- Introduction
- Ancient Writing Systems
Elymian is one of the non-Hellenic languages of ancient Sicily and it proves the existence of indigenous and alphabetized communities in the northwestern part of the island. The Elymian language belongs to the Indo-European family, as can be seen by the existence of an inflected system and of a -a singular nominative. Moreover, the linguists who studied the Elymian corpus found the greatest structural similarities of the Elymian either with the Anatolic group (see R. Ambrosini, U. Schmoll, V.I. Georgiev, who sees a relationship with the Hittite, see "The Elymian Language", in M. Mayrhofer et al. Hrsgg., "Antiquitates Indogermanicae. Studien zur Indogermanischen Altertumskunde und zur Sprach- und Kulturgeschichte der indogermanischen Völker". Festschrift für H. Güntert, Innsbruck 1974, pp. 409-411) or with the Italic group (as far as this group is concerned, the opinions are varying. According to M. Lejeune, Elymian would represent the southern offspring of a general Italic; according to M. Durante it would be an idiom very similar to Sicel; according to G. Alessio it would be very close to Latin). With reference to ancient historiographical memory, both these hypotheses can be linked to the two best-known traditions about the Elymians: Thucydides (Thuc. VI, 2), according to whom Egesto and Elym fled from Troy, and Hellanicus of Mytilene (fr. 53 M), who supports the Italic origin of the Elymian language.
The hypothesis of an affinity with the languages of the Anatolian group does not find any confirmation and Elymian is thought to be linguistically far closer to the Italic groups, even if unarguable evidence is still missing. From this perspective, for northwestern Sicily Elymian is depicted as an Italic language even though it is different from the Italic speech found in the east and in the documents coming from Montagna di Marzo or from Mendolito di Adrano.
The linguistic reconstruction of a fragmentary language (Restsprache) is challenging; interpreters have exegetical problems due to the overall exiguity of the sources and the recurrence of few textual structures. This is true of Elymian, to which few documents can be attributed with a high degree of certainty. In any case, the intense and ongoing debate (see IAS, pp. 123-168; Agostiniani 1988-1989, pp. 359-362; 1992, pp. 142-143; 2000) has managed to attribute a certain amount of phonological and morphosyntactic traits to the Elymian language through the analysis of the Elymian corpus, which contains coin legends and all the vase inscriptions which can be identified as non-Hellenic with a high degree of certainty.
Through the analysis of the Elymian alphabet, a vowel system made of five vowel phonemes has been recognised together with two diphthongs, i.e. [ai] and [ou], which are respectively noted as <αι> and <ου>. The diphthongs [oi], [ei] and [au] (probably represented as <αο>) cannot be isolated and identified for sure as Elymian (see Agostiniani 1992, p. 142).
Because of the existence of some reduplicated graphs such as <αα>, two different hypotheses have been formulated: that it is the realization of a long vowel phoneme or rather the /aa/ biphoneme representation of an original /aja/ with the drop of an intervocalic /j/ (see Durante 1980, p. 886). From this second hypothesis it could be derived that a palatal sound corresponding to /w/ seems to be lacking in the Elymian system. Moreover it is worth pointing out that this lack is not thought to be graphic, but that it would be the mirror of the Hellenic habit of representing both [i] and [j] through <ι> (about this theme see Agostiniani 1988-1989, pp. 360-361; 1992, p. 143).
As far as the consonants are concerned, the tendency is to rebuild a phonological system which presents the plosive voiceless series of /p/ /t/ /k/ and the voiced one /b/ /d/ /g/. This system involves also the nasal phonemes /m/ and /n/, the lateral one /l/, the resonant /r/ and the labiovelar approximant /w/ as well; among the fricative consonants in the Elymian corpus, /s/ and /h/ can be recognised. According to L. Agostiniani, there is presumably a labial consonant (/f/ o /β/ o /Φ/), which is represented through the sign ; this is different from the opinion of G. Alessio, R. Ambrosini, R. Arena, L. Dubois, M. Durante and M. Lejeune (see section i) Coin legends) who consider it a vowel.
As far as morphosyntactic traits are concerned, it is possible to identify Elymian as an inflecting language, which employs suffixes to express different grammatical functions cumulatively.
The suffix -(α)ζι- performs a derivational function and it has been found only in the coin legends expressed in -αζια and in -(α)ζιβ, where it creates adjectives from the toponyms of Segesta and Eryx, two of the most important centres of the Elymian ethnos. The suffix -(α)ζι- bears the same functional content of the suffixes -αιο- ed -ινο-, which form the ethnonyms inflected at the plural genitive EΓEΣTAIŌN ed EΡΥKINŌN (see section i) Coin legends).
It is possible to notice some inflectional morphemes as well. Some of these recur in formulaic structures and this is linguistic evidence of the cultural hegemony of the Greek world and of the existence of a strong koiné (Agostiniani 1992, pp. 363-364), which involved both the non-Hellenic and the Hellenic components in the same area. Among the inflectional morphemes the most frequent ones are: -α(α)ι, which can occur both alone (e.g. IAS no. 305: ḥαλενιαι; no. 278: ]ιλααι) or associated to εμι (e.g. IAS no. 319: αταιτυκαι εμι; no. 306: ]λενααι εμ[) in the most recent pottery inscriptions as well (such as ατιιαι εμι from Grotta Vanella); -β, which in the coin legends of Segesta and Eryx characterizes the -(α)ζι- derivational morpheme and which can occur isolated (σεγεσταζιβ; ιρυκαζιβ) or, but exclusively in Segesta, followed by εμι (σεγεσταζιβ εμι); -α, occurring as singular nominative ending both in the legend σεγεσταζια and in the pottery inscriptions (for the indications on the functional values see the sections i) Coin legends and ii. Vase inscriptions).
On the contrary, it is extremely complex to give the status of endings to the sequences -οι, -ει, and even -ι, while it seems more likely to consider -ρ a singular nominative ending linked to themes such as -αρ, -ερ, -ιρ.
Both in the inscriptions and in the coin legends εμι is unanimously regarded as an independent word. It can be said with the same degree of certainty that the non-Hellenic εμι performs the same textual function of the Greek εἰμί in the respective expression of possession. However, there is an ongoing debate among the linguists about the options of regarding εμι as a purely Elymian form, that is the indigenous outcome of the Indo-European *esmi (see, for example, M. Lejeune, "Notes de linguistique italique, XXV. Observations sur l'épigraphie élyme", in Revue des Etudes Latines, XLVII, 1969, p. 166, who highlights the difference with Oscan and Latin) or as a Greek element which entered the indigenous lexicon through the Hellenic formulaic scheme as a loanword. This thesis is supported by L. Agostiniani (IAS, pp. 138-139, 150-152; Agostiniani 1988-1989, p. 366; 1992, p. 145), but see also different interpretations, such as R. Arena (2003).
Notwithstanding the fragmentary, incomplete and restricted nature of the documents, there are far more words linked to personal onomastics than to the area of common nouns. This assessment has been definitely stated thanks to the comparison between several non-Hellenic and Grecian onomastic themes found in different areas of Sicily and ancient Italy. As far as Hellenic onomastics is concerned, this is the case of hερμων (inflected at the genitive case in IAS no. *371b) which clearly shows that the Segestan sanctuary was attended both by resident and non-resident Greeks. Also the themes TITELA- (IAS no. 322) and BOTULA- (IAS no. 289; 317) belong to the study of non-Hellenic onomastics (these themes can be compared to the recurring ones of Selinunte, i.e. TITEL- and BOTUL-).