In the scanty sources for ‘the end of Roman Britain’ there is a small clutch of apparent references to continuing city-life. Orosius writes of a certain municeps — usually understood as a leading townsman — called Gratian, who was for a short time supported as emperor by the troops on the island.Footnote ¹⁰⁹ Zosimus refers directly to poleis, ‘towns’ — in Latin, civitates — which, after the fall of Constantine III, were successfully defended by their inhabitants against barbarian attack, or were urged by Honorius to resort to self-help.Footnote ¹¹⁰ From these references, historians have been ready to draw wide-ranging conclusions, including the observation that there was by now clearly no layer of administration between the central government and the civitates of the region (including northern Gaul), and that the cities were the only surviving entities with a chance of preserving what remained of Roman life. It is on the basis of such sources that, Esmonde Cleary, for example, declares: ‘the municipalities were the only surviving authority after the ejection of Constantine III's officials.’ Chrysos comes to a similar conclusion in his view that ‘the ineffective administration of imperial officials was replaced by those of the cities’.Footnote ¹¹¹

Sadly, closer examination of the sources reveals that they offer no significant information on the North-Western cities. Thus it becomes plain that Orosius' mention of a ‘municeps eiusdem insulae’ in no way categorises Gratian as a member of the municipal aristocracy.Footnote ¹¹² It is just a circumlocution — carefully chosen, and borrowed from Cicero — for ‘an inhabitant of Britain’.Footnote ¹¹³ Likewise, Zosimus' two references to poleis in Britain / northern Gaul cannot automatically be assumed to indicate ‘cities’. Let us examine them again:

1. 1. The inhabitants of Britain react to barbarian attacks after the fall of Constantine III by taking up weapons and seeing to their own defence: ‘eleutherosan ton epikeimenon barbaron tas poleis’. This may be translated as ‘they freed the cities from the threatening barbarians’.Footnote ¹¹⁴

2. 2. Honorius writes a letter to the ‘cities of Britain’ (‘pros tas en Bretannia … poleis’), telling them that they must defend themselves.Footnote ¹¹⁵

In both cases, the translation of poleis as ‘cities’ (i.e. civitates) is, at first glance, self-evident. However, the matter is not straightforward since elsewhere, in similar contexts, Zosimus uses the plural of polis, ‘city’, with a wholly different meaning. At 4.3.1 he describes the partition of the Empire between Valentinian and Valens in a.d. 364. He writes that Valentinian gave Valens the East, deciding that he himself would withdraw to Italy and take over Illyria. With regard to the last, Zosimus' words are: ‘labonta de tas en Illuriois poleis’. Though one might well translate these as ‘to take over the cities of Illyria’, their clear meaning is that Valentinian aimed at taking on the diocese of Illyria, not its individual cities: ‘en Illuriois poleis’ is just shorthand for the whole region.Footnote ¹¹⁶ Comparable is Zosimus 4.20.6, which concerns Goths fleeing before Huns in a.d. 376. Roman commanders on the Danube sought a decision from Valens as to whether they should allow these people to enter imperial territory. Zosimus' term for these commanders is ‘hoi tas epi tou Istrou poleis phrourein tetagmenoi’: ‘those charged with protecting the cities on the Danube’. However, it is plain that he had in mind not commanders of individual cities, but the military hierarchy of the whole region.Footnote ¹¹⁷

The same may well be true of Britain, i.e. in the two passages in Book 6 of his ‘History’, Zosimus has in mind not particular civitates, but the entire province. On this basis, the first translates as ‘they freed their land from barbarians’ and the second as ‘he wrote a letter to the Britons’. In any event, caution must be exercised in drawing any far-reaching conclusions from these passages.

There is, however, a possible objection to this. While in Book 4 Zosimus' source is Eunapius, for Book 6 he uses Olympiodorus, and so it could be argued that his choice of words simply reflects that of his authorities. The force of poleis might, therefore, vary from case to case, preventing any generalisation. However, although Zosimus keeps very close to his sources in content, in his writing he displays, both in language and style, a remarkable uniformity. His clear aim was to re-work his sources into a language that was easily understood by his target audience.Footnote ¹¹⁸ So, for example, his syntax is uniform throughout his ‘History’, which makes direct distortion of his terminology by his sources unlikely.Footnote ¹¹⁹

Finally, of course, it goes without saying that rejection of these passages as references to city-life in no way amounts to disproof of the continuing existence of civitates in fifth-century Britain and the North-Western Empire. What it does suggest is that solid evidence for this must be sought elsewhere.