There is a list in Caesar, eleven names long, set down in 56 BC like an administrative memo. Most of these peoples then drop out of history as though someone had closed a door. One did not. Second in line stand the Bigerriones, a small people on the upper Adour at the northern foot of the Pyrenees. Their name never disappeared. It became a fortress, then a landscape, then a county, and anyone speaking today of the Bigorre in south-western France is using, without knowing it, a word older than Roman rule in Gaul. This article asks why. Because the answer is not the one you would expect.
The rule in this series has so far been simple: peoples disappear. A name surfaces once in an ancient author, perhaps twice, and then there is silence. We reconstruct from a single line, a milestone, a votive altar. The Bigerriones are the exception, and a very clear one. Their name appears not only in the ancient lists but on today's map. The Bigorre is a living landscape in the département of Hautes-Pyrénées, with its own name, its own sense of itself, its own cooking. It is named after a people Caesar mentioned in 56 BC.
That is unusual enough to be worth asking how it happened. The obvious answer would be: Rome turned the people into a civitas, the civitas became a diocese, the diocese a district and then a county. Administrative continuity. That answer is not wrong, but it is imprecise. Because the moment you open the decisive late antique source, the tribal name is not there as the name of the administrative unit. Something else stands there instead. And the whole story hangs on exactly that.
In the summer of 56 BC Publius Licinius Crassus, one of Caesar's legates and son of the triumvir, campaigned in Aquitania. After a battle against an army Caesar puts at 50,000 men, said to have included fighters from beyond the Pyrenees, resistance collapsed. Caesar sums up the outcome in a single sentence: Hac audita pugna maxima pars Aquitaniae sese Crasso dedidit obsidesque ultro misit – on news of this battle the greater part of Aquitania surrendered to Crassus and sent hostages of its own accord. Then comes the list: Tarbelli, Bigerriones, Ptianii, Vocates, Tarusates, Elusates, Gates, Ausci, Garumni, Sibusates, Cocosates. Eleven names. The Bigerriones stand second, directly behind the Tarbelli, with whom they also appear in close company in Pliny.
The list is not a text Caesar wrote in order to do ethnography. It is a report of success. The names are negotiating partners, providers of hostages, entries in a ledger. And they are not cleanly transmitted: the manuscripts vary, and several of the eleven appear differently in different editions, such as Ptianii beside Preciani and Vocates beside Vocasates. For the Bigerriones the reading is stable. That is not a given, and for what follows it matters.
A point that popular accounts almost always lose: the Bigerriones were not Celts. Caesar himself writes in the opening sentence of his work that the Aquitani differ from the Gauls in language, institutions and laws. Strabo adds that in speech and physique they resemble the Iberians more than the Gauls. This is not an ancient quirk; it can be tested linguistically.
Aquitanian is a Vasconic language, that is, related to Basque. No continuous text survives. What we have are personal and divine names carried along in Latin inscriptions of the imperial period – on current estimates some 200 personal names and about 60 names of deities, mostly from the first to the third century. Luis Michelena first described this material linguistically in 1954, Joaquín Gorrochategui collected it in full in 1984. The elements are readable from Basque: andere "woman", ume "child", sahar "old", gorri "red". The Bigerriones belong to this world. Anyone who turns them into Celtic warriors is mistaking neighbours for relatives.
About 120 years after Caesar, Pliny the Elder writes out his roll of the Aquitanian communities. It is a long, dry, partly undecipherable sequence of names, and it is nonetheless one of the most important sources we have for this region. The Bigerriones appear there in a slightly different form.
Aquitanicae sunt Ambilatri, Anagnutes, Pictones, Santoni liberi, Bituriges liberi cognomine Vivisci, Aquitani, unde nomen provinciae, Sediboviates. mox in oppidum contributi Convenae, Begerri, Tarbelli Quattuorsignani, Cocosates Sexsignani …
"To Aquitania belong the Ambilatri, the Anagnutes, the Pictones, the free Santoni, the free Bituriges surnamed Vivisci, the Aquitani, from whom the province takes its name, the Sediboviates. Then those gathered into one town: the Convenae, the Begerri, the Tarbelli of the four standards, the Cocosates of the six standards …"
Pliny, Naturalis historia IV,108; translation: Glanz und Gravur
The form here is Begerri, not Bigerriones. The manuscripts also offer Begerbi, Bebergi and Bergebi – the kind of slips a copyist makes in a list of names he has never heard spoken. Paul-Marie Duval, who studied this list in depth in 1955, takes Caesar's Bigerriones and Pliny's Begerri to be one and the same people. Scholarship has followed him.
More interesting than the name is the phrase before it: mox in oppidum contributi, "then those gathered into one town". The participle is plural and nominative, and the real question is how far it reaches. Does it apply only to the Convenae, named immediately after? To Convenae and Begerri together? Or does it introduce the entire following sequence?
Most editions and translations refer the phrase to the Convenae, and that fits well: their name means roughly "those who came together", and their town of Lugdunum Convenarum is the textbook case of a Roman administrative act turning scattered people into a civitas. Extend the phrase further and you get a quite different statement, namely a series of peoples whom Rome assembled into towns. For the Bigerriones that would mean they too entered the Roman order not as a grown unit but as one shaped by it. To be honest: the Latin does not settle this. It can be punctuated either way, and it has been. Anyone selling one of these readings as certain is selling an interpretation.
The territory of the Bigerriones lay along the upper Adour, in the plain around present-day Tarbes and in the foreland of the high Pyrenees. On Philippe Gardes's assessment the Adour formed the backbone of their country; their original lands probably covered much of the later civitas, taking in the Lavedan and the upper Adour valley. To the east, towards the valley of the Aure, their boundary ran against the Convenae.
This is a frontier zone in more than one sense. To the north lie the Ausci, to the east the Convenae and the Garumni on the upper Garonne, to the west the Tarbelli with their access to the Bay of Biscay. Mountains, rivers, pasture: the Bigerriones sit where the plain meets the range and where the routes out of the valleys converge. Such positions are rarely unimportant and just as rarely spectacular. They do not produce great battles. They produce endurance.
Archaeologically, about a dozen fortified sites are known in Bigerriones territory, many of them strung along the plateau edges above the Adour valley. The chief one is the spur of Castetbielh at Saint-Lézer, an enclosure of some seven hectares commanding the river. Occupation has been assumed as far back as the Bronze Age; in the early Iron Age a fortified hilltop settlement arose here.
No protohistoric rampart survives, however. What is visible is the ancient wall, and above it the medieval motte, and scholars suppose that both partly reused the older line. That is an important methodological point: the continuity of this site is a plausible reconstruction from the lie of the ground and the later structures, not an excavated pre-Roman find. At Tarbes itself, traces of early occupation point instead to an open settlement in the plain.
What makes this place decisive for our question is its late antique phase. After the incursions of the early fifth century a fortress was laid out on the spur: the Castrum Bigorra. Its wall enclosed several hectares, its length is given as around 940 metres, reinforced by several towers. It protected the population of the neighbouring vicus – later Vic-en-Bigorre – and of the surrounding settlements.
So here, in the middle of the collapse of the Roman order, stands a structure carrying the tribal name. Not the town, not the administrative district: the fortress. And that fortress became the most important place in the region for centuries. Coins were still being struck there in the sixth century. Into the early Middle Ages the site counted as the capital of the Bigorre; a count of Bigorre later built a castle on part of the ground, and below the walls a priory was founded that was placed under Cluny in the eleventh century. Landslips have shifted the remains of the wall, which are now accessible along an archaeological trail.
Now to the source that forced a correction to the brief for this article. The Notitia Galliarum, the administrative survey of the Gallic provinces from around AD 400, lists twelve communities for Novempopulana: Auch at the head, then Aquae, Lectoure, Saint-Bertrand-de-Comminges, Saint-Lizier, the Boiates, Béarn, Aire, Bazas – and in tenth place the entry that matters: civitas Turba, ubi castrum Bigorra, spelt Bogorra in Otto Seeck's edition.
There is, in other words, no civitas Bigerrionum in this list. The community is named after the place Turba, not after the people. The tribal name appears only as the name of the fortress. Measured against our opening question that is an almost paradoxical finding: the name that survives in the end is precisely not the official name of the administrative unit in the one late antique list we have. The Bigorre owes its name not to the district but to the castle.
The passage is famous partly because it is oddly built. Other entries in the Notitia simply name a civitas. Here we have a double entry: town plus fort. Ferdinand Lot held the transmitted text to be corrupt and proposed reversing it: civitas Bigorra ubi castrum Tarba – the city of Bigorra, where the fort of Tarba lies. Saint-Lézer would then have been the chief town, and Tarbes merely the fortified height.
Jean-Pierre Bost, by contrast, keeps the text as it stands and takes Tarbes to have been the chief town from the first century onward; Philippe Gardes identifies the Notitia's castrum Bigorra with the Roman fort at Saint-Lézer. The dispute is unresolved, and it is more than a philologists' quarrel. At stake is whether in this landscape the administrative seat moved from the plain onto the spur or the other way round. All that is certain is that around 400 both places were important enough to be named in a single entry.
A good deal is taken for granted about Tarbes that turns out, on inspection, to be unattested. No ancient author names the town. The earliest mention is that same Notitia entry around 400, in the form Turba. The meaning of the name is regarded as obscure; in the Middle Ages it appears as Tarbe, Tursa, Tarvia. Neither Caesar nor Pliny nor Ptolemy knows it.
This also clears up a stubborn error that has already come up in our article on the Tarbelli: Tarbes has nothing to do with them. The town lies in Bigerriones territory, not in Tarbelli territory, and the similarity of the names is exactly that – a similarity. Among the Tarbelli no landscape name survives; among the Bigerriones one does. The two neighbouring peoples stand side by side in one line of Caesar and have precisely opposite afterlives.
What this region has left in abundance is divine names. On Duval's count the Bigerriones are among the Aquitanian peoples with the most inscriptions; some sixteen are assigned to them. The most impressive site lies at Montsérié in the Hautes-Pyrénées: a sanctuary that yielded a very fine bronze mask and more than twenty votive altars – on a precise count twenty-four dedications to the god Erge, together with seven to Mars and one to Jupiter. Erge is thus one of the best attested native gods of the whole Pyrenean region, and he is known essentially from this one place.
Other native deities are scattered across the area: Belgo at Gazost, a water divinity; Baigorix at Cadéac, there equated with Jupiter; Fagus between Tibiran-Jaunac and Générest, a deified tree; Ageio in the iron-working country of the Hautes Baronnies. A correction is needed here, because the brief for this article had it wrong: Abellio, Leherennus and Ilixo are precisely not characteristic of the Bigerriones. Abellio is attested above all in the Haute-Garonne, Leherennus practically only at Ardiège, and Ilixo is the spring god of Luchon. That is the Comminges, the country of the Convenae. Fail to separate the sacred landscapes cleanly and you credit a people with its neighbour's cult.
Bagnères-de-Bigorre, the second great town of the landscape, has lived off its thermal springs since antiquity. When the Grands Thermes were built in 1824 the remains of Roman baths came to light; that people bathed here in Roman times is beyond doubt. The town's ancient name, on the other hand, is very much in doubt. The literature offers Aquae Convenarum and Vicus Aquensis; both are contested, and some authors regard Vicus Aquensis as a generic term for a spa rather than a transmitted proper name.
This is typical of the region and worth holding on to: we have the baths, we have the gods, we have the ramparts – and for many of these places we have no secure ancient name. Against that background the tribal name of the Bigerriones is the unusually robust quantity.
Around 400 the name stands in the Notitia as the designation of a fortress. Barely two centuries later it has become something else. Gregory of Tours, writing in connection with the Treaty of Andelot in 587, names a civitas Begorra. The name has passed from the building to the community – and with it to the region.
That is the real process, and it can be generalised. A name survives not because a people continues, but because it is attached to something that is needed: a fort you flee into, a court district, a diocese, a title of lordship. The Bigerriones as a people are no longer traceable as actors after the first century. By then their name had long ceased to be an ethnonym and had become a location. Locations are tough.
In the early Middle Ages the Bigorre appears as a Carolingian county within the duchy of Gascony, with ties into the Navarrese-Pyrenean world. Popular accounts begin the line of counts early, invoking the so-called charter of Alaon, according to which a Gascon duke granted the Bigorre to a son at the start of the ninth century. That document is a seventeenth-century forgery. Scholarship retains the person in question but moves him to the end of the ninth century; the first reliable mention of a count of Bigorre dates from 945.
After that the county passes through the great houses of the south-west – Foix, Béarn, Marsan, Comminges, and in the thirteenth century Montfort. The name does not travel. It stays while the lordship changes hands, and that is precisely the property that carries it to the finish. We know the pattern from this series already: the Vellavi gave their name to the Velay, the Gabali to the Gévaudan. The Bigorre belongs to the same family of landscape names.
The county of Bigorre ends as an administrative unit with the French Revolution. In 1790 the département of Hautes-Pyrénées is created, with Tarbes as its chief town. The political form in which the name had last lived thereby disappears. And still the name does not. Bigorre remains in use as a name for the country, in place names such as Vic-en-Bigorre, Bagnères-de-Bigorre or Rabastens-de-Bigorre, in the names of clubs, in product labels, in everyday speech. After more than two thousand years the name has detached itself from every institution that once carried it and is simply what the region is called.
Run through Caesar's eleven names once more. The Elusates live on faintly in Eauze, the Ausci in the town name Auch, the Convenae in the Comminges. Of the Sotiates, subdued in the same campaign of 56 BC, little remains but a place name. Most of the rest fade out. So why the Bigerriones?
The honest answer has three parts, and none of them has to do with size or power. First: the name was attached to a structure people needed. A fortress on a spur above the Adour was, in the fifth century, somewhere you went; an administrative district was something you wrote down. Second: it was attached to a closed natural region. The Bigorre is a basin ringed by mountains, and such regions keep their label because the people living in them need one. Third, and this is the underrated point: there was no competitor. Tarbes enters the record so late that the name of the country was already fixed. Where a strong town was there early, its name overwrote the tribal name – which is exactly what happened to the Ausci. Where there was none, the field stayed open.
For us as a workshop that matters beyond antiquarian interest. What survived here is not an object, not a grave, not a piece of armour. It is a word that people held on to because they could use it. Two syllables, older than any wall in that landscape, and no Roman administrative list ever took them as seriously as the people who lived there.
The Tarbelli · The Garumni · The Ausci · The Convenae · The Sotiates · The Vellavi · The Gabali · The Celtic Warriors.
All History & Finds articles →
Long before the Edda was written down, the people of Gotland told their stories in stone – in images that still puzzle …
Read more →About 30 kilometres west of modern Stockholm sits a small, unremarkable island in Lake Mälaren. Walk across Björkö …
Read more →Since 1878, Birka’s grave Bj 581 had been the textbook example of a high-status Viking warrior – the „ultimate Viking" …
Read more →