«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 ...»
Such names are not uncommon in the twelfth century and are found later: Gilbert Fathevedsteppeson (1307 Wak), John le Personesneve (1324 FFEss), Richard Hannebrothir, Ameria Ibbotdoghter (1324 Wak), Amabilla Hannewyf (1327 ib.), William Maisterneue (1327 SRSf), John Prestebruther, Johanna Raweswyf (1332 SRCu), Emma Rogerdaughter, Robert Prestcosyn, Marjoria Vicar neys (1381 PTY), Isolda Peersdoghter (1430 FeuDu). The only names of this type to survive are a few compounds of -magh ‘brother-in-law’: Hickmott, Hitchmough, Hudsmith, Watmough.
Names in -son In Old English, patronymics were formed by adding -ing to the stem or -sunu to the genitive of the personal name: Dudding ‘son of Dudda’, Ēadrīcessunu ‘son of Ēadrīc’.
The latter type was used as a patronymic adjunct: Hering Hussan sunu (603 ASC), a type found also in the eighth and ninth centuries and not uncommon in the names of the festermen of Peterborough (963–92): Godwine Ælfrices suna. This was also a common Rolfes sune. In his Old English Bynames,13 Tengvik
has collected 146 examples, of which 111 are English and 24 Scandinavian. In twelfthcentury London Ekwall has noted a further eleven examples.14
Similar formations, though less common, were based on the mother’s name: Eadric Wynflæde sunu (c1015 ASCh), Siwardus Leuerunessone (1066 Winton), Edric’ Modheuesune (1137 ELPN).15
The frequency of names in -son in the North has been commonly attributed to Scandinavian influence,16 but examples are rare or non-existent from the twelfth to the fourteenth centuries in north and south alike. The common form of both patronymics and metronymics in the twelfth and thirteenth centuries is Willelmus filius Hugonis, Ricardus filius Agnetis (1185 Templars), a form found side by side with compounds of -sunu in the eleventh century and, less commonly, of -son in the fourteenth. It is hardly conceivable that surnames like Godricsone (1066 DB) ceased to be used for a couple of centuries and were then suddenly revived. The formula Rogerus filius Radulfi may be merely a description, ‘Roger son of Ralph’, or it may be a translation of Roger Fitz Ralph, a form rare in documents. Radulfus filius Godrici may similarly be a translation of Godricson.
That the formula was merely descriptive is proved by the fact that a man could be named both Willelmus filius Hermanni (1134–40) and Willelmus Hermannus (1141–9 Holme), whilst there are a few examples of the equation of a simple christian name as a surname with a compound of -sunu: Aluuinus Dode, Aluuinus Dodesune (1066 DB). Names like Willelmus filius ƒabri (1219 AssY), Hugo filius clerici (1185 Templars, Gl) are common and descriptive, ‘son of the smith or the clerk’; they are found in the fourteenth century as Smythson and Clerkessone and still survive.
In the Cumberland Subsidy Roll (1332) we find Alan Malleson, John Diksson, etc., side by side with Adam son of Alan, John son of Robert (presumably translations of filius Alani, filius Roberti) and Thomas Prestson. It would appear that the form in common use was Diksson, Helewisson, Heliotesson, etc., that the twelfth- and thirteenth-century scribes translated all such names by filius Roberti, etc., and that by the fourteenth century there was a growing tendency for the clerk to use the spoken form, particularly with the common pet-names Dick, Hob, etc. Thirteenth-century examples noted are: Adam Saresone 1286 LLB D, William Marysone 1298 ELPN (Willelmus filius Marie 1292 SRLo), William Paskessone 1293 FFC, Thomas Wummanesone 1297 Coram (C).
In the Sussex Subsidy Rolls there are no examples of -son in 1296, 4 in 1327, 13 in 1332, all metronymics. In other Subsidy Rolls we have in 1327, in Worcestershire 11 (including 4 metronymics); Somerset 8 (5 metronymics); Cambridgeshire 17 (6 metronymics); in 1332, in Surrey 6 (4 metronymics); Lancashire 23 (7 metronymics);
Cumberland 22 (7 metronymics); in Yorkshire 2 in 1297, 5 in 1301, 10 in 1327, including 3 metronymics; in FrY (1272–1381) 14, all patronymics—earliest example 1323; in Suffolk (1327) 7 (3 metronymics), (1381) 12 (5 metronymics). In all these sources there are additional examples of Reveson, Smithson, etc. Occasionally the surname is based on the father’s surname. In Cambridgeshire John Brunnison and William Broun occur in the same parish as do Richard and William Lawisson and Henry Law. In Cumberland, Hugh Moserghson was presumably the son of Thomas de Mosergh.v. also p. xliv.
The surnames in -son form a very small proportion of the whole and are more common in the north. In Lancashire the number assessed is only about one-fifth that in Suffolk. In the northern counties the number of individuals with no surname or described as filius— —is much greater than in the south where the development of surnames was more advanced. But it is clear that in the fourteenth century, when surnames in -son begin to appear again, they were not limited to the north. It is unlikely that in Somerset, Sussex and Surrey these names should be due to Scandinavian influence. Tengvik notes that in his Old English material examples are found at a date when ‘we can hardly reckon with any important Scandinavian influence’. The local distribution of the type, too (especially in Devonshire), points to a native origin.17 In the north we may have to reckon with Scandinavian influence also, but the frequency of the type may be due, in part at least, to the late development there of hereditary surnames. We find such names as Henry Dicounesson de Clesnesse 1359 Pat (Nb), ‘Henry, son of Dicoun de Clesnesse’, Richard Jeffson Nanneson de Radford 1385 NottBR, a type found also in Yorkshire and Lancashire: Robert Tomson Watson, Robert Stevenson Malynson, Thomas Robynson Richardson 1381 PTY, John Robynson Diconson 1408 AD v (La), John Atkynson Jonson 1433 ib. (Y). It is doubtful whether the latter should be interpreted ‘John, son of Atkyn, son of John’ or ‘John, son of Atkyn Jonson’. The occurrence together of John Prestson, Agnes ancilla Johannis Prestson and Geoffrey Jonson Prestson (1379 PTY) points clearly to a surname but that it was hereditary is doubtful. The frequency of the type and the common addition of -wyf (sometimes added to the christian name), -doghter, -man, -maiden, -servant, give a very strong impression that these were not real surnames in the modern sense but patronymic descriptions in a constant state of flux.
Font-names as Bynames
Less common in the early twelfth century than names like Symon filius Ricardi, but steadily increasing in number, are names of the type Johannes Gerard, Henricus Bertram, in which a font-name is added to the christian name as a byname. Tengvik has noted ten examples in the eleventh century in which, in seven instances, the byname is Scandinavian, in one, Old English, and in two, French. In Domesday Book there is a great increase in the use of French personal-names (40), as well as Old English (28) and Scandinavian (18), with three Celtic and five Latin.18 The general opinion is that such surnames are due to the dropping of filius.19 Tengvik has noted five instances which seem to support this view: Osbern Hauoc, Osbern filius Hauoc; Rainaldus Croc, Rainaldus filius Croc, in which two of the personal names (Hauoc and Dudde) are Old English, two (Baderon and Clamahoc) Breton, and so introduced by Normans, and one (Croc) Scandinavian, though the christian name Rainaldus suggests a possible Norman origin.
It is difficult to believe in this ‘dropping of filius’ theory. A name of the type Johannes filius Willelmi was never used in everyday life by either Englishmen, Frenchmen or Scandinavians. It is Latin and a documentary form. Where the font-name is French, it might be a translation of Fitzwilliam, though such names are unknown in France and rare in English sources. If the font-name is English, filius Dudde might be a translation of Duddesunu, but it is unlikely that one-sixth of the Suffolk peasants of c1095 bore such names. There seems no alternative to regarding these forms as scribal descriptions.
Walter Dudde was known to be the son of Dudda and he was so described in writing, in the clerk’s Latin, filius Dudde. But in ordinary conversation, when his full name was needed, he was called Walter Dudde. Three early examples of this type have been noted by von Feilitzen: the Scandinavian Sendi Arfast (c1044), the French William Ingelram (1088) and Ieduue Ialdit (c1100–30), from OE 20 Nine other examples occur in Suffolk (c1095 Bury), all English: Aldwine Ælfuine, Lemmer Brihtmer, Ordric Wihgar, etc.
The origin of surnames of this type cannot, at present, be definitely decided. The majority of such names are not, as Smith states, from personal names of Scandinavian origin. There are numerous examples from English and French personal names and a smaller number from Celtic. Scandinavian influence may be partly responsible, especially in the Danelaw. The type does not become common in England until after the Conquest and we may be concerned with a French custom introduced by the Normans. Similar names are found in northern France in the tenth century and in the south and south-west in the previous century.21 The fact that similar formations from Old English personal names are common in the south of England in the late thirteenth and the fourteenth centuries and in eleventh-century Suffolk suggests an independent formation in English.
The frequency of the type may be due to all three influences, combined with its simplicity for everyday use and the analogy of similar simple attributes in the form of nicknames and occupation names. It is noteworthy that such surnames from very common christian names like William, John, etc., are late formations.
Post-Conquest Survival of Old English Personal Names
The Norman Conquest revolutionized our personal nomenclature. The Old English namesystem was gradually broken up, Old English names became less and less common and were replaced by new names from the Continent, a limited number of which gradually became more and more popular. The general trend is well known, but many of the definite statements on the relative frequency of various names are based on insufficient evidence, often from late sources which can have little or no bearing on the history of surnames. Fashions in names varied among different classes and in different parts of the country. Most of the early documents deal with the upper classes. Names of peasants are less common, rarely occur in large numbers, and have largely been ignored. Intensive work on the abundant material in manuscript and on that already printed will ultimately throw much new light on the history of our names. For the present, we must be content with a study of selected material from varied districts.
In the Holme Cartulary, some 75 per cent of the twelfth-century names are those of witnesses and grantees of charters, monks, clerics, and tenants of the abbey, and these reflect the new nomenclature introduced by the Normans. The personal names are of French or continental Germanic origin, with Norman forms of Scandinavian names brought from Normandy. The remaining 25 per cent, the names of peasants, represent some 30 English and 35 Scandinavian names, some more than once repeated.
In the twelfth century the most popular names were William (10 per cent) and Robert (7 per cent), followed, with variation of order in different documents and counties, by Richard, Ralph, Roger, Hugo and Walter. John (3 per cent) was much less popular.22 In a thirteenth-century collection of deeds relating to Aveley (Essex), John shares the top place with William (20 per cent) and these, with Robert, Richard, Geoffrey and Thomas, were the names of 160 out of 250 individuals. There were 41 other French names in use, including the Breton Alan, Hervey and Wygan, shared by 86 persons. In addition, 28 persons shared 16 English names, including an unrecorded Weorðing.23 In the fourteenth century Old English names were fewer and much smaller in proportion, 7 persons sharing 5 names. Some 460 persons shared 35 French names. John (34 per cent) was now much more popular than William (18 per cent). Then came Thomas (9 per cent), Richard and Robert (6 per cent), Henry, Roger and Geoffrey, these 8 names being borne by 375 persons. The remaining 32 names were shared by 85 persons.