Of Romano-British Canterbury, we know both
much and little. The name of the place is happily quite certain. It is
Durovernum, which seems to be good Celtic for ‘the fortress or
possession of Vernos,’ Vernos being an otherwise well-attested Celtic
personal name. The evidence for it is excellent. The Antonine Itinerary
states that Roman roads ran from the three Kentish ports, Dover, Lympne
and Richborough to Durovernum, 14 miles from Dover, 16 miles from Lympne,
and 12 from Richborough, and then a single road continued on to Rochester
(27 miles) and London. The roads can still be followed to their common
centre, Canterbury, and the distances agree. Other ancient sources lend
confirmation, though with slightly distorted spelling. Ptolemy, as already
stated, names ‘Darouernon’ as a town of the Cantii. The anonymous
Ravennas puts ‘Duroaverno Cantiacorum’ in close neighbourhood to
Dubris and Rutupis, Dover and Richborough.10 The
Peutinger Table or Map (P1. VI), which
contains a fragment of eastern Britain, marks ‘Duroaverus’ in an
appropriate position in East Kent.11
The name is further and more strongly attested by a proof
which is rarely available for Romano-British place-names. Canterbury is
one of the few English sites whose Romano-British names survived into
English times in their Roman forms. Most of the names which lived on, did
so in truncated shapes, generally as the first syllable of a later English
appellation. Such, for instance, are Richborough, Winchester, Litchfield,
which embody the first letters of Rutupiae, Venta and Le(c)tocetum. At
Canterbury and at two or other sites in Kent, and rarely elsewhere in
England, the English kept the old name, practically unaltered, when they
wrote Latin. An ‘abbas de Doroverno attended the Council of Paris in
614. Pope Boniface V, writing to Justus of Canterbury in 622, uses the
phrase civitas Dorobernia. An early gold coin, minted probably in
the 7th century, bears the legend Dorovernis civitas. Early
charters of that and succeeding centuries mention the civitas
Doruvernis, and Bede is only following a false analogy in grammar when
he uses Doruvernis and Doruverni as genitive and ablative of
the name for Canterbury.12 The name Doruvernum was
plainly familiar to the early English ; and when after A.D. 600 they began
to use Latin freely, they used that name for Canterbury.13
10 Cantiaci occurs only in
the Ravennas and is probably a late equivalent for Cantii. Two parallels
may be cited, Parisiaci for Parisii, which appears for Parisii—often as
an adjective and sometimes as a noun—in late writers, and once (as an
adjective) on an early 1st-century altar at Paris (C. xii 3026 nautce
Parisiacl); and Osismiaci, which the Notitia twice substitutes for
Qsismii. The termination in all three cases is probably (as Schulze, Lateinische
Eigennamen, init. thinks) not the Greek iakos, but the Celtic âcos.
The proper function of this suffix is, indeed, to make place-names,
but it seems to have been also used sporadically to form tribe-names. An
early parallel is supplied by the British tribe Segontiaci, mentioned in
Ceasar’s Gallic War, who must have derived their name from some
Segontium or Segontios.
11 Itin. Ant. 472, 473;
the best MSS. read Durouerno, others Duroruerno, Durarueno,
Duraruenno, or the like, always with u in the first syllable.
Ptolemy ii, 3, I 2, Δapovepvov,
a spelling which is followed by Holder in his Sprachschatz, but
which seems wrong. Ravennas, p. 428, Peutinger Table, ed. Miller. For the
derivation see D’Arbois de Jubainville, Mots Gaulois, p. 210.
12 For the council of Paris see Mon. Hist.
Germ. Concilia aevi Meroving, ed. Maassen, p. 192; for the papal
letter, Wilkins, Concilia, i, 32, hence Mansi Nova coll. x,
553. The coin is discussed by Longpérier, Numism. Journal, ii
(1838), 232, and Revue Numism. 1841, p. 437; Cartier, Numism.
Chron. ii (1839), 204; D. H. Haigh, ibid.. (1841), 120; C. F.
Keary, English Coins in the Brit. Mus., Anglo-Saxon series, introd.
etc.; its date seems fairly certain. Dorobernia and similar forms abound
on Kentish coins of the 9th and following centuries, though one of the
earliest of those usually cited, that inscribed DOROBREBIA CIBIT, contains
the Roman name of Rochester, p. 81. For early charters, see Kemble, nos.
vii, ix, xiv, etc. Post-Conquest writers use Dorobernia often for Dover:
so Will, of Malmesbury (Rolls Ser.), p. 376; Higden, ibid.
ii, 56; Flor. of Worc. Chron. A.D. 1051.
13 Faussett, Arch. Journ. xxxii,
382, suggests that the knowledge of the name Durovernum was reimported
into Britain by continental clergy. But there is no known source from
which these clergy could have learnt it on the Continent. Certainly they
did not possess the Itinerary. See Engi. Hist. Rev. Oct. 1895.