Attested: Olca clavis, Olci clavis, Oleclavis (in the 3 manuscripts) at position 188 in the Ravenna Cosmography. Richmond & Crawford picked Olcaclavis as the best consensus spelling.
Where: The Cosmography appears here to be following an ancient road, roughly equivalent to the modern A1, in Northumberland from Alnwick to Berwick-on-Tweed (Bishop, 2014:166). No Roman fort is known in that area, but there are several Iron-Age enclosures, the name Outchester is suggestive, and Bamburgh castle on the coast at NU18333497 may be built over the capital of a post-Roman British kingdom that later became the Anglian Bernicia (a name of uncertain origin).
Name origin: Rivet & Smith focussed on the possible word division and wanted to emend the name to *Orea *classis, representing a Roman naval stores depot analogous to, but different from, Horrea Classis. Retaining initial Olea- would point towards Latin -oleo ‘to grow large’ (only in compounds) or oleo ‘to smell’ (Northumberland kippers anyone?). Neither analysis is convincing, and it seems better to follow Richmond & Crawford in focussing on late Latin olca, which Gregory of Tours defined as fertile ground with these words “campus tellure fecundus - tales enim incolae olcas vocant”. Olca also formed part of an early place name in northern Spain Οτταουιολκα (Ptolemy 2,6,51) and shows up in a Celtiberian inscription. The survival of olca into French ouche ‘kitchen garden’ is well documented, for example here. Most discussions of this word assume that it was originally Celtic, possibly from *polka losing its initial P. However, once a lost initial letter is invoked, it might be H. Greek ὁλκας (holcas) ‘ship which is towed, hence, trading vessel, merchantman’, would suit the coastal location, even conceivably a diolkos-like ship portage. Or maybe a W got lost (as in Woden/Wotan/Odin and the *Votadini), which draws attention to the other spelling osca for olca, which hints at OE wase ‘mud’ or The Wash.
Perhaps the most likely source of olca was PIE *elk-es- ‘wound, ulcer’, which dictionaries currently show as distinct from PIE *el- ‘red, brown’ (the source of elk) but maybe they really belong together. Then, hypothetically, the ‘kitchen garden’ meaning of olca might refer to the fertility of dark earth (caused by waste in the soil).
The second element might derive from PIE *klau- ‘hook, peg’ (which led to Latin words for key, club, and nail), presumably referring to the hook-like shape of Lindisfarne. Latin clavus also meant the tiller of a ship. This seems a more likely source than PIE *kleu- ‘to listen’ (which led to *klewos- ‘fame’ and to Cleopatra) favoured by Richmond & Crawford.
The spelling Oleclavis could naturally translate as ‘bent nail’ if its initial Ole- was similar to the start of Ολικανα.
Notes: All the possible analyses of olca seem to point to the vicinity of Lindisfarne (Holy Island), which has an interesting history and whose name has been much discussed (see Coates, 2000c). Mediaeval armies marched to and from Scotland, like the Romans before them, along the course of the A1 main road and/or the “Devil's Causeway” almost parallel; their course existed in 1675. Much land to the east of the A1 would have been marsh, judging by modern place names. Kentstone Hill at NU04394145 is topped by a recently discovered multivallate hillfort, which shows up particular clearly in Google's aerial photos. Kent- in a name often hints at some kind of water-channel engineering (as archetypally at Cunetio), and it is possible that in the generally wetter Roman Britain that hillfort would have had a harbour near Haggerston Castle at the apex of an estuary represented now by the small rivers called Low. The other side of the fort would also have been protected by water, now represented by Dean Burn. Elconio is a possible parallel. It is probably just an amusing coincidence that the mud flats exposed at low tide around Lindisfarne and in the adjacent Budle Bay are shaped rather like a gigantic Roman key (clavis), and that there was a later diolkos-like Waggonway.
You may copy this text freely, provided you acknowledge its source as www.romaneranames.uk, recognise that it is liable to human error, and try to offer suggestions for improvement. Last edited 13 April 2020 To main Menu