Semantic Network in the Lari Language

This descriptive-analytical research examines sense relations between the lexemes of the Lari language, the continuation of the Middle Persian and one of the endangered Iranian languages spoken in Lar, Fars province, to determine its semantic network, for the first time, in the framework of lexical semantics. To do this, 2500 words (the research data), marked by “L” (Lary) by Eghtedari (1992), were extracted from among 5000 original words he had gathered and classified into different semantic fields. Then, the sense relations existing between them were identified, a plan was presented for each of the relations in the semantic network, and the number of words participating in each relation and their percentages were determined. The results revealed that different sense relations, including hyponymy, synonymy, homonymy, homograhy, polysemy, opposition and semantic contrast, meronomy, member-collection relation and portion-mass relation, were present in Lari semantic network, while homophony was absent. This indicates Lari’s less ambiguity compared to the Standard Persian. Among these relations, hyponymy (%15.28) and semantic contrast (%0.8) had the highest and the lowest frequencies, respectively. The results also indicate that Lari semantic network is idiosyncratic and its biggest part has a hierarchical order while it is heterogeneous in the rest parts.
 
Extended Abstract
1.Introduction
The Lari language, the continuation of the Middle Persian, is one of the southwest Iranian languages spoken in Larestan, located in the southeast of Fars province, Iran. It has more than 20 different dialects, such as Gerashi, Evazi, Bastaki, and Bikhei. In various studies on this endangered language, it has been mentioned under different titles such as "Larestani language", "Lari accent/dialect", "Larestani dialect" and "Lari language". In the present study, a variety of the language that is spoken in Lar city, the center of Larestan, is investigated which is sometimes called “Standard Lari Language”. Although the language has been studied significantly from different synchronic and diachronic views, there is no systematic, comprehensive and coherent semantic study about it, and, as far as the authors know, there is no research considering its semantic network. Therefore, the goal of this descriptive-analytical study is to identify the sense relations between the lexemes of the language and to determine a part of the semantic network (semantic memory) in the mental lexicon of Lari speakers in the framework of lexical semantics.
2.Theoretical framework
Lexical semantics is a part of the language theory in which the semantic structure of words is studied by applying two different methods. One of them, used in this research, is based on the existing sense relations between words wherein their meanings are related to each other through a network that clearly shows their connections. A semantic network is a knowledge base that represents the semantic relationships between concepts in a network and includes sense relations -- the relationships between words that give them meaning - such as hyponymy, synonymy, polysemy, homonymy, etc., and any other semantic relations between them.
3.Methodology
The method of this study was descriptive-analytical and its tool was a checklist. Cluster sampling method was applied to gather the research data; i.e. from among 5,000 original words presented in Ancient Larestan and Larestani Culture that made the research population, 2,500 words which were marked by “L” (Lari) were extracted and divided into twelve categories according to their  subjects (semantic fields). These categories included "social relations", "objects", "body parts", "verbs", "places", "animals", "food", "family relations", "nature", "agriculture", "occupations" and "types of adjectives and adverbs". Pronouns were also classified in a separate category. Then, the number and the percentage of words in each category were calculated compared to the total extracted words.
The research sample consisted of those words that had sense relations with each other, and these relations included hyponymy, synonymy, polysemy, homonymy, semantic opposition, semantic contrast, meronymy, member-collection, and portion-mass. Finally, while availing from descriptive statistics, the number and the percentage of these words were determined.
4.Results and discussions
Among the 2500 words, the highest number (522 words; 20.88%) was in "social relations" category, while the lowest number (9 words; 0.36%) belonged to “occupations”. The results of studying the existing sense relations revealed that 382 words (15.28%) entered hyponymy, 91 words (3.64%) entered synonymy, 50 words (2%) were in polysemy, 154 words (6.16%) were in homonymy, 30 words (1.2%) were homograph, 33 words (1.32%) entered semantic opposition, 20 words (0.8%) participated in semantic contrast, 103 words (4.12%) entered meronymy, 80 words (3.2%) were in member-collection and 28 words (1.12%) were in portion-mass relations, while no evidence of homophony was found. Moreover, it was found that hyponymy had the highest (15.28%) and semantic contrast had the lowest (0.8%) frequencies, respectively, among Lari sense relations.
5.Conclusion and suggestions
Thematic classification of the Lari words indicated that the largest number of them belonged to "social relationships" category while the lowest one belonged to the "occupations" one; and therefore, it can be claimed that Lari speakers pay specific attention to the relationships between people. In addition, the lower number of words for occupations can be due to the originality and antiquity of Lari language.
The results of the study reveal that the most part of the semantic network in Lari language has a hierarchical order, and in some parts, it is heterogeneous: the sense relations of hyponymy, meronymy and member-collection (22.6% of all relations) have a hierarchical order and the rest of them (16.24%) are mostly heterogeneous. In addition, while the results of the present study, regarding hyponymy, are in line with those of Nasiripour and Almasi, which dealt with the sense relations in Laki Noorabadi dialect and a part of the Gilaki dialect of Lahijan city, Iran, it is quite different from Mohammadi Ivatlu’s results in Azerbaijani Turkish. Since this language belongs to the Ural-Altaic language family, polysemy has the highest frequency (18.44%). Furthermore, the Lari language and Laki Noorabadi dialect both lack homophony and this indicates less ambiguity in them compared with the standard Persian language and some other Iranian languages and dialects such as Gilaki.
The highest frequency for hyponymy in the Lari language and Laki Noorabadi and the above-mentoned Gilaki dialects with different frequencies reveals that hyponymy differs from one language to the other and depends on what each language selects from the world. Lari, Laki Noorabadi, and Gilaki are all sub-branches of Iranian languages, while Azerbaijani Turkish belongs to another language family, and this may be the reason why polysemy has the highest frequency among its words. The difference in the semantic networks of the dialects in question and lari language can be due to the difference between the root language of Lari and these dialects. In addition, the current research and its comparison with other studies, according to the differences found, shows that the semantic network of Lari language is different from the semantic networks in the above-mentioned studies, and at the same time, unique, due to the existence of semantic relations in it, and the frequency of occurrences of each relation.
Finally, it is suggested that: 1) the semantic network of different dialects of Lari language be studied before it becomes completely extinct, and 2) analogous to FarsNet (Farsi Lexical Network), LarNet ( Lari Lexical Network) for the Lari language be prepared, so that this original and valuable language can be accurately documented.
Select Bibliography
Almasi, H.The Study of Lexical Relations of Laki Dialect in Noorabad. MA Thesis of General Linguistics, Faculty of Humanities. Tehran: Payame Noor University, 2010. [in Persian]
Eghtedari, A. Keshte-ye Xwish: Collection of 50 Article. Tehran: Toos Publications, 1978. [in Persian]
Dianat, L. Homonyms and homographs in Lari. The 8th International Conference on Language and Literature Studies of Nations, 2023; pp.1-8. Retrieved from https://icll.bcnf.ir/.  [in Persian]
Malchanova, A.K. Lari Dialects. M. Ehsani, (Trans.).  Nameh Farhangistan, 2001; 18: 183-7. [in Persian]
Miller, G. et al. Introduction to WordNet: An On-line Lexical Database. International Journal of Lexicography, 1991; 3(4): 235-244. (Revised: August 1993)
Mohammadi Ivatloo, E. An Investigation in Lexical Relations in Azerbaijani Turkish Language. MA thesis of General Linguistics. Tehran: Payame Noor University. 2009. [in Persian]
Nasiripour, B. Lexical relations in the Field of Aquatic and Birds in Gilaki Dialect of Lahijan. MA Thesis of General Linguistics. School of Literature and Humanities. Tehran: Payame Noor University. 2009. [in Persian]
Rovshan, B. Lexical Semantics: The Classification of Persian Verbs. Ph.D. Dissertation of General Linguistics. Faculty of Humanities. Tehran: Tehran University. 1998. [in Persian]
Salami, A. Ganjine-ye Gooyeshshenasy-ye Farsi (A Treasury of the Dialectology of Fars). Tehran: Academy of Persian Language and Literature. 2004. [in Persian]
Vossugi, M.B.  Padzooheshi dar Zabanshenasi-e Larestan (A Research in Larestan Linguistics). Tehran: Kalameh Publications. 1990. [in Persian]