and the latter, Dutch duist, but not found in German. I know no tradition of up sounded as oup, but the German auf points in that direction, and the Anglo-Saxon word is often found accented (úp). Just so Chaucer's form ous for us becomes intelligible now by means of the Anglo-Saxon ús.

35 Objection from Palsgrave.

36

But there are two or three objections to look at. First, Palsgrave speaks of cowe, mowe, sowe, as sounded almost like the Italian u and the French ou; and Mr. Ellis asserts that we certainly know what sound these symbols represent. But without insisting on the uncertainty of this knowledge, and that there may be much meaning in that "almost;" I would urge that both Salesbury and Sir Thomas Smith were nearly contemporary with Palsgrave, the former of whom seems to have known no other word spelt with -owe and pronounced (uu) but the word wowe (woo); and the latter expressly makes the nouns mow and sow (moou) and (soou). I am in the dark as to Palsgrave's meaning. Only I would observe that certainly in Old French, possibly therefore in Palsgrave's time, the symbols o, u, ou, &c., at least before n, had not the same sound as now. Thus words like baron, raison, which in the Chanson de Roland are spelt with un or on, are rarely assonant with other u words, never with other o words: they for the most part stand by themselves. It is therefore not unlikely that the sound struck the English ear as approximating to our -oun.

vation

were

I have said approximating, but the approximation was probably somewhat close. Two reasons lead to this conWords in on clusion. First, in Chaucer the English repreof French deri- sentatives of French words in -on, habitually with rhyme with down, town, soune, brown, gown, &c., (doubtless having ceased to be sounded "somethyng in the noose,” as Palsgrave would say); while they scarcely ever rhyme with words in (oon)—as now

sounded

(ǝun).

* I

say "ceased" on the supposition that this anuswára, which exists also in modern Welsh, is correctly regarded as among the Celtic, and therefore the earliest, elements of French.

pronounced, and never, so far as I have observed, with words in (uun).* Secondly, many-indeed I believe allof the older English derivatives of French words containing on have now the sound of (ǝun), as abound, fountain, mountain, to mount, to found, confound, profound, noun, renown, renounce, pronounce, announce, round, and the vulgar Mounseer for Monsieur; there being very few, if any, exceptions to this rule except where the final unaccented -oun has shrunk and withered into (en), as in mention, nation, extension. Balloon and caissoon are but modern words. It thus becomes an almost necessary conclusion that if the original French words were not sounded exactly with (ǝun), at any rate-for this is the point we are seeking to ascertain their descendants in English were so sounded. 37 Second objec

38

cer and the Cuc

koo Song.

Then there is Chaucer's rhyme of cuckow tion from Chau- with how, swow, now, and thou, while it does not rhyme with words like do, like go, or like know; as in the Cuckoo Song also the same word, spelt cucu, rhymes with cu and nu; the forms cow and now not having yet come into fashion. From these rhymes my conclusion is simply this, that Chaucer and the writer of the Cuckoo Song pronounced the word with (ǝu). But cuckoo, it is urged, is an imitative word, and the final vowel is (uu), not (ǝu). True, but if we insist on the natural sound as (uu), how are we to account for the Greek KóκKUέ and KÓKKU? Were not these in all probability sounded not with (uu), but (yy), and later (ii)? And is it really the case in our or in any language that the imitative words are exact imitations? What then of our neigh, bellow, cackle, laugh, the Dutch brieschen, hoest, the French rire, tousser, glousser, &c. &c.? To say that they were once correctly imitative words in an earlier stage of their existence is to say nothing; for Chaucer's cuckow was not in its earliest stage, and might as easily appear in a corrupt form as our now familiar laugh.

Let us look however at the derivation of cuckoo. It is * Alone, bone, stone, to gon, to grone, &c.; and noon, soon, moon, to don, &c. These classes of words are discussed below in § 54.

imitations of the

=

Cuckow and not from the A.S. gæc nor the O.N. gaukr, but prow imperfect from the French cucu, now coucou; which leads French originals. me to observe that there is another old French word, prowe or pru profit, which in Chaucer, in the form prow, rhymes repeatedly with now and you, which we now pronounce (Juu). Now the very form of prowe (given by Kelham), though I can find no evidence for a similar byform of cucu, seems to imply some diphthongal sound which may-as I have suggested in a like case at the end of § 35 -have struck the English ear as resembling the English (ǝu) their nearest indigenous sound to represent it. (Compare ponch as the French representative of our familiar punch.) But here is the point: weighty evidence will be adduced by and by (§§ 51 to 56) tending to prove that do, to, shoe, were sounded by Chaucer with (uu), just as they are sounded at present; and Chaucer, whether it pleases our taste or not, did not make cuckow and prow rhyme with do, &c., but with thow, &c.

Of one of these words a derivative survives in our language, namely prowess, retaining, as I believe, the same diphthong as the root word had in Chaucer. The other, cuckowe, has changed; but is this difficult to account for? Not to mention our greatly increased intercourse with our French neighbours, and that at any rate their modern name for the bird is in accordance with the actual cry, we ourselves every spring take lessons in music from the songster himself, and have thus arrived at a more perfect onomatopoeia than that which we first imported.

39 Negative argu

words will not

Mr. Ellis however takes this written ou to be ments: first, ou (uu). But besides the positive evidence above rhyme with (uu) adduced to show that in a large class of words it stood for (ǝu), or some modification of that sound; the negative argument may also be fitly urged, and reasons be assigned for believing that ou was not (uu).

words.

First then, we shall presently find another class of words which contained and contain (uu), and the sound is not expressed by ou, so that that sound is independently provided for; and moreover these words are never found

rhyming with those that in A.S. are written with ú, and in E.E. with ou or ow: an argument of great weight, though it needs but few words to express it.

40 Orthoepists do

41

not make our ou

Then again, Smith, Hart, and other orthoeItalian . pists are in the habit of referring to foreign languages to illustrate the nature of English sounds; but, except only the doubtful and limited assertion of Palsgrave above alluded to, no one of them makes the English ou or ow to be equivalent to the French ou or the Italian or German u.

Cheke and

of ov.

Thirdly, both Cheke and Smith take the Smith's account Greek ov (and Cheke, and possibly Smith,* the Latin u also) to be sounded as our ou in foule and houle, and expressly describe this as a compound sound. Smith says: "Ov diphthongus Græca, ou et wv. Ex o breui and u, diphthongum habebant Latini, quæ si non eadem, vicinissima certè est ov Græcæ diphthongo, et proxime accedit ad sonum u Latinæ," &c. See Ellis, p. 151 (where I think the "transliterations" are correct.) Smith's o brevis is exemplified elsewhere in the words smock, horse, hop, sop, not, rob, bot, pop; and I can see no sufficient ground for believing these words to have been at all otherwise sounded 300 years ago than they are now. His description therefore is at least approximately correct for (ǝu). Smith's expression, speaking of ov (de Ling. Græc. Pron., p. 38), "u facit Latinum quando producitur," I take to mean that when the (ǝu) sound is prolonged, the prolonged part of the compound is (uu), which it undoubtedly is. 42 Smith's (uu) But fourthly, when Smith elsewhere (de Ling. Angl. Script., p. 12) speaks of the Latin u— and in this passage I agree with Mr. Ellis (p. 167) that it is really (u) or (uu) that is intended—it is very remarkable that not one of the English words given as examples is spelt with ou or ow. Here they are: "Brevis but, sed;

words not written

with ou.

* I mean that it is possible that Smith may have been inconsistent with himself, and entertained this notion when he wrote the passage quoted just below, which however he certainly did not entertain when he wrote that quoted in the next paragraph.

luk, fortuna; buk, dama mas; mud, limus; ful, plenus; pul, deplumare; tu, ad: longa, büt, ocrea; lük, aspicere; bük, liber; müd, ira aut affectus; fül, stultus; pül, piscina; tü, duo, etiam." I need hardly explain that the words which Smith thus writes are, in the usual spelling, but, luck, buck, mud, full, pull, to; boot, look, book, mood, fool, pool, two or too,-all, I believe with Mr. Ellis, sounded with (u) or (uu) in Smith's time. It is strange indeed if, when a certain sound was to be exemplified, and a multitude of words in ou, as Mr. Ellis supposes, contained that sound, not one of the fourteen examples was so spelt.

grammarians dis

as in modern

43 Fifth. Other Lastly, the grammarians on whose authority tinguish (uu) and Mr. Ellis so much relies, and whom, as I believe, (ǝu) words just he so frequently misinterprets, furnish this English. additional argument against him, that they habitually observe a distinction in spelling between these (uu) and (ǝu) classes of words which Mr. Ellis confounds, and that distinction is just the same as in modern English. Thus Bullokar, 1580, writes intoo, whoo, stool, tool, good, boot, broom, doo, dooth, look, crooked, &c.; but how, ground, douteth, found, towel, your, about, sound, bow (vb.), vowel, bowel, sower (i.e. sour, German sauer). Bullokar's spelling is phonetic,* which greatly adds force to my argument; and other phoneticists-Gil and Butler who were later than Bullokar, and Smith and Hart who were earlier-all mark the same distinction. And so do other grammarians, who did not adopt a peculiar orthography; Palsgrave for example. And so did Chaucer: both his rhymes and, I may add, his spelling convict Mr. Ellis of error.

44

In a few words

Were there then in Early English no words ou was= (uu). spelt with ou or ow and sounded otherwise than with (ǝu)? Certainly there were some sounded with (u) or (uu) or with (0) or (oo), chiefly words of French derivation and imperfectly Anglicised, many of which are still exceptionally pronounced. For example, touch, country, double, trouble, course, discourse, flourish, courage: so far

* Though I fail to discover any difference between the forms which he marks with the cedilla.