Kawabata's Master of Go

The Master of Go
http://images.amazon.com/images/P/0679761063.01.LZZZZZZZ.jpg [Cover art: Detail from ''Boating in the Shade of the Willow Tree'', artist unknown, The Museum of Fine Arts, Spfld., MA., Raymond Bidwell Collection]
By: Kawabata Yasunari
Publisher: Vintage International?, 1996
ISBN10 0-679-76106-3

Kawabata, a Nobel prizewinner for Literature, published this novel in 1951. It describes the last game, in 1938, of a Go master (actually Shusai Meijin) and the younger challenger (actually Kitani Minoru). It was first published as Meijin in the Shincho magazine in 1951. It was printed in book form in 1954.

The version translated into English by Edward G. Seidensticker is a shorter form preferred by Kawabata, since it was the one included in the most recent edition of his complete works (Vol. 11 as of 1980). Some material was cut from between the end of the match and the master's death. Kawabata's name is ordered in the Western order (Yasunari Kawabata) in the English-language version of the book. Seidensticker's translation is published by Wideview/Perigee Books, of G.P. Putnam's Sons, ISBN 0-399-50528-8 and Vintage International, ISBN 0-679-76106-3.

The diagrams have been inverted from the Japanese edition, and some stones lose their move numbers in some diagrams. The group of four captured White stones become four unnumbered Black stones in later diagrams. The translation itself has won good reviews.

Other editions exist, as well as translations to other languages, listed here in the order of appearance (language: title, translator, publisher, year, ISBN if available):

  1. Korean: Myeong In, Ming Byeong Son (閔丙山), Shingu Munhwa Sa, Seoul, 1969
  2. English: The Master of Go, Edward G. Seidensticker, Alfred A. Knopf, New York, 1972, ISBN 0-394-47541-0
  3. French: Le Maître ou Le tournoi de Go, Sylvie Regnault-Gatier, Albin Michel, Paris, 1975, ISBN 2-253-04673-6
  4. Serbo-Croatian: Velemajstor (part of Velemajstor, Snežna zemlja), Ljiljana Đurović, Slovo Ljubve, Beograd, 1981; published also by Logos, Beograd, 2007, ISBN 978-86-85063-36-7)
  5. Chinese: Mingren, Liu Hua Ting (劉華亭), Xingguang Chubenshe, Taipei, 1985
  6. Dutch: De meester van het go-spel, Annemarie van Frankenhuysen (translated from the English translation), Uitgeverij BZZTôH, 's-Gravenhage, 1987, ISBN 90-6291-304-0
  7. Czech: Meidžin (part of Tanečnice z Izu a jine prózy), Vlasta Winkelhöferová und Miroslav Novák, Odeon, Praha, 1988
  8. Turkish: Go ustasi, Belkıs Çorakçı (Dişbudak) (translated from the English translation), Remzi Kitabevi, İstanbul, 1992, ISBN 975-14-0335-9
  9. Italian: Il Maestro di Go, Cristina Ceci, Arnoldo Mondadori, Milano, 1995, ISBN 88-7710-513-5
  10. Chinese: Mingren, Ye Wei Qu (葉渭渠), Zhongguo Shehui Kexue Chubenshe, Beijing, 1996 (also published in Taiwan by Ecus Publishing House, Sindian, 2002, ISBN 986-7897-25-0)
  11. Russian: Mejdzin, Boris Kornilov, Arkadij Bogatskij, Kiev, 2003 (published already 1998 in Moscow and 2004 in the internet at [ext] lib.ru)
  12. Polish: Meijin - Mistrz go, Henryk Lipszyc, Wydawnictwo Elay, Bielsko-Biała, 2004, ISBN 83-916749-4-0
  13. Spanish: El Maestro de Go, Amalia Sato, Emecé Editores S.A., Buenos Aires, 2004, ISBN 950-04-2615-3
  14. Romanian: Maestrul de Go, Flavius Florea, Humanitas Fiction, Bucuresti, 2007, ISBN 978-973-50-1624-1
  15. German: Meijin, Felix Heisel, Brett und Stein Verlag, Frankfurt, 2015, ISBN 978-3-940563-22-4

[ext] SGF of game in book

John Fairbairn has published a book The Meijin's Retirement Game devoted to this game. In addition to a detailed commentary on the moves the book discusses the historical context of the game and the various translations that have been published.


I've just read the book. I HolIgor was a little surprised to actually find the diagrams of the game there. As a piece of high literature it could just mention the author's interpretation of the value of moves. But Kawabata takes a more technical approach, still keeping his excellent style and writing more about people than the game. In the situations where his amateur judgement is not sufficient he cites Go Seigen.

The moral issue of the book is the conflict between the old artistic values and the new pragmatic young approach. Kawabata takes the conservative side and is definitely Meijin's fan. For balance though he shows his sympathy to the challenger (Kitani Minoru - the changed name won't fool anybody). His school is mentioned as well.

The highest point of the book is, of course, Black 121 (B2 in the diagram below). That was a sealed move. The next session was in two days. The referee opened the envelope and could not find the move. Then he muttered: "Ah..."

I looked at the diagram to find it. When I finally found the move I was disgusted. Disgusted was Kawabata. Kawabata did not see the Meijin's reaction but that day the Meijin committed multiple errors and lost the game. During the dinner he said that the challenger spoiled the game and he wanted to resign immediately.

Interestingly, pros were on the challenger's side seeing nothing wrong with sealing a forcing move (or making a forcing move just to gain some time for the real problem on hands). Go Seigen, for example, was more critical of the Meijin's automatic reply. A different defensive move was better in his opinion. It seems that the Meijin's fast reaction was just a sign of his anger.

2 (121 in the game) was a sealed move  

Here is the crucial position from the book. White played an empty triangle 1. Perhaps Black was too afraid to spoil his chance, but 2 was the sealed move. At the beginning of the next session, disgusted White replied 3 without thinking. Go Seigen thinks that White had to play a. My opinion is that all three of these moves were poor.

Looking through the diagrams I had a strong impression that White had a considerable lead up to White 130 after which the game became close. The problem is in komi, of course, to be more precise, in its absence. Even after all the Meijin's mistakes Black won by 5 points only. Komi 5.5 would make White the winner. However, this consideration is not applicable. With komi 5.5 Black would play far more daringly and riskily from the very beginning of the game. Black played to win by the existing rules.

Bill Spight: The attitudes of the players softened over time. I don't think that Kitani ever admitted publicly that he made the play so he could think about the game during the recess. However, in an article about the game for a Go magazine not long before his death, Kitani was asked about the move, and the controversy it caused. Kitani said, "Well, he answered it," and chuckled.

Andrew Grant It's unfair to criticise Kitani for playing a forcing move in this situation (certainly anything as strong as "disgust" is inappropriate). Shusai was notorious for playing tricks like this himself. In the old days the stronger player had the right to suspend play for the day as long as it was his turn; Shusai used to take full advantage of this, suspending play whenever he faced a tough decision so that he could analyse the position during the recess with his pupils. Kitani's friend Go Seigen was just one victim of this practice. One game Shusai played against Karigane in 1920 ended up taking 20 playing sessions over a period of six months. This was the main reason why Kitani insisted on fixed adjournment times with sealed moves. If he took the opportunity of the sealed move to give Shusai a dose of his own medicine, it's hard to blame him. Certainly for Shusai to complain that the game had been spoilt reeked of hypocrisy. Don't make Shusai out to be some kind of injured innocent.

Migeru Ah, that explains why at the beginning of chapter 5 it says _"On the first day there were only five plays, from black 101 to black 105. A dispute arose over scheduling the next session. Otake rejected the modified rules the Master had proposed for reasons of health, and said that he would forfeit the game."_

Another curious aspect of the game was the fact that Meijin played an empty triangle in the center. Kawabata, who was about 5k level I (HolIgor) guess, understood that the shape was bad, so he explained the fact by the unwillingness of the Master to give in any point in a match that was so close. The result turned out to be disastrous. Kitani used his opponent's bad shape effectively.

John Fairbairn That seems unnecessarily rude about Kawabata. The GoGoD collection has two games by him. One marks his promotion to 2-dan in 1954, on 6 stones against Iwamoto. In 1963 he played a sponsored serious game for publication against the Meijin-Honinbo Sakata as 3-dan, also on 6 stones, and won by 6.

Anonymous: I like to add, that (in Master of Go, the 1981 Perigee reprint) he is writing about himself giving a 13k western amateur six handicap stones. It reads: "I was Grade Thirteen," he said with careful precision, as if doing a sum. He was an American. | I first tried to give him a six-stone handicap. He had taken lessons at the Go Association, he said, and challenged some famous players. He had forms down well enough, but he had a way of playing thoughtlessly, without really putting himself into the game. (Kawabata:115-16)

BobMcGuigan: I think some attention should be paid to Kawabata's literary intent in writing "Meijin". It is not strictly a piece of journalism and hence it may not be appropriate to make too strong a connection between what the real Kitani and Shusai said about the real game and what the characters Otake and the Master say in the book. Many of Kawabata's works are concerned with the decline of traditional Japanese cultural icons. "Yukiguni" (Snow Country) treats the decline of the institution of the geisha, "Sembazuru" (Thousand Cranes) the decline of the tea ceremony, and "Meijin" (The Master of Go) the decline of the traditional system of Go. Kawabata is mourning the loss of traditional Japanese cultural values, with the Master representing tradition and the character Otake representing the new wave. One scene in "Meijin" that brings this out is the scene at the inn where the master has to sit in a chair and drink Western style tea. The supposed flap over the sealed forcing move is also a contrast between the old way and the new.

Ren: Does anyone have the full SGF of this game or know in which collection it could be found? Would be nice to read through.

Uwe Schmidt: Gobase.org has it. Search for Shusais games. [ext] http://www.gobase.org

Ben: I think there might be a mistake with the translation. It seems that white's group would die if he played at "a" in the above diagram.


Other response  

Bill: I checked the book (Knopf, 1972). What Wu (Go Seigen) said was that if Black waits until after White has played W1 - B4, White can answer B6 at W7, and Black does not have the ko threat or threats that are there in the actual game. (There was no diagram in the book, this is my rendition.)

Also, Kitani pointed out elsewhere that after W7, B10 is not sente. This appears to be an infinitesimal difference in chilled go, a very subtle point.

John F. Bill, there are several baffling aspects to your comment. You say you checked the book, but do not say which one. But whether you are citing the Japanese or the English one, there are major discrepancies with my versions of the books.

First, my edition of the Seidensticker translation (Penguin 1976, page 124) says: "Wu touched only lightly on the play. After a diagonal and connection on White's part at E-19 amd F-19, he said, 'White need not respond as the Master did with 122 even to Black's 121, but could defend himself at H-19. Black would thus find the possible ko threats more limited.'"

My edition of the Japanese version (Shincho Bunko 1982, page 121) has no inline diagrams (the game record is given only in bulk at the end of the book), and there is accordingly no possibility of reference to the format W1, W3 no hanetsugi. The portion that relates to this in the Japanese original is, instead, just as point-specific as Seidensticker's version: "kono shiro kara, 'yo-ichi', 'ka-ichi' to, hanetsugi wo utareta ato de wa,..."

Also there is no reference to not having a ko threat or threats. The Japanese is koudate ga kikinikuku naru - a pedantic difference but the English is quantitative, the Japanese qualitative. Seidensticker's version seems to capture that.

Assuming, you are referring to an English book, this raises the possibility that there is more than one Seidensticker version around. My Penguin edition mentions that the translation was copyright Knopf 1972 and a note at the beginning mentions that "new diagrams, showing the progress of the match, were prepared for this Penguin edition by Stuart Dowsey, Director of the London Go Centre."

Now I remember talking to Stuart about this translation. It was 30 years ago, and I have to rely on memory, but my strong recollection is that Stuart tried very hard to get Seidensticker to change parts of the translation but was totally unsuccessful (and had there been any changes they would surely also have been reported or acknowledged). There may have been changes in later editions but it beggars belief that 'and' would be changed to 'or' on the say-so of a go player. There are certainly, from a go player's point of view, many infelicities in the translation (which is despite that a masterful work), but I'm not sure I recall anything that might go down as a pure mistranslation of this order.

Someone else refers on this page to "other translations and editions." Can they be more specific, please? If it is other English transaltions they mean, this is interesting because I have a recollection Seidensticker claimed to have secured a monopoly on translaing Kawabata's work.

If we are talking about other editions, well, the plot thickens - is there really an alternative Seidensticker text out there?

Bill: Thanks, John. :-) I checked the Knopf (1972) version. Here is the exact wording on p. 165:

"Wu touched only lightly on the play. After a diagonal linking on White's part at E-19 or F-19, he said, 'White need not respond as the Master did with 122 even to Black's 121, but could defend himself at H-19. Black would thus find the possible plays from kou more limited.'" (The text has a long o in kou instead of ou.)

This English plainly needs translation for go players. I adapted the diagram and took "W1, W3 no hane-tsugi" from Kitani's commentary. I think that it is something like "'yo-ichi', 'ka-ichi' to, hanetsugi". :-)

Anyway, thanks to your reply I have revised my comments above. :-)

John F. Interestinger and interestinger. Thanks, Bill - can anyone shed any light on other changes to the translation?. At this stage I'd have to assume changes were made for the Penguin edition, on the grounds that they listened to Dowsey regarding diagrams and they were in close touch with him regarding the Iwamoto Go for Beginners, but they omitted to acknowledge the changes to him and/or the general public. Maybe this was the only change. I do know that there was at elast one change that Dowsey wanted that was not made (regarding oyogu).

Reprints of the book in the UK may well follow the Penguin edition, but do modern editions in the USA follow Knopf or Penguin? And what about non-English translations?

Bob McGuigan: My copy of the Seidensticker translation is published as a Perigee softcover, copyright 1981, same ISBN as the one mentioned at the beginning of this page, but it has a different cover. It is simply a reprinting of the Knopf version. My copy of the Japanese is the same as John's.

senbazuru: I just added all the translations of "Meijin" that I am aware of, with a few question marks about the Russian and Serbian(?) one. If there is any other translation out there, please give me a note - for my collection of translations that I am editing in the Japanese Wikipedia version of [ext] http://en.wikipedia.org/wiki/The_Master_of_Go

senbazuru: Concerning the above mentioned sentence that had variations in the Knopf and Penguin editions of the English translation, here are two other translations:

Variations of translations
Language Page Text
Japanese 121 呉六段の解説では、この白から「よ一」、「か一」と、はね粘ぎを打たれた後では、「黒が百二十一と打っても、白は百二十二に受けないで『を一』と生きる。すると黒から劫立てが利きにくくなる。」と、黒百二十一の意味にあっさり触れただけだった。大竹七段もその意味で打ったにはちがいなかった。

Go rokudan no kaisetsu de wa, kono shiro kara "yo ichi", "ka ichi" to, hanetsugi wo utareta ato de wa, "Kuro ga hyakunijūichi to utte mo, shiro wa hyakunijūni ni ukenaide 'wo ichi' to ikiru. Suruto kuro kara kōdate ga kikinikuku naru." to, kuro hyakunijūichi no imi ni assari fureta dake datta. Ôtake shichidan mo sono imi de utta ni wa chigainakatta.

English (Knopf) 165 Wu touched only lightly on the play. After a diagonal linking on White's part at E-19 or F-19, he said, 'White need not respond as the Master did with 122 even to Black's 121, but could defend himself at H-19. Black would thus find the possible plays from kō more limited. No doubt Otaké's explanation would have been similar.
English (Penguin) 124f Wu touched only lightly on the play. After a diagonal and connection on White's part at E-19 and F-19, he said, 'White need not respond as the Master did with 122 even to Black's 121, but could defend himself at H-19. Black would thus find the possible threats more limited. No doubt Otaké's explanation would have been similar.
French 142f Go Sei-gen n'y fit qu'à peine allusion. D'après lui, les Blancs s'étant joints par la diagonale à E-19 ou F-19, «rien ne les contraignait à réagir, même à ce 121, par le 122 du Maître. Il existait une défense en H-19. Les Noirs auraient alors trouvé les possibilités de ko très restreintes». Nul doute qu'Otaké ne l'eût expliqué de la même manière.
Italian 157 E Go Seigen puntualizzò che, se i bianchi avessero fatto un hanetsunagi in E-19, «avrebbero poi potuto sopravvivere in F-19 invece di rispondere in quel modo con il 122 al 121 nero. Così facendo, avrebbero ridotto di molto le possibilità di dei neri» Go Seigen sorvolò sul senso profondo del 121 e Ōtake avrebbe di certo concordato con la sua analisi.

[ext] ilanpi's review of La Joueuse De Go also contains some comments on the French translation of this book.

John F. An interesting table. The French and Italian read as if translated from the English, so I think they can be safely ignored. But Seidenstecker's attempt seems not only to miss the go aspects but also to miss a couple of important points of Japanese grammar.

To get the go points out of the way first, the "diagonal linking" for hanetsugi is highly misleading, and he should have known that Go Seigen had naturalised and taken the official name Go, and that had already been in long use anyway, so Wu is a case of not knowing go history. On top of that, it was a commentary not an explanation (wrong nuance). Go was in a sanatorium not far away and the latest moves were ferried to him for his newspaper column. And we don't need all these "dans."

The more interesting points are the grammar. The first is that he appears to ignore the adversative passive utareta, i.e. if Black "suffers" from White playing the hanetsugi, which we can better render as "if Black allows White to play the hanetsugi". Secondly, "No doubt Otaké's explanation would have been similar" invents an "explanation" when the Japanese says "played". I suspect he was thrown by "mo" and wanted it to make it apply to the noun. Since only one player can play a particular move, he obviously couldn't say "Otake also played..." so he changed the verb. However, mo cam also be used to apply to other parts of a sentence than the noun (as a sort of pseudo-adverb) and it seems clear that the meaning here is "No doubt Otake played this with the same intent." It might seem to say the same thing in the end, but using the "explanation" version seems to add a nuance of Otake/Kitani feeling he had to, or might have to, explain himself, which is not there at all. In any case, "... to iu imi de utsu" is a pretty normal go expression for "to play with ... intent".

So, in view of all of that, my suggested translation would be: "In his commentary, Go touched briefly on the significance of Black 121. He said, 'If Black had allowed White first to play the hanetsugi E19 and F19 here, then should Black still play 120 White would live with H19 and not answer at 121. Which means that Black's ko threat there would no longer work.' No doubt Otake played this with the same intent,"

It will be evident from this that, in my view, Kawabata is not trying to point up the fuss about sealing a forcing move. Rather he is implying that for the pros Go and Otake there was no fuss because it was a sensible play.

senbazuru: Quite a lot of good points, too many to let them go unnumbered.

  1. Other than the Dutch and Turkish translations, which are made from the English one, the French claims explicitely on its title page: "Roman traduit du Japonais par Sylvie Regnault-Gatier". Although I forgot details I had looked up earlier, there are indeed some places in the translation with a choice of words which prove that they were translated from the Japanese. But at many places the similarity to Seidensticker's wording and sentence structure is very high. I am pretty sure that SRG had Seidensticker's book constantly open to check what the master hade made of it.
  2. The Italian translation is interesting because its base is not the 41 chapter version of "Meijin" which was used by Seidensticker, Ye and Kornilow, but Ceci took the 47 chapter edition that was published in book form in 1954. Here, no furigana were provided with はね粘ぎ like in the modern Shinchô bunkô paperback edition, so she translated it as hanetsunagi. She very much let technical Go terms untranslated, but also other Japanese words. Take chapter one, second sentence: 数年六十七であった。 / Kazoedoshi rokujûshichi de atta. / He was sixty-seven years old by the Oriental count. / Il entrait dans sa soixante-septième année. / Aveva sessantasette anni, secondo la chronologia kazoedoshi. Whether Ceci referred to the Seidensticker translation once in a while is anybody's guess, but at least with the additional 6 chapters she was on her own.
  3. Seidensticker was well aware that Go Seigen is Go Seigen. He writes in note 9 (referring to chapter 10, page 35, where Go is mentioned first in the Penguin edition): "Wu Ch'ing-yüan, born in Fukien Province in 1917. He is far more famous under the Japanese version of his name, Go Sei-gen; but he will be called Wu throughout this translation. The game of Go and Goi the newspaper reporter seem to introduce quite enough possibilities for confusion without Go the player of Go." OK, Go Seigen was born in 1914, but Seidensticker makes a good point; after all, he was translating Meijin for the uninitiated American mass market reader.- I didn't know that (or forgot if it is written in Meijin) that Go was in a sanatorium at that time; from where is this info?
  4. Seidensticker doesn't need all these "dans". He often translates 大竹七段 as "Otaké of the Seventh Rank" or just as "Otaké". This is a fundamental dilemma of readability in English: Make it "Otake 7 dan" (or 7 Dan, 7-Dan, 7-dan) and make a footnote, or make a cumbersome literal translation, or just delete this extra stuff. How would you translate it in a novel? I have only your "Kato's Attack and Kill" at hand where you use the "Otake 7-dan" style. (Good work, by the way :-)
  5. Same question for hanetsugi versus "diagonal and connection" of Seidensticker/Dowsey/Penguin in a novel.
  6. Concerning the "mo" in "Ôtake shichidan mo sono imi de utta ni wa chigainakatta": to which part of the sentence does it apply if not to the noun "Ôtake shichidan"? What do you mean with the pseudo-adverb? BTW: is it shichidan or nanadan?
  7. In your suggestion, you omitted "dake" and made two sentences. Shouldn't it be: "In his commentary, Go touched only briefly on the significance of Black 121, saying ..."? And the part "If Black had allowed White first to play the hanetsugi E19 and F19 here" is actually not in the quoted commentary in the original, while "would no longer work" for "kikinikuku naru" is also slightly abrupt ...

Notwithstanding those nuances, your conclusion about Kawabatas intentions seem convincing. A fine detailed analysis of this dense Kawabata style!

John F. senbazuru I have never researched foreign editions of Meijin so I cannot add anything to your useful comments on these. What you say seems sound.

On your other points: Of course Seidensticker knows about Go Seigen. His origin is mentioned in the novel at the end of Chapter 28 (Shincho Bunko Japanese edition), and Chapter 29 explains that Go was in the sanatorium, although this is well known from other sources. Kawabata also says he was keen to see Go's commentary, so Seidensticker ought to have realised some of the points I made above (or maybe he did, and didn't agree, but does his treatment of the go aspects inspire confidence?).

As to his choice of Wu to avoid confusion with Goi and go, I don't feel it is a good point nowadays, and probably not even then. The references to Go Seigen involve either use of the full name Go Seigen or are located close to this, he's not a major character, and the reporter is even less of a character. In fact, I can't even remember the very odd name Goi. There was a reporter called Sunada, I think, but he was a fictional character. I imagine the real-life reporter would have been Mitani Suihei.

On dans, maybe Seidensticker is sparing in their use - I've only read his version once, many years ago. But in general, I'd say they don't belong in a novel or descriptive text, but do have a place in a technical book because you need to know whether the opinions or examples quoted can be trusted.

By pseudo-adverb I mean acting as an adverb here. Therefore it modifies the predicate (i.e. played with a certain intent) rather than the subject. I suggested "same" to convey that. This usage is not uncommon.

My omission of "only" was not intentional and it can usefully be inserted. I found it very hard to edit this page and ended up trying to memorise the Japanese so I could translate it down below. This also explains my abrupt "no longer work", though I'm not sure it really needs changing. If you want you might say, "would now be unlikely to work", but English tends to avoid these Japanese softenings and using them can even cause confusion: implying an extra nuance that is not really there.

On a different tack, there have been references by Kitani and Go that suggest they did not see eye to eye with Kawabata's interpretations. Indeed, Kawabata felt obliged later to explain some of his choices (such as using the fictional name Otake). Inevitably this meant a certain amount of revisionism about the novel took place, quite apart from the rewriting. That seems to make it even more important to place this book properly in its go context. I've covered some of the context in my forthcoming book on the Kamakura match, so I'll skip further comment on that.

Kawabata's Master of Go last edited by senbazuru on October 22, 2015 - 15:09
RecentChanges · StartingPoints · About
Edit page ·Search · Related · Page info · Latest diff
[Welcome to Sensei's Library!]
Search position
Page history
Latest page diff
Partner sites:
Go Teaching Ladder
Login / Prefs
Sensei's Library