The Singular

Book Nav

96. Nominative Singular. The final -ς is retained after vowels and mutes, but lost with stems ending in ρ, as πατήρ, μήστωρ.

Stems ending in ν either

  1. take final -ς (with loss of ν).

    εἷς (for ἑν-ς)
    θίς accusative θῖν-α
    μέλας genitive μέλαν-ος

  2. or do not take -ς, but lengthen a preceding vowel.

    χθών genitive χθον-ός
    ποιμήν genitive ποιμέν-ος

So with stems in -ντ: δούς genitive δόντ-ος, but ἰδών. Originally it seems that all monosyllables took -ς and all others -ν (J. Schmidt, K. Z. xxvii. 392). If so, χθών, φρήν, etc. are forms due to the -ν of the oblique cases: and on the other hand διδούς, τιθείς, etc. have followed the analogy of corresponding monosyllabic words, δούς, θείς, etc.

There is a remarkable group of masculine stems in -ᾱ(η), with nominative singular in -ᾰ, viz:

Titles of gods

εὐρύοπα (Ζεύς)
ἀκάκητα (Ἑρμείας ἀ.)
κυανοχαῖτα (Ποσειδάων)

Titles of heroes

ἠπύτα (κῆρυξ)

One proper name, Θυέστα (Il. 2.107).

Except Θυέστα these words are only found as adjectives: thus we have

αἰχμητὰ Λυκάων
κυανοχαῖτα Ποσειδάων

but αἰχμητής, κυανοχαίτης when the same words are substantives.

The accent generally follows the forms in -η-ς where such forms exist; thus ἱππότα, αἰχμητά, like ἱππότης, αἰχμητής

But it is thrown back in εὐρύοπα, μητίετα, ἀκάκητα–ancient epithets only known from the traditional Homeric use.

These are in reality vocatives which have been turned into nominatives. That is to say, they belonged originally to certain established forms of address—μητίετα Ζεῦ, κυανοχαῖτα Ποσείδαον, ἱππότα Πηλεῦ, etc.—and were not inflected when the names to which they were attached came to be used in the nominative In this way the rhythm, which doubtless had a traditional sacredness, remained unaltered, and the whole phrase retained something of its vocative character. The feeling which might lead to this is that expressed by Eumaeus in Od. 14.145 ff.

τὸν μὲν ἐγών, ὦ ξεῖνε, καὶ οὐ παρεόντʼ ὀνομάζειν
αἰδέομαι· περὶ γάρ μʼ ἐφίλει καὶ κήδετο θυμῷ·
ἀλλά μιν ἠθεῖον καλέω καὶ νόσφιν ἐόντα.

Ι call him by the title ἠθεῖος even in his absence—the word ἠθεῖος being only used as a form of address. Cp. also § 111.2. The nominatives in -ᾰ are evidently part of the archaic and conventional style of epic poetry. They are commoner in the Iliad than in the Odyssey in the proportion of 3 to 1. The ancient grammarians regarded them as Aeolic, but without sufficient reason.

The form εὐρύοπα also appears as an accusative, and has accordingly been explained from a nominative εὐρύ-οψ. It is improbable however that it is a different word from the nominative vocative εὐρύοπα. Probably the fact that it had the appearance of an accusative of one of the numerous compounds in -οψ led to an extension of use.1

97. Accusative Singular. The ending -ᾰ is found after consonants and the diphthongs ηυ, ευ.

νηῦ-ς, νῆα (for νηυα, νηϝα)
βασιλεύ-ς, βασιλῆα
Τυδεύ-ς, Τυδέα.2

Otherwise the accusative takes -ν.

πόλι-ν, ἰχθύ-ν, βοῦ-ν

But εὐρύ-ς makes εὐρέα in the phrases εὐρέα πόντον, εὐρέα κόλπον: the common form being εὐρύ-ν.

The original ending is -m, which becomes -ν after a vowel and -ᾰ (for ) after a consonant. The preference for -ᾰ after ηυ, ευ is due to the semi-consonantal nature of the υ in these combinations. We may compare the aorists ἔκηα (for ἐκηυ-α), ἔχευα (also ἔχεα), etc., and on the other hand ἔδυ-ν, ἔφυ-ν.

Several stems form the accusative in -ιν and also in -ιδα.

ἔριδα and ἔριν (Od.)
φυλόπιδα (Od. 11. 313) and φύλοπιν
γλαυκώπιδα (Il. 8. 373) and γλαυκῶπιν (Od. Il. 156)
ἀνάλκιδα and ἄναλκιν (Od. 3. 375)
ὄπιδα and ὄπιν
Κύπριδα and Κύπριν

Cp. also χάρι-ν (for χάριτ-α), and κόρυ-ν (for κόρυθ-α), found in the line Il. 13.131 (= 16.215):

ἀσπὶς ἄρʼ ἀσπίδʼ ἔρειδε, κόρυς κόρυν, ἀνέρα δʼ ἀνήρ.

In Attic there are many more such forms; ὄρνιν, etc.

Note that no oxytones form the accusative in -ιν.

The accusatives ζαῆν (Od. 12. 313), Ἄρην, Μέγην are probably formed directly from the nominative ζαής, Ἄρης, Μέγης, on the analogy of masculine nouns in -η-ς. On the other hand Ζῆν (Ζεύς), βῶν (βοῦς), are very ancient forms, answering to the Sanskrit dyâm, gâm (Joh. Schmidt in K. Z. xxv. 17): see § 106.2.

A final δ is lost in the neuter pronouns ὅ, τό, τοῦτο, ἐκεῖνο, ἄλλο (Lat. id, is-tud, illudaliud), and in τί (Lat. quid): perhaps also in the personal pronouns, accusative singular ἐμέ (με), σέ, ἕ, dual νώ, σφώ, σφῶε, plural ἄμμε, ὔμμε, σφέ (Curt. Stud. vi. 417 ff.; Max 11. Müller, Chips, iv. 44).

98. Gentive Singular. The stems in -ο form the genitive in -οιο, -οο, -ου. Of these forms only -οιο and -ου are read in the existing text of Homer; but there are sufficient traces of
-οο, and indeed several places where it is called for by the meter. Thus we must read

Il. 2.518 υἱέες Ἰφίτοο μεγαθύμου

Il. 15.66 (= 21. 104) Ἰλίοο προπάροιθεν

Il. 22.313 ἀγρίοο, πρόσθεν δὲ κτλ.

Od. 10.36 δῶρα παρʼ Aἰόλοο μεγαλήτορος

Od. 60 βῆν εἰς Αἰόλοο κλυτὰ δώματα

II. 9.440, etc. ὁμοιΐοο πτολέμοιο (for ὁμοιΐου πολέμοιο)

Il. 2.325 ὅο κλέος οὔποτʼ ὀλεῖται (for ὅου)

Od. 1.70 ὅο κράτος ἔσκε μέγιστον (for ὅου)

Il. 2.731 Ἀσκληπίοο δύο παῖδε

Il. 15.554 ἀνεψιόο κταμένοιο

Il. 5.21 ἀδελφεόο κταμένοιο

so in—

Il. 6.61 ( = 7.120, 13.788) ἀδελφεόο φρένας ἥρως

Od. 14.239 χαλεπὴ δʼ ἔχε δήμοο φῆμις

Also in the two lines

Il. 6.344 εἵνεκ’ ἐμεῖο κυνὸς κακομηχάνου ὀκρυοέσσης

Il. 9.723 ὃς πολέμου ἔραται ἐπιδημίου ὀκρυόεντος

since ὀκρυόεις does not occur elsewhere, but κρυόεσσα (Il. 5.740), κρυόεντος (Il. 9.2), κρυερός etc., we should probably read

. . . κακομηχάνοο κρυοέσσης

. . . ἐπιδημίοο κρυόεντος

A trace of -οο may also be found in the fact that nouns in -αος sometimes form the genitive in -εωο, which is for -αοο; e. g. Πετέ-ωο, Πηνελέ-ωο.

Masculine stems in -ᾱ(η) form the genitive in -ᾱο (original -ᾱσι̯ο), less commonly -εω (by transference of quantity). This -εω is often scanned as one syllable; after another vowel it is written -ω.

Βορέ-ω (for Bορέ-εω)

(So in Ionic, Curt. Stud. v. 294., viii. 172.)

The pronominal stems in -ε, viz. ἐμε (με), σε (for τϝε), and ἑ or ἑε, form the genitive in -ε-ιο, -εο and (by contraction) -ευ. Thus we find

ἐμεῖο, ἐμέο (Il. 10.124), ἐμεῦ
σεῖο, σέο, σεῦ
εἷο, ἕο, εὗ

For σεῖο there is also a longer form τεοῖο (Il. 8.37 = 468), and for ἕο in one place (Il. 19. 384) Zenodotus read ἑοῦ.

99. Dative Singular. In Homer the ι of the dative is sometimes long (as in Latin), chiefly in forms which otherwise could not be easily brought into the verse

in the Iliad:


in the Odyssey:


But we find also Ζηνῒ μενεαίνομεν (Il. 15.104), πὰρ νηΐ τε μένειν (Od.) See § 373.

The dative of neuters in -ας was commonly written -ᾳ; but the long α is anomalous, and
-αι is now read by La Roche from good MSS. (in σέλαι, κέραι). The forms in -ᾳ appear to have become established in later Greek (Hdn. Il, 316, 10, ed. Lentz).

Stems in -ι, genitive -ι-ος, form the dative in -ῑ

νεμέσσι (with v. ll. νεμέσσει Il. 6.335)

So Bekker restored the forms

πόλι (Il. 5.686, etc.)
ἀγύρι (Il. 16.661)

for which the common texts give forms in -ει.

Stems in -υ, genitive -υ-ος, form the dative in -υι (a diphthong which in later Greek can only occur before a vowel).

πληθυῖ (Il. 22.458)

But δρῦ-ς, σῦ-ς form the disyllables δρυ-ΐ, συ-ΐ.

It is possible, however, that the datives in -ῑ are instrumental forms, and similarly that the datives in -υι have taken the place of instrumentals in -ῡ. For the Vedic and Zend instrumental in -ῑ, -ῡ see Osthoff, d. 7. ii. 139.

Sanskrit nouns in -an and -as sometimes form the locative from the stem without any case ending (Whitney, 425, c). Traces of this are to be found in Greek in the form αἱέν (cp. αἰεί), and the infinitive in -μεν and -εω (§ 85).

Stems in -ο sometimes form a locative in -οι, as well as the true dative in -ῳ, e.g. οἴκο-ι. So χαμα-ί and perhaps πάλα-ι. Cp. the adverbial ending -ει (§ 110).

Pronominal stems in -ε form -οι; ἐμοί (enclitic μοι, σοί (encl. τοι), ἑοῖ and οἶ. For σοί there is another form τείν (Il. 11.201): so in Doric we find ἐμίν and ἑΐν, ἵν.

99*. Plural. Several stems in -ο which are masculine (or feminine) in the singular form a neuter plural.

plural κέλευθοι and more commonly κέλευθα

plural μηροί and μῆρα

plural κύκλοι and κύκλα

plural ἰοί and ἰά

plural Τάρταρα (Hes.).

There is probably a slight change of meaning, the neuter expressing vague mass or quantity rather than plurality: cp. δρυμά thicket, and post-Homeric δεσμά, θεσμά, σῖτα, Latin loca, joca. Thus κέλευθὰ means a group of paths, and could not be used (e. g.) in such a passage as

Il. 10.66 πολλαὶ γὰρ ἀνὰ στρατόν εἰσι κέλευθοι

So κύκλα of a set of wheels, Τάρταρα of one place so called, etc.

Note— To the examples of metaplastic neuter plural used with collective meaning add ἕσπερα evening time (Od. 17.191), νεῦρα sinews (used in Il. 16.316 of one bowstring), πλευρά side (Il. 4.468), παρειά cheeks (neuter plural in Il. 22.491 according to Aristarchus). It may be suspected that ἐμετμά oars belongs to this group, since the singular in later Greek is always ἐρετμός, and a neuter ἐρετμόν (Od.), for which we can read εὐήρε’  ἐρετμόν.

  • 1. It will be shown hereafter (§ 116.2) that the masculine nouns in -της are probably derived from feminines in -τη, of abstract or collective meaning. Hence it is possible that the Homeric nominatives in -τᾰ come directly from these feminines: so that (e.g.) μητίετα meant literally counsel rather than counsellor. The abstract word may have been used as a title, like βίη Πριάμοιο and the like. According to Joh. Schmidt (Pluralb. p. 400) εὐρύοπα is originally a Neuter: see § 107.2.
  • 2. The forms Τυδῆ (Il. 4.348) and Μηκιστῆ (Il. 15. 339) are probably false: see Nauck, Mél gr.-rom. iii. 222.