Iberian language

Geographic distribution

The Iberian language was widely spoken along the Mediterranean coast of the Iberian Peninsula.

In the north, the Iberian language inscriptions reached the south of France up to the Hérault river. 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. Towards 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 not farther.

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.

History

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, according to Menéndez Pidal.

Writing

The oldest Iberian inscriptions date to the 4th 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:

Northeastern Iberian script

Dual variant (4th century BC and 3rd century BC)

Non-dual variant (2nd century BC and 1st century BC)

Southeastern Iberian script

Greco-Iberian alphabet (most of the aforementioned Leads of La Serreta are written in this version).

Northeastern (or Levantine) Iberian script

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.

Southeastern (or Meridional) Iberian script

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.

Greco-Iberian alphabet

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.

Current extent of linguistic knowledge

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.

Vowels

Iberian appears to have 5 vowels commonly transcribed as a e i o u. Other modern languages on the peninsula such as Basque and 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 front vowels (a, e, i) appear more frequently than the back vowels. 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 ēta) as opposed to the short epsilon (Greek ἔψιλόν épsilón).

Diphthongs

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.

Semivowels

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.

Consonants

Vibrants: There are two vibrants r and ŕ. Iberian specialists do not agree about the phonetic values assigned to either vibrant. Correa (1994) hypothesized that ŕ was an alveolar flap [ɾ] and r was a "compound vibrant", that is, a trill [r]. Later, Rodríguez Ramos (2004) suggested that ŕ was an alveolar flap [ɾ] and r is a retroflex flap [ɽ] in line with Ballester (2001) who thought that r represents a uvular fricative [ʁ]. However, Ballester (2005) later changed his hypothesis and took r for an alveolar flap [ɾ] and ŕ for the alveolar trill [r]. Neither r nor ŕ occurs word-initially, which is also the case in Basque.

Sibilants: There are two sibilants s and ś. The distinction is unclear, and there are multiple proposals. Ballester (2001) theorizes that s was an alveolar [s] and ś was an alveolo-palatal [ɕ]. Rodríguez Ramos (2004) proposes that ś was alveolar [s] and s was an affricate, either dental [ts] or palatal [tʃ] (like English "ch"). This proposal coincides with the observation by Correa on adaptations of Gallic names in Iberian texts.

Laterals: The lateral l is normally interpreted as [l]. It is extremely rare in final position and it could be that the distribution is on occasion complementary with ŕ: aŕikal-er ~ aŕikaŕ-bi.

Nasals:

The n was probably alveolar [n].

m: Researchers studying Iberian do not agree on the kind of nasal represented by this letter. The letter m rarely occurs word-initially. Velaza (1996) hypothesizes it could be an allophone of medial n, as shown in the example of iumstir/iunstir. José A. Correa (1999) suggests it may be a geminate or strong nasal. Ballester (2001) considers it to be a labialized nasal in Iberian and in Celtiberian. Rodríguez Ramos (2004) mentions that it could be an allophone of n where it nasalizes the preceding vowel.

There is some controversy over the sound ḿ. While it's thought to be some type of nasal, there is no certainty as to its value. Several linguists agree on the value [na], based on similarities with texts written in the Greek alphabet, as there are similarities between the suffixes -ḿi / -nai, and in the onomastic elements -ḿbar- / -nabar-. Another part of this theory seems to contradict itself with the transcription of ḿbar-beleś into Latin as VMARBELES. Correa (1999) proposes that it was a labialized nasal. It is not even clear that the sign is always pronounced in the same form. Rodríguez Ramos (2004) considers it a nasalized vowel, produced by progressive nasalization.

Plosives: There are five plosives.

The evidence indicates the non-existence of a phoneme p as it is not documented either in the Greek alphabet or in the dual Iberian systems. It is only found in Latin inscriptions naming native Iberians and is thought to be an allophone of b. It has been suggested that the phoneme b would on occasions have been pronounced similar to w (this would be explained by the frequency of the sign bu), and as such it could have had a nasalized pronunciation.

Morphology

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.

-ar: applied to proper names to mark possession. -en: of a similar or identical use to -ar. -ka: seems to indicate the person who receives something -te: seems to indicate the ergative -ku: seems to indicate the ablative Possibly related to the Basque local genitive -ko. -ken / -sken: usually understood as genitive plural because of its use on coins in ethnical names (with parallels on Latin and Greek coins). -k: has been proposed on occasions to mark the plural. -k is a plural marker in Basque.

Lexicon

There are some words for which there has been surmised a more or less probable meaning:

aŕe take as akin to the Latin formula hic est situs ("here he is") (Untermann 1990, 194) because of a bilingual inscription from Tarragona C.18.6

eban and ebanen as equivalent to the Latin coeravit ("he cared [to be done]") in tombstones (Untermann 1990, 194), because of a bilingual inscription from Sagunto F.11.8

iltiŕ and iltun as typical Iberian toponyms for city names, meaning something like "city" / "town"

ekiar: verb or verbal noun with a meaning like "to do" / "to make" compared with the Basque verb egin (Beltrán 1942; Correa 1994, 284). likine-te ekiar usekerte-ku with a meaning akin to "made by Likinos of Osicerda" (Correa 1994, 282)

seltar and siltar as meaning something like "tomb" on tombstones (Untermann 1990, 194).

śalir as meaning something like "money" / "coin", because of its use in coins (as iltiŕta-śalir-ban) and its use in lead plaque inscriptions besides numbers and quantities (Untermann 1990, 191).

Personal names

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), 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.

Iberian and Basque

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 their 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:

the suffixes -sken / -ken on Iberian coins (which were compared to the genitive plural on similar ancient coins) with the Basque plural (-k) and genitive (-en) endings

Iberian town names containing ili (particularly iliberri), where parallels were drawn with Basque hiri ("town") and berri ("new").

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:

Phonetics: Proto-Basque phonology, first proposed by Michelena, appears to be very similar to what is known about the Iberian phonological system. It has been claimed that the lack of /m/, common to both Proto-Basque and Iberian, is especially significant ).

Onomastics: Aquitanian-Latin inscriptions contain personal and deity names which can clearly be related to modern Basque words, but also show structural and lexical resemblances with Iberian personal names. But Iberian influence on the Aquitanian name system, rather than a genetic link, cannot be dismissed either.

In Iberian iltiŕ and iltur, ili is read "city". Modern Basque hiri, "city", is derived from the very similar Proto-Basque root *ili

The Iberian genitive ending -en and maybe the genitive plural-(s)ken, compared to the Basque genitive -en and the Basque genitive plural *ag-en as reconstructed by Michelena. But Michelena himself was sceptical about this comparison.

An Iberian formula which frequently appears on tombstones, aŕe take, with variants such as aŕe teike, which on a bilingual inscription from Tarragona may be equivalent to the Latin hic situs est ("here is"), as proposed by Hübner. This was compared by Schuchardt (1907) with Basque "(h)ara dago" “there is/stays”.

The Iberian word ekiar, explained as something akin to “he made”, proposed to be linked with the Basque verb ‘egin’ "make"

The Iberian word śalir explained as “money”, “coin” or “value”, proposed to be linked to Basque word ‘sari’ (probably Proto-Basque *sali) meaning “value”, “payment”, “reward”.

Numerals

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:

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-):

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 have 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 protobasque 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 they combine as numerals and appear in contexts where numerals are expected. And remarks that Lakarra (2010) does 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.