A Branch Is Not a Year

This article is about the second half of the book only. That restriction is the whole point, so it goes first.

Before 象34, this commentary dates things properly. 象6 gives 至德二載 — the second year of Zhide, 757 AD — with the month, the commanders, and the city retaken. 象17 gives 景德元年, 1004 AD, with the invasion and the minister who argued for it. 象19 gives 熙寧, the Xining era, with the memorial that started the trouble. Reign era, year, name, outcome: you can look every one of them up.

Then comes the leaf where he says he is switching from verifying to inferring, and the dating changes with everything else. Nothing below is a claim about the book as a whole. It is a claim about what replaced the reign eras once he stopped having them.

象39 and the 酉 year

象39’s commentary ends:

必至酉年始得平定也

It must reach a 酉 year before it is settled.

酉 is the tenth of the twelve Earthly Branches. It is not a year. It is a position in a twelve-year cycle, in exactly the way that “a Tuesday” is a position in a seven-day cycle. Saying an event will happen in a 酉 year is saying it will happen on a Tuesday, and the useful question is: which one?

Take the century around the printing of this edition. The 酉 years are:

1861 · 1873 · 1885 · 1897 · 1909 · 1921 · 1933 · 1945 · 1957 · 1969

Ten of them in a hundred and nine years — a branch returns every twelve years, so about eight times a century. Extend the list backwards and forwards and it never thins out; there has never been a stretch of thirteen years without one, and there never will be.

Which is the one everybody picks

象39 is commonly read as the Second World War and the surrender of Japan, and 1945 is a 乙酉 year. It fits — but being a 酉 year is a property one year in twelve has, whatever else is happening in it. 1885 is on the list too, and the Sino-French War ended in it.

That reading is a popular one, and it is fair to say where it comes from: these are web sources, not scholarship. The reference works make no leaf-level identification at all. That is worth sitting with — the confident leaf-by-leaf tables found online have no counterpart in the scholarly literature, which declines to assign leaves to events.

The structure of the inference is what matters. You do not read 必至酉年 and derive 1945. You already know 1945, and you check that it is a 酉 year, and it is, because one year in twelve is. The precision is supplied by the reader. The text supplies a filter that removes eleven-twelfths of the candidates; the reader supplies the answer and experiences the filter as a hit.

It is not one leaf doing this

If 象39 were unique you could call it a quirk. The same move runs through the speculative section, in two different shapes.

象42 ends 發現或在卯年 — its appearance is perhaps in a 卯 year. One branch again, so once every twelve years, and this time with a 或 — perhaps — sitting in front of it.

象48 gives 龍蛇相鬥想在辰巳之年 — dragon and snake struggling; one supposes in the 辰 or 巳 years. That is two adjacent branches rather than one, which makes it looser, not tighter: two years in every twelve instead of one. It goes on to add 其建都或在南方, the capital perhaps in the south.

象43 does something different, and it is the most interesting of the three. It gives no cyclical label at all. It says 必在三十年後 — it must be thirty years after — thirty years after an event which the commentary never dates. A reader cannot place that even in principle. There is no cycle to consult; there is an interval measured from an unspecified point, which is not a date but the shape of one.

You can read all four leaves in context: 象39, 象42 and 象43 in bookofchanges.academy/texts/tuibeitu/03-xiang-31-45, and 象48 in the following section.

This is not a trick

It would be easy to write the previous section as an exposé, and it would be the wrong reading. The commentator is not smuggling vagueness past anyone. He told you, on 象34, that from there on he was inferring rather than verifying, and asked to be judged leniently for it. A cyclical branch is exactly the right instrument for a commentator in that position: it says this kind of year, which is as much as he claimed to know.

The interesting failure is downstream. The branch was a hedge when written and gets read as a specification — and once you have picked your 酉 year, the leaf appears to have named it. That gap between what the page constrains and what a reader takes from it is the whole subject of this series.

Five leaves of this particular copy carry later handwriting or marks in an ink that is not the printed ink — and on three of them, the gap did not stay open. 象39, the leaf above, is not one of the three: what it carries is a mark, not a legible identification.