Letters are conventional marks or visible signs of the elemental sounds of spoken language. The earliest symbols of sounds represented syllables rather than simple sounds, and it was only gradually that syllables were reduced to their ultimate elements, all alphabets yet bearing marks of the syllabary origin and displaying various imperfections of excess and defect. The distinction between vowels (φωνόντα and φωνήντα) and consonants (ἄφωνα γράμματα) is as old as Plato's Cratylus, but the earlier methods of classification, which offered no definite line of demarcation, have given place to a more scientific method and more precise terminology. A consonant is the sound or noise resulting when the breath is closely squeezed or stopped at some part of the mouth or breath passage. This passage has two outlets; one at the lips, the other at the nose. In the case of some consonants the passage is closed at a given point—e.g. at the lips in p, b; at the teeth in t, d; at the palate in k, g; while the nose-passage is closed by the bottom (uvula) of the soft palate. In the case of other consonants the passage is not closed but only narrowed, and the breath sounds against the narrow walls, as in f, s, ch, &c. The nose channel is left open in pronouncing n, m. In the case of vowels the breath is not checked or closely narrowed at a given point, but passes freely through the mouth-passage, and they may be classified according to the place of their articulation in this passage owing to the horizontal or vertical movements of the tongue. They may also be distinguished as open (or wide) and shut (or narrow); in the latter there is more convexity of the tongue and a sense of effort in pronunciation. Consonants, again, are classified according to the point in the mouth-passage at which they are articulated. Furthest back we have the guttural sounds, with articulation of the tongue and the soft palate (velum), hence called the Velar Gutturals, as the sounds of k and g before the vowel sounds a and o, ch in German ach, and ng in sing. Further forward we have the Palatals, which result from the contact of the tongue and hard palate, as k and g before e and i. Next come the Dentals, where the tongue, teeth, and front of the palate are the instruments of articulation—e.g. t, th, d, sh, r. The Labials, or lip-sounds, are those which are formed either by both the lips or by the under lip and upper teeth, p, b, m, f, v.
Again, consonants may also be classified according to the form of their articulation. When there is a complete check or closure of the mouth-passage at the point of articulation, consonants are called Mutes or Stopped or Explosives, as k, g, t, d, p, b. The ancient grammarians divided mutes into Ténues and Mediæ, the former including k, t, p, the latter g, d, b. When the mouth-passage is narrowed but not completely closed at the point of articulation, we have the Open or Fricative consonants, such as r, s, f, z, v, the last four of which are also called Spirants. When the passage is stopped in the middle but kept open at the sides, we have a Divided sound, such as l. When the nose-passage is left open we have the Nasals m and n. N takes its character from the nature of the neighbouring sounds—e.g. it is guttural in sink, but dental in tent.
Again, consonants may be divided into Voiced and Voiceless. The voiced or soft consonants are pronounced with a vibration of the vocal chords. To this class belong the Nasals n, m; the Liquids l and r; the Mediæ g, d, b; the Aspirated Mediæ gh, dh, bh; and the Spirants z, j, v. The Voiceless or Surd consonants are the Tenues p, t, k, q, the Aspirated tenues kh, th, ph, and the Spirants s, f.
There are some sounds which can play the part of either consonants or vowels, as the Semi-vowels i and u. In the same way, m, n, r, l between vowels are purely consonantal, but before or after consonants they can be vocalised or made sonant. Besides all these different sounds we have the introductory glides, represented by the smooth and rough breathings.
Such is the variety and complexity of the different sounds which are attempted to be represented in the letters of alphabets. These, however, merely satisfy roughly practical needs, and in no case represent all the sounds actually employed in any language. Accentuation, which is an important element of speech, is rarely denoted, and still more rarely is quantity. The alphabetic symbols remain comparatively unchanged, while the sounds they represent are constantly changing. Hence the literary spelling often corresponds but indifferently with the actual sounds of the words, and in English especially this has grown into so heavy an additional burden on the memories of learners that many scholars have been led to advocate the adoption of a scientific phonetic system of spelling. Thus, in English there are at least forty-three easily distinguished sounds, while the written alphabet has only twenty-six letters or symbols to represent these. Again, our alphabet is redundant, containing three superfluous letters, c, g, x, so that there are actually but twenty-three letters wherewith to represent forty-three sounds!
Many persons amongst ourselves are unable to pronounce certain letters, as l and r; others change r or l to d, and we observe that children for some time habitually substitute dentals for gutturals, as tut for cat, tiss for kiss. This is quite distinct from phonetic diversification which follows certain definite laws, the observation of which formed the basis of scientific etymology. We find that some languages lack certain sounds which to us seem indispensable. Thus, the Mohawks and Hurons employ no labials; the Society Islanders are destitute of gutturals (the name of Captain Cook became Tute), and the Australian dialects of s, as are also several of the Polynesian languages, where its place is taken by h (cf. Lat. sul and Gr. hals). Again, d is never used in Chinese, Mexican, or Peruvian; n is absent in the language of the Hurons; and even in so perfect a language as Sanskrit we have no f, no soft sibilants, no short e and o. Greek has no y, no w, no soft sibilants; Latin has no soft sibilants, no native θ, φ, χ; English is deficient in guttural breathings like the German ach and ich, although these are plentiful enough in Scotland. High German has no w like the English w in wind, no th, dh, ch, j. While Sanskrit has no f, Arabic has no p. The letter f is absent in Finnish (despite the name—given it by its neighbours), Lithuanian, Tamil, Mongolian, and Burmese. No Chinaman ever pronounces r, Christ being rendered Ki li sse tu. It is also absent in the language of the Hurons, Mexicans, and Kaffirs. Max Müller gives the following enumeration of the consonants in a few alphabets: Hindustani has 48 consonants, of which 13 are classical Sanskrit aspirates, nasals, and sibilants, and 14 Arabic letters. Sanskrit has 37 consonants (or with the Vedic l and lh, 39); Turkish, 32 (of which only 25 are really Turkish); and Persian, 31 (of which 22 are really Persian, the rest Arabic). Arabic has 28; the Kaffir (Zulu), 26, besides the clicks. Hebrew has 23; English, 20; Greek, 17 (3 compound); Latin, 17 (1 compound); Mongolian, 17 or 18; Finnish, 11; Polynesian, 10 native consonantal sounds; some Australian languages, 8; of the Melanesian languages the poorest, the Duauru, has 12, others 13, 14, and more. Again, some races find exceeding difficulty in distinguishing some of our sounds. Thus, the Sandwich Islanders habitually confuse k and t, and we find the same word written by Protestant missionaries with k, by French missionaries with t. Even in Canada the lower classes say mékier and moikié for métier and moitié. And even so respected a lexicographer as Noah Webster actually says in the Introduction to his Dictionary that in English the letters cl are pronounced as if written tl, and gl as if dl.
The foregoing is mainly taken from the following: J. E. King and C. Cookson's admirable Principles of Sound and Inflection (1888), their Introduction to the Comparative Grammar of Greek and Latin (1890), and Max Müller's Lectures on the Science of Language (2d series, lect. iv. 1864). See also the articles in the present work on ALPHABET, GRAMMAR, GRIMM'S LAW, PHONETICS, SPELLING, and on each of the letters of the alphabet.