This post is doing great on diagrams but it needs more formulas and math equations.
Yoshikuni Ichiro and the others would be proud of this pursuit of rationality that does not in any way impair the richness of Japanese Go. The problems that may have arisen in the past have been eliminated and we can enjoy Go just as we ever did. The rationality of Japanese Go, which has a long history and tradition with courtesy and dignity, has been preserved. Well done.
we're doomed
GERMAN interpretation of J89's intended contents
-
John Fairbairn
- Oza
- Posts: 3724
- Joined: Wed Apr 21, 2010 3:09 am
- Has thanked: 20 times
- Been thanked: 4672 times
Re: GERMAN interpretation of J89's intended contents
Yes, I am. And so, it seems, are you.Apparently, you are still fine with utilising ambiguous technical terms in rule texts.
There can be ambiguity when there is no context, but almost always there is context and in most such cases the ambiguity is purely theoretical and not a problem for humans, though computers may choke on it.
However, you are also walking the ambiguity minefield in that (a) you are using technical words like 'point' and (b) you are implicitly accepting the ambiguity in most Japanese nouns, which do not distinguish between singular and plural, or in the lack pf indefinite and definite articles.
I don't have to imagine this sort of scene. I have lived it. On my first trip to Germany, by train, there was a border stop and customs officers came on board, checking everyone's bags. In my bag I had a sandwich. The German officer asked if it contained butter. I said yes and he confiscated my lunch. Apparently it was against the law to import butter into Germany without a licence (this was pre-EU of course). At first I thought he was re-living the war, but he turned out to be quite a nice chap. Still, when I suggested common sense should be used, he looked at me as if I'd come from Mars: "This is regulations!" (If I'd used margarine instead, would it still have been a Butterbrot?)Just imagine if there were a punishable ban on eating apples in public in Germany.
Usually, Japanese tourists (quite fond of pears) have to pay the € 100 fine that is due for violating this ban.
Police officer: "Oh, you didn't know that everyone in Germany knows that "林檎" also means "梨"? Nevertheless, ignorance does not protect from punishment."
I was going to add some other things but kvasir has covered some of them. I'll therefore just say that I ventured into this discussion when CDavis was being criticised for saying, basically, on linguistic principles, that of course the Japanese understand their own rules. He was challenged precisely on those linguistic principles and at that point I, as a linguist, got interested. It seemed to continue as a linguistic exercise, and this appeared to be confirmed when we were offered a German translation.
However, it seems to have unravelled quickly since then and it seems that in reality Cassandra (a prophetess cursed to issue true prophecies never to be believed, as I recall?) was far more interested in logic and computer programming than in language - all the references to languages were used just to support the algorithms. At that point I suddenly lost interest. I am already in the habit here of skipping over anything (by any writer) that smacks of formulas, numbers, maths and even diagrams. I have zero interest in the logic of rules. Unless language is allowed to raise its Hydra-like heads again, there is little more for me to say. But I do continue to maintain that the Japanese do understand their own rules, even with their alleged ambiguities. Which is not quite the same as saying they understand all the applications of those rules, incidentally - otherwise we wouldn't get so many Japanese people asking pros how to handle this or that unfamiliar position. In that respect, what it now appears Cassandra is really doing - producing algorithms - may be useful, but I'm not the one to judge. I, for example, haven't got a clue what the latest Japanese rule on bent four is. Though I can tell you what the Japanese terms for it are.
-
CDavis7M
- Lives in sente
- Posts: 716
- Joined: Tue Jun 15, 2010 2:18 pm
- Rank: Shokyu
- GD Posts: 0
- Universal go server handle: CDavis7M
- Has thanked: 109 times
- Been thanked: 140 times
- Contact:
Re: GERMAN translation of J89's intended contents
Your statement jumped out at me while scanning down the thread.jann wrote:why bother with the disguise of logical rules at all? Just include 25 precedents and done.
Did you catch on to what the Japanese committee was doing or was this a joke?
It seems pretty clear to me that this is what the committee (chairmen Yoshikuni, Oeda 8dan, etc.) planned to do, and what they did. Life and death is merely defined. There's no need to play it out, which is why the later examples don't bother. Bent 4 is dead, seki collapse is a thing, so bent 4 and seki collapse is easy to assess.
This is how I read the statements about the revision 日本囲碁規約改定の経過について and the revision outline 日本囲碁規約改定の概要（ご参考）. They defined life and death and merely provided a (possible) rationale for why it is the way it is. They even took back torazu sanmoku.
-
Cassandra
- Lives in sente
- Posts: 1326
- Joined: Wed Apr 28, 2010 11:33 am
- Rank: German 1 Kyu
- GD Posts: 0
- Has thanked: 14 times
- Been thanked: 153 times
Re: GERMAN interpretation of J89's intended contents
It is no wonder that Japanese understand Japanese rules, written in Japanese, which were clearly designed for the application by Japanese Go players.John Fairbairn wrote:But I do continue to maintain that the Japanese do understand their own rules, even with their alleged ambiguities.
Apparently, Nihon Kiin's main concern was to get rid of the weaknesses, shortcomings, contradictions and errors of J49.
"Internationalisation of Go" and similar, although mentioned in the preface, was just a sideshow, a nice-to-have. Otherwise, they would have published an AUTHORISED translation into English in parallel.
I doubt that the Nihon Kiin had any truly altruistic interest in developing the game of Go in the West. Rather, the focus was probably on getting ahead of other East Asian countries / competitors on that field.
----------
Once upon a time, the Japanese were firmly convinced that no one other than native Japanese could achieve masterful perfection in Ikebana (same for any other of the Japanese arts).
A few decades ago, this view changed somewhat to the effect that non-Japanese had to study Ikebana in Japan for a long time to achieve this mastery.
On the occasion of the Japan Days 2008 in Leksand (the European Go Congress was held at the same time) my wife, together with two other Ikebana ladies, had organised a large Ikebana exhibition and filled it with arrangements.
In our hotel, she was constantly addressed in Japanese by Japanese women who wanted to express their appreciation. My wife had to apologise and the conversation had to be continued in English; she did not know any Japanese and had not been to Japan at that time. Years later, on the occasion of our trip to Japan, it turned out that she was at war with Japanese anyway, this language was somehow not really hers.
----------
It may be Japanese culture that all faults in J89 shall be cured by applying "Japanese culture". However, it would have been more in line with the Japanese striving for perfection not to make any mistakes in the first place. As I already mentioned before, the work on the formulation of the rules lacked a professional approach.
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)
Igo Hatsuyōron #120 (really solved by KataGo)
Re: GERMAN translation of J89's intended contents
This is about the opposite of the truth. In reality J89 took a huge step forward and found and established a logical theory behind Japanese L/D with the new ko rule. It also performs well in examples. Torazu3 actually changed, consistently with the new theory. My comment was a response to the mentality of ignoring basic counterexamples for random new inventions that simply don't work.CDavis7M wrote:Your statement jumped out at me while scanning down the thread.jann wrote:why bother with the disguise of logical rules at all? Just include 25 precedents and done.
Did you catch on to what the Japanese committee was doing or was this a joke?
It seems pretty clear to me that this is what the committee (chairmen Yoshikuni, Oeda 8dan, etc.) planned to do, and what they did. Life and death is merely defined. There's no need to play it out, which is why the later examples don't bother. Bent 4 is dead, seki collapse is a thing, so bent 4 and seki collapse is easy to assess.
This is how I read the statements about the revision 日本囲碁規約改定の経過について and the revision outline 日本囲碁規約改定の概要（ご参考）. They defined life and death and merely provided a (possible) rationale for why it is the way it is. They even took back torazu sanmoku.
-
CDavis7M
- Lives in sente
- Posts: 716
- Joined: Tue Jun 15, 2010 2:18 pm
- Rank: Shokyu
- GD Posts: 0
- Universal go server handle: CDavis7M
- Has thanked: 109 times
- Been thanked: 140 times
- Contact:
Re: GERMAN translation of J89's intended contents
I think we are starting off saying the same thing. You say "logical theory behind" life and death status and I said "rationale." But then I take it a step further.jann wrote:This is about the opposite of the truth. In reality J89 took a huge step forward and found and established a logical theory behind Japanese L/D with the new ko rule. It also performs well in examples. Torazu3 actually changed, consistently with the new theory.CDavis7M wrote:...It seems pretty clear to me that this is what the committee planned to do, and what they did.jann wrote:why bother with the disguise of logical rules at all? Just include 25 precedents and done.
I think that the "theory/rationale" really is just that. The life and death rulings/definitions already existed and there is just a new theory/rationale to provide the most consistency. This is the stated goal of the committee. They were not trying to create a procedure to define life and death status. They were trying to rationalize the existing life and death rulings/definition with consistency. Inconsistent rulings were thrown out and made consistent.
Moreover, the correct resolution of the compound examples near the end of the list can really only be reached by layering several of the ruling/definitions from the earlier examples. In the compound positions, if you just try to "play it out" using Article 7-2 you can get weird results where teire would be required (when it isn't needed), or you end up in a loop, or you somehow get the wrong life and death status. Like in Examples 16, 17, and 18, etc. For example, after the first move in confirming life and death status, a double-ko loop can start up where one player cannot afford to NOT pass for the ko to play elsewhere. Though I think this doesn't happen if you follow the "playstyle" of the examples (which is never explained).
So, these are just life and death definitions (based on a rationale theory) and you can apply them and layer them as needed:
• Bent-4 in the corner makes stones dead, even with other kos.
• An eye beats no eyes in a triple ko, even with other kos.
• A ko for temporary life is dead, even with other kos.
• Mannen ko is seki.
• Sometimes Teire is needed.
• Sometimes Teire is not needed.
So, why bother with the disguise of
Last edited by CDavis7M on Tue Oct 19, 2021 6:46 pm, edited 1 time in total.
Re: GERMAN translation of J89's intended contents
https://www.lifein19x19.com/viewtopic.p ... 02#p267502CDavis7M wrote:in confirming life and death status, a double-ko loop can start up where one player cannot afford to not pass for the ko to play elsewhere.
-
CDavis7M
- Lives in sente
- Posts: 716
- Joined: Tue Jun 15, 2010 2:18 pm
- Rank: Shokyu
- GD Posts: 0
- Universal go server handle: CDavis7M
- Has thanked: 109 times
- Been thanked: 140 times
- Contact:
Re: GERMAN translation of J89's intended contents
I answered you over there, but that suggestion is ignoring basic counterexamples for random new inventions.jann wrote:https://www.lifein19x19.com/viewtopic.p ... 02#p267502CDavis7M wrote:in confirming life and death status, a double-ko loop can start up where one player cannot afford to not pass for the ko to play elsewhere.
-
Cassandra
- Lives in sente
- Posts: 1326
- Joined: Wed Apr 28, 2010 11:33 am
- Rank: German 1 Kyu
- GD Posts: 0
- Has thanked: 14 times
- Been thanked: 153 times
Re: GERMAN interpretation of J89's intended contents
Continued.RobertJasiek wrote:How / with which principles have you derived / created the German version?
Approach #4: Topology of "valuable regions".
Already subdivided regions
The aim of the game of Go is to create value through securely fencing in board points.
In the diagram above, the points
are securely fenced in by White's 
stones.The points
create value for White.Black is unable to prevent White from creating value.
A Black move
, intending to capture White's stones with another move at A, is forbidden by the rules.A division of White's formation into two parts is sufficient.
Why "95"?
Black's forbidden move is a "no-go".
"Nine" in English sounds like "nein" in German, which is "no" in English.
"五" = "go" is "five" in Japanese.
----------
Further subdivision of White's formation does not fundamentally strengthen the stability of the fence (there are exceptions, of course, but which we will not discuss here).
The points
create value for White.Please note that White could -- in principle -- transform this formation into the basic one.
As a matter of course, White would do without it in practice so as not to destroy her own values.
----------
A larger size of a fenced part of the board does not fundamentally affect the stability of the fence.
The points
create value for White.Again, Black is hindered to prevent White from creating value.
And again, White would be able to create the basic formation.
----------
A Black stone
inside the fenced region enhances the value of this region for White.The points
and 
create value for White.The value of
is worth two times the value of .Again, Black is hindered to prevent White from creating value.
And again, White would be able to create the basic formation.
----------
Here, Black's fenced-in stones
also fence something in the corner.However, White is able to prove that Black's fence is not solid enough.
The points
and create value for White.The value of
is worth two times the value of .----------
Obviously White was busy mending her fences elsewhere on the board.
Again, Black's stones
in the corner also fence something.White fenced one point
at the upper edge.Black fenced one point
in the corner.Black cannot prevent White from creating value.
White cannot prevent Black from creating value.
However, both fences share one point
poisoned no man's land that makes the fenced areas nearby unsellable.The upper area of the board has no value for anyone.
Undivided regions
Be aware! Newbies at work!
White's
stones fence something in the corner.However, Black is able to prove that White's fence is not solid enough.
There is nothing of value for White in the upper left corner.
----------
White's
stones fence something in the corner.However, Black is able to prove that White's fence is not solid enough.
Please note that the sequence could have been stopped after
, as this move reached a position, of which we already know that it does not contain any value for White.There is nothing of value for White in the upper left corner.
----------
White's
stones fence something in the corner.The foreman came by just in time to put the finishing touches on the White construction.
Thereafter, the points
create value for White.Another area on the board was not so lucky.
Black professionals came by and started dismantling the White fence.
White's newbies quickly stopped their efforts to repair their fence, realising that an already known state of their remaing fence was reached.
After this incident, there is nothing of value for White in the upper left corner.
To what extent the upper left corner contains value for whom has not yet been decided.
----------
From here one, we will see skilled workers at work again.
In the diagram above, the points
are securely fenced in by White's stones.The points
create value for White.Black cannot prevent White from creating value.
With
, White returns to the previous section.White could also do so on her own initiative, but will refrain from doing so to not unnecessarily destroy any value.
----------
A Black stone
inside the fenced region enhances the value of this region for White.The points
and create value for White.The value of
is worth two times the value of .Black is hindered to prevent White from creating value.
White would be able to return to the previous section.
Double fenced regions
Something unexpected happened inside the White fence during the lunch break.
White is unable to dismantle the beginnings of a Black fence inside her own one.
Black also has no success in trying to remove the interfering White fence.
The bosses of both companies agree to treat the two points
, which have obviously been fenced off by both sides, as no man's land. Cloak-and-dagger operations to dismantle the fence of the other company in moonlight will therefore not take place.----------
Somewhere in the foothills of the desert.
White is unable to dismantle the section of Black's fence inside her own one.
Black also has no success in trying to remove the interfering section of White's fence.
The bosses of both companies agree to treat the two points
, which have obviously been fenced off by both sides, as no man's land. and will remain in the area as memorials.----------
To be continued ... (simply scroll down)
Last edited by Cassandra on Sat Oct 23, 2021 10:19 pm, edited 2 times in total.
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)
Igo Hatsuyōron #120 (really solved by KataGo)
-
Cassandra
- Lives in sente
- Posts: 1326
- Joined: Wed Apr 28, 2010 11:33 am
- Rank: German 1 Kyu
- GD Posts: 0
- Has thanked: 14 times
- Been thanked: 153 times
Re: GERMAN interpretation of J89's intended contents
Continued.RobertJasiek wrote:How / with which principles have you derived / created the German version?
Approach #5: Territory! Territoryyyyy!!! Where is Territory? Has anyone seen Territory? Does anyone know where my dog iiiiis???
In the previous posting, it was stated that VALUE can only be created within called-group-by-the-common-people formations that either already contain at least two permanently-for-the-opponent-forbidden-board-point construction elements, none of which is the sole connection between the called-group-by-the-common-people formation's stones along the lines of the board, which are compatible to called-true-eye-by-the-common-people construction elements, or that can be transferred into such a called-group-by-the-common-people formation with at least two permanently-for-the-opponent-forbidden-board-point construction elements, none of which is the sole connection between the called-group-by-the-common-people formation's stones along the lines of the board, which are compatible to called-true-eye-by-the-common-people construction elements, even if the opponent plays first, and which enclosed board intersections are either unoccupied or occupied only by opposing can-be-made-to-disappear-permanently-even-if-the-opponent-plays-first called-group-by-the-common-people construction elements, which are compatible to called-dead-by-the-common-people called-group-by-the-common-people construction elements, which are compatible to called-two-eye-formations-by-some-rule-specialists construction elements, as well as to called-independently-alive-by-the-common-people called-group-by-the-common-people construction elements.
However, we must not forget the stubborn company managers who are unable to apply universal principles across the board, but only realise AFTER carrying out a lost labour of love action that the previous use of their workers was merely a waste of labour and working time.
Two application examples may illustrate this.
----------
Called-snap-back-by-the-common-people formation.
The points
and at the left create value for White.The value of
is worth two times the value of .In total, the value for White is seven times the value of
.The points
at the upper edge create value for Black.In total, the value for Black is two times the value of
.Black's company manager insist on dismantling White's fence at the top permanently, but fails.
In the end, every single part of White's fence is where it originally was.
White's value on the board is five times the value of
.The value in White's prisoner lid is two times the value of
.In total, White's value is seven times the value of
, the same as before Black's desperate deed.----------
Called-nakade-by-the-common-people formation.
The points
and create value for White.The value of
is worth two times the value of .In total, the value for White is 15 times the value of
.Please understand the lower part of the diagram being value for Black.
Black's company manager insist on dismantling White's fence at the top permanently, but fails.
In the end, every single part of White's fence is where it originally was.
White's value on the board is nine times the value of
.The value in White's prisoner lid is nine times the value of
.In total, White's value is 18 times the value of
.Black's value on the board has decreased by three times the value of
.The value in Black's prisoner lid is six times the value of
.In total, Black's value is three times the value of
.White has a surplus value of 15 times the value of
, which equals her value before Black's desperate deed.----------
It should have become evident that
is a compound of one IN the board and one in a layer ABOVE the board.A player's "territory" is a direct property of the board's intersections, and as such equivalent to the set of all
of that player.----------
To be continued ... (simply follow the link)
Last edited by Cassandra on Wed Oct 20, 2021 10:58 pm, edited 4 times in total.
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)
Igo Hatsuyōron #120 (really solved by KataGo)
-
Cassandra
- Lives in sente
- Posts: 1326
- Joined: Wed Apr 28, 2010 11:33 am
- Rank: German 1 Kyu
- GD Posts: 0
- Has thanked: 14 times
- Been thanked: 153 times
Re: GERMAN interpretation of J89's intended contents
Continued.RobertJasiek wrote:How / with which principles have you derived / created the German version?
Approach #6: Ambiguities brought about deliberately, which can only be removed again with a very high level of effort.
「石」
English translation: "stone".
----------
Japanese grammar does not know an explicit plural.
Consequently, in the game of Go 「石」 can mean:
"one stone". "more than one stone".Either individually or firmly connected to each other (aka "chain / string").
"one chain / string" (of stones).
"more than one chain / string" (of stones).Either individually or seen as a sub-component of a larger compound (aka "group").
"one group" (of stones).
"more than one group" (of stones).----------
Use of 「石」 within J89:
Article 2 (Play):
「石」 means
seen individually.Article 3 (Point of play):
「石」 means
.Article 4 (Stones that may exist on the board):
「石」 means
and .Article 5 (Capture):
「石」 means
(and seen individually) and (and seen individually) and .Article 6 (Ko):
「石」 means
.Article 7 (Life and death):
「石」 means
and and .Article 8 (Territory):
「石」 means
and and .Article 9 (End of the game):
「石」 means
and seen individually and and seen individually and and .Article 10 (Determining the result):
「石」 means
and seen individually and and seen individually and and .Article 13 (Both players lose):
「石」 means
and and at least.----------
Article 7's "if capturing" (or whatever translation you prefer) enforces the verification of life & death being done for every single
and at least (you already know that = with a length of 1).One exemplary evidence.
However, when it comes to Article 8's "possess dame" (or whatever translation you prefer), 「石」 gets its main (i.e. most decisive) reference to
.One exemplary evidence.
----------
To be continued ... (simply follow the link)
Last edited by Cassandra on Tue Oct 26, 2021 9:30 am, edited 2 times in total.
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)
Igo Hatsuyōron #120 (really solved by KataGo)
-
John Fairbairn
- Oza
- Posts: 3724
- Joined: Wed Apr 21, 2010 3:09 am
- Has thanked: 20 times
- Been thanked: 4672 times
Re: GERMAN interpretation of J89's intended contents
They are not brought about deliberately (by whom? why?).Approach #6: Ambiguities brought about deliberately, which can only be removed again with a very high level of effort.
Strictly, they are not even ambiguities. And even if they were, they require of a Japanese person a close-to-zero amount of effort. The root (as we all know) is Latin ambiguus, meaning doubtful or uncertain. Since you have been able to identify every reference of ishi (and I bet with minimal effort if you know as much Japanese as you claim), there is no doubt or uncertainty.
It is of course true that the word used in English in each instance may vary. That's just trivial variation between languages, not ambiguity.
Claiming this sort of thing is ambiguity is below the level of children's word play: "When is a door not a door? When it's ajar." Which even has an extra level of linguistic sophistication. A similar joke by the great Chic Murray: "She opened the door in her nightgown. I didn't realise nighties had doors." adds yet another level, of lewdness. Humans cope with both the basic word play and the extra nuances with essentially no effort at all - certainly with no need of an algorithm. To make a pun of my own, it's child's play.
The different references in English (or German) do matter for us, admittedly, and so there is value in re-writing a Japanese text. But let's not pretend it's the fault of the Japanese or the Japanese language. This is what I mean by Japan bashing. It's unedifying.
And Japanese does have plural markers. -tachi, -domo, -ra, -rui and duplication just for starters. They just choose not to use them if they are not needed. As we do sometimes in English (a tank full of fish, NOT a tank full of fishes).
I would suggest an alternative heading: #6. Japanese words that require multiple different references in English.
-
Cassandra
- Lives in sente
- Posts: 1326
- Joined: Wed Apr 28, 2010 11:33 am
- Rank: German 1 Kyu
- GD Posts: 0
- Has thanked: 14 times
- Been thanked: 153 times
Re: GERMAN interpretation of J89's intended contents
I know of a JAPANESE rule text written by a JAPANESE that contains the following formulations, among others:John Fairbairn wrote:I would suggest an alternative heading: #6. Japanese words that require multiple different references in English.
一団の石
a group of stones
一団の白石
a group of White stones
着手をしてはならない
must not make a move
白1子
one White stone
黒の1子
Black's one stone
盤面のすべての石
all the stones on the board
死に石をすべて
all dead stones
黒の一団の石
a Black group of stones
この一団の石
this group of stones
この石
these stones
一団の黒石
a group of Black stones
黒石
Black stones
I do not think that a change in the heading would be justified.
Especially because there are some more ambiguities to come.
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)
Igo Hatsuyōron #120 (really solved by KataGo)
-
CDavis7M
- Lives in sente
- Posts: 716
- Joined: Tue Jun 15, 2010 2:18 pm
- Rank: Shokyu
- GD Posts: 0
- Universal go server handle: CDavis7M
- Has thanked: 109 times
- Been thanked: 140 times
- Contact:
Re: GERMAN interpretation of J89's intended contents
Don't forget this ambiguity on 点.
残り1点（入荷予定あり）
Now where did I put my credit card...
残り1点（入荷予定あり）
Now where did I put my credit card...
-
Cassandra
- Lives in sente
- Posts: 1326
- Joined: Wed Apr 28, 2010 11:33 am
- Rank: German 1 Kyu
- GD Posts: 0
- Has thanked: 14 times
- Been thanked: 153 times
Re: GERMAN interpretation of J89's intended contents
Yes, the username was chosen carefully.John Fairbairn wrote:Cassandra (a prophetess cursed to issue true prophecies never to be believed, as I recall?)
However, it is not at all unusual for people not to want to hear the truth.
This does not require the god of oracles, truth and prophecy, healing and diseases, archery, music and dance, poetry, sunlight, knowledge, herds and flocks, and protection of the young to have cursed the speaker.
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)
Igo Hatsuyōron #120 (really solved by KataGo)