"that archaic stylistic voice that appeals to you has absolutely nothing to do with Heian Japan or the original voice of the author"
I am happy to have been wrong. Of course T1's voice would have appealed to me bc of its familiar tone. But leaving out a whole chapter is criminal, and does a disservice to the author. Faithfulness to the text is one thing but making the prose sound good while trying to convey the same emotional meanings is another. I am looking forward to this new translation.
Well, you're not wrong about the appeal of Waley's voice. I still can't fathom how he was able to do what he did with the tools and information available to him at the time.
But it's important to know the truth in order to make optimal decisions. There would be absolutely nothing wrong with using the Waley translation. It would certainly be a lot easier. But there is more potential upside in doing our own translation.
I didn't stop to consider my own biases before making my decision until a few hours after, with having known nothing about any of the translations besides those selections. I would still like to read Waley though, you're right, there is a huge upside to doing your own translation, and you are adept enough to try.
The Edwardian register is nice regardless of its faithfulness to the original Japanese, but on the other hand leaving out a whole chapter because it's "not essential"?!
This is why I rely on people more knowledgeable about these things. Thank you for the explanation and the translation. I look forward to seeing the final product!
Interesting analysis from the AI's. They line up with my impressions of 1 and 2, which were my favorites for the same reasons the AI gives them their various ranks.
It's going to be interesting to see what kind of possibilities AI translations give in the future. I suspect awful translations by those that don't know what they're doing while providing wonderful ones from those that do; which will be wonderful for those texts that people have never translated into English.
It won't be long before complaining that something was "written by AI" will sound as bizarre as complaining that something was "written by a word processor". The tool doesn't, and cannot, define the performance of the task.
Fascinating peek behind the curtains of the translations. I am really looking forward to the Castalia translation; I love the "Edwardian register" but I also seek a purer translation of the authors original voice. Especially excited that it won't be an "expurgated version."
What a shockingly wonderful surprise to see that the translation most of us preferred is from Castalia 2025! Pretty clear that this is an excellent strategy moving forward. Your East Asian Studies bears fruit yet again, SDL. Thank You for your herculean efforts. This is going to be brilliant.
I'm completely sold on moving forward with the Castalia House translation. It's a brilliant idea, and I'm very excited to see the final product. Based on all of Vox's work to date that I have been fortunate enough to read, I wouldn't trust any one more with this task of creating a faithful, and literarily beautiful translation. I had some trepidation of moving away from the masterpiece which is Waley, but this post today cleared up my misconceptions about Waley and ameliorated all those fears I had yesterday. Thank you.
Fascinating. Love seeing these kinds of comparisons, especially with the AI breakdown. It makes you wonder how an LLM truely gauges something like 'emotional resonance' versus stylistic patterns. There's alot of nuance human readers bring to a text.
"The prose is muscular, image-driven, and emotionally immediate."
I'm sure this is what gave me the YA vibe. It is a very emotionally driven literature. The excerpt didn't fit my idea of a very constrained Japanese culture. But obviously Vox understands Japan better than I do, so I'll trust the process.
Thanks for sharing your process here, it's really interesting to learn more about the different translations. Reading through yesterday's post, I could see my daughter being interested in reading the Castalia translation eventually, but probably none of the others. Well done.
"that archaic stylistic voice that appeals to you has absolutely nothing to do with Heian Japan or the original voice of the author"
I am happy to have been wrong. Of course T1's voice would have appealed to me bc of its familiar tone. But leaving out a whole chapter is criminal, and does a disservice to the author. Faithfulness to the text is one thing but making the prose sound good while trying to convey the same emotional meanings is another. I am looking forward to this new translation.
Well, you're not wrong about the appeal of Waley's voice. I still can't fathom how he was able to do what he did with the tools and information available to him at the time.
But it's important to know the truth in order to make optimal decisions. There would be absolutely nothing wrong with using the Waley translation. It would certainly be a lot easier. But there is more potential upside in doing our own translation.
I didn't stop to consider my own biases before making my decision until a few hours after, with having known nothing about any of the translations besides those selections. I would still like to read Waley though, you're right, there is a huge upside to doing your own translation, and you are adept enough to try.
The Edwardian register is nice regardless of its faithfulness to the original Japanese, but on the other hand leaving out a whole chapter because it's "not essential"?!
Oh dear.
This is why I rely on people more knowledgeable about these things. Thank you for the explanation and the translation. I look forward to seeing the final product!
Interesting analysis from the AI's. They line up with my impressions of 1 and 2, which were my favorites for the same reasons the AI gives them their various ranks.
It's going to be interesting to see what kind of possibilities AI translations give in the future. I suspect awful translations by those that don't know what they're doing while providing wonderful ones from those that do; which will be wonderful for those texts that people have never translated into English.
It won't be long before complaining that something was "written by AI" will sound as bizarre as complaining that something was "written by a word processor". The tool doesn't, and cannot, define the performance of the task.
I glad to hear thar translation 2 is from Castalia! This has an enormous potential for future translations and I'm excited to see where it goes.
Also, thank you for the lesson that the language used is not fully indicative of the quality of the translation.
Fascinating peek behind the curtains of the translations. I am really looking forward to the Castalia translation; I love the "Edwardian register" but I also seek a purer translation of the authors original voice. Especially excited that it won't be an "expurgated version."
What a shockingly wonderful surprise to see that the translation most of us preferred is from Castalia 2025! Pretty clear that this is an excellent strategy moving forward. Your East Asian Studies bears fruit yet again, SDL. Thank You for your herculean efforts. This is going to be brilliant.
I'm completely sold on moving forward with the Castalia House translation. It's a brilliant idea, and I'm very excited to see the final product. Based on all of Vox's work to date that I have been fortunate enough to read, I wouldn't trust any one more with this task of creating a faithful, and literarily beautiful translation. I had some trepidation of moving away from the masterpiece which is Waley, but this post today cleared up my misconceptions about Waley and ameliorated all those fears I had yesterday. Thank you.
T2 was my number two after Royall Tyler. It is good.
Am I correct in understanding that you teach Japanese literature?
If so, would you mind sending me an email? I have a question for you.
Indeed.
Asian Studies professor, Japan specialist.
Emailed vox + sent a direct message via substack.
Sent you an email back. Let me know if it's of interest to you.
Fascinating. Love seeing these kinds of comparisons, especially with the AI breakdown. It makes you wonder how an LLM truely gauges something like 'emotional resonance' versus stylistic patterns. There's alot of nuance human readers bring to a text.
"The prose is muscular, image-driven, and emotionally immediate."
I'm sure this is what gave me the YA vibe. It is a very emotionally driven literature. The excerpt didn't fit my idea of a very constrained Japanese culture. But obviously Vox understands Japan better than I do, so I'll trust the process.
Thanks for sharing your process here, it's really interesting to learn more about the different translations. Reading through yesterday's post, I could see my daughter being interested in reading the Castalia translation eventually, but probably none of the others. Well done.