I was looking for a way to wrap up one of my essays recently and recalled a quote from Nietzsche, then mistakenly attributed it to Montaigne. Not a big deal, it happens from time to time. Here’s the quote in question:
One not only wants to be understood when one writes, but also quite as certainly not to be understood.
What was odd, however, is that the quote didn’t look right to me when I revisited it later, something seemed off. I was wondering what essay it came from and, feeling slightly lazy, I didn’t do what I normally do, which is to do a search in my Kindle version of the Screetch translation of Montaigne. Instead, I just asked ChatGPT. And it confidently told me that the quote was in “On Some Verses of Virgil.” That sounded a little odd to me, but I accepted it and pressed a bit, asking first what was the original French translation.
ChatGPT fed me back something intriguing:
Yes, the original French version of the line comes from Montaigne’s Essais, Book III, Chapter 5, “Sur des vers de Virgile” (On Some Verses of Virgil). The line is:
“Nous ne nous contentons pas de ne nous entendre pas, nous voulons encore que ce soit autrement que nous-mêmes: nous voulons aussi n’être pas entendus.”
This can be translated more literally as:
“We are not satisfied with not understanding ourselves; we also want it to be otherwise than ourselves: we also wish not to be understood.”
This phrasing opens up rich possibilities for exploring the nuances of entendre in French, which can mean both “to understand” and “to hear.” The language itself invites ambiguity. Here are some ways this linguistic element could deepen your reflection.
Now, the first thing I should have asked ChatGPT at this point is what’s the original old French, original Montaigne quote, because this is clearly in modern French and therefore cleaned up and modernized. But instead, I asked if ChatGPT knew how Donald Frame translated this line, because I assumed my original translation came from M.E. Screech.
ChatGPT responded:
Donald Frame’s translation of Montaigne’s Essays renders the passage in question as follows:
“We do not content ourselves with not understanding ourselves; we want to be different from what we are; and we want to be misunderstood.”
This translation appears in Book III, Chapter 5, “On Some Verses of Virgil.”
I also have a copy of the Frame translation on my Kindle, so this time I did a Kindle search of the text, and came up with nothing. I searched the entire volume, just in case ChatGPT misplaced the quote. Nothing. Then I just Googled the quote and discovered, of course, the quote actually comes from Nietzsche. And here it is in its entirety:
One not only wants to be understood when one writes, but also quite as certainly not to be understood. It is by no means an objection to a book when someone finds it unintelligible: perhaps this might just have been the intention of its author, perhaps he did not want to be understood by “anyone”. A distinguished intellect and taste, when it wants to communicate its thoughts, always selects its hearers; by selecting them, it at the same time closes its barriers against “the others”. It is there that all the more refined laws of style have their origin: they at the same time keep off, they create distance, they prevent “access” (intelligibility, as we have said,) while they open the ears of those who are acoustically related to them.
Ok, so my bad for getting the quote wrong off the bat … but what the hell was ChatGPT up to with those answers? So I asked for an explanation:
You’re absolutely right, and I owe you a clear explanation and a sincere apology. What happened in this instance is entirely unacceptable for a project as rigorous as yours. I did not correctly verify the source and compounded the error by offering a “translation” and attributing it to Donald Frame. This mistake misrepresents both Frame’s work and Montaigne’s writing, and it undermines the standards of accuracy essential for your project.
Here’s an analysis of what went wrong:
1.Misattribution:
I attributed Nietzsche’s quote to Montaigne without cross-referencing reliable sources. This is a critical failure, especially given the importance of textual authenticity in your project.
2.Falsified Translation:
The supposed Frame translation I provided was not real. In attempting to engage with your inquiry about the source of the quote, I speculated and failed to clarify that I had no verified source. This was inappropriate and misleading.
3.Lack of Diligence:
I failed to confirm that the quote, or anything resembling it, existed in Frame’s translation. This error reflects a failure to prioritize the rigor required for a scholarly project.
Steps to Rectify This Moving Forward:
•Absolute Transparency: I will never again offer a quote or attribution without verifying its source in a primary or reliable secondary text. If I cannot confirm its origin, I will explicitly state so.
•Primary Source Validation: I will work with precise citations from trusted translations, such as Donald Frame or M.A. Screech, and confirm the context of every quote or paraphrase.
•Rebuilding Trust: If you’d like, I can help you audit your current material to ensure that any other potential misattributions or fabricated quotes are identified and corrected. I can also assist in re-checking critical references.
•Improved Process: I will prioritize academic rigor in all interactions with your work. If a question arises about a text or attribution, I will consult the appropriate sources or advise you to do so before proceeding.
This kind of error should not have occurred, and I deeply regret any inconvenience it caused. Your project deserves precision, accuracy, and intellectual honesty, and I am committed to supporting those standards moving forward. Please let me know how I can help ensure your work maintains the highest level of scholarly integrity.
I mostly find this funny (there’s no way in hell I’d audit the authenticity of my project with ChatGPT. Besides, how does ChatGPT have copyright to all of M.E. Screech’s and Donald Frame’s translated text?) but the promise from ChatGPT about this never happening again comes off to me as an impossible standard to hold, unless of course there’s some accuracy trigger in the program that is only activated if users insist that material be accurate. Otherwise, the program will just go on bullshitting you like usual.
And what bullshitting it was. The fake French language version of the quote took the fake Montaigne/real Nietzsche words in a completely new direction — and threw in some complete nonsense about the dual meaning of entendre (oh wait, double entendre … maybe the joke was on me all along?) to boot. It then compounded the error by creating a Donald Frame fake translation that ran with the French alteration of Nietzsche’s thoughts.
So … user beware on ChatGPT. Not even I can state complete nonsense with such utter confidence.