Further information

Searchable categories

The data can be searched under the following headings: name; printed source; type of source (epigraphic, numismatic, literary); type of object (e.g. coin, stone inscription, lead tablet, samian ware, etc.); place of find (location and county); tribal area (for coins); geographical area (using the area division employed in the find-reports in recent volumes of Britannia); date of find (BC or AD or BC/AD (for periods crossing the BC/AD line), including the facility to search over a date-span); name-type (e.g. tria nomina, single name, patronymic, etc.); gender; name-elements (first element, second element, or suffix); and finally whether the name (or a similar form) occurs elsewhere in the Roman Empire (usually by reference to OPEL; cf. also now CLI). For each name, there is a section for notes where additional comments on a name are supplied, for instance the contents of the inscription. Several search types can be used together to sift the data; for example, it would be possible to take all the names in the database from Bath, and then sift them for names found in the curse-tablets, and search them by gender to extract the female names. Alternatively, one could begin from a particular name-element, e.g. seno- 'old' and then sort for date or location.

Criteria for inclusion

The database contains all the names attested in Roman Britain which are thought to contain Celtic elements. That deceptively simple sentence begs a series of questions: what is meant by 'attested in Roman Britain', 'Celtic elements', etc.? We began by including all the names containing Celtic elements found in the epigraphic and numismatic sources, but it soon emerged that to have a database of Celtic personal names from Roman Britain without Boudicca, Cassivellaunus or Caratacus in it was simply perverse, though justifiable as these names only occur in the literary sources for Roman Britain written a long way from Britain and preserved in much later manuscripts. The decision was made to include them for the sake of completeness. Also attested in Roman Britain are individuals with Celtic names whose origin can be traced to the Continent, for example, some potters and some military personnel; where we were confident of their continental origins, they have been omitted.

The issue of the Celticity of a name is more problematic. Essentially, we regard a name as being Celtic if it contains at least one element which is arguably attested in a Celtic language, whether ancient, such as Gaulish or Celtiberian, or later, such as Irish or Welsh. In some cases, that Celtic element may only be the suffix (typically the suffix -a:ko- (see Russell, '-āko-')), and clearly the Celticity of the complete name is of a different quality than a name such as Matugenus where both elements are demonstrably Celtic. We are, however, fortunate that considerable work has been done recently in identifying Celtic place-names in Continental Europe (see, for example, Patrick Sims-Williams's ACP-N for methodological discussion and references) and we have been able to make use of some of this; for example, we have benefited from the lists of Celtic elements in work arising from the AHRC-funded project run by Professor Patrick Sims-Williams in Aberystwyth, such as those in AILR, Ptol., and CLI (I & S) together with the list of Gaulish elements in DLG and NPC. For each name element, basic bibliographical details have been supplied, sufficient to access the scholarly literature. As ever, decisions have had to be made about inclusion and exclusion at the fringes of the data. As a matter of policy, we have erred on the side of inclusion, indicating our doubts with a liberal sprinkling of question marks; it was felt that such names needed to be included so that users of the database were aware of their existence and could choose to exclude them for themselves. For discussion of the criteria for inclusion, see Mullen 2007a.


Any such database is necessarily provisional. First, it will be updated when new evidence comes to light. Secondly, our understanding of the early stages of the Celtic languages is far from perfect, and it may be that names we have excluded will turn out to be Celtic when more evidence for certain name-elements emerges, and conversely some names may need to be excised. Thirdly, there may be archaeological reasons for adjusting or providing a dating, or indeed for removing a name from the database. All such modifications will be undertaken at regular intervals.

Details of updates:
18 Dec 2008: names added from B 39 (2008).
15 Dec 2009: names added from B 40 (2009) and re-readings and references from RIB III (Es[u][vius (RIB III 3079) has been removed).
9 Dec 2010: names added from B 41 (2010), and especially from TV IV.1 (initially published in this volume, pp. 187-224).
31 Dec 2011: names added from B 42 (2011), and especially from TV IV.2 (initially published in this volume, pp. 113-44); references to NPC also added.
24 May 2013: names added from B 43 (2012).
1 Nov 2013: names added from B 44 (2013).
24 February 2015: names added from B 45 (2014) with some consequential changes.
8 March 2016: names added from B 46 (2015).
14 Sept 2019: names added from B 47 (2016), B 48 (2017), B 49 (2018); references to RLFV also added.

Any comments, suggestions or corrections should be addressed to Paul Russell (pr270@cam.ac.uk) or to Alex Mullen (Alex.Mullen@nottingham.ac.uk).