111. For the purpose of accentuation nouns may be divided into those in which the accent remains on the stem (and as far as possible on the same syllable of the stem), and those in which it passes in the genitive and dative to the case ending.
Nouns of the vοwel declensions generally belong to the first of these groups. The last syllable if accented has the acute in the nominative and accusative, the circumflex in the genitive and dative, and in the adverbs in -ου and -ως.
καλός, καλοῦ, καλῷ etc.
Acc. Plur. καλούς
On the Nouns in -ᾰ, see § 96.
One or two feminines with nominative singular in -ᾰ accent the ending in those cases in which the last syllable is long.
Nom. ταρφύς thick
Nom. ἄγυια street
Plur. ἀγυιαί, ἀγυιάς
So θαμειαί and θαμειάς answer to a nominative singular θαμεῖα, masculine *θαμύς (cp. θαμέ-ες, θαμέας) ; and καυστείρης (Il. 4.342. etc.) is genitive of καύστειρα.
αὔτως in the very way (from αὐτός), is made barytone by the authorities. The word is only Homeric, and the original accentuation αὐτῶs had evidently been lost, perhaps by a confusion vwith οὕτως.
The oxytone adverbs in -ει and -ι, as αὐτονυχεί, ἀσπουδί, μελειστί, may date from a time when the locative of the ο-declension was regularly oxytone—the accent determining the appearance of ε for ο.
The second group consists οf
- Nouns with monosyllabic stem.
πούς, ποδ-ός, ποδ-ί, ποδ-οῖϊν, ποδ-ῶν, ποσσί
θήρ, θηρ-ός, θηρ-ί, θηρ-ῶν, θηρ-σί
- The words πατήρ, μήτηρ, θυγάτηρ, ἀνήρ, γαστήρ
Gen. πατρ-ός, μητρ-ός, θυγατρ-ός, ἀνδρ-ός, γαστρ-ός etc.
The accent οf μήτηρ and θυγάτηρ is anomalous: cp. the Accusatives μητέρ-α, θυγατέρ-α. Probably the nominative singular was originally oxytone. The change of accentuation may be explained by supposing that the nominative was influenced by the accent of the vocative—that in fact the vocative pro tantο took the place of the nominative (cp. § 96). It is evident that the vocative of these words would be especially familiar to the ear.
The dative ending -εσσι never takes the accent; hence πόδ-εσσι, νή-εσσι, ἄνδρ-εσσι, κύν-εσσι, etc. The reason doubtless is that these are forms that have followed the analogy of the stems in -εσ, as ἔπεσ-σι, βέλεσ-σι, etc.
The Genitives παίδων, δᾴδ-ων, Τρώ-ων, δμώ-ων, θώ-ων, are barytone; perhaps becaαse the stems are originally disyllabic.
It appears that in an earlier stage of the language the shifting of the accent to the case ending was always accompanied by "weakening" of the stem (§ 106), The few instances of the type of κύων, genitive. κυv-ός, and πατήρ, genituve πατp-όs, are to be regarded as surviving examples of the older declension.
112. The vocative in the consonantal declension sometimes retracts the accent.
Proper names with a long vowel in the penultimate are often properispomena.
Otherwise they are mostly proparoxytone, as Ἀγάμεμνον, Ἄπολλον.
Oxytones in -εύς form the vocative in -εῦ, as Ζεῦ, Ὀδυσεῦ. This may be regarded as a retraction of the accent, since the circumflex stands for a double accent, viz. an acute followed by a grave in the same syllable (Ζεῦ = Ζέὺ).
Originally the vocative, unless it stood at the beginning of a sentence, vwas enclitic. Hence the barytone accent is to be explained as in the case of the verb (§ 87), as the result of an original οss of accent.