Establishing a Corpus

Modern forms of place-names are not a reliable guide to their etymology (the words which made up the names when they came into existence, as meaningful descriptions of places). Place-name scholars collect early attestations of the names—the earlier the better—in order to move back in time, as close as possible to the point of naming (and therefore to transparency in meaning). To take just one example, the modern name Englefield Green, Surrey, looks as though it could well belong to a corpus of Engle-names. All but one of its medieval attestations, though, lack <l> (e.g. Hingefelda 967 [S 752; 13th-cent. MS], and Ingefeld, recorded in 1291, 1295, 1317, and so on), which suggests that the etymology of the name is more likely to be the Old English personal name Inga + OEfeld ‘open land’: ‘Inga’s open land’ + modern English green.²⁰ Only with early spellings therefore, is it possible to suggest a likely etymology for a name. Even with a good record of early attestations, there may be further complications in determining with confidence the words which made up the name at its inception. The discussion below assesses place-names attested before 1400 whose early forms suggest that they could be included in a corpus of Engle-, Seaxe-, and Ēote-place-names. Each set of place-names is analysed in turn, with indications given of the level of confidence we have in an individual name’s suggested etymology. The intention is to establish a (relatively) reliable corpus of relevant names, which can then form the basis of a discussion of what these ethnonyms may have signified when they were used as place-name elements.

OEEngle or ONEnglar

The identification of names in OEEngle or Old Norse (ON) Englar is relatively straightforward. As the specific (first element) in a place-name, the ethnonym would appear in the genitive plural form, and, for both the Old English and Old Norse, this is Engla ‘of the Engle’.²¹ Over time, the unstressed genitive inflection -a would have become less distinct (> /ə/), with this represented in the written record by Engle-. Neither Engla- nor Engle- is close in form to other place-name-forming elements, and there are no English or Scandinavian personal names with which they might be confused. In other words, early place-name spellings which begin Engla- or Engle- are generally accepted as containing forms of the ethnonym Engle/Englar. It has also been suggested that an uninflected root-form of the ethnonym, Engel, might appear in Ængelhamstæde, one of the bounds of Wootton St Lawrence, in Hampshire,²² but parallels for such a use of uninflected ethnonyms seem to be confined to names for territories or kingdoms.²³ A singular form of the ethnonym (and possible personal name), Angel, has been suggested as the first element of Anglesey (Cambridgeshire);²⁴ it is not, however, independently attested as a personal name in early medieval England and an alternative etymology for this name is more likely.²⁵ In what follows each of the Engle-names is listed under its generic (second element), with relevant comparative material.

With OE generics

OEburna ‘stream’

Great Englebourne (Devon; O.S. SX 775565)²⁶

Engleborna 1086, Engleburne 1250

Compare Walburn in the North Riding of Yorkshire, whose first element may be Walh ‘Briton’.²⁷

OEfeld ‘open land’

Englefield (Berkshire; O.S. SU 625725)²⁸

(on) Englafelda c. 900, Inglefelle, Englefel 1086, Englefeldia 1154×89

The co-occurrence of OEfeld with ethnonyms is notable: compare Merce ‘Mercians’ in Markfield, Leicestershire, and Markingfield, West Riding of Yorkshire, Daynesfeld in Cheshire (singular Dene ‘a Dane’, or possible personal name), and the two possible Seaxe examples, below.²⁹

OEhām-stede ‘homestead’

Ængelhamstæde (Hampshire; O.S. SU 592532)³⁰

(on) Ængelhamstæde 990 (S 874; mid-12th-cent. MS)

Karl Inge Sandred suggested that <Ængel> here stands for Engel (note <stæde> for stede), a root form of Engle, and that this name could therefore be interpreted as “the homestead of the Anglians.” This root form seems not to occur elsewhere in place-names, nor is OEhām-stede otherwise found with group-names, so a question remains as to whether this is indeed an instance of the ethnonym.

OElane ‘lane’

Ing Lane (now Maiden Lane; London; O.S. TQ 297804)³¹

Englenelane 1282, Inggelenelane 1310–11, Ingelane 1320

This is not quite as straightforward as Eilert Ekwall’s statement, that “[t]he earliest instance clearly indicates a meaning ‘the lane of the Angles,’” suggests.³² If this is indeed the case, it seems to be the only example of the ethnonym with an n-stem genitive ending (*Englena). Given the tendency of other ethnonyms to vary in declension between i- and n-stem, a variant n-stem form is by no means impossible. It is unusual, though of course not unique, for as full a form as Englene- to be preserved in the written record as late as the late 13th century, even for group-names for which we have better evidence of an n-stem derivation. Contextual evidence may support Ekwall’s proposed etymology, as there is a small corpus of ethnonym + routeway names: Denver, Norfolk (‘Danes’ passage’, OEfær); the compounds with ONgata ‘road’ listed below under Sash Gate; there are two possible instances of Bretar/Brettas ‘Britons’ with OEstrēt ‘main road’ in the north west.³³

OEtūn ‘farm, estate’

Engleton Hall (Staffordshire; O.S. SJ 898101)³⁴

Hengleton 1204, Engleton c. 1206–30

The ethnonym + tūn structure is recurrent, although given the general ubiquity of tūn-names this is not surprising. Group-names with tūn include: Brettas/Bretar ‘Britons’, Frīsan/Frísir ‘Frisians’, Īras/Írar ‘Irish, Goidelic speakers’, Norðmenn ‘Scandinavians, Norwegians’, Seaxe/Saksar ‘Saxons’ (see below, p. 134), and probably Cumbre ‘Welsh, Britons’. The distribution of these names as a group has a marked Danelaw weighting, and it is possible that many of them date to the period of Scandinavian settlement or later.³⁵

OEwudu ‘wood’

Inglewood Forest (Cumberland; O.S. NY 453391)³⁶

Englewod’ c. 1150; Engelwode c. 1158; Englewud 1227

Compare Wychwood, Oxfordshire, whose first element is the group-name Hwicce.³⁷

With ON generics

ONbý ‘farm, village’

Ingleby (Lincolnshire; O.S. SK 894778)³⁸

Englebi 1086, c. 1115, Engleby 1166

Ingleby (Derbyshire; O.S. SK 352269)³⁹

Englabý 1009, Englebi 1086, Engleby 1216

Ingleby (Greenhow parish, Yorkshire, North Riding; O.S. NZ 581063)⁴⁰

Englebi, Engleby 1086, 1203–7

Ingleby (Barwick parish, Yorkshire, North Riding; O.S. NZ 441137)⁴¹

Englebi 1086, Caldengilbi 1279, Kaldingelby, Caldingelby 1283

Ingleby (Arncliffe parish, Yorkshire, North Riding; O.S. NZ 447009)⁴²

Englebi, Engleby 1086, 1231, Engelby juxta Erneclyf 1303

While these names have been listed as containing an ON generic, it should be remembered that, in places which had seen Scandinavian linguistic influence as a result of settlement, bý ‘farm, village’ may have been adopted into the local toponomasticon regardless of the cultural or linguistic identity and heritage of those areas’ inhabitants. Nevertheless, bý is certainly of Scandinavian origin, it has a consistent Danelaw distribution, and it compounds with a markedly high proportion of Scandinavian first elements, including identifiably Scandinavian personal names.⁴³ In many cases, bý-names would have arisen in Norse-speaking environments, and can be interpreted as good evidence of Scandinavian-speaking communities in the late 9th century and early 10th.⁴⁴ On contextual grounds an Old Norse origin for these names is most likely—the names occur within clusters of other early-attested bý-names in areas of considerable Scandinavian linguistic influence, and their etymology, ‘farm or village of the English/Angles’, clearly points in this direction. Ethnonym + ONbý is recurrent: examples include compounds in ONBretar ‘Britons’, ONDanir/OEDene ‘Scandinavians, Danes’, ONFrísir/OEFrīsan ‘Frisians, ONÍrar/OEĪras ‘Irish’, and ON/OENorðmenn ‘Scandinavians, Norwegians’.

OESeaxe and ONSaksar

As indicated above, identifying OESeaxe- or ONSaksar-names presents considerably more difficulty than identifying Engle-names. Early spellings which suggest these ethnonyms can also suggest two other (related) possibilities, and there are few definitive characteristics by means of which to differentiate the three.

(1) The weak masculine personal names OESeaxa and ONSaxi are themselves derived from the ethnonyms. We would expect both the ethnonym and personal name, when found as the specific elements in place-names, to appear in the genitive case, and these forms are as follows:

OESeaxe (ethnonym, n-stem, plural) ~ Seax(e)na⁴⁵

OESeaxa (personal name, n-stem, singular) ~ Seaxan

At first glance, this looks promising—the plural and singular genitive forms are distinct, and therefore possible to tell apart. That this difference might, however, be difficult to detect in place-names can be demonstrated by looking at some of the early spellings for Exton in Hampshire.⁴⁶

æt east seaxnatune 940 (S 943; 12th-cent. MS)

Essessentune 1086

Estsexentun c.1127

Exton 1182

In the earliest, pre-Conquest attestation, it is very clear that we are dealing with seaxna-, the genitive plural form of the ethnonym Seaxe. We have here a diagnostic spelling, and seaxnatune must mean ‘settlement/estate of the [East] Saxons’. By Domesday Book in 1086, though, this distinctive ending has reduced to -en. Were it not for the pre-Conquest spelling, we could not be confident, on formal grounds at least, that it is a reduction of -ena; it could represent a reduced form of genitive singular -an (although see further discussion under Saxondale, p. 126). The 1182 form serves as a further warning that complete loss of a genitive plural inflection (or indeed a genitive singular inflection) was also possible by a relatively early date. Original -ena- might only survive in Middle English (ME) spellings as <e>, or it might disappear altogether. In short, if we do not have pre-Conquest attestations we are unlikely to have diagnostic spellings which point clearly to underlying -ena or -an, and relatively few settlement-names are attested before the Domesday Survey.

We should also probably reckon with an alternative OE genitive plural Seaxa, modelled on i-stem ethnonyms, including Engle.⁴⁷ This may be evidenced in suþsaxa londc.848 (S 1193; 13th-cent. MS).⁴⁸ Again, here we have a pre-Conquest spelling which seems to preserve the distinctive -a ending of this i-stem genitive plural, but the quality of the vowel in compounds would have become less distinct and represented in later medieval sources as <e>. A run of late medieval spellings with medial -e- is therefore indicative of an earlier inflection; but it could be the remnant of the plural forms -ena, -a, or of singular -an.

It has also been argued that the root of an ethnonym such as OESeaxe might sometimes have been used uninflected in place-names as Seax-.⁴⁹ Ekwall suggested this for Saxham in Suffolk, Saxton in West Yorkshire, and Saxon Street in Cambridgeshire, but it should be noted that most of his suggested analogues for such uninflected ethnonyms are names of territories or kingdoms (Seaxland, Scotland and Francland) rather than settlement-names.

The Old Norse equivalents, plural ethnonym Saksar and singular personal name Saxi, have indistinguishable genitive forms, Saksa and Saxa (using <ks> and <x> to distinguish is a modern spelling convention)—we simply cannot tell, on formal grounds, which we are dealing with. And again, these -a inflections would reduce over time to -e-, or disappear altogether.

(2) The neuter nouns OEseax, ONsax ‘dagger, sword’ (the ethnonym derives ultimately from the etymon of these two cognates), may be used in place-names with a transferred sense, usually taken to describe (a) a stream, hill or other landscape feature which in some way resembles a dagger; (b) dagger-like flora; (c) cultural practices of weapon deposition in watery features. The first of these, especially in reference to streams, is the more securely attested (see Saxcemede, below). In such cases, when used as the first element in compounds, we would expect spellings in seax-, whether from an originally uninflected form or from a genitive singular seaxes reduced through dissimilatory loss. It should here be noted that the ON cognate of seax might also be reckoned with in English place-names, given its occurrence in Scandinavian river-names (see Saxcemede, below, and references cited there). The second application, posited by Kitson (see Seaxa broc, below), might entail use in the genitive plural (seaxa), which would render it formally indistinguishable from the i-stem genitive plural of the ethnonym noted above. The third possibility has not generally been considered in the place-name literature, but ritual deposition of this nature is attested in early medieval and post-Conquest England.⁵⁰ If such practices could give rise to seax place-names, then this application might be reckoned with for any place-name whose generic refers to a feature associated with wetland or rivers. A meaning ‘cliff, stone, rock’ has been proposed, but is not widely accepted.⁵¹

To summarise, Seax(e)na-forms must derive from the ethnonym Seaxe, but the reduced form Seaxen- might derive either from Seaxe or from the personal name Seaxa; while the reduced forms Seaxa-, Seaxe-, and Seax- might derive from either of those or from OEseax. S(e)axa-, S(e)axe-, and S(e)ax-forms might alternatively derive from either of the Old Norse equivalents, Saksar and Saxi. As the personal names Seaxa and Saxi presumably arose as bynames bestowed on individuals considered ethnically Saxon, the implication of ethnic otherness might still be present in place-names whose first element is this personal name. It could be argued, then, that place-names containing either the ethnonym or the personal name should be considered together. Place-names which contain seax ‘dagger, sword’, on the other hand, are of a different order, even if that word is thought to be the origin of the ethnonym.

The name-by-name analysis which follows takes the above considerations into account, and draws on comparative material, in order to suggest which names might usefully be considered part of the corpus of Seaxe-names.

With OE generics

OEbrōc ‘stream’

Seaxa broc (lost) (boundary point of an estate at Conington, Cambridgeshire; O.S. TL 158876)

Seaxa broc 957 (S 649; 10th-cent. MS)

Formally, Seaxa could be the genitive plural of the group-name or of seax ‘knife’. Kitson suggested that the stream, seaxa broc ‘knives’ brook’, was named from the sharpness of its rushes.⁵² Ritual deposition of weapons might also be reckoned with. The stream has been identified with Holme Brook,⁵³ and forms part of the northern boundary of the estate at Conington, which ran lang broces on gyruwan fen ‘along the brook to gyruwan fen’.⁵⁴ Whitsun Brook in Worcestershire (wixena ‘of the Wixan’) and—less certainly—Walbrook in Middlesex (weala ‘of the Britons’) are possible comparable names.⁵⁵

OEbrycg ‘bridge’

Sackbridge (lost) (in Bottisham, Cambridgeshire; O.S. TL 546606)

Saxbriggemore 1391, Saxbrugge 1429

P.H. Reaney takes this to be uninflected Seax, thus perhaps ‘Saxon-bridge’.⁵⁶ The forms are late, but possible parallels to such a compound can be found: la Denebrygg’ 1303 is closely analogous if its first element is OEDene ‘Danes’ (gen. pl. Dena); so too, perhaps, Weala brucge (944 [S 500; MS c.1240]), with OEwalas ‘Welsh (speakers)’ (gen. pl. weala),⁵⁷ although in each of these cases a genitive inflection appears to be preserved. Also of relevance are the several instances of ethnonyms compounded with other elements referring to river crossings, such as OEford in Denford (Northamptonshire; Dene) and Orford (Lincolnshire; OEĪras ‘Irish; Goidelic speakers’).⁵⁸ Given the possibility that OEseax was used as a river-name (see Saxcemede, below), a similar explanation should also be entertained here, although compounds of brycg with river-names are relatively rare.⁵⁹

OEcot ‘cottage, hut’

Sascott (Shropshire; O.S. SJ 425118)

Saxcote 1086, Sayscot 1274, Saxecot 1308

Sessacott (Devon, O.S. SS 355165)

Saxecot’ 1219

There is little early evidence of an original inflection in the forms for the Shropshire name, but OEseax ‘dagger’ is an unlikely recurrent first element with OEcot ‘cottage, hut’, even in a transferred sense; so an uninflected form of Seaxe or early reduction of a medial inflection might be considered. That means both ethnonym and personal name are possible for these place-names, and the latter was preferred both by J.E.B. Gover, Allen Mawer and Frank Stenton, and by Margaret Gelling.⁶⁰OEcot is very rarely compounded with an ethnonym. An apparent exception is OEwala-cot, but while the primary sense of OEwalh was ‘Welsh (speaker)’, it developed a secondary meaning ‘serf’, and this might be the relevant sense in those compounds. A compound meaning ‘cottage(s) of the serfs’ would be directly comparable with recurrent compounds such as gebūra-cot ‘cottage(s) of the (tied) peasants’ and ceorla-cot ‘cottage(s) of the peasants’, swāna-cot ‘cottage(s) of the herdsmen or peasants’,⁶¹ and not therefore relevant in the present case. Compounds of OEcot with personal names, on the other hand, are relatively common. The personal name Seaxa seems most likely, then, but the double occurrence of the compound in places with later survival of Welsh and Cornish cultures is noteworthy, as is the proximity of Saxworthy (discussed below), just across the River Torridge from Sessacott.

In each case, the development /saks(ə)kot/ > /sas(ə)kot/ can be explained by dissimilatory deletion of the first /k/. There may also have been some Brittonic influence on the development of each name: the apparent diphthong of the 1274 form for Sascott might indicate influence from Welsh Sais ‘Englishman’ (compare Seisdon, below);⁶² while it may be Brittonic prosody—placing stress on the penultimate syllable—that has preserved the medial syllable in the modern form of Sessacott.

OEdæl / ONdalr ‘valley’⁶³

Saxedele (lost field-name in Gayton Thorpe, Norfolk; approx. O.S. TF 745185)

Saxedele e. 13th

Saxedale (lost; in Withcall, Lincolnshire; approx. O.S. TF 285839)

Insley treats the first element of the Norfolk name as OESeaxe or ONSaksar, thus ‘valley of the Saxons’. Gover, Mawer and Stenton note a mention of the apparently identical Saxedale in the Kirkstead cartulary, but provide no further information.⁶⁴

OEdenu ‘valley’

Saxondale (Nottinghamshire; O.S. SK 683400)

Saxeden 1086, Saxenden 1316; Saxendala c.1130, Saxendale 1199

This is a rare instance where post-Conquest forms may be diagnostic. There is some evidence to suggest that an original inflectional /n/ was more likely to survive in the late medieval reflexes of place-names when immediately followed by a vowel, as would have been the case with medial -(e)na-. The /n/ in place-name compounds with genitive singular -an seems generally to survive only when the following element begins with a vowel or /h/ plus vowel. So, we might argue that the early (and current) spellings for Saxondale suggest that OESeaxena-denu ‘valley of the Saxons’ is the most likely interpretation of the name.⁶⁵ The second element seems to have interchanged for a while with OEdæl ‘pit, hollow; valley’ or ONdalr ‘main valley’, which eventually supplanted it.⁶⁶

OEdūn ‘hill’

Seisdon (Staffordshire; O.S. SO 836947)

Saisdon hvnd’, Saisdone hvnd’, Seiesdon hvnd’ 1086, hdr’ de Saiesdona 1130, Seidon’hundredum 1182, Seisdon’hundredum 1185; Seisdone 1086, Seisdun 1160–1206 (personal name), Seyxdun 1235, 1236, Seydon’ 1236, Seisdon’ 1242

Association of Seisdon with Seaxe or Seaxa is problematic. Anderson and others have attributed the development of /ks/ > /s/ or /z/ (indicated by the surviving forms and modern pronunciation) to Anglo-Norman influence.⁶⁷Although <s> forms begin early and are almost universal, early Anglo-Norman influence on the name of an administrative centre is feasible (Seisdon was the name of a Domesday hundred).⁶⁸ Unlike most other Seaxe or Seaxa place-names, however, the spellings listed above suggest preservation of the first vowel as a diphthong. Attempts to identify it with Welsh Sais ‘Englishman’, which might explain the diphthong, have not been received entirely enthusiastically, and it is hard to account for an English place-name compound with Welsh Sais as first element, or for a Welsh place-name compound with OEdūn as the generic.⁶⁹ Welsh influence, as suggested for Sascott (above), seems rather less likely this far east, unless it occurred very early. It is hard to conclude with much conviction that Seisdon is a Seaxe place-name.⁷⁰

OEēa ‘river, stream’

Saxcemede (lost, nr Puddletown, Dorset; O.S. SY 755945)

Saxcemede 1270 (1372 MS)

Formally both Seaxe and the personal name Seaxa are possible, given the lateness of the only surviving attestation, but A.D. Mills prefers OE *Seax-ēa ‘stream called Seax’, where Seax means something like ‘the sword, the bright one’, noting use of the ON cognate sax in Scandinavian river-names.⁷¹ Ritual deposition of weapons might be an alternative explanation of the use of seax in such names. (Compare also Seaxa broc, above.) Either of these is probably more attractive than derivation from a compound such as Seaxa-mēd ‘meadow of the Saxons’, since OEmēd is rare as the second element in compounds with an ethnonym. Compare Common Mead in Dorset (Comermede 1300), where the first element might be Cumbre ‘Britons’, but Mills prefers a personal name Cumbra;⁷² and Walmead in Wiltshire (Walemade 1100–35), with OEwalh ‘Briton; serf’ (and see the discussion of wala-cot under Sascott, above).⁷³

OEfeld ‘open land’

Saxfield (Cheshire; O.S. SJ 823901)

Saxefeld 13th cent., c.1290, Saxe Felde 1578

Saxifield (Lancashire; O.S. SD 862352)

Saxifeldyk 1324, Saxxefeld 1425, Saxyfeld 1507

For the first of these names, either OESeaxa-feld or Seaxena-feld ‘open land of the Saxons’ is formally acceptable,⁷⁴ but so is Seaxan-feld ‘open land of a man named Seaxa’. For Saxifield, spellings with medial <i> and <y> seem problematic and earlier attestations would be required to establish their precise significance. Ekwall suggests a personal name as first element, either OESeaxa or ONSaxi.⁷⁵ The genitive inflections would normally be reduced to /ə/, giving ME spellings in <e>, as also (in most instances) the genitive plural inflection of OESeaxe. The <i> and <y> may simply stand for /ə/, or else we may be dealing with an unattested OE *seaxing ‘knife-shaped feature’, if the topography is appropriate (much of the area is now built up). It is worth noting that personal names often form the specific in compounds with OEfeld,⁷⁶ but the twin occurrence of the collocation of Sax- with OEfeld is perhaps suggestive of the ethnonym rather than a personal name. It might be compared with Englefield (above), and the other examples of ethnonym + feld noted there.

OEhalh ‘nook’

Saxehale (lost, nr Stutton in Tadcaster, West Riding of Yorkshire; O.S. SE 479415)

Saxehale, Saxhale, Saxhalla 1086

This is another of the rare occasions when the post-Conquest forms may be partly diagnostic. With no more than Domesday to go on, caution is required, but the surviving spellings probably rule out both Seaxena-halh and Seaxan-halh. When followed by /h/ plus vowel, some trace of the /n/ might be expected. This might therefore be OESeaxa-halh ‘nook of the Saxons’, or perhaps OEseaxa-halh ‘daggers’ nook’, if characterized by particularly dagger-like flora (compare Seaxa broc, above).⁷⁷

OEhām ‘homestead, settlement’

Great/Little Saxham (Suffolk; O.S. TL 794632)

Saxham 1086, 1197, Sexham 1086

There is no evidence for an originally inflected first element, and the forms count against Seaxena- or Seaxan- (see the preceding entry), casting doubt on Walter Skeat’s interpretation of the first element as a personal name. Ekwall suggests OE *Seax-hām ‘homestead of the Saxons’, rejecting OEseax in a topographical application and (on formal grounds) the personal name.⁷⁸ Victor Watts also prefers the ethnonym, as do Keith Briggs and Kelly Kilpatrick, though they offer the personal name as an alternative.⁷⁹ It should be noted that OEhām ‘homestead’, or OEhamm ‘enclosed land’ (from which hām can be hard to distinguish in place-names), occur rarely in compounds with ethnonyms: there are few if any certain instances with reference to groups of Angles, Jutes, Danes, Frisians or Northmen; only with OESwǣfe ‘Swabians’ has such a compound been proposed with any great force.⁸⁰

OEmearc ‘boundary’

Seaxana meare (lost, nr Warehorne (?), Kent; O.S. TQ 985325)

Seaxana meare 845 for 830 (S 282; 11th-cent. MS)

Paul Cullen follows Wallenberg in taking the second element to be OEmearc, and the first element is clearly a weakly inflected genitive plural of OESeaxe, thus ‘boundary of the Saxons’.⁸¹

OEsēað ‘pit’

seaxe seað (Slackstead, Hampshire; approx. O.S. SU 396257)

on seaxea seað of seaxe seaðe, on seaxe seaðe 900 (S 360; early 11th-cent. MS); on seaxes seaðe 899×924 (S 381; 12th-cent. MS)

Grundy takes this to be a personal name, or rather a by-name, thus ‘Saxon’s Pit’.⁸² We may then be dealing with an unattested strong personal name *Seax, comparable with the Scandinavian name discussed under Saxthorpe, if the form in S 381 is the more reliable. The other forms might result from reanalysis of seaxes seaðe as seaxe seaðe, but the process could be the other way round. Kitson notes an association of sēað with watery features, so another instance of seax-ēa as a river-name is possible (whatever the motivation, see Saxcemede), or a reference to dagger-like rushes (see Seaxa broc), although the spellings do not necessarily indicate a genitive plural form.

OEstede ‘place’

Saxstead (Suffolk; O.S. TM 265655)

Saxsteda (Skeat gives Saxteda)

Skeat is followed by Ekwall (who considers but rejects OE *seax ‘rock’), Mills, Watts, and Briggs and Kilpatrick in assuming a personal name Seaxa, with OEstede ‘place’.⁸³ Few forms listed by any of these commentators could be considered diagnostic of that personal name in preference to the ethnonym, so OESeaxe is at least a possibility. The only other possible instance of ethnonym + stede is Irstead in Norfolk. Its early forms do not rule out a form of OEĪras ‘Irish’ as first element, but OEgyr ‘mud, filth, marsh’ has been preferred by commentators.⁸⁴

OEtūn ‘farm, estate’

Saxton (West Yorkshire; O.S. SE 476369)

Saxtun, Saxtuna, Saxton, Saxtuna 1086

Saxon Street/Saxton Hall (Cambridgeshire; O.S. TL 675595)

Sextuna, Sextone 1086, Sexton, Sextone 1208 Cur (personal name), Saxton, Saxtone 1236

For the first of these names, A.H. Smith noted the possibility of a connection with Saxons, but preferred a personal name.⁸⁵ For the second, Reaney was unconvinced by Seaxe for lack of ME forms with medial <e>, and felt that OE *seax ‘hill’ was topographically inappropriate;⁸⁶ he therefore posits the root Seax- used in compound to mean ‘Saxon’.⁸⁷ Certainly there is no reflex of an original inflection in the forms listed here, although, as noted above, that need not rule it out. Compounds of the type ethnonym + tūn are recurrent (see discussion of Engleton, above).

OEworðig ‘enclosure’

Saxworthy (Devon; O.S. SS 375175)

Sexaworthi 1330

Gover, Mawer and Stenton take the first element to be a personal name, but OESeaxe cannot be dismissed on formal grounds.⁸⁸ Saxworthy is located just across the River Torridge from Sessacott, and a connection with that name is possible.

With Brittonic generics

Welsh pen ‘chief, head; promontory’

Pensax (Worcestershire, O.S. SO 725695)

Pensaxan (dat.) 11th cent.⁸⁹

This is a Brittonic place-name, formed as a name-phrase from the ancestors of Welsh pen ‘head, top’, and Sais ‘Englishman, Saxon’.⁹⁰ This accounts for the order of the elements—the generic element pen precedes the ethnonym specific. It should be considered alongside the lost name Pensousen (1582; Cornwall), and Pennersax (or Pennersaughs) in Annandale (Middlebie, Dumfries, Scotland; Penresax 1194, Pennyrsax 1194),⁹¹ name-phrases containing reflexes of the same elements.

Sixpenny Farm (Dorset; O.S. ST 844169)

seaxpennes 932 (S 419; 15-cent. MS), Sexepenne 1340 (personal name)

Sixpenny was also the name of an 11th-century hundred (Sexpene 1084), part of the later hundred of Sixpenny Handley.⁹² R.E. Zachrisson suggested ‘Saxons’ mountain top’, but this was rejected by Anton Fägersten, who preferred OE*seax in the sense ‘cliff, stone’.⁹³ While that sense is unrecorded, it is clear that the first element could be an instance of the transferred use of OEseax in reference to a topographical feature. Mills, however, follows Zachrisson’s interpretation of the first element as the uninflected root of OESeaxe; the generic is ultimately of Brittonic origin, but was perhaps borrowed into some OE dialects.⁹⁴ It is difficult, however, to consider this name without reference to Pensax, Pensousen and Pennersax. In each of those examples, the name is apparently a reference by Brittonic speakers to a headland or similar feature associated with one or more Saxons, and the recurrence of the association with pen suggests that the collocation may have been a feature of Brittonic lexis. In view of that, it is tempting to see Sixpenny as a further example of the same Brittonic elements present in those names, with the same meaning, but formed as a true compound, with specific preceding generic. Compounds seem generally to have been favoured in the Brittonic languages up to the 5th century, name-phrases from the 5th century onwards, which would require Sixpenny to have been coined very early in the Anglo-Saxon period; but Oliver Padel points out that this general rule is not a hard and fast one, and that true compounds could be used in the coining of place-names at a later date.⁹⁵ Chronologically, there seems little reason to rule out the possibility that Sixpenny originated as a Brittonic place-name, even if its surviving forms are the result of Anglicization. A difficulty with this interpretation, however, is the dearth of evidence for pen as the generic in early Brittonic place-names.⁹⁶

With ON generics

ONbý

Saxby (Leicestershire; O.S. SK 820201)

Saxebi 1086, Sessebia c.1130, Saxenebi 1175, 1198

Saxby All Saints (Lincolnshire; O.S. TF 004861)

Saxebi 1086, 1166

Saxby (Lincolnshire; O.S. TF 005862)

Sassebi 1086, Saxsabi, Saxsebi c.1115, Saxebia c.1155

These three compounds in ONbý ‘farm, settlement’ should be treated together, although different etymologies have been proposed. Formally they could all contain a reference to Saxons. The late 12th-century forms for the Leicestershire Saxby point to the genitive plural form of OESeaxe (Seax(e)na),⁹⁷ although it is possible that these spellings represent a ‘translation’ or reanalysis of an underlying Old Norse specific. All other spellings could represent either OESeaxe or ONSaksar. While an ethnonym interpretation was favoured by Fellows Jensen,⁹⁸ Ekwall and others have preferred the ON personal name Saxi, citing the frequency with which this personal name occurs in eastern England as an explanation for the recurrence of the same compound.⁹⁹ The compound Ingla- or Engla-bý, discussed above, is also recurrent, as are Dena- or Dana-bý ‘settlement of the Danes’, Frīs(en)a- or Frísa-bý ‘settlement of the Frisians’. The Saxby names fit within this pattern and should probably be derived from an ethnonym, whether OESeaxe or ONSaksar.

ONgata ‘road’ (or late OEgate)

Sash Gate (Lincolnshire; O.S. TF 062947)

Saxgate 1280×85

Cameron offers both the ON personal name Saxi and uninflected OESeax- or ONSaks-. If the latter, the place-name would mean ‘the road of the Saxons’.¹⁰⁰ONgata is several times compounded with other ethnonyms (see also Ing Lane, above), including Jubbergate (Brettegata 1145–55) in York, from ONbretta-gata ‘road of the Britons’;¹⁰¹ four instances in Lincolnshire of ON *Dana-gata ‘road of the Danes’ and three possible instances of *Norðmanna-gata, two in Leicestershire and one in Northamptonshire;¹⁰² Flemingate in Beverley, Yorkshire (ONFlæmingjar ‘Flemings’); the lost Scotgate in Lincolnshire (OEScott ‘Irish person, Scots Gael’).¹⁰³ This makes a compound ONSaksa-gata ‘road of the Saxons’ a strong possibility.

ONþorp ‘secondary settlement’

Saxthorpe (Norfolk; O.S. TG 115305)

Saxthorp, Saxiorp, Sastorp 1086, Saxtorpe 1146–8, Sakesthorp 1244¹⁰⁴

Sandred takes this to contain the personal name Saxi, but Insley believes the spellings to be indicative of a strong form *Saks, which seems to occur in Danish place-names.¹⁰⁵ONþorp is compounded with ethnonyms in a small number of cases, such as Friesthorpe (Lincolnshire) and Danthorpe (Yorkshire),¹⁰⁶ but is common with personal names (especially within the Danelaw).¹⁰⁷

OE *Ēote (*Ῡte, *Īote)

Alistair Campbell stars these, presumably because the nominative form is, essentially, hypothetical;¹⁰⁸ an indication of just how rarely Jutes seem to be mentioned in Old English texts. The corpus of place-names traditionally identified as containing OEĒote is small and not especially well attested. It is set out below in alphabetical order, rather than grouped by generic.

Bishopstoke (Hampshire; O.S. SU 465195)

(apud) itinstokan 959 (14th-cent. MS), (æt) yting stoce 960 (S 683; 12th-cent. MS), Stoches 1086¹⁰⁹

Richard Coates cautiously notes that the import of the forms “is obscure, but they may contain a reference, of very great antiquity, to Jutes.” He points out, however, that the form Yting “appears to be a singular -ing name here,”¹¹⁰ which weighs against that etymology. Gelling suggested a personal name derived from the ethnonym (compare Eatenden, below), and Insley takes the compound to mean ‘place associated with the Jute(s)’.¹¹¹ The final element is OEstoc ‘outlying settlement’, which otherwise does not seem to be attested with ethnonyms.

Eadens near East Meon (Hampshire; approx. O.S. SU 679221)

Ytedene 1263, 1301, 1453, Itedene 1350, Eadens 1840

Gover takes this to be ‘valley of the Jutes’, OE *Ῡte (the West Saxon variant of *Ēote) with OEdenu.¹¹² The forms certainly allow for an original Ῡtena- or Ῡta-denu, but the personal name posited for the preceding entry is again possible.

Eatenden (Mountfield, Sussex; O.S. TQ 730186)

Itintune 12th cent., Ytenton, Itenton c.1200, Ytintun 1261

This does not seem previously to have been considered in connection with OEĒote. It has been interpreted as an OE -ingtūn compound based on an unattested personal name *Ita, a pet form of dithematic personal names in Id- or Ith-, and a cognate of ohgIzo. The place-name could thus be interpreted as ‘Ita’s farm or settlement’.¹¹³ The forms, however, bear comparison with those listed under Bishopstoke and Eadens, and it therefore merits inclusion in this corpus. There is no reason to assume that the initial vowel was a short one—indeed, the modern form suggests a long vowel. The early spellings, as they stand, would not rule out OE *Ῡtena-tūn ‘farm or settlement of the Jutes’, or an -ing formation comparable with that proposed for Bishopstoke.¹¹⁴

Ytene, associated with the New Forest (Hampshire)

in Nova Foresta, quae lingua Anglorum Ytene nuncupatur c.1115

The form Ytene, taken from the chronicle of John of Worcester, seems to represent West Saxon OE *Ῡtena ‘of the Jutes’, and Ekwall suggested expanding it to ME *Ytenewode ‘woodland of the Jutes’.¹¹⁵ In the same chronicle we have in provincia Jutarum, in Nova Foresta.¹¹⁶

It is unfortunate that the earliest attested of these names is so problematic as an example of OEĒote. A compound Ēote + -ing + stoc would be a very unusual one on two counts: -ing-suffixed ethnonyms are otherwise unattested, and OEstoc is rare with ethnonyms. Even if, as Gelling suggests, the first element is a personal name derived from the ethnonym, it is not direct evidence that the inhabitants were Jutes, unless it arose as a byname *Ῡting ‘son of the Jute’. Eadens is similarly problematic, since it could be derived from the same personal name. If that purported personal name were only identified in place-names located in and around the area thought to have been settled by Jutes, this might cast doubt on its existence; but it does seem to occur elsewhere. Edford Brook (Lilleshall, Shropshire), for instance, is Eotanford 963 in the bounds of Church Aston, apparently ‘Eota’s ford’. The unrecorded personal name *Ēota would be the normal west midland form of West Saxon Ῡta. If Ῡta has an ethnonymic function in the Hampshire names—if, for instance, it is a byname meaning ‘the Jute’—then we should reckon with it also in Shropshire. That leaves only John of Worcester’s Ytene. The association in the same text of provincia Jutarum with the New Forest makes it very likely that this is a genuine reference to Jutes; but, in the form in which it has been passed down, it is not a place-name. The chronicler may have taken it from an OE text in which it was part of a hypothetical *Ytenewald, but that is not evidence for its more general currency in spoken contexts.

Discussion

It is clear from the analysis above that identifying Seaxe- and *Ēote-names is especially challenging. However, patterns of compounding shared with other group-name place-names may support the inclusion in our corpus of a considerable number of these names, even when the early forms allow for other interpretations. Determining the sense and significance of the ethnonyms within the corpus of place-names assembled above is not straightforward. What follows is an assessment of the possible significances of the more securely identified group-name place-names.

Many—if not all—of the place-names under discussion have been interpreted by scholars as containing references to distinctive minority groups, communities who were notably different in some way from their surrounding populations. It is this assumption that lies behind Ekwall’s statement, “[i]t is a priori obvious that places were not generally named from the tribe or nation in whose territory they are found.”¹¹⁷ These names are usually interpreted as having arisen in the speech of surrounding populations, rather than as self-identifying labels.

Even within one subset of our corpus—the Engle/Englar names, for example—the cultural and/or linguistic identity that made the referents distinct was unlikely to have been uniform. As was noted above, the group-names could have varying referents in the documentary record. The same variation in application might be observable in the place-names, and indeed place-name usage might depart significantly from that of written documents, whose contexts of production were very different. What, if anything, can be said about the ways in which these group-names were used in place-names?

For the English-language Engle-names, it is usually the more specific—more Bedan—sense of ‘Angle’ that is thought to be active. A.H. Smith summarised, “Outside East England, the E[ast] Midl[ands], and the N[orth] C[ountry], the word refers in p[lace]n[ame]s. to an isolated settlement or group of Angles.”¹¹⁸ They tend to be seen as evidence of migration from further north or east in England.

Englebourne in Devon can readily be interpreted as one such ‘Anglian’ name. Ekwall—and Smith—writing in the 1950s, saw the Engle here as a distinctive minority group within Wessex.¹¹⁹ An English-language name must of course have been formed at some point after the West Saxon expansion into this area in the late 7th to early 8th centuries. Englebourne’s genesis was probably considerably later than the initial movements westwards, as the scant evidence there is suggests that, in Devon, Old English speakers initially adopted Brittonic place-names, which were only later replaced by Old English formations.¹²⁰ This replacement was largely complete by the time of the Domesday Survey, at least for the types of settlement-name contained therein, but it is evidenced in texts from the 9th century.¹²¹ And by the late 9th century Englebourne was, if not exactly West Saxon heartland, very nearly so—a number of nearby estates were bequeathed by King Alfred, in his will, to members of his family.¹²² Angles may well have been notable in such a context. Back in 1931, however, Gover, Mawer, and Stenton, the English Place-Name Society (epns) editors for Devon offered a rather different reading. They suggested that Englebourne “may mean ‘bourne or stream of the English’, possibly because at one time it formed the boundary between the British and the invading Anglo-Saxons.”¹²³ The notion of a fixed boundary between Britons and Anglo-Saxons is of course problematic from a 21st-century perspective,¹²⁴ but it is perhaps worth exploring the idea that the name referred in some way to the (notional) extent of English control. At first glance, Gover, Mawer, and Stenton’s suggestion, with its reference to “invading Anglo-Saxons,” seems to rest on the assumption that the name was formed at the time they were moving westwards into this area. The name is perhaps unlikely to date from this period for the reasons mentioned above, and the use of Engle at this early date is also problematic. If the (English) name had been formed during a period of West Saxon expansion into Brittonic-speaking regions, it would of course have to have been self-identification on the part of the English speakers. Self-identification as Engle, rather than Seaxe, at a time earlier than that for which we have evidence that Engle could denote English speakers regardless of regional origin is improbable. Gover, Mawer, and Stenton may, though, have envisaged a later context for the formation of Englebourne, and it is possible to imagine it arising, with this ‘boundary’ sense, as a name which preserved—or invented—a folk-memory of an earlier border between English and British, at a time after Engle had come to mean English speakers. If English-language names in this part of the world originated, on the whole, in the 9th and later centuries, then this is at least a possibility.

Englefield in Berkshire is the other Engle-name for which the “contrasting” population is thought to be West Saxon.¹²⁵ Berkshire was contested territory, and came under Mercian rule in the late 8th and early 9th centuries. Margaret Gelling pointed out that the phonology of the county’s place-names was not obviously affected by these periods of Mercian rule—the names owe their forms to the West Saxon dialect. This suggested to her that Mercian overlordship was just that, and “did not affect the composition of its population,” and that “[t]he Angles of Englefield […] would not have been mentioned in a place-name if they had not been exceptional.”¹²⁶ Citing Englefield as evidence of the lack of any settlement concomitant with Mercian dominion has the effect, probably unintended, of identifying the name’s Engle with Mercia. Engle did not necessarily refer to settlers from “core” Mercian territory. Their linguistic or cultural distinctiveness may have had its origins well outside that territory—that Engle could refer in place-names to a non-Mercian Anglian identity may be suggested by its use in Engleton in Staffordshire.¹²⁷ Englefield’s Engle could have been Middle or East Angles.

A memorializing function for Englefield was suggested by Ekwall, who offered a “mere guess” that “the name commemorates an early battle between Angles and Saxons,”¹²⁸ and by Andrew Reynolds, who suggested that the name may have been a deliberate bestowal by the West Saxons on the place where in 871 Æthelwulf, ealdorman of Berkshire, defeated two jarls of the Great Viking Army.¹²⁹ Reynolds’ contention is that the term Engle here reflects a conscious use of language which stresses English unity. This has most commonly been detected in the use of Angelcynn, from the late 9th century and primarily in texts associated with Alfred’s court, to refer to both Mercians and West Saxons as a unified people.¹³⁰ These arguments rest, at least in part, on feld here having the sense of battle-field, and this is indeed one of the word’s senses.¹³¹ However, it seems unlikely to have been the one at work in Englefield when we look at the name in context. It is one of a swathe of feld-names found in West Berkshire and neighbouring Hampshire (Fig. 6.1).¹³²Feld here seems to have been used in contrast to forested land just to the east. Stripped of commemorative function, it is difficult to understand how a general sense of Engle ‘English’ would work in this particular location. A more specific ‘Anglian’ sense seems more likely, especially as feld is relatively common in the earliest attested place-names, and may belong to a fairly early stratum here rather than to the late ninth century (or later).¹³³

View Full Size

Feld-names in Berkshire and Hampshire map produced by Richard Jones

• Download Figure
• Download figure as PowerPoint slide

Engleton in Staffordshire is situated, as Ekwall observed, in “old Anglian territory;” he therefore saw the names as having arisen at some point after Staffordshire’s identity had become distinctively Mercian.¹³⁴ For Ekwall, these Engle were incomers from further east, Middle or East Angles, defined against—and presumably by—a surrounding population whose primary identity was at that point not “Anglian.” This population need not necessarily have identified as Mercian, of course; more local identities may have been at work. More recently David Horovitz has suggested that this area was “mainly British” at the point at which the name came into existence, noting that the British-OE hybrid name Brewood (Brit. brigā ‘hill’, OEwudu ‘wood’) is less than 2 km away.¹³⁵ While the name Brewood is not necessarily in itself compelling evidence for the late survival of a British-speaking community,¹³⁶ Engleton is certainly far enough west to imagine significant continuity in population in this area from pre-English times through to the later Anglo-Saxon period; retention of British speech was not necessarily part of that continuity. Ekwall and Horovitz envisage significantly different social and linguistic environments and processes of naming. The first is an entirely English cultural environment, with the place-name arising descriptively in speech. The second is a situation of cultural contact, and possibly one where the place-name was imposed by the settlement’s inhabitants, an example of self-identification. It was noted above (pp. 119, 130–31) that many place-names of the type ethnonym + tūn occur in the Danelaw and suit a context after the Scandinavian settlements. This adds further uncertainty to any interpretation of Engleton. It could suggest a later context for the name, which in turn would count against a linguistically or “culturally” British environment.

Inglewood (Forest) in Cumberland is perhaps the only Engle-name for which we can suggest with some confidence a particular context of production. There seems to be general agreement that here we have English or Angles defined against a population which was in some sense British. Ekwall suggests that these Engle were “in Welsh territory,” Smith that they were “amongst the Cumbrian Britons,” and the epns editors and Watts plump for the Britons of Strathclyde.¹³⁷ The area so named would have been part of the kingdom of Northumbria before its collapse in 867, and though enclaves of Brittonic speakers may have survived Northumbrian rule these were unlikely to have been land-owners.¹³⁸ A significant Brittonic-speaking population re-emerged only in the late 9th or 10th century, with the southwards expansion of the kingdom of Strathclyde.¹³⁹ Inglewood, between Carlisle and Penrith, formed part of the southernmost extent of this kingdom, “an inland border zone that comprised Inglewood Forest and the Eamont/Penrith area.”¹⁴⁰ Given that the population here and further north and west comprised speakers of several different languages—English, Scandinavian, and Goidelic, as well as Brittonic—it seems more likely that Inglewood is a name referring to political rather than linguistic identity, one that was given in recognition of the fact that this area was where British political control petered out. This is perhaps not a case of a ‘minority population’ name, but one used with a particular political significance to mark a borderland area.

View Full Size

OEEngle-namesmap produced by Richard Jones

• Download Figure
• Download figure as PowerPoint slide

It is perhaps of interest, then, that when we look at the distribution of the more secure OEEngle-names—leaving, for a moment, Ing Lane—they are all located in what might be termed borderland areas, where one set of historic identities or allegiances intersected with another (Fig. 6.2). Englebourne lies within Anglo-Saxon territory but not far from areas where Brittonic speech endured; Engleton, in the west midlands, lies in an area in which British identities, cultural and linguistic likely lasted longer than areas not much further east; and the name Inglewood seems to have arisen in the context of British territorial expansion into English speaking (or linguistically mixed) areas. Englefield lies within territory claimed at various points by Mercia and Wessex. There is perhaps something more to these names than their referring ‘observationally’ to minority groups, then. Inglewood looks like a reference to the limits of political power, and, while the other names may not have this precise function, it may be that ethnic identifiers arose in historically contested areas, even if that contest was a relatively distant memory at the point at which the name came into existence.

View Full Size

Seaxe distribution mapmap produced by Richard Jones

• Download Figure
• Download figure as PowerPoint slide

A distribution map of OESeaxe-names is strikingly different from that of Engle-names, both because of the markedly greater number of possible instances (Fig. 6.3), and because of the number of names located within Scandinavian-influenced areas. Some of these names may fit a ‘borderland’ interpretation.

Seaxana meare refers directly to a boundary, of course, and is possibly therefore a special case. The estate of Flodhammas, whose southern border it comprised, lay within Walland Marsh and the parishes of Brookland, Brenzett, and Fairfield, all of which abut Kent’s boundary with Sussex. Seaxana here may be elliptical, or it may represent an everyday “vernacular” reference to South Saxons. The more specific Suþsaxa lond (see above, p. 115) lay less than 10 miles to the east, and may represent a detached part of South Saxon territory. It is perhaps of interest that we have a “full” form of the group-name in such a context, with what may be a direct reference to land-holding within Kentish territory by a political unity.

Sessacott and Saxworthy are located on the western periphery of Anglo-Saxon England, close to the Anglo-Cornish border, and their motivation is a matter of some curiosity, if, as discussed above, they contain the personal or by-name Seaxa rather than the ethnonym Seaxe. It may be that the social conditions of the border region gave rise more readily to use of the ethnonym in given names or as a byname. The use of OESeaxa ‘the Saxon’ as a by-name in Sascott, however, on the edge of supposedly Anglian Mercia, is harder to interpret in this way, unless we reckon with some sort of semantic influence from Brittonic *Saχso-, whose referents were English rather than specifically Saxon. It is not difficult to imagine such influence on the English spoken in this part of the west midlands, whether that of “original” Germanic-speakers or of those who had learned it under political exigency. The other west midland name in the corpus, Pensax, a Brittonic place-name, is very likely to contain the ethnonym, whether in reference to an individual Saxon or a community of Saxons. In Della Hooke’s view, the name arose because “a group of Anglo-Saxon settlers were sufficiently distinctive in north-west Worcestershire to give their name … to the estate of Pensax … in the woodland beyond Wyre.”¹⁴¹ Whether this distinctive group was an isolated community surrounded by Brittonic speakers, or dwellers in the borderland at the limits of Anglo-Saxon expansion, the name seems likely to have arisen at an early stage, when Anglo-Saxon culture or Old English language were distinctively rare in this part of Worcestershire. In Gelling’s judgment, Old English had come to dominate even in districts much further west by the end of the 9th century.¹⁴² An early date was also proposed by Kenneth Jackson, on phonological grounds, although Alan James demurs.¹⁴³ Full discussion of this name belongs with a detailed analysis of comparable Brittonic material, and is beyond the scope of this chapter. Whether Sixpenny should be treated separately is moot. Its location is much less clearly culturally liminal, except in the very earliest of Anglo-Saxon contexts—Anderson posits “an ancient Saxon boundary” in explaining the name’s motivation.¹⁴⁴

Another group of Seaxe names might be considered to have arisen in a border environment: Saxon Street, Saxham, and Sackbridge. Ekwall considered them remnants of an early Saxon migration to this otherwise predominantly Anglian area,¹⁴⁵ but all three are located about 10 miles north of the modern county boundary of Essex. If that is a good approximation of the extent of East Saxon territory in the Anglo-Saxon period,¹⁴⁶ then it is not inconceivable that control of parts of southern Cambridgeshire and Suffolk was contested by the East Saxons at various times, or that East Saxons migrated into these nearby areas after the initial settlement period. It is harder to make such an argument for Saxstead, around 20 miles north of the boundary, but in any case the personal names is probably more likely than the ethnonym here.¹⁴⁷

The remaining OESeaxe-names are harder to read as expressions of border dynamics, and the conventional interpretation, as isolated communities of Saxons in areas whose identities Bede classed as collectively Anglian, may be preferable. Dodgson considered Saxfield to commemorate “a settlement of Saxons among the Anglian population of north-east Ch[eshire];”¹⁴⁸ Gover, Mawer and Stenton considered Saxondale in Nottinghamshire to record “some isolated Saxon settlement in this Anglian territory.”¹⁴⁹ Gover et al. went further, in fact, and suggested that the presence of a Saxon population might account for use of OEdenu, otherwise rare in Nottinghamshire; this seems not to be entirely supported by more detailed analysis of that element,¹⁵⁰ but the recurrence of generics referring to valleys and nooks, in Saxondale, Saxehale, and the two instances of Saxedale, is worthy of note. Parallels can be found in the place-name Walden (Essex, Hertfordshire), from OEWala-denu ‘valley of the Britons’, and Eadens in Hampshire (above, p. 136) might also be compared.¹⁵¹ The historical context for these isolated communities is impossible to reconstruct, since place-names are generally their sole evidence and we cannot date their coining very precisely. Does a name such as Saxfield, near the border of Lancashire and Cheshire, tell us that Saxons and Angles were both present in that region in the early to mid Anglo-Saxon period, or that a group of Saxons migrated to this otherwise Anglian land at a later date? The name Exton shows that movements of this kind could take place and be recorded in place-names after the establishment of a distinctively East Saxon identity. In an otherwise apparently non-Saxon region, OESeaxe might sometimes have been short for Ēast, Sūð or West Seaxe, as may have been the case in Seaxana meare on the border with Kent. On the other hand, if many of the names originated early on, they presuppose a much more mixed settlement of Angles and Saxons that belies Bede’s later rationalization.

The exceptions here are the Jutes, whose name seems to be confined at most to one Jutish area only, and is not apparently found elsewhere. This may be because they have not been looked for beyond their Bedan districts; or because the Jutish identity was subsumed by a West Saxon one early on. John Insley contends that “[t]he separate Jutish identity [of parts of Hampshire and of the Isle of Wight] lasted for a considerable time.”¹⁵² This rests on seeing Asser’s reference to Alfred’s maternal grandfather, Oslac, as de Gothis et Iutis ‘from the Goths and Jutes’ as reflecting a late 9th-century consciousness of Jutish origins in those regions, to which Oslac’s family belonged.¹⁵³ It is hard to know, however, whether this is anything more than a reflection of learned tradition or acquaintance with Bede. The scarcity of *Ēote-names does not especially weigh in favour of the Hampshire/Wight population continuing to identify with the Jutes for an extended period of time. If this scarcity is indeed a result of that identity being subsumed by a West Saxon one, it might support an interpretation of many Seaxe- and Engle-names as having arisen at a relatively late date; they arose, perhaps, at a time when those identities were more clearly geographically defined. A late interpretation may also more easily explain those Engle- and Seaxe-names located in borderlands, whose semantic value required an understanding that there were Anglian or Saxon polities.

Chronology is less of a problem for those names which show Scandinavian influence, and which therefore must date from the late 9th century or later. They are best discussed together, particularly as eight of them are bý-names. Ethnonym + bý is recurrent in Danelaw areas of England: the generic is found with specifics referring to Britons (Bretar), Flemings (Flæmingjar), Frisians (Frísir), Irish/Goidelic speakers (Írar), Scandinavians (Danir and Norðmenn), as well as Englar and Saksar. For this reason, the Saxby-names are very likely to contain the ethnonym, despite there formally being other possibilities. Scholars who have preferred an ethnonym interpretation have argued that the Saxbys, like the Inglebys, refer to English-speaking communities. Gillian Fellows-Jensen offers two possibilities:

the Danes must either have been able to distinguish between the Angles and the Saxons or have used the two national terms synonymously. The latter alternative seems the more likely.¹⁵⁴

Barrie Cox adds that, “by the date of such a name’s formation, the Danes were unlikely to have been able to distinguish between Angles and Saxons.”¹⁵⁵ The traditional argument holds that Saxby- and Ingleby-names referred to English-speaking communities in areas dominated by Scandinavian speakers, and indeed they certainly occur within areas whose place-names are good evidence for heavy Scandinavian linguistic influence. However, given that the element bý compounds elsewhere with ethnonyms referring to peoples of the European mainland, the possibility should be borne in mind that the Saxby-names in fact refer not to (Anglo-)Saxons but to ‘Old’ Saxons, from Saxony. OESeaxe refers sometimes to Continental Saxons; in Scandinavian contexts, ONSaksar refers only to Continental Saxons. If the Saxby-names were formed by Scandinavian speakers, as seems likely,¹⁵⁶ why would Saksar have been the term chosen to denote English speakers? There is no Scandinavian textual evidence to suggest that it was used with this sense, and, had Scandinavian communities in the Danelaw learned the vocabulary for English speakers from their Anglo-Saxon neighbours, we may wonder why these neighbours would have self-identified as Seaxe. If we allow that Flemings and Frisians appear in English place-names,¹⁵⁷ then we must allow that Continental Saxons may too. It is worth noting that one of the Lincolnshire Saxbys is just over a mile from a lost Frísa-bý site,¹⁵⁸ and Saxby in Leicestershire is approximately 8 miles from Frisby on the Wreake, one of three Frísa-býs in that county.¹⁵⁹ If the Saxbys do indeed refer to Continental Saxons, this makes better sense of a situation in which both Englar and Saksar bý-names are found in the east midlands at no great distance apart, in one case fewer than 10 miles apart (Fig. 6.4).

View Full Size

East midlands namesmap produced by Richard Jones

• Download Figure
• Download figure as PowerPoint slide

Considering the names within the broader onomastic context might also cast light on whether these names did indeed refer to minority communities surrounded by one dominant group. The recurrent names in *Engla-bý—which can readily be interpreted as Old Norse formations¹⁶⁰—are, for example, usually identified by scholars as groups of English speakers surrounded by Scandinavians.¹⁶¹ Ethnonym + bý-names tend to occur in clusters, however, which rather suggests that these names were characteristic of areas whose populations were thoroughly mixed and ethnically dynamic,¹⁶² but where Old Norse was the dominant language in the toponymicon at the time that these names arose. Fig. 6.5 shows one such cluster in the North Riding of Yorkshire.

View Full Size

North Riding clustermap produced by Richard Jones

• Download Figure
• Download figure as PowerPoint slide

Less can be said about the remaining two possible Saksar-names, the lost Sash Gate, in Owersby parish, Lincolnshire, and Saxthorpe in Norfolk. The personal name Saxi is a formal possibility in both of these, and is accepted by most scholars as the specific of Saxthorpe. As there are just two other possible ethnonym names in ONþorp but very many personal names,¹⁶³ this is probably the safest assumption. Sash Gate, on the other hand, sits with ten or more other names comprising ethnonym + ONgata, plus several additional names comprising ethnonym + routeway. For Sash Gate, as for the Saxby-names, it has been argued that the Saksar were Anglo-Saxons.¹⁶⁴ However, it is possible again to point to instances in which gata compounds with specifics referring to Continental European peoples: there is a Flæmingja-gata ‘road of the Flemings’ in Beverley, as well as recurrent instances of Norðmanna- and Dana-gata. A possible reference to Continental Saxons must at least be considered. What the relationship between the Saxons (of whatever origin) and the road might be is not entirely clear. There are three surviving instances of ethnonym + gata: Jubbergate in York (Bretar), Flemingate in Beverley, and Deansgate (Danir) in Grimsby. These are also the earliest attested of the ethnonym + gata compounds, all appearing in 12th-century documents. The roads so-named are within substantial settlements: Ekwall suggested that Britons were Jubbergate’s main inhabitants, and Smith that Flemingate was occupied by Flemish traders.¹⁶⁵ Occupation or trading sites are certainly possible motivations for such names, and may lie also behind Ing Lane (if the name does indeed contain Engle), which Ekwall saw as “the result of an early Anglian immigration into London.”¹⁶⁶ Use of the road in question is an alternative explanation, as in Alan James’s assessment of the two Bret(t)a-strǣt names in the north west, which “presumably refer to Cumbrian drovers or traders.”¹⁶⁷ Ekwall explained these names as referring to their perceived builders, but that rested on their having been applied, as he thought, to Roman roads.¹⁶⁸ Such an explanation seems unlikely to apply to Sash Gate, which lay within the Lincolnshire village of Owersby, or indeed to any of the gata names.

Conclusions

In considering the material set out in this essay, we are inevitably indebted to and constrained by Bede’s ethnographical framework. In many cases, we can only judge the motivation of an Engle, Seaxe or Ēote place-name against his parameters. Yet the reality of early Anglo-Saxon identities must have been ethnically and geographically more complex: ethnically complex because of participation in the settlement of England by other Germanic ‘tribes’ not mentioned by Bede in this passage;¹⁶⁹ geographically complex because it is hard to imagine an uncontrolled, early medieval mass migration with such a neat outcome. Barbara Yorke’s cautionary observation, when assessing Bede’s description of the languages of Britain, is just as valid here:

… Bede seems to show awareness that the identity of Angles and Saxons (and presumably the Jutes), although partly the result of heavy migration from certain areas in the Germanic homelands, was also the product of a certain rationalization and regrouping of identity which occurred within Britain itself…¹⁷⁰

With our material there is an additional complication, that of chronology—we can rarely date the coining of these place-names with any great precision. Even so, it is clear that a number of them post-date the Scandinavian settlement of northern and eastern England, another time of migrations to and within Britain. If some of the place-names discussed here arose during the 5th and 6th centuries, or reflect patterns of settlement established at that time, other contexts should also be sought in explaining the full range of names. If we are aware that Bede’s description of the settlement of Angles, Saxons and Jutes may be a rationalization of a very complex reality, we must also avoid simplification in explaining the toponymic afterlives of those ‘tribes’.