Page images
[ocr errors]

The long and short sounds of a vowel were probably identical in quality. In English they are always different.

& as in Italian, i.e. as in father; not as in fate.
å the same sound shortened, as in French chatte; not as in hat.

as Italian open o, nearly as in dot.
Ő as Italian open o, or the Cumberland pronunciation of home,

a sound nearer to English aw than is the ordinary o in dote,

or in the ordinary English home. (pp. Ixiii-lxix.) ů as in Italian, i.e. as French ou in poule, nearly as in pull;

not as in lull. ū as in Italian, i.e. oo in pool; not with a prefixed y-sound, as

in pule, mule. è as Italian open e; nearly as in pet, met. ē the same sound lengthened; not as in peat, mete. (pp.lxiii.—lxix.) ī as Italian i, i.e. as in machine ; not as in shine, pine. I the same . sound shortened: but practically the ordinary

English short I may be used, as in pin. y as Germ. ü, but inclining to i, e.g. Müller, which is nearer

Miller than Muller.

[ocr errors]

[This pronunciation of 7 and ē is recom

mmended, partly because it appears more probably to be right than the sound of French au and French é: partly because the ordinary English long o and long a, which might be otherwise used, are usually diphthongs, (see $ 21.)]

A long vowel was pronounced long, and a short syllable short, whether by itself or before one or more consonants, e.g. lüx, lūce; påter, pătre; māter, mātre; amânt, regúnt, &c. (p. lxxiii.)

A vowel before ns or nf was pronounced long (§ 167).

In unaccented syllables, each vowel probably had its proper sound, instead of their being all alike reduced as in English to the sound in mention, paper, label, turban, &c. (p. lxxiii.)

When est followed a vowel or m, the e was omitted ($ 721).


The right rule for pronouncing diphthongs is to pronounce the constituent vowels as rapidly as possible in their proper order. (See a more exact account in App. A. xi. xii.) This will give as follows: au as in Germ. haus, i.e. a broader sound than ow in cow;

not as au in cause. eu as in Italian Europa, i.e. as ow in Yankee town. ae nearly as (the single vowel) a in the Somerset pronunciation

of Bath, i.e. as in bat lengthened. (p. lxix.)
oe as a diphthong. (p. lxx.)
ei nearly as in feint, but with the stress on the latter vowel;

not as long English i. (Cf. § 267.)
ui (in huic, cui) as French oui. (p. lxx. and § 222.)

The diphthongs ou (§ 251) and oi ($ 263) are found only in early Latin.

[blocks in formation]

c always hard, as k in kitty; not as c (=s) in city. (pp. xliii —

1.) g always hard, as g in give; not as in gin. (p. li.) ng as ng +g i.e. as in anger (i.e. ang-ger); not as in hang-er.

So nc, ng, as ng +c, ng +q. (p. liii.) as English y, in year; not as English j in jeer. ($ 138.) v as English w in wine, or perhaps sometimes French ou in oui;

not as v in vine. (pp. xxxii-xlii.) qu as in English, e.g. queen. But quu should be avoided, and

e. g. quom or cum uttered. On qui- see p. lxxi. r always trilled, never vocalized as commonly in English when

a vowel does not follow. (See App. A. xiii.-xvii.) Thus per should be sounded as in perry, not as in pert; ēre as English ā-ry, not airy: ire as (English) ee-ry, not eary.

8 always sharp as in hiss; not (like z) as in his. (pp. liv

lvii.) The mispronunciation by Englishmen occurs most

when s follows e or n. bs as ps, not as bz. (p. liii.) x always as ks, as in axe; not gz, as in exact. (p. liii.) ti always tee (long or short as the case may require), not (as

before a vowel, e.g. natio) as sh or she. (p. lii.) ph, ch, th were not like English f, German ch, English th,

but as p+h, k+h, t+h: sounds somewhat difficult to Englishmen, but often heard from Irishmen ($ 132).

[ocr errors]

In prepositional compounds assimilation in pronunciation was usual in certain cases : ad was completely assimilated to all consonants, except b

and m. (§ 160. 9.) sub, ob were completely assimilated to c, f: and became sup,

op, before sharp consonants. ($ 78.) com was completely assimilated to 1, r; became co before gn

and h; and became con before all other consonants, ex

cept labials. ($ 85. 4.) in was completely assimilated to 1, r, and became im before

labials. (S$ 168. 1. 2; 176. 1; 184. 1.) per was completely assimilated to l. ($ 176. 1.) On other cases see Book I.

The other consonants in Latin were probably pronounced as we now pronounce them. But final m was sometimes not sounded, or perhaps gave only a nasal sound to the vowel. (p. lxxiv.)

An observance of the Latin rules for accentuation does not involve much which is different from the usual English practice (p. lxxv). On the division of the words into syllables, see S$ 15, 232 ; pp. Ixxv--lxxviii.


A few examples will show plainly the great difference between the ordinary English, and what is here represented to be the Roman, pronunciation. To express the pronunciation I have thought it best to follow no exact system, but to select, where possible, common English words or syllables. I have however used ah, æ, eh, and ò for what I suppose to be the true sounds of Latin a, a, e, and o as defined above; err for the sound in herring, not in English err; ay for the ordinary English long a.

[blocks in formation]

The division of syllables in the above is, in order not to embarrass the reader, accommodated in the main to the ordinary view.

Observations on Book ii.

Noun-stems ending in e.


I believe the general doctrine of grammarians may be represented to be, that the stems commonly forming the fifth declension have the genitive and dative singular, except occasionally in poetry, in ei; that the ei is a dissyllable; and that the e is usually long, e.g. diēī, but short, if it follows a consonant, e.g. fiděī. And accordingly it is common enough to find modern writers using such words as materiēī, and referring (e. g. Corssen, II. 723) without hesitation to words like faciēī, notitiēī, amicitiei, as if they were of common and undoubted occurrence. Now, putting aside the Latin authors subsequent to the silver age, into whose usage on this point I have made but little investigation, and speaking of the older period, that which alone I regard in this volume, I believe all the above parts of the ordinary doctrine to be quite unfounded. I do not profess to have read through all the writers of the gold and silver ages with a view to this inquiry, but I have used such other means as were available, and have had the point before me for some years. The result is stated in SS 340-—343, and 357 and 360. The kernel of the whole matter is to be found in Gellius, ix. 14, and in Quintilian's significant question (v. 6. § 26) quoted in the note to p. 116; and the inference, which may be thence drawn, is confirmed by Neue's collection of the facts of actual usage. The great mistake commonly made is in starting from the assumption, derived from Roman grammarians, that a dissyllabic ei is the regular ending, and consequently only noticing what are supposed to be deviations. In $S 357, 360 will be found all the instances that I have been able to collect of the use of a genitive or dative singular of an e stem at all, It will be seen that dies, res, spes, fides and plebes, are the only words which are found in these cases, except quite sporadically,

« PreviousContinue »