The Atlantic Celts: Ancient People or Modern Invention? 
by Simon James.
British Museum, 160 pp., £6.99, March 1999, 0 7141 2165 7
Show More
Show More

Celtic Studies, or Celticism, has rarely been disinterested. In 1884 one James Cruikshank Roger published Celticism: A Myth. The title was ‘intended to express’ his ‘conviction that the assumption of Celtic civilisation and Celtic art is utterly without foundation’. Encouraged by the press reception, Roger showed his hand in the second edition: ‘Let him who will, deduce his origin from the shiftless savage of the British isles, I am content to believe myself of that great Teutonic stock, which has ruled the world in the past, and will rule it to the end of time.’ More sympathetic studies of the Celts, many of them inspired by Matthew Arnold, were actually more insidious, forcing Irish, Scottish, and Welsh culture into an untraditional and externally imposed category. As W.J. McCormack argues in his Ascendancy and Tradition in Anglo-Irish Literary History (1985), this is a characteristic which ‘Celticism’ shared with ‘Orientalism’. McCormack concedes that ‘There is no doubt that scholarship in the area of Celticism constitutes perhaps the most refined and intellectually demanding exercise of the academic mind, but this should not blind us to the specific ideological and historic origins and affiliations of the discipline.’ Practitioners of Celtic Studies with an English background like myself are obliged to note McCormack’s final clause, not just his flattering preamble.

The latest phase in the debate about the meaning of the word ‘Celtic’ has been going on for twenty years, but no one has been as successful as Simon James in bringing it before the general public. The Atlantic Celts: Ancient People or Modern Invention? has been greeted by the press in England with headlines such as ‘Celts Were “Really Just a Scotch Myth” ’ (the Independent) and ‘The Tribe that Never Was’ (the Guardian). This attention was invited. The British Museum Press wanted a book ‘to coincide with the first elections to the new Scottish Parliament and Welsh Assembly, which provides an opportune moment to raise some issues about insular Celtic identity’. James, an Iron Age and Roman archaeologist, was the natural choice to give the view from London: he defines his own identity as ‘Westerner, European, Briton, Englishman, Southerner (by birth), Londoner (by family origin and long residence)’. His book gives a good insight into such a person’s difficulties in seeing the Celtic point of view. Why is it that English scholars can show extraordinary sympathy for Greek, Roman, Egyptian or Aztec culture, and yet be so suspicious of the Celts on their doorstep? The unpalatable answer is blindingly obvious to everyone but themselves.

Colonialism, as Siân Jones explained in Archaeology of Ethnicity (1997), commonly involves ‘the critical scrutiny of a minority group’s identity and history by the dominant society, rather than vice versa, ultimately perpetuating the relations of power between groups’. It would be odd if English and French study of the Celtic-speaking peoples of Britain, Ireland and Brittany were somehow a benign exception – especially as the negative tone had already been set by classical writers. The Greeks and Romans were naturally terrified by the arrival of the Celts at the gates of Delphi and Rome, and exacted their revenge by leaving Celtic culture ‘outside the horizon of the traditional civilised Western world’ (as Momigliano had it). The core-periphery model, which puts the Celts permanently on the ‘fringe’ of someone else’s centre (Athens, Rome, Paris, London), started with the Greeks. Greek writers, such as Ephorus in the fourth century BC, placed themselves in the centre of the world, and assigned the peripheries, in clockwise order, to Celts, Scythians, Indians and Ethiopians. The beauty of such a model is that it puts the ‘fringe’ in a permanent no-win situation. In Pro Fonteio Cicero laughed at the Gauls for wearing barbarous trousers instead of civilised tunics. By 1747 the Highlanders had got their act together, from Cicero’s point of view; unfortunately, however, trousers were now the ‘civilised’ fashion and the ‘barbarous’ kilts had to be banned by Act of Parliament.

The next move for the ‘fringe’ people will be either to conform (i.e. buy some trousers), conceding that they may not be so different in other ways either, or else to fetishise their distinctiveness as perceived from the centre. In the latter case, the revenge of the centre is sweet: send in the debunkers; Hugh Trevor-Roper, who has a selective nose for fakes, got the kilt job in The Invention of Tradition (1983). He made much of the difference between the authentic ‘plaid’, starting from the shoulders, and the modern ‘kilt’, starting from the waist. This ‘sound, if not entirely novel’, argument overlooked the obvious similarity between the two garments, as Malcolm Chapman has commented; besides, ‘recent social anthropological approaches to the “infinite sequence of rememorisations” by which the present appropriates the past, are at once more subtle, and more general, than any “invention of tradition” directed at obvious and easy targets.’

Siân Jones notes that ‘minority groups are subjected to a relentless discourse which requires them, in one form or another, to possess a traditional homogeneous culture and identity stretching back in a continuous and unilinear fashion into the past. Many will inevitably fail such a requirement.’ While most peoples, including the modern Greeks, have had some difficulty in demonstrating long-term continuity, it would generally be thought that the Celts fare relatively well. According to James, however, the island ‘Celts’ of Britain and Ireland fall at the first fence. First, their culture differs from that of the continental Celts mentioned by the classical writers, and second, they only became Celts thanks to an 18th-century mistake: ‘the insular Ancient Celts never existed.’ Both these points are familiar from Chapman’s The Celts: The Construction of a Myth (1992), as well as from works unknown to James such as Terence Brown’s Celticism (1996), but James gives them a new polemical slant. The facts, as opposed to the interpretation, seem to be as follows.

Classical writers from the time of Herodotus referred to ‘Celts’ in Western and North-Western Europe. Some writers distinguished them from other ethnic groups in the region such as Ligurians and Cynesians, while others, like Ephorus, used ‘Celts’ as a blanket term for the whole area, by implication including Britain and Ireland (pace James). As the complexity of the ethnic situation became clearer, the term ‘Celtic’ – and its apparent synonyms, ‘Gallic’ and ‘Galatian’ – came to be confined for the most part to peoples whom we also label ‘Celtic’ on the basis of the language of their names and inscriptions – important evidence unmentioned by James. There were groups of such Celts from Iberia to Asia Minor, but the greatest concentration known to classical writers was in Gaul. In fact, Julius Caesar found that a large group of Gauls actually ‘called themselves Celtae’. A corollary of this was that ancient geographers distinguished between Celtica, meaning the Continental mainland, and the various offshore islands such as Britain and Ireland. Thus in the fourth century BC, Pytheas spoke of sailing south from Kent to Celtica, and three centuries later Strabo compared the customs and appearance of the Britons and Irish with those of the ‘Celts’, implying that the former were not to be labelled as ‘Celts’. This usage survives in Paradise Lost: Milton distinguishes between the ‘Celtic’ fields and the ‘utmost isles’.

So Britain and Ireland were perceived as distinct from Celtica in the same way that, say, Ceylon was always seen as distinct from India. From this basically geographical viewpoint, the inhabitants of Britain and Ireland were not Celts like the Gauls, any more than Sri Lankans are Indians. On the other hand, there were clearly plenty of linguistic and cultural connections, some of them noted by Caesar, Tacitus and others, between Gaul and Britain (and Ireland, to the extent that it was known). Does ‘Celtic’ not remain a convenient label for these, at least as valid as most such labels (‘Germanic’ for instance, which is applied to many peoples who never regarded themselves as Germani)?

Here the problems of interpretation begin. James writes as a ‘settlement’ archaeologist, and attaches more importance to the round (versus rectangular) shape of the Britons’ houses than to the fact that they spoke virtually the same language as the Gauls, had a comparable range of place, tribal and personal names, practised the same druidical religion, and even shared a similar range of motifs in the high art of the élite (here regarded as an ‘aristocratic package’ irrelevant to the ‘real’ peasant population). James’s point of view may be relevant within settlement archaeology, but it is barely relevant to his main polemic. I have yet to meet anyone whose ethnic identity was comprised in the groundplan of his house. On the other hand, language, about which James is resolutely ignorant, is as important to modern Celts as it is to modern Greeks. It is glib to argue that ‘Celtic identity is not dependent on speaking a Celtic language, otherwise millions of people today who regard themselves as culturally or ethnically Celtic are disqualified.’ Surely, the vast majority of those millions would claim to have fairly recent ancestors who spoke one of the relevant languages?

The importance of language in ‘Celticity’ reflects the re-emergence of the term in 18th-century linguistics in connection with the discovery that ancient British (the ancestor of Welsh, Cornish and Breton) and ancient Gaulish, which had been known to be related since the time of Tacitus, were also related to the Gaelic language of Ireland and Scotland. The first convincing demonstration of this fact was made by the polymath Edward Lhuyd, of the Ashmolean Museum. Lhuyd published his linguistic material in his stupendous Archaeologia Britannica in 1707. James comments that ‘it is, at the very least, a remarkably fortuitous coincidence that the concept of Celtic speakers, and by implication ethnic Celts, in Britain, Brittany and Ireland was first published in 1707. That was the very year that the Treaty of Union between England and Scotland saw the creation of a new political identity called “British”.’ I would agree with James that ‘Celtic’ (sparingly used by Lhuyd) eventually came to provide a handy umbrella-term for those threatened by Englishness; no such term had been available before, despite 20th-century historians’ tendency to describe various medieval attempts to ally the Welsh and Scots against the English as ‘Celtic alliances’. I would disagree, however, with James’s idea that Lhuyd himself ‘had a political agenda’, Welsh patriot though he was. His great work had been many years in the making in 1707, to the chagrin of the subscribers, and the basic idea of aligning Irish with the Gaulish-British family had already been advanced by Leibniz in his Celtica ad Jobum Ludolfum in 1699 and elsewhere, with full credit to ‘Lloydius’. We are not dealing with a political conspiracy, but a correct linguistic observation by two great scientists.

Modern philologists and Simon James agree that Lhuyd got it right: the kinship of the Irish and British languages, forgotten during the Middle Ages (so far as we know), is genuine. Since these Insular languages are clearly related to ancient Continental languages which were called ‘Celtic’ in ancient and medieval times, it is convenient to call the whole family ‘Celtic’ and its speakers ‘Celtic-speakers’ (in the same way as the Scandinavian languages are labelled ‘Germanic’, and so on). The real trouble started after Lhuyd’s day, when people began to talk of Celtic-speakers as ‘Celts’ for short, and to associate them rightly or (often) wrongly with various archaeological manifestations such as megaliths, La Tène art etc, with various stereotyped literary conventions (Matthew Arnold and his unacknowledged mentor Ernest Renan were the chief culprits), and even with certain biological characteristics – ‘Celtic’ genes for phenylketonuria, cystic fibrosis and HLAA1, the B8 tissue antigen being the current fashion. Lhuyd cannot be held responsible for these accretions.

When all the accretions are stripped away, we seem to be left with an ancient people – ‘Atlantic Celts’, if you will – definable principally by a distinctive dialect of Proto-Indo-European, by whatever socio-historical factors resulted in them speaking that language, and by whatever cultural interactions the shared language made possible. We simply do not know whether they had a common ethnonym for themselves or even any sense of community. James is doubtful whether ethnic identities beyond the kin-group were possible in the Iron Age – ‘we should not speak of the “Ancient Irish” or the “Ancient Britons”, let alone insular “Celts”,’ he warns; but there is no firm evidence either way. In other words, the Celts are not very different from other early Northern European groups on the fringes of the documented, literate world. From our modern standpoint, their identity seems stronger than that of the Germans but weaker than that of the Hellenes, but this may have more to do with the spread of literacy and hence documentation than with the social dynamics of the cases in question.

Simon James has succeeded in hanging a question-mark over glib popular generalisations about Celts, and this will be salutary for readers who have not seen Malcolm Chapman’s two books or Terence Brown’s important collection, as well as those (like James himself) who are unfamiliar with the work of Celtic Studies departments. But does all this ancient history have ‘profound political implications’ for the devolution issue? Hardly. Take Wales: the subsidiary historical case for devolution (as opposed to the major democratic one) is based on priority of arrival. ‘And how are you finding it here?’ the Welsh farmer replied to the gentleman who claimed his family ‘arrived with William the Conqueror’. The more distant Celtic kinship with the Irish is no more or less relevant to Welsh identity than Germanic is to English. In politics Celtic is a more important label at the ‘centre’ (in this case London) than at the peripheries.

James’s book may help to discourage simplistic generalisation, but it does a disservice by underestimating the defining role of language in the study of Celticity, ancient and modern, and by overestimating the importance of archaeology. As James himself writes: ‘Typically, human “cultures” are not sharply bounded and neither are they usually homogeneous internally.’ It is hardly surprising or significant, then, that the ‘Celts’ fail to make the grade as a discrete archaeological culture. That is not particularly disconcerting if one takes a long view. James also allows that ‘the assumption that Ancient Celts had existed in Britain and Ireland before the Romans was already taken for granted before anyone had discovered how to identify with confidence traces belonging to the period.’ In other words, the assumption of Celticity was originally based on linguistic and historical evidence. The idea that it can be proved or disproved archaeologically rests on dubious epistemological foundations.

Send Letters To:

The Editor
London Review of Books,
28 Little Russell Street
London, WC1A 2HN

Please include name, address, and a telephone number.

Read anywhere with the London Review of Books app, available now from the App Store for Apple devices, Google Play for Android devices and Amazon for your Kindle Fire.

Sign up to our newsletter

For highlights from the latest issue, our archive and the blog, as well as news, events and exclusive promotions.

Newsletter Preferences