teration in the position, or any motion of the organs of speech, from the moment the vocal sound commences till it ends, while a consonant is an interruption of the effusion of vocal sound, arising from the application of the organs of speech to each other."*

But as Müller,† from whose researches we shall borrow largely in the succeeding pages, observes, the 'difference between vowels and consonants is, however, much less considerable than this, for all vowels, as well as the consonants, can be produced without a vocal tone; and, morever, one whole class of consonants can be uttered with a vocal sound as well as without this sound.'

Our grammarians next proceed to divide the consonants according to the organs by which they are supposed to be formed. Now, a little attention to the organic formation of these letters would serve to show the absurdity of the distinction which is attempted, and this creates the difficulty of explaining the conversions of consonants in several languages springing from the same root. In his attempt accurately to classify the sounds of articulate speech, Müller first corrects an error which has pervaded all previous systems, and lays it down as a fundamental principle, necessary to enable us to recognize the essential properties of the articulate sounds, that we must first examine them as they are produced in whispering, and then investigate which of them can also be uttered in a modified character, conjoined with vocal tone. By this procedure,' he continues, we find two series of sounds; in one, the sounds are mute, and cannot be uttered with a vocal tone; the sounds of the other series can be uttered independently of voice, but are also capable of being uttered in conjunction with it.'

6

6

There are certain sounds which require, for their production, a difference in the form of the aperture through which the voice escapes. They are continuous tones, and are capable of prolongation during any time that the breath can sustain them: this is the proper definition of a vowel. Such sounds are, therefore, the simplest which can be formed, requiring, for their accurate production, merely that the learner should know the position of the mouth on which they severally depend; they are, therefore, perhaps the best with which to commence instruction. In the English language, we improperly admit six, or even seven vowels by name; and, taking into account the different powers which several of these possess, the number is greatly increased. 4, for example, has four distinct sounds, as heard in the words fate, far, fall, and fat, respectively. But ap

* Walker's Principles of English Pronunciation.

+ Elements of Physiology, translated by William Baly, M.D., Part 4.

plying the definition which we have given of vowel sounds, several of these must be rejected. Thus, the short A, as in fat, is incapable of prolongation, while the German A, as heard in ball, fall, &c., might appropriately be represented by au, which, however, is not to be considered as a diphthong. E has two sounds, of which the words theme, and red may serve as examples. I has no title whatever to be considered as a vowel. It is formed by the combination of the Italian A, as heard in Father, and the long E, run into one another. O is usually regarded as possessing three sounds. Sheridan instances these as occurring in the words not, note, prove, respectively. Walker adds a fourth, as in love, a fifth, as in for, and a sixth as in woman; but it is plain that of these six, only two can be regarded as true vowel sounds. As an example of the first we may give the word cold, while the second is instanced in the word cool. The only sound of U not adequately represented by the combination of some of the foregoing vowels is that which occurs in rude, but testing it by the definition we have given of a vowel, it will be seen that it does not present the characteristics there laid down, but is a true diphthongal sound. Y might be omitted from the alphabet altogether, its long and short sounds being identical with those of the letter I, and subject to the remarks made on them. Though we have thus omitted I, U, and Y from our list of proper vowels, yet when the varieties of several of those retained are taken into account, it will be seen that five vowel sounds are still allowed as heard in the following words, ah, name, theme, cold, cool. Kratzenstein* and Kempelent have pointed out that the same sound produced in the larynx is altered into the different vowels by changes in the size of the opening of the mouth, and of the space bounded by the cheeks, the palate, and the tongue. Admitting five degrees of size, both of the opening of the mouth (' oral opening') and of the space so bounded (oral canal') Kempelen states the dimensions of these parts in the different vowels as follows:

• Tentamen resolvendi problema, &c. 1780.

64321

32146

5

+ Mechanismus der Menschlichen Sprache, nebst der Beschreibun gseiner sprech

enden Maschine. Wien, 1791.

The close affinity which exists between the several vowel sounds is clearly proved by the interchange which so frequently occurs between them in rendering words from one language into another, which has led Sir W. Jones to protest against the licentiousness of etymologists, in totally disregarding the vowels,' in their attempts to assign a common origin to words.*

It is well illustrated in the following table, slightly altered from Webster:-+

But we need not have sought examples in the transmutation of words in passing from one language to another, the same is observed in the different inflections of the same word, in the same language.

Examples of this are numerous in Latin. Facio, Feci; Sto, Steti; Ago, Egi; Vello, Vulsi. Nothing is more certain,' observes Webster, 'than that the Welsh goryz and the English wood are the same word, although there is one letter only common to them both.'

To communicate then the proper pronunciation of the five simple vowel sounds, it is only necessary to bear in mind, that the vocal tone produced in the larynx by the vibrations of the vocal chords is the same for all, and that it is by changes in the oral canal and opening that it is modified into the different vowels. The nature of these changes is so well exhibited in the foregoing table of Kempelen, that farther illustration is unne

cessary.

In endeavouring still farther to point out the connection subsisting between spoken and written language, and how each may be made to contribute to the correct acquirement of the other, we turn next to the consonants; but in attempting any arrangement of them, we encounter great difficulties. The two great

Asiatic Researches, vol. iii. p. 480.

A Dictionary of the English Language, &c. &c, New York.

6

By Noah Webster, LL.D.,

6

est defects in the orthography of any language,' observes Halhed, are the application of the same letter to several different sounds, and of different letters to the same sound.'* To these we may add another defect, peculiarly brought out by any comparison between spoken and written language, viz. that our words are, in many cases, very far from being pronounced as they are written. A very different rule was laid down by Quinctilian'Ego nisi quod consuetudo obtinuerit, sit scribendum quodque judico quomodo sonat; hic enim usus est literarum, ut custodiant voces et velut depositum reddant legentibus, itaque id exprimere debent, quod dicturi sunt.' In our language we find innumerable examples of words spelt differently and pronounced alike, and of words spelt alike and pronounced differently: the termination ough affords an amusing instance of the latter, which has nine different pronunciations; it has the sound of up in hiccough; of aou in bough; of o in dough; of off in cough; of och in lough; of uff in tough; of ou in through; of o short in thorough, and of aw in thought. It would tend greatly to facilitate this part of our subject, could we lay down distinctly the number of simple elementary sounds which the human voice is capable of uttering. This has been attempted by Wachter,§ who thinks ten characters comprehend all; by Bibliander, who admits thirteen; by Wilkins, who swells the number to thirty-four; by Harris,** who enumerates twenty; by Franklin,++ twenty-six; by Sheridan, twenty-eight.

Without attempting the solution of a question which has baffled the ingenuity of so many able men, we shall content ourselves with simply pointing out some of the redundancies noticed by Franklin, and then endeavour to classify the remaining consonants according to the modifications of those organs by which they are formed.

In his alphabet, Franklin omits c, its hard sound being represented by k, and its soft by s. Qis supplied by k, and ks take the place of x. U is sounded oo, and therefore renders w unnecessary. Y has its place supplied either by i or by two vowels, and jis better represented by an aspirated s, sh.

In examining the manner in which the remaining consonants are formed, and attempting to classify them accordingly, we find,

* Grammar of the Bengal Language. Ed. Encyc. vol. i. p. 565.

+ Inst. Orat. lib. i. cap. 7.

§ Naturæ et Scripturæ Concordia.

De Ratione Communi Omnium Linguarum.

Essay on a Real Character, and Philosophical Language.

** Hermes, Book iii. cap. 2.

Rhetorical Grammar, page 9.

++ Works, vol. ii.

that while some are only of momentary duration, taking place during a sudden change in the conformation of the mouth, and are, therefore, incapable of prolongation; others, like the vowels, can be prolonged ad libitum, as long as a particular disposition of the mouth and a constant expiration are maintained.* The former are termed explosive, the latter continuous. It is an additional distinction between these two classes, and one which tends much to simplify any attempt to acquire them by means of studying the disposition of the organs essential for their right formation, that, in pronouncing the former class, the communication between the nostrils and the mouth is entirely closed, and the modifications of sound are produced in the mouth itself, by interrupting the current of air passing through its cavity. This is effected in three ways. First, the middle of the tougue is approximated to the arch of the palate, and this momentary occlusion ceases with the escape of the breath. In this way, the hard sound of g, as in gun, and k is produced. Second, The tongue is brought to the front of the palate, or to the upper teeth, as in the production of d and t; and, Third, The lips are approximated, and the mouth thus almost entirely closed, as in the letters 6 and p. It will be seen, that in each of these classes, two sounds are arranged. The distinction between these is, that one in each class is produced by a stronger aspiration than is necessary for the formation of the other. K, p, t, are just modifications of b, d, and g, effected by means of a stronger aspiration during their utterance. These letters seem in very general use among all nations. B and d are, however, unknown to the Chinese. G is not pronounced by the Mexicans,§ and p is not found in the Arabian language.

The second class, or continuous consonants, may be arranged in three orders: 1st, Where the whole oral canal is open. This is only in sounding the letter h. 2d, Where the mouth is completely shut, and the air passes simply through the nasal aperture. M, though generally considered a labial, from the lips being closed during its pronunciation, is one of this class. To it may be added n and ng, which letters in combination produce a

• Müller, op. cit. p. 1045.

It is a curious circumstance, noticed by Müller, in reference to the subject of stammering, that if the consonant preceding the impediment be one of the explosive class, (b, d, g, and p, t, k,) which do not admit of a continuous pronunciation, until the formation of the vowel is attained, it is more prone to be frequently repeated; thus, a person who stammers, says for bitter,' b-b-b bitter. But if it be a continuous consonant, as m, n, ng, f, ch, sh, r, l, or s, it is not necessarily repeated, because the sound of a continuous consonant can be prolonged ad libitum, until the vowel follows it; for example, in the word 'laughing.'

Wilkins' Essay, part iii. chap. 14.

§ Vossius, de Gr. cap. 27.