The history of atic research into Suletian prehistory is relatively recent. With the exception of the anecdotal report provided by R. Gombault in 1914 about the six burial mounds at Egizuri, Irati and Larrau, most of the research has been carried out during the second half of the 20th century. It was from the middle of the century onwards that there was a sustained development of prehistoric research. 1949 and 1950, three of the most emblematic prehistoric sites in Zuberoa were discovered: the cave dwellings of Haregi and Gatzarria and the rock ‘sanctuary’ of Etxeberri. If P. Boucher can be considered the initiator of this new phase, the contribution made at that time by J.M. Barandiarán and, especially, G. Laplace, alongside him, was equally important in terms of its initial materialisation and subsequent consolidation.
Since then, Zuberoa's rich heritage has undergone a considerable increase in terms of archaeological references. This has been mainly due to the diverse finds made at prehistoric sites (Palaeolithic decorated caves, Mesolithic shelters, dolmen complexes, etc.) and, in particular, protohistoric sites (with an important catalogue of cromlechs, burial mounds and fortified camps). This has provided essential source material for beginning to unravel the prehistoric past. This has resulted in a better understanding of that remote antiquity, thanks to the control of the sites through appropriate excavation s and techniques, the proper processing of the available data in the laboratory in order to deduce relevant information, and the involvement of different specialists and research teams in this work. In addition to the ongoing fieldwork carried out by P. Boucher and the palaeontological interpretation by G. Laplace, we must also mention the efforts that, since the 1970s, have been made, especially in the field of prehistoric archaeology by D. Ebrard and in protohistoric archaeology by J. Blot and F. Gaudeul.
Despite these achievements, there are still significant gaps and shortcomings in our knowledge of prehistory in Zuberoa that future research will need to address. The absence or very limited knowledge of certain prehistoric cultures is compounded by the lack of information available on some sites that were excavated years ago... Furthermore, from a global spatial perspective, prehistoric research in Souletine, like that in the other provinces of Iparralde, has not participated in the solid dynamic that has been experienced in the southern territories of Euskal Herria since the early decades of the 20th century: general structuring of research, initially around the interdisciplinary team made up of T. Aranzadi, J.M. Barandiarán and E. Eguren and later through J.M. Barandiarán and his most direct disciples; definition of general research guidelines (focused on specific cultural stages or phenomena, on settlement processes in specific geographical settings, etc.) within a global conception of the Basque territory; intense and fruitful development of field action programmes (search, identification and excavation of sites) and, progressively, the atic application of qualified multidisciplinary analyses... All of this, logically, will have a significant impact on the coherence of the prehistoric discourse.
Remnants of that era also remain in the form of ‘gaztelus’, fortified or simple enclosures. Their study and inventory were undertaken by General F. Gaudeul. They are generally located on hilltops and can be classified into two categories: parapet enclosures and stepped enclosures. They are often surrounded by earthen embankments or stone walls. Some of the simplest ones were undoubtedly used as sheepfolds, but most were built as defensive structures to protect humans and animals from invaders, as castles would later do; they are often surrounded by moats and embankments. General Gaudeul points out that ‘several other constructions can generally be seen from each construction, so that the whole complex could constitute an excellent observation and warning network’. These fortifications are difficult to date as they have been reused at various times and still need to be excavated further.
A significant number of the sites studied by General Gaudeul are located in Zuberoa. Here is the list he provides with the location of the study:
Protohistoric sites in Zuberoa according to the table published by General Gaudeul
(B.S.P, 1981-1982, pp. 32-33)
1. Enclosures with parapets
Earth parapets.
Gazteluzarre, in Ordiarp (B.S.P., 1974)
Gaztellaia, in Chéraute (Revue de Pau et du Béarn, 1975)
Gaztelari, in Etcharry (Ibidem, 1976)
Gazteluxage, in Arrast-Larrebieu (B.S.P., 1976)
Gaztelugain, in Gotein-Laruns (Ibidem)
Larlete, in Berrogain-Laruns (Ibidem)
Txoikantegia, in Idaux-Mendyy (Ibidem)
Lexegita, in Trois-Villes, Barcus (Ibidem)
Kurku, in Nabas (Ibidem, 1977)
Haitzhandialtea, in Barcus (Ibidem, 1980)
Koxugaina, in Esquiule (Ibidem, 1983)
B. Stone parapets.
Maidekoralia, in Alçay (Ibidem, 1975)
2. Stepped enclosures
Montarei, in Etchebar (Ibidem, 1976)
Gaztalaria, in Sauguis-Saint-Etienne (Ibidem, 1977)
Gaztelu, in Aussurucq (Ibidem)
Gaztelugaña, in Sainte-Engrâce (Ibidem, 1980)
3. Unclassified constructions
Gaztelharriko Botxia, in Lacarry-Arhan (Ibidem, 1977)
Charritte-de-Flaut (ibidem)
La Madeleine, in Tardets, Trois-Villes, Barcus (‘Bulletin du Musée Basque’, 1976)
Château-Fort de Mauléon (Ibidem)
Cumbre de Erretzu, Tardets, Montory
All these protohistoric remains are the only evidence we have of human activity during that period. They are mostly located in areas of transhumance: ‘high-altitude pastures, ridges and hills accessible for part of the year’ (Blot, 1976). In spring, shepherds leave the lowland pastures, which have been depleted during the winter, with their flocks and, as the snow melts, they move towards higher peaks in a north-south direction. Dr Blot points out that the shepherds of protohistory did nothing more than follow in the ancient footsteps of the nomadic hunters of prehistory, as evidenced by the Palaeolithic flint tools found along these tracks. These same paths would later become Roman roads and the routes to Compostela.
The Basques made their entrance into history with the Greco-Roman authors. Caesar, in his Commentaries on the Gallic War (I, 1 and 1, from 58 to 51 BC), wrote: ‘Gaul as a whole is divided into three parts, one inhabited by the Belgians, another by the Aquitani, and the third by those who in their native language are called Celts, whom we call Gauls’. He adds: ‘All of them differ from each other in language, customs and laws.’ Strabo (Geography, IV, 1) insists on this difference the Aquitani and the Gauls: ‘The Aquitani are a people entirely apart, not only in language, but also in physical appearance...’ They were, it seems, less marked by the Celtic influence than the other peoples of Gaul. See Aquitaine.
According to Caesar, ‘Aquitaine stretches from the Garonne to the Pyrenees and the part of the ocean that bathes Hispania’ (Caesar, Op. cit. I, 1). Its name comes from its ancient capital, ‘Aquae Tarbellicae’ (Dax). It was conquered by Crassus, Caesar's lieutenant, in 56 BC.
‘Caesar orders Publius Crassus to go to Aquitaine with 12 cohorts of legionaries and a large cavalry in order to prevent the peoples of this country from sending aid to Gaul and such large nations from uniting’
(Caesar, Idem, III, 11).
This was an event of great importance, as it was the first example of Basque unity against an invader, and the Suletians were part of this coalition, along with all the other Aquitanians and even the Vascones, who had crossed the Pyrenees to join them. The battle took place near Dax or Tartas, but the Romans were masters in the art of war and Crassus won the victory over the Aquitanian coalition.
‘At the sound of this battle, a large part of Aquitaine surrendered to Crassus and sent ages on their own initiative. Among them were...the Sibuzates... A small number of states, trusting in the advanced season, did not follow this example’
(Caesar, ibid., III, 27).
Some have thought that these ‘Sibuzates’ mentioned by Caesar correspond to the ‘Sibyllates’ mentioned by Pliny the Elder (Natural History, IV, 108-109) in his list of Aquitanian peoples. For others, the former refer to the habitatues of Saubusse (Rat, M.: La guerre des Gaules, G. Flammarion, 244), Dax and Bayonne, but all agree that the latter, the sibyllates, are the inhabitants of Soule or Subola (Zuberoa in Basque). In this case, the Suletini are part of these peoples who, trusting in the winter, neglected to send ages to Crassus. It is likely, in fact, that the Suletini, belonging to those peoples far from the scene of combat, retreated to their mountains after the failure of the coalition and did not consider the urgency of submitting to the Romans. This independence was short-lived, however, as in 51 BC, Caesar himself, having subjugated Gaul, set out for Aquitaine.
‘This expedition, like the others, had a swift and happy outcome. All the states of Aquitaine sent him deputies and provided him with ages’
(Caesar, Op. cit., VIII, 46).
Later, in 27 BC, Emperor Augustus extended Aquitaine northwards to the Loire, but again, on a date that cannot be precisely determined, undoubtedly in the 2nd or 3rd century, the part of Aquitaine located south of the Garonne was separated from Celtic Aquitaine and took the name of Novempopulania or Federation of the Nine Tribes. Its date of birth is engraved on the famous stone of Hasparren. See Novempopulania, Verus.
Excerpt from the study by Gerhard Rohlfs, in Le Gascon. Études de philologie Pyrénéenne, on the Basque substratum in primitive Novempopulana Aquitaine
It is only in the domain of the ancient tribe of the Vasates, that is, around the town of Bazas and in the vicinity of Bordeaux, that place names ending in -ac reappear. They become more numerous further north the Garonne and the Dordogne, intermingled here with names ending in -an. We can therefore conclude that Gallo-Roman colonisation, even in these northern territories of ancient Aquitaine, did not succeed in overcoming the indigenous element.
We also note the extreme rarity of place names ending in -dunum in certain regions of Aquitaine. The few localities established along the course of the Garonne (Verdun, Lagdunum Convenarum, Salardú), in the Gers (Tourdun, Gaudun, Monlezun) or in the vicinity of Mont-de-Marsan (two Besaudums) are contrasted by a total absence of such names in the regions further west and south; See Xavier Ravier, VD, XII, 1963, p. 58. Compounds with -magus “field” (after dunum, the most common element in Gallic compound place names) are also particularly rare. I can only cite Condom Condatomagus (Gers) and Argenton (Argentomagus) a short distance north of Condom.
I have indicated on our map, by means of dotted lines, the areas of Gallo-Roman infiltration and colonisation. This dotted line marks the outer limit of the compact distribution of place names formed with the suffixes -anum and -acum, without taking into account the very sporadic isolated pockets. We can therefore conclude that vast areas of ancient Aquitaine must have remained untouched by Roman or Gallo-Roman influences.
Place names ending in -òs. Another family of place names, ending in -òs, provides even more striking evidence to support this view.
Through the latest research conducted by Mr. Séguy and myself on place names ending in -os, we now know that the Aquitaine tribes used a suffix in the formation of place names that must have had the same value as the Latin -anus and the Gaulish -acos. [See Jean Séguy, Le Suffixe toponymique -os en Aquitaine, in the Proceedings and Memoirs of the Third International Congress of Toponymy, vol. II, 1951, pp. 218-222; G. Rohlfs. Sur une couche pré-romane dans la toponymie de Gascogne et de l'Espagne du Nord, RFE, tome 36, 1952, pages 209-256].
It is the suffix -ossu that appears in Aquitaine inscriptions, attached to names of people or deities. We find, for example, the personal name Andossus derived from the cognomen Andus, Apinossus alongside the personal name Apinus, the personal name Billicatidossus alongside the personal name Billicatus; Ilunnossus, derived from Ilunnus, attested as the name of an Aquitanian deity. We can guess the function of our suffix when we remember that in Latin there existed:
Aurelianus alongside Aurelius, Octavianus alongside Octavius, Domitianus alongside Domitius.
Similarly, among the Gauls, the suffix -akos was used to form new anthroponyms. In Gallic inscriptions, we find the following personal names: Artacus alongside Artus, Cariacus alongside Carius, Verdinacus alongside Vedinus.
We know that both formations (Latin -anus, Gaulish -akos), used to name a person (Aurelianus, belonging to the Aurelii family), later became specialised to designate a landowner's estate. It is therefore reasonable to assume that the suffix -os, so widely represented in Gascon toponymy, constituted the Aquitaine process of deriving the name of an estate from the name of an owner, following the example of the Latin -anus and the Gaulish -akos.
In my work published in the Revista de filología española, volume 36 (1952), I analysed around 200 Gascon place names ending in -os. As a result, I was able to establish that almost all of them contain a person's name as their root. For example, Andernòs is derived from the Gallic personal name Andernus, Argelòs (from Argailo), Arròs (from Arro), Baliròs from Valerius, Bernòs from Bernus, Bournòs from Burnus, Cabidòs from Capitus, Gaillagòs (from *Galliacus), Julòs (from Jullus), Lauròs (from Laurus), Mouliòs (from Molinus), Pudòs (from Putus), Sabalòs from Sapalus, Vidalòs from Vitalis.
Our map s all the names ending in -os that I have been able to gather and locate. It can be seen that they are most concentrated in the Béarn plain south of the Adour, then around Pau (Anòs, Ardòs, Arròs, Baliròs, Bizanòos, Bruscòs, Carròs, Caubiòs, Guindalòs, Lagòs, Mansòs, Ziròs, Uzòs, Ylòs), Orthez and Bayonne, and in the district of Argelès. North of the Adour, the density decreases significantly. The vast, heavily wooded and sparsely populated region north of Mont-de-Marsan has no names of this type. The Landes coastline has a few examples that serve as a link with the Gironde department: Biscarrosse, Garrosse, Mézòs, Pissòs, Souprosse, Ygòs. Here, to the west and south-east of Bordeaux, the number of names ending in -òs is once again considerable. In this region, the density reaches its peak in the Bazadais, the former land of the Vasates (an Aquitaine tribe). To the north, the Garonne river forms a fairly significant boundary. I know of only five villages with the ending -òs that lie beyond the river: Carcòs, Carròs, Libòs, Moulós, Picòs. [To this small group were later added half a dozen localities in the Libournais region, none of which are more than 20 kilometres from the river: Bardos, Bellos (or Ballos), Garosse, Gueyrosse, Marosse and Sambosse; see R. Lafon, in “La Dordogne et sa région” (Bordeaux 1959), p. 127]. To the east, beyond the Gélise, names ending in -òs disappear. The entire vast plain Auch and Toulouse is devoid of them.
Towards the Pyrenees, east of the Garonne, names ending in -òs are very rare. I know of two cases in Haute-Garonne: Arguenòs (Argynnus) and Génòs (Genna), and two cases in Ariège: Alòs (Allus) and Miglòs (Miccalus?).
Broadly speaking, we can say that in the same areas where formations ending in -anum and -acum are becoming rarer, names ending in -òs are beginning to prevail. They almost completely replace Gallo-Roman formations west of Tarbes, in the Gave de Pau and Cave d'Oloron valleys. They also replace both Gallo-Roman types in the western part of Gironde. It is only in a few areas that formations ending in -òs intermingle with Gallo-Roman derivatives. This is the case in Bazadais (where we find Auròs, Baulòs, Bernòs, Billòs, Carcòs, Cudòs, Giscòs, Insòs, Ladòs, Sauròs, Targòs, Triscòs), north of Pau and in a large part of the Hautes-Pyrénées department [this situation is perfectly consistent with the epigraphic data: 'the epigraphic texts of the Ausci region and other parts of central Gascony contain many more names with a Gallic appearance than those of the mountains. Aquitaine onomastics become rarer as one approaches the middle and lower reaches of the Garonne. The region immediately to the left of the river appears to have been largely Celticised, as was the Bordeaux region' (Lisop, p. 108). 'This toponymy, which is related to Basque, becomes rarer in the plains of Aquitaine as one moves away from the Pyrenees and closer to the Garonne. There, it takes on a more Italo-Celtic or Celtic character' (ibid. 112)].
The main impression that emerges from our map is that Gallo-Roman colonisation, at a time that is difficult to pinpoint, must have come to a halt in the face of strong resistance from the indigenous population. Since the latter refused to accept Gallo-Roman derivatives, there is every reason to believe that the ancient Aquitanian language survived much longer than has been believed until now.
Derivatives ending in -os do not stop at the Pyrenees mountain range. They can be found on the other side of the Pyrenees in Upper Aragon and in the Pamplona region. They reach their maximum density north of Huesca and especially Jaca and Pamplona, i.e. in an area where there is an extraordinary density of Basque survivals. Due to a particular revolution in Spanish phonetics, our suffix appears here in the form of -ués: Aragués, Arascués, Arbués, Bagüés, Barbués, Bernüés, Binués, Chisaüés, Escabués, Gallués, Garrués, Gordués, Larués, Nardués, Sagüés, Sigüés, Undués, Urdués, Virués. It is often the same anthroponyms that form the roots of these formations on both sides of the Pyrenees: Angòs: Angués, Bernós: Bernués, Binòs: Binúes, Biscarosse: Bisccarrués, Garròs: Garrués, Urdòs: Urdués, Virós: Virués.
This striking similarity attests to the close linguistic relationship that must have existed Aquitaine and the ancient inhabitants of Upper Aragon. Since our suffix also belongs to the Basque language, where it appears in the form -oz or -otze (Aloce, Andoce, Arroce, Bardoce, Bildoce, etc., in Spain Escaroz, Iloz, Imoz, Madoz), the hypothesis that the Basque language and the language of ancient Aquitaine derive from the same root becomes more likely [the suffix appears to be the same as that which appears in the Basque Country in the form -oz(e) or -otz(e) (Michelena, On. 442)].
In any case, the results of our research must rule out an ancient Romanisation in Aquitaine in the territories where derivatives in -òs resisted Gallo-Roman formations. The continuity of the indigenous element in the French Basque Country can hardly be doubted [for another group of place names formed with the ending -un, concentrated in western Pyrenean Gascony (Aucun, Azun, Eygun, Lescun, Orcun) and attributable to a pre-Roman Pyrenean layer, see Xavier Ravier, VD, XII, 1963, p. 58 ff. As for Lescun (in patois Lascù, called Lascún by the Spanish = a. 1077 Lascun), it may be linked to the Basque laskunde “expansion” (López Mendizabal)].
Cette nouvelle région administrative comprenait, entre autres, la ville d'Iluro (Oloron), dont le nom vient de ili (ville) et ur (eau). Iluro serait donc la ville des eaux et, en effet, Oloron se trouve au confluent de deux rivières (Menjoulet, 1864, I, 16). Sous les Romains, une ville était bien plus qu'un village ; c'était le chef-lieu d'un district, dont le territoire était divisé en « pagi », sortes de cantons eux-mêmes divisés en « vici » ou bourgs. Ces termes ont subsisté sous la forme de « pays » et de « vic ». Il existait également des villes appelées « urbes », des places fortes « oppida » et des domaines ruraux « villae ».
The ‘pagos’ of Zuberoa depended on the aforementioned Iluro. The Roman administration adopted the natural regions, so each Pyrenean valley corresponded to a ‘pagos’. The Sibyllates or inhabitants of Zuberoa, ‘Subola’, occupied the ‘Pagos Solensis’, which would later become the ‘Pays de Soule’. As ethnic groups tended to respect Roman administrative divisions, it is reasonable to assume that the Iluronenses and the Sibyllates considered themselves to be quite close.
There are no known Roman towns in Zuberoa itself. Mauléon, which has retained a name that is more Latin than Basque, located on top of a hill and in the centre of the ‘pagus’, must have been both an oppidum or stronghold and the capital of the region. There are few remains from that period; however, we can mention the inscription on Mount Madeleine and the Barcus treasure. The Roman inscription embedded in the wall of the Madeleine chapel in Tardets reads as follows:
FANO
HERAUS
CORRITSE
HE.SACRUM
C.VAL.VALE
RIANUS
‘Fano Herauscorritse H(oc) E(rigit) Sacrum C(aius) Val(erius) Valerianus’
(Urrutibéhéty, 1977, 206).
Cayo Valerius Valerianus erected this altar consecrated to the temple of Herauscorritse." Who could this Cayo Valerius Valerianus have been who erected this altar on Madeleine Hill? Undoubtedly, he was a wealthy landowner from Upper Zuberoa, steeped in Latin culture, but who preferred to pray to the Basque god Herauscorritse rather than his counterpart in the Latin pantheon, Jupiter, in order to protect himself from storms and hail. As for the Barcus treasure, discovered in 1879 in an old pot hidden in the ground, it consisted of nearly 1,800 Celtiberian silver denarii, dating from 400 to 30 BC. It is generally believed that this treasure was abandoned by Roman soldiers during Pompey's campaigns against Sertorius.
This scarcity of remains, as well as inscriptions on tombstones, the few traces left in place names and family names, seem to reflect a certain resistance to the Roman language and customs. Romanisation was slower and less profound in the mountains and poor regions such as Zuberoa than in the plains and rich regions of the north of the Adour. Thus, the Basques were the only ones who retained their language, while their neighbours adopted that of the conquerors. Rome was far away, and the Romans must have allowed the Basques a great deal of autonomy. The Basques seemed to adapt quite well to this occupation and remained at peace for several centuries.