Pronunciation

Vowels.—Short α, ι, υ differed in sound from the corresponding long voweis only in being less prolonged; ε and ο probably differed from η and ω also in being less open, a difference that is impossible to parallel in English as our short vowels are more open than the long vowels. ᾰ: as a in Germ. hat. There is no true ă in accented syllables in English; the a of idea, aha is a neutral vowel. ε: as é in bonté; somewhat similar is a in bakery. η: as ê in fête, or nearly as e in where. ῐ: nearly as the first e in meteor, eternal. ο: as o in Fr. mot, somewhat like unaccented ŏ in obey or phonetic (as often sounded). ω: as o in Fr. encore. Eng. ō is prevailingly diphthongal (o^{u}). υ was originally sounded as u in prune, but by the fifth century had become like that of Fr. tu, Germ. thür. It never had in Attic the sound of u in mute. After υ had become like Germ. ü, the only means to represent the sound of the old υ (oo in moon) was ου (25). Observe, however, that, in diphthongs, final υ retained the old υ sound.

Diphthongs.—The diphthongs were sounded nearly as follows:

In ᾳ, ῃ, ῳ the long open vowels had completely overpowered the ι by 100 B.C., so that ι ceased to be written (5 a). The ι is now generally neglected in pronunciation though it may have still been sounded to some extent in the fourth century B.C.—The genuine diphthongs ει and ου (6) were originally distinct double sounds (ĕh’-i, ŏh’-oo), and as such were written ΕΙ, ΟΥ in the Old Attic alphabet (2 a): ΕΓΕΙΔΕ ἐπειδή, ΤΟΥΤΟΝ τούτων. The spurious diphthongs ει and ου (6) are digraphs representing the long sounds of simple ε (French é) and original υ. By 400 B.C. genuine ει and ου had become simple single sounds pronounced as ei in vein and ou in ourang; and spurious ει and ου, which had been written E and O (2 a), were now often written ΕΙ and ΟΥ. After 300 B.C. ει gradually acquired the sound of ei in seize. ευ was sounded like eh’-oo, ηυ and ωυ like ēh’-oo, ōh’-oo, pronounced rapidly but smoothly. υι is now commonly sounded as ui in quit. It occurred only before vowels, and the loss of the ι in ὑός son (43) shows that the diphthongal sound was disliked.

Consonants.—Most of the consonants were sounded as in English (1). Before ι, κ, γ, τ, ς never had a sh (or zh) sound heard in Lycia (Λυκίᾱ), Asia (Ἀσίᾱ). ς was usually like our sharp s; but before voiced consonants (15 a) it probably was soft, like z; thus we find both κόζμος and κόσμος on inscriptions. —ζ was probably = zd, whether it arose from an original σδ (as in Ἀθήναζε, from Ἀθηνα(ν)σ-δε Athens-wards), or from dz, developed from dy (as in ζυγόν, from (d)yυγόν, cp. jugum). The z in zd gradually extinguished the d, until in the Hellenistic period (p. 4) ζ sank to z (as in zeal), which is the sound in Modern Greek.—The aspirates φ, θ, χ were voiceless stops (15 b, 16 a) followed by a strong expiration: π^{h}, τ^{h}, κ^{h} as in upheaval, hothouse, backhand (though here h is in a different syllable from the stop). Thus, φεύγω was π̔εύγω, θέλω was τ̔έλω, ἔχω was ἔ-κ̔ω. Cp. ἐφ᾽ ᾧ for ἐπ(ὶ) ᾧ, etc. Probably only one h was heard when two aspirates came together, as in ἐχθρός (ἐκτ̔ρός). After 300 A.D. (probably) φ, θ, and χ became spirants, φ being sounded as f (as in Φίλιππος Philip), θ as th in theatre, χ as ch in German ich or loch. The stage between aspirates and spirants is sometimes represented by the writing πφ (= pf), τθ, κχ, which are affricata.—The neglect of the h in Latin representations of φ, θ, χ possibly shows that these sounds consisted of a stop + h. Thus, Pilipus = Φίλιππος, tus = θύος, Aciles = Ἀχιλλεύς. Modern Greek has the spirantic sounds, and these, though at variance with classical pronunciation, are now usually adopted. See also 108.