Some more play with ChatGPT

This evening, my wife and I were playing with ChatGPT to see how it translates from Japanese and Chinese. I remembered that I had once translated a passage about rhetoric from Tseng Yueh-nung, the first president of Tunghai University. It's written in a more literary style (although it is fairly understandable if you read modern Chinese), so I wanted to test ChatGPT on it. Here are the results:

Me: Translate: 希臘羅馬政治即臻全盛，日中則昃。於是道德淪亡，人懷自競之心，挾數任術，以獵取功名而致富貴。雄辯學 (rhetoric)，逐為必修之科。整體教育反失去重心，一時消極與失望，詭詐與荒瑩，籠罩全歐，不可救藥，如是者數百年。

ChatGPT: During the heyday of Greek and Roman politics, the situation in China and Japan was declining. As a result, morality decayed, and people harbored a sense of rivalry, employing any number of schemes to hunt for fame, wealth, and status. The study of rhetoric became a mandatory subject. Overall, education lost its focus, leading to passivity and despair, deceit and confusion, enveloping all of Europe for hundreds of years, with no hope of recovery.

Me (somewhat rudely): 日中則昃 doesn't have anything to do with Japan and China. It's talking about the sun at its zenith declining.

ChatGPT: You are correct, I apologize for the mistake. "日中則昃" is a Chinese idiom that means "the sun is at its zenith and then begins to decline," and it has nothing to do with Japan and China as I previously stated. Here's the corrected translation:

At the height of Greek and Roman politics, the sun was at its zenith and then began to decline, just as the morality of society began to decline. People became competitive, using various means to pursue fame and fortune. The study of eloquence and rhetoric became a required subject. Overall, education lost its focus, leading to a period of passivity and despair, deceit and confusion, that enveloped all of Europe for several hundred years, and seemed irreparable.

I think it did a decent job on the revised translation. It makes the comparison between the sun's decline and the decline of morality in society more obvious with "just as," which is not (I don't think) in the Chinese original (於是 is more like "therefore" or "as a result," not "just as"). So it's taking an implicit comparison and making it explicit. 

We tried ChatGPT with 陋室銘 by 劉禹錫. My wife got a similar kind of situation with the last line, "孔子云：「何陋之有？」", which it translated as something like "What does it matter if it's humble as long as I am at peace?" I generated 10 translations but never got that version. Here's what I got for the last sentence:

1.  Confucius said, "What does it matter if the room is humble?"
2. Confucius said: "What does it matter that the room is simple?"
3. Confucius said, "What does it matter if it's humble?"
4. Confucius said, "How humble it is, yet how significant!" [Hmmm... this one is interesting...]
5. Confucius said, "What humble abodes can also have admirable qualities?" [Hmmm...]
6. Confucius said, "How can a humble abode be considered inferior?"
7. Confucius said, "What does being shabby matter?"
8. Confucius said, "What is humble can also be distinguished."
9. Confucius said, "What does it matter that it is humble?"
10. Confucius said, "What is humble about it?" [I think ChatGPT was getting tired here...]

I looked for some English translations online. This one by Abby Lu notes that "Confucius actually said: ‘What is ugly about the home of a gentleman?’, the latter of which the author cut out of humility." (子曰：君子居之，何陋之有？). See also this translation (an interesting website, by the way!). 

Anyway, I'm getting off-track now, and I should go to bed so I can get back to my book review tomorrow. My book review is coming along well, by the way (thanks for asking); I think I'll be able to get a good draft of it done by Friday. 

ChatGPT's suggestions about my list of comparative rhetoric readings

Just a footnote the previous post: after I finished it, I plugged the comparative rhetoric readings into ChatGPT to get its response to how to organize those readings. It didn't complain as much about the diversity of topics, but it just spat out the list in a different order. I asked it how it decided on the order, and it gave me the following response:

I organized the list of articles and books in a way that provides a coherent flow of ideas and themes. I began by grouping together articles that have similar themes or authors, and then arranged them in a way that makes sense chronologically or thematically.

I started with some articles that introduce the concept of comparative rhetoric and translingual approaches to meaning-making, such as Cushman's "Translingual and Decolonial Approaches to Meaning Making" and Cousins' "Self-reflexivity and the Labor of Translation." Then, I moved on to articles that focus on decolonization and pluriversal rhetorics, such as Cushman et al.'s "Decolonizing Projects: Creating Pluriversal Possibilities in Rhetoric" and Johnson and Endres' "Decolonizing Settler Public Address."

Next, I included articles that examine non-Western rhetorical traditions, such as Lipson and Binkley's "Ancient Non-Greek Rhetorics," Lu's "Rhetoric of the Chinese Cultural Revolution" and "The Rhetoric of Mao Zedong," and Tomasi's "Rhetoric in Modern Japan." I also included articles that discuss the development of comparative rhetoric and its future, such as Mao's "Beyond Bias, Binary, and Border" and "Thinking beyond Aristotle" and Wang's "Comparative Rhetoric, Postcolonial Studies, and Transnational Feminisms."

Finally, I included articles that explore other related themes such as posthumanism, hybridity, and materialist approaches to activism. The order of the articles and books is not set in stone, and it could be adjusted based on the reader's interests and needs.

This is interesting to me for two reasons: how it says it decided how to organize the sources (sounds pretty reasonable) and the fact that it basically appears to be reflecting on its process. For the fun of it, I asked it what it learned from the process, and it said,

As an AI language model, I do not learn in the same way that humans do. However, I can tell you that organizing sources in a systematic and coherent way helps to make information more accessible and easier to understand. By categorizing sources and grouping them together, it becomes easier to identify patterns, compare and contrast different viewpoints, and draw more informed conclusions. Additionally, organizing sources can help to identify gaps in knowledge or areas where further research may be needed. Overall, a well-organized set of sources can facilitate the research process and ultimately lead to more robust and accurate conclusions.

Well, that's sort of a relief... 

Books to the ceiling, books to the sky...

I've quoted this poem a couple of times already. One of the many things that I have trouble with is deciding what book to read next. Sometimes I just grab something off the shelf that looks interesting. Sometimes I choose something that I've had for a very long time because I feel guilty that I've neglected it (hi, John Shepherd's Statecraft and Political Economy on the Taiwan Frontier, 1600-1800!). Sometimes I choose something written in Chinese because I feel guilty that I haven't read a book in Chinese for a long time (hi, Yang Tsui's (楊翠) 《永不放棄：楊逵的抵抗、勞動與寫作》!). Sometimes I choose a biography or a work of fiction because I find them easier to read than a more scholarly or theoretical work, and sometimes I choose a more scholarly or theoretical work because I feel guilty that I'm making life too easy for myself by reading something that isn't torture to get through. (My Calvinist feelings of total depravity even motivate my reading choices, it seems.)

But I've got a long list of readings that I've given myself for that proposed semester leave that I mentioned (no word on that yet), and I'm trying to decide the order of the readings. One way might be just to jump from one book to another than seems relevant based on what I learn from the first book. Another might be to look through the books to find themes that connect them, sort of like writing a syllabus for myself. I haven't found any good answers on the internet yet (probably because I'm not using the right search terms). Maybe I should ask ChatGPT?

OK, here's what I asked the old Chatter:

I have a lot of books that I need to read for a research project. They're on a variety of topics, some more closely related to each other than others. How should I decide the order to read them in?

And they said,

When deciding the order to read your books for a research project, it's important to keep in mind your goals and priorities for the project. Here are some steps you can take to decide the order: 

1. Identify the most important books: Take a look at all the books you have and identify the ones that are most important to your research. These may be the ones that are most closely related to your topic or the ones that are most highly recommended. 

2. Create categories: Once you've identified the most important books, try to categorize them based on their topic or theme. This can help you to see how the books relate to each other and which ones should be read together.

3. Prioritize: Decide which categories of books are most important for your research and prioritize them. This can help you to focus your reading and ensure that you're getting the most important information first.

4. Balance: As you decide the order in which to read your books, try to balance the categories so that you're not reading too many on one topic before moving on to another. This can help you to stay engaged and avoid burnout.

5. Make a plan: Once you've decided on the order in which to read your books, create a reading plan. This can help you to stay on track and ensure that you're making progress towards your research goals. You can also adjust your plan as needed based on your progress and any new information you learn.

Overall, by identifying the most important books, categorizing them, prioritizing, balancing, and making a plan, you can create an effective reading order that helps you achieve your research goals. 

Hmmm... all in all, not bad advice. (Wonder where they're getting it from?) It sounds basically like what I said about writing a syllabus for myself, though. (With the possible exception of the point about "balance".)

I'm going to try something more specific, though. I'm going to feed my reading list into ChatGPT and see what it suggests about that... Wait a second...

Well, it didn't know quite how to answer that: "Wow, that's quite a list of books! Since the topics of the books are quite diverse, it's difficult to suggest a specific order for you to read them in." It did give suggestions about how to read some of the books, though. Back to the drawing board...

Practice with Chat GPT