simplifying paradigms/tables

Greek Learning Greek
1

Hey folks

So, I have to learn Greek for my degree, which is fine, however, I do not have the time to learn hundreds of tables/paradigms.

I’ve learnt the rules of contraction, thus simplifying the verb tables a hell of a lot, but I can’t find any way to simplify the 3rd declension down to less than 7:

νυξ, νυκτος
πραγμα, πραγματος
τριηρης, τριηρους
πολις, πολεως
αστυ, αστεως
βασιλευς, βασιλει
οφρυς, οφρυι

it’s deeply frustrating because they’re quite similar, but have 1,2,3, forms which are different. I’ve managed to get present, future, imperfect, and aorist verbs (both active and middle) down to 3 paradigms, from the 30 that were presented, not including irregular verbs like ειμι and οιδα of course. I really want to do the same for the third declension, but i can’t find any one or two common set of endings. I thought I had it down to 2:

m/f
ς ες
α ες
ος ων
ι σι
(ε)

plus the neuter

but that doesn’t work for πολις or νυξ, because πολε + ς wouldn’t be πολις, πολε + α wouldnt be πολιν, etc

any help would be appreciated

2

The genitive singular of βασιλεύς is βασιλέως. βασιλεῖ is the dative. Again ὀφρύι is the dative of ὀφρῦς, the genitive being ὀφρύος. Anyway, have a look at the tables on this web site:

http://www.danlj.org/eaj/greek/

3

The good news is that they really do reduce down to (almost) one set of endings for masculine/feminine and one for neuter…

… The bad news is, that only works in Proto-Greek. And Homeric Greek, which is close enough for these purposes to Proto-Greek.

The Homeric declension of πόλις is:

πόλις, πόλι-ος, πόλι-ι, (πόλιν)
πόλι-ες, πολί-ων, πόλι-σι, πόλι-ας

That πόλιν… yeah, it’s an exception. Like Old Man Smyth says (§246): Masculines and feminines usually add α to stems ending in a consonant; ν to stems ending in ι or υ (but not ευ).

The alternation between ν and α is explained in Proto-Indo-European as a syllabic -m̥: turns into an -n in Greek after a vowel, turns into a vowel, -α, after a consonant. So there quite isn’t a single set in Proto-Greek, but there is a single set in Proto-Indo-European.

The Attic declension results from some unfortunate hybridisation of this declension with the declension of βασιλεύς.

The Homeric declension of βασιλεύς is:

βασιλεύ-ς, βασιλῆ-ος, βασιλῆ-ι, βασιλῆ-α
βασιλῆ-ες, βασιλή-ων, βασιλεῦ-σι, βασιλῆ-ας

which you can derive, with digamma deletion and compensatory lengthening, from:

βασιλέϝ-ς, βασιλέϝ-ος, βασιλέϝ-ι, βασιλέϝ-α
βασιλέϝ-ες, βασιλέϝ-ων, βασιλέϝ-σι, βασιλέϝ-ας

Attic shortens the eta back to epsilon, but not before its notorious quantitative metathesis, which you will already have seen at play in the Attic declension: ηο > εω, so βασιλῆος > βασιλέως. And of course ει is the pronunciation of εε, once the original diphthong ει became a monophthong.

No problem for νύξ at all, as long as you allow a proto-Greek nominative of *νύκτ-ς. Greek does not allow consonant clusters at the end of words.

… I’ve been daydreaming in recent months on doing some lectures online, explaining noun declensions in terms of proto-Greek.

4

Before you do, a few notes:

  1. Compensatory lengthening means the situation where a consonant loss would cause a long syllable to become short. This is compensated by lengthening the vowel so that the prosodic pattern of the word is preserved. E.g. *esmi (Sanskrit asmi) > Attic εἰμί. That’s not what happens in your example of βασιλεύς.

  2. The declension of -ευ stems isn’t normally reconstructed as you do. Instead a suffixal Ablaut is posited.

  3. Of course ξ is a consonant cluster and it is allowed at the end of words (as is ψ).

5

Greek resists simplifying, even when the rules for formation are well understood. The good news is that once you have worked with the language for a bit, you can usually recognize the forms even if you can’t exactly formulate them on the fly. BIG HINT with nouns: there is a definite article in Greek, and when that article is used with the noun, you can always parse the noun correctly if you have your definite article memorized… :sunglasses:

Just out of curiosity, what degree are you getting that requires Greek?

6

7

And yes, I would do my homework much more attentively if I ever do.

(Looks more attentively at Sihler.)

Osthoff’s Law: long diphthongs shorten before a resonant, so -ēws, -ēwos > -ews, -ēos rather than -ews, -ewos > -ews, -ēos. That’s why I wanted to stick to internal reconstruction: it’s easier. :slight_smile:

(Looks even more attentively at Sihler.)

So either -ēw- as an ablaut variant of hēr-ōws > ἥρως, or an ablaut variant of hēd-us, using an -ew-/-w- suffix. Actually, Sihler’s recent alternative proposal does seem to do lengthening after all, but analogically:

  • ɡʷhone-w-s, ɡʷhone-w-m̥, ɡʷhone-ow-os, ɡʷhone-ew-es
  • ɡʷhone-w-s, ɡʷhone-w-m̥, ɡʷhonēw-s, ɡʷhonēw-es
  • φονεύς, φονέϝα, φονεύς, φονῆϝες
  • φονεύς, φονῆϝα, φονῆϝος, φονῆϝες

That’s meant to be an improvement on what went before. I’ll… have to be persuaded of that.

Well yeah, apart from those %-)