The Iberian language and Palaeohispanic scripts
Iberian was the language of the peoples of the Mediterranean coast of Spain and southern France before Latin replaced it. It is the best-attested of the Palaeohispanic languages, the pre-Roman languages of the peninsula, with about 2,300 inscriptions on lead sheets, pottery, stone, coins and silverware, dating from the end of the fifth century BC to the first century AD[1]. Most are written in a native script, the north-eastern Iberian semi-syllabary, in which vowels and continuous consonants have their own letters but stops are written as syllables[1].
The script is read. Manuel Gómez-Moreno worked out its semi-syllabic structure and the values of most signs in 1922, and later work refined it, notably Joan Ferrer i Jané's demonstration that the script originally used an extra stroke to mark voiceless dental and velar stops[1][2]. The language is not. It is not Indo-European, it has no well-understood close relative, and bilingual inscriptions are few and short[1]. Scholars can identify personal names, which are built from about 150 recurring elements, place names on coins, some suffixes, funerary formulas and probable numerals, but cannot translate a longer text[1].
The question that divides the field is whether Iberian is related to Basque. A list of Iberian elements that look like the Basque numerals, proposed in 2005 and 2009, is accepted by many as more than chance, but whether it reflects common descent or contact is disputed[3][4][1]. A computational decipherment model tested in 2021 found no strong signal of a relationship[5]. Of the 4 questions on this site, 0 are answered. The most likely answer to “How far did the dual system of the Iberian scripts extend?”, not yet confirmed: duality marked only the voiced and voiceless dental and velar stops.
Key questionsset by the editors
The object
Iberian texts are found from the River Hérault in Languedoc to the province of Almería, with a strong extension up the Ebro valley[1]. About 2,000 are in the north-eastern script, about 70 in the south-eastern script, and a small group in an adaptation of the Greek alphabet, the Graeco-Iberian script[1]. The oldest datable inscription is a graffito on an Attic cup from Ullastret, near the end of the fifth century BC; the latest date from the Augustan period and the first century AD[1]. The longest texts are on lead sheets, many of them commercial or private documents[1].
Two corpora hold the material. Jürgen Untermann's Monumenta Linguarum Hispanicarum edits each inscription with its analysis, and the Hesperia databank, maintained by a team from the Complutense, Basque Country, Zaragoza and Barcelona universities, records all Palaeohispanic inscriptions, coin legends and native names with texts, critical apparatus, images and archaeological context[6][7][1].
Timeline
What is known
The sound system can be partly reconstructed: five vowels, three series of stops of which the dental and velar series show a two-way contrast, and a set of diphthongs[1]. Personal names were understood from the Ascoli Bronze, a Roman grant of citizenship of 89 BC to Iberian cavalrymen, whose names show that most Iberian names join two elements from a shared stock of more than 150[1]. Bilingual funerary inscriptions from Tarragona suggest that the formula aŕe teki or aŕe take renders the Latin hic situs est, "lies here"[1]. Coin legends give place names and suffixes such as -ken and -sken[1].
Numerals are the most discussed lexical result. Eduardo Orduña in 2005 and Ferrer i Jané in 2009 proposed that Iberian ban, bi(n), lau(r), bors(te), śei, sisbi, sorse, (a)baŕ and oŕkei mean 1, 2, 4, 5, 6, 7, 8, 10 and 20, matching Basque bat, bi, lau(r), bortz, sei, zazpi, zortzi, (h)amar and (h)ogei[3][1]. A lead sheet from Casinos (Valencia) with numerals written as words and metrological expressions was published in support[8]. The 2020 survey judged that the formal similarities "cannot be random", while noting the hypothesis is still under discussion[1].
What has been tested
The numeral hypothesis has been tested against reconstructed Proto-Basque. Joseba Lakarra argued in 2010 that the proposed Iberian forms do not fit reconstructed Proto-Basque, and offered etymologies that make several Basque numerals later formations[4]. Orduña replied that the Iberian numerals are compatible with the best-established Proto-Basque forms and that some of Lakarra's etymologies are improbable[9]. In December 2025 Orduña restated the case for a Basque-Iberian relationship in nominal and verbal morphology and numerals, and proposed a set of Iberian kinship terms related to Basque[10].
A computational test was run in 2021. Luo and colleagues built a decipherment model that jointly segments an unsegmented script and aligns it with a known language using phonetic embeddings. It identified the correct relatives of Gothic and Ugaritic in controlled tests, but for Iberian it found no strong evidence that Basque is related, in line with the cautious scholarly view[5].
A new find bears on the question. A bronze hand from the Iron Age site of Irulegi in Navarre carries an inscription in a variant of the Palaeohispanic script. Its editors read it as Vasconic, the first text in an ancestor of Basque, and as an apotropaic wish for good fortune, comparing its first word with Basque zorion[11]. A comparative-method test of that comparison found the evidence inconclusive[12].
The readings
Three positions are held on Iberian's affiliation. The first, the Basque-Iberian hypothesis, holds that Iberian and Basque, or its ancient form Aquitanian, are related, on the evidence of the numerals, some morphology and shared name elements[10]. The second holds that the resemblances are not evidence of common descent, because the Iberian forms do not fit reconstructed Proto-Basque; the computational test points the same way[4][5]. Contact between neighbouring languages could explain shared numerals and name elements without descent, but no study has set out that case in detail. The third is the cautious middle: a relationship is not impossible, but at present there is "no point" in comparing the two languages to translate Iberian texts[1]. On the script, the main open point is how far the dual marking extended, to vowels and continuants as Ferrer has proposed, or only to the stops[1][2].
Published claims
| Year | Who | Claimed | Status | Reason |
|---|---|---|---|---|
| 1821 | Wilhelm von Humboldt[13] | Basque descends from the language of the ancient Iberians, so the old place names of Hispania can be explained from Basque. | Ruled out | Once the script was read, Iberian texts proved impossible to understand through Basque, and the specialist survey sees no point in using Basque to translate them[1]. |
| 1922 | Manuel Gómez-Moreno[14] | The Iberian script is a semi-syllabary: vowels and continuous consonants are letters, and stops are written as syllables. | Accepted | His sign values are the basis of every reading of the texts since, refined but not overturned[1]. |
| 2009 | Joan Ferrer i Jané[3] | Following Orduña, the elements ban, bi(n), lau(r), bors(te), śei, sisbi, sorse, (a)baŕ and oŕkei are Iberian numerals for 1 to 20. | Accepted | A lead sheet with numerals written as words supports the values, and the standard survey judges the pattern cannot be random[8][1]. |
| 2010 | Joan Ferrer i Jané[2] | The south-eastern Iberian script also had a dual system, wider than the north-eastern one, marking a sibilant, a vibrant and the nasal as well as the stops. | Still possible | Duality in the stops is accepted, but how far it extended to other sounds is still under discussion[1]. |
| 2010 | Joseba Lakarra[4] | The proposed Iberian numerals do not fit reconstructed Proto-Basque, and several Basque numerals are later formations. | Disputed | Orduña replied that the Iberian forms fit the best-established Proto-Basque reconstructions, and the question is still open[9][1]. |
| 2011 | Eduardo Orduña[9] | The Iberian numerals are compatible with the best-established Proto-Basque forms, and some of Lakarra's etymologies are improbable. | Disputed | Bascologists working from Proto-Basque reconstruction still reject the comparison, and the survey calls it unresolved[4][1]. |
| 2014 | Joan Ferrer i Jané and Vicent Escrivà[8] | A lead sheet from Casinos writes numerals as words beside metrological expressions, confirming their values. | Accepted | The standard survey accepts the numeral values that the Casinos lead supports[1]. |
| 2020 | Noemí Moncunill and Javier Velaza[1] | The Iberian and Basque numerals are too alike to be chance, but at present there is no point in using Basque to translate Iberian texts. | Most likely | It is the position of the standard survey, and the 2021 computational test found nothing to move it[5]. |
| 2021 | Jiaming Luo and colleagues[5] | A decipherment model that finds the right relatives for Gothic and Ugaritic finds no strong evidence that Basque is related to Iberian. | Still possible | It is one model, and its result agrees with the cautious view of the specialist survey rather than settling the question[1]. |
| 2023 | Julen Manterola and Céline Mounole[12] | By the comparative method, the Irulegi word sorion cannot be shown to be Basque zorion, so a Basque reading is not confirmed. | Still possible | The editors maintained the Vasconic reading in their 2024 publication, and no longer text has been found to decide it[11]. |
| 2024 | Mattin Aiestaran, Javier Velaza, Joaquín Gorrochategui and Carmen Usúa[11] | The Irulegi bronze hand is inscribed in Vasconic, an ancestor of Basque, with a wish for good fortune comparable to Basque zorion. | Disputed | A comparative-method test of the key word found the evidence inconclusive[12]. |
| 2025 | Eduardo Orduña[10] | Besides the numerals, Iberian shares nominal and verbal morphology and a set of kinship terms with Basque. | Still possible | It is too recent to have been answered; the test it has to pass is Lakarra's objection from reconstructed Proto-Basque[4]. |
What would settle it
A substantial bilingual, an Iberian text with a Latin or Greek translation of more than a formula, would give the first secure vocabulary and grammar. Lead sheets with numerals and totals, of which more are found each decade, could confirm the numerals by arithmetic, as the fraction values of Linear A were confirmed. On affiliation, a test that compares Iberian with reconstructed Proto-Basque using regular sound correspondences, with a control of unrelated languages, would show whether the matches exceed chance. The Hesperia database, if released as structured data, would let such tests run on the whole corpus.
Questions
0 of 4 questions are answered.
Evidence 14 findings
Artifacts
The films, photos, papers and objects this mystery rests on, each with its own key and its distance from the original.
See also
References
- 1.Moncunill Martí, N., Velaza Frías, J. 2020. Iberian. Palaeohispanica 20, 591-629.doicited 21 times hereSRC0D6C2WA618XX
- 2.Ferrer i Jané, J. 2010. El sistema dual de l'escriptura ibèrica sud-oriental. Veleia 27.doicited 5 times hereSRC0D6C2XH81VS1
- 3.Ferrer i Jané, J. 2009. El sistema de numerales ibérico: avances en su conocimiento. Palaeohispanica 9, 451-479.doicited 5 times hereSRC0D6C2X1M599J
- 4.Lakarra, J. 2010. Haches, diptongos y otros detalles de alguna importancia: notas sobre numerales (proto)vascos y comparación vasco-ibérica. Veleia 27.doicited 5 times hereSRC0D6C2XS2447M
- 5.Luo, J., Hartmann, F., Santus, E. et al. 2021. Deciphering Undersegmented Ancient Scripts Using Phonetic Prior. Transactions of the Association for Computational Linguistics.doicited 9 times hereSRC0D8MS0BA506A
- 6.Untermann, J. 1990. Monumenta Linguarum Hispanicarum III: Die iberischen Inschriften aus Spanien. Reichert, Wiesbaden.cited 3 times hereSRC0D6C2WST6SNA
- 7.Hesperia team (Universidad Complutense de Madrid and partners) n.d.. Hesperia: Banco de Datos de Lenguas Paleohispánicas. hesperia.ucm.es.cited 6 times hereSRC0D6C2WJ01T8F
- 8.Ferrer i Jané, J., Escrivà Torres, V. 2014. Un plomo ibérico de Casinos (Valencia) con numerales léxicos y expresiones metrológicas. Palaeohispanica 14, 205-227.doicited 4 times hereSRC0D6C2X9E22GP
- 9.Orduña Aznar, E. 2011. Los numerales ibéricos y el protovasco. Veleia 28, 125-139.doicited 1 time hereSRC0D6C2Y0W2CXS
- 10.Orduña Aznar, E. 2025. La relación entre el vasco y el ibérico: más allá de los numerales. Palaeohispanica 25, 329-357.doicited 4 times hereSRC0D6C2Y8P4903
- 11.Aiestaran, M., Velaza, J., Gorrochategui, J. et al. 2024. A Vasconic inscription on a bronze hand: writing and rituality in the Iron Age Irulegi settlement in the Ebro Valley. Antiquity 98(397), 66-84.doicited 7 times hereSRC0D6C2YRA0YA5
- 12.Manterola, J., Mounole, C. 2023. An assessment of the Basque Interpretation Hypothesis on the inscription contained in the hand of Irulegi. Fontes Linguae Vasconum 136, 583-595.doicited 3 times hereSRC0D6C2Z0455K6
- 13.Humboldt, W. 1821. Prüfung der Untersuchungen über die Urbewohner Hispaniens vermittelst der vaskischen Sprache. Berlin.cited 1 time hereSRC0D6C2Z7Y4ZFV
- 14.Gómez-Moreno, M. 1922. De epigrafía ibérica: el plomo de Alcoy. Revista de Filología Española 9, 341-366.cited 1 time hereSRC0D6C2ZFR7T96