The Origins of the British: The New Prehistory of Britain (9 page)

Having at last taken this decision, Collis makes a magnificent job of debunking 150 years of archaeological conviction-scholarship. As I mentioned above, James gives us three timorous lines of politely concealed incredulity, obliquely referring to unspecified details of ‘time and place’ as orthodox evidence for Celtic origins north of the Alps.
⁵⁸
By contrast, the more senior author Collis devotes three and a half chapters out of eleven to deconstructing such evidence, naming and shaming the authorities responsible for this academic castle of cards.

I shall sketch this part of Collis’s book here, while recommending the reader to his more thorough treatment. After
cover ing the classical writers and early linguists, Collis enigmatically entitles his fourth chapter ‘Race and time’. ‘Time’ here refers to the realization in the nineteenth century of the vast time depth in prehistory, in contrast to Bishop Ussher’s rather shorter Biblical chronology, commencing in 4004
BC
. Collis uses the term ‘race’ to introduce early concepts, growing during the eighteenth and nineteenth centuries, of racial differentiation and a new sense of nations, all mixed up with linguistics and comparative craniometry (head-measuring).

Collis quotes one anatomist, the otherwise brilliant Paul Broca (who gave his name to the speech area in our brains), confidently stating in 1864 that the Celts had arrived in central Gaul introducing the ‘more civilized’ Bronze Age to Europe from Asia in the East. Broca associated this event with a hypothetical change in head shape in the population. His strident picture of racial migration takes on a surreal tint, however, when he admitted that opinions varied as to whether ‘broad-heads’ replaced ‘long-heads’ or vice versa.
⁵⁹

Against this background of early and distinctly flaky anthropology, elements of which were later to be developed and used by the Nazis to justify the Holocaust, the new ‘Celts’ were conceived and underwent a strange gestation. During the eighteenth and nineteenth centuries the term ‘Celtic’, unearthed initially by linguists, became re-established, after a gap of over a thousand years, as a reincarnated ethnic label – although clearly not meaning quite the same as could be inferred through the smoky window of the classical texts.

John Collis contrasts the views of mainly French authors of the nineteenth century on the origins of the Celts. The most conservative of these, Amédée Thierry, rather wisely stuck to
the classical sources such as Livy, Caesar and Strabo, coming up unsurprisingly with Caesar’s Middle Gaul as the original location of Gauls. In his
Histoire des Gaulois
(1827), Thierry gives the location of the Celtic homeland as south of the Seine and the Marne and well west of the Rhine, missing out most of the south-west third of France and the pre-existing Roman province of Narbonensis, presumably because of Caesar’s vagueness farther south.
⁶⁰
But Thierry argued for a Gallic presence in France from at least 4,000 years ago, with early invasions of Spain some time between 3,700 and 3,500 years ago, and of northern Italy by 3,400 years ago, based on some interesting interpretation of certain texts.

So far, so good. Thierry gives a text-based reconstruction that could be challenged rationally; but in the year Thierry published his book, another Frenchman, Henri d’Arbois de Jubainville, was born who subsequently had a profound effect on the nature of recon structions of ‘Celtic prehistory’, replacing reason with conviction.
⁶¹

D’Arbois de Jubainville published books and papers in the last quarter of the nineteenth century and into the twentieth, but somehow managed to infuse the rest of the twentieth century with his views on the issue of the Celts. His chronology was based mainly, but loosely, on the classics rather than the new archaeological interpretations; but his reconstructions were altogether more creative than either, accepting mythological statements in parallel with those of classical historians. He also made extensive but selective use of linguistics. D’Arbois de Jubainville saw four phases of colonization, or ‘empires’. The first (undated) consisted of hunters and herders living in caves. The second empire was Iberian, dating from 6000
BC
,
with hunters and herders from Atlantis, speaking a non-Indo-European language, forced by the disappearance of their land to colonize new territory (i.e. Spain). He dated the third of these exotic empires at 1500
BC
, formed by Ligurians (à la Himilco) as the first of the Indo-European speakers, bringing agriculture and colonizing from the East, thereby displacing the Iberians. The fourth empire was embodied by the Celts arriving from southern Germany in 500
BC
, representing the second wave of Indo-European-speakers from the East.

The first three of d’Arbois de Jubainville’s mythical empires seem to have been conveniently forgotten or dismissed, because, as Collis says, they were ‘fanciful conjecture’. But this makes it all the more surprising that the fourth has survived in name, date and location into the twenty-first century, let alone as the orthodox paradigm of Celtic origins, since the way he abused the evidence from the classics in its creation was equally fanciful.

Collis appears to ‘give face’ to d’Arbois de Jubainville for the fourth empire by saying that the reasons for his belief in the late arrival of Celts ‘need more careful consideration’. In this gentlemanly faint-defence Collis cites the Frenchman’s arguments: ‘Firstly … the evidence of the introduction of Indo-European languages, the assumption of their easterly origin, and their relatively late appearance in western Europe … secondly … claims [of] evidence for population change in the historical record’.
⁶²
But when we find that the sources for these arguments are taken from Herodotus and the
Periplus
of Himilco as quoted by Avenius, Collis’s cover as apologist is blown.

As we have seen, those tantalizing classical fragments are insufficient to form any but the flimsiest of hypotheses about the origins and dates of classical Celts or any other West European tribes, let
alone their languages. And then we hear that this evidence was what d’Arbois de Jubainville used to construct the story ‘that the south of France was taken over by the Gauls around 300
BC
’ from Central Europe
⁶³
– in the face of all the classical sources that give a much earlier date (see above). We may wonder if Collis is being tongue in cheek or just kind to his colleagues, who were all taken in by it for the next hundred years.

After this, Collis’s real sceptical stance becomes apparent. Still citing d’Arbois de Jubainville, he continues: ‘For the eastern origin of the Celts, he uses Herodotus’ assertion about the Danube rising in the territory of the Celts [e.g. in Central Europe] as well as Polybius’ statement that the Gauls invaded northern Italy from the other side of the Alps [e.g. from the north].’
⁶⁴
As I had guessed when reading these texts myself, Herodotus’ misunderstanding of the source the Danube is what underpins the whole house of cards of the German homeland theory. It is further used to prop up an indef ensible interpretation of Polybius, since ‘the other side of the Alps’ clearly refers to France and not to Germany. There is more discussion of d’Arbois de Jubainville’s distortion of his other sources, and some description of French archaeologists who carried this torch of Celtic origins ultimately through to this century, but Collis’s message is clear.

In his fifth chapter, ‘Art and archaeology’, Collis explains how this basic misunderstanding of classical sources was then combined with the exciting new nineteenth-century finds of exotic jewellery and weaponry from Hallstatt and La Tène in Austria and Switzerland to create the cultural-archaeological picture of the Celts we have today. The unjustified assumption of origins of Celts (whoever they were) in Central Europe,
and their late expansions into the rest of Europe from there, led to further unjustified assumptions of cultural asso ciations, followed by inferences about migration and ethnic replacement. British archaeologists then come under the torch for conflating the glorious art and precious metalwork of Iron Age Britain (Plate 2) and Ireland with this view of invading Celts in 300
BC
.

At the same time, Collis recapitulates the whole history of western archaeology in a revealing and cathartic way. At the same time as these Romantic ‘evolutionary’ notions of human ethnicity were developing, the principles of modern archaeology with its own folklore were being established. The concept of three stages of ‘civilization’ – the ‘three Ages’ of Stone, Bronze and Iron – was first conceived in 1823.
⁶⁵
In 1863, an earlier Stone Age, or Palaeolithic, represented in the caves and river gravels of France and the more southerly part of Europe, was distinguished from the more recent or Neolithic Stone Age remains found throughout Europe.
⁶⁶
These crude chronological divisions were reinforced by stone tool typology, an approach that lent itself to relative chronology. Chronology in those pre-carbon-14 days also depended on context, known historic sites and battles, dated coins and pot typology, and on careful stratigraphy and type comparison between sites.

Like most people, I find it quite amazing that so much useful information was inferred in the past – and still can be inferred – by using these practical, reason-based, pre-hi-tech methods. That is what makes field archaeology still such compulsive television viewing today, much more exciting than the whizbang special effects used by film producers to imply technical wizardry in their Palaeo-documentaries. We all have a bit of the tracker in us.

Worthy though Collis’s charting of the growth of knowledge and ingenuity among archaeologists may be, it remains an inadequate veil to obscure his colleagues’ persistence in sticking to d’Arbois de Jubainville’s particular myth-structure of Celtic origins. Collis seems to paint a Swiftian picture of the dons of his profession as erudite, supremely articulate historian-scholars. Archaeologists, in spite of their more practical leanings, which dominate most of their day-to-day work, would seem to create their grand reconstructions in cycles of sudden leaps of conviction and group-persuasion, followed by detailed consolidation and prolonged stasis. On top of these cycles of theory are imposed further cycles of fashion and process. Fads such as migrationism (enthusiasm for movements of people) and diffusionism (enthusiasm for movements of culture) move in and out of fashion, irrespective of the weight of evidence for or against particular events. There is now a school of what is called processual archaeology, which concentrates less on cultural history and more on processes such as the way humans do things (behaviour) or the way things decay. Then there is post-processualism … However, in spite of Collis’s criticisms, their heavy load of other ‘isms’ and a false start on Celts, archaeologists would seem to the observer to have achieved the most extraordinary advances in European prehistory over the last hundred years.

Geneticists also come in for a deserved pasting, with well-aimed punches from Collis in his frenzy of debunking. He derides both the research in the context of the ‘Celtic’ question and papers pub lished by geneticists seeking to ‘prove’ the later Anglo-Saxon replacement.
⁶⁷
I hope to resuscitate some of the relevance of genetics from such onslaught later on.

*
To avoid having endlessly to repeat the distinction between, on the one hand, the terms ‘Celt’, ‘Celtic’, ‘Celtae’ and ‘Keltoi’ as used by whomever and however loosely, and, on the other hand, the modern definition of ‘celtic languages’, from here on I shall use lower-case ‘c’ for the celtic languages and a capital for everything else Celtic (except in passages quoted from other sources). The reason for this particular distinction is the sceptics’ doubt that Celt and celtic languages have any solid or meaningful connection.

Having demolished the evidence for a Central European homeland and any specific association of Celts with the Hallstatt and La Tène cultures, John Collis moves back briefly, in the penultimate chapter of his Celto-sceptic book, to language. Here he mainly bemoans the unhealthily close relationship between linguists and archaeologists in the nineteenth century, which has persisted in some quarters until today. Along with the obvious problem that ancient artefacts without writing do not identify the language of the maker or wearer, there is the tendency to force both language and culture into similar monolithic racial stereotypes, where the reality is of diversity, difference and mixture.

The twelve conclusions that Collis reaches in his last chapter are mainly deconstructions of the struts of the modern Celtic
myth, but the only hint of reconstruction, albeit half-hearted, comes from his seventh conclusion, which refers to language:

One interpretation of the historical and linguistic evidence also seeks the origin of the Celts in south-west Germany, but other interpretations of the classical sources are also possible, indeed perhaps more likely, and would include central and western France.
¹

This brings us conveniently back to the point in the last chapter at which I digressed into archaeological homeland myths. I had just referred to Caesar’s comment that the people he called ‘Celts’ in Middle Gaul called themselves Celts ‘in their own language’ (
Figure 2.1a
). Since Caesar’s assertion contradicts those who claim that Celts did not use the term for themselves, and broaches the whole question of the identity of celtic languages, it might be worth asking whether there is any evidence for the linguistic affiliation and identity of the language Caesar calls ‘Celtic’. In other words, is there a systematic record from Roman times, of a dominant, indigenous non-Latin language, in Caesar’s Celtic Gaul, which could be related or linked to modern insular-celtic languages?