Remerciez-le!

Remerciez @Admin pour avoir partagé cet document gratuitement, de la manière la plus simple, en partageant sur les réseaux sociaux.

SUMERO-INDO-EUROPEAN LANGUAGE CONTACTS Aleksi Sahala 2009, University of Helsin

SUMERO-INDO-EUROPEAN LANGUAGE CONTACTS Aleksi Sahala 2009, University of Helsinki (Corrections made in 2013) aleksi.sahala@helsinki.fi | http://www.ling.helsinki.fi/~asahala/ Originally written as a course essay in 2009. ABSTRACT: Albeit the genetic affinity of the Sumerian language is still lacking consensus, some vocabulary related to Sumerian may be found from various language families including Indo-European, Kartvelian, Semitic, Dravidian and Uralic. Where the Semitic contacts are well attested, contacts to other families have often regarded controversial. In this paper I will present and briefly review 30 words attested in the Sumerian and Indo-European languages which may share a common etymology, including some which have already been proposed by J. Pokorny and G. Whittaker. Of the presented lexical data, ~9 words can be tentatively considered as direct borrowings and 6 as a proto-historical adstrate. The rest can be regarded partly as wanderworts or perhaps even as relics of the debatable Nostratic macrofamily. However, in some cases the semantic and phonetic resemblance may be purely coincidental. Introduction Sumerian language was spoken in ancient Mesopotamia from the 4th millennium BC to the Old Babylonian period (1900 BC) during which the Sumerians gradually assimilated into Akkadian speaking Babylonians. By the end of the 17th century BC Sumerian was no longer spoken as a first language but it was still studied by Akkadian scholars as a classical language and its literary tradition continued for almost two millennia. The latest written memorials of the Sumerian language date back to the post- Seleucid era 1st century AD (Hayes 1997: 4). In the 1850s after the rediscovery and partial decipherment of the Sumerian language by Henry Rawlinson and Edward Hincks, arose an 1 intriguing question of its genetic affinity. Although the typological features of Sumerian reflect those found in many agglutinative languages of Eurasia1, finding any genealogical links to other languages proved to be an insuperable task. Regardless of numerous attempts to connect Sumerian with Caucasian, Semitic, Ural-Altaic2, (Elamo-)Dravidian, Basque and Indo-European languages, by the vast majority of scholars it is still regarded as a language isolate with no known relatives (Edzard 2003). It has also been suggested that the Sumerian language descended from a late Paleolithic creole (Høyrup 1992). However, no conclusive evidence, excluding some typological features cannot be found to support Høyrup's view. Even though Sumerian has no known genealogical relatives, some Sumerian vocabulary can be identified from various languages such as Akkadian and Assyrian3, and also from their modern relatives e.g. Arab. هيكل (haykal) 'temple' ← Akk. ekallum ← Sum. é.gal 'palace'; Hebr. עיר ('ir) 'town' ← Sum. iri 'city; town'. Where the language contacts with Semitic languages are well studied and practically undeniable, it becomes more complicated to find convincing evidence on Sumerian language contacts with families located outside Mesopotamia, such as Indo-European. The key problem is, that due to distribution of possible Sumerian loan words in IE languages the contacts must have taken place before the diverging of the Proto-Indo- European language (PIE), which according to the present knowledge4 took place before the Sumerian migration into Mesopotamia. Consequently, in order to explain the distribution one is tempted to assume that either (1) Sumerian or its earlier language stage was once spoken in the proximity of the PIE urheimat located in the Pontiac-Caspian Steppe, or (2) that the common vocabulary was not directly transmitted from Sumerian to PIE (or vice versa), but was borrowed through unknown prehistoric languages spoken between the PIE and Sumerian homelands (and perhaps partly 1 Features include a developed case system, lack of grammatical gender or articles, complex finite and non-finite verb conjugation, extensive use of compounding etc. See Edzard 2003: 1; Hayes 1997: 6–7 or Thomsen 1984: 48–51 for a synopsis of the general characteristics of the Sumerian language. 2 Bobula 1951; Zakar 1971; Gostony 1975; S. Parpola 2007 (work in process). 3 See Lieberman 1976. 4 See Kurgan hypothesis by Gimbutas 1956. 2 even originated from them). I would personally consider the latter a more credible option as we know next to nothing about the Sumerian homelands before their migration into the Southern Mesopotamia. Despite Kramer's (1963) Transcaucasian hypothesis, i.e. a Sumerian migration into Mesopotamia from the north, ultimately from the Caucasian or Transcaucasian region is acknowledged as the most plausible option (see Ziskind 1972), the actual hard evidence for it is extremely difficult to find. Kramer based his hypothesis mostly into Sumerian chronicles, cultural features and their expertise in metal working5. The hypothesis also loosely supported by later genetic studies on the Iraqi people, which point to their close relationship with Kurds, Caspian Iranians and ultimately the Svani Georgians of The South Caucasus (Cavalli-Sforza et al. 1994: p. 242), but as the genetic relationship between the modern Iraqi people and the Sumerians are uncertain, this cannot be taken as a hard evidence. Alas, also from the linguistic point of view evidence for "Caucasian" origin is practically nonexistent. Sumerian and Kartvelian certainly share some typological features including ergativity and heavy verbal prefixation, but yet both can be explained as a late development in Sumerian. Common vocabulary is minimal and consists only of few uncertain similar lexical items (see Klimov 19986), which despite of their phonological and semantic similarities are problematic as the Kartvelian cannot be reconstructed beyond the Georgian-Zan level (ca. 2600 BC). Where the urheimat problem makes it difficult to give any certain time or place for the possible loan words between Sumerian and PIE, the phonological inventories complicate this matter even further. The Proto- Indo-European sound system is completely based on reconstruction and thus reflects the "pronunciation" on a very abstract level. The situation is not much easier with the Sumerian, as the exact quality of its phonemic inventory is very uncertain. The uncertainty is a result of the 5 The two last mentioned can point Sumerian origins to any mountainous region, not necessarily to Caucasus. 6 Klimov considers the few similar lexical items between Kartvelian and Sumerian as "phonetic symbolism" and not real etymologically connected words. However, he has failed to notice few possible items, e.g. Sum. kur 'mountain; (foreign) land' ~ GZ. *gora 'mountain; hill'; Sum. -da '(comitative case marker), side, with, and' ~ GZ. *-da 'and'. 3 decipherment process based on the Akkadian phonetic values of the Sumerian cuneiform signs, whereupon misleadingly the phonemic inventories of these languages seem to be almost identical. By internal reconstruction it is possible to reveal some of these lost phonemes, but unfortunately it is often impossible to locate their distribution among the vocabulary on a larger scale. E.g. we know that the Sumerian language featured a phoneme /dr/7 of uncertain quality as well as probable earlier (Proto-Sumerian) labio-velar stop8 /gw/ and perhaps two liquids9 /l1/ and /l2/ but there are only a handful of words where they are known to possibly exist. Even more complicated is the case of Sumerian vowel inventory of which only four vowels <i e a u> are distinguished on graphemic level, but where each of these graphemic vowels hide a subset of vowel phonemes tentatively reconstructed10 as <i> = /i/; <e> = /e ε/; <a> = /a/ and <u> = /o ɔ u/ (Smith 2007). Nevertheless, in this paper I assume that the current reconstructions of the Sumerian and PIE sound systems are adequate enough to be used for lexical comparison. In general, Sumerian words are represented in their graphemic forms, but also more detailed phonemic reconstructions are presented in cases they are available. Vocabulary This section consists of 30 Sumerian and Proto-Indo-European words and roots, which could possibly share a common etymology. Arrows (→) refer to the direction of borrowing, tilde (~) stands for “possibly corresponds with”. In the case of borrowing, the arrows point rather to language families than individual languages. For example OCS osl; OE assa; → Finn aasi 'donkey' means that the word has been borrowed to Finnish from Indo-European languages, not from Old Church Slavonic or Old English. 7 Quality of this phoneme is uncertain. Jagersma (2011) suggests /tsh/, but also some kind of flap, tap or spirant have been suggested. 8 This would explain the correspondence <b> ~ <g> between Emesal and Emegir, as well as other dialectal variation. 9 Jagersma (2011) posits only one liquid in the Sumerian phonemic inventory. 10 This is a very controversial topic often ignored or discussed very briefly in Sumerian grammars. 4 (1) Sum. anše 'equid; donkey; ass' ~ Hitt. (ANŠE) /?/; HLuw. (ASINUS)-na 'donkey; mule'; Arm. ēs11 'donkey'; Lat. asinus; PCelt. *assin 'ass'; Lith. asilas; OCS osl; OE. assa; → Finn. aasi 'donkey' The word dates back to the domestication of the African wild ass in North-East Africa around 4500 – 4000 BC wherefrom it was introduced to Mesopotamia and Levant ca. 2800 – 2500 BC. By this time the Proro-Indo- European had already diverged and consequently the word was only borrowed into western IE languages through Anatolia. Unfortunately, despite Hittites and Luwians wrote 'donkey' using a Sumerian logogram ANŠE, the word is unattested syllabically and thus the actual pronunciation is unknown. Phonetically anše seems to be a plausible source of borrowing for western uploads/Geographie/ sumero-indo-european-language-contacts-by-a-sahala 1 .pdf