«P.H.REANEY Litt.D., Ph.D., F.S.A. Third edition with corrections and additions by R.M.WILSON M.A. LONDON AND NEW YORK First published as A Dictionary ...»
No satisfactory explanation has been given of this final -s in surnames formed from personal-names. Fransson’s examples are late (1310). He regards them as elliptic genitives. As all his examples but one (Roger le Persones) are women, he must take all four to mean ‘servant of Robert, the parson, etc.’.38 Ewen’s account is confused.39 He cites Willelmus Johannis (1159–60) and three similar forms of 1229–35 as examples of ‘inflected genitives’ due to ‘filius and filia having fallen into disuse’. But surnames of the type Willelmus filius Johannis are common long after 1235. He notes also two undated metronymics, Johanna Mariote and Willelmus Margarete, adding ‘but the English nominative form, as Henry Maynard or John Rogers was also used, and the genitive ending (es, is, ys, or s) also begins to be noticeable, and at first most frequently in the names of women, thus Robertus filius Radulfi became Robertus Rolle (Raoul), but Matilda filia Radulfi was written Matilda Rolles…There was no precise rule, many surnames of women are without the final sibilant, which is occasionally found added to the second names of men.’40 He does not explain why filius Radulfi becomes Rolle when the name is a man’s, but Rolles when it is that of a woman.
The Latin type is more common than Tengvik and Ewen would lead us to think. Over 50 examples have been noted between 1130 and 1240, all except one (Emma Philippi 1240 Rams, Nf), names of men, usually from French personal names: Hugo Oillardi 1130 P (Sr), Willelmus Walkelini 12th DC (Lei), Willelmus Luce 1185 Templars (K),
Johannes Jeremie 1196 P (Y); occasionally from Old English or Scandinavian names:
Willelmus Ailrici 12th DC (L), Robertus Edwini 1229 Pat (So), Alanus Torberti 1212 Cur (Ha). Three are formed from names of women: Arnaldus Mabilie 1185 Templars (Ess), Robertus Margerie 1195 P (Gl). These names can only mean ‘John, son of Jeremiah’, ‘Arnold, son of Mabel’, etc., literal translations of the vernacular, just as the clerk translated ‘Edward of Salisbury’ by Edwardus Saresberiae 1100–35 Rams (Hu).
The English forms are early examples of the elliptic genitive, Edricus Keteles ‘Edric Ketel’s (son)’, parallel to Personnes, Prestes above. That this interpretation is correct is proved by the following. In 1281 we have mention of Robert de Rokesle junior who is twice called Robert Dobes in 1305. His father was Robert de Rokesle senior who must often have been called by his pet-name Dob. Hence his son’s surname Dobes which must mean ‘son of Dob’, i.e. of Robert.45 Toward the end of the thirteenth century, this type of name becomes more common and steadily increases in the fourteenth, but there is a marked difference in its frequency in different counties. In the Worcestershire Subsidy Roll for 1275 there are only 7 examples, 5 being names of women; in that for 1327 we find 138 men and 30 women so named. In Somerset in 1327:128 men, 73 women; in Warwickshire (1332), 161 men, 20 women; in Suffolk (1327), 27 men, 1 woman; in Surrey (1332), 5 men, 11 women. In other Subsidy Rolls the number is negligible: Sussex (1296), 6; (1327) 2; (1332) 3;
Cambridgeshire 4; Lancashire 2; Cumberland 2; Yorkshire (1297) 6; (1301) 3; (1327) 3, all men. In Essex there are 5 men in 1295 ParlR, a number in the Colchester Court Rolls (1311–45) and about 12 (all women) in the 1349 Fingrith Hall Court Rolls. The surnames are usually the common christian names in use, often pet-forms, rarely Old English and almost invariably masculine. Some few are nicknames or occupational names (these sometimes preceded by le, occasionally la): Ysabella Barones 1275 SRWo, Hugh Rabuckes 1301 SRY, Claricia le Parkeres 1327 SRSo, Juliana la Kinges 1285 Ass (Ess), Amiscia la Wrihtes 1333 ColchCt. Occasionally we have a place-name: John Dounes 1327 SRWo.
The interpretation of these surnames is more difficult than one would expect. It is clear that in the twelfth century Segarus Aileves meant ‘son of Aileve’ and that is probably the meaning in the fourteenth century in the names of men. But John le Cokes (1327 SRWo) may well have been the cook’s servant or assistant and names like William Hogges may have lost the article and have a similar meaning. So with women’s names. Claricia le Parkeres may have been the servant of the parker and Isabella la Chancelers ‘the servant of (a man named) Chanceler’. But such an interpretation is unsatisfactory for Avice la Schepherdes (1311 ColchCt) and Juliana le Smithes (1279 RH), for shepherds and blacksmiths were unlikely to have servants. Where the surname is a place-name, ‘servant of a man named Bylton’, etc., is probable.
In the Colchester Court Rolls (1311 ff.), large numbers of women were regularly fined at court after court for selling ale at too high a price. They were usually described as ‘the wife of John Carpenter’, etc., but a certain number are mentioned by name, which almost invariably ends in -es (Joan la Warneres, Alice Sayheres). It is a reasonable presumption that they were widows and that this type of women’s surname denoted either a widow or a married woman. Matilda Candeles (1327 SRSx) was probably the wife of Ralph Candel, for in 1332 she is described as ‘Matilda relicta Candel’. Margery la Mazones was the wife of Walter le Mazoun (1311 ColchCt). Agnes Rickemannes (1329 Husting) was probably the widow of Rickeman le Chaumberleng (1292 SRLo) and if so, we have to reckon with the fact that some of these women’s names denote the christian name and not the surname of their husbands. It is not uncommon in these documents to find pairs of names like Nicholas le Knyt and Cecilia Knyctes (1297 MinAcctCo), who, we may fairly assume, were husband and wife. Amisia Hugines (1327 SRWo) was probably the wife of William Hugyns and here the surname means ‘son of Hugyn’. Thus, a surname like Stevens may mean ‘son of Stephen’, ‘servant of Stephen’, or ‘servant at Stephen’s house’, or it may be a metronymic derived from a form Stevenes ‘Stephen’s wife’. The only certainty is that atte Stevenes means ‘servant at Stephen’s house’. The -s of local surnames may be a plural inflexion (or a sign of French origin), but more often falls into one or other of the above classes. Sometimes, in late additions, it may be a dialectal pronunciation, with excrescent s.46
Already in Old English we find pet-names in use: Tuma for Trumwine in the seventh century and Ælle for Ælfwine in the tenth, and such forms as Wine and Wulfa for Winefrīð and Wulfwine.47 Names of this type continued to be formed and a number still survive in surnames, some otherwise unrecorded. But most of the pet-names in modern surnames are of post-Conquest formation and some are difficult to identify. Examples are found in the twelfth century; they become more numerous in the middle of the thirteenth and in the fourteenth are common. They are found among all classes and are derived from Old English, Scandinavian and French personal names alike. Cudd (1358) and Cutt (1279) are from OE Cūðbeorht, Ugga (1212) from Asti (1203) is a pet-form of ON Ásketill; Lamb (1161) is for Lambert, Gibbe (1179) for Gilbert, Lina (1181) for Adelina or Emelina, whilst the Breton Sanson, Samson has given Sanne (1260) and Samme (1275). Not all pet-names are so easy to identify. Hudd(1177) and Hulle (1227) are undoubtedly for Hugh, but Huddis also used for Richard. Pelle (1274) is a pet-name for Peter. Hann is undoubtedly for John (from Jehan), but is just as certainly for Hanry (Henry) and is said to have been used also for Randolph.
Some of these forms follow normal phonetic laws of assimilation: Judd from Jurd (Jordan); Fippe from Philip, Bette from Bertin and Bertelmeu (Bartholomew), Penne, a shortening of Pennel, from Pernel, Cuss from Cust (Custance, Constance), Ibb for Isabel.
In others, the name begins by anticipating the following medial consonant: Dande (1246) for Andrew, Biby (1240) for Isabel. Some pet-names are formed from the second syllable of the full name: Pot (1115) from Philipot, itself a diminutive of Philip; Coll (1247) from Nichol, Belle (1279) from Isabel; Sander (1248) from Alexander. Voiced and voiceless consonants were used indiscriminately: Dicke, Digge; Hikke, Higge; Gepp, Gebbe;
Vowels were unrounded: Rob, Rab; Dobb, Dabb; or rounded: Malle, Molle (Mary); Magge, Mogge (Margaret), whilst the changes were rung on the consonants:
Robb (1196), Hobbe (1176), Dobbe (1202), Nabbe (1298), all for Robert. In some names we find a combination of more than one of these features: Libbe (Elizabeth), Pogge (Margaret).
The clue to the explanation of these pet-names is given by Napier and Stevenson when they suggest that OE names such as Lilla, Bubba and Nunna are due to ‘regressive assimilation’ and have their origin in the speech of children.48 Scandinavian scholars call them ‘Lall-names’. According to this theory, Lilla is a short form of some compound of the stem Bil-, such names as Bilheard or Bilnoth. Stenton is disinclined to accept this on the ground that it implies the contemporaneous existence of two sharply contrasted conceptions of nomenclature. ‘The state of mind which produced the compound names with their far-fetched significance is hardly compatible with one which allowed infantile attempts at expressing a name to pass into permanent use.’ He admits, however, that this theory has ‘the great merit of proposing an intelligble connection between these meaningless names and compounds of the normal Germanic type. Its chief weakness is the remoteness of the sound-association between the original compound name and the suggested simple derivative’.49 Children are children and parents are parents, whether we are concerned with the eighth or the twentieth century. The process of learning to speak is the same—trial and error by imitation of sounds heard and there are innumerable examples of common words which have been corrupted in form through misdivision, mispronunciation and misunderstanding. When an Anglo-Saxon named Æðelstan and his wife Wulfgifu deliberately named their son, the future Bishop of Worcester, Wulfstan, a combination of the first theme of the mother’s name and the second of the father’s, they were not concerned with the meaning of the compound—‘wolf-stone’, any more than were those who named their children Fripuwulf ‘peace-wolf’ or Wīgfrið ‘war-peace’. Names had become names and their meaning was a matter of no concern. Association—here, a perpetuation of themes common to the two families—is more important than meaning.
Detailed studies of the early speech-habits of children would throw much light on the eccentric forms of many pet-names. A wreath sent recently by the Queen was from Lilibet, the name by which Her Majesty is known in the family circle, a deliberate perpetuation of her early attempts to pronounce her own name. My own daughter still answers to the name of Titt, a shortening of Titter, her first attempts at sister. A newlywedded wife of my acquaintance regularly addressed her young husband as sweetheart, which gradually became weetheart, sweetie, weetie and finally weet, a pet-name used for many years. There must be many similar pet-names confined to a particular family and never seen in print. With such developments, it is not difficult to realize that a pet-name may have more than one origin, and that a single name may give rise to a variety of petnames.
A few diminutives in -uc of OE origin survive (Haddock, Whittock, Willock), but most
are derivatives of French names. The most common suffixes are -ot, -et, -un, -in, -el:
Philpot, Ibbott; Hewett, Jowett; Paton, Dickens; Rankin, Higgins; Pannel, Pottell.
Double diminutives are formed from these suffixes:
-el-in: Hamlin, Hewlins, Jacklin
-el-ot: Giblett, Roblett
-in-ot: Adnett, Rabnott
-et-in (rare): Turkentine The variety of surnames resulting from these different forms of pet-names may be seen
from the following (varieties of spelling ignored):
Richard (pronounced Rich-ard and Rick-ard):
Rich, Richings, Ritchie; Hitch, Hitchcock, Hitchen, Hitchman, Hitchmough Ricard, Rick, Ricky; Hick, Hicken, Hicklin, Hickman, Hickmott, Hickox; Higgett, Higgins, Higgs; Dick, Dickels, Dicken, Dickin, Dicketts, Dickie; Digg, Diggen Robert: Rabb, Rabbets, Rabjohn, Rablen, Rabnott; Robb, Robbie, Ropkins, Robins, Robjant, Roblett, Roblin; Dabbs, Dabin, Dabinett; Dobb, Dobbie, Dobbin; Hob, Hobbins, Hobday, Hobgen, Hoblin, Hopkin; Nabb, Nap, Nobbs, Nopp Hugh: Hugo, Hue, Hew, How; Hewell, Hewett, Hewlett, Hewlins; Houchen, Howett, Howlett, Howlin (g); Hudd, Hudden, Huddle, Hudman, Hudsmith; Huelin, Huett, Huot;
Huggett, Huggin, Huggon, Huglin; Hukin, Hewkin, Howkins; Hull, Hullett, Hullot;
Hutchin William: Will, Wilkin, Wilcock, Willet, Willott; Willmott, Wellemin, Wellerman, Willmin, Willament Gill(ham), Gilliam, Gillet, Gillman, Guillerman, Gelman v. also Henry, Jack, John, Maud, Paul, Philip.
The Suffixes -cock and -kin