Chapter Iii. Spelling and Sound

"Do you spell it with a V or a W?" inquired the judge.

"That depends upon the taste and fancy of the speller, my lord," replied Sam. "I never had occasion to spell it more than once or twice in my life, but I spells it with a V."

(Pickwick, ch. xxxiv.)

Many people are particular about the spelling of their names. I am myself, although, as a student of philology, I ought to know better. The greatest of Englishmen was so careless in the matter as to sign himself Shakspe, a fact usually emphasized by Baconian when speaking of the illiterate clown of Stratford-on-Avon. Equally illiterate must have been the learned Dr. Crown, who, in the various books he published in the latter half of the seventeenth century, spelt his name, indifferently Cron, Croon, Croun, Crone, Croone, Croune. The modern spelling of any particular name Is a pure accident. Before the Elementary Education Act of 1870 a considerable proportion of English people did not spell their names at all. They trusted to the parson and the clerk, who did their best with unfamiliar names. Even now old people in rural districts may find half a dozen orthographic variants of their own names among the sparse documentary records of their lives. Dugdale the antiquary is said to have found more than 130 variants of Mainwaring among the parchments of that family. Bardsley quotes, under the name Blenkinsop--

"On April 2 3, 1470, Elizabeth Blynkkynesoppye, of Blynkkynsoppe, widow of Thomas Blynkyensope, of Blynkkensope, received a general pardon"--

four variants in one sentence. In the List of Foreign Protestants and Aliens in England (1618) we have Andrian Medlor and Ellin Medler his wife, Johan Cosen and Abraham Cozen, brethren. The death of Sarah Inward, daughter of Richard Inwood, was registered in 1685.

VARIANT SPELLINGS

Medieval spelling was roughly phonetic, i.e. it attempted to reproduce the sound of the period and region, and even men of learning, as late as the eighteenth century, were very uncertain in matters of orthography. The spelling of the language is now practically normalized, although in conformity with no sort of principle; but the family name, as a private possession, has kept its freedom. Thus, if we wish to speak poetically of a meadow, I suppose we should call it a lea, but the same word is represented by the family names Lea, Lee, Ley, Leigh, Legh, Legge, Lay, Lye, perhaps the largest group of local surnames we possess.

In matters of spelling we observe various tendencies. One is the retention of an archaic form, which does not necessarily affect pronunciation. Late Mid. English was fond of y for i, of double consonants, and of final -e. All these appear in the names Thynne (thin) and Wyllie (wily). Therefore we should not deride the man who writes himself Smythe. But in some cases the pronunciation suffers, e.g. the name Fry represents Mid. Eng. fri, one of the forms of the adjective that is now written free. Burt represents Anglo-Sax. beorht, the normal result of which is Bright. We now write subtle and perfect, artificial words, in the second of which the pronunciation has been changed in accordance with the restored spelling; but the older forms survive in the names Suttle and Parfitt--

"He was a verray parfit, gentil knyght."

(A, 72.)

The usual English pronunciation of names like Mackenzie, Menzies, Dalziel, is due to the substitution by the printer of a z for an obsolete letter that represented a soft palatal sound more like y. [Footnote: This substitution has led one writer on surnames, who apparently confuses bells with beans, to derive the rare surname Billiter, whence Billiter's Lane in the City, from "Belzetter, i.e., the Bell-setter." The Mid. Eng. "bellezeter, campanarius" (Prompt. Parv.), was a bell-founder, from a verb related to geysir, ingot, and Ger. giessen, to pour. Robert le bellegeter was a freeman of York in 1279.]

We have an archaic plural ending in Knollys (Knowles), the plural of knoll, and in Sandys, and an archaic spelling in Sclater for Slater or Slatter, for both slat and slate come from Old Fr. esclat (eclat), a splinter. With Knollys and Sandys we may put Pepys, for the existence of the dims. Pipkin, Peppitt, and Peppiatt points to the medieval name Pipun, corresponding to the royal Pepin. Streatfeild preserves variant spellings of street and field. In Gardiner we have the Old Northern French word which now, as a common noun, gardener, is assimilated to garden, the normal French form of which appears in Jardine.

Such orthographic variants as i and y, Simons, Symons, Ph and f, Jephcott, Jeffcott, s and c, Pearce, Pearce, Rees, Reece, Sellars (cellars), ks and x, Dickson, Dixon, are a matter of taste or accident. Initial letters which became mute often disappeared in spelling, e.g. Wray, a corner (Chapter XIII), has become hopelessly confused with Ray, a roe, Knott, from Cnut, i.e. Canute, or from dialect knot, a hillock, with Noll, crop-haired. Knowlson is the son of Nowell (Chapter IX) or of Noll, i.e. Oliver.

Therefore, when Mr. X. asserts that his name has always been spelt in such and such a way, he is talking nonsense. If his great-grandfather's will is accessible, he will probably find two or three variants in that alone. The great Duke of Wellington, as a younger man, signed himself Arthur Wesley--

"He was colonel of Dad's regiment, the Thirty-third foot, after Dad left the army, and then he changed his name from Wesley to Wellesley, or else the other way about"

(KIPLING, Marklake Witches);

and I know two families the members of which disagree as to the orthography of their names. We have a curious affectation in such spellings as ffrench, ffoulkes, etc., where the ff is merely the method of indicating the capital letter in early documents.

The telescoping of long names is a familiar phenomenon. Well-known examples are Cholmondeley, Chumley, Marjoribanks, Marchbanks, Mainwaring, Mannering. Less familiar are Auchinleck, Affleck, Boutevilain, Butlin, Postlethwaite, Posnett, Sudeley, Sully, Wolstenholme, Woosnam. Ensor is from the local Edensor, Cavendish was regularly Candish for the Elizabethans, while Cavenham in Suffolk has given the surname Canham. Daventry has become Daintree, Dentry, and probably the imitative Dainty, while Stepson is for Stevenson. It is this tendency which makes the connection between surnames and village names so difficult to establish in many cases, for the artificial name as it occurs in the gazetteer often gives little clue to the local pronunciation. It is easy to recognize Bickenhall or Bickenhill in Bicknell and Puttenham in Putnam, but the identity of Wyndham with Wymondham is only clear when we know the local Pronunciation of the latter name. Milton and Melton are often telescoped forms of Middleton.

DIALECTIC VARIANTS

Dialectic variants must also be taken into account. Briggs and Rigg represent the Northern forms of Bridges and Ridge, and Philbrick is a disguised Fellbrigg. In Egg we have rather the survival of the Mid. English spelling of Edge. Braid, Lang, Strang, are Northern variants of Broad, Long, Strong. Auld is for Old while Tamson is for Thompson and Dabbs for Dobbs (Robert). We have the same change of vowel in Raper, for Roper. Venner generally means hunter, Fr. veneur, but sometimes represents the West-country form of Fenner, the fen-dweller; cf. Vidler for fiddler, and Vanner for Fanner, the winnower.

We all the difficulty we have in catching a new and unfamiliar name, and the subterfuges we employ to find out what it really is. In such cases we do not get the help from association and analogy which serves us in dealing with language in general, but find ourselves in the position of a foreigner or child hearing unfamiliar word for the first time. We realize how many imperceptible shades there are between a short i and a short e, or between a fully voiced g and a voiceless k, examples suggested to me by my having lately understood a Mr. Riggs to be a Mr. Rex.

We find occurring in surnames examples of those consonantal changes which do not violate the great Phonetic law that such changes can only occur regularly within the same group, i.e. that a labial cannot alternate with a palatal, or a dental with either. It is thus that we find b alternating with p, Hobbs and Hopps (Robert), Bollinger and Pullinger, Fr. boulanger; g with k, Cutlack and Goodlake (Anglo-Sax. Guthlac), Diggs and Dix (Richard), Gipps and Kipps (Gilbert), Catlin and Galling (Catherine); j with ch, Jubb or Jupp and Chubb (Job); d with t, Proud and Prout (Chapter XXII), Dyson and Tyson (Dionisia), and also with th, Carrodus and Carruthers (a hamlet in Dumfries). The alternation of c and ch or g and j in names of French origin is dialectic, the c and g representing the Norman-Picard pronunciation, e.g. Campion for Champion, Gosling for Joslin. In some cases we have shown a definite preference for one form, e.g. Chancellor and Chappell, but Carpenter and Camp. In English names c is northern, ch southern, e.g. Carlton, Charlton, Kirk, Church.

There are also a few very common vowel changes. The sound er usually became ar, as in Barclay (Berkeley), Clark, Darby, Garrard (Gerard), Jarrold (Gerald), Harbord (Herbert), Jarvis (Gervase), Marchant, Sargent, etc., while Larned, our great-grandfathers' pronunciation of "learned," corresponds to Fr. Littri. Thus Parkins is the same name as Perkins. (Peter), and these also give Parks and Perks, the former of which is usually not connected with Park. To Peter, or rather to Fr. Pierre, belong also Parr, Parry and Perry, though Parry is generally Welsh (Chapter VI). The dims. Parrott, Perrott, etc., were sometimes nicknames, the etymology being the same, for our word parrot is from Fr. pierrot. To the freedom with which this sound is spelt, e.g. in Herd, Heard, Hird, Hurd, we also owe Purkiss for Perkins; cf. appurtenance for appartenance.

The letter l seems also to exercise a demoralizing influence on the adjacent vowel. Juliana became Gillian, and from this, or from the masculine form Julian, we get Jalland, Jolland, and the shortened Gell, Gill (Chapter VI), and Jull. Gallon, which Bardsley groups with these, is more often a French name, from the Old German Walo, or a corruption of the still commoner French name Galland, likewise of Germanic origin.

We find also such irregular vowel changes as Flinders for Flanders, and conversely Packard for Picard. Pottinger (see below) sometimes becomes Pettinger as Portugal gives Pettingall. The general tendency is towards that thinning of the vowel that we get in mister for master and Miss Miggs's mim for ma'am. Littimer for Lattimer is an example of this. But in Royle for the local Ryle we find the same broadening which has given boil, a swelling, for earlier bile.

APHESIS

Among phonetic changes which occur with more or less regularity are those called aphesis, epenthesis, epithesis, assimilation, dissimilation, and metathesis, convenient terms which are less learned than they appear. Aphesis is the loss of the unaccented first syllable, as in 'baccy and 'later. It occurs almost regularly in words of French origin, e.g. squire and esquire, Prentice and apprentice. When such double forms exist, the surname invariably assumes the popular form, e.g. Prentice, Squire. Other examples are Bonner, i.e. debonair, Jenner, Jenoure, for Mid. Eng. engenour, engineer, Cator, Chaytor, Old Fr. acatour (acheteur), a buyer--

"A gentil maunciple was they of a temple, Of which achatours mighte take exemple" (A. 567),

Spencer, dispenser, a spender, Stacey for Eustace, Vick and Veck for Levick, i.e. l'eveque, the bishop, Pottinger for the obsolete potigar, an apothecary, etc.

The institution now known as the "orspittle" was called by our unlettered forefathers the "spital," hence the names Spittle and Spittlehouse. A well-known amateur goal-keeper has the appropriate name Fender, for defender.

Many names beginning with n are due to aphesis, e.g. Nash for atten ash, Nalder, Nelms, Nock, atten oak, Nokes, Nye, atten ey, at the island, Nangle, atten angle, Nind or Nend, atten ind or end. With these we may compare Twells, at wells, and the numerous cases in which the first part of a personal name is dropped, e.g. Tolley, Bartholomew, Munn, Edmund, Pott, Philpot, dim. of Philip (see p.87), and the less common Facey, from Boniface, and Loney, from Apollonia, the latter of which has also given Applin.

When a name compounded with Saint begins with a vowel, we get such forms as Tedman, St. Edmund, Tobin, St. Aubyn, Toosey, St. Osith, Toomey, St. Omer, Tooley, St. Olave; cf. Tooley St. for St. Olave St. and tawdry from St. Audrey. When the saint's name begins with a consonant, we get, instead of aphesis, a telescoped pronunciation, e.g. Selinger, St. Leger, Seymour, St. Maur, Sinclair, St. Clair, Semark, St. Mark, Semple, St. Paul, Simper, St. Pierre, Sidney, probably for St. Denis, with which we may compare the educated pronunciation of St. John. These names are all of local origin, from chapelries in Normandy or England.

Epenthesis is the insertion of a sound which facilitates pronunciation, such as that of b in Fr. chambre, from Lat. camera. The intrusive sound may be a vowel or a consonant, as in the names Henery, Hendry, perversions of Henry. [Footnote: On the usual fate of this name in English, see below.]

To Hendry we owe the northern Henderson, which has often coalesced with Anderson, from Andrew. These are contracted into Henson and Anson, the latter also from Ann and Agnes (Chapter IX). Intrusion of a vowel is seen in Greenaway, Hathaway, heath way, Treadaway, trade (i.e. trodden) way, etc., also in Horniman, Alabone, Alban, Minister, minster, etc. But epenthesis of a consonant is more common, especially b or p after m, and d after n. Examples are Gamble for the Anglo-Saxon name Gamel, Hamblin for Hamlin, a double diminutive of Hamo, Simpson, Thompson, etc., and Grindrod, green royd (see p. III). There is also the special case of n before g in such names as Firminger (Chapter XV), Massinger (Chapter XX), Pottinger (Chapter XVIII), etc.

EPITHESIS AND ASSIMILATION

Epithesis, or the addition of a final consonant, is common in uneducated speech, e.g. scholard, gownd, garding, etc. I say "uneducated," but many such forms have been adapted by the language, e.g. sound, Fr. son, and we have the name Kitching for kitchen. The usual additions are -d, -t, or -g after n, e.g. Simmonds, Simon, Hammond, Hammant, Fr. Hamon, Hind, a farm labourer, of which the older form is Hine (Chapter XVII), Collings for Collins, Jennings, Fr. Jeannin, dim. of Jean, Aveling from the female name Avelina or Evelyn. Neill is for Neil, Nigel. We have epithetic -b in Plumb, the man who lived by the plum-tree and epithetic -p in Crump (Chapter II).

Assimilation is the tendency of a sound to imitate its neighbour. Thus the d of Hud (Chapter I) sometimes becomes t in contact with the sharp s, hence Hutson; Tomkins tends to become Tonkin, whence Tonks, if the m and k are not separated by the epenthetic p, Tompkins. In Hopps and Hopkins we have the b of Hob assimilated to the sharp s and k, while in Hobbs we pronounce a final -z. It is perhaps under the influence of the initial labial that Milson, son of Miles or Michael, sometimes becomes Milsom, and Branson, son of Brand, appears as Bransom.

The same group of names is affected by dissimilation, i.e. the instinct to avoid the recurrence of the same sound. Thus Ranson, son of Ranolf or Randolf, becomes Ransom [Footnote: So also Fr, rancon gives Eng. ransom. The French surname Rancon is probably aphetic for Laurancon.] by dissimilation of one n, and Hanson, son of Han (Chapter I), becomes Hansom. In Sansom we have Samson assimilated to Samson and then dissimilated. Dissimilation especially affects the sounds l, n, r. Bullivant is found earlier as bon enfaunt (Goodchild), just as a braggart Burgundian was called by Tudor dramatists a burgullian. Bellinger is for Barringer, an Old French name of Teutonic origin. [Footnote: "When was Bobadil here, your captain? that rogue, that foist, that fencing burgullian" (Jonson, Every Man in his Humour, iv. 2).] Those people called Salisbury who do not hail from Salesbury in Lancashire must have had an ancestor de Sares-bury, for such was the earlier name of Salisbury (Sarum). A number of occupative names have lost the last syllable by dissimilation, e.g. Pepper for pepperer, Armour for armourer. For further examples see Chapter XV.

It may be noted here that, apart from dissimilation, the sounds l, n, r, have a general tendency to become confused, e.g. Phillimore is for Finamour (Dearlove), which also appears as Finnemore and Fenimore, the latter also to be explained from fen and moor. Catlin is from Catherine. Balestier, a cross-bow man, gives Bannister, and Hamnet and Hamlet both occur as the name of one of Shakespeare's sons. Janico or Jenico, Fr. Janicot may be the origin of Jellicoe.

We also get the change of r to l in Hal, for Harry, whence Hallett, Hawkins (Halkins), and the Cornish Hockin, Mal or Mol for Mary, whence Malleson, Mollison, etc., and Pell for Peregrine. This confusion is common in infantile speech, e.g. I have heard a small child express great satisfaction at the presence on the table of "blackbelly dam."

METATHESIS

Metathesis, or the transposition of sound, chiefly affects l and r, especially the latter. Our word cress is from Mid. Eng. kers, which appears in Karslake, Toulmin is for Tomlin, a double dim., -el-in, of Tom, Grundy is for Gundry, from Anglo-Sax. Gundred, and Joe Gargery descended from a Gregory. Burnell is for Brunel, dim. of Fr. brun, brown, and Thrupp is for Thorp, a village (Chapter XIII). Strickland was formerly Stirkland, Cripps is the same as Crisp, from Mid. Eng. crisp, curly. Prentis Jankin had--

"Crispe here, shynynge as gold so fyn"

(D. 304);

and of Fame we are told that

"Her heer was oundie (wavy) and crips."

(House of Fame, iii. 296.)

Both names may also be short for Crispin, the etymology being the same in any case. Apps is sometimes for asp, the tree now called by the adjectival name aspen (cf. linden). We find Thomas atte apse in the reign of Edward III.

The letters l, n, r also tend to disappear from no other cause than rapid or careless pronunciation.

Hence we get Home for Holme (Chapter XII), Ferris for Ferrers, a French local name, Batt for Bartholomew, Gatty for Gertrude, Dallison for d'Alencon. The loss of r after a vowel is also exemplified by Foster for Forster, Pannell and Pennell for Parnell (sometimes), Gath for Garth (Chapter XIII), and Mash for Marsh. To the loss of n before s we owe such names as Pattison, Paterson, etc., son of Paton, the dim. of Patrick, and Robison for Robinson, and also a whole group of names like Jenks and Jinks for Jenkins (John), Wilkes for Wilkins, Gilkes, Danks, Perks, Hawkes, Jukes for Judkins (Chapter VI), etc. Here I should also include Biggs, which is not always connected with Bigg, for we seldom find adjectival nicknames with -s. It seems to represent Biggins, from obsolete biggin, a building (Chapter XIII).

The French nasal n often disappeared before r. Thus denree, lit. a pennyworth, appears in Anglo-French as darree. Similarly Henry became Harry, except in Scotland, and the English Kings of that name were always called Harry by their subjects. It is to this pronunciation that we owe the popularity of Harris and Harrison, and the frequency of Welsh Parry, ap, Harry, as compared with Penry. A compromise between Henry and Harry is seen in Hanrott, from the French dim. Henriot.

The initial h-, which we regard with such veneration, is treated quite arbitrarily in surnames. We find a well-known medieval poet called indifferently Occleve and Hoccleve. Harnett is the same as Arnett, for Arnold, Ewens and Heavens are both from Ewan, and Heaven is an imitative form of Evan. In Hoskins, from the medieval Osekin, a dim. of some Anglo-Saxon name such as Oswald (Chapter VII), the aspirate has definitely prevailed. The Devonshire name Hexter is for Exeter, Arbuckle is a corruption of Harbottle, in Northumberland. The Old French name Ancel appears as both Ansell and Hansell, and Earnshaw exists side by side with Hearnshaw (Chapter XII).

The loss of h is especially common when it is the initial letter of a suffix, e.g. Barnum for Barnham, Haslam, (hazel), Blenkinsop for Blenkin's hope (see hope, Chapter XII), Newall for Newhall, Windle for Wind Hill, Tickell for Tick Hill, in Yorkshire, etc. But Barnum and Haslam may also represent the Anglo-Saxon dative plural of the words barn and hazel. A man who minded sheep was once called a Shepard, or Sheppard, as he still is, though we spell it shepherd. The letter w disappears in the same way; thus Greenish is for Greenwich, Horridge for Horwich, Aspinall for Aspinwall, Millard for Millward, the mill-keeper, Boxall for Boxwell, Caudle for Cauldwell (cold); and the Anglo-Saxon names in -win are often confused with those in -ing, e.g. Gooding, Goodwin; Golding, Goldwin; Gunning, Gunwin, etc. In this way Harding has prevailed over the once equally common Hardwin.

BABY PHONETICS

Finally, we have to consider what may be called baby phonetics, the sound-changes which seem rather to transgress general phonetic laws. Young children habitually confuse dentals and palatals, thus a child may be heard to say that he has "dot a told." This tendency is, however, not confined to children. My own name, which is a very uncommon one, is a stumbling-block to most people, and when I give it in a shop the scribe has generally got as far as Wheat- before he can be stopped.

We find both Estill and Askell for the medieval Asketil, and Thurtle alternating with Thurkle, originally Thurketil (Chapter VII). Bertenshave is found for Birkenshaw, birch wood, Bartley, sometimes from Bartholomew, is more often for Berkeley, and both Lord Bacon and Horace Walpole wrote Twitnam for Twickenham. Jeffcock, dim. of Geoffrey, becomes Jeffcott, while Glascock is for the local Glascott. Here the palatal takes the place of the dental, as in Brangwin for Anglo-Sax. Brandwine. Middleman is a dialect form of Michaelmas (Chapter IX). We have the same change in tiddlebat for stickleback, a word which exemplifies another point in baby phonetics, viz. the loss of initial s-, as in the classic instance tummy. To this loss of s- we owe Pick for Spick (Chapter XXIII), Pink for Spink, a dialect word for the chaffinch, and, I think, Tout for Stout. The name Stacey is found as Tacey in old Notts registers. On the other hand, an inorganic s- is sometimes prefixed, as in Sturgess for the older Turgis. For the loss of s- we may compare Shakespeare's parmaceti (1 Henry IV. i. 3), and for its addition the adjective spruce, from Pruce, i.e. Prussia.

We also find the infantile confusion between th and f e.g. in Selfe, which appears to represent a personal name Seleth, probably from Anglo-Sax, saelth, bliss. Perhaps also in Fripp for Thripp, a variant of Thrupp, for Thorp. Bickerstaffe is the name of a place in Lancashire, of which the older form appears in Bickersteth, and the local name Throgmorton is spelt by Camden Frogmorton, just as Pepys invariably writes Queenhive for Queenhythe.

Such are some of the commoner phenomena to be noticed in connection with the spelling and sound of our names. The student must always bear in mind that our surnames date from a period when nearly the whole population was uneducated. Their modern forms depend on all sorts of circumstances, such as local dialect, time of adoption, successive fashions in pronunciation and the taste and fancy of the speller. They form part of our language, that is, of a living and ever-changing organism. Some of us are old enough to remember the confusion between initial v and w which prompted the judge's question to Mr. Weller. The vulgar i for a, as in "tike the kike," has been evolved within comparatively recent times, as well as the loss of final -g, "shootin and huntin," in sporting circles. In the word warmint--

"What were you brought up to be?"

"A warmint, dear boy"

(Great Expectations, ch. xl.),

we have three phonetic phenomena, all of which have influenced the form and sound of modern surnames, e.g. in Winter, sometimes for Vinter, i.e. vintner, Clark for Clerk, and Bryant for Bryan; and similar changes have been in progress all through the history of our language.

In conclusion it may be remarked that the personal and accidental element, which has so much to do with the development of surnames, releases this branch of philology to some extent from the iron rule of the phonetician. Of this the preceding pages give examples. The name, not being subject as other words are to a normalizing influence, is easily affected by the traditional or accidental spelling. Otherwise Fry would be pronounced Free. The o is short in Robin and long in Probyn, and yet the names are the same (Chapter VI). Sloper and Smoker mean a maker of slops and smocks respectively, and Smale is an archaic spelling of Small, the modern vowel being in each case lengthened by the retention of an archaic spelling. The late Professor Skeat rejects Bardsley's identification of Waring with Old Fr. Garin or Warin, because the original vowel and the suffix are both different. But Mainwaring, which is undoubtedly from mesnil-Warin (Chapter XIV), shows Bardsley to be right.