Identifying the Roman names for locations in modern Britain is not so easy as may first appear. The survival of chester, caistor, or caster place names usually indicates Roman sites, but not the actual original names of those sites. Chester, we know, was once Deva, and Silchester is former Calleva. Similarly the place name element strat- or stret- is evidence of the streets, Roman roads that is, which once linked their towns and fortifications. Stratford and Streatham are obvious examples of names of this type, and with Chester le Street you have the best of both worlds. Modern place names are normally of little help. The castle in Papcastle probably does refer to an abandoned Roman fortification but papa is seemingly Old English or Old Norse for the religious hermit or monk who possibly lived there. In any case the name bears no resemblance to the Roman Derventio. Perhaps Viroconium (illustrated) did mutate to Wroxeter, or the nearby hill known as The Wrekin.
It is true that a very few names have survived fifteen hundred years with little mutilation. Examples of these preserved names include London (Londinium), Kent (Cantium), Lincoln (Lindum Colonia), Catterick (Cateractonium) and perhaps Speen, Berkshire (Spinis). If you reflect how relatively common Roman inscriptions are it is surprisingly rare for the name of a site to be identified by epigraphy. Banna, Birdoswald fort on Hadrian’s Wall, is one example where this does seem to have occurred. The inscription Deo Sancto Silvano venatores Banniess, found at this fort, seemingly refers to the religious activities of the ‘hunters of Banna’. Artefacts like the Rudge Cup and the Staffordshire Moorlands Pan may have been produced quite consciously in the Roman period as souvenirs, and they are inscribed with the contemporary names of several Hadrian’s Wall forts. In consequence such names are known with a high degree of certainty and it is quite conceivable that Britain’s most famous Roman monument was originally ‘the Aelian Wall’.
The names of Roman sites frequently have a meaning which may still be intelligible today if the Latin is translated. The attribution of Trimontium in the Scottish lowlands would be difficult if three hills were not still visible from Newstead. The presence of salt springs at Northwich or Droitwich (Salinae) is wholly credible. We may not be certain who Sulis was but it seems sensible that her waters, Aqua Sulis, should be at Bath. Unfortunately in Roman Britain there were plenty of ‘oak tree glades’ so the derivative name, Derventio, may have been applied to Papcastle but it was also the name of Malton in North Yorkshire and the origin of numerous rivers Derwent.
The most valuable resources for historians are four sets of Roman documents: Ptolemy’s Geography, the Antonine Itinerary, the Ravenna Cosmography, and the Notitia Dignitatum. All these works give lists of Roman sites in Britain. The Antonine Itinerary in particular seems to have been composed for the benefit of those moving from point A to point B and wishing to know the names of, and distances to, intermediate stopping points. Consequently ‘iter V’ in this document takes the traveller from London to Lugovalium (Carlisle). The Roman roads between these two places are known and some of the intermediate stops are easy to guess. Camulodunum (‘the fort of Camulos’) is generally taken to be Colchester. Venta Icenorum must be in the territory of the Iceni and Caister St Edmund fits the bill, the element ‘venta’ being considered to mean a marketplace. This route, I appreciate, is a somewhat indirect way of reaching Carlisle; perhaps it took one by the scenic route!
Lindum has already been mentioned and Eboracum has always been recognised as modern York. The first stopping place after York, Isurium Brigantum, must be be in the territory of the Iron Age tribal confederacy of the Brigantes and the important walled Roman town at Aldborough is likely to fit the bill.
Despite decades of careful scholarship many uncertainties remain. The fort nearest to Bradford is at modern Ilkley. This was first identified with an entry in Ptolemy’s Geography, Olenacum, which is elsewhere spelled as Olicana. But an altar stone found at Ilkley mentions Verbeia who is considered to be the protective goddess of the river Wharfe. Roman settlements often have river related names. This arrangement links the Dee with Deva, or Usk and Isca. If so perhaps Verbeia was also the name of the Ilkley fort which would then free-up Olenacum for the next fort west, a farmland site now known as Elslack.
St Patrick, in the story of his life, mentions that he grew up in ‘Bannavem Taburniae’. Scholars have long considered that this name has been mutilated during the copying of the document. One very definite possibility would be that the original was Banna Venta Burniae. The fort of Banna has already been mentioned and archaeological study has demonstrated post-Roman occupation. It might well have developed into a marketplace with Burniae being added (not necessarily even in Roman times) to distinguish it from some other Banna Venta. Unfortunately Banna fort is well inland and not an obvious place for Irish pirates to be capturing slaves. An alternative possibility is that St Patrick originally wrote Glannoventa, which is thought to be the fort and civilian settlement at Ravenglass. This certainly had an exposed position on the west coast of Cumbria within fairly easy reach from Ireland by sea.
If you want to pursue this whole subject further let me thoroughly recommend the classic publication The Place Names of Roman Britain by ALF Rivet & C Smith (Batsford 1979). I can guarantee hours of fun for all the family.