Attested:  (1) Ptolemy 2,3,4 Τρισαντωνος river mouth
(2)  Possibly Tacitus Annals 12,31 castris antonam emended to cis trisantonam

Where:  (1) The river formerly called Tarrant, but now Arun, in Sussex, which reaches the sea by Littlehampton, at TQ027011, but Ptolemy's river mouth might have been nearer modern Arundel, probably the Cosmography's Leucomagnus, at about TQ012068.
(2)  The river Trent, variously spelled Treenta, Treanta, Treontan etc by Bede in AD 731

Name Origin:  Greek τρισ ‘threefold, triple’ plus αντιος ‘opposite’ plus the common river-name ending -ona yields a meaning much like ‘three tributaries’.  Much the same could have arisen in Latin and other European languages.  See the discussion under Aventio.  Ekwall (1928:417-418) explained how an original *trisanton developed, by losing its S and then metathesizing around the R, into early English forms.

Notes:  The Sussex Arun is formed by convergence of three main rivers at its former limit of navigation (and Roman road junction) near Pulborough: the Rother, the Chilt, and the upper Arun; see map here.  Exactly how it reached the sea in Roman times is uncertain, but the flood-risk map for that area shows two more drainage routes to the sea through the Bognor-Littlehampton conurbation, and two side-streams converging on a point that might have been the head of an ancient estuary.  Three possible ways to explain why Trisantona was a triple!
  Much the same holds for the Trent.  At the Roman-era head of the river Humber three rivers converged: the Trent, the Ouse, and the Don (before it was diverted by the Romans into joining the Aire and thence into the upstream Ouse, as described by Jones, 1995).  Each of those individual rivers results from a fractal pattern of inflowing tributaries, but the Trent has a particularly definite threefold merging of the major rivers Derwent, Soar, and upper Trent roughly where the modern M1 crosses the Trent. 
  Other rivers that show the 3-tributary pattern include the Trannon in Wales, also beside a Roman road, and the Trisanna in Austria.
  This analysis rejects the Celticist-inspired guess by Ekwall (1928:415-8), picked up by Rivet & Smith pp 476-8, that the name should be divided Tri-santona, with a first part meaning ‘very’ or related to ‘trans’ and a second part from PIE *sent- ‘to go’.  Coles (1994) thought that the name Trisantona arose from these rivers' ancient importance as transport routes.  Breeze (2021) argued that the second part resembles Celtic *suanto ancestral to Welsh chwant ‘desire’.
  That emendation to Tacitus Annals is highly debatable.  Pearce (2000) explained that all extant manuscripts go back to one medieval master, which seems not to be visible online to check.  Oxford Classical Texts edition of the Annals prints detrahere arma suspectis cunctaque castris Avonam [inter] et Sabrinam fluvios cohibere parat, translated as ‘prepared to disarm the suspect and to overawe the whole district on this side of the Trent and Severn’.   That appears to result from several emendations of what was originally castris Antonam et Sabrinam fluvios.  There seems to have been no discussion of geographical issues, such as a link (etymologically or by later confusion) to torrent, derived from Latin torrens, and possibly referring to tricky tidal currents at Humberhead.

You may copy this text freely, provided you acknowledge its source as, recognise that it is liable to human error, and try to offer suggestions for improvement.
Last edited 15 February 2023     to main Menu