First attempts at laying down the law in go

In the course of adding several articles on rules to our GoGoD CD, it dawned on us, with some surprise, that historically important items such as Yasunaga's Constitution and Shimada's Primitive Rules of Go appear not to exist in an English form.

Although rules discussions do not excite us, we felt this information should not go unrecorded for a western audience. We dare to tread on the territory of mathematicians and logicians only with great trepidation, but in compensation we hope this will be new for many of you.

First, Yasunaga's Constitution. He called it a Draft Proposal. Others dubbed it a constitition. Either way, it was the first attempt to codify the rules of go in writing. It was triggered by the famous 10,000-year ko problem of 1928 in a game between Segoe Kensaku and Takahashi Shigeyuki. Yasunaga Hajime, a go writer and quasi-professional, first published proposals to create a basis for the rationalisation of go in Kido in 1929, and in 1932 followed up with the following draft.

1  Go board: a flat surface on which are inscribed 19 parallel lines both vertically and horizontally.

2  Go stones: pieces used to mark one's moves on the board. Normally there are two kinds of stones, black and white.

3  Two people play and each takes the stones of one colour.

4  Each player has the right to move alternately.

5  When a move is played on the last liberty of any of the other side's stones, those stones of the other side are removed from the board.

6  Except when stones of the other side are removed from the board, it is not allowed to play on the last liberty of any of one's own stones.

Liberty  A stone's liberties are defined as the four neighbouring points alongside the intersection it occupies. This applies likewise to groups of two or more stones.

7  Unless the right to move alternately is given up, repetition of a previous position is prohibited.

8  End of the game: occurs when the right to move has been given up three times in succession.

9  At the end of the game the result is determined by the following means.

* Stones which cannot be removed from the board or which can prevent being removed are referred to as being alive.

* Stones which are not alive are removed from the board at the end of the game.

* Areas of the board which can be surrounded by stones which can live independently are referred to as territory. In this case, each intersection is considered to be one point of territory.

* Surrounding an intersection refers to the case where there are stones of one side on the four neighbouring intersections (liberties) along the lines from that intersection.

* Stones which have been removed from the board are substracted from the total number of points of the side owning those stones.

* Victory is decided according to the relative size of the territories thus obtained and to the extent of the difference between them.

Although valid criticisms of this draft are easy, some which look valid are unfair. It was written for the existing go world in Japan rather than beginners, and the style of putting definitions after rather than before their first use was (and is) typically Japanese and works for them. Whatever else this Draft did, it legitimised discussion of the rules and was probably the source of the immensely useful shorthand of three passes to end a game.

The first treatment of go rules in a more rigorous mathematical style seems to be due, in Japan, to Shimada Takuji. He was possibly stimulated by the Americans Karl Davis Robinson and John M. H. Olmsted, though there was an earlier mathematical work on go in Japan, but not to do with rules. This was due to Fukuzawa Sanpachi (presumably the son of the famous educator Fukuzawa Yukichi) who wrote on "Go and shogi in two or more dimensions" in 1934. Shimada also covered toroidal, cylindrical and 3-D boards in his book Igo no Suuri (Mathematical Theory of Go), but now we turn to his attempt to improve on Yasunaga.

Next >>