Basque esponda, ezponda “slope”
Azkue lists the following:
(AN-Arakil, BN-Salazar, Z, R) esponda “slope of the face of a wall or piece of land”
(G-Burunda-Etxandi-Aranaz, BN-Salazar, Z, R) ezponda “slope, riverbank”
Basque esponda, ezponda seems to be present right across Euskal Herria though it is missing from Bizkaia. The variance between orthographic /s/ and /z/ (the apical and laminal sibilants, respectively) is a little bit mysterious. As we shall see, the /s/ form appears to be original.
Attempts have been made to derive Basque esponda, ezponda from Latin sponda “bedframe, bed, couch”, but this is misconceived. A slope is the opposite of a bed, which is generally something flat or at least low-lying (like a bed for sleeping in, or a river-bed, a sea-bed or a flower-bed).
The misconception is a very old one. An 11th century scribe called Navarrenx (Pyrenées-Atlantiques) sponda Navarrensis. This is a twofold false Latinisation. First, the scribe mistakes Basque esponda, ezponda for Latin sponda. Secondly, he confuses the toponymic suffix, -enx (quite frequent in this region), with the Latin local genitive, -ensis. Then in the 16th century, the poet, Joanes Ezponda (1557-1595) was known in French as Jean de Sponde.
Basque esponda, ezponda seems to have spread into neighbouring Romance languages. There is a Gascon esponne, which appears to have a similar meaning to Basque. It is often to be found as a name for steep mountainsides in the Western French Pyrenées. Italian sponda “slope, edge” cannot be derived from Latin sponda “bedframe, bed, couch”, because the meanings are opposites. The appearance of these words in Romance languages is an example of a superstrate borrowing. Western Romance languages share numerous words and forms that do not derive from Latin. Most are Germanic, a few are Arabic and one or two are Vasconic. Within Romance they are capable of spreading around quite freely. An example is Spanish zanca, Italian zanca, both “leg”, and Basque, zanka, zango, etc, “leg”. Another is Spanish guinda, French guigne, both “sour cherry”, and Basque ginda “cherry”.
Dutch sponde “bed” does actually derive from Latin sponda.
Basque esponda, ezponda plays little part in the toponymy of Euskal Herria. However, it is quite possibly present in the following toponyms further afield:
Esponzues, a village of Cantabria overlooking the Pas Valley south of Santander: <*espon-su “place of the slope”.
This would suggest an Iberian *espon. It is possible that the final -da in esponda, ezponda is augmentative -ta/-da. (It is important to note here that Iberian west of the Nerbioi discontinuity is evidenced only in onomastics. There are no native inscriptions.)
Espondeilhan (Herault, France), which was recorded as Espondeilla in 1170: <*esponda-ili “settlement on the slope”, to which the Roman -anum has been added. (The village is in fact located on a small hill rising above the coastal plain.)
In Iberian, there appears to have been, in addition to the *espon preserved in Esponzues, a form that had undergone component metathesis. It is recorded in both Roman and East Iberian Scripts, but is missing from the modern language. In some dialects at least, an original *es-pon appears to have become *bon-es. Or perhaps an original *bon-es became *espon. Consider the following:
Roman Script, Upper Garonne Valley
These texts were written in Latin but include many Iberian anthroponyms and theonyms. The following anthroponyms consist of Iberian *bones alone:
BONEXSI (CIL 13 00178)
BONNEXI (CIL 13 00071)
In the Upper Garonne texts, XS and X appear to denote the sounds represented in Basque by /ts/, /tz/, /x/ and /tx/.
East Iberian Script, Catalonia
The following bipartite anthroponyms are recorded on the Lead Foil of Palamos, a roll of donors of grave goods:
Adinbones (adin “age, intelligence”)
Bilosbones (bilos “round”)
The following bipartite anthroponym is recorded on the Lead Foil of Ullastret, another roll of donors of grave goods:
Biurbones (biur “bent”)
The No 2 S-sign of the East Iberian Script appears to denote the sounds represented in Basque by /s/, /ts/, /x/ and /tx/.
Note how the Upper Garonne and Catalonia agree on the form of this word but disagree with Basque. This is one of a group of words where this happens. The language of the Upper Garonne is demonstrably closer to the language of Catalonia than it is to modern Basque. Yet the bunch of cranks that we call orthodox Vascologists insist that the Upper Garonne language is the antecedent of Basque while Iberian is wholly unrelated to Basque.
*bon-es puts in a single appearance within the borders of Euskal Herria. Consider the following:
Arbonies (Navarra), a village in the far east of Navarra close to Lumbier which was recorded as Arbones in 1110: <*ar-bones “stone slope”. The meaning of Iberian bones is thereby narrowed down do something geomorphological that can be made of stone, like a slope.
Component metathesis is the placing in a different order perceived lexical components. They may not actually be lexical components, they may just be perceived as such. Sometimes it is quite hard to say which coupling is correct. So when one is dealing with a component metathesis it can be very difficult to look for cognates in Vasconic languages further afield, let alone in other non-Vasconic Dene-Caucasian languages.
Greek sfondülios “vertebra, spinal column” might possibly derive from a Pelasgian cognate of Basque esponda, ezponda. If so, *es-pon-da is likely to be original and *bones secondary.
However, if we look for possible cognates in other non-Vasconic Dene-Caucasian languages, we might conclude that Basque esponda, ezponda is a tripartite compound dating back to the Proto-Vasconic stage, and that -pon- or -bon- is the primary component. Consider the following:
Sino-Tibetan
Preclassic Old Chinese paŋ “side, quarter, place, region, square, regular thing, etc”
Lushai paŋ “body, side, flank”
Limbu ku-buŋ “foot of a mountain, bottom”
Sadly, the data that would allow a complete analysis of Basque esponda, ezponda have yet to be unearthed.