Recent Developments in the Study of Place-Names and the Anglo-Saxon Settlement
RECENT DEVELOPMENTS IN THE STUDY OF PLACE-NAMES AND THE ANGLO-SAXON SETTLEMENT By K. J. EDWARDS, M.A. THE apphcation of place-name evidence to studies of the nature and evolution of Anglo-Saxon settlement has long been accepted as a valuable method of historical investigation.1 In the last few years exciting detective work has produced many new developments, and this short paper is intended to present the substance of this work to a wider audience. The first indication of a relative chronology for place-names and hence settlement evolution came from Kemble in the last century.2 He suggested that place-names terminating in the Old Enghsh (OE) stem -ingas (the plural form of the -ing suflix as found in the settlement names of Coohng and Malling for instance), denoted the areas first settled by immigrating Angles, Saxons, and. Jutes. This observation was ratified and the concept developed by EkwaU3 and Smith* to the point where various -ing place-names were located in a chronological progression based on their Old Enghsh phUologic form, viz. (i) -ingas, (ii) -inga- (medial -ing element as in Farningham, TriUinghurst), (Hi) -ing2 (a singular -ing stem found in the names Kemsing and Welling). 6 Those place-names in OE -ingas and -inga- have been taken to signify the earhest pagan phases of settlement. The archseological evidence, however, is at variance with this proposition. Smith,6 and Myres before him7 had noted the lack of coincidence between these supposedly archaic place-names and pagan burial-sites. The problem 1 Examples include S. W. Wooldridge, 'The Anglo-Saxon Settlement', in H. C. Darby (Ed.), An historical Geography of England before A.D. 1800, Cambridge, 1936, 88-132, and P. H. Beaney, 'Place-names and early Settlement in Kent', Arch. Cant., lxxvi (1962), 58-74. s J. M. Kemble, The Saxons in England, I, London, 1849, 35-71. 8 E. Ekwall, English Place-names in -ing, 2nd Edn., 1962. * A. H. Smith, English Place-name Elements, Pt. 1, vol. 25, English Placename Society (1966a); A. H. Smith, 'Place-names and the Anglo-Saxon Settlement', Proc. Brit. Acad., xiii (1956b), 67-88. * Plaoe-name study is a hazardous field without expert guidance, and attempts to guess at the derivations of various settlement names should be resisted. Beference to county monographs, and where available those of the English Place-name Sooiety in particular, is strongly recommended. 0 A. H. Smith, op. cit. (1966b), 84. 7 J . N. L. Myres, 'Britain in the Dark Ages', Antiquity, ix (1936), 455-64. 81 K. J. EDWABDS was tackled by Dodgson8 who considers that the spatial disparity between place-names in -ingas, -inga-, and the pagan-burials in southeast England has a temporal cause—that the pagan-burials are located in districts of Anglo-Saxon immigration whereas the -ingas, -inga-, centres arose from a subsequent colonization period. The techniques of nearest-neighbour analysis and ordination are brought to bear on this problem by Kirk.9 The results of her examination of the data for Kent generaUy support Dodgson, and it is suggested that the earhest stratum of Anglo-Saxon settlement might be indicated by the -ing2 place-names, though these were not taken into consideration. Dodgson10 has in fact stratified the -ing place-names on a phUological basis into a time-sequence which is at variance with the conventional wisdom, viz. (i) -ing2, (ii) -ingas, (Hi) -inga-. A fourth category, -ingi, together with assorted hybrid forms was current throughout the time period involved and is therefore unsuitable as a chronological indicator. The association between place-names in -ing2 and pagan burials remains to be investigated, but strong competition as the earhest Old EngUsh place-name type has come from outwith the -ing group. Gelling11 examined the twenty-nine probable instances of place-names derived from the Old EngHsh compound wlcham in England, and discovered that twenty-five of these are situated on, or close to, a Roman road. Furthermore, twenty of the settlements are within five mUes of an Anglo-Saxon pagan cemetery, and Gelling tentatively suggests that the place-names in wicham might indicate settlements of the Germanic laeti, employed towards the end of the Roman period to provide protection for villages and viUas from marauding native Britons. Doubts about the archseological evidence for continuance of settlement, however, led Gelling to exercise caution over such speculation. The terminal element in uMcham is CE -ham ('homestead', 'village'), long accepted as an archaic place-name form supposedly succeeding those in -ing. A recent paper by Cox12 persuasively argues that placenames in -hem represent the earhest pagan phases of Anglo-Saxon settlement. Cox points out the close relationship between place-names in -ham (and frequently those in the compound -ing form, -ingahdm) 8 J. McN. Dodgson, 'The Significance of the Distribution of the English Place-name in -ingas, -inga, in south-east England', Med. Arch., x (1966), 1-29. 9 S. Kirk, 'A Distribution Pattern; ingas in Kent', J.E.P.N.S., iv (1972), 37-59. 10 J. MoN. Dodgson, 'The -ing in Place-names like Birmingham and Altrinoham', Beitrage zur Namenforschung, N.E.2 (1967), 221-45; 'Various Forms of Old English -ing in English Place-names', ibid., 326-96; 'Various English Placename Formations containing Old English -ing, ibid., N.F.3 (1968), 141-89; also see S. Kirk, op. cit., 38. u M. Gelling, 'English Plaoe-names derived from the Compound wtaham', Med. Arch., xi (1967), 87-104. 12 B. H. Cox, 'The Significance of the Distribution, of English Plaoe-names in -ham in the Midlands and East Anglia', J.E.P.N.S., v (1973), 15-73. 82 BECENT DEVELOPMENTS IN THE STUDY OF PLACE-NAMES to the ancient trackways, Roman roads, Romano-British settlements and Anglo-Saxon pagan burial-sites of the Midlands and East Anglia. Cox, like Gelling, suggests that the place-names in -Mm mark the settlements of Roman-employed Germanic mercenaries. The acceptance of this theory would, of course, impose a new stratification on early place-names. Cox elaborates on this theme by putting forward a chronologic progression which may be simply stated as: (i) topographic term-f-Mm. (as in Higham), (H) personal name-f-Mm (as in Meopham, 'Meapa's homestead'), (iii) personal name+-ingaham, (iv) -ingas. For the area under study, it is intimated that the -ham names belong to the pagan period; those in -ingaham denote a transition from the pagan to Christian phases (sixth to seventh centuries A.D.); and that placenames in -ingas, -inga, relate largely to the post-pagan era. In this reversal of the orthodox place-name progression, Cox observes that the -ham suffix is only one immigration or pagan-phase place-name which he has identified and others may exist. Detailed research jn north-west Kent13 tends to support Cox as well as throwing some Hght, perhaps, on additional early place-name types. This work shows that six of the nineteen pagan burials in the area have four place-names in -ham, -ingaham (Betsham, Wioham, Wrotham and Farningham), as their nearest Anglo-Saxon settlements. Centres bearing place-names signifying a water relationship appear as the closest settlements to another five burials—Sharbroke ('dirty brook'), South Darenth ('south oak river'), Northfleet ('north river'), Otford ('Otta's ford') and Littlebrook. A tenuous aqueous relationship may also be exhibited in the place-names of two riverside settlements, Horton Kirby ('settlement on muddy land') and Strood ('marshland'), which He in close proximity to another three pagan burial-sites. It would be very interesting if future research reveals a widespread degree of association between pagan burials and certain water-based place-name forms. It would be expected, of course, that the initial pagan settlers would locate their homes as closely as possible to a readily available water supply. The study of the eariiest settlements and place-names of the Anglo-Saxons is in a healthy state of flux and advancing steadfly. The amount of research whioh has been carried out into the place-names of the post-pagan phase is somewhat less. No reliable place-name succession for this period of settlement expansion has been demonstrated although Hsts appear in print14 and use is made of supposed 18 K. J. Edwards, Aspects of the Geography of Anglo-Saxon Settlement in North-west Kent, unpublished M.A. thesis, University of St. Andrews (1972), and additional unpublished material. 11 J. A. Everson and B. P. Fitzgerald, Settlement Patterns, London, 1969. 83 K. J. EDWABDS progressions.15 The suffix 'ton' ((E-tun) for instance, does not necessarily mark the immediate post-immigration phase of settlement and was stiU being appended to new settlements in the thirteenth century.16 In addition many place-names are known to have disappeared. It must also be stressed, however, that the scope of place-name based settlement studies is capable of further development than has hitherto been the case. It is not necessary, and possibly misleading, to examine only those place-names whioh contain Old Enghsh elements in common, such as -ing, -tun, -leak, -feld and -denn. It is only when all settlement names are considered, whether they incorporate repetitive or nonrecurrent place-name elements, that it is possible to perceive the pattern and full extent of the Anglo-Saxon communities. It may not be possible to assemble the total set of place-names into a sequence as yet, but the areas and sites which received them may be investigated. In north-west Kent,17 a consideration of settlements which feature repetitive place-name forms alone would reveal a pattern consisting of about 50 centres. This would represent one-third of the total which would be detected by considering every Old Enghsh settlement placename. This in turn has repercussions on inferences regarding the nature of settlement. In the case of north-west Kent, the presence of almost 150 settlements in as many square mUes signifies that one is deahng with dense settlement of a hamlet type as opposed to the nucleated viUage form of the Enghsh Midlands—a finding which supports earHer documentary investigations into Jutish society.18 The method detailed above also aUows an investigation into the locational behaviour of the Anglo-Saxons. In the same part of Kent, it was found that the largest group of settlements (37 per cent) were located on river-side sites, a category computed to occupy only 17 per cent, of the land area. Furthermore, 60 per cent, of settlements were located upon fertfle loam soUs which occupy only 29 per cent, of the soils area, compared to those centres on chalk sofls which provide figures of 16 and 45 per cent, respectively. It is possible that this evidence for locational preference wfll be valuably augmented by the introduction of a social dimension in explanatory models of settlement evolution. The place-name based model of ElHson and Harriss19 which embodies the concept of uniform territorial expansion goes some way towards achieving this aim. The use of place-name evidence to investigate the nature and 15 A. Ellison and J. Harriss, 'Settlement and Land Use in the Prehistory and early History of southern England: a Study based on locational Models', in D. L. Clarke (Ed.), Models in Archceology, London, 1972, 911-62. 18 A. H. Smith, op. cit. (1966b), 81. " K. J. Edwards, op. cit., in note 13. 18 J. E. A. Jolliffe, Pro-feudal England: The Jutes, London, 1933. 10 A. Ellison and J. Harriss, op. cit,, in note 16. 84 EECENT DEVELOPMENTS IN THE STUDY OF PLACE-NAMES diffusion of Anglo-Saxon settlement raises many problems. It is undeniable that many difficulties relate to the phflological side of such study, and it is probably outwith the competence of the average settlement researcher to make a direct linguistic contribution to tbis field. It is quite possible, however, that locational work may caU attention to settlement names which demand greater scrutiny from the place-name scholar. The fact that questions continue to be posed by the inter-disciplinary approach which early settlement study entails might justifiably be considered to be a measure of its success. The propagation of new ideas must needs foUow, and this is clearly being provided by place-name based settlement research. 85