Collaborative translation game / exercise

  1. Participants translate sentence by sentence. (Split if too long.) For the first/base (possibly incomplete) translation, they take turns.
  2. discussions/refinements; or
  3. “ok” counter +1; if everyone is ok, move on to the next sentence
  4. (After a paragraph, another round of "ok"s?)

EDIT: Updated version

This might work better as a game or exercise than as an attempt at making a high quality translation.

For platform, github is an option. (I made a minimal tutorial on how to start using it.)

Specific to Japanese...

Japanese copyright “50 years after the author’s death … 2018 … 70 years … not applied retroactively”

青空文庫 Japanese analogue of Project Gutenberg

Some examples of available works...

竹内浩三 poems あきらめろと云うが, しかられて, 愚の旗
山本周五郎 short story 泥棒と若殿
寺田寅彦 essays 芝刈り, 自画像

I’ve no idea if this would actually work. Any thoughts?

Not sure I understand what’s being proposed.


I’m indeed not sure if I understand the project… It almost sounds like work :stuck_out_tongue:

青空 is very nice, but most of the books are pretty difficult to read for a non-native, being that they are literature or poetry, and usually 70+ years old…

1 Like

I think this sufficient reason to flag, flog, keel haul and ban this person for ever.
Okay, let’s add an icon.
No the other one. :grin:

1 Like

I think given this is a Go forum, maybe just collaboratively translating a Go book might be more enticing.

Also I imagine the wording might be easier to understand, since western usage has adopted a good few japanese terms, and that there are some books in english dedicated to learning enough japanese to read Go books :slight_smile:


Well, there is such a project.
@JethOrensin’s Multilingual Go Book Project exactly does that.
More info on
And sure you can help.

@yebellz’s thread is about this project.

1 Like

At the beginning, assign colors (one for each base translator; base translation is a possibly incomplete or rough translation) to all sentences. This allows working out of order and in parallel.

Editting of base translation can be participated by other base translators or even beginners of the source language as long as they know target language.

There needs a way to approve or say “enough” to a sentence.

But I’m afraid @Vsotvep is right.

From 趙治勲 in Wikipedia:

  • 「入門者・初心者は、最初は少ない路数の碁盤で学ぼう」という考えから、「よんろの碁」が登場するかなり前の、1994年に発行された『発想をかえる 囲碁とっておき上達法』や『はじめて打つ碁―誰でも楽しく碁が打てるようになる』では、他棋士があまり扱わない3路盤・4路盤・5路盤を詳しく解説している。

Those two books seem interesting to translate. He’s still alive though I wonder if we’d get into trouble.

I’m still confused what you are talking about. What books are you referring to? Are you talking about translating from Japanese to English? Or some other languages? How is this game? Why are you assigning colors? What’s the purpose? Learning? Creating the translated product? Friendly competition?


I only know English and Japanese but for other people it can be in other languages.

I don’t exactly know what the purpose should be or if it would actually work. It might be too tedius like @Vsotvep imagined, or too difficult to coordinate since things can get very difficult or near impossible to explain especially when it’s a person’s native language, in which case all the person can say might be “it sounds better for me”.

The two books are 『発想をかえる 囲碁とっておき上達法』and 『はじめて打つ碁―誰でも楽しく碁が打てるようになる』(both 1994). According to the description I quoted, they teach beginners with 3 x 3, 4 x 4, and 5 x 5 boards, which seems unique.

There might be a more flexible way than assigning colors that allows for people coming and going. Coloring or equivalent might be able to take into account each base translator’s level of source language.

Possibly, if the discussions are detailed enough and a person understands one of the languages, spectating the discussions might be an interesting way to learn a language.

1 Like

Oh no, it’s not tedious. I should really try translating, to boost my Japanese. If only I weren’t so lazy…
What I meant with it resembling “work” is that I don’t see the purpose.

The big question we have is:


  • Is it just for fun?
  • Is it a game (if so how is it a game)?
  • Is it to learn a language?

I would be interested in translating Japanese go content, but it would be nice if it’s something that (a) has not been translated, (b) is modern enough to be relevant nowadays © would bring new information to the English language that was previously not available and (d) preferably is open source, so that the translation can be distributed.


Since I was mentioned (thank you @Atorrante :slight_smile: ) and I am working on something similar, I want to say that this actually works pretty well. Assuming that the amount of people participating does not exceed 12-15 people (which might be too many to keep up with and would make the chosen colours become practically indistinguisable), then a google doc where you keep the original text in black letters and all the proposed translations are written underneath the original in various colours, works great.
Any discussions or proposals about differences in translations, or meaning, can be carried out by making comments on those particular parts of the text and the people involved can talk about any issue they might be having with it.

This is actually a very valid opinion though, for a native speaker. Since I have no knowledge of any of the languages currently under translation in the Multilingual Go project, it is a standard by now that if the translators feel that the underlying English text is not written in the same way as it would have been written in their language and a direct translation would have resulted in sentences that would feel “unnatural” to a native reader, then the translators change the sentence in what feel natural for them. If they like it better if a sentence is changed, then I am happy with it as well. They know best. :slight_smile:

A translation is meant to be made for the people that will read the text. So, I think that this automatically means that we should be taking into account what the readers would feel more natural to read and we should not focus on what could be more convenient for us the translators.

Good luck with your endeavor! :slight_smile:


I also wonder which of these purposes is the goal. If the language was Japanese, and the barrier to entry was set as “give it your best shot; no better way to learn than doing”, then I would participate. I think an entire book is biting off too much, though. I think that would end up being 0-3 people working on it alone very shortly.


I mostly agree with this, though I think there are exceptions and edge cases, such as translating for pedagogical purposes, or translating for an audience such as anime viewers with subs or Christians with the Bible who value a bit more accuracy in exchange for slightly unnatural phrasing and translator notes.


Sorry for the wrong use of the word. I was reading Winesburg, Ohio recently and realized I have problem distinguishing synonyms in English.

Maybe the purpose can vary project to project. “Game” may have been the wrong word.

(a) Shūgorō Yamamoto (-1967) is well-known in Japan but I was surprised to find that only one of his works–not even one of his major ones–seems to have been translated into English (Goodreads).

(b) I’m not confident enough to judge the relevance of a work. Or do you rather mean the language is close to how it’s currently used?

© I’m not sure what this can be. Historical or anthropological information that’s been buried? Probably not fiction?

(d) I guess that’s safer.

As for quality, since it can be edited forever, we don’t have to worry about it. Or am I missing something?

One advantage of assigning sentences beforehand that I’d like to highlight is that everyone gets the chance to translate from scratch and have their work corrected, especially if someone figures out a way to assign according to each person’s level (unassigning a tough sentence from oneself might work as a coarse approximation).

1 Like

From a learning perspective, I think the most interesting thing would be a fairly modern work of practical prose, on a somewhat Japanese subject like:

  • pottery
  • smithing
  • sailing
  • fishing
  • cooking
  • calligraphy



And limited in length; we don’t have that large a community here, so all participants would be committing to a large portion of whatever was tackled.

1 Like

To estimate difficulty of translating a sentence:

  1. Max (?) of each word’s frequency bracket, specific or adapted to the field–e.g., “劫(こう)” is common in Go but uncommon in general.
  2. Grammatical structure.
  3. (Difficulties of surrounding sentences.)

  • Is this in the right direction?
  • Any more factors?
  • Any thoughts on how each factor can be implemented? Or an existing library in some programming language?

Surrounding the nasty sentences seems like a good approach for go players in a translation exercise :stuck_out_tongue_winking_eye: