|Native to||Modern Spain and France|
|Region||Mediterranean coast of the Iberian Peninsula|
|Extinct||1st–2nd century AD|
The Iberian language was the language of an indigenous western European people identified by Greek and Roman sources who lived in the eastern and southeastern regions of the Iberian Peninsula in the pre-Migration Era (before about 375 AD). The ancient Iberians can be identified as a rather nebulous local culture between the 7th and 1st century BC. The Iberian language, like all the other Paleohispanic languages except Basque, became extinct by the 1st to 2nd centuries AD, after being gradually replaced by Latin due to the Roman conquest of the Iberian Peninsula.
Iberian is unclassified: while the scripts used to write it have been deciphered to various extents, the language itself remains largely unknown. Links with other languages have been suggested, especially the Basque language, based largely on the observed similarities between the numerical systems of the two. In contrast, the Punic language of Carthaginian settlers was Semitic, while Indo-European languages of the peninsula during the Iron Age include the now extinct Celtiberian language, Ionic Greek, and Latin, which formed the basis for modern Iberian Romance languages.
Iberian inscriptions are found along the Mediterranean coast of the Iberian Peninsula, reaching up to the river Hérault in the south of France. Important written remains have been found in Ensérune, between Narbonne and Béziers in France, in an oppidum with mixed Iberian and Celtic elements. The southern limit would be Porcuna, in Jaén (Spain), where splendid sculptures of Iberian riders have been found. Further inland the exact distribution of the Iberian language inscriptions is uncertain. It seems that the culture reached the interior through the Ebro river (Iberus in Latin) as far as Salduie (Zaragoza) but no further.
Among the pre-Roman peoples of the Iberian Peninsula the following might have spoken the Iberian language: Ausetani (northeastern Catalonia), Ilergetes (Lleida and Huesca up to the Pyrenees), Indigetes (coast of Girona), Laietani (Barcelona), Cassetani (Tarragona), Ilercavones (Murcia and Levante up to Tarragona), Edetani (Valencia, Castellón and Teruel), Contestani (Valencia, Alicante, Cartagena and Albacete), Bastetani (Granada, Almería and Murcia) and Oretani (Jaén, Ciudad Real, Albacete and Cuenca). Turduli and Turdetani are believed to be of Tartessian language.
For some scholars, such as Velaza (2006), Iberian could have been the language spoken by the autochthonous population of these territories, while for others, such as De Hoz (1993), Iberian could have been more of a lingua franca.
The origin of the language is unknown. Although Iberian ceased to be written in the 1st century AD, it may have survived in some areas until the Visigothic period (ca. 500s to 700s), according to Menéndez Pidal.
There are several theories about the geographical origin of Iberian. According to the Catalan theory, the Iberian language originated in northern Catalonia, where the earliest Iberian inscriptions are documented (600 BC) (Ullastret). Its expansion towards the north and south would have been due to broad population movements in times not long before the first written documents, from the 11th to the 10th century BC, given that the Iberian language appears homogeneous in Iberian texts and, if it were of greater antiquity, dialectalization should be evident. The presence of non-interpretable elements such as Iberian anthroponyms amongst inscriptions in this area is not considered statistically significant.
Main article: Iberian scripts
The oldest Iberian inscriptions date to the 6th century BC or maybe the 5th century BC and the latest ones date from the end of the 1st century BC or maybe the beginning of the 1st century AD. More than two thousand Iberian inscriptions are currently known. Most are short texts on ceramic with personal names, which are usually interpreted as ownership marks. Many coins minted by Iberian communities during the Roman Republic have legends in Iberian. The longest Iberian texts were made on lead plaques; the most extensive is from Yátova (Valencia) with more than six hundred signs.
Three different scripts have remained for the Iberian language:
The northeastern Iberian script is also known as the Iberian script, because it is the Iberian script most frequently used (95% of the extant texts (Untermann 1990)). The northeastern Iberian inscriptions have been found mainly in the northeastern quadrant of the Iberian Peninsula: chiefly on the coast from Languedoc-Roussillon to Alicante, but with a deep penetration into the Ebro valley. This script is almost completely deciphered.
All the paleohispanic scripts, with the exception of the Greco-Iberian alphabet, share a common distinctive typological characteristic: they use signs with syllabic value for the occlusives and signs with monophonematic value for the remaining consonants and for vowels. From a writing systems point of view they are neither alphabets nor syllabaries; rather, they are mixed scripts that are normally identified as semi-syllabaries. Regarding their origin there is no agreement among researchers; for some they are linked only to the Phoenician alphabet, while for others the Greek alphabet played a part.
The southeastern Iberian script is a semi-syllabary too, but it is more similar to the Tartessian script than to the northeastern Iberian script. The southeastern Iberian inscriptions have been found mainly in the southeastern quadrant of the Iberian Peninsula: eastern Andalusia, Murcia, Albacete, Alicante and Valencia. This script is not completely deciphered.
The Greco-Iberian alphabet is a direct adaptation of an Ionic variant of a Greek alphabet to the specificities of the Iberian language. The inscriptions that use the Greco-Iberian alphabet have been found mainly in Alicante and Murcia.
Very little is known for certain about Iberian. The investigation of the language is past its initial phase of transcription and compiling of material, and is currently in the phase of identifying grammatical elements in the texts.
The hypotheses currently proposed are unconfirmed, and are likely to remain so unless the discovery of a bilingual text allows linguists to confirm their deductions.
Iberian appears to have five vowels commonly transcribed as a e i o u. Some other languages on the peninsula such as Basque and modern Spanish also have such systems. Although five-vowel systems are extremely common all over the world, it has been suggested that this may point to a Sprachbund amongst the ancient languages of the Iberian peninsula.
The unrounded vowels (in frequency order: a, i, e) appear more frequently than the rounded vowels (u, o). Although there are indications of a nasal vowel (ḿ), this is thought to be an allophone. Judging by Greek transcriptions, it seems that there were no vowel length distinctions; if this is correct then Iberian uses the long ē (Greek: ῆτα, romanized: ēta) as opposed to the short epsilon (Greek: ἔψιλόν, romanized: épsilón).
It seems that the second element of diphthongs was always a closed vowel, as in ai (śaitabi), ei (neitin), and au (lauŕ). Untermann observed that the diphthong ui could only be found in the first cluster.
It is possible that Iberian had the semivowels /j/ (in words such as aiun or iunstir) and /w/ (only in loanwords such as diuiś from Gaulish). The fact that /w/ is lacking in native words casts doubt on whether semivowels really existed in Iberian outside of foreign borrowings and diphthongs.
There are a number of known affixes, especially applied to last names. For the Iberian language these seem to be postpositional, and apparently more agglutinative than fusional.
The best-known are the following:
There are some words for which there has been surmised a more or less probable meaning:
Thanks to the Latin Inscription of the plaque of Ascoli, which includes a list of Iberian cavalry soldiers in the Roman army (the Turma Salluitana attested in the Bronze of Ascoli), the forms of Iberian proper names have been unraveled. Iberian names are formed mainly by two interchangeable elements, each usually formed of two syllables, which are written together (Untermann 1998). For example, the element "iltiŕ" can be found in the following names: iltiŕaŕker, iltiŕbaś, iltiŕtikeŕ, tursiltiŕ, baiseiltiŕ or bekoniltiŕ. This discovery was a giant step: from this moment it was possible to identify with some kind of confidence the names of persons in the texts. Nevertheless, the list of components of Iberian names varies between researchers. The basic list comes from Untermann (1990) and was recently updated by Rodríguez Ramos (2002b); complementary data and criteria can be found in the Faria papers (the last two: 2007a and 2007b).
The following list includes some of the elements proposed as components of Iberian names: abaŕ, aibe, aile, ain, aitu, aiun, aker, albe, aloŕ, an, anaŕ, aŕbi, aŕki, aŕs, asai, aster, ata, atin, atun, aunin, auŕ, austin, baiser, balaŕ, balke, bartaś, baś, bastok, bekon, belauŕ, beleś, bels, bene, beŕ, beri, beŕon, betan, betin, bikir, bilos, bin, bir, bitu, biuŕ, bolai, boŕ, boś, boton, ekes, ekaŕ, eler, ena, esto, eten, eter, iar, iaun, ibeś, ibeis, ike, ikoŕ, iltiŕ, iltur, inte, iskeŕ, istan, iunstir, iur, kaisur, kakeŕ, kaltuŕ, kani, kaŕes, kaŕko, katu, keŕe, kibaś, kine, kitaŕ, kon, koŕo, koŕś, kuleś, kurtar, lako, lauŕ, leis, lor, lusban, nalbe, neitin, neŕse, nes, niś, nios, oŕtin, sakaŕ, sakin, saltu, śani, śar, seken, selki, sike, sili, sine, sir, situ, soket, sor, sosin, suise, taker, talsku, tan, tanek, taneś, taŕ, tarban, taŕtin, taś, tautin, teita, tekeŕ, tibaś, tikeŕ, tikirs, tikis, tileis, tolor, tuitui, tumar, tuŕś, turkir, tortin, ulti, unin, uŕke, ustain, ḿbaŕ, nḿkei.
In some cases, linguists have encountered simple names, with only one element for a suffix: BELES, AGER-DO and BIVR-NO are in the plaque of Ascoli, neitin in Ullastret and lauŕ-to, bartas-ko or śani-ko in other Iberian texts. More rarely there have been indications of an infix, which can be -i-, -ke- or -ta- (Untermann used oto-iltiŕ in front of oto-ke-iltiŕ or with AEN-I-BELES). In rare cases Untermann also encountered an element is- or o- prefacing a proper name (is-betartiker; o-tikiŕtekeŕ; O-ASAI).
In the elements that formed Iberian names it is common to encounter patterns of variation, as in eter/eten/ete with the same variations as in iltur/iltun/iltu; kere/keres as lako/lakos; or alos/alor/alo and bikis/bikir/biki).
Some Iberian onomastic elements have look-alikes in Aquitanian or Basque. This has been explained by Vascologists like Mitxelena as an "onomastic pool". However, since the meaning of most Iberian words remains opaque to date, the connection remains speculative except in a very small number of cases. An ancient sprachbund involving these two languages is deemed likely by some linguists. But as Trask notes, Basque has been of no help in translating Iberian inscriptions.
Main article: List of Spanish words of Iberian origin
Whether Iberian and Basque are two languages of the same language family is still a much-debated question. Many experts on Iberian suspect that there is a relationship of some sort between Iberian and Aquitanian, a precursor of the Basque language. But there is not enough evidence to date to ascertain whether the two languages belong to the same language family or whether the relationship is due to linguistic borrowing. Lexical and onomastic coincidences could be due to borrowing, while the similarities in the phonological structures of the two languages could be due to linguistic areal phenomena (cf. the similarities between Basque and Old Spanish in spite of being languages of two different families). More scientific studies on Iberian language are needed to shed light on this question.
From a historical perspective, the first features where a relationship between Basque and Iberian was claimed were:
Although other pairs have been proposed (such as eban, ars, -ka, -te), the meanings of these Iberian morphs are still controversial. The main arguments today which relate to coinciding surface forms between Basque and Iberian are:
In 2005 Eduardo Orduña published a study showing some Iberian compounds that according to contextual data would appear to be Iberian numerals and show striking similarities with Basque numerals. The study was expanded upon by Joan Ferrer (2007 and 2009) based on terms found on coins, stating their value, and with new combinatorial and contextual data. The comparison proposes the following:
|Iberian||Iberian meaning||Proto-Basque||Modern Basque and meaning|
|erder / erdi-||"half"||erdi "half"|
|ban||"one"||*badV / *bade?||bat "one" (but cf -n final compound forms such as bana "one each")|
|bi / bin||"two"||biga||bi (older biga) "two" (also cf -n final compound forms such as bina "two each")|
|borste / bors||"five"||bortz / *bortzV?||bost (older bortz) "five"|
|abaŕ / baŕ||"ten"||*[h]anbar ?||hamar "ten"|
The basis of this theory is better understood if we compare some of the attested Iberian compounds with Basque complex numbers (the dots denote morpheme boundaries and are not normally written in Basque; also note that the final -r in numbers 3 and 4 also occurs in bound forms in Basque i.e. hirur- and laur-):
|Iberian word||Basque comparison||Basque Meaning||Basque analysis|
|oŕkei-irur||hogei.ta.hiru||"twenty three"||"20 and 3"|
|oŕkei-ke-laur||hogei.ta.lau||"twenty four"||"20 and 4"|
|oŕkei-abaŕ||hogei.ta.(ha)mar||"thirty"||"20 and 10"|
|oŕkei-(a)baŕ-ban||hogei.ta.(ha)maika||"thirty one"||"20 and 11"|
Even so, Orduña does not claim this comparison to be a proof of a family relation between Iberian and Basque, but rather owing to Iberian loanwords in the Basque language. In contrast, Ferrer believes that the similarities could be caused due to both the genetic relationship or the loan, but indicates that the loan of the entire system of numerals is rare (but has known to occur such as the case of Middle Chinese numeral being borrowed wholesale into Vietnamese, Japanese, Korean and Thai).
Joseba Lakarra (2010) has rejected both hypotheses: loan or genetic relationship. Lakarra’s arguments focus almost exclusively on the field of Basque historical grammar, but also arguments, following de Hoz (1993) hypothesis, that the hypothesis of the borrowing has already turned out implausible due to the limited and remote extension of the territory where Iberian was spoken as first language in South-East Spain.
Javier de Hoz (2011, pp. 196–198) considers plausible the internal contextual and combinatorial arguments that would support the hypothesis that these Iberian elements could be interpreted as numerals. In fact, concerning the specific values, he considers valid the proposed equivalences between Iberian ban with 'one' and between Iberian erder with 'half', according to the marks of value found in coins, while he considers that the rest of the proposed equivalences are a working hypothesis. Regarding the equivalence between the possible Iberian numerals and the Basque numerals, he agrees with Lakarra (2010) that the shape of the documented Iberian forms does not fit the expected Proto-Basque forms. Finally, he considers that the greatest difficulty in accepting this hypothesis is, paradoxically, its extent and systematic nature, because if it was correct, it would result in a close relationship between Iberian and Basque, which should allow the identification of other relationships between Iberian and Basque subsystems, as clearly as this one, relationships that no investigator using reasonable linguistic arguments has been able to identify.
Eduardo Orduña (2011) insists that the Iberian elements proposed as numerals are not only similar to the Basque numerals, but also combine as numerals and appear in contexts where numerals are expected. He observes Lakarra (2010) not dispute these arguments [neither does de Hoz (2010)]. As regards the de Hoz hypothesis about considering the Iberian language as a lingua franca, Orduña remarks its hypothetical character, although Lakarra presents that hypothesis as an established fact. The problems of this hypothesis have been collected by Ferrer (2013) in a later work. Regarding the phonetic difficulties indicated by Lakarra, Orduña argues that its proposals are compatible with the Proto-Basque reconstructed of Michelena, which is for chronology and security the reconstruction that an iberist has to consider, while the hypothesis of internal Basque reconstruction of Lakarra has a vague chronology and a much lower degree of security. Finally, contrary to his first opinion in favor of the loan, concludes that the most economical hypothesis to explain the similarities between the Iberian numeral system and the Basque numeral system is the genetic relationship.
Francisco Villar (2014, 259) notes that the similarities between Iberian numerals and Basque numerals are of the same order as those documented among Indo-European languages and consequently argues that the only sustainable hypothesis at this point is the genetic relationship between Iberian and Basque. Villar also believes that if the reconstruction of Proto-Basque proposed by Lakarra (2010) is incompatible with the evidence derived from the numerals, the reconstruction must be corrected, as like all reconstructions, is hypothetical and perfectible.