In thinking about this a little more, my fervent plea is to not rely on AI to answer a question like this. I would be suspect whether it was actually parsing an ancient Greek word or defaulting to modern Greek definitions. In this specific test case, it at least stuck with ancient Greek although in examining the sources references we find:
1. http://www.ldysinger.com/Evagrius/03_Pr…_prayer_all.htm - "Webpage was created for a workshop held at Saint Andrew's Abbey, Valyermo, California in 1997" Item 92 has the word in question: not become agitated but these are all Christian prayers. The main page http://www.ldysinger.com/ shows this is from a Christian seminary workshop.
2. https://sites.temple.edu/dwolf/files/20…-Telos-6.10.pdf - This is more promising as it is actually a paper on Epicurus by a professor of philosophy at Temple University in Philadelphia. The word in question is actually quoted from the letter to Menoikeus. This could be worth reading in full as the author says "I claim that Epicurus does not hold the view that telic pleasure is simply an absence of pain or disturbance." However, if we pass this over in a reliance on AI to simply scrape it for our question, we might miss out on something valuable.
The New Testament question is interesting, but I would suggest caution. I've done this myself, but we also have to remember that word meanings change over time and there are a couple centuries between Epicurus and Paul and the other writers of the New Testament. In thinking about this a little more, it might be more applicable to ask where words in Epicurus show up in the Septuagint, the Greek translation of the Hebrew Bible, since it was composed closer to Epicurus' time. That just occurred to me as I was writing. In any case, the sources "referenced" by the AI are all simply different websites that reference Strong's Concordance entry #5015. That's a fine source, but the AI is pointing to multiple sites with the same information, making it look like it found several different citations when it really only found Strong's.
If you have a question about an ancient Greek word in a text, my suggestion would be to:
- Copy and paste the word into Wiktionary: In this case, the word itself doesn't come up, BUT if you start chopping from the end, you get https://en.wiktionary.org/wiki/%CF%84%CE…%83%CF%83%CF%89
- Just type the word into Google and assess the actual search results: https://www.google.com/search?client=…bih=559&dpr=1.5
In this case, Logeion comes up first and also a site with Strong's Concordance.
Adding in site:edu brings up several academic papers with the word, including ones using the Letter to Menoikeus. You can even use a Google search like (try it, copy and paste this into a Google search bar):
ταράττεσθαι site:edu intext:menoeceus -ai
and that puts the word into context within some academic papers.
I simply don't trust a generative AI using an LLM to provide trustworthy answers in and of itself. Okay, tracking down the actual sites it has scavenged from across the Internet may be helpful, but its extruded text based on word probabilities... I'm not going to trust it to "compose" a text-based answer. I'm going to evaluate WHERE it's getting its words that it smooshed together, worked on probabilities for what tokens come after each other, and strung together what it came up with. And, if that's the case, I'm going to search for the sources themselves from the start rather than put my trust in an automaton mediating access to actual sources.
)
