Encyclopédie Marikavel-Jean-Claude-EVEN/Encyclopaedia/Enciclopedia/Enzyklopädie/egkuklopaideia

d'ar gêr ! ***** à la maison ! ***** back home !

Noms de lieux

Noms de personnes



blason ou logo en attente 




Whilton Lodge


page ouverte en 2003 forum de discussion

* forum du site Marikavel : Academia Celtica

dernière mise à jour 02/06/2012 17:34:43

Définition : commune d'Angleterre, en Northamptonshire. Ancienne station romaine : Bannaventa.


Extrait de la carte Ordnance Survey : Map of Roman Britain.

Histoire : Le territoire fait apparemment partie de celui des Catuvellauni.

(en préparation)

Armoiries; blason


Patrimoine. Archéologie

seules les fenêtres ouvertes ont des liens actifs



A) BANNAVENTA : selon Rivet & Smith

- Itinéraire d'Antonin 4705 (Iter II) : BANNAVENTA, variante BENNAUENTA

- Itinéraire d'Antonin 4771 (Iter VI) : ISANNAVANTIA;

- Itinéraire d'Antonin 4795 (Iter VIII) : BANNAVANTO.

The correct form is certainly Bannaventa. The form in Iter VI is much the same in all MSS and probably goes back to the archetype, but is none the less a misreading; perhaps compare the reverse, B- for S-, in TP's Baromaci for (Ce)saromaci.

DERIVATION. For *Banna, see BANNA. Venta is a well-known problem. It is found again as a second element in Glannoventa, and as an independent word in the city-names Venta Belgarum, Venta Icenorum and Venta Silurum. Neither Latinists nor Celticists wish to claim it. Those who have favoured a Latin origin have not considered their Du Cange carefully enough. They seck a Latin venta ' market' and find it ('Locus, ubi merces vcnum exponuntur') copiously documented as a medieval Latin legal word; they also find in Du Cange venda, venta in the sense 'toll on goods for sale'. But these are mere medieval relatinisations of vernacular French vente 'sale', etc. Vente itself proceeds from a late Latin vendita abstracted from vendere, this vendita replacing classical venditio. It is just possible in theory for vendita > venta by the time of Ptolemy, who records Venta Belgarum and Venta Icenorum, in spoken Latin; elision of both pre-tonic and post-tonic vowel is recorded in the graffiti of Pompeii, e.g. oriclas for auriculas, and was repeatedly censured by grammarians. This elision is abundantly registered in Latin texts of Britain; among the few datable examples are some from the second century, such as Hercli for Herculi (RIB 2177), piissma (RIB 369), perhaps sanctissmo (RIB 600). Yet it is almost inconceivable, for several reasons, that the British Venta names should really show a derivative of this Latin vendita 'market': (I) because these names are recorded for us in official, that is classical Latin forms; (2) because Ptolemy nowhere offers us a vulgar rather than a classical form; (3) because vendita itself is not registered as a Latin word until long after the period in question, and it is virtually impossible to suppose that a usage arose within the Latin of this province alone. Most telling of all is the fact that no Continental places have Venta as an independent element. Holder I. 174 mentions Ventadour < *Ventadornum (Corrèze, France), but there are no ancient forms recorded for this and it may well be of different origin. Spanish venta, both 'sale' and 'inn, hostelry' has been temptingly mentioned in this connection but is not relevant; it is either a native word proceeding from vendita, or a later borrowing from French vente; the numerous place-names in Iberia having venta, the best-known of which is Venta de Baños (Palencia, Spain), are medieval names based on the sense 'inn', in no way Roman. Nor does an origin in venio and a sense of 'coming together, assembly' seem possible; from the supine ventum were formed adventus, conventus, but both in -us and both with prefix.

If on all these grounds the word cannot be Latin, it can hardly be other than Celtic. Certainly it is easier for a word to belong exclusively to Insular Celtic than for it to be uniquely British Latin. Williams discussed the problem under Venta Belgarum and less fully under Glannoventa. He noted that venta produced Welsh gwent (which does not of course prove that venta was originally Celtic, since many words were borrowed into British and preserved as this evolved into Welsh), and continued: 'I suggest that gwent in Welsh meant "field". Cadwent occurs for "battle", and since cad- itself means "battle" or "host", -went can mean "field". Llinwent in mid-Wales is llin- "flax", with an element -went which should mean "field". Arddunwent, in which Arddun is a woman's name, is naturally explained as " Arddun's field".' He concluded that venta was a Celtic word for' field', perhaps with a secondary meaning 'market-place', comparing magus with similar sense-development. The meaning of Bannaventa is, then, 'prominent field', 'market on the spur', or the like, acceptably enough; a similar sense for this element suits Glannoventa and the civitas-names.

Jackson in Britannia, I (1970), 80, objects very properly to the proposed Latin origin of venta and observes that 'no convincing Celtic etymology has been proposed'; that is, no Indo-European roots are visible, no cognates known in Gaulish, etc., and no derivatives of supposed British *uenta appear in Cornish or Breton. He also objects to Williams's reasoning out the sense ' field' and thinks his evidence 'slight', concluding 'One may well guess that it means something like "town", but this can be nothing but a speculation.' The problem, then, remains. It is diminished, but not solved, if we note that in both Bannaventa and Glannoventa the word is associated with impeccably Celtic elements, and that in the three  civitas-names the venta element is distinctive, unusual in our toponymy — that it may have been brought in exceptionally to designate artificial tribal centres newly founded under Roman pacification of the Belgae, Iceni and Silures. We have seen that there are no names on the Continent having Venta as an independent element; but in form there is very close resemblance to such centres as Forum Gallorum (now Gurrea, Huesca, Spain), Forum Gigurnium (now La Rua S. Esteban, Orense, Spain), and Forum Segusiavorum > Feurs (near Lyon, Rhône, France), the sense offorum in all being 'market (of the tribe)'. We lean, then, towards Williams's solution, while not denying that the problem remains.

A possible solution lies in Rostaing's identification of a pre-Indo-European *vin-' hill', which with the common infix -t- figures in such compounds as Ventabren (a mountain in Alpes-Maritimes, France) and Ventabrun (a mountain in Basses-Alpes, France), and in the names of a village and three other hills outside Provence: ETP 295-96. Possibly French Ventadour, noted earlier, belongs with these. Unfortunately there are only medieval, not ancient, records for Rostaing's examples. A sense-development 'hill > flat hill > flat place > field' would be by no means impossible; for the sense 'field > market' compare then magus. Rostaing ETP 86-87 has a further element from pre-Indo-European, *ban- *ben-, with the same meaning (see BANNA), citing among examples Beneventum of S. Italy and British 'Bennaventa' (which is more properly Bannaventa). If we are right in following Rostaing (and recalling Gaulish *benno- indicated by Dottin) we are in sight of a solution : Bannaventa is equivalent to Beneventum; both could mean 'spur-hill' or 'promontory-hill', but equally both have further sense-developments (especially *uenta within Celtic) which carry the names some distance from that primitive meaning. There are also hints in Rostaing's work which explain the reluctance of Celticists to accept venta as Celtic. The word - like *banna, probably - was indeed not Celtic originally, but was adopted by Celts in contact with pre-Celtic peoples presumably on the Mediterranean fringes, leading thereafter a somewhat tenuous existence (being little known in Gaulish) until emerging as an element with a very specific sense in Romano-British toponymy. It might also have some relation to Vendum, a place of the Japodes people (described by Strabo as a mixture of Celts and Illyrians) which J. G. von Hahn in Albanesische Studien (Jena, 1854, p. 243) identifïed as having a root meaning ' Ort, Land, Platz', a word related in turn by E. Cabej to Venta in the British names (Studi Linguistici in onore di Vittore Pisani (Brescia, 1969) I, 175-76), that is, another originally non-Celtic element absorbed into Celtic speech; see further Pokorny, Zur Urgeschichte der Celten und Illyrier (Halle, 1959), 136. On the whole, one feels that to follow Rostaing's trail produces better results, mainly because his examples are numerous and because some of them are in Gaul, providing a geographical link - though hardly a close one - with Britain.

In this context, a note by E. McClure (Archaeologia Cambrensis, 6th series, IX (1909), 239-40) has its importance. This note has been dismissed, rightly, because

McClure insisted on a Latin origin of venta, on very weak arguments. But he did have the merit of assembling relevant names. To S. Italian Beneventum he adds a Mutatio Beneventum (Bordeaux Itinerary, 55810), between Brescia and Verona in N. Italy; three modern names from Spain - Benaventa south of Astorga (Zamora), Benavent north of Lérida (Lérida), Benavente near Oviedo (Oviedo) - and one from Portugal, Benavente near the mouth of the Tagus. There are several other such names in Iberia. McClure also adds all the recent Spanish Venta names, which, as noted above, are to be removed from the discussion. McClure's interprctation of the name was erroneous, naturally, since of the S. Italian Beneventum he observes that Pliny NH III, II took the name as an auspicious one, it formerly having been ' Maleventum' — but this was mere amateur etymologising.

Even if further evidence should show that we are right in arguing the claims of one of the non-Celtic words taken into Celtic and known in British, and in the senses we have applied to it, it remains true that unlike Bannaventa, Glannoventa, which seem natural cnough, the three civitas-names with Venta are a puzzling feature considered as an isolated and un-typical naming act.

IDENTIFICATION. The Roman settlement at Whilton Lodge, Northamptonshire (SP6164); the name was presumably transferred from the Iron Age hill-fort on Borough Hill, Daventry, two miles to the south-west.


B. WHILTON LODGE : selon Eilert Ekwall.

"Woltone, DB; Whelton 12 NS, Hwelton 1254 Val.; Identique à Wheelton, en Lancashire. Vieil-Anglais hweol 'roue', désigne  ici une colline de forme ronde".  


* Eilert EKWALL : The Concise Oxford Dictionary of English Place-Names. Clarendon Press. 1980.

* ALF RIVET & Colin SMITH : The Place-Names of Roman Britain. Batsford Ltd.  London. 1979.

Liens électroniques des sites Internet traitant de Whilton Lodge / Bannaventa :  

* lien communal : 

forum de discussion

* forum du site Marikavel : Academia Celtica

hast buan, ma mignonig vas vite, mon petit ami

go fast, my little friend

Retour en tête de page