Presented here is an early statement on the progress of, and the implications arising from, the collection of ‘wet’ place-names in Wales written by Kelly Kilpatrick.

The process of data collection can be laborious and time-consuming, requiring hours of systematic data entry and problem solving. However, it is always well worth the effort when results are generated! Data collection is just as important in place-name studies as it is in other disciplines. Place-name specialists often want to collect and analyse specific types of name-related data. For example, we might collect data relating to the element(s) that make up a name, the language(s) of the element(s), the dates and historical sources recording the name, as well as administrative, geographical and environmental information. This type of information can enable the toponymist to determine the meaning of the name (sometimes with reference to habitational or environmental features in the local landscape), the language spoken by the people who created the place-name, and the name’s linguistic development over time. When such place-name studies are applied to large areas, they can provide considerable information about past peoples (including migrations, languages and dialects, social, religious and administrative customs) and historic environments. Place-names are essentially a window into the past, and collecting place-name data is one of the principal steps to opening this window.

My role in the Flood and Flow project is to collect data on place-names in Wales that contain references to water in their meaning. The aim of the project is to provide relatively comprehensive coverage of the major place-names in Wales that relate to water. By ‘major-names’ we mean the names of parishes, larger settlements, villages, and hamlets, as well as names of historical or etymological interest. However, Wales has not benefited from a nation-wide survey of place-names such as that conducted for England by the English Place-Name Society. Therefore, to compile a thorough survey of the major ‘wet’ names across Wales I have had to rely on primary data collection and a variety of resources, both published and unpublished. The scope and coverage of information also varies considerably in the resources available. Nevertheless, in a span of nearly eight months, and after sifting through an initial 4,175 outline place-name entries in our database, a total of 1,128 entries were identified as ‘water-related’. The relevant place-name data described above was then added for each entry. It turns out that nearly 30% of all the place-names in the database for Wales are associated with water, either directly or indirectly. In addition to these, after a second phase of data collection, a further 361 water-related place-names have been added to the database, bringing the total to 1,489 ‘wet’ place-names in Wales.

Wales 1

Collecting data for a place-name survey can present a number of challenges, but working with this material is always interesting and a lot of fun. Each name, or data entry, is a puzzle in itself, and the place-name scholar has to assemble the evidence to uncover the name’s meaning, trace its history in documentary sources, and analyse its relationship to the local environment. As these individual puzzles are completed they each form pieces of a larger puzzle, and once a sizeable body of place-name data has been generated, then the real fun begins—analysing the data. Now that we have assembled a considerable amount of our data set, a variety of initial results have emerged, suggesting potential avenues for future research.

In earlier stages of the project we identified a total of 259 Welsh water-related place-name elements. However, not all of these have been found in the major place-names across Wales, and some of these elements are only found in minor-names and place-names more recent in origin (both fall outside the time-frame and scope of this project). For example, the element rhewyn (‘gutter, ditch, channel, drain; streamlet, rivulet, brook’) is first attested between the fifteenth and sixteenth century according to the Geiriadur Prifysgol Cymru. This element has (thus far) only been found in two nineteenth-century field-names in Letterston/Treletert and Clydai parishes in Pembrokeshire (Charles 1992, vol. 1: 220, 379), and because these are minor-names and modern, they are not included in the database.

In the major-names of the database, we have so far encountered at least 159 different water-related elements in the place-names of Wales, around 131 of which are in Welsh and the remaining 28 primarily in English and very occasionally Old Norse. Some Welsh place-name elements are very common and widespread.

Wales 2

Our research has shown that the most common water-related Welsh place-name elements are:

aber ‘river-mouth, estuary, confluence; river, stream brook’

blaen with the sense ‘source of a river or stream, upper reaches of a river or stream’

dôl ‘meadow (often by a river), dale, field, pasture, valley; bend, ox-bow’

dyffryn ‘valley, vale, bottom’

ffynnon ‘spring, fountain; well, source, origin’

glan ‘river-bank, brink, edge; shore; slope, bank, hill-side, mountain’

glyn ‘glen, dingle, dale, dell, (wooded) valley’

gwern ‘alder-tree(s); alder-grove, alder-marsh, swamp, quagmire; damp meadow’

llyn ‘lake, pool, pond, puddle, fishpond, moat’

melin ‘mill, manufacturing mill’

nant ‘river, stream, brook, rivulet; torrent, ditch; valley, glen, dale; ravine, gorge’

pont ‘bridge’

porth with the sense ‘port, harbour, haven; estuary, landing-place, ferry’

pwll ‘pool, puddle, pond; hole, pit, depression’

rhos ‘(upland) moor, heath(land), down, meadow on high land; marshland’

rhyd ‘ford’

ynys ‘island; river-meadow’

ystrad ‘river-valley; level lands bordering a river’

Included in this list are some common terms for valleys, which naturally have watercourses flowing through them.

Some elements such as rhos and dôl have a range of meanings, both related and unrelated to water, and their precise interpretation can only be determined by careful landscape analysis. Elements such as aber, nant, rhyd, and pont, to name but a few, are practically always directly related to water. There are some rare elements, however, whose relationship to water is uncertain, and these pose vexing questions for the data collector. For example, Llanddowror in Carmarthenshire is first attested in the Book of Llandaff (also known as the Liber Landavensis) in a charter dated to before c. 1025 (Coe 2001: 493-5). Llanddowror is located a short distance south of the River Taf. The name itself is composed of llan ‘church, churchyard’ + dyfrwr, a compound of Old Welsh dubr ‘water’ and guir ‘men’ (Coe 2001: 494), which is defined by the Geiriadur as ‘one who lives in the water; one who drinks water only; water-carrier; water-bailiff; waterman; boatman’. The church of Llanddowror is dedicated to St Teilo. In the Life of St Teilo in the Book of Llandaff, the place-name Llanddowror is said to be derived from an incident involving seven brothers who were rescued and baptized by St Teilo after their father had attempted to drown them in the River Taf. Afterwards, the brothers continued to live as monks in St Teilo’s community. In the Life, the term dyfrwr in the place-name is said to derive from the brothers’ religious life and because they depended on fish alone (nisi aquatilibus piscibus) for sustenance (Evans 1893: 128-9; Rees 1840: 121, trans. 368) and ‘were called Dyfrgwyr, because they were found in the water, escaped from the water, and were maintained by fishes of the water. Dwfr gwyr in the British language, signifying, men of the water’ (Rees 1840: 121, trans. 368). How is the data collector to interpret such a place-name? It is certainly associated with water, but this association is probably not a direct response to the natural environment, but rather to the medieval monastic diet—which in this case, certainly relied on the natural resources from rivers and streams.

Wales 3

St Teilo’s Church, Llanddowror Image © Copyright John Lord and licensed for reuse under the Creative Commons Licence.

Wales 4

What can the place-name data tell us about human responses to water and living with water in early medieval Wales? This is one of the main questions we hope to answer in this project, but therein lies another of the persistent problems that affects the study of place-names in Wales—namely, the lack of sources surviving from the early medieval period (c. 700-1100). After querying the database, it is clear that most place-names are first attested in sources dating to the post-Conquest period. Six place-names are first attested in Classical sources ranging from the fourth to fifth century, 10 between the eighth and tenth centuries, and 37 in the eleventh century (largely in Domesday Book entries for place-names near the English border). There is a significant increase in the twelfth century (170), with the greatest spike of all in the thirteenth (376), closely followed by the fourteenth century (288). Just because more place-names are first attested in the later Middle Ages, however, does not necessarily mean that they are later in origin. This pattern of attestation  reflects the survival rate of manuscripts and documentary sources. In the words of Thomas Charles-Edwards (2013: 245), ‘The history of early medieval Wales often has to be written from the twelfth century backwards, since most of our sources survive from collections of texts assembled after the Norman Conquest.’

Unlike England, from which a larger number of early medieval sources preserving place-names survive, Wales has appreciably few early sources—and even fewer surviving that record place-names. Some of the earliest Old Welsh survives in interlinear glosses, marginalia, or in short tracts. Examples include the marginalia and ‘Surexit Memorandum’ of the Lichfield Gospels (Jenkins & Owen 1983 part I: 37-66; Jenkins & Owen 1984 part II: 91-120), the marginal poems of the ninth-century Juvencus Manuscript (Cambridge Juvencus MS Ff.4.42), glosses in the ninth-century ‘St Dunstan’s Classbook’ (Oxford Bodleian MS Auctarium F.4.32), and in the ninth- to early tenth-century Computus Fragment (Williams 1926-7: 245-72). The Cynfeirdd poetry is generally considered to be early, although the dates of the Canu Taliesin and Y Gododdin—which do contain place-name information—are hotly contested, as the earliest surviving manuscripts date to the Middle Welsh period. Furthermore, even though the Canu Taliesin and Y Gododdin contain toponymic information, their geographies are largely focused on the Old North and not Wales.

Fortunately for those interested in studying early medieval ‘wet’ place-names in Wales, the Book of Llandaff, though compiled in the twelfth century, contains a collection of charters that record numerous place-names, and can be shown on linguistic grounds to date from between the late sixth/early seventh to the early twelfth centuries (for which see Coe 2001: 6-50 and Davies 1979). The place-names of the charters in the Book of Llandaff take us directly to the early medieval period. What is more, a cursory comparison of the ‘wet’ elements in these place-names and those in our database provides evidence for a potential early stratum of common water-related elements in early medieval Welsh place-names. For example, aber is first attested in our database in the ninth century (Abergele in Denbighshire), and relatively early in the Book of Llandaff (c. 720 in the now lost name Aper Menei). In contrast, blaen-names have a tendency to occur slightly later. In our database, the first example of a blaen name is found in Blaen Camddwr in Cardiganshire (1184), while in the Book of Llandaff, 27 of 28 place-names containing this element are first attested in the eleventh century, the outlier occurring in the tenth. This suggests that river-mouths, estuaries or river-confluences were more suitable sites for habitation in the early Middle Ages, rather than the sources or upland reaches of a river. Though this aspect of the research is still in its infancy, once further data is collected, comparison and analysis of the ‘wet’ elements in the Book of Llandaff with the names in our database has the potential to provide crucial information about historic settlement patterns in relation to water.

These are just some of the many exciting discoveries that can result from data collection, and the finished database will enable us to study the relationship between people and water in early medieval Wales and beyond. Our future research will include detailed landscape analysis of the ‘wet’ place-names in Wales, so stay tuned for future b(l)og entries and new discoveries.


Coe, Jonathon Baron 2001, ‘The Place-Names of the Book of Llandaf’. Unpublished PhD            Thesis, University of Wales, Aberystwyth. Available on:

Charles, B. G. 1992, The Place-Names of Pembrokeshire, 2 vols. Aberystwyth: National   Library of Wales.

Charles-Edwards, Thomas M. 2013, Wales and the Britons 350-1064. Oxford: University Press.

Davies, Wendy 1979, The Llandaff Charters. Aberystwyth: The National Library of Wales.

Evans, J. Gwenogvryn & Rhŷs, John (eds.), 1893, The Text of the Book of Llan Dâv          Reproduced from the Gwynsaney Manuscript. Oxford: Imprinted at Gloucester by           John Bellows.

Jenkins, Dafydd & Owen, Morfydd E. 1983, ‘The Welsh Marginalia in the Lichfield         Gospels. Part I’, Cambridge Medieval Celtic Studies 5, pp. 37-66.

— 1984, ‘The Welsh Marginalia in the Lichfield       Gospels. Part II: The “Surexit”           Memorandum’, Cambridge Medieval Celtic Studies 7, pp. 91-120.

Rees, William J. 1840, The Liber Landavensis, Llyfr Teilo, or the Ancient Register of the     Cathedral Church of Llandaff. Llandovery.

Williams, Ifor 1926-7, ‘The Computus Fragment’, Bulletin of the Board of Celtic Studies 3,            pp. 245-72.