The Celtic Veneration Of Water From The Late Bronze Age To The Medieval Period, And The Search For The Lost Celts Of Britain

By Robin Melrose

Most of the remaining speakers of Celtic languages live in the British Isles, in Wales, northwest Scotland and Ireland. Until the Anglo-Saxons began settling in England in the 5th century, Celts were spread throughout Britain, but today almost all that remains of these Celts is place-names derived from early Celtic. Almost all, becaus

• Language: English
• Publisher: Robin Melrose
• Release date: Mar 10, 2023
• ISBN: 9781915889584

Book preview

INTRODUCTION

Most of the Celtic languages that still survive today, known as the Insular Celtic languages – Irish, Welsh and Scottish Gaelic – are spoken in the British Isles, the exception being Breton, which is spoken in Brittany, western France. The Celtic languages belong to the Indo-European family of languages spoken in Europe from Ireland and Portugal to Russia, south into Iran and east into the Indian subcontinent. The homeland of the Indo-European languages has been discussed for over a century, but the most popular hypothesis today is that Indo-European languages originated in the Pontic-Caspian steppe. This vast region extends from the northern shores of the Black Sea as far east as the Caspian Sea, from Ukraine to western Kazakhstan, and can be associated with the Yamnaya culture, which flourished in the late Copper Age and Early Bronze Age in the Pontic Steppe, between the Black Sea and the Caspian Sea.

How the Indo-European languages spread west throughout Europe remains the subject of lively debate, but it is likely that the Bell-Beaker culture played a part in this. The Bell-Beaker culture followed the Neolithic Corded Ware Culture, which stretched from the Rhine in the west to the Volga in the east, from Germany to the northern part of Ukraine and the European part of Russia. According to Wolfgang Haak and his colleagues¹, the Corded Ware people were genetically closely related to the people of the Yamnaya culture of the Pontic steppe, documenting a massive migration into the heartland of Europe from its eastern periphery.

This would suggest that the Beaker people of central Europe were the product of migrations from the Pontic Steppe. Recently a study has shown that steppe-related ancestry was introduced into Britain in the Early Bronze Age, presumably by Beaker people. In a recent study, Iñigo Olalde and his colleagues presented data from 400 Neolithic, Copper Age and Bronze Age Europeans, including 226 individuals associated with Beaker-complex artefacts. The authors note²:

We detected limited genetic affinity between Beaker-complex-associated individuals from Iberia and central Europe, and thus exclude migration as an important mechanism of spread between these two regions. However, migration had a key role in the further dissemination of the Beaker complex. We document this phenomenon most clearly in Britain, where the spread of the Beaker complex introduced high levels of steppe-related ancestry and was associated with the replacement of approximately 90% of Britain’s gene pool within a few hundred years, continuing the east-to-west expansion that had brought steppe-related ancestry into central and northern Europe over the previous centuries.

In other words, the Beaker people of Britain were related to the people of the Pontic steppe where Indo-European languages were first spoken, and these Beaker people gradually replaced the Neolithic Britons with whom they came in contact.

Although most of the surviving Celtic languages are spoken in the British Isles, we know little about the Celts of Britain before the early medieval period. Fortunately, however, Greek and Roman writers have left us some information about the Iron Age Celts of Gaul (France), who spoke a Continental Celtic language known as Gaulish. Thanks to these authors writing in the 1st century BC and the early centuries AD, we know that the Celts of Gaul practised a cult of the head, gave lavish feasts, and deposited precious objects in watery places. The cult of the head and lavish feasts ended with the coming of the Romans and the growth of Christianity, but watery places continued to play an important role in France and Britain until the Middle Ages.

In this book I want to focus on the role of watery places through the ages. In Chpter 1 I explore the possible origin of the Celts in the Late Bronze Age, around 1000 BC, examine some examples of the Continental Gaulish language, and look in more depth at the cult of the head, feasting among the Gauls, and Gaulish sacred lakes. In Chapter 2 I turn to Late Bronze Age Britain, looking at sites where feasting took place, and looking at the deposition of precious objects in watery places, including rivers, fens and lakes, from the Somerset Levels to the Cambridgeshire fens and the lochs of Scotland. Chapters 3 and 4 cover the Iron Age in Gaul and Britain. In Chapter 3 I briefly survey hillforts in Gaul mentioned by Julius Caesar in his conquest of Gaul, and look in more depth at hillforts in Britain, often associated with a British version of the cult of the head, and at the vitrified hillforts of Scotland. In Chapter 4 I refer to coastal promontory forts alluded to by Caesar in his description of the Veneti of Brittany, then discuss coastal promontory forts in Cornwall and Devon, Wales and Scotland, before turning my attention to the Somerset Levels, the Cambridgeshire fens, and the crannogs (dwellings on artificial islands) found in Scottish lochs in regions as widely spread as Galloway, Argyll, and Perthshire.

Chapter 5 is a brief interlude examining the Belgae of northern Gaul, who may not have been Celts and migrated to southeastern Britain in the 1st century BC, playing a significant role in Hampshire, parts of Wiltshire, Sussex, Kent, Essex and Hertfordshire. Chapter 6 is a survey of watery places in Roman times, in Gaul and more especially in Britain. There were shrines in Roman Britain associated with water, from Jordan Hill Roman temple in Weymouth (Dorset), which overlooked the English Channel, to Aquae Sulis (Bath in Somerset) with its sacred spring, as far north as Northumberland, with a well dedicated to the goddess Coventina. Beyond Hadrian’s Wall there were crannogs in Ayrshire, Perthshire, and Inverness, and a sacred cave on the shore of the Moray Firth.

Chapter 7 brings us to the period after the last Roman legions left Britain, and pagan Anglo-Saxons began settling in the east of what is now England. The western parts of Britain were initially unaffected, and Christians living there often established monasteries on islands: like Lundy Island off the north coast of Devon; Caldey Island off the coast of Pembrokeshire, southwest Wales; Holyhead, Anglesey off the north coast of Wales; Inchmarnock off the coast of southwest Scotland; and Iona off the west coast of Scotland. In addition there were two early medieval crannogs in Argyll, one of which may have had links to the nearby royal centre at Dunadd, where Gaelic was the main language spoken. In Chapter 8 I consider the Picts, whose centre of power lay on the Moray Firth in northeast Scotland. Judging from tales in Adomnan’s Life of Columba, water played a significant role in Pictish life, and they inhabited coastal promontory forts like Burghead in Moray and Dunnicaer in Aberdeenshire, as well as crannogs in Perth and Kinross and Aberdeenshire. Although they spoke a Celtic language, there are hints that they may have once spoken a language related to Germanic.

Chapter 9 looks at the early Anglo-Saxon kingdoms of Wesssex, East Anglia, Mercia and Northumbria, where watery places were embraced by the newly Christian kingdoms. In the late 7th century King Ine of Wessex established a monastery at Glastonbury close to an old Roman well (St Joseph’s well), and also founded a monastery at Wells in Somerset, near the sacred spring now known as St Andrew’s Well. Æthelthryth, the daughter of king Anna of East Anglia, became abbess of a monastery at Ely in the Cambridgeshire fens, and Guthlac, the son of Penwalh of Mercia, set up a hermitage on an island in the fens of south Lincolnshire. In the kingdom of Northumbria a bishopric was established on the island of Lindisfarne (Northumberland), and there was a hermitage on Bass Rock in the Firth of Forth.

Finally in Chapter 10 I briefly survey early medieval holy wells in France, then examine in depth the later medieval holy wells of Britain. As might be expected, holy wells tend to be found in areas associated with watery places in the prehistoric or Roman period. There are numerous holy wells in Cornwall and Wales; there are holy wells in Dorset and Somerset, but not in Hampshire and Wiltshire. There are holy wells in northern England and southern Scotland, where the kingdom of Northumbria once ruled; but in the Pictish and Gaelic areas of Scotland, medieval holy wells are harder to find.

CHAPTER 1

THE CELTS OF IRON AGE GAUL: THE CULT OF THE HEAD, FEASTING, AND DEPOSITING PRECIOUS OBJECTS IN SACRED LAKES

The Celtic Languages Today

The Celtic languages are part of the Indo-European family of languages, spoken from Europe to Iran and large parts of India. Today most Celtic languages are spoken in the British Isles. Irish was widely spoken in Ireland until the 18th century, and is still spoken in the west of Ireland (the Gaeltacht); Cornish was widely spoken in Cornwall until the mid 16th century, when it was gradually displaced by English; Welsh was the main spoken language of Wales until around 1800, and even today one in five inhabitants of Wales speak Welsh; Scots Gaelic was spoken in north-western Scotland (the Highlands) until 1500, and is still spoken in the Outer Hebrides. In addition, a language called Cumbric, related to Welsh, was spoken in northwest England and southwest Scotland, possibly until the 10th or 11th century. In addition, Breton, which is related to Cornish, is still spoken in Brittany (France), mostly in the west of the region.

Where Did the Celts Come From?

Celts from the East?

We all know where the Celts are today – Ireland, Wales, Scotland, Brittany in France and, until recently, Cornwall in England, but we still don’t know where they came from. Starting in the 19th century, three theories have been proposed to explain Celtic origins. For most of the 20th century, scholars have supposed that the Celts originated in the Late Bronze Age Urnfield Culture of central Europe and spread west to France, Iberia and the British Isles. However, this supposition is contradicted by the evidence from classical writers. According to Greek and Roman authors, the Celts entered history around 400 BC, when they began migrating from north of the Alps into the Mediterranean lands, as documented by the Greek historian Polybius (2nd century BC) and the Roman historian Livy (59 BC-17 AD). Livy says the migrations began in 600 BC, while Polybius places them 200 years later. It is possible that both are right: small scale movements of people may have begun in 600 BC and resulted in the Lepontic Celts of Lake Como and Lake Maggiore on the Swiss/Italian border, while the large-scale migrations may have occurred at the end of the 5th century BC¹.

Julius Caesar, writing in the 1st century BC, says that the Volcae Tectosages were settled north of the Danube in ‘the most fertile areas of Germany, around the Hercynian forest’. Their tribal name is linguistically Celtic (‘property seekers’), as is the name of the Hercynian forest (‘oak forest’). Caesar says that the Volcae Tectosages had migrated eastwards across the Rhine from Gaul², possibly from the region around Toulouse, in the southwest of France.

The Gaulish origin of the Celts in northern Italy is confirmed by archaeology and by other classical writers. An early Celtic presence in the Marseilles hinterland is provided by Celtic personal names on 5th- and 4th-century pottery from Lattes (Hérault) and Ensérune (Hérault) , and by Apollonius of Rhodes’ allusion in the 3rd century BC to Celts and Ligurians along the river Rhône.

Celts from the West?

Recently an alternative view of Celtic origins has been proposed by the authority on Iron Age Britain, Barry Cunliffe, and the scholar of Celtic languages and culture, John Koch, that Celtic originated in the west of Europe, along the Atlantic ocean. The shift from a central European homeland for the Celts began with Colin Renfrew. In 1987 Renfrew published his controversial hypothesis that by 4000 BC Proto-Indo-European had already spread, with the Neolithic farmers, from Anatolia to France, Britain, Ireland, and probably ‘much of Iberia also’, and that the Celtic languages emerged in situ, ‘essentially in those areas where their speech is later attested’³.

In 2001, in Facing the Ocean, Barry Cunliffe theorized that Celtic developed as a lingua franca spoken ‘from Portugal to Britain by the middle of the first millennium BC’, and could be argued to have ‘developed gradually over the four millennia that maritime contacts had been maintained, perhaps reaching its distinctive form in the Late Bronze Age’. In time Cunliffe came to envisage a deeper chronology: that Celtic developed in the Atlantic zone as early as the fourth millennium BC, when contact along the seaways began to get under way. In the new edition of The Ancient Celts (2018), he favours the hypothesis that a Proto-Indo-European language from Anatolia had reached the Atlantic along with Neolithic farming by 5500 BC. He prefers this to the more popular theory that Proto-Indo-European first arrived three millennia later from the Pontic-Caspian steppe and now hypothesizes that ‘Atlantic Celtic’, having evolved in situ from Renfrew’s hypothetical Neolithic Proto-Indo-European, was the primary form of Celtic, which then spread eastwards with the Maritime Bell Beaker culture in the third millennium⁴.

According to Patrick Sims-Williams, Cunliffe believes that the Celtic language group split from other Indo-European languages around 4000 BC, and that the division into Brythonic (e.g. Welsh) and Goidelic (e.g. Irish) languages occurred around 900 BC. To reach these conclusions, he relies on glottochronology, which Sims-Williams calls a pseudo-science⁵ (glottochronology is a statistical technique that typically uses the percentage of shared cognates between languages to calculate divergence times by assuming a constant rate of lexical replacement) . Sims-Williams rejects these conclusions, and he also rejects John Koch’s argument that Tartessian (a language spoken in southwest Spain, and written in the Southwestern script) is Celtic⁶, citing a number of scholars in the field who are almost unanimous in rejecting Koch’s identification of Tartessian as Celtic.

Celts from Gaul?

Sims-Williams is skeptical about the Celtic from the west hypothesis, and rejects the Celtic from the east argument; instead he suggests that Celtic originated in the 2nd millennium BC somewhere in Gaul. One argument for this is that it keeps Celtic fairly close to Italy, which suits the view that Italic and Celtic were in some way linked in the second millennium. During the 1st millennium BC, says Sims-Williams, Celtic spread into eastern Iberia …, into northern Italy (as first evidenced by the Lepontic inscriptions in the sixth century), into Britain, and perhaps already into Ireland⁷.

Sims-Williams’ belief that Celtic spread into Britain in the 1st millennium BC is supported by a recent study8. According to the study, the largest analysis of ancient DNA to date has revealed a mass migration of people from what is now France into England and Wales during the late Bronze Age, which may have spread Celtic languages to Britain. The study notes that two large migrations of people into Britain were previously known, the first taking place around 6000 years ago9.

The ancestry of these people came mostly from a group known to archaeogeneticists as Early European Farmers, with around 20 per cent from another group called Western European Hunter-Gatherers. This migration led to the replacement of most of the existing local hunter-gatherer ancestry.

Around 4500 years ago, at the start of the Bronze Age, there was a second migration that consisted of descendents of livestock farmers from the Pontic-Caspian steppe – grassland that spans from present-day Bulgaria to Kazakhstan. Ancestry from this group eventually formed at least 90 per cent of the genetic make-up in Scotland, England and Wales.

The first migration was the Neolithic one that first brought agriculture to Britain, the second was that of the Early Bronze Age Beaker people, who probably brought the first Indo-European language to Britain. The study goes on to say that people living in England and Wales today have more ancestry from Early European Farmers than people in the Early Bronze Age did, suggesting a third migration from Europe may have occurred more recently.

Ian Armit at the University of York in the UK and his colleagues sequenced the genomes of nearly 800 individuals from the Bronze Age to the Iron Age whose remains were found at archaeological sites in Britain and in western and central Europe. They looked at the proportion of Early European Farmer ancestry in these ancient people over time.The team found evidence of a third mass migration into Britain from France that took place between 1000 BC and 875 BC, during which Early European Farmer ancestry increased from around 30 per cent to roughly 36 per cent on average in southern Britain by the late Bronze Age. In the Iron Age, this stabilised at nearly half of the ancestry in populations of England and Wales.

Continental Celtic: The Gaulish Language

Inscriptions in Gaulish

The Gaulish origin of the Celts is reinforced by the existence of the Gaulish language. Gaulish is a Continental Celtic language well attested in the early Roman period in a number of inscriptions – in fact, it is virtually the only Celtic language with a written history before the Middle Ages. The Gaulish inscriptions include the Larzac tablet from L’Hospitalet-du-Larzac in the Aveyron department of southern France, which in Roman times was a village on the road from Lodève (Roman Luteva) to Millau (Roman Condatomagus), with a cemetery containing 250 burials; the Chamalières tablet from Chamalières near Clermont-Ferrand, which in Roman times was Augustonemetum, with a population of 15,000-30,000; and the Coligny calendar from Coligny in the Ain department of eastern France, not far from the border with Switzerland, which was found along the ancient Roman road from the important Roman city of Lugdunum (Lyon) to Lons-le-Saunier, renowned for its saltworks.

The Larzac Tablet

The lead-tablet from Larzac, which is dated to around 100 AD, contains one of the longest of the Gaulish inscriptions. It is two pieces, each written on both sides, and seems to be a magical inscription about women casting spells, although much of it is extremely difficult to interpret. Among the words that can be interpreted is bnanom, which may be the genitive plural of the word for ‘woman’ (cf. Old Irish ben ‘woman’ (genitive singular mná), and bricto[m] a word for a spell (cf. Old Irish bricht ‘spell’). Anuana is probably the plural noun ‘names’; cf. Old Irish ainm, Welsh enw ‘name’. The latter part of this side of the inscription seems to be a list of the women involved, with the Gaulish words for both ‘mother’ matir and daughter duxtir, and this is confirmed by the fact that we can see pairs of names, thus adiega matir aiias (l. 14) ‘Adiega m other of Aiia’ beside (l. 11) aia duxtir adiegias ‘Aia daughter of Adiega’¹⁰.

The Chamalières Tablet

Although not all of the Chamalières tablet (50 BC – 50 AD) is understood, some words can be interpreted. The verb of the first sentence is uediiu-mi I beseech, pray (cf. Old Irish guidiu, Old Welsh gweddiaf). It then invokes Mapon Arverniiatim, probably meaning [the god] Maponos of the Arverni tribe The third line includes the phrase brichtia anderon by a magical spell of underworld beings. Between lines 7 and 8 is the phrase toncnaman tonsciiontio, which has been compared with the formulaic oaths, Old Irish tongu do dia toinges mo thuath I swear to the god by whom my tribe swears, and Welsh tyghaf tyghet I swear a destiny. The text ends with the repeated formula Luge dessu-mmi-iis By Lugus I prepare them (set them right).¹¹

The Coligny Calendar

Some elements of the Coligny calendar (2nd century AD) can be compared to known Irish or Welsh words. The calendar begins with SAMONI, which is cognate with Old Irish Samain (a harvest festival on November 1), and contains the root for summer (Old Irish samrad). GIAMONI, six months later, contains the word for winter (Old Irish gemred, Old Welsh gaem). The other month names are more difficult to explain with the exception of EQUOS horse (Old Irish ech horse, Welsh ebol colt), a month-name also attested in some Greek dialects (Calabrian Hippios, Aetolian and Thessalonian Hippodromios)¹².

The Pottery from La Graufesenque

On a more mundane level, we have the pre-firing graffiti on the pottery from La Graufesenque near Millau in the Aveyron department in the south of France. Among the graffiti are the Gaulish ordinal numbers¹³:

cintux (Welsh cyntaf) – first

allos (Welsh ail) – second

tritios (Welsh trydydd) – third

petuarios (Welsh pedwerydd) – fourth

pimpetos (Welsh pumed) – fifth

suexsos (Welsh chweched) – sixth

sextametos (Welsh seithfed) – seventh

oxtumetos (Welsh wythfed) – eighth

nametos (Welsh nawfed) – ninth

decametos (Welsh degfed) – tenth

The Ratiatum Lead Tablet

We can compare these numbers to those found on a lead tablet at Ratiatum (now Rezé), to the south of the Loire and opposite Nantes in western France, in the territory of the Pictones. Occupied from the reign of Augustus on, the city of Ratiatum had many monuments of which little remains today except in the writings of local scholars of the last century: in 1853 a great colonnaded portico was found near the west portal of the present-day church¹⁴. The lead tablet at Ratiatum contains several Gaulish numbers, including the following¹⁵:

suexxe (La Graufesenque, suexsos)

ali (La Graufesenque, allos)

trilu (this includes tri three)

paetrute (La Graufesenque, petuarios)

pixte (La Graufesenque, pimpetos)

The number pixte is unexpected - according to David Stifter, an authority on Old Irish and Continental Celtic languages, pixte appears to be related to Latin quintus, Greek pentos, Old Prussian piencts and Lithuanian penktas¹⁶.

The Customs of the Iron Age Gauls

The Cult of the Head

We know nothing of Celts in the British Isles until the early medieval period, and everything we know about early Celts come from classical writers describing the customs of Celtic Gauls in the south of France, which was under Roman control from about 120 BC. Sometime in the early 1st century BC the Greek philosopher, geographer and historian Posidonius, who died around 51 BC, travelled through the south of France, presumably starting in the Greek colony of Massalia (Marseille), and wrote a work on the customs of the local Celts, now lost but quoted by other writers. For example, the Greek geographer Strabo, in his Geography, written between 23 BC and 20 AD, refers to cult of the head among the Iron Age Gauls, quoting Posidonius¹⁷:

when they depart from the battle they hang the heads of their enemies from the necks of their horses, and, when they have brought them home, nail the spectacle to the entrances of their homes. At any rate, Poseidonius says that he himself saw this spectacle in many places, and that, although at first he loathed it, afterwards, through his familiarity with it, he could bear it calmly.

Evidence for this cult of the head can be found at the oppidum (defended settlement) of Entremont near Aix-en-Provence, 19 miles north of Marseille. This oppidum was founded at the beginning of the 2nd century

BC

as a small settlement of about two acres, and was enlarged around 150

BC

. Between 125 and 100 BC, a monumental hall 65 feet long and 16 feet wide was erected. The ground floor of this building was a hypostyle hall with a packed mud floor. The pillars of the facade rest upon a stylobate (platform) made up of reused elements from at least one earlier religious sanctuary. These elements, which may date to as early as 500

BC

, include representations of human heads without mouths, and cephalic concavities destined to receive human skulls or parts of skulls. Around twenty skulls of aged men, which were pierced for suspension, were found dispersed around the stylobate of the hall. They were probably originally attached to the wood of the facade. The only remains found on the floor are fragments of Italian amphoras, which suggest the practice of libation rituals¹⁸.

This was not the only such sanctuary in the region. At Pech Maho, Sigean, near Narbonne in the Aude department, the excavator Yves Solier, who investigated the site in the 1960s and 1970s, discovered a number of human skulls that had apparently been nailed to a pillar, sometime in the 4th century BC. The oppidum was destroyed around 200 BC, and at subsequent excavations between 2004 and 2011, archaeologists uncovered a post -destruction layer filled with a huge number of horse bones mixed with human remains¹⁹.

It seems that the cult of the head has its roots in the Early Iron Age, in a form of burial known as excarnation and secondary burial. In 1996 archaeologists excavated an oval enclosure dating to the 8th-7th century BC (the Early Iron Age) at Vestric-et-Candiac in the Gard department, to the southwest of Nîmes. Just outside the enclosure they found two burials²⁰. In Grave 1 they found a skull, the remains of the arms (but no hands), part of the pelvis, and the upper part of the two femurs (thigh bones). The burial was accompanied by two pottery vessels. Some 45 feet to the east was Grave 2. It contained a skull, two arms (again without hands), the upper part of the two femurs, and the two tibias (shinbones). The burial was accompanied by a pottery sherd and two flint flakes.

In this region known as the Garrigues of Languedoc, which includes parts of the Gard and Hérault departments, excarnation and secondary burial were very common in the Early Iron Age. For example, at the Sadoulet cemetery at Pompignan to the west of Nîmes, Tumulus 1 contained a cranium, long bones of arms and legs and a mandible (jaw bone), together with an iron dagger and an iron knife, dating from the early 6th century BC²¹.

Excarnation and secondary burial, first studied in a 1907 paper by the French sociologist Robert Hertz, is a process in which the body is allowed to decompose and the dry bones are given a second (or secondary) burial. Among many people who practise this burial rite even today, or practised it until very recently (and that includes parts of rural Greece), it is believed that the soul of the deceased cannot leave the body until the flesh has decomposed. While the flesh is decaying, the deceased is neither alive nor dead, and is unable to enter the society of the dead, and leads a shadowy existence on the fringes of human habitation. Once the flesh is decayed, the dry bones are recovered, a great feast is held, the bones are given a second burial, and the deceased is then free to enter the land of the ancestors²².

Gaulish Sacred Lakes

The Greek writers Strabo and Athenaeus touch on other Gaulish customs. In discussing the Tectosages of southwest France, Strabo reports that they were involved in the sack of Delphi in 279 BC. When the Roman statesman and general Caepio plundered the temples of Tolosa (Toulouse in southwestern France) in 105 BC, part of the treasure he stole came from Delphi, which led to a series of misfortunes for Caepio. However, says Strabo²³,

the account of Poseidonius