The Aquitani

← Encyclopedia

At a spring-sanctuary in the foothills of the western Pyrenees, sometime in the second century of the common era, a man of the country had a stone cut and set up beside the warm water. The work is entirely Roman. The altar is a squared block with a molded cornice; the letters are good imperial capitals; the formula is the one used from Britain to Syria — the dedicator names the god in the dative, records that he paid his vow gladly and deservedly, and signs his own name. Everything about the monument says that its maker lived inside the Roman world and worshiped in its manner. Everything except the god. The name in the dative is neither Latin nor Celtic. It is a word like Baigorrixo, or Ilurberrixo — and a Basque shepherd from a valley two mountains east could read the roots of it today. Berri means “new.” Gorri means “red.” The god has a name built out of the living language of a people whom Rome believed it had absorbed three centuries before, and the stone that carries it is, in a strict and literal sense, one of the oldest surviving pieces of Basque.

The people who raised such stones were the Aquitani, and this is the paradox at the center of their history. They left no book, no myth written in their own hand, no line of verse. What they left was their names — their own and their gods’ — caught in the amber of somebody else’s language, on altars they raised in somebody else’s fashion. It was not their swords that outlasted them. It was the sound of their words.

Caesar’s line at the Garonne

The Aquitani enter the written record at the very opening of one of the most-read books in Latin. “All Gaul,” Julius Caesar begins, “is divided into three parts, one of which the Belgae inhabit, the Aquitani another, and the third a people called in their own tongue Celtae, in ours Galli.” He wrote it around 58 BCE, at the head of his account of the wars that would make him master of Rome, and it is the first sentence generations of schoolchildren met when they opened the Gallic War. What follows is the crucial line for anyone who wants to know who the Aquitani were: “These all differ from one another in language, customs, and laws.” Caesar then fixes the boundary with a geographer’s precision — the river Garonne separates the Gauls from the Aquitani — and defines their country as the wedge running “from the river Garonne to the Pyrenean mountains and to that part of the ocean which is near Hispania.” It is the far southwestern corner of Gaul: roughly modern Gascony and the French Basque country, hemmed by the Garonne to the north, the mountains to the south, and the Atlantic to the west.

Caesar had a soldier’s reason to notice the difference, but he was not alone in noticing it. The geographers who came after him — Strabo above all — insisted that the Aquitani were not Celts at all. In build, in bearing, and in speech, Strabo wrote, they resembled the Iberians across the mountains more than they did the tall fair Gauls to the north. Modern readers should not press ancient ethnography too hard; its categories were loose and its physiognomy was guesswork. But the linguistic observation was sound, and the last two thousand years have borne it out. The Aquitani spoke a tongue that was neither Celtic nor Italic nor anything else in the great Indo-European family that had spread across Europe. It was something older, something that had been in this ground before the Celts arrived, and it had a single distant descendant that would still be spoken when Latin itself had died into French.

They were not one nation but many. Strabo counted “about twenty peoples” among them; taken together, Caesar, Pliny the Elder, and the geographer Ptolemy yield more than thirty distinct tribal names — the Tarbelli of the coast around modern Dax, the Sotiates and the Vocates and the Tarusates of the interior, the Ausci whose name survives in the town of Auch, the Bigerriones behind whom stand the modern Bigorre. It was a loose weave of hill and river communities, without a single center or a common capital, the kind of political landscape that is easy to conquer piecemeal and hard to hold whole.

Rome took it by force in 56 BCE. Caesar, campaigning far to the north, detached his young legate Publius Licinius Crassus — son of the triumvir, the immensely rich Crassus who would die at Carrhae three years later — and sent him south with, in Caesar’s own reckoning, twelve cohorts and a large body of cavalry. Crassus reinforced these from the old Roman province along the Mediterranean and marched into the territory of the Sotiates, who met him with horse and foot, fought hard, and were beaten; he then stormed their stronghold. The other Aquitanian peoples, alarmed at how fast a single corner had fallen, drew together into a coalition. They sent to Hispania for help and put themselves under leaders who had served through the long guerrilla war of Quintus Sertorius against Rome and were reckoned masters of the craft. Caesar numbers the host they raised, out of Aquitania and from among the Cantabri beyond the mountains, at some fifty thousand. Crassus broke them in the field, and when his cavalry ran the fugitives down across the open plains, Caesar records, scarcely a fourth part of the fifty thousand survived. After that the resistance collapsed. The greater part of Aquitania sent hostages and submitted — the Tarbelli, the Bigerriones, the Vocates, the Tarusates, and the rest — and the country passed, with a speed that still reads as brutal, into the Roman order.

The tribes and towns of Gaul on the eve of the Roman conquest; the Aquitani hold the southwest, between the Garonne and the mountains. Feitscherg, via Wikimedia Commons. CC BY-SA 3.0.

The gods with Basque names

For most conquered peoples that would be the end of the story: a paragraph in a commentary, a list of tribes on a triumphal inscription, silence. The Aquitani are the great exception, and the reason is buried in the way they worshiped once Rome had made them provincials.

Across the Pyrenean valleys, at the thermal springs and mountain shrines the Romans loved, the Aquitani took up the imperial habit of the votive altar. They dedicated in Latin, in the correct formulae, to gods with correct-sounding case endings. But the gods themselves were their own. When the philologists began to gather these dedications — several hundred of them, scattered through the sanctuaries of the Comminges, the Bigorre, the Béarn, and the coastal plain, and dating from roughly the first to the third centuries CE, a few perhaps later — they found a fossil language. The surviving corpus amounts to about four hundred personal names and some seventy names of gods, the figures cited by the Basque philologist Joaquín Gorrochategui and by the linguist R. L. Trask; more cautious summaries put it lower, at around two hundred personal and sixty divine names, but even the conservative count preserves a substantial vocabulary. And that vocabulary is, to a reader of Basque, uncannily legible.

The pioneer was a young French scholar, Achille Luchaire, who in the 1870s first set the Aquitanian names beside medieval and modern Basque and saw the correspondences. The work was placed on a rigorous footing in the 1950s by Koldo Mitxelena — Luis Michelena, the towering figure of twentieth-century Basque linguistics — and given its definitive corpus by Gorrochategui, whose 1984 study of the indigenous onomastics of Aquitania remains the standard reference. What they showed is that the Aquitanian words read straight through Basque. The element sembe- recurs in the names as “son,” and Basque for son is seme. Cison or gison means “man,” and Basque for man is gizon. Andere and its diminutive anderexo mean “lady, woman,” Basque andere. Nescato means “girl,” Basque neska. The little suffixes that pepper the names, -co and -xo, are the Basque diminutives still in use. These are not scattered look-alikes of the kind that turn up between any two languages by chance; they are the ordinary, high-frequency words of kinship and person, the last vocabulary a language surrenders, appearing exactly where a Basque speaker expects them.

The gods carry the same signature. Alongside Baigorrixo and Ilurberrixo — the berri “new” and gorri “red” that a shepherd could still parse — stand Ilunno, whose root recalls Basque ilun “dark”; Artahe, in which readers hear arte, the holm oak; and Leherenno, and a long file of others. These were local powers, gods of a particular spring or peak or wood, named in the tongue of the valley that honored them and then, when the fashion came, carved onto Roman stone. The worshiper had changed his script and his ritual grammar without changing his god’s name, and in doing so he froze that name for the philologists to find.

The conclusion the specialists draw from all this is as firm as anything in the study of a language with no ancient literature can be. Aquitanian is either the direct ancestor of Proto-Basque or a very close sister to it — the two are not spoken over any great distance apart in time or space — and it is by a wide margin the strongest evidence that Euskara, the Basque language, is a pre-Indo-European tongue native to this ground since before history. Very few language isolates can be traced at all into the ancient world; Basque can be followed, through these altars, two thousand years into a documented past. That is a rare and precious thing, and it belongs to the Aquitani.

A Roman votive altar at Bagnères-de-Luchon in the central Pyrenees, of the class on which Aquitanian personal and divine names were preserved. Eugène Trutat (1840–1910), via Wikimedia Commons. Public Domain.

It is worth drawing one sharp line here, because the strength of the Aquitanian-Basque case has repeatedly been borrowed to prop up a far weaker one. In the 1990s the German linguist Theo Vennemann advanced what he called the Vasconic substrate hypothesis, gathered in his 2003 volume Europa Vasconica–Europa Semitica: the claim that a whole family of Basque-related languages once blanketed prehistoric Western Europe, spreading out of the Ice Age refuges of Aquitania to name the continent’s rivers and hills before the Indo-Europeans arrived. It is a seductive picture, and it is not the mainstream view. Historical linguists have largely rejected it — Trask, Joseba Lakarra, P. R. Kitson, Jürgen Udolph, and Stefan Georg among the critics — on the grounds that its etymologies compare modern Basque to ancient place-names rather than to reconstructed Proto-Basque, segment words arbitrarily, and read chance resemblance as descent. The narrow claim and the broad one must be kept apart. That Aquitanian is Basque’s ancestor is well founded and widely held; that a Vasconic family once covered Europe is a fringe proposition that the discipline does not accept. The one should never lend its credit to the other.

The name that outlived the people

The word Aquitania had a long career ahead of it, and most of that career had less and less to do with the people who first bore the name. Augustus, reorganizing conquered Gaul after his census of 27 BCE, enlarged the province of Gallia Aquitania far to the north, past the old ethnic frontier, all the way to the Loire, sweeping in a mass of genuinely Celtic peoples who had nothing to do with the Aquitani proper. The name now covered a vast administrative space defined by a bureaucrat’s convenience rather than by any nation. When Diocletian carved the region up again at the end of the third century, one of the new units was called Novempopulania — “the nine peoples” — and it corresponded, tellingly, almost exactly to Caesar’s original Aquitania between the Garonne and the mountains, as though the administration retained a faint memory of where the real ethnic line had run. The name kept traveling. In the seventh century a Duchy of Vasconia formed in the southwest, and Vasconia would wear down through the mouths of centuries into Gascogne, Gascony; the medieval Duchy of Aquitaine, which Eleanor carried to two thrones, preserved the ancient word in the grandest form of all. By then the Aquitani as a distinct people were long gone, Romanized out of their language and into the Latin that became Gascon and French. The name outlived the tongue that gave it — everywhere, that is, except in the valleys where the tongue itself survived.

For the Basques, who have no ancient literature of their own, the Aquitanian inscriptions became the cornerstone of an entire scholarly tradition. They are the fixed point from which linguists reconstruct the older stages of the language and against which every theory of its history must be tested. Chief among those theories is the vexed question of “Vasconization”: whether Basque speech was long native to both slopes of the western Pyrenees, or whether it spread — north across the Garonne, or west along the range — only in the late and post-Roman centuries, as the imperial order dissolved. The debate is genuine and unsettled; serious scholars hold different views, and the Aquitanian evidence is invoked on more than one side of it. It should be presented for what it is, an open problem, not a solved one.

The altars themselves testify to something subtler than conquest or resistance: a people who let Rome reshape the outward forms of their worship while keeping the names at the heart of it, dedicating to indigenous gods in the invader’s language and script. It is the ordinary syncretism of a province, and folklorists have read it alongside the later Basque sacred world — the mountain powers, the goddess of the caves and the weather whose full story belongs to the Basques’ own tradition — with due caution about the many silent centuries in between. That later history, including the long tragedy in which the old religion was recast as witchcraft and hunted, is told where it belongs and is not to be read back onto these stones.

New evidence keeps arriving. In 2021, on a hilltop at Irulegi near Pamplona — south of the mountains, in the country of the Vascones rather than the Aquitani proper — excavators lifted from the burned debris of an Iron Age settlement a flat bronze in the shape of a hand, inscribed in a Palaeohispanic script. Its first word reads sorioneku, and a modern Basque hears in it zorioneko, “of good fortune.” Dated to the first century BCE, the Hand of Irulegi is the earliest Vasconic writing yet found, and it pushes the documented life of this language back another step, on the far side of the Pyrenees from the votive altars but pointing to the same deep continuity. Modern Basque cultural life — the immersion schools, the standardized written language, the pride of a small nation in a tongue with no living relative on earth — draws real strength from this evidence of antiquity, and rightly so. It also draws bad company: nationalist and New-Age writing that overstates the case, that miscites Vennemann’s rejected continent-wide claim as if it were settled fact, that wants the stones to prove more than stones can. The care the scholars take is part of the tradition’s dignity, not a qualification of it.

Set the Aquitani beside the peoples Caesar conquered and forgot, and what distinguishes them is not their fate — which was the common one, defeat and absorption — but the strange durability of the smallest thing about them. A tribe can be broken in an afternoon on an open plain; fifty thousand men can be run down by cavalry and reduced to a quarter, and their country can send its hostages and vanish from the record. But a word for “son,” a word for “red,” the name a valley gave its spring-god, can slip through the fingers of an empire and go on being spoken. The Aquitani are best understood not as a lost tribe but as the ancestors who left, on Roman stone in the Roman alphabet, the first legible sentences of a language their descendants still speak.

→ Related: The Basques · The Vascones · Mari Goddess · Sugaar · Basajaun · Lamiak · Akerbeltz · Anboto · Aralar · Zugarramurdi · Txindoki