of the bequest of Charles James O’Donnell: ‘Was the Atlantic Zone the Celtic Homeland?
Early Linguistic Evidence’ (University of Edinburgh 24 April), ‘People called Keltoi, the La
Tène Style, and Ancient Celtic Languages: the Threefold Celts in the Light of Geography’
(Swansea University 29 April, Bangor University 6 May, Aberystwyth University 13 May)
AN APPROACH ABOUT KELTOI (A) – arganthoonios = him who flourishes like or with argyrus(argent,silver,lt.bright,slow),arganthodannos=him who delivers the flowers of argyrus
The modern concept of the Celts equates three categories of evidence:
• (first) people called Keltoí by the Greeks,
• (second) the ancient Celtic languages, and
• (third) the Hallstatt and La Tène archaeological cultures.
To locate the origins of the Celts in time and space, the last category has been
given priority, which gives priority to central Europe, specifically the lands about
the Alps and the Watershed Zone where the great rivers come together—the Danube, Rhine, and Marne and Seine, Loire, Rhône and Saône, and the Po. The key period is the Early Iron Age, about 750–400 BC.
The traditional Celtic studies narrative has therefore tended to view Britain as
peripheral, a zone of secondary expansion. Like Ireland, Armorica, and the Iberian
Peninsula, or Galatia in the east—Iron Age Celtic Britain is often represented as
one spoke of several emanating from the central European hub.
Although this model is not often spelled out in words in this way, it is an accurate
description of many maps drawn to show Celtic expansion, in which radiating
black arrows still seem to imply invasions or mass migrations, even though the
‘Invasion Hypothesis’ has been obsolete for forty years in British archaeology. The
Iron Age central-European model of Celticization is such an ingrained habit of
thought—underpinning innumerable orthodox subtheories—that neither maps nor explicit words are necessary and alternative explanations tend to be overruled quickly on the basis of any number of things that we think we already know.
The main point that I want to make in this lecture is that for two of the three
components of the synthetic Celts—namely, as people called Keltoí and as speakers
of Ancient Celtic languages—there is earlier and better evidence in the Atlantic
west than in the central European Watershed Zone. Therefore, we should no longer
associate those three traditional indicators of Celticity so closely as to assume that
the Celtic languages must have originated in the same time and place as the Hallstatt
and La Tène cultures. In fact, a shift of focus from Iron Age central Europe to the
Atlantic Bronze Age is indicated as the more meaningful starting place for a new
narrative Story of the Celts. Such a new account of Celtic origins will have more
to do with the exchange of ores, ingots, and prestige metalwork and less with
expansionist warbands—though both factors were present in both periods.
The proposed paradigm shift would affect how we understand early Britain in a
number of ways. Not only would Britain have become Celtic, in the sense of Celtic speaking, at an earlier date than usually allowed, but probably by different mechanisms and possibly from a different direction. West Britain, as well as remaining
Celtic speaking longer, possibly became Celtic speaking earlier.
After uncoupling Hallstatt and La Tène from the Celtic proto-language, there is
no longer any a priori reason to rule out the possibility that Britain was an integral
part of the region in which Celtic first evolved from Indo-European. In other
words, the question ‘when did the Celts come to Britain’ may have built into it an
assumption that is no longer valid.
You’ve now heard my conclusions. Let’s proceed to the evidence, starting with
Herodotus’s famous statements about the Keltoí. Writing around 430 BC, he stated
that they inhabited the lands near the source of the river Ister, that is, the Danube. §2.34. I am willing to believe that [the Nile] rises at the same distance from
its mouth as the [Danube], which has its source amongst the Keltoí at and flows right through the middle of Europe, to reach the Black Sea at Miletos’s colony of Istri. The Keltoí live beyond the Pillars of Hercules, next
to the Kuneesioi who are the most westerly people of Europe.
§4.48. . . . the [Danube], that mighty stream which, rising amongst the Keltoí,
the most westerly, after the Kuneetes, of all the European nations, traverses
the whole length of the continent before it enters Scythia.
[adapted from translations of de Sélincourt]
References to Keltoí on the upper Danube coincide well with the established theory of central-European origins, bringing us close to Hallstatt itself and such important Hallstatt and Early La Tène sites as the Heuneberg hillfort and
Dürrnberg-bei-Hallein. And many modern writers have drawn attention to this
apparent confirmation. Unfortunately, the passages themselves alert us to the likelihood that Herodotus was profoundly ignorant about the upper Danube. He wrote that the river was as long as the Nile, flowed across the whole of Europe, and that
its source lay beyond the Straits of Gibraltar and near Pyrene, a name apparently
connected with the Pyrenees.
One possible explanation is that he knew of people called Keltoí on both the upper Danube and in the Iberian Peninsula, but being
unaware of the course of the river, he assumed these were the same group and that
the Danube therefore began in Spain. Given this level of geographical confusion,
fine tuning and second guessing today cannot rehabilitate these passages as solid
evidence for Keltoí in the centre of Europe in the 5th century BC.
In these same passages, Herodotus gives much clearer information about the
other Keltoí in the extreme south-west, where they lived next to the Kuneesioi or
Kuneetes, the westernmost people in Europe. From later sources we know that the
Kuneetes inhabited the Algarve and that their eastern limit was at or near the river
Anas, now the Guadiana, which forms the southern border between Portugal and
Spain. According to Trogus Pompeius (44.4), the Cunetes inhabited the forests of
the Tartessians. Thus, we can place this group of Herodotus’s Keltoí exactly where
later classical sources located the Celtici, in south-west Spain. Before 350 BC Ephoros
likewise wrote that Keltikoi extended as far south-west as Gades near Gibraltar.
According to Strabo (3.3.5), other tribes also called Celtici (Greek Keltikoi) inhabited north-west Spain and shared a common origin with the Keltikoi on the Anas in the south-west [cf. p. 101 above].
In tending to ignore the Keltoí who were certainly in the far south-west in favour of those doubtfully at the source of the Danube, there has also been a general tendency to overlook the fact that Kuneetes looks like a Celtic name. We may compare it to the place-name Cuneetio in Roman Britain, or Old Welsh Cinuit, the name of the founder of the leading dynasty of Dark Age Strathclyde. This same Cynwyd is also a place-name in north Wales. [Cf. pp. 98–9 above.]
Unlike the upper Danube, Herodotus was apparently well informed about the
kingdom of Tartessos in what is now south-west Spain and southern Portugal.
This is not surprising, since Greek imports were common in the rich orientalizing
archaeological culture of Early Iron Age Tartessos, about 775–550 BC. These finds
include ceramics and other manufactured luxuries from Cyprus, Phokaia, Rhodes,
Samos, and Attika, alongside Phoenician imports from Tyre and Tyre’s colonies in
north Africa and southern Spain. A key factor was Tyre’s colony at Cádiz (Phoenician
Gadir) near the Straits of Gibraltar and on the southern edge of Tartessos.
As well as information on a species of Tartessian weasel (4.192), Herodotus provides accounts of two remarkable Greek voyages to Tartessos. The first was led by a ship’s captain named Kolaios, who sailed from Samos between 650 and 638
BC. Kolaios returned from metal-rich Tartessos with silver worth 60 talents, a tenth
part of which was spent to commission an enormous bronze ritual vessel (4.152).
The second expedition came from Phokaia, when this Greek state was under
threat from the Medes of western Iran during the 550s BC.
§1.163 [The] Phokaians were the earliest of the Greeks to make long sea voyages:
it was they who discovered the Adriatic Sea, and Tyrrhenia, and Iberia, and
Tartessos . . . When they came to Tartessos they made friends with the king of
the Tartessians, whose name was Arganthonios
[trans. Godley].
Arganthonios is a transparently Celtic name or title, meaning something like ‘agent
of divine silver’ (*Arianhonydd if the name existed in Welsh today). Arganthonios is,
in fact, the only clearly Celtic personal appellation in all of Herodotus’s Histories.
The basis of the fabled wealth of Tartessos was metals, silver most especially, but
also gold and copper, and tin transhipped from Galicia, Brittany, and Cornwall. It
was need of great quantities of silver, demanded as tribute by the Assyrians, that
had impelled sailors from Tyre in what is now Lebanon to Tartessos.
According to the Roman historian Velleius Paterculus, the Tyrians founded Gadir
80 years after the Trojan war, or about 1100 BC [see p. 16]. Phoenician metalwork
occurs together with Atlantic Late Bronze Age types in the Huelva deposition of
about 950 BC. But the Phoenician colony of Gadir is not archaeologically detectable until about 770 BC, early in the Tartessian Orientalizing Phase of the Iberian First Iron Age. In 573 BC, Babylon conquered Tyre, and there was afterwards a
downturn of eastern luxuries reaching the Tartessian aristocracy, which explains
Arganthonios’s eagerness, about 20 years later, for the Phokaians to found a colony
‘anywhere they liked’ in Tartessos.
As well as the etymology, a further indication that Arganthonios ‘Agent of Divine
Silver’ might be a title or office rather than a mere name is Herodotus’s improbable
statement that Arganthonios ruled Tartessos for 80 years. Since the Phokaians’
visit came near the end of this long reign, one possible explanation is that the
Tartessian potentate who had enriched Kolaios 80 years before was also called
‘Agent of Divine Silver’ and Herodotus assumed that this was the same Arganthonios.
Note that A r g a n t o d a n n o s, meaning ‘silver minister’, was a Gaulish title and
recurs on pre-Roman silver coinage.
For the present subject, the most significant import from the east was alphabetic
writing. The first written language in western Europe occurs on roughly 75 inscribed stones concentrated in south Portugal and a further 15 from south-west Spain. This script and language are sometimes called simply ‘south-western’, referring
to their location in the Peninsula, though that name also suits their situation
in Europe overall. Alternatively, they are lately often called ‘Tartessian’, which is
the correct historical and archaeological identification if we understand Tartessos
to refer to a sizeable region rather than narrowly to that civilization’s chief protourban
concentration at Huelva. There are closely datable contexts for a few of the
Tartessian inscriptions. However, many are associated with necropolises of the
Iberian First Iron Age, about 800–550 BC. Some scholars see an admixture of Greek
orthographic principles behind the Tartessian script, but the most obvious primary
source was the Phoenician alphabet, and in some letter forms, specifically a version
of the Phoenician alphabet in use about 825 BC. It is also relevant to the questions
of origins and dating that the Tartessian inscribed stones continue a Late Bronze
Age tradition of pre-literate funerary stelae depicting weapons, armed warriors,
and sometimes wheeled vehicles, dancing figures, and lyre-like musical instruments.
Since the 1990s, the sounds represented by most of the letters of the Tartessian
alphabet have been known. We don’t have enough time to go into all the particulars
of the writing system, but one key detail is that the voiceless and voiced stop
consonants—t and d, k and g—are not distinguished. A second peculiarity is that
the stop consonants have different forms depending on what vowel follows—thus
—even though the redundant vowel is written afterwards:
[the consonant-vowel pairs are printed here in their more common
right-to-left arrangement]. In other words, the system itself implies consonant
qualities broadly reminiscent of the sound system of Goidelic.
José Correa, a classical philologist from Seville, had proposed that the inscriptions
contained Celtic proper names and titles. No one until now has unreservedly
agreed with this idea, and Correa now regards Tartessian as an unclassified language.
In the light of the early references to Keltoí in the south-west and the Celtic names
Kuneetes and Arganthonios, as well as many further Ancient Celtic place- and group
names recorded in the region in Roman times, I decided that the possibility that
the language of the Tartessian inscriptions was Celtic was worth re-examining. At
present, the established list of Ancient Celtic languages is as follows:
• Lepontic (the oldest, attested in northern Italy from about 500 BC),
• Celtiberian in east-central Spain,
• Gaulish in France and central Europe,
• Goidelic (the ancestor of Scottish Gaelic and Irish),
• Brittonic or British (the ancestor of Welsh, Breton, and Cornish),
• and Galatian (in the east, about Ankara in central Asia Minor).
Having studied the inscriptions for about a year now, I think that the Tartessian
language should be added to this list. It is therefore the first attested of the Celtic
languages. It is also about 2,000 kilometres west of Hallstatt and 1,500 from La
Tène. Tartessian has affinities with Celtic names attested in Galicia as well as with
Celtiberian—as might be expected—but also with Gaulish and even Gaelic and
Brittonic. It is the longer, most complete, and best preserved inscriptions where
Celticity is most apparent. And most of the elements—names, common nouns,
verbs, and prepositional preverbs—are not rare but occur in the core vocabulary of
one or more of the other Celtic languages.
(TO BE CONTINUED)