28 February 2010

Celtic *dūno- 'fort'

Celtic *dūno- 'fort' is found as a toponym suffix in areas which were formerly Celtic-speaking as France and Great Britain, where it takes the form -ton. It was also borrowed into Germanic (English town, Dutch duin 'garden', German Zaun 'enclosure').

The origin of this word has been disputed, and some Indo-Europeanists have suggested a possible relationship with Latin fūnus 'burial' (from the unattested meaning of 'burial mound'), thus reconstructing a PIE root *dhuHno-1. But IMHO this etymology fits better to Germanic *dūnō 'hill, dune' (English down, Dutch duine, borrowed into English dune). I regard this word as a Paleo-European root *dhaunV ~ *dhūnV 'hill, mound' also reflected in Etruscan θaur(a) 'tomb', with n > r.

A connection with PNC *dompe 'edge, bank', which is the source of Greek táphos (native)  and túmbos (a Pelasgian loanword) 'tomb', is possible, although phonetical developments are not clear. 

From my own experience2, I know the meaning 'enclosure' is related to 'thick, dense, packed', so I link the Celtic word to Semitic *duhn- 'fat'.
1 See for example Mallory & Adams (2006): The Oxford Introduction to Proto-Indo-European and the Proto-Indo-European World, p. 223.
2 For example, Spanish cerrar 'to close', Catalan serrar, French serrer 'to tighten', Basque (Biscayan) zarri 'dense', zarra-tu 'dense; to close', ultimately from PNC *=utɕE(rV) 'thick, fat'.


  1. Any idea what the basis is for assuming that the Germanic word is borrowed from Celtic?

  2. Germanic received a number of Celtic cultural loanwords like 'iron', 'land', 'lead', 'leather', etc.

  3. But how do we know it's a Celtic loan rather than, say, a common substratal borrowing or a PIE word retained only in Celtic and German?

  4. I'm sure there's no PIE etymology for this word.