Lauren Hutton/PA Images/Getty Images

Leo Varadkar, the prime minister of Ireland, and his partner, Matthew Barrett, at an LGBTQ Pride festival, Dublin, June 2017

In June 2018, the Irish prime minister, Leo Varadkar, made a formal apology to men who had been convicted under laws against homosexual acts that were finally repealed in 1993. In the course of his speech in parliament, he noted that “Aristotle wrote that the Celts openly approved of same-sex relationships.” The implication was that homophobic laws were not merely wrong, they were un-Celtic, and thus especially inexcusable in that most Celtic of countries, Ireland.

Varadkar was tapping into an assumption that has been common since the mid-nineteenth century. It is taken for granted not only that the Irish (along with the Scots, the Welsh, the Bretons, and the Galicians of northern Spain) are Celts, but that this Celticity carries with it a profound sense of difference, a sensibility fundamentally at odds with classical rationality. Aristotle, in the Politics, does indeed underline this sense of difference when, writing about female-dominated regimes, he adds, in parenthesis, “excluding the Celts and any others that openly honor sexual relations among males.” It is a lovely notion that the people of whom Aristotle was aware, the Central European tribes the Greeks called keltoi, are the forebears of today’s inhabitants of the Atlantic fringes of Europe—all the more so when these racial ancestors give their blessing across the millennia to contemporary Ireland’s enthusiastic embrace of liberal values.

But these Atlantic Celts are an oddly slippery people. The superbly refurbished Neues Museum in Berlin, which opened its doors in 2009 and surely reflects the most current views of early European history, is a case in point. There is an obvious anomaly in its depiction of the ethnic makeup of the islands of Britain and Ireland during and after the fall of the western Roman Empire. On the map showing the movement of the Germanic peoples in the fourth and fifth centuries of the Christian era, most of England is inhabited by Angeln and Sachsen—Angles and Saxons. Scotland, Cumbria, Wales, Cornwall, the Isle of Man, and Ireland are colored green and marked Kelten—Celts. However, the next map, showing Europe around 800, has Ireland inhabited by Iren (the Irish), Scotland by Pikten (Picts), and western England and Wales by Briten (British). So what happened to those occidental Celts in those four hundred years? How did they become merely Irish, Pictish, and British? Since we know of no great cataclysm that hit the western regions of the Atlantic archipelago in those centuries, the even more awkward question arises: Were the Celts ever there in the first place?

If we look to the commercial marketplace and to popular culture for an answer, they will tell us emphatically that Ireland, Scotland, and Wales have always been the Celtic homelands, and still are. The Irish economic boom of 1998 to 2008 (before its brand was tarnished) was the Celtic Tiger. New Age Celtic spirituality rubs up against Celtic Christianity. You can light your nuptials with Celtic candles and seal them with Celtic wedding rings. You can wash yourself with Celtic herbal soap and dry yourself with Celtic linen. You can sip Celtic whiskey under Celtic art posters illustrating Celtic myths while listening to Celtic music (performed by groups like the Celtic Tenors and Celtic Woman). You can record your consequent insights in Celtic leather–bound journals (“for a Celtic writing experience”). You can support the Boston Celtics or Glasgow Celtic or the Galician soccer team Celta Vigo. You (or your pets) can be buried under Celtic crosses. The term is now so elastic that it can extend from the spaced-out Celtic sounds of Enya to the testosterone-fueled antics of the mixed martial arts star Conor McGregor.

As Norman Davies put it in 1999 in his pioneering The Isles: A History, Celticity “appeals to all those people who feel the strains of modern civilization, and who seek, however impractically, to recover the benefits of the world before civilization.” Yet this free-floating signifier is anchored in an assurance of authenticity. “Celtic” means Irish, Scottish, and Welsh (with an occasional nod to Brittany and Galicia). Why spoil it with the dark secret revealed by scholars in recent decades that the Celtic identity of the Irish, Scots, and Welsh is a very recent invention?

According to the nineteenth-century model that still has a grip on popular accounts of the early history of the islands, Britain and Ireland were invaded by Celts from continental Europe sometime between 500 and 300 BC. These invaders conquered the indigenous populations, who then disappeared from history. The Celts brought with them the technologies that shifted the islands from the Bronze Age to the Iron Age, the distinctive artistic styles that blossomed in the Atlantic air, and the language that eventually became the Gaelic of Ireland (and later of Scotland and the Isle of Man) and the Brythonic that split from it and survives now as modern Welsh, Cornish, and Breton.


The Celts did exist, at least as a large-scale cultural nexus in continental Europe. From around 500 BC, beginning with the work of the Greek geographer Hecataeus of Miletos, there is a notion that the area roughly equivalent to France is inhabited by the keltoi, a people who were also recorded as sacking Delphi in 279 BC. There is early knowledge of the existence further west of an archipelago generally called the “Tin Islands.” By around 320 BC, the explorer and scientist Pytheas, who seems to have sailed through the Irish Sea, could distinguish the islands of Britain and Ireland lying in the great ocean to the west of the Celtic lands of France. The problem is that neither he nor any subsequent classical writer suggested that the inhabitants of these islands were themselves Celts. (Tacitus, for example, reckoned that the inhabitants of Scotland must be Germans.)

Crucially, neither did these people themselves: neither Old Irish nor Old Welsh has any word equivalent to “Celt” or “Celtic.” Julius Caesar, writing in the mid-first century BC, tells us that the Gaulish tribes living between the Seine and Garonne rivers did specifically call themselves Celts, so this kind of self-identification must have been possible. In the first century AD, the Roman satirist Martial, born in central Spain, proudly claimed to be descended from both Iberian and Celtic stock, so it was equally possible for individuals in Western Europe to claim a Celtic identity. But there is no evidence that the inhabitants of the Atlantic islands ever did so. In fact, right up to the modern period, the Irish—now the most “Celtic” of peoples—went to literal great lengths to avoid any such identity: their medieval pseudohistories insisted that they were Iberians in origin, or even Scythians.

What is not in doubt is that the Gaelic language of Ireland and Scotland is very closely related to Welsh, Cornish, and Breton, and that all of them are in turn related to the group of languages that were once widespread in Atlantic Europe: Gaulish in France, Lepontic in northern Italy, Hispano-Celtic languages in Iberia, even Galatian as far east as Turkey. The ancestors of the Irish and Welsh tongues may have been synthesized from several different sources, including much older indigenous vernaculars. But quite how the “Celtic” languages of Britain and Ireland developed remains a matter of speculation.

The distinguished American Celtic scholar John Koch posits that a Celtic language reached Ireland in the period 1200–600 BC—much earlier than any supposed invasion of Celts. Barry Cunliffe argues in Europe Between the Oceans (2008) that Celtic emerged in the Bronze Age as a “lingua franca spoken along the Atlantic seaways” that connected western Iberia, western France, Britain, and Ireland. This would imply a revolutionary shift of perspective, with Celtic languages subsequently spreading from west to east rather than, as has always been imagined, from east to west. It is not easy to imagine how a trader’s and sailor’s argot could be adopted by entire populations. Nor do we really know when these Celtic languages were thoroughly adopted in the Atlantic islands—native forms of writing did not emerge until the fifth century AD, so the written record is of little help.

What seems clear, however, is what did not happen: there was no Celtic invasion, no large-scale movement of people from continental Europe who imposed their language and culture on the existing peoples of Britain and Ireland. DNA evidence shows no significant migration of people into the islands during the Iron Age. The archaeological record contains nothing of what would be expected if one people had displaced another: new building methods, new domestic objects, new burial practices. The humble, utilitarian evidence left by new peoples—fittings for horses and carts, pottery, belt buckles—has not been found.

What appears instead are high-end, often spectacular objects in the La Tène style (named after an archaeological site in Switzerland). This is what we think of as Celtic art: the fantastical abstraction of curving, flowing, twisting forms, of alternating symmetries and asymmetries, of a playfulness and vitality that are yet redolent of mystery and a sense of the sacred. There is no doubt that this style did arrive in the Atlantic islands around 300 BC or that it flourished, especially in Ireland, for many centuries after it declined in continental Europe. It even survived into Christian imagery, most notably in the astounding Book of Kells, created in Scotland and Ireland around 800 AD.

Even in Ireland, however, the blossoming of La Tène metalwork does not support the idea of a Celtic invasion. It seems obvious that traders and craftworkers brought the new forms from the outside world. But by far the most striking thing about the La Tène objects found in Ireland is that they are overwhelmingly of local manufacture. They are distinctive, insular adaptations of the Continental style to local tastes. Outside influence is certainly at work, and it would be surprising if migrants were not carrying that influence, but the overall impression is of a native aristocratic, religious, and warrior elite using the new forms to create exquisite expressions of status and power.


The vast majority of the natives almost certainly had no access to this high-end weaponry and art. Even for those who did, there is no evidence that it made them feel “Celtic.” It certainly didn’t make them Celts—any more than building a Bauhaus villa in California makes you German or driving a Lamborghini in New York makes you Italian. Indeed the paradox is that what made Scotland and Ireland seem “Celtic” in retrospect was the decline of the continental Celts—with the political and cultural dominance of Rome on the European mainland, the non-Romanized Atlantic zones took on a more idiosyncratic and archaic character. The common culture that had connected those zones to continental Europe now made their inhabitants stand out as barbarians on the fringes of the civilized world. But the people who lived there never thought about themselves in this way.

As a way of imagining a collective cultural and political identity, insular Celticism is essentially a phenomenon of the second half of the nineteenth century—the title of Caoimhín De Barra’s The Coming of the Celts, AD 1860 is provocatively witty but accurate. In his finely researched and lucidly written study, De Barra details the rise, ebb, and flow of the idea of a common Celtic identity linking Ireland and Wales. One of his most salient empirical findings is the enormous increase in the occurrence of the term “Celtic” in nineteenth-century newspapers. In Ireland it appeared an average of ten times per newspaper in the first twenty years of the century. By 1870–1879, this had risen to 291, and by 1890–1899, it had declined to 188. The Welsh newspapers used the word fifteen times each on average in the first decades of the century, rising to 130 times in 1890–1899. It is striking that the overall usage of the term is still relatively sparse, but clear nonetheless that it really enters public discourse in the Victorian era.

The idea itself had been around since George Buchanan’s The History of Scotland (1582), in which the Scottish scholar correctly identified the common origins of the Irish, Scottish, and Welsh languages and their connections to Gaulish. He drew the not unreasonable conclusion that these insular peoples therefore “appear… to have sprung from the Gauls.” Early-eighteenth-century scholars, notably the Breton abbot Paul-Yves Pezron and the Anglo-Welsh linguist Edward Lhuyd, revived and developed these insights, though neither explicitly described the insular peoples as Celts. The process of reasoning from Caesar’s evidence that some of the Gauls called themselves Celts to the application of the term to their westerly linguistic cousins was sporadic and slow—it took nearly three centuries to establish itself as accepted fact.

Even when it did so, the fashion for Celticism was a product of contradictory impulses, a strange mixture of aggrandizement and disparagement. On the one hand, there was French Romanticism. The Académie celtique was founded under Napoleon in Paris in 1804, not least because France’s imperial ambitions could be imagined as the revival of the ancient Celtic hegemony in Europe. (Napoleon was dubbed, among other things, “Emperor of the Celts.”) Ernest Renan, in his highly influential La Poésie des races celtiques (1854), fixed the idea of the Bretons, Highland Scots, Welsh, and Irish as the survivors of this once-dominant people and expressed wonder at “the peculiarity of this fact of an ancient race living, until our days and almost under our eyes, its own life in some obscure islands and peninsulas in the West… still faithful to its own tongue, to its own memories, to its own customs.” Traveling into this zone, Renan wrote, was like “entering on the subterranean strata of another world.” In societies undergoing the homogenizing processes of the creation of the industrial nation-state, the allure of this ancient and subterranean world was becoming powerful.

It was also, however, double-edged, for alongside Romanticism there was a Social Darwinist “scientific racism” in which being unevolved was not such a good thing. Matthew Arnold, in his On the Study of Celtic Literature (1867), resolved this tension by creating a kind of ethnic division of labor for the United Kingdom, with hard-headed, rational, Germanic Anglo-Saxons racially primed to build empires and run things and Celts providing a necessary emotional and spiritual balance. Arnold characterized the Celt as poetic, demonstrative, sentimental, and imaginative. He added, of course, that these commendable qualities made the Celts innately incapable of real achievement:

Balance, measure, and patience, these are the eternal conditions… of high success; and balance, measure, and patience are just what the Celt has never had. Even in the world of spiritual creation, he has never, in spite of his admirable gifts of quick perception and warm emotion, succeeded perfectly, because he never has had steadiness, patience, sanity enough to comply with the conditions under which alone can expression be perfectly given to the finest perceptions and emotions.

The late-nineteenth-century Celtic Revival in Ireland broadly accepted Arnold’s division of labor, even while it strove with great success to challenge his implication that Celts could dream dreams but only Saxons could rule. In the hands of its presiding genius, W.B. Yeats, Celtic difference was not just an argument for self-government, it was a way of saving the world from the tyranny of industrial rationalism. For Yeats, Renan’s subterranean Celtic stratum was a mineshaft into the collective unconscious. Yeats’s Celtic Twilight (the title of his 1893 collection of folkish tales) was later famously mocked by James Joyce in Finnegans Wake as the “cultic twalette.” Larry Doyle, Bernard Shaw’s alter ego in John Bull’s Other Island (1904), rages that “when people talk about the Celtic race, I feel as if I could burn down London.”

Erich Lessing/Art Resource

A Celtic warrior; detail of a terra-cotta temple frieze, Sassoferrato, Italy, early second century BC

But in spite of this skepticism, there was something irresistibly flattering for a small and subject nation in the belief not merely that its ancestors had once controlled much of Europe, but that this heritage made it uniquely qualified to prevent the descent of civilization into a cold and mechanical logic. (The Pan-Slavs in Russia and Eastern Europe had much the same beliefs, naturally, but they were conveniently far away.) For nationalists in Wales and Scotland, the prestige of the Irish literary and theatrical revival was an invitation to climb on board the colorful Celtic wagon.

Even so, the flowering of a Celtic identity on the Atlantic islands was a bit like the Celts themselves—immensely significant in principle but hard to pin down in reality. As De Barra shows in impressive detail, some Welsh intellectuals did indeed begin to identify themselves with the Irish as fellow Celts. But he also shows how unsteady this identification could be. Even Welsh speakers were often uncomfortable with the Irish—as nonconformist Protestants they tended to be suspicious of Irish Catholicism, and their own demands for Home Rule were generally tempered by a desire to preserve their place in the United Kingdom and the British Empire. Even at the height of Pan-Celtic enthusiasm, the supposed Celts tended to look to each other for different things, the Irish admiring the relatively vigorous health of the Welsh language in contrast to the rapid decline of their own Gaelic, and the Welsh hoping to piggyback on Irish political militancy to press their own demands for land reform and the disestablishment of the Anglican Church.

And as De Barra usefully reminds us, the greatest “Celtic” statesman of the age was David Lloyd George, the reforming Liberal chancellor of the exchequer and prime minister from 1916 to 1922. Lloyd George was a native speaker of Welsh (the only British prime minister to date whose first language was not English) and considerably more fluent in this Celtic tongue than most militant Irish nationalists were in Irish. But he used this fluency to articulate a very powerful argument against “Celtic” separatism: If Welsh could thrive in the UK, why could Ireland not revive its own culture while staying within the union? If he felt any affinity with his “Celtic” brethren across the Irish Sea, it did not stop him from unleashing the infamous Black and Tan and Auxiliary irregulars against the Irish population in an effort to crush the Irish Republican Army or defending their indiscriminate “reprisals” against Irish civilians. When it mattered, common “Celtic” identity didn’t amount to much. It has never had any purchase on real politics.

Nor is it entirely innocent. Celticism may be broadly associated with touchy-feely New Age spirituality, but it is also increasingly adopted by the far right. One of the oldest and largest neo-Nazi forums on the Internet, Stormfront, uses a Celtic cross as its logo. The Anti-Defamation League notes in its database of hate symbols that “the white supremacist version of the Celtic Cross, which consists of a square cross interlocking with or surrounded by a circle, is one of the most important and commonly used white supremacist symbols.” It is easy to see why: Celticism offers whiteness without guilt, a white identity that is associated not with persecution but with victimhood. This chimes perfectly with the self-pitying narrative of the far right: the claim that white identity, once dominant like the continental Celts, is now being swamped and pushed out to the margins.

But the best answer to this abuse is not to drop the whole Celtic business altogether. It is to actually understand it. For what it really shows us is that collective identities are not racially determined. They are inventions. They have almost nothing to do with genetic purity and everything to do with the needs of particular historical moments. They exist in the realms of the imagination, in both its hatefully destructive and its joyfully playful forms. If the Celts are alive in our historical moment, it is not because we know who they are, and still less because they help any of us to really know who we are. It is because they can be summoned to sanction anything from white supremacy to gay liberation, just as in the nineteenth century they could be called on to argue either for or against movements for national independence. And who knows but that, in the long and turbulent aftermath of Brexit, a new Pan-Celticism may yet be invented as an argument for keeping Scotland (and perhaps Wales) in the European Union alongside Ireland?

Matthew Arnold was nearly right about one thing. He imagined that the great racial characteristic of Celtic people was the “indomitable reaction against the despotism of fact.” But it is not the Celts who do, or ever did, this. It is we who use the Celts to help us react against the despotism of fact. That reaction may be an invitation to dangerous racial fantasy. Or it may be a salutary reminder that identity is like one of those great pieces of Celtic art, never moving in a straight line, but turning around and back on itself in dizzying twists and unpredictable swerves.