Just how big is the 3rd declension?

Greek Learning Greek
1

I’m relearning all the Greek we were death-marched through in college because I still can’t read it in any meaningful sense. Now I’m taking a slower and much more methodical approach, including really getting my paradigms down solid and understanding what sound shifts/contractions/etc. explain why a particular paradigm exists.

This approach worked fine - until the third declension appeared. I then noticed that my new textbook divides it up differently than my original one did… as did another textbook, and another. The problem for me is that I can’t tell just how “big” the whole declension is and how many paradigms really need to be memorized to master it. The Greek Paradigm Handbook (Geannakis et al.) had the most discrete paradigms explicitly provided:

  1. Labial stems (κλόψ, κλοπός)
  2. Palatal stems (φύλαξ, φύλακος)
  3. Feminine δ- stems (πατρίς, πατρίδος)
  4. Masculine τ- stems (ἄρχων, ἄρχοντος)
  5. Neuter τ- stems (σῶμα, σώματος)
  6. Masculine θ- stems (ὄρνις, ὄρνιθος) - but I thought ὄρνις could also be feminine?
  7. Nasal stems (δαίμων, δαίμονος)
  8. Liquid stems (ῥήτωρ, ῥήτορος)
  9. ερ- stems (μήτηρ, μητρός)
  10. ρ- stems (ἀνήρ, ἀνδρός)
  11. Masc/Fem σ- stems (τριήρης, τριήρους)
  12. Neuter σ- stems type A (μέρος, μέρους)
  13. Neuter σ- stems type B (γέρας, γέρως)
  14. ι- stems (πόλις, πόλεως)
  15. υ- stems (ἄστυ, ἄςτεως)
  16. ευ- stems (βασιλεύς, βασιλέως)
  17. αυ- stems (ναῦς, νεώς)
  18. ου- stems (βοῦς, βοός)

Are there other paradigms in the declension that are somehow missing? And are there any truly, hopelessly irregular nouns that should be on the list? I never got a chance to learn this properly the first time around, so any help in getting the fullest picture of it this time is appreciated - χάριν ἔχω!

2

Hopeless irregularities:

http://www.perseus.tufts.edu/hopper/text?doc=Perseus%3Atext%3A1999.04.0007%3Asmythp%3D285

You really don’t need to memorize the entire paradigms–in most cases just learn the nominative and genitive singular. There are some irregularities in the dative plural, but you will pick these up over time. You should probably pay most attention to ##9 and following–these do have some forms that need to be noted.

You might run through the sections of Smyth that are adjacent to 285, above. In fact, if you’re really serious about mastering Greek, you might want to buy a copy of Smyth–get the 1956 Harvard University Press edition edited by Gordon Messing, not one of the cheaper ones. It will last you a lifetime.

Smyth will give you some of the information about how the various forms evolved, but to fully understand all (or most) of them, you would need a book on the prehistory of Greek, such as Sihler, or in French, Lejeune and Chantraine; in German, Rix. For ##9 and 10 on your list you would need to understand the Indo-European accentual system. It’s really not worth your time to dive that deep at this point, when you’re simply trying to get the forms down.

3

Most of the forms can be explained by just a few phonetic processes:

the disappearance of sigma and digamma (the “w” sound represented by F in archaic Greek) (Smyth 118-123): http://www.perseus.tufts.edu/hopper/text?doc=Perseus%3Atext%3A1999.04.0007%3Asmythp%3D118

vowel contractions (Smyth 59): http://www.perseus.tufts.edu/hopper/text?doc=Perseus%3Atext%3A1999.04.0007%3Asmythp%3D59

quantitative metathesis (## 14-17; Smyth 34): http://www.perseus.tufts.edu/hopper/text?doc=Perseus%3Atext%3A1999.04.0007%3Asmythp%3D34

the spelling of πς as ψ and κς as ξ (##1 and 2)

the disappearance of dental consonants before sigma (##3 and 6)

compensatory lengthening (##7 and 8 )

If you’re aware of these rules, you really don’t need to waste a lot of time memorizing the entire paradigms.

But if your aim is to learn Greek, brace yourself for lots of hopeless irregularities.

4

Here’s another infuriating third-declension irregularity: θρίξ, τριχ-ός (hair). The explanation is in Smyth 125:

http://www.perseus.tufts.edu/hopper/text?doc=Perseus%3Atext%3A1999.04.0007%3Asmythp%3D125

5

I can’t add to Qimmik’s list of extra forms but I would recommend this site
http://www.lexigram.gr/lex/arch/
As it automatically declines nouns for you, it is good for checking which variant a specific noun belongs to.

I would add that when I get stuck it is far more likely to be because of an unusual verbal form than an odd noun form. Hence attempting to learn all the third declension forms now is probably not a good strategy. Learn the forms you encounter reasonably often and learn the less frequent forms when you are reading fast enough to encounter them reasonably often.

6

Most (but not all) of the “paradigms” are valid only for a limited subset of nouns. Learn to recognize the basic 3rd declension endings and then learn the nominative and genitive of individual words as you encounter them. That will simplify the process of mastering the 3rd declension.

#5 covers a very large class of verbal nouns ending in -μα. Once you have the endings down, all you need to know about these is that the stem ends in -ματ-, but in the nominative and accusative singular, there is no ending, so -τ, a consonant which generally can’t end a word in Greek, drops off. The same principle is operative in γάλα, γάλακτ-ος: -κτ cannot end a word in Greek, so it drops off in the nominative and accusative singular.

7

Thank you for the responses. Greek is tough enough without having to delve into Proto-Indo-European linguistics and the like; it’s enough for me to know that the τριήρης, τριήρους forms result from an invisible σ dropping out. Third declension endings actually strike me as pretty regular - it’s all the dropped/combined/contracted letters that make it tricky, especially once you get into the genitive forms that suddenly have omegas pop up. Dative plurals are an exception… but it would not be Greek unless there were exceptions somewhere.

I also appreciate the links to Perseus and the recommendation for Smyth. I’ve got a couple grammars (the meatiest of which are Godwin and Crosby), but I couldn’t get their descriptions of the declension to square with all the others.

Qimmik, I think your second response really gets at what I’m trying to understand, so I’ll take a close look at those links. And daivid, I think your link will be a big help once I’m trying to read again and can’t quite figure out just what form I’m looking at. If nothing else, at least I will know which part of speech I’m not looking at. :slight_smile:

8

This is one I stumbled on today:
οἶς sheep.

When I see it declined it is clearly behaving like πολις but the genitive singular is οἰός.
I guess it isn’t surprising that οἰ+ός should stay that way even though ι+ος ends up as εως.
It is sensible advice to learn the ways that the few phonetic processes that cause all these
surprises but those processes are hard to learn without first learning the declensions.
The processes on their own are unfortunately too abstract to sink in for me.

9

I think οἶς can be explained by the fact that it originally had digamma, ὄϜις. Then there are some irregularities in Attic that are understandable if you know Homeric Greek. Then there those that comparative linguists disagree about, I think the rule that two aspirated consonants can’t follow each other in consecutive syllables is one of those.

Anyway, I think it’s enough to learn the basic patterns first. Then you’ll be able to guess what a form is when encounter it in a text. Come back to them later on, if you’re interested in the history of language.

10

Well, you’ve just mentioned my favorite hopelessly irregular 3rd declension noun, χάρις, χάριτος, where you would expect the accusative singular to be χάριτα (and indeed this regular form irregularly occurs,) but the regular form is the irregular χάριν.

Just how big is the 3rd declension?

How big is it? It’s so big, as Johnny Carson might say, that I once heard one Greek professor say to another, “Hey, is that a list of all the hopelessly irregular Greek third declension verbs in your pocket, or are you just happy to see me?”

χάριν ἔχω!

καὶ χάριν σοι, φίλε, ἀποδίδωμι δή!

11

Markos, a number of nouns in -ις that are not properly -ι stems, like χάρις, χάριτος, have alternative accusatives in -ιν, for example, ὄρνις, by analogy with -ι stems like πόλις and οἶς.

http://www.perseus.tufts.edu/hopper/text?doc=Perseus%3Atext%3A1999.04.0057%3Aentry%3Do)%2Frnis

οἶς, οἰός is relatively regular, as these things go. It’s πόλις that requires elaborate explanation, which is beyond my comprehension, except to note that πόλεως is from πόληος, not πόλιος, by the process known as “quantitative metathesis” – the change of long η to short ε, coupled with the change of short ο to long ω (i.e., the e vowel and the o vowel “exchange” quantity). Don’t ask me where πόληος came from because I have no idea. There’s an explanation of this in Sihler, New Comparative Grammar of Greek and Latin (New York and Oxford, 1995), sec. 304 (p. 313), which goes into proto-Indo-European, if you’re really curious, but it seems largely based on conjecture about pre-Greek forms.

Unfortunately, other nouns in -ις (but not all of them) have found the irregularities of πόλις attractive, including a large class of verbal nouns in -σις, such as ποίησις, πρᾶξις (κ+σ>ξ), etc., so you need to learn this paradigm, or at least be able to recognize the forms when you encounter them.

12

Just to drive everyone crazy, here’s LSJ’s list of alternative forms for πόλις found in various texts. Note that three alternatives for the genitive singular (four if you count πτ-) are found in the Iliad alone:

Ep. forms, acc. sg. “πόληα” Hes.Sc.105, Call.Aet.Oxy.2080.62; gen. πόλιος or πτόλιος, Il.2.811, 4.514, al., “πόληος” 16.395, al. (also Thgn.757), “πόλεος” Il.21.567; dat. “πόλει” 5.686, al., “πτόλεϊ” 17.152, “πόληϊ” 3.50 (also Tyrt.12.15): pl., nom. “πόλιες” Od.15.412, “πόληες” Il.4.45; gen. “πολίων” 1.125, al.; dat. “πολίεσσι” Od.21.252; acc. “πόλεις” Il.2.648, al., “πόληας” Od.17.486, Call.Fr.9.70 P.(scanned ^ ^ ^_ IG12.826), πόλιας (disyll.) Od.8.560,574, (trisyll.) Il.4.308 (s.v.l., “πόλεας” Aristarch.): Ion. forms, gen. “πόλεως” IG12(8).356（Thasos), GDI5653a13 (Chios), etc., also Xenoph.2.9,22, v.l. in Thgn.1043; written “πόλειως” GDI5532.19 (Zeleia); πόλεος ib.5339.41 (Orop.), IG12(7).103 (Amorgos), Thgn. 776, etc., “πόλιος” Hdt.1.26, al., Herod.2.8, al., “πόληος” Thgn. (v. supr.), cj. in Hippon.47, cf. An.Ox.1.361; dat. mostly πόλει, but “πόλι_” Hdt. 2.60, al., πόληϊ (or -ῃ) SIG169.3 (lasos, iv B.C.): pl., usu. πόλεις, πόλεων, πόλεσι, but in Hdt. “πόλιες” 1.142, al., πολίων ib.6, al., πόλισι ib. 151, al.; acc. “πόλι_ς” 2.177,al., “πόλιας” 1.142, 2.102, al.: Dor. gen. sg. “πόλιος” SIG615.3 (Delph., ii B.C.); dat. sg. “πόλι” IG4.839 (Calaurea, iv B.C.); dat. pl. “πολίεσι” Pi.P.7.8; πολίεσσι Foed.Lac. ap. Th.5.77 (v.l. πολίεσι), 79, IG42(1).74.4 (Epid., iii B.C.); Cret. “πόλιθι” GDI5019.3: Aeol. gen. “πόλιος” IG12(2).526a8 (πόληος is an Epicism in Alc. Supp.17.6); gen. pl. “πολίων” IG11(4).1064b20; dat. pl. πολίεσσι ib. 12(2).1.6: Trag., gen. πόλεως disyll. (as also in Com., exc. Ar.Eq. 763), thrice “πόλεος” A.Ag.1167 (lyr.), S.Ant.162, E.Or.897:—Att. Inscrr. earlier than 350 B.C. sts. have dat. sg. πόλῃ, IG12.108.35,22.17.10,42.5, 53.7; Att. dual “πόλη” Isoc.8.116, πόλη or “πόλει” Aeschin. Socr.8 (where Choerob. cites both forms, in Theod.1.314, 136 H.); gen. “τοῖν πολέοιν” Isoc.4.73: Elean nom. sg. “πόλερ” Schwyzer425.16; gen. πόλιορ ib.20 (iii/ii B.C.)

http://www.perseus.tufts.edu/hopper/text?doc=Perseus%3Atext%3A1999.04.0057%3Aentry%3Dpo%2Flis

13

I’d recommend just learning

  • τίς τίνος τίνι τίνα,
    τίνες τίσων τίσι(ν) τίνας
    and the neut variation
    τί ̔— — τί,
    τίνα — — τίνα.

Learn to think these question words. Then ask yourself questions about what you are reading.
Σαῦλός ἔγραψεν τήν ἐπιστολήν τῷ φίλῷ αὐτοῦ.

  • τίς ἔγραψεν;
    τίνα ἔγραψεν;
    τίνος φίλου ἔγραψεν Σαῦλος;
    etc.

When you have this paradigm fully internalized, you will more readily recognize the cases of the odd 3rd Declension nouns.

οἱ πατέρες ἔπεμψαν τὴν θυγατέρα αὐτοῦ τῷ μητρί.

  • τίνες ἔπεμψαν; πατέρες
    τίνα ἔπεμψαν πατήρ; θυγατέρα
    τίνι ἔπεμψαν πατήρ; μητρί
14

Literary dialects other than Attic seem to have levelled out the paradigm of πόλις to a relatively regular declension based exclusively on the ι- stem :

http://www.perseus.tufts.edu/hopper/text?doc=Perseus%3Atext%3A1999.04.0007%3Asmythp%3D268+D

But the weird Attic forms apparently persisted into the koine:

http://www.perseus.tufts.edu/hopper/searchresults?doc=Perseus%3Atext%3A1999.01.0155&inContent=true&target=greek&all_words=po%2Flews&phrase=&any_words=&exclude_words=&search=Search

15

χαῖρε, Σαῦλε, and thank you for the suggestion to memorize τίς, τίνος instead. (That paradigm was on my list to be memorized anyway.) But your idea - coupled with Qimmik’s bite-sized description of the linguistic changes that resulted in most of the declension’s oddities - will be a great shortcut in learning the third declension. Thanks to all for your suggestions and help!

16

Now… What’s the genitive plural of τις? :slight_smile: