The Raetic language

From Thesaurus Inscriptionum Raeticarum
Jump to navigationJump to search

The Raetic language as documented in inscriptions written in the alphabets of Sanzeno (Bolzano) and Magrè has turned out to be much more homogenous than expected (or hoped) (see Modern research on the Raeti and Raetic). Despite the fact that its uniformity, though long suspected, has been demonstrated only in the 1990ies, linguistic criteria today make for a better basis for the definition of the Raetic corpus than epigraphic parameters.

As is generally the case with fragmentary languages, our knowledge of Raetic is restricted by the limited input not only in quantity, but also in quality. The prevalent Raetic text types are:

  • votive inscriptions / dedications, mostly formulaic,
  • owner's inscriptions,
  • possibly funerary inscriptions.

The handful of longer and more complex inscriptions remain mostly obscure due to lack of material for comparison; on the other end of the scale, a great number of inscribed characters do not appear to encode linguistic entities and cannot at this point be used for the purpose of researching the Raetic language. As a consequence, we primarily know names. The identification of derivational and inflectional suffixes and of lexical items is usually owed to comparison with Etruscan.



The absence of Omikron from Raetic inscriptions demonstrates that Raetic had a quadripartite vowel system like Etruscan. The lack of an /o/-phoneme results in increased articulatory leeway for /a/ and /u/.

front/unround back/round
high /i/ /u/
non-high /e/ /a/

For Etruscan, it has been argued that the original vowel system (in Archaic Etruscan) was symmetrical, featuring rounded /a/. Lowering and unrounding of /a/ during the 6th century appears to have led to lowering of /u/ during the 2nd/1st century. (Details and argumentation in Wallace 2008: 32 f.; differently De Simone 1968: 48 f.) A similar development may have taken place in Raetic. Detailed research into foreign language material in Raetic inscriptions, especially concerning the question what Raetic substitutes for non-Raetic /o(:)/, is still necessary. So far, there are no clear pieces of evidence for /o(:)/ in the Raetic corpus. Raetic φausu quite likely is the counterpart of the name bauso found in Latin inscriptions, but we do not know from which language bauso was borrowed into Raetic; besides, bauso itself is a latinised name, and when this name was latinised, it was clearly made a Latin n-stem; and it is impossible to tell what was the quality of the last vowel before the name was latinised. There is one instance of a letter Omikron in a Raetic context, but this is from a Latinoid inscription, which bears the name ossurie. However, this single instance does not necessarily mean that there was phonemic /o/ in Raetic: instead, this Omikron might as well represent a rounded allophone of Raetic /a/ or the like.

There is no evidence that there were distinctively long vowels in Raetic. After all, Archaic Etruscan had no distictively long vowels, and length only became a distinctive feature after long vowels had emerged via monophthongisation processes (Wallace 2008: 33). Not enough is known about Raetic historical phonology to judge the situation in Raetic; for the time being, vowels in Raetic will be treated as not distinguished by length.


Nominal inflection

As shown in the table below, a considerable number of the endings of nominal inflection of Etruscan are attested in Raetic. The table gives three paradigmata: The endings reconstructed for Proto-Etruscan (pre-apocope) (following Rix 1985: 23 ff.), which are used for the morpheme pages in TIR, the endings attested in Etruscan inscriptions, and the endings attested in Raetic inscriptions. While there is only one allomorph for the casus rectus and the locative, respectively, the genitive, and consequently the cases based on it (ablative = genitivus genitivi, pertinentive = locativus genitivi), have two allomorphs each.

Etruscan mm Raetic
mm Nom./Acc. Locative Genitive Ablative Pertinentive mm mm Nom./Acc. Locative Genitive Ablative Pertinentive
reconstructed -i I -si I -si-s I -si-i mm
II -la II -la-s II -la-i mm
synchronical -i I -s I -s I -si mm I -s I -s ? I -si
II -l II -las II -le mm II -l ? II -le

Due to the fact that nouns in Raetic are mostly identified via their endings, the null morpheme of the casus rectus can be conclusively argued in only one instance, where a plural ending marks the nominal base. Another possible case is aχvil, if correctly identified as a noun. Locatives in -i have not so far been identified, the only feasible option is tanini (from taniun). The genitive I -s is amply documented, while the genitive II -l(a) is more problematic, as both Neo-Etruscan -(a)l and Archaic Etruscan -a can be argued to be attested (see the morpheme page). The most frequent case in Raetic inscriptions is the pertinentive, pertinentive I -si being used with individual names, pertinentive II -le mostly with patronyms in -nu (for exceptions see the morpheme page). The single attestation of the ablative I -s (with umlaut of the stem vowel) is somewhat doubtful. The distribution of allomorphs has not so far been made sense of in Etruscan, the same is true for the much more sparsely documented Raetic.

Also attested is the plural ending -r(a), both alone and in a morpheme syntagma with the pertinentive I. All securely attested endings are formally identical to their Etruscan counterparts, having undergone pre-historic apocope (Rix 1985: 117). Additionally, the pertinentive plural form demonstrates that Raetic, like Etruscan, inflected by agglutination.

Verbal inflection

  • 3rd singular preterite -ke, as in Etruscan; it is questionable whether the spelling with Kappa or Khi (in some cases) reflects a distinction of voice.


  • Derivational suffix -na / -nu, used in Raetic for deriving patro-/matronyms, comparable with the Etruscan derivational suffix -na.
  • Derivational suffix -u forming nominal derivations from verbal stems in -ke.


A small subset of recurring and non-recurring words can be identified with lexemes known from Etruscan.

  • þinaχe (-ke preterite) ~ Etr. zinake 'made'.
  • eluku (-ku participle), possibly to Etr. ilucu 'sacrifice' (?) → 'sacrificed'.
  • aχvil ~ Etr. akvil 'present'.
  • sφura ~ Etr. spura 'community'.
  • þal ~ Etr. zal 'two'.
  • possibly enclitic -ka ~ Etr. -ka 'and'.

In the absence of Etruscan comparanda, we have to rely on our often doubtful segmentation of inscriptions and the evidence of recurring sequences. The appearance of known suffixes may help.

  • utiku (-ku participle), probably 'given', 'donated' vel. sim.
  • taniun, attested several times, but obscure.
  • terisna, attested several times; probably formed with the derivational suffix -na, but otherwise obscure.
  • φuter, clearly a noun in the plural (see above), but semantically opaque.

For more sequences which are likely to represent words, see Category:Word.


Personal names

A full Raetic name consists of two parts: an individual name and a patronym (or possibly also a metronym). The latter is derived from an individual name by suffixation of -nu or -na; on the questions of the relation to Etruscan -na and whether the two variants reflect gender, see the morpheme pages. We have reason to believe that the Raetic patronymic system was productive at the time of it's documentation (Rix 1998: 18 f.) – a number of names are attested both as individual names and as base of a patronym, but most important is the testimony of ST-1, ST-2 and ST-3 with the names of three related men. The following names are attested as both individual name and patronym:

Names attested more than once as individual names:

Names attested only once, as individual names or bases for patronyms, some dubious to varying degrees:

It can be remarked that most of the names end in a vowel, or, more precisely, in -i/-e/-ie; -a and -u are also represented. Of the six names in -u, two are dubious; of two, the auslaut is uncertain. Of the remaining two, one is attested elsewhere with an auslaut -o, the other is probably Celtic. Auslauting -u might therefore be suspected to appertain to non-Raetic names and represent /o/. Whether -a is the auslaut for female names is questionable: φrima may have a suitable parallel in Venetic, but names in -a are regularly combined with the patronymic suffix -nu (vaþanu, velχanu, helanu, θianu, ketanu, laśanu), which would imply metronyms for men (if -nu is indeed masculine). No name attested on its own (without suffixation) ends in a consonant – while it cannot be excluded that such names do exist, but are not recognised as names by circular reasoning, the statistical preponderance of auslauting vowels suggests that where suffixed names have a consonant before the suffix, some sort of phonetic simplification (syncope/haplology) has taken place.

Even under the assumption that a considerable number of the names attested in the Raetic corpus are foreign, the lack of parallels with Etruscan in the sphere of onomastics is surprising.


Despite the fact that at least three of the find places of Raetic inscriptions were sanctuaries (Magrè, Montesei di Serso, Sanzeno), we do not know the names of the deities. Unlike, for example, in Venetic, where the name of the adressee / recipient of the votive gift is regularly mentioned in the inscriptions, the theonym is not part of the Raetic dedication formulae. Nevertheless, there are three candidates for theonyms, only one of them from sanctuary context:

  • tianu(s): Of doubtful auslaut, the name appears four times in the Non valley; the interpretation as a theonym (in the benefactive genitive) is based mainly on NO-15. Possibly the deity of the Sanzeno sanctuary?
  • θiuθi: Attested only once, in a context which might not be sacral; the interpretation as a theonym is due to its appearing in the (benefactive) genitive and the lack of a patronym, which would be expected in the name of a secular recipient. Note that Rix reverses the functions of genetive and pertinentive in the inscription, and consequently assumes the sequence śniχe to represent the recipient/deity. Accordingly, he also interprets eθsu* (in the pertinentive) on the Paletta di Padova as a theonym; the respective genetive form here is (a)θaris.
  • kusenku: Suspected to be a theonym for exactly the same reasons as θiuθi (and (a)θari), but mind the dubious segmentation of the sequence.

The Celtic theonym Taranis might be attested in the problematic FI-1.


So far, only one place name can be identified in the Raetic inscriptions, owing to the combination with the noun sφura 'community'. The community Entu*, arguably the home town of the person who left the respective votive at the Serso sactuary, can of course not be identified.

Text formulae

The most frequent syntagma is a name in the pertinentive case with a -ku participle, translated as 'X-ed of/by XY'. It occurs on its own, completely and with a two-part name on WE-3, NO-3 and NO-15; the combination of pertinentive ending and -ku is also found on SZ-14, SZ-30, WE-4, BZ-4, HU-7, NO-17 and PA-1. The syntagma is known from Etruscan, for example ???, and served as the starting point for the systematic comparison of the two languages as cognates (see Rix 1998: ???).

Relationship to Etruscan and the Tyrsenian language family

So far as the limited documentation of Raetic allows conclusions, it appears to be close to the Etruscan of the oldest inscriptions.


De Simone 1968 Carlo de Simone, Die griechischen Entlehnungen im Etruskischen, Wiesbaden: 1968.