In the Cambridge lexicon we find:
ἑλέ-πολις , ep. ἑλέπτολις , ι , gen. εως adj.
I’m not sure what that iota refers to. I would think it would be the neuter nominative ending but those forms do not exist. I was also thinking that perhaps that it might be trying to distinguish nouns which end in ις but have the accusative in ιν versus those that have it in α but that theory does not work.
ἄ-πολις ι , gen. ιδος (Ion. ιος ) adj.
δικαιόπολις ι , gen. ιος dial.adj.
ἄ-χαρις ι , gen. ιτος adj.
δυσ-γάργαλις ι , gen. ιδος adj.
δύσ-ελπις ι , gen. ιδος adj.
δύσ-ερις (also δύσηρις Pi.) ι , gen. ιδος adj.
δυωδεκά-πολις ι , gen. ιος dial.adj.
ἑλέ-πολις , ep. ἑλέπτολις , ι , gen. εως adj.
ἔμ-πολις ι adj. only acc. ἔμπολιν
ἐπί-χαρις ι , gen. ιτος adj. acc. ἐπίχαριν
It’s the neuter ending of the adjective.
1 Like
ok, that makes sense. The problem was that I wasn’t doing enough research. So here is a list of all the neuter adjectives that end in ι and their non-neuter counterparts with their frequency. There are only about 30 of these words. As you can see the neuter ending in iota is attested. Sometimes neither the non-neuter nor the neuter are attested but that’s because the Cambridge Lexicon quotes 9 authors which my corpus does not have.
αθεμις 0
αθεμι 1
ακιθαρις 0
ακιθαρι 0
αμαντις 3
αμαντι 0
αχαρις 48
αχαρι 118
δικαιοπολις 7
δικαιοπολι 6
δυσγαργαλις 1
δυσγαργαλι 0
δυσελπις 26
δυσελπι 8
δυσερις 36
δυσερι 11
δυσηρις 3
δυσηρι 0
δυωδεκαπολις 0
δυωδεκαπολι 0
ελεπολις 6
ελεπολι 0
ελεπτολις 1
ελεπτολι 0
εμπολις 0
εμπολι 1
επιχαρις 20
επιχαρι 25
ερυσιπτολις 1
ερυσιπτολι 5
ευελπις 50
ευελπι 20
ευχαρις 18
ευχαρι 23
πολυδηρις 0
πολυδηρι 0
τροφις 0
τροφι 17
φιλοπατρις 15
φιλοπατρι 3
φιλοπολις 45
φιλοπολι 1
φιλοπτολις 1
φιλοπτολι 0
Have you noticed that almost every adjective on your list is a compound? Originally PIE had only two genders: animate and inanimate. For some reason in Greek compound adjective retained those two genders.
Νο, Ι διδν᾽τ excuse me, I didn’t notice that, thanks for pointing that out. I did a little chatgpt research on 3 genders. Here’s what it said: * Proto-Indo-European (~4500–2500 BCE)
PIE had at least three grammatical genders:
- Masculine
- Feminine
- Neuter
These were originally linked to animacy and natural classes, not sex alone.
- Proto-Germanic (~500 BCE)
As PIE evolved into Proto-Germanic, it retained the three genders. The system was used to inflect adjectives, pronouns, and articles.
Ok, so does the ancestor of Greek have three genders? Here’s what it says about Mycanean:
- Mycenaean Greek (used ca. 1400–1100 BCE) is preserved in the Linear B script, a syllabic writing system used mainly for administrative records.
- While the script is limited and does not represent all phonological distinctions clearly, scholars have reconstructed forms that show agreement in gender, just as in later Greek.
- Noun endings and adjective agreement in surviving tablets show masculine, feminine, and neuter patterns consistent with those of Classical Greek.
So can you point to me something that shows Greek had a two gender system?
What’s the point in asking ChatGPT when all you get is a confident and plausible-sounding answer without sources that might as well be totally incorrect, for all you know?
A better starting point for actual research is the Wikipedia article on Proto-Indo-European nominals. The section on gender mentions Greek two-termination adjectives as one of the reasons for positing an early stage of PIE with no feminine, and cites a couple of sources.
I don’t think anyone’s claimed that Greek itself has had a two-gender system during its documented existence, only that it retained traces of the earlier two-gender system in PIE.
1 Like
You don’t need to be defensive. I was merely asking you where you got your info. If someone asks you to cite your sources then that is no reason to become upset. In my experience Chatgpt is far more reliable than people I’m unfamiliar with posting stuff on the internet. In any case, thank you for teaching me something new. I did not know that there was a two class gender and now I do and it thanks to people like you that I have this info.
I’m sorry, it’s just that I’ve been seeing too much lazy use of ChatGPT online and in real life lately. In my work as a librarian, I get people asking for books that don’t exist (hallucinated by AI) and I see forum threads (not saying this one!) fill up with non-sequitur AI answers when people are looking for human judgment on difficult issues.
There are ways to use ChatGPT to get more reliable answers, but I wouldn’t call it research – at best a starting point to get ideas for what to look for. Either way, the whole question-answer dynamic of “This is what ChatGPT says, can you prove it wrong?” in a discussion gets tedious quite fast, since a priori I have no reason to assume that the AI info is better than nonsense. Although admittedly in this case I don’t see anything absurd in its claims!
Please avoid posting ChatGPT content on the site. It trains on textkit.com data like everything else, and posting hallucinations here poisons the well.
2 Likes
You can’t trust ChatGPT. I tried using it (with an Ancient Greek Addon made by hoomans) just under three years ago to try to verify the first three unit’s exercises from the Hansen and Quinn : Intensive Course (as there was nowhere else and noone else for me to do this with). I did the exercises and painstakingly typed in all my answers but then spent more time correcting the mistakes it made then actually doing the exercises! It frequently made mistakes with declensions, conjugations and even made spelling mistakes! The only thing that came from it is that at the end of the process I was absolutely certain that my answers were 100% correct.
I then tried using it as a glorified search engine. I asked it to track down certain words and list containing passages etc. After an extensive process of tracking down the passages (in greek rather than translations) I found that a lot of the passages it cited never actually contained the word! It tells you it doesn’t have access to this and that but when you give it some token to show that you do have such access, yourself, suddenly it ‘knows everything’ about that subject, with complete confidence and ‘authority’.
I tried it on a coding task. It trolled me for hours by giving me standard object oriented methods in a non-OO specialised language. ChatGPT cant even tell you how many characters are in a sentence.
I tried Deepseek. It randomly dropped some chinese characters in the middle of an english sentence. When I made a joke with it about its Chinese Heritage showing through it proceeded to tell me it wasn’t Chinese because it was Claude AI written and trained by Anthropic!
You can, at best, use AI to point you in a direction but anything meaningful will always come from your own research and hard work.
For websites, use robots.txt file to restrict AI from accessing or training on your site.
It’s better to be on Textkit where everyone can tell you you’re wrong instead of with AI trolling you that you’re completely right, now you’re thinking like a true expert.
Just my two-cents on the subject. I ended up making myself a number of tools to make my life easier. I wrote a Polytonic Greek Keyboard for my PC and made a Wordpress plugin based on the Logeion LSJ which gives me a pop-up definition of any greek word on-page, including inflections and morphology. These two things mean I can complete tasks in a fraction of the time and with a fraction of the browser-tabs. I also trained a Tesseract OCR to digitise myself a copy of the H&Q book so I can copy and paste from it’s pages. It’s about 85% accurate.
I also created a Music Decoder which takes any greek words as input and allows you to play them in music within a sandbox based on theory by different authors, like Pythag etc. It’s just to investigate a theory I have about the definition of Psalms, the missing music in Ancient greek literature and why there are no vowels in Hebrew. But that’s a different discussion entirely.
I’ll stop rambling now.