The sound, spelling, and even the meaning of a word are often perverted by influences to which the collective name of folk-etymology has been given. I here use the term to include all phenomena which are due to any kind of misunderstanding of a word. A word beginning with n sometimes loses this sound through its being confused with the n of the indefinite article an. Thus an adder and an auger are for a nadder (cf. Ger. Natter) and a nauger, Mid. Eng. navegor, properly an instrument for piercing the nave of a wheel. Apron was in Mid. English naprun, from Old Fr. naperon, a derivative of nappe, cloth. The aitch-bone was formerly the nache-bone, from Old Fr. nache, buttock, Vulgar Lat. *natica for nates. Nache is still used by French butchers. Humble-pie is a popular perversion of umble-pie, i.e., a pie made from the umbles, or inferior parts of the stag. But umble is for earlier numble, Old Fr. nomble, formed, with dissimilation, from Lat. lumbulus, diminutive of lumbus, loin; cf. niveau (p. 58). Thus humble-pie has etymologically no connection with humility. Umpire represents Old Fr. non per (pair), not equal, the umpire being a third person called in when arbitrators could not agree. This appears clearly in the following extract from a medieval letter—

For the sense we may compare Span. tercero, "the third, a broaker, a mediator" (Percyvall). An eyas falcon is for a neyas falcon, Fr. niais, foolish, lit. nestling, related to nid, nest. Rosenkrantz uses it in the literal sense—

Somewhat similar is the loss in French of initial a in la boutique for l'aboutique, Greco-Lat. apotheca, and la Pouille for l'Apouille, Apulia, or of the initial l in ounce, a kind of tiger-cat, from Fr. once, earlier lonce, "the ounce, a ravenous beast" (Cotgrave), taken as l'once. It is almost a doublet of lynx.

The opposite has happened in the case of a newt for an ewt and a nick-name for an eke-name. Eke, also, occurs in the first stanza of John Gilpin. It is cognate with Ger. auch, also, and Lat. augere, to increase. Nuncle, the customary address of a court fool to his superiors—

is for mine uncle. We also find naunt. Nonce occurs properly only in the phrase for the nonce, which is for earlier for then ones, where then is the dative of the definite article. Family names like Nash, Nokes are aphetic for atten ash, at the ash, atten oakes, at the oaks. The creation of such forms was perhaps helped by our tendency to use initial n in Christian names, e.g., Ned for Edward, Noll for Oliver, Nell for Ellen.

Agglutination of the definite article is common in French, e.g., lingot, ingot, lierre, ivy, for l'ierre, Lat. hedera, and the dialect lÉvier, sink, for Évier, Lat. aquarium, whence Eng. ewer. The derivation of Fr. landier, andiron, is unknown, but the iron of the English word is due to folk-etymology. Such agglutination occurs often in family names such as Langlois, lit. the Englishman, Lhuissier, the usher (see p. 90), and some of these have passed into English, e.g., Levick for l'ÉvÊque, the bishop.

The two words alarm and alert include the Italian definite article. The first is Ital. all'arme, to arms, for a le arme, and the second is all'erta for alla (a la) erta, the last word representing Lat. erecta. With rolled r, alarm becomes alarum, whence the aphetic larum—

Ger. LÄrm, noise, is the same word. In Luther's time we also find Allerm.

We have the Arabic definite article in a great many words borrowed from Spanish. Alcalde, or alcade, and alguazil, common in Elizabethan literature, are two old friends from the Arabian Nights, the cadi and the wazir, or vizier. The Arabic article also occurs in acton, Old Fr. auqueton, now hoqueton, for al qutn (cotton), because originally used of a wadded coat—

In alligator, Span. el lagarto, the lizard, from Lat. lacertus, we have the Spanish definite article. See also lariat, p. 24.

A foreign word ending in a sibilant is sometimes mistaken for a plural. Thus Old Fr. assets (assez), enough, Lat. ad satis, has given Eng. assets, plural, with a barbarous, but useful, singular asset. Cherry is for cheris, from a dialect form of Fr. cerise, and sherry for sherris, from Xeres in Spain (see p. 51). Falstaff opines that—

Pea is a false singular from older pease, Lat. pisum. Perhaps the frequent occurrence of pease-soup, not to be distinguished from pea-soup, is partly responsible for this mistake. Marquee, a large tent, is from Fr. marquise. With this we may class the heathen Chinee and the Portugee. Milton wrote correctly of—

It has been ingeniously suggested that Yankee has been derived in the same way from Du. Jan Kees, John Cornelius, supposed to have been a nickname for early Dutch colonists. It is more probably the Dutch dim. Janke, i.e. Johnny. The vulgarism shay for chaise[90] is of similar formation. Corp, for corpse, is also used provincially. Kickshaws is really a singular from Fr. quelque chose—

Cotgrave spells it quelkchoses (s.v. fricandeau).

Skate has a curious history. It is a false singular from Du. schaats. This is from escache, an Old French dialect form of Échasse, stilt, which was used in the Middle Ages for a wooden leg. It is of German origin, and is related to shank. Cf., for the sense development, Eng. patten, from Fr. patin, a derivative of patte, foot, cognate with paw. Skates are still called pattens by the fenmen of Cambridgeshire. We also had formerly a doublet from Old Fr. escache directly, but in the older sense, for Cotgrave has eschasses (Échasses), "stilts, or scatches to go on." Row, a disturbance, belongs to rouse, a jollification—

of uncertain origin, but probably aphetic for carouse, drink carouse being wrongly separated as drink a rouse. The bird called a wheatear was formerly called wheatears, a corruption of a name best explained by its French equivalent cul blanc, "the bird called a whittaile" (Cotgrave). We may compare the bird-name redstart, where start means rump.

Conversely a word used in the plural is sometimes regarded as a singular, the result being a double plural. Many Latin neuter plurals were adopted into French as feminine singulars, e.g., cornua, corne, horn; labra, lÈvre, lip; vela, voile, sail. It is obvious that this is most likely to occur in the case of plurals which are used for a pair, or set, of things, and thus have a kind of collective sense. Breeches or breeks is a double plural, Anglo-Sax. brec being already the plural of broc. In Mid. English we still find breche or breke used of this garment. Trousers was earlier trouses, plural of trouse, now trews, and was used especially of Irish native costume. The latest researches throw doubt on the identity of these words with Fr. trousse, a page's short breeches. The etymology which now finds most favour is Irish and Gaelic triubhas, from Late Lat. tubracci or tribracci, which is supposed to be a corrupted compound from tibia, leg, shank, and braccÆ, breeches. Bodice is for bodies, as pence is for pennies. Cotgrave explains corset by "a paire of bodies for a woman," and the plural sense occurs as late as Harrison Ainsworth—

Trace, of a horse, is the Old Fr. plural trais[91] (traits) of trait, "a teame-trace" (Cotgrave). Apprentice is the plural of Fr. apprenti, formerly apprentif, a derivative of apprendre, to learn, hence a disciple. Invoice is the plural of the obsolete invoy, from Fr. envoi, sending.

In the Grecian steps, at Lincoln, we have a popular corruption of the common Mid. Eng. and Tudor grece, grese, plural of Old Fr. grÉ, step, from Lat. gradus. Shakespeare spells it grize—

Scot. brose, or brewis, was in Mid. Eng. browes, from Old Fr. brouez, plural of brouet, a word cognate with our broth. From this association comes perhaps the use of broth as a plural in some of our dialects. Porridge, not originally limited to oatmeal, seems to be combined from pottage and Mid. Eng. porrets, plural of porret, leek, a diminutive from Lat. porrum. Porridge is sometimes used as a plural in Scottish—

and in the northern counties of England people speak of taking "a few" porridge, or broth. Baize, now generally green, is for earlier bayes, the plural of the adjective bay, now used only of horses; cf. Du. baai, baize. The origin of the adjective bay, Fr. bai, forms of which occur in all the Romance languages, is Lat. badius, "of bay colour, bayarde" (Cooper). Hence the name Bayard, applied to FitzJames' horse in The Lady of the Lake (v. 18), and earlier to the steed that carried the four sons of Aymon. Quince is the plural of quin, from the Norman form of Old Fr. coin (coing), which is derived from Gk. ??d?????. Truce is the plural of Mid. Eng. trewe (lit. truth, faith) with the same meaning. Already in Anglo-Saxon it is found in the plural, probably as rendering Lat. induciÆ. Lettuce, Mid. Eng. letows, seems also to be a plural, from Fr. laitue, Lat. lactuca.

Earnest in the sense of pledge—

has nothing to do with the adjective earnest. It is the Mid. Eng. ernes, earlier erles, which survives as arles in some of our dialects. The verb to earl is still used in Cumberland of "enlisting" a servant with a shilling in the open market. The Old French word was arres or erres, now written learnedly arrhes, a plural from Lat. arrha, "an earnest penny, earnest money" (Cooper). The existence of Mid. Eng. erles shows that there must have been also an Old French diminutive form. For the apparently arbitrary change of l to n we may compare banister for baluster (see p. 60).

The jesses of a hawk—

were the thongs by which it was held or "thrown" into the air. Jess is the Old Fr. jes, the plural of jet, from jeter, to throw. In Colman's Elder Brother we read of a gentleman who lounged and chatted, "not minding time a souse," where souse is the plural of Fr. sou, halfpenny. From Fr. muer, to moult, Lat. mutare, we get Fr. mue, moulting, later applied to the coop or pen in which moulting falcons were confined, whence the phrase "to mew (up)"—

When, in 1534, the royal mews, or hawk-houses, near Charing Cross were rebuilt as stables, the word acquired its present meaning.

Chess, Old Fr. esches (Échecs), is the plural of check, Fr. Échec, from Persian shah, king. By analogy with the "game of kings," the name jeu des dames was given in French to draughts, still called dams in Scotland. Draught, from draw, meant in Mid. English a "move" at chess. The etymology of tweezers can best be made clear by starting from French Étui, a case, of doubtful origin. This became in English etwee, or twee, e.g., Cotgrave explains estui (Étui) as "a sheath, case, or box to put things in; and (more particularly) a case of little instruments, as sizzars, bodkin, penknife, etc., now commonly termed an ettwee." Such a case generally opens book-fashion, each half being fitted with instruments. Accordingly we find it called a surgeon's "pair of twees," or simply tweese, and later a "pair of tweeses." The implement was named from the case (cf. Fr. boussole, p. 127), and became tweezers by association with pincers (Fr. pinces), scissors, etc.

The form of a word is often affected by association with some other word with which it is instinctively coupled. Thus larboard, for Mid. Eng. ladeboard, i.e. loading side, is due to starboard, steering side. Bridal, for bride-ale, from the liquid consumed at marriage festivities, is due to analogy with betrothal, espousal, etc. A 16th-century Puritan records with satisfaction the disappearance of—

Rampart is from Old Fr. rempar, a verbal noun from remparer, to repair; cf. Ital. riparo, "a rampire, a fort, a banke" (Florio). By analogy with Old Fr. boulevart (boulevard), of German origin and identical with our bulwark,[93] rempar became rempart. The older English form occurs in the obsolete rampier or rampire, which survive in the dialect ramper, embankment, causeway. For the spelling rampire we may compare umpire (p. 113). The apple called a jenneting, sometimes "explained" as for June-eating, was once spelt geniton, no doubt for Fr. jeanneton, a diminutive of Jean. It is called in French pomme de Saint-Jean, and in German Johannisapfel, because ripe about St John's Day (June 24). The modern form is due to such apple names as golding, sweeting, codlin, pippin.

In the records of medieval London we frequently come across the distinction made between people who lived "in the city," Anglo-Fr. deinz (dans) la citÉ, and "outside the city," Anglo-Fr. fors (hors) la citÉ. The former were called deinzein, whence our denizen, and the latter forein.[94] The Anglo-French form of modern Fr. citoyen was citein, which became citizen by analogy with denizen. The following passage from a medieval London by-law shows how rigid was the division between "denizen" and "foreign" traders—

Even words which have opposite meanings may affect each other by association. Thus Lat. reddere, to give back, became Vulgar Lat. *rendere by analogy with prendere (prehendere), to take away; hence Fr. rendre. Our word grief, from Fr. grief, is derived from a Vulgar Lat. *grevis, heavy (for gravis), which is due to levis, light.

The plural of titmouse is now usually titmice, by analogy with mouse, mice, with which it has no connection. The second part of the word is Anglo-Sax. mase, used of several small birds. It is cognate with Ger. Meise, titmouse, and Fr. mÉsange, "a titmouse, or tittling" (Cotgrave). Tit, of Norse origin, is applied to various small animals, and occurs also as a prefix in titbit or tidbit. Cf. tomtit (p. 37).

The Spanish word salva, "a taste, a salutation" (Percyvall), was used of the pregustation of a great man's food or drink. We have given the name to the tray or dish from which the "assay" was made, but, by analogy with platter, trencher, we spell it salver. In another sense, that of a "salutation" in the form of a volley of shot, we have corrupted it into salvo. With the use of Span. salva we may compare that of Ital. credenza, lit. faith, "the taste or assaie of a princes meate and drinke" (Florio), whence Fr. crÉdence, side-board, used in English only in the ecclesiastical compound credence table, and Ger. credenzen, to pour out.

In spoken English the ending -ew, -ue, of French origin, has been often changed to -ee, -ey. Thus pedigree was formerly pedigrew (see p. 77). The fencing term veney—

also spelt venew, is from Fr. venue, "a venny in fencing" (Cotgrave). Carew has become Carey, and Beaulieu, in Hampshire, is called Bewley. Under the influence of these double forms we sometimes get the opposite change, e.g., purlieu, now generally used of the outskirts of a town, is for purley, a strip of disforested woodland. This is a contraction of Anglo-Fr. pour-allÉe, used to translate the legal Lat. perambulatio, a going through. A change of venue[96] is sometimes made when it seems likely that an accused person, or a football team, will not get justice from a local jury. This venue is in law Latin vicinetum, neighbourhood, which gave Anglo-Fr. visnÉ, and this, perhaps by confusion with the venire facias, or jury summons, became venew, venue.

In the preceding examples the form has been chiefly affected. In the word luncheon both form and meaning have been influenced by the obsolete nuncheon, a meal at noon, Mid. Eng. none-chenche, for *none-schenche, noon draught, from Anglo-Sax. scencan,[97] to pour. Drinking seems to have been regarded as more important than eating, for in some counties we find this nuncheon replaced by bever, the Anglo-French infinitive from Lat. bibere, to drink. Lunch, a piece or hunk, especially of bread, also used in the sense of a "snack" (cf. Scot. "piece"), was extended to luncheon by analogy with nuncheon, which it has now replaced—

The term folk-etymology is often applied in a narrower sense to the corruption of words through a mistaken idea of their etymology or origin. The tendency of the uneducated is to distort an unfamiliar or unintelligible word into some form which suggests a meaning. Some cases may have originated in a kind of heavy jocularity, as in sparrow-grass for asparagus or sparagus (see p. 66), or Rogue Riderhood's Alfred David for affidavit—

In others there has been a wrong association of ideas, e.g., the primrose, rosemary, and tuberose have none of them originally any connection with the rose. Primrose was earlier primerole, an Old French derivative of Latin primula; rosemary, French romarin, is from Lat. ros marinus, sea-dew; tuberose is the Latin adjective tuberosus, bulbous, tuberous. Or attempts are made at translation, such as Sam Weller's Have his carcase for Habeas Corpus, or the curious names which country folk give to such complaints as bronchitis, erysipelas, etc. To this class belongs Private Mulvaney's perversion of locomotor ataxy—

Our language is, owing to our borrowing habits, particularly rich in these gems. Examples familiar to everybody are crayfish from Fr. Écrevisse, gilly-flower from Fr. giroflÉe, shame-faced for shamefast. Other words in which the second element has been altered are causeway, earlier causey, from the Picard form of Fr. chaussÉe, Lat. (via) calciata, i.e., made with lime, calx; penthouse, for pentice, Fr. appentis, "the penthouse of a house" (Cotgrave), a derivative of Old Fr. appendre, to hang to. Fr. hangar, a shed, now introduced into English by aviators as unnecessarily as garage by motorists, may also contain the same idea of "hanging."

In hiccough, for earlier hickup, an onomatopoeic word, the spelling, suggested by cough, has not affected the pronunciation. Surcease is Fr. sursis, past participle of surseoir, "to surcease, pawse, intermit, leave off, give over, delay or stay for a time" (Cotgrave), Lat. supersedere. Taffrail has been confused with rail, its older form being tafferel, from Du. tafereel, diminutive of tafel, picture, from Lat. tabula. It meant originally the flat part of the stern of a ship ornamented with carvings or pictures. This is called tableau in nautical French. Fr. coutelas, an augmentative of Old Fr. coutel (couteau), knife, gave Eng. cutlass, which has no more etymological connection with "cutting" than a cutler, Fr. coutelier, or a cutlet, Fr. cÔtelette, little rib, Lat. costa. Cutlas was popularly corrupted into curtal-axe, the form used by Rosalind—

We have a similar corruption in pick-axe, Mid. Eng. pikeys, Old Fr. piquois, picquois, "a pickax" (Cotgrave), from the verb piquer. The word posthumous has changed its meaning through folk-etymology. It represents the Latin superlative postumus, latest born. By association with humus, ground, earth, it came to be used of a child born, or a work published, after its author's death, a meaning which the derivatives of postumus have in all the Romance languages.

The first part of the word has been distorted in pursy, short-winded—

Fr. poussif, from Lat. pulsus, throbbing. It was formerly used also in connection with horses—

Arquebus, Fr. arquebuse, is a doublet of hackbut, Old Fr. haquebute, "an haquebut, or arquebuse; a caliver" (Cotgrave). The corruption is due to arcus, bow. Both arquebus and hackbut are common in Scott—

The origin is Du. haakbus, hook-gun, the second element of which appears in blunderbuss. The first part of this word has undergone so many popular transformations that it is difficult to say which was the original form. Ludwig has Donner-bÜchs, Blunder-bÜchs, oder Muszketon, "a thunder-box; a blunder-buss; a musketoon; a wide-mouthed brass-gun, carrying about twenty pistol bullets at once." It was also called in German Plantier-bÜchs, from plantieren, to plant, set up, because fired from a rest. Du. bus, like Ger. BÜchse, means both "box" and "gun." In the bushes, or axle-boxes, of a cart-wheel, we have the same word. The ultimate origin is Greek p????, the box-tree, whence also the learned word pyx. Fr. boÎte, box, is cognate, and Fr. boussole, mariners' compass, is from the Italian diminutive bossola, "a boxe that mariners keepe their compasse in. Also taken for the compasse" (Florio).

Scissors were formerly cizars (cf. Fr. ciseaux), connected with Lat. cÆdere, to cut. The modern spelling is due to association with Lat. scissor, a cutter, tailor, from scindere, to cut. Runagate is well known to be a corrupt doublet of renegade, one who has "denied" his faith. Recreant, the present participle of Old Fr. recreire, Vulgar Lat. *recredere, to change one's faith, contains very much the same idea; cf. miscreant, lit. unbeliever. Jaunty, spelt janty by Wycherley and genty by Burns, is Fr. gentil, wrongly brought into connection with jaunt.

In some cases of folk-etymology it is difficult to see to what idea the corruption is due.[98] The mollusc called a periwinkle was in Anglo-Saxon pinewincla, which still survives in dialect as pennywinkle. It appears to have been influenced by the plant-name periwinkle, which is itself a corruption of Mid. Eng. pervenke, from Lat. pervinca; cf. Fr. pervenche. The material called lutestring was formerly lustring, Fr. lustrine, from its glossiness. A wiseacre is "one that knows or tells truth; we commonly use it in malam partem for a fool" (Blount, Glossographia, 1674). This comes, through Dutch, from Ger. Weissager, commonly understood as wise-sayer, but really unconnected with sagen, to say. The Old High Ger. wizago, prophet, is cognate with Eng. witty. The military and naval word ensign is in Shakespeare corrupted, in both its meanings, into ancient. Thus Falstaff describes his tatterdemalion recruits as—

while Ancient Pistol is familiar to every reader. A cordwainer, from Old Fr. cordouanier, "a shoomaker, a cordwainer" (Cotgrave), worked with cordouan, "Cordovan leather; which is properly a goat's skin tanned." The modern French form cordonnier is due to association with cordon, a thong, bootlace, etc. Witch-elm has nothing to do with witches. It is for older weech-elm, wiche-elm, and belongs to Anglo-Sax. wican, to bend. Service-tree is a meaningless corruption of Mid. Eng. serves, an early loan word from Lat. sorbus.

In the case of a double-barrelled word, folk-etymology usually affects one half only, e.g., verdigris is for Fr. vert-de-gris, for Old Fr. vert de Grece, Greek green. The reason for the name is unknown. Cotgrave calls it "Spanish green." Mid. English had the more correct vertegresse and verte Grece (Promptorium Parvulorum, 1440). The cavalry trumpet-call boot and saddle is for Fr. boute-selle, lit. "put saddle." Court card is for coat card, a name given to these cards from the dresses depicted on them. Florio has carta di figura, "a cote carde." The card game called Pope Joan would appear to be in some way corrupted from nain jaune, lit. "yellow dwarf," its French name.

But occasionally the results of folk-etymology are literally preposterous.[99] The Fr. choucroute is from surkrut, a dialect pronunciation of Ger. Sauer-kraut, sour cabbage, so that the first syllable, meaning "sour," has actually been corrupted so as to mean "cabbage." Another example, which I have never seen quoted, is the name of a beech-wood near the little town of Remilly in Lorraine. The trees of this wood are very old and curiously twisted, and they are called in French les jolis fous, where fou (Lat. fagus) is the Old French for "beech" (fouet, whip, is its diminutive). This is rendered in German as tolle Buchen, mad beeches, the fou having been misunderstood as referring to the fantastic appearance of the trees.

Forlorn hope is sometimes used metaphorically as though the hope were of the kind that springs eternal in the human breast. In military language it now means the leaders of a storming party—

but was earlier used of soldiers in any way exposed to special danger. Cotgrave has enfans perdus, "perdus; or the forlorne hope of a campe (are commonly gentlemen of companies)." It is from obsolete Du. verloren hoop, where hoop, cognate with Eng. heap, is used for a band or company. In 16th-century German we find ein verlorener Haufe. Both the Dutch and German expressions are obsolete in this sense.

The military phrase to run the gauntlet has no connection with gauntlet, glove. The older form gantlope—

It is a punishment of Swedish origin from the period of the Thirty Years' War. The Swedish form is gatlopp, in which gat is cognate with Eng. gate, in its northern sense of "street," and lopp with Eng. leap and Ger. laufen, to run.

The press-gang had originally nothing to do with "pressing." When soldiers or seamen were engaged, they received earnest money called prest-money, i.e., an advance on "loan," Old Fr. prest (prÊt), and the engagement was called presting or impresting. Florio explains soldato (see p. 154), lit. "paid," by "prest with paie as soldiers are." The popular corruption to press took place naturally as the method of enlistment became more "pressing."

The black art is a translation of Old Fr. nigromance, "nigromancie, conjuring, the black art" (Cotgrave); but this is folk-etymology for nÉcromantie, Greco-Lat. necromantia, divination by means of the dead. The popular form nÉgromancie still survives in French. To curry favour is a corruption of Mid. Eng. "to curry favel." The expression is translated from French. Palsgrave has curryfavell, a flatterer, "estrille faveau," estriller (Étriller) meaning "to curry (a horse)." Faveau, earlier fauvel, is the name of a horse in the famous Roman de Fauvel, a satirical Old French poem of the early 14th century. He symbolises worldly vanity carefully tended by all classes of society. The name is a diminutive of Fr. fauve, tawny, cognate with Eng. fallow (deer). (See also p. 192, n.)

A very curious case of folk-etymology is seen in the old superstition of the hand of glory. This is understood to be a skeleton hand from the gallows which will point out hidden treasure—

(Ingoldsby, The Hand of Glory.)