نشریه پژوهش های زبانی
سال دهم شماره 2 (پاییز و زمستان 1398)

All languages in the world use different ways to create new words. One of these ways is Compounding. In many languages, Compounding (also called composition) is the most frequently used way of making new lexemes. Its defining property is that it consists of the combination of lexemes into larger words. Compounds are words that are composed of two (or more) bases, roots, or stems. In other word, compounding consists of the combination of two words, in which one word modifies the meaning of the other, the head. This means that compounds have a binary structure. The productivity of compounding in many languages is largely due to its semantic transparency and versatility. When a new compound is formed, we already know the meaning of its constituents, and the only task we face is to find out the semantic relation between the two parts. In Persian, compounding and derivation are the most frequently used way of making new words. Many Iranian linguists have written articles about compounding. Classification of compounds has been and still is a much-debatable issue in the linguistic literature. In this article, we will review other classifications and then introduce a new classification which is based on a morphosyntactic approach and focus on the structure of words. According to this new classification, there are two types of compounds: root compounds and synthetic compounds. Synthetic compounds have v element in their structure but root compounds don’t have this element. Presence of v element in the structure of the compounds creates a dual situation which logically doesn’t have a third option and this is not reputable. Indeed it cannot be violated.

This study is an investigation of multiple–râ constructions as an instance of so-called split DP in Persian in the Minimalist framework. In principle, DP splitting may give rise to another phenomenon, namely, syntactic doubling. This phenomenon, as its name suggests, involves the occurrence of more than one syntactic marker in a syntactic object. In Persian, this marker is the accusative Case marker –râ. To account for the multiple occurrence of –râ, we follow Kahnemuyipour's (2014) treatment of adjectival Ezafe construction as involving a head final NP. We then analyze split DPs in Persian as a byproduct of topicalization and focalization of inner elements of DPs. So given the Topic and Focus interpretation of syntactic objects in the constructions under investigation and their relative order, we posit TopP and FocP projections in a domain between head D and NumP.  It is worth noticing that in linguistics theorizing the Split DP construction corresponds to split CP as proposed by Rizii (1997(. With respect to accusative Case marker then we adopt Darzi’s (2006) treatment of –râ as the morphological realization of dependent Case in the domain of V + I in Persian. Under Baker and Vinokurova's (2010), Preminger's (2014) and Baker's (2015) proposals, the dependent Case marker appears on a DP which is c-commanded by another DP in a certain domain if the latter is not yet valued for Case. In this respect, finally, we employ the mechanism of subextraction from the main DP as a result of which the extracted element altogether carries other Case marker to a higher DP upon extraction. With regard to nominative Case, after adopting Levin and Preminger’s (2014) proposal, namely, unmarked Case, which in turn appears as a result of the interaction between I/T and being in the domain proposed by Bobaljik (2006,2008) and Darzi (2006), we maintain that subextraction is also accounted for constructions with multiple nominative Case.

Nowadays, errors made by language learners are not seen as signs of language incompetency, rather than are considered as measures for assessing second language , competency in learning a second language. In the present research, it is attempted to analyze the written errors of Kazakh- Persian students based on Corder’s model. The main question of the study is the determination of frequent errors in the writing of Kazakh students, and the hypothesis is that the grammatical errors are more than lexical errors in the written text of the Kazakh language students. The present research is descriptive-analytic and its theoretical framework is the proposed Corder model. The statistical population of this research is non-Persian speaking Kazakh who study Persian language as a second language in Persian language at the Ablai Khan University in Kazakhstan. The research tool is a collection of 24 written articles from Kazakh undergraduate students that relates to the final test of their writing course with the right to choose a topic from among the five suggested subjects. On average, each entry contains 150 words.  First, the language errors of Kazakh students were identified and extracted based on the proposed Corder model. Then, based on the type of errors, they were categorized and analyzed using statistical software. The frequency of errors were determined and finally analyzed. The research findings confirmed the hypothesis. The results of this research show that out of a total of 466 errors identified in the Kazakh written articles, the most errors are due to the non-observance of grammatical points with 231 cases and 47% of the total errors, and then the graphological – phonological errors with 182 cases and 41% of The sum of the errors is the worst mistake. Lexico-semantic errors with 43 occurrences and 10% of the total errors, in the next rank are the frequency of errors, and the lowest errors with 10 occurrences and 2% errors, are related to the cognitive- application errors.

The aim of this research is to study the effect of Persian language on Azerbaijani compound verb construction. Data were collected from Azerbaijani resources and the interactions of 30 speakers of this language in Ardabil province (Ardabil, Meshkin Shahr and their surrounding regions). Native Azerbaijani compound verbs plus those borrowed from Persian were extracted from the collected data and their frequency was determined. The research method is descriptive-analytic, because at first the data were classified in the framework of the verb accommodation strategies in the borrowed language and they were analyzed according to the common approaches in language contact literature. Results show that no Persian finite verb has entered Azerbaijani language. Azerbaijani language uses two strategies to accommodate Persian finite verbs. The light verb construction and indirect insertion are two important strategies that Azerbaijani speakers use to accommodate Persian verbs. The frequency of light verb construction (Persian non-verbal element + Azerbaijani light verb) in the speakers’ speech is very high in comparison to indirect insertion strategy (Persian noun or adjective + Azerbaijani verb making affixes) and Azerbaijani native compound verbs. In spite of some previous studies considering Azerbaijani compound verbs as patterns borrowed from Persian, this research finding shows that light verb construction is not a new pattern in Azerbaijani and this construction exists in the language itself. Azerbaijani speakers use the extending strategy to increase the frequency of this construction matching that of Persian. Borrowed Persian non-verbal element causes Azerbaijani speakers to choose among alternative verb making constructions (Persian non-verbal element + Azerbaijani light verb and Persian noun or adjective + Azerbaijani verb making affixes) the one corresponding to Persian construction. This change resulting in Azerbaijani progressive convergence to Persian is a natural consequence of language contact.

Affixoid is a frequent word attached to different bases, plays the role of derivational affixes and makes derivational words. These linguistic elements are morphologically independent but by the passage of time and under the influence of different factors, they attain different figurative meanings and are present in derivational words. In this research, we study 2 Persian prefixoids,"khosh-, bad-" which have opposite meanings and have ahigh frequency in Persian, in cognitive morphology framework and based on Hamawand (2011). It is corpus-based and the corpus used is Sokhan Dictionary. Firstly all words containing these elements are listed and the classified and analyzed. These elements have all three cognitive operations; categorization, configuration and conceptualization. Based on categorization operation, these elements besides their denotational meanings, have other meanings some of which are prototypical and some are peripheral. Prefixoid “khosh-“has 2 prototypical meanings, beauty and goodness, and 5 peripheral meanings, abundance, faithfulness, lack, skill, ease, and prefixoid “bad-“ has 1 prototypical meaning, ugliness, and 5 peripheral meanings, abundance, unfaithfulness, hardship, severity, positive attitude.   Based on configuration operation, different domains are studied. Each of these prefixoids has specific domains and separate facets. Prefixoid “khosh-“has 2 domains, quantity and manner, and “bad-“ has distinction domain. And finally based on conceptualization operation, we study words which have the same bases but synonymous prefixes are attached to them and different construals are attained such as “more than usual vs. excessive” by the use of prefixoids “khosh-“ and “por-“ attached to the base “Namak”, “quantity vs. quality” by the use of the previous mentioned prefixoids attached to the base “kar” and the last construal “humor vs. insult” which is made by the use of prefixoids “bi-“ and “khosh-“ by the base “Gheirat”. Based on the findings of this study, all linguistic elements have senses that represent speakers’ different mental experiences and prefixes, generally, and prefixoids, specifically, are not exceptional.

Numerous studies in different languages have shown that, in general, subject relative clauses are easier and faster to process than object relative clauses. However, there are discourse expectations that are specific to object relative clauses and influence their processing. According to Fox and Thompson (1990), object relative clauses have a strong association with a discourse grounding function, while subject relative clauses are associated with other functions, such as introducing new information about modified noun phrases. So, the difficulty in processing object relative clauses could be due to the unexpectedness of encountering new information, and more specifically, to the embedded noun phrase not referring back to the topic of conversation. On this basis, if subject relative clauses occur isolated from the discourse context and their embedded noun phrases are not mentioned previously, no difference will be found in their processing speed. By contrast, when the referent of the embedded noun phrase of object relative clauses is the ongoing topic in the preceding context, their processing speed probably would be faster. The main aim of this research is to investigate the comprehension speed and accuracy of full noun phrase subject and object relative clauses in the presence of a context (a sentence before subject and object relative structures). Researchers investigated the processing speed of different regions of subject and object relative clauses in two conditions, using self-paced reading experiment. In the first condition, the embedded noun phrase of the subject and object relative clauses was not mentioned in the preceding sentence (neutral context), and in the second one, the embedded noun phrase of the subject and object relative clauses was the topic of the preceding sentence (topical context). 40 mono lingual, Persian speaking normal adults participated in this research. The findings indicated that discourse factors would help to ease the processing of object relative clauses, and processing object relative clauses preceded by the topical context is faster than subject relative clauses in a similar context.

Metadiscourse as a key term in discourse analysis involves coherensive and interpersonal features which helps the connection between the text and the context with the aim of communicating with the audience, organizing the text and interpreting it by the audience. To achieve a coherent text in academic discourse, authors should use metadiscourse, its types including interactive and interactional metadiscourse and their strategies appropriately.  So, the nature and the distribution of metadiscourse is of significance regarding their related scientific fields. The present study seeks to explore “interactive metadiscourse” based on Hylandʼs model (2005) and corpus-based approach in the specified instance of Persian academic discourse which is referent researches. Interactive metadiscourse is used to organize the propositional informations of the text with five strategies including transitions, frame markers, endophoric markers, evidentials and code glosses. The aim of this study is determining and comparing the frequency of occurrences of different kinds of interactive metadiscourse markers in referent researches. Hence, we identified and extracted the interactive metadiscourse markers in 120 Persian referent researches of various scientific fields on humanities, basic science and engineering using Mahak Samim corpus and AntConc software program then we investigated them by manual method too and classified them in five groups of metadiscoursal strategies. Data analysis shows that there is a significant difference in the use of interactive metadiscourse markers in Persian referent researches of three scientific disciplines. This result indicates that although scientific articles belong to a unitary genre (academic discourse), they are also affected by the nature of the affiliated fields. Likewise, surveying interactive metadiscourse strategies shows that only “frame markers” are distributed equally among three scientific fields and the significant differences in the frequency of occurrences of transitions, endophoric markers, evidentials and code glosses among studied scientific fields are confirmed.

This article is an EEG/ERP study based on psycholinguistic models of reference resolution to study the discourse comprehension and coherent contexts through co-referencing between an anaphor and its antecedent in the same spoken context. How brain is able to establish reference between an anaphor and potential antecedents in discourse has become of great interest. Based on Informational Load Hypothesis by Almor (1999), an auditory task using participants with no neurological or medical disorder was designed to compare coreference resolution of a referential pronoun and its acceptable antecedent in two-sentence discourse context. Based on this theory, and considering the morpho-syntactic limitations of Persian language for processing the information comes from the pronominal anaphor, discourse function and processing cost of reference resolution analysis was examined by situations such as ambiguity in referring expressions. Research method in this study is based on psycholinguistic experimental model which indicates the sub-processes of brain functions responding to co-reference resolution. The model used in this research represents the balance between discourse function and processing costs of reference resolution between referential coherence and referential ambiguity conditions. The ERP components shown in this project indicate that the processing of anaphor resolution in different situations imposes different processing cost on working memory. The two components Nref and Left Anterior Negativity (LAN) in frontal lobe were elicited during referential ambiguity of the pronominal anaphor (third person) which has no gender feature. Nref is a sustained referential negativity which indicates the ambiguity of the referring anaphor. (LAN) in previous researches has shown the increasing activity of working memory. This result which is incompatible with previous studies, has indicated that lower salience of the antecedent can cause higher processing cost for the working memory in order to establish a cross-reference bridge between the new information and the already existing representations of antecedent.

Discourse analysis examines the impacts of human, gender and cultural interactions and social values on discourse, and studies the way these relationships are presented in the text. Hence, in this article, which deals with the phenomenon of bargaining in an Iranian culture, Kermanshah, the approach of critical discourse analysis is used to study the discourse of bargaining regarding the mind control process of the audience. According to Foucault, power has a non-predetermined structure, hence because of its elusive nature, participants in a discourse can play different roles. This research data is gathered utilizing a questionnaire (164) and recording some conversations (850 minutes), in three different social class areas in Kermanshah. The analysis of the data shows that discourse participants use different linguistic mechanisms to reach their goals in bargaining, such as topicalization, short single sentences, exclamatory, imperative and interrogative structures, besides prosodic prominence strategies such as changes in stress, intonation and juncture which can end in shifts in the focus of the information structure of the sentences. However, simple words with ideological connotative meanings have the highest frequency. Moreover, according to the findings, three highlighted behavorial strategies frequent in bargaining include the usage of religious beliefs, having ethnic tradition biases, and stating economic, social, and class problems. The results indicate that the various roles pertaining to the bargaining participants –buyers and sellers- put bargaining in the category of reverse discourse, as Foucault calls it. The seller, in the power hierarchy, is the superordinate through the ownership position, while the buyer is the subordinate. However, in mind controlling, the buyer’s attempt to control the seller secretly makes them the controller, the one who applies power in a bottom-up order. Given that it is impossible to control the seller's mind without persuading them or without changing their worldview, bargaining is categorized as a rhetorical task, and the buyer, in this discourse, is the rhetor.

This article investigates the verbal ellipsis (iḍmār al-fi’l) in Sībawayhi’s (c. 760–796) Kitāb. First, the terminology of this research domain has been explicated and it is stated that the term “ḥaḏf” (deletion) in the medieval Arabic (Sībawayhi era) is often dedicated to the phonological level of the language whereas the term of “iḍmār” (suppression) belongs to the syntactic (and pragmatic or even cognitive level). For this reason, the term “iḍmār al-fi’l” is used in this article to refer to the (verbal) ellipsis. Thereafter Sībawayhi’s data are introduced and a new classification of them has been proposed (based on the ellipsis ground). The presence of the speaker and the hearer in a common situational or cultural context and the presence of a grammatical reason can be accordingly enumerated as the grounds for ellipsis. Moreover, the data are investigated differently, i.e. in the framework of the Parallel Architecture of the Grammar (PA) which is a formal model to analyze the natural language. Although the data are historical, each modern language includes similar ones that can be compared with our data and therefore the proposal which is made here, can also be used to analyze them. For the analysis of these data which don’t have linguistic antecedents, it has been proposed to consider an intermediate cognitive tier which binds the elliptical utterance with its non-linguistic antecedent so that its minimal syntax will be licensed. This mechanism of licensing which is known as “indirect licensing” makes this binding possible. It looks similar to an interpretive process in medieval Arabic grammar which is called taqdīr (suppletive insertion). Our proposal is also compatible with a more general model, the structure of the mind, proposed by Jackendoff (1997, 2002, 2010) and can clarify some aspects of contextual ellipsis in the medieval Arabic which is imparted by the term “ḥāl” (discourse equivalence).