---

The present research explores the “wæl”, “tɑ” and “ærɑ” prepositions in Gurani Kurdish within the framework of cognitive semantics. One of the recent models in semantic analysis of prepositions is Principled polysemy model (Tyler & Evans, 2001, 2003; Evans & Tyler, 2004a, 2004b; Evans, 2004, 2005, 2006) that suggests explicit and applicable criteria for determining the distinct senses as well as the primary sense of prepositions. We aim to analyze the semantics of three mentioned prepositions, to determine their distinct senses and primary sense, and to represent their semantic networks. The results suggest that the primary sense of “wæl” is “companionship” and its semantic network contains 9 distinct senses and 2 clusters of senses. The primary sense of “tɑ” is “ending” and its semantic network contains 4 distinct senses and 1 cluster of senses. The primary sense of “ærɑ” is “for and sake” and its semantic network contains 4 distinct senses and 1 cluster of senses.

---

The current paper focuses on polysemy of one highly flexible spatial term in Persian, namely "zir", from the Principled Polysemy Model perspective. The studied applications of "zir" are mostly taken from 'Farhang-e Sokhan-e Anvari' and the Principled Polysemy Model has served as the analytical tool. Besides discussing the semantic aspects of spatial "zir", the study aims to investigate the applicability and efficiency of Principled Polysemy model in semantic analysis this specific term. In fact, we seek to know what strong or weak points the principled polysemy framework shows regarding primary sense specification and distinct senses discrimination. After the model's application on uses of "zir", its primary spatial sense was specified and next, four distinct extended senses were determined and  all of them  constituted the semantic network of "zir". The results of its semantic analysis within principled polysemy framework confirmed this model's two points of advantage compared to some other leading cognitive investigations especially Lakoff (1987); one in determining the primary sense due to considering certain linguistic criteria in the procedure, and the other in restricting the number of distinct senses by adopting a moderate view towards polysemy. However, the semantic analysis of "zir" in the adopted framework faced some challenges too, among which two more significant issues included the psychological reality of distinct senses and their high level of context-dependency.
 

---

This article explores the polysemy of four negative non-verbal prefixes in Persian language (zedd ‎‎'against, opposite of', bi 'without', nā 'not' and qejr 'not, non-') based on Principled Polysemy ‎framework (Tyler and Evans 2001, 2003). First, the primary sense of each prefix is determined ‎and then it is explained how non-primary senses are derived from the primary one, hence ‎demonstrating the semantic network of each prefix as a radial category. In this research, using ‎AntConc software (Anthony, 2014), first all the occurrences of the four prefixes were extracted ‎from the Hamshahri Corpus Version 2 (AleAhmad, Amiri, Darrudi, Rahgozar & Oroumchian, ‎‎2009) and then in order to analyze research data, some of them were randomly selected. The ‎findings of the study indicate that only in three of the four prefixes under study, polysemy is ‎observed, and that the frequency of use, ease of derivation and predominance in the semantic ‎network are the best criteria for determining the primary sense. The conceptual phenomenon ‎involved in the polysemy of these prefixes is metonymy. Data analysis shows that metonymical ‎shift occurs at two levels: at the level of morpheme/prefix sense and at the level of word-‎formation, and that the former leads to more straightforward relations within the semantic ‎network. This is due to the fact that the latter requires a more complex line of imagination which ‎automatically translates into a corresponding complexity of relations in the semantic network and ‎significant reduction in the type frequency of non-primary senses as a whole.‎