been thought of as delimiting the seasons. Wikipedia calls this cross-quarter system the “Traditional reckoning”, and the equinoctial and solstitial system the “Astronomical reckoning”. But it also adds a third way of reckoning the seasons, the more locationalised “Meteorological reckoning”, based on temperatures.

Using temperature seems like a more descriptive way of marking the seasons, and its results fall in the middle of the other two reckonings. The months of January, April, July, and October mark the midpoints of the Meteorological reckoning, but they’re roughly the ends of the Traditional seasons and the beginnings of the Astronomical seasons. Anyway, none of the three systems are really tied to the months themselves: months have no correlation with either the sun or the earth’s climate. So, how do we find out when the seasons can be said to start in the Meteorological reckoning, and what does that tell us?

The MET Office maintains a list of the monthly mean temperatures for Central England dating back to 1659. The average for the twelve months, I’ve ascertained, for 1659 to 2006 is: 3.22, 3.85, 5.3, 7.9, 11.21, 14.32, 15.96, 15.63, 13.32, 9.68, 6.03, and 4.08°C. January and August are the coldest and warmest months, marking midwinter and midsummer in the Meteorological reckoning, but the mildest points are obscure. So I plotted these values on a graph where the X values were the middles of the months, in order to find where the centres of spring and autumn fall. Since you have to round them up, the midmonths fall on the days of the year 16, 46, 75, 106, 136, 167, 197, 228, 259, 289, 320, and 350. The magic average temperature of 9.59°C falls on the 121st and 290th days of the year: the 1st May and the 16th October.

In other words, those are the Meteorological equinoxes. To find the cross days, and thus the start of the Meteorological seasons, we take the average of the low and mean temperatures, and the mean and high temperatures, which are 6.41 and 12.78°C respectively, and find those on the graph. That gives us 28th March (for Spring), 31st May, 19th September, and 11th November. So in the English Meteorological reckoning, Winter and Summer are much longer, 137 and 112 days long, than Spring and Autumn, which last 63 and 53 days.

Of course there are drawbacks to this approach, and my scribbled calculations may be riddled with mistakes. At least you can check them. One of my favourite pieces of scholarship is a footnote in Samuel Schoenbaum’s Shakespeare: A Compact Documentary Life, regarding the plague: “According to Stow it claimed 10,775 lives (Actually Stow gives 10,675 as his total in his Annales of England [1601], p.1274, but he has incorrectly summed up his own figures.)” Schoenbaum must have resummed Stow’s figures! As well as human error, the graph software that I used didn’t smooth the graph well, but thankfully it did show that the data’s maxima and minima are close to the graphed maxima and minima. Nor did I take account of the changes in climate over time, which would make interesting further study.

But none of this detracts from my point and conclusion that the seasons aren’t as straightforward as we pigeonhole them to be. Even for Britain’s maritime temperate climate, the seasons are as fuzzy and irregular as our calendar is arbitrary. We simplify them in our minds, but that causes us to miss their actual patterns. This is especially bad when we pander to the ‘science’ of the equinoxes and solstices because in fact they’re not even regular themselves: due to the diameter of the sun there is more daytide during the equinoxes than night. And when you look at the meteorological data, as I’ve shown, the picture gets an order of magnitude more complex.

The MET Office records start, I’ve found, coincidentally just eight years after the odd first appearance of the word ‘seasonal’. It’s one of a few antedatings of the word that I found to keep up my current record of antedating my title words. This oldest usage is in the work that John Keats (speaking of his ‘daytide’) called his favourite book, on p.386 of the lengthy medical tome called The Anatomy of Melancholy: “To be Sea-sick first is very good at seasonall times.” Though it was written by Robert Burton in 1621, the catch is that this wording only appears in the 1651 revision; I believe that this revision was made by Burton himself but I can’t refind the source, and since Burton died in 1640 it must at least have been by that date if at all.

The sentence does anyway appear in the earlier versions of the same work, but in a different and telling form, as from p.465 of the original edition: “To be Sea-sicke, first is very good at seasonable times.” So clearly ‘seasonall’ is either a synonym for ‘seasonable’, or merely an interesting compositor’s mistake or invention. The sense of seasonal as a synonym for seasonable does not appear to be in the OED to date.

Another antedating closer to the OED’s current oldest usage by Robert Mudie in 1838 comes in the Magazine of Natural History: “The birds that live constantly with us are the first of the animal creation which are actuated by seasonal changes.” The piece is called the Indicatorial Calendar, and it appears in the Magazine of Natural History (1829, p.99), edited by Edward Charlesworth. The credit is literally to “J.M.”, apparently of Chelsea given the citational style of other entries. This is the first modern use of the word that I can find, unsurprisingly the word of a naturalist.

As well as that, I also found an antedating of Mudie’s own use of the word, in his Botanic Annual for 1832: “Now, as the action of the tree in any one country, in England, for instance, depends on the seasons—is called forth by a certain degree of heat, and rendered quiescent by a certain degree of cold—it follows that, by observing the seasonal aspects of the trees in one country, and knowing the climates of other countries, the habits of trees there may be determined.” This points to yet another method of reckoning the seasons, even more complex than the Meteorological one: that of observing nature’s cumulative progress.

by Sean B. Palmer, 2007-04-02

For older articles, please see the site contents. To submit material to Lo and Behold! please read the submission guidelines and details. When it becomes available, you may purchase a paper counterpart; see the publications page for more information.