The history of a word has to be studied from the double point of view of sound and sense, or, to use more technical terms, phonetics and semantics. In the logical order of things it seems natural to deal first with the less interesting aspect, phonetics, the physical processes by which sounds are gradually transformed. Speaking generally, it may be said that phonetic changes are governed by the law of least resistance, a sound which presents difficulty being gradually and unconsciously modified by a whole community or race. With the general principles of phonetics I do not propose to deal, but a few simple examples will serve to illustrate the one great law on which this science is based.

The population of this country is educationally divided by the letter h into three classes, which we may describe as the confident, the anxious, and the indifferent. The same division existed in imperial Rome, where educated people sounded the aspirate, which completely disappeared from the every-day language of the lower classes, the so-called Vulgar Latin, from which the Romance languages are descended, so far as their working vocabulary is concerned. The anxious class was also represented. A Latin epigrammatist[42] remarks that since Arrius, prophetic name, has visited the Ionic islands, they will probably be henceforth known as the Hionic islands. To the disappearance of the h from Vulgar Latin is due the fact that the Romance languages have no aspirate. French still writes the initial h in some words by etymological reaction, e.g., homme for Old Fr. ome, and also at one time really had an aspirate in the case of words of Germanic origin, e.g., la honte, shame. But this h is no longer sounded, although it still, by tradition, prevents elision and liaison, mistakes in which are regarded much in the same way as a misplaced aspirate in English. The "educated" h of modern English is largely an artificial restoration; cf. the modern hotel-keeper with the older word ostler (see p. 164), or the family name Armitage with the restored hermitage.

We have dropped the k sound in initial kn, as in knave, still sounded in Ger. Knabe, boy. French gets over the difficulty by inserting a vowel between the two consonants, e.g., canif is a Germanic word cognate with Eng. knife. This is a common device in French when a word of Germanic origin begins with two consonants. Cf. Fr. dÉrive, drift, Eng. drive; Fr. varech, sea-weed, Eng. wrack. Harangue, formerly harengue, is Old High Ger. hring, Eng. ring, the allusion being to the circle formed by the audience. Fr. chenapan, rogue, is Ger. Schnapphahn, robber, lit. fowl-stealer. The shallop that "flitteth silken-sail'd, skimming down to Camelot," is Fr. chaloupe, probably identical with Du. sloep, sloop.

The general dislike that French has for a double consonant sound at the beginning of a word appears also in the transformation of all Latin words which began with sc, sp, st, e.g., scola > escole (École), spongia > esponge (Éponge), stabulum > estable (Étable). English words derived from French generally show the older form, but without the initial vowel, school, sponge, stable.

The above are very simple examples of sound change. There are certain less regular changes, which appear to work in a more arbitrary fashion and bring about more picturesque results. Three of the most important of these are assimilation, dissimilation, and metathesis.

Assimilation is the tendency of a sound to imitate its neighbour. The tree called the lime was formerly the line, and earlier still the lind. We see the older form in linden and in such place-names as Lyndhurst, lime wood. Line often occurred in such compounds as line-bark, line-bast, line-wood, where the second component began with a lip consonant. The n became also a lip consonant because it was easier to pronounce, and by the 17th century we generally find lime instead of line. We have a similar change in Lombard for Ger. lang-bart, long-beard, or, according to some, long-axe. For Liverpool we find also Litherpool in early records. If the reader attempts to pronounce both names rapidly, he will be able to form his own opinion as to whether it is more natural for Liverpool to become Litherpool or vice-versÂ, a vexed question with philologists. Fr. vÉlin, a derivative of Old Fr. veel (veau), calf, and venin, Lat. venenum, have given Eng. vellum and venom, the final consonant being in each case assimilated[43] to the initial labial. So also mushroom, Fr. mousseron, from mousse, moss.

Vulgar Lat. circare (from circa, around) gave Old Fr. cerchier, Eng. search. In modern Fr. chercher the initial consonant has been influenced by the medial ch. The m of the curious word ampersand, variously spelt, is due to the neighbouring p. It is applied to the sign &. I thought it obsolete till I came across it on successive days in two contemporary writers—

Children used to repeat the alphabet thus—"A per se A, B per se B," and so on to "and per se and." The symbol & is an abbreviation of Lat. et, written &.

Dissimilation is the opposite process. The archaic word pomander—

was formerly spelt pomeamber. It comes from Old Fr. pome ambre, apple of amber, a ball of perfume once carried by the delicate. In this case one of the two lip consonants has been dissimilated. A like change has occurred in Fr. nappe, cloth, from Lat. mappa, whence our napkin, apron (p. 113), and the family name Napier.

The sounds most frequently affected by dissimilation are those represented by the letters l, n, and r. Fr. gonfalon is for older gonfanon. Chaucer uses the older form, Milton the newer—

Gonfanon is of Germanic origin. It means literally "battle-flag," and the second element is cognate with English fane or vane (Ger. Fahne). Eng. pilgrim and Fr. pÈlerin, from Lat. peregrinus, illustrate the change from r to l, while the word frail, an osier basket for figs, is due to a change from l to r, which goes back to Roman times. A grammarian of imperial Rome named Probus compiled, about the 3rd or 4th century, A.D., a list of cautions as to mispronunciation. In this list we find "flagellum, non fragellum." In the sense of switch, twig, fragellum gave Old Fr. freel, basket made of twigs, whence Eng. frail; while the correct flagellum gave Old Fr. fleel (flÉau), whence Eng. flail. A Vulgar Lat. *mora, mulberry, from Lat. morus, mulberry tree, has given Fr. mÛre. The r of berry has brought about dissimilation in Eng. mulberry and Ger. Maulbeere. Colonel has the spelling of Fr. colonel, but its pronunciation points rather to the dissimilated Spanish form coronel which is common in Elizabethan English. Cotgrave has colonel, "a colonell, or coronell; the commander of a regiment."

The female name Annabel is a dissimilation of Amabel, whence Mabel. By confusion with the popular medieval name Orable, Lat. orabilis, Annabel has become Arabel or Arabella. Our level is Old Fr. livel, Vulgar Lat. *libellum, for libella, a plummet, diminutive of libra, scales. Old Fr. livel became by dissimilation nivel, now niveau. Many conjectures have been made as to the etymology of oriel. It is from Old Fr. oriol, a recess, or sanctum, which first occurs in an Anglo-Norman poem of the 12th century on Becket. This is from a Late Latin diminutive aulÆolum, a small chapel or shrine, which was dissimilated into aurÆolum.

Sometimes dissimilation leads to the disappearance of a consonant, e.g., Eng. feeble, Fr. faible, represents Lat. flebilis, lamentable, from flere, to weep. Fugleman was once flugelman, from Ger. FlÜgelmann, wing man, i.e., a tall soldier on the wing who exaggerated the movements of musketry drill for the guidance of the rest.

Metathesis is the transposition of two sounds. A simple case is our trouble, Fr. troubler, from Lat. turbulare. Maggot is for Mid. Eng. maddok, a diminutive of Anglo-Sax. maÞa; cf. Ger. Made, maggot. Kittle, in the phrase "kittle cattle," is identical with tickle; cf. Ger. kitzeln, to tickle. One theory for the origin of tankard is that it stands for *cantar, from Lat. cantharus, with which it corresponds exactly in meaning; e.g., cantharus, "a pot, a jugge, a tankerd" (Cooper); cantharo, a "tankard or jug that houldeth much" (Florio); canthare, "a great jugge, or tankard" (Cotgrave). The metathesis may be due to association with the name Tankard (Tancred).

Wattle and wallet are used indifferently in Mid. English for a little bag. Shakespeare no doubt had in mind the wattles of a cock or turkey when he made Gonzalo speak of mountaineers—

Fr. moustique is for earlier mousquite, from Span. mosquito, a diminutive from Lat. musca, a fly. Tinsel is Fr. Étincelle, spark, earlier estincele, which supposes a Lat. *stincilla for scintilla. The old word anlace, dagger, common in Mid. English and revived by Byron and Scott—

has provoked many guesses. Its oldest form, anelas, is a metathesis of the common Old Fr. alenas, dagger. This is formed from alÊne, of Germanic origin, cognate with awl; cf. cutlass, Fr. coutelas (p. 126). Beverage is from Old Fr. bevrage, or beuvrage, now breuvage, Vulgar Lat. *biberaticum, from bibere, to drink. Here, as in the case of level (p. 58), and search (p. 57), English preserves the older form. In Martello tower, from a fort taken by the British (1794) in Mortella, i.e., Myrtle, Bay, Corsica, we have vowel metathesis.

It goes without saying that such linguistic phenomena are often observed in the case of children and uneducated people. Not long ago the writer was urged by a gardener to embellish his garden with a ruskit arch. When metathesis extends beyond one word we have what is known as a Spoonerism, the original type of which is said to be—

We have seen (p. 57) that the letters l, n, r are particularly subject to dissimilation and metathesis. But we sometimes find them alternating without apparent reason. Thus banister is a modern form for the correct baluster.[44] This was not at first applied to the rail, but to the bulging colonnettes on which it rests. Fr. balustre comes, through Italian, from Greco-Lat. balaustium, a pomegranate flower, the shape of which resembles the supports of a balustrade. Cotgrave explains balustres as "ballisters; little, round and short pillars, ranked on the outside of cloisters, terraces, galleries, etc." Glamour is a doublet of grammar (see p. 145), and flounce was formerly frounce, from Fr. froncer, now only used of "knitting" the brows—

Fr. flibustier, whence our filibuster, was earlier fribustier, a corruption of Du. vrijbuiter, whence directly the Eng. freebooter.[45]

All words tend in popular usage to undergo a certain amount of shrinkage. The reduction of Lat. digitale, from digitus, finger, to Fr. dÉ, thimble (little thumb) is a striking example. The strong tonic accent of English, which is usually on the first, or root, syllable, brings about a kind of telescoping which makes us very unintelligible to foreigners. This is seen in the pronunciation of names such as Cholmondeley and Marjoribanks. Bethlehem hospital, for lunatics, becomes bedlam; Mary Magdalene, taken as a type of tearful repentance, gives us maudlin, now generally used of the lachrymose stage of intoxication. Sacristan is contracted into sexton. Fr. paralysie becomes palsy, and hydropisie becomes dropsy. The fuller form of the word usually persists in the literary language, or is artificially introduced at a later period, so that we get such doublets as proctor and procurator.

In the case of French words which have a prefix, this prefix is very frequently dropped in English, e.g., raiment for arrayment; while suffixes, or final syllables, often disappear, e.g., treasure trove, for Old Fr. trovÉ (trouvÉ), or become assimilated to some familiar English ending, e.g., parish, Fr. paroisse, skirmish, Fr. escarmouche; cartridge, Fr. cartouche, partridge, Fr. perdrix. A good example of such shrinkage is the word vamp, part of a shoe, Old Fr. avant-pie (pied), which became Mid. Eng. vampey, and then lost its final syllable. We may compare vambrace, armour for the forearm, Fr. avant-bras, vanguard, Fr. avant-garde, often reduced to van—

and the obsolete vaunt-courier, forerunner—

When the initial vowel is a-, its loss may have been helped by confusion with the indefinite article. Thus for anatomy we find atomy, for a skeleton or scarecrow figure, applied by Mistress Quickly to the constable (2 Henry IV., v. 4). Peal is for appeal, call; mend for amend, lone for alone, i.e., all one. Peach, used by Falstaff—

is for older appeach, related to impeach. Size, in all its senses, is for assize, Fr. assise, with a general meaning of allowance or assessment, from Fr. asseoir, to put, lay. Sizars at Cambridge are properly students in receipt of certain allowances called sizings. With painters' size we may compare Ital. assisa, "size that painters use" (Florio). We use the form assize in speaking of the "sitting" of the judges, but those most familiar with this tribunal speak of being tried at the 'sizes. The obsolete word cate, on which Petruchio plays—

is for earlier acate, an Old French dialect form corresponding to modern Fr. achat, purchase. The man entrusted with purchasing was called an acatour or catour (whence the name Cator), later cater, now extended to caterer, like fruiterer for fruiter, poulterer for poulter and upholsterer for upholdster or upholder.[46]

Limbeck has been squeezed out by the orthodox alembic—

and prentice has given way to apprentice. Tire and attire both survive, and maze persists by the side of amaze with the special sense which I have heard a Notts collier express by puzzle-garden (cf. Ger. Irrgarten). Binnacle is a corruption, perhaps due to association with bin, of earlier bittacle, from Lat. habitaculum, a little dwelling. It may have come to us through Fr. habitacle or Port. bitacola, "the bittacle, a frame of timber in the steerage, where the compass is placed on board a ship" (Vieyra, Port. Dict., 1794). As King of Scotland, King George has a household official known as the limner, or painter. For limner[47] we find in the 15th century lumner and luminour, which is aphetic for alluminour, or enlumineur. Cotgrave, s.v. enlumineur de livres, says, "we call one that coloureth, or painteth upon, paper, or parchment, an alluminer."

But confusion with the article is not necessary in order to bring about aphesis. It occurs regularly in the case of words beginning with esc, esp, est, borrowed from Old French (see p. 56). Thus we have squire from escuyer (Êcuyer), skew from Old Fr. eschuer, to dodge, "eschew," ultimately cognate with Eng. shy, spice from espice (Épice), sprite from esprit, stage from estage (Étage), etc. In some cases we have the fuller form also, e.g., esquire, eschew; cf. sample and example. Fender, whether before a fireplace or slung outside a ship, is for defender; fence is always for defence, either in the sense of a barrier or in allusion to the noble art of self-defence.[48] The tender of a ship or of a locomotive is the attender, and taint is aphetic for attaint, Fr. atteinte, touch—

Puzzle was in Mid. Eng. opposaile, i.e., something put before one. We still speak of "a poser."

Spital, for hospital, survives in Spitalfields, and Spittlegate at Grantham and elsewhere. Crew is for accrewe (Holinshed). It meant properly a reinforcement, lit. on-growth, from Fr. accroÎtre, to accrue. In recruit, we have a later instance of the same idea. Fr. recrue, recruit, from recroÎtre, to grow again, is still feminine, like many other military terms which were originally abstract or collective. Cotgrave has recreuË, "a supplie, or filling up of a defective company of souldiers, etc." We have possum for opossum, and coon for racoon, and this for arrahacoune, which I find in a 16th-century record of travel; cf. American skeeter for mosquito. In these two cases we perhaps have also the deliberate intention to shorten (see p. 66), as also in the obsolete Australian tench, for the aphetic 'tentiary, i.e., penitentiary. With this we may compare 'tec for detective.

Drawing-room is for withdrawing room, and only the final t of saint is left in Tooley St., famed for its three tailors, formerly Saint Olave Street, and tawdry. This latter word is well known to be derived from Saint Audrey's fair. It was not originally depreciatory—

and the full form is recorded by Palsgrave, who has Seynt Andries (read Audrie's) lace, "cordon." The verb vie comes from Old Fr. envier, to challenge, Lat. invitare, whence the phrase À l'envi l'un de l'autre, "in emulation one of the other" (Cotgrave); cf. gin (trap), Fr. engin, Lat. ingenium. The prefix dis or des is lost in Spencer (see p. 165), spite, splay, sport, stain, etc.

In drat, formerly 'od rot, zounds for God's wounds, 'sdeath, odsbodikins, etc., there is probably a deliberate avoidance of profanity. The same intention appears in Gogs—

Cf. Fr. parbleu for par Dieu, and Ger. Potz for Gottes.

This English tendency to aphesis is satirised in a French song of the 14th century, intentionally written in bad French. Thus, in the line—

Glais is for Anglais and vauchier is for chevauchier (chevaucher), to ride on a foray. The literary language runs counter to this instinct, though Shakespeare wrote haviour for behaviour and longing for belonging, while such forms as billiments for habiliments and sparagus for asparagus are regular up to the 18th century. Children keep up the national practice when they say member for remember and zamine for examine. It is quite certain that baccy and tater would be recognised literary forms if America had been discovered two centuries sooner or printing invented two centuries later.

Many words are shortened, not by natural and gradual shrinkage, but by deliberate laziness. The national distaste for many syllables appears in wire for telegram, the Artful Dodger's wipe for the clumsy pocket handkerchief, soccer for association, and such portmanteau words as squarson, an individual who is at once squire and parson, or Bakerloo for Baker St. and Waterloo.

The simplest way of reducing a word is to take the first syllable and make it a symbol for the rest. Of comparatively modern formation are pub and Zoo, with which we may compare Bart's, for Saint Bartholomew's, Cri, Pav, "half a mo'" bike, and even paj, for pageant.

This method of shortening words was very popular in the 17th century, from which period date cit(izen), mob(ile vulgus), the fickle crowd, and, pun(digrion). We often find the fuller mobile used for mob. The origin of pundigrion is uncertain. It may be an illiterate attempt at Ital. puntiglio, which, like Fr. pointe, was used of a verbal quibble or fine distinction. Most of these clipped forms are easily identified, e.g., cab(riolet), gent(leman), hack(ney), vet(erinary surgeon). Cad is for Scot. caddie, errand boy, now familiar in connection with golf, and caddie is from Fr. cadet, younger. The word had not always the very strong meaning we now associate with it. Among Sketches by Boz is one entitled—

where cad means conductor. On tick, for on ticket, is found in the 17th century. We may compare the more modern biz and spec. Brig is for brigantine, Ital. brigantino, "a kinde of pinnasse or small barke called a brigantine" (Florio). The original meaning is pirate ship; cf. brigand. Wag has improved in meaning. It is for older waghalter. Cotgrave has baboin (babouin), "a trifling, busie, or crafty knave; a crackrope, waghalter, etc." The older sense survives in the phrase "to play the wag," i.e. truant. For the "rope" figure we may compare Scot. hempie, a minx, and obsolete Ital. cavestrolo, a diminutive from Lat. capistrum, halter, explained by Florio as "a wag, a haltersacke." Modern Ital. capestro is used in the same sense. Crack-rope is shortened to crack. Justice Shallow remembered Falstaff breaking Skogan's head—

Chap is for chapman, once in general use for a merchant and still a common family name. It is cognate with cheap, chaffer, and Ger. kaufen, to buy, and probably comes from Lat. caupo, tavern keeper. We have the Dutch form in horse-coper, and also in the word coopering, the illicit sale of spirits by Dutch boats to North Sea fishermen.[50] Merchant was used by the Elizabethans in the same way as our chap. Thus the Countess of Auvergne calls Talbot a "riddling merchant" (1 Henry VI., ii. 3). We may also compare Scot. callant, lad, from the Picard form of Fr. chaland, customer—

and our own expression "a rum customer," reduced in America to "a rum cuss." Hock, for Hochheimer, wine from Hochheim, occurs as early as Beaumont and Fletcher; and rum, spirit, is for earlier rumbullion, of obscure origin. Gin is for geneva, a corruption of Fr. geniÈvre, Lat. juniperus, with the berries of which it is flavoured. The history of grog is more complicated. The stuff called grogram, earlier grograyne, is from Fr. gros grain, coarse grain. Admiral Vernon (18th century) was called by the sailors "Old Grog" from his habit of wearing grogram breeches. When he issued orders that the regular allowance of rum was henceforth to be diluted with water, the sailors promptly baptized the mixture with his nickname.

Sometimes the two first syllables survive. We have navvy for navigator, brandy for brandywine, from Du. brandewyn, lit. burnt wine, and whisky for usquebaugh, Gaelic uisge-beatha, water of life (cf. eau-de-vie), so that the literal meaning of whisky is very innocent. It has a doublet in the river-name Usk. Before the 18th century usquebaugh is the regular form. In the following passage the Irish variety is referred to—

Canter is for Canterbury gallop, the pace of pilgrims riding to the shrine of St Thomas. John Dennis, known as Dennis the Critic, says of Pope—

In bugle, for bugle-horn, lit. wild-ox-horn, Old Fr. bugle, Lat. buculus, a diminutive of bos, ox, we have perhaps rather an ellipsis, like waterproof (coat), than a clipped form—

Patter is no doubt for paternoster—

and the use of the word marble for a toy sometimes made of that stone makes it very probable that the alley, most precious of marbles, is short for alabaster.

Less frequently the final syllable is selected, e.g., bus for omnibus, loo for lanterloo, variously spelt in the 17th and 18th centuries—

Fr. lanturelu was originally the meaningless refrain or "tol de rol" of a popular song in Richelieu's time. Van is for caravan, a Persian word, properly a company of merchants or ships travelling together, "also of late corruptly used with us for a kind of waggon to carry passengers to and from London" (Blount, Glossographia, 1674). Wig is for periwig, a corruption of Fr. perruque, of obscure origin. With the 17th century 'varsity, for university, we may compare Sam Weller's 'Tizer, for Morning Advertiser.

Christian names are treated in the same way. Alexander gives Alec and Sandy, Herbert, 'Erb or Bert. Ib (see p. 172) was once common for Isabella, while the modern language prefers Bella; Maud for Matilda is a telescoped form of Old Fr. Maheut, while 'Tilda is perhaps due to unconscious aphesis, like Denry—

Among conscious word-formations may be classed many reduplicated forms, whether riming, as hurly-burly, or alliterative, as tittle-tattle, though reduplication belongs to the natural speech of children, and, in at least one case, Fr. tante, from ante-ante, Lat. amita, the baby word has prevailed. In a reduplicated form only one half as a rule needs to be explained. Thus seesaw is from saw, the motion suggesting two sawyers at work on a log. Zigzag, from French, and Ger. zickzack are of unknown origin. Shilly-shally is for shill I, shall I? Namby-pamby commemorates the poet Ambrose Philips, who was thus nicknamed by Pope and his friends. The weapon called a snickersnee—

is of Dutch origin and means something like "cut and thrust." It is usually mentioned in connection with the Hollanders—

Here the reduplication is only apparent, for the older form was to stick or snee, representing the Dutch verbs steken, to thrust, snijden or snijen, to cut. The initial of the first verb has been assimilated to that of the second—

Reduplication is responsible for pickaback, earlier pickpack, from pack, bundle. The modern form is due to popular association with back.

Occasionally we have what is apparently the arbitrary prefixing of a consonant, e.g., spruce for pruce (p. 48). Dapple gray corresponds so exactly to Fr. gris pommelÉ, Mid. Eng. pomeli gris, Ger. apfelgrau, and Ital. pomellato, "spotted, bespeckled, pide, dapple-graie, or fleabitten, the colour of a horse" (Florio), that it is hard not to believe in an unrecorded *apple-gray, especially as we have daffodil for earlier affodil, i.e., asphodel. Cotgrave has asphodile (asphodÈle), "the daffadill, affodill, or asphodill, flower." The playful elaboration daffadowndilly is as old as Spenser.

---