sign in sign up
<<
Home , Damnonii, Decantae, Dobunni, Epidii, Lugi, North Britain

Proc Antiqc So Scot, (1987)7 11 , 85-91

Ptolemy, Tacitus and the tribes of Joh MannnC David *an dBreezeJ t

ABSTRACT Ptolemy's list of places northin Britain, arranged tribe, by include may both native sitesand Roman forts. Unallocated fort names havemay been added Ptolemyby list the ofto what thoughthe was the appropriate tribe, possibly not always correctly. Separation of the two groups of names allows a re-allocation of tribal territories attempted.be to While Caledonesthe gave their landnamethe to north RomanForth,ofthe the by period havetheymay been pushed Greatbackthe upperto the Glen and glens of the mountains by later in-comers.

INTRODUCTION knowledgr Ou namee th locationd f eo s an tribe e th f so living nortTyne-Solwae th f ho y isthmus e arriva e Romatime th th a th f tf eo o l n armies depends primarily upo sourceso ntw , Ptolemy's Geography Tacitus'd an s Agricola. e campaignTh voyaged san f luliuso s Agricola, governof o r Britai 83,8o t 8 85clearlr n7 e 4 o r fro,ar o maie 7 my7 th n sourc informatiof eo lattere th r n.fo Ptolemy used material gathered by Marinus of Tyre, which in turn was probably derived from Agricola's operation Nortn i s h Britain (Rive Smit& t h116)accounte & 1979 4 Th . 11 , f botso h Ptolemd yan Tacitus suffer from grave defects. Bot second-hane har d record than si t neither autho visited ha r d Britain. This might naturally lead to mistakes. Ptolemy in a way was operating at third-hand, correctin improvind gan informatioe gth n already gathere Marinuy db s (Rive Smit& t h 1979, 106). Ptolemy's record also suffers as a result of a basic error whereby Britain north of the Tyne-Solway isthmus is turned eastward through 90°. The major problem in Tacitus's account of Britain is his sparing use of geographical terms, only eight names being mentioned north of the Tyne-Solway isthmus.

ANCIENE TH T SOURCES (illu) s1 Ptolemy, Geography II, 3, 5-7 (the correct locations are given in brackets): There norther e dwelth n o westernle (i n ) side, belo peninsule wth same th f ea o namee th , , among who these mar e places: Lucopibia, Rerigonium. Beneat hSelgovae,e theth e mar among who these mar e places: Carbantorigum, Uxellum, Corda, To the east (ie north) of these are the more northern , in whose territory are these places: Colania, Vindogara, Coria, Alauna, Lindum, Victoria t Catherine'S 8 *2 s Avenue, Bletchley, Milton Keynes T HBM Brando0 2 , n Street, 18 19 20 21 22 23 24 25 24 27 28 29 30 31 32 3

-"~\ ^ T H- V L E 43

62

61

60

5»

ILLUS 1 Ptolemy's map of north Britain. Based on the Map of , reproduced by kind permission of the Ordnance Survey MANN AND BREEZE: PTOLEMY, TACITUS AND THE TRIBES OF NORTH BRITAIN | 87

The Otadini are more to the south (east), and among them are these places: Curia, Alauna, Bremenium Afte e Damnonii,th r toward e easth s t (north t morbu ) e northerly (west), from Epidium Promontory eastwards (north) are the , after whom further east (north) are the Creones; the Carnonacae,e nth the Caereni;e nth furthesd an t east (north lasd Cornavii;e th tan ) from Lemannonian Gulf Vararo t EstuaryCaledones,e th e ar abovd an e the Caledoniane mth Forest, and east (north) of them the , after whom are the , adjoining the Cornavii, and abov Lugi,e e th Smertae. e th Belo Caledonese w th ,e th e ar among who these mar e places: Bannatia, Tameia, Pinnata Castra, Tuesis Below these towards the west (south) are the , in whose territory is this place: Orrea More towards the east (north) are the and this place: Devana. Tacitus, Agricola 10 refers to Caledonia; 11 the inhabitants of Caledonia; 25 the inhabitants of Caledonia, which from the context is situated beyond Bodotria (the Firth of Forth); 38 the Boresti.

COMMENT Few of the names listed by Ptolemy can be identified. Trimontium can be equated with the fort at Newstead, at the foot of the (three) Eildon Hills. Bremenium is the Roman fort at High Rochester (RIB 1262 and 1270; 464.3). Victoria and Orrea are presumably Roman places, and possibly Pinnata Castra also. Rivet and Smith (1979, 116) 'assume tha placee tth s name nortn di h Britai mose nar t likel"Romane b o yt "tha- place, tis s which beend wereha r , o ,occupie Romae th y db n army. Ther evidenco n s ei e that Ptolemy name purelsa y "native" settlement anywher Britain'n ei . lase tTh statemen reversee th s i o trues ,s i tthat bu , t therevidenco n s ei e that Ptolemt no d ydi nam purelea y 'native' settlemen Britainn i t certainle H . y named non-Roman settlement Irelann si d and althoug placeso hn , either Roma nativer no namee ,ar d Moranorte th f ho y Firth (Rive Smitt& h 1979,116 simple b ) thiy ysma because Roma npenetratet forceno d sha d ther werd ean e consequently unable to name any native settlements within the northern tribes (this may indicate,pace Henderson 1984, that Mons Graupiu t foughno s t swa Morae nort th f ho y Firth certainls i t i : y interestino gt compar levee e th f informatio o l interioe th n nf northero o r n wit interioe h th tha n f o t o r , which is rather better documented). The possibility that some names are native places and others Roman forts opens up the question of the source of the primary information. Ptolemy may have had two lists: one of tribes with native places within their territories, the other of Roman forts. The information for the two groups will have been gleane different da t times tribae nativ;d th lan e place-names durin advance gth e northwardd san the information on forts during the subsequent consolidation (Tacitus, Agricola 22 emphasizes the distinction between thes activitieo etw remarkiny sb gthire thath dn t i seaso campaigninf no g Agricola eve timd nbuilha o e t d forts). Nevertheles activitiee sth Agricolf so presumable aar y responsiblr efo all of this knowledge - of tribal and native place-names, the geographical names and the forts at least as far north as the Tay: if Victoria refers to the Roman victory at Mons Graupius this place-name may possibly date slightl true samyb Pinnat f e e y o later th ed ma ,an a campCastra t no thi.forf ad a i s i an t Rive Smitd an t h (1979, 121) suggest 'that Ptolemy located the tribes merely by reference to their administrative centres (whether in the military or in the succeeding civilian phase) and that the tribal names, when written on the original map, attracted to themselves all the places in their areas when the co-ordinates were read off', SOCIET ANTIQUARIEF YO SCOTLANDF SO , 1987 thus creatin opportunitn ga errorr yfo Northern I . n Britain this hazy knowledg tribaf eo l boundaries may have been compounde existence th listo y dnamesf b stw o f possibles o i t I . e that Ptolemy simply tagged the unallocated fort names on to the list of what he thought was the appropriate tribe. Thus Bremeniu assignes m Votadine wa th o d t Trimontiu d ian Selgovaee th mo t , etc.t Howeverno y ma e ,h always have been correct. In southern England Ptolemy certainly seems to have been mistaken in placing Aquae Calidae (Bath) in the territory of the rather than the (Rivet & Smith 1979, 121 & 256). j t THE LOCATION OF THE TRIBES (illus 2) Separation of the list of 'tribal' names into two categories and the release of the fort names from tribal territories allows a reconsideration of the location of the tribes. Uxellum in the territory of the mout oughe closNite e th li th ho o ef t t h o (18°30' , 59°20 18°20'd an ' , 59°30' respectively). Carbantorigum and Corda lie to the east (ie north on correcting Ptolemy's 90° turn); they thus may

CO ^ DECANTAE/^

fiMOVANTAE ^~=s

ILLUS 2 The proposed location of Ptolemy's tribes in north Britain MANN AND BREEZE: PTOLEMY, TACITUS AND THE TRIBES OF NORTH BRITAIN 89

Nithsdalbotn i e hli Annandaler eo woult I . d make sens termn ei humaf so n geograph Trimontiuyf i m in did not lie within the territory of the Selgovae as this valley forms a distinct entity looking eastwards and not to the south or south-west. The uplands forming the watershed between the Annan and Tweed basins were (and are) sparsely populated while many cross-dykes control the passes at the headwaters of the (Barber forthcoming). Thus the Selgovae could have occupied either Tweeddale or and , but are unlikely to have occupied both areas e relationshiTh . p between Uxellu moute e Nitth th hf d ho woulm an d appea pulo t e r th l Selgovae into aren a , a well populate Iroe th n n di Age . bees Iha t n suggested (Breez Dobsoe& n 1987, 45-6) thaBrigantee th t s occupie lowee dth r Annan valley. This is argued on the basis of the existence of outpost forts to the north of Hadrian's Wall, planned t wouli , d seem, frodedicatio a firste d mth goddesan e , th o nt s Briganti t Birrenaa s (RIB 2091) purpose .Th thesf eo e definitelt fortno s si y known onln surmisede ca y b t :i . Troops could have been stationed there to protect those members of the cut off from the rest of the province by the construction of the Wall. Alternatively, the purpose of these units may have been simply to provide advance warning of any impending attacks on the Wall. However, it might be thought that this function would normally have been carried out by scouts. The inscription, although dedicated to Brigantia, may not have been erected in Brigantian territory. The dedicator was an architect, probably from York, and thus may have become a devotee of Brigantia while stationed at lattee th r place paralle.A dedicatioa r lfo tribaa o nt l goddes anothen si r tribal territor foune b n dyca in Germany. In the territory of Cologne are a series of dedications (mostly by soldiers) to the goddess Sunuxsalis nothind witha o d o ho g t wh tha , t area, bein goddese gth triba f so e living further west around Aachen. Thus neithe e outposth r e dedicatiotth fortr no sprooe nar f thae Briganteth t s stretched nort Hadrian'f ho s Wall. Nevertheless, eve thef ni therd ydi e woulprobleo n e db m about populoue on s tribal territory touching another thud possibilite ,an s th y tha northere tth n fringee th f so Brigantes lay north of the Solway cannot be ruled out. Finally, divorcing Tweeddale from the Selgovae would remov line eth k betwee tribae nth modere l namth d enan nam Selkirkf eo Sele Th -. element in both names would be reduced to a coincidence and no more. SelgovaeNovantae e wese th Th theif th f o o to t e re place-namese,on li , Rerigonius, being tied to the Rerigonian Gulf, presumably to be identified with Loch Ryan. The peninsula of the Novantae ought to be the Rhinns of , and the cape of the Novantae the Mull of Galloway. Thus the Novantae would occupy the area now known as Galloway, but nothing further east. Votadine Th more iar e difficult. Cori southere ath appearn i e nli paro st theif to r territord yan Alauna Forthnorte th repea a Alaune f ht indeeo f th ,i f no to s i at d i liste d unde Damnonie rth i (Rivet & Smith 1979,245). All that can be said is that they appear to lie along the east coast. If they stretch from north Northumberlan e Fortth r beyondo ho t d thud an ,s includ e Tweeth e d basin, then Trimontium ought to lie on their territory. In later centuries part of their territory appears to have lain Forthnorte th f namhe o :th thaf i e s Clackmannati n retains referenc Manae th o et u (Rivet & Smith 1979, 509; Steer 1958,107) late thae th e b n .firsi t y Thut ma centurt si tribe stretcyd th edi h beyond the Forth. The Damnonii have an abundance of place-names on their territory: five in all, minus Victoria. These places stretch from Vindogara, which ough beside li gulo e t t Vindogar f eth fo a (21°20'd an ° ,60 21°20', 60°30' respectively) Alauno t , Lindud aan m Forth-Clyde nortth f ho e isthmus t seemI . s unlikely thatribe ton e occupied this large area, middl e sinconlt Campsith e no yn eth e i eli e Felld san Kirkpatrick Hills, but also the boggy valley of the river Kelvin and the mosses of Stirlingshire, all areas largely devoi populatiof do Iro(Breeze e th nAg n i e 1985, 225 possibilite , Th illu . s2) a f yo mistake here must be considered: perhaps Ptolemy (or a later scribe) omitted a tribal name and Alauna and Lindum lay within the territory of a more northerly tribe (Breeze 1982, 32). 90 SOCIETY OF ANTIQUARIES OF SCOTLAND, 1987 I NortForth!e th f ho , alon ease gth t Veniconecoaste th Taexalie e li ,th d latee san th , r relateo dt the promontory of the Taexali, either Kinnairds Head or Rattray Head. The place-names of the Vacomagi apparently straddle the Highlands, stretching from Angus to . It is highly unlikely that they covered this vast area since it is cut in two by the unpopulated mountain rangMounthe th f eVacomago e othere .Th side inclusioth e r eon o .Th n i o ough e f no li o t Tuesis in their list, which Ought to lie on the river Tuesis, generally identified with the river Spey, must pull their territory sout e Mountnorte th th hf o ht o shorMora e n ho th f eo y estuary (26°45', 59°10d an ' 27°respectively)° 59 , locatioe placee Th . th f no s Bannati Tameiad aan , however, suggests thae th t tribe's territory stretched a considerable distance up the Spey Valley. Pinnata Castra is near Tuesis, and musMorae th tn yalso e Firtoli h (Mann 1968, 307). The Caledones are Specifically located by Ptolemy. They lie between the Lemannonian Gulf on the west coast and the Varar estuary on the east. The latter is to be identified with the Moray and Beauly Firths formee Th . r ought, fro positios mit Ptolemy'n i s list, betwee Clyde nth Kintyred ean , to be Loch Fyne or possibly Loch Long. However, it has been argued that Ptolemy may have been confused by two sources into stretching the Caledones to the Lemannonian Gulf rather than the mouth of the river Longus, which should be Loch Linnhe (Rivet & Smith 1979,141). The Great Glen runs from Loch Linnh Beaule th o et y Firth thid whers si an , Caledonee eth generalle sar y locatey db modern scholars. Hind (1983, 375) commenting that 'the land coul t suppordno populatioa t n that would be able to lead so, powerful a confederacy as that which opposed Agricola' therefore located Caledonee th s along the; souther easterd nan n Highlanedgee th f so d massif, from Loch Lomono dt Morae th y Firth. Supporj thir fo ts locatio Caledonee th f no claimes si modere th n di n place namef so Perthshire: , Rohallion and Schiehallion (Jackson 1954, 14-16; Hind 1983, 376). Further, Hind sees the Vacomagi', Taexali, Venicones and possibly the Damnonii as 'part of the Caledonian groupin tribes'f go Caledoniae pare th i , f o t n confederacy. Ther , however!eis distinctioa , n Ptolembetweey wa e ynth Caledonee referth e o t sth d san manne whicn ri h Tacitus^ othed an , r early writers, mention Caledonia Ptolemo .T Caledonee yth e sar a people, but to Tacitus!Caledonia is a region, apparently indeed the whole of Britain north of the Forth-Clyde isthmus. Tacitus refer Caledoniamo st incolentes (Agricola r Caledoniamo ) 25 habitan- tium (Agricola 11). This careful avoidance of Caledones needs to be explained. Furthermore, when this trib mentiones ei Ptolemyy db carefulls i t ,i y distinguished fro Decantae mth othed ean r northern hane fro d Vacomagie on people dman th e th n so , Venicone Taexal d othere san th n .o i Caledones therefore cannot be explained as a group name for these peoples. They must be a distinct people. When Jackson (1955,135) asserts that Caledonia 'cannot be proved to be Celtic and may therefore very possibly be pre-Celtic' it seems reasonable to suggest that the pre-Celtic Caledones, who had give countre nth names y it bee d ha ,n pushed bac Celtiy kb c invader s- Venicones , Vacomagd an i Taexal thao s ti- whil 'popularn ei ' usage Caledonia survived Caledonee tribe th th f , eo s sank into comparative insignificanc Manf e(c n 1974, 36). They welthuy slma have been restricte more th eo dt barren lands of the Great Glen, and possibly the upper glens. One final point may be made. It is by no means certain that Ptolemy listed all the tribes of northern Britain. Tacitiis's Boresti certainly seem to have escaped his notice, and there may have been more alse H .o demonstrably made mistakes elsewhere. Ptolemy's Geography thus shoult dno be accepted as either correct or complete.

REFERENCES j Barber, J' forthcomin | Linearg earthworks Scotland.n i (=Edinburgh Univ Dept Archaeol Occasp Pa Ser, 15.) j MAN BREEZED NAN ! PTOLEMY , TRIBEE TACITUTH NORTF D SO SAN H 1 BRITAI9 N

Breeze J 198 D ,Northerne 2Th Frontiers of Roman Britain. London. Breeze J 198D , 5 'Roman force nativd san e populations', Proc Antiqc So Scot, (1985)5 11 , 223-8. Breeze, D J & Dobson, B 1987 Hadrian's Wall. Harmondsworth. Henderson R 198,AA 4 'From8 19833o t : onthetrailof MonsGraupius' Deesidee Th , Field, 18(1984), 23-9. HindF 198G ,J 3 'Caledoni occupatios it d aan n unde Flavians'e rth , Proc Antiqc So Scot, (1983)3 11 , 373-78. Jackson, K H 1954 'Two early Scottish names', Scot Hist Rev, 33 (1954), 14-19. Jackson, K H 1955 The Pictish language', in Wainwright, F T (ed), The problem of the , Edin- burgh, 129-66. Mann C 196J , 8 Revie Richmondd Ogilvief wo an M R (eds), A I , vitae d , Agricolae, Oxford 1965n i , Archaeol Aeliana, 4 ser, 46 (1968), 306-8. Mann, J C 1974 'The Northern Frontier after AD 369', Archaeol J, 3 (1974), 34-42. B CollingwoodRI Wright& G P 196R ,R , Roman e 5Th Inscriptions f Britain,o : InscriptionsI n o stone. Oxford. Rivet, A L F & Smith, C 1979 The place-names of Roman Britain. London. Steer, K A 1958 'Roman and native in North Britain: The Severan reorganisation', in Richmond, I A (ed), Roman and Native in North Britain, Edinburgh, 91-111. This paper publishedis granta of from aid with the Historic Buildings Monumentsand (SDD)