In October 1908 the Classical Association adopted a number of recommendations made by its Greek Pronunciation Committee, and has since published them for the use of teachers and others. They are put forward “not as constituting a complete scientific scheme, but as approximations which, for teaching purposes, may be regarded as practicable, and at the same time as a great advance on the present usage, both for clearness in teaching and for actual likeness to the ancient sounds.” The period (the early fourth century B.C.) to which they are intended mainly to apply is one whose literature Dionysius studied rather than that in which he lived (cp. pages 43-46 above). But his scattered hints are of great moment in the whole inquiry; and if they are read with care and with reference to their bearing, not only on disputed points, but on points which (largely through the evidence they furnish) are undisputed, it will be seen how much we owe to them when making any attempt to reconstruct the pronunciation of the classical period. The principal passages of Dionysius’ text which throw light upon the question of Greek pronunciation and accentuation will be found on pages [126]-130, [136]-150, [218]-224, [230] above. The following are the suggestions made by the Classical Association:—
Vowels
ᾱ and α, ῑ and ι, ε and ο, η and ω may be pronounced as the corresponding vowels in Latin, i.e.
ᾱ, as a in father,
α, as a in aha.
ῑ, as ee in feed.
ι, as i in Fr. piquet, nearly as Eng. i in fit.
ε, as e in fret.
ο, as o in not.
η (long e), as e in Lat. mēta, Eng. a in mate.
ω (long o), as o in Lat. Rōma, Eng. home.
The pronunciation recommended for η and ω is dictated by practical considerations. But in any school where the pupils have been accustomed to distinguish the sounds of French è and é, the Committee feels that the open sound (of è in il mène), which is historically correct for η, may well be adopted. In the same way there is no doubt that the pronunciation of ω in the fifth century B.C. was the open sound of oa in Eng. broad, not that of the ordinary English ō. But since the precise degree of openness varied at different epochs, the Committee, though preferring the open pronunciation, sees no sufficient reason for excluding the obviously convenient practice of sounding ω just as Latin ō. For both Greek and Latin the diphthongal character of the English vowels in mate and home, i.e. the slight ĭ sound in mate and the slight ŭ sound in home, own, is incorrect. But the discrepancy is not one which any but fairly advanced students need be asked to notice, unless indeed they happen to be already familiar with the pure vowel sounds of modern Welsh or Italian.
υ as French ŭ in du pain.
ῡ as French ū in rue or Germ. ü in grün.
In recommending this sound for the Greek υ, the Committee is partly guided by the fact that its correct production is now widely and successfully taught in English schools in early stages of instruction in French and German. But in any school where the sound is strange to the pupils at the stage at which Greek is begun, if it is felt that the effort to acquire the sound would involve a serious hindrance to progress, the Committee can only suggest that, for the time, the υ should be pronounced as Latin u (short as oo in Eng. took, long as oo in Eng. loose), though this obscures the distinction between words like λύω and λούω.
Diphthongs
αι = α + ι nearly as ai in Isaiah (broadly pronounced), Fr. émail.