[NB. The reader will probably find it easier to understand this page if he or she has already acquainted himself or herself with the contents of Chapters 1 and 2. Chapter 1 concerns the basic building-blocks used in Romano-British place-names and Chapter 2 explains how those building blocks were assembled by the Celts to form compound names. In addition, the reader should not worry about any identification made below - all identifications are explained on other pages of this website, more particularly in Chapters 10-16, which concern the Romano-British place-names in different regions of Britain.]
1 The river-names of the Esk, Exe, Axe and Usk family are generally considered to be derived from a river-name Isca, this being considered to be derived from a hypothetical root isca assumed to mean ‘river’, though there has been much learned discussion of other possible roots, including esca and eisca, the latter assumed to mean ‘fishy water’ and being derived from yet another hypothetical root peisko, assumed to mean ‘fish’. The reader is referred to the first Isca entry in Rivet and Smith’s Alphabetical List (Rivet and Smith 1979) for a detailed discussion of the many hypothetical roots brought into the discussion by various scholars. Rivet and Smith appeared reluctant to recommend any particular one of the many derivations offered, and rightly so, since it does appear that many, perhaps even all, of the Esk, Exe, Axe and Usk family of river-names are in fact derived from an isca-type place-name, isca being an inversion-type land-name element meaning ‘hill steep’.
1.1 This is most easily demonstrated by considering the Cumbrian river Esk. The Ravenna list goes from Camulodono (111) at Skipton to Cantiventi (115) at Watercrook (the modern river-name Kent being derived from Cantiventi) and then on through Ambleside to Gabrocentio (117) at Hard Knott pass. The next-following Ravenna name is Alavna (118), which must have been the fort at Ravenglass. But Alavna is a river-name, so quite clearly the Romans simply adopted this river-name and applied it to an early fort which they built on the banks of the river at Ravenglass. In other words, the river now called the Esk was in fact called the Alavna when the Romans arrived in that part of England. The fort at Ravenglass was presumably abandoned at some point and later rebuilt or re-commissioned under the land-name Isca (or another name from the family of isca-type names), which the Romans themselves then transferred to the river, now the Esk. There is in fact a steep hill, the meaning of isca, behind the fort at Ravenglass, the hill leading up to Muncaster Castle, but the Ravenglass fort is described as standing at the top of a cliff which is being eroded by the sea. If this description is correct then that cliff is more likely to be the natural feature which gave rise to the name Isca. And of course it is not in the least surprising that there should be a place with the land-name Isca on the banks of a river called the Alavna, since the latter is a compound river-name (discussed in Chapter 5) of which the second element was originally Abona, and b is the river-letter corresponding to the hill-letter s in Isca. The above name-changing process is identical to what happened in the case of the fort called Rumabo (190), discussed in Chapter 16: Roman place-names in Scotland. Rumabo is a river-name and it was adopted by the Romans and applied by them to a Flavian fort which they built on the banks of the river Almond in Midlothian, either at Cramond or at some point upriver. The fort was later abandoned, along with all the other Roman forts in Scotland, and later the Romans built Antonine Cramond. This fort was given a land-name of the form Lamond or Almond and this name was subsequently transferred by the Romans to the river, now the Almond. Note that the river-letters corresponding to the hill-letters m and n of Lamond are r and m (for major rivers), both of which are present in the river-name Rumabo.
1.2 Another example of an isca-type name being transferred to a river is afforded by Ptolemy’s Caelis river. As explained in Chapter 19: the rivers of Roman Britain, this name will originally have been Iscaelis and can be identified as the fortlet at Inverquharity. Iscaelis is a land-name (referring to the steep river embankment on the north side of the fortlet) which was transferred by the Romans to the river now called the South Esk. It may be that at some date after the departure of the Romans the el element of the name was mistaken for the river-letter l, the proper name of the river then being taken to be Isca, this then surviving in the modern name Esk.
1.3 Another example is the rather odd Ravenna name Evidensca (189) (taking Rivet and Smith’s reading of the Ravenna text rather than Richmond and Crawford’s). This was the fort at Inveresk and the name was earlier apparently Levidanisca (actually Lecidanisca: the c → v change is discussed later). This is an unusual name in including the dan element (a dunum-type element) in the middle of the name, but assume for the moment that the spelling is correct. The name, or just the tail-end of it, was then transferred by the Romans to the river, now the Esk. It is, however, possible, as explained in Chapter 16: Roman place-names in Scotland, that the name of the fort was actually Levidanum and the Romans themselves added Isca to distinguish Levidanum more clearly from the Leviodanum (220) at Doune. This might be explained by assuming that a vexillation of the Legio II Augusta built or occupied the fort at Inveresk and they applied the name of their old base at Exeter to their new base at Inveresk, just as they appear to have applied it earlier to their new fortress at Caerleon.
1.4 Another example is the Devon Axe, this name being derived from Ravenna’s river-name Traxula. Traxula is a river-name of the kind in which a river-element, in this case the compound of the river-letters t and r, is used as a prefix to a land-name, in this case Axula. In other words there was a place called Axula on the banks of the river, the place-name was incorporated into the river-name Traxula and the latter in due course yielded the modern river-name Axe.
1.5 Another example is the Somerset Axe, which appears to be derived from Ptolemy’s land-name Iscalis (in the territory of the Belgae). This is one of a number of names which appear to have been reversed at some stage, in this case yielding a name like Lacsis. The fort-name was transferred to the river by the Romans, and later the l was presumably misinterpreted as the river-letter l, the proper name of the river being seen to be Acs, which turned into Axe.
1.6 A further example is the Devon Exe, which took its name from Isca at Exeter, the name again referring to the steep slope down to the river. At some stage this name, too, was reversed and a grammatical um added to the end, this yielding Acsium, which appears in Ravenna’s river-list as Axium, though in this case the initial a has changed to e in the modern name of the river.
1.7 The Usk in South Wales presumably takes its name from Isca at Caerleon, and the Yscir river, a tributary of the Usk, appears to take its name from the fort called Ypo(s)cessa at Y Gaer, just west of Brecon.
1.8 No clear evidence is available for the Esk which flows into the Solway, though it appears that the fort at Broomholm was called Eburocaslum (184), and it may be that caslum is simply a latinised version of Celtic uxelum/ucselum, this referring to the steep slope at the top of which the fort was built. This element may have been reversed at some stage to give lescum and the esc of this form may have yielded the modern river name Esk.
1.9 No evidence is available for the river Esk in North Yorkshire. However, there was a fort at Lease Rigg, very close to the point where Margary road 81b crosses the river (Margary 1967). The fort stands on a ridge bounded by steep slopes down to the river Esk to the north and to the river Murk Esk to the east, so a name of the isca-type - a name such as Isca, Iscalis, Loxa, Uxelis or Uxellum - would be entirely appropriate for that fort. The Romans will then have transferred the name to the river, now known as the Esk.
2 As noted above, the reader should not worry at this stage about any identification made in this chapter – all of these identifications will be discussed in more detail later in the appropriate chapters. The purpose of the present chapter is merely to demonstrate that the river-names of the Esk, Exe, Axe and Usk family are derived from a land-name of the isca-type and not from any hypothetical British root isca or esca, assumed to mean ‘river’, and still less from a hypothetical root peisko assumed to mean ‘fish’!
[NB. For further discussion of the Uxella, Iscalis, Loxa family of names see Updates: 22 August 2015]
[This page was last modified on 14 May 2021]