Exploring the cohesive devices in written and spoken texts of “Let’s Learn Persian” based on Halliday and Hassan's (1976) model

Previous studies on cohesion within Halliday and Hassan's (1976) Functional Grammar framework have mostly explored the use of one or two cohesive devices in various educational texts. The limited scope of these studies has impeded the development of an applicable model for all cohesive devices. Through an in-depth contrastive analysis, this study explored the relationship between the number of cohesive devices in written and spoken texts of “Let’s Learn Persian” textbook (Zolfaghari et al., 2008) and the simplicity/complexity of those texts. Through systematic random sampling, in each of the three levels nine lessons were selected. Since each lesson contained a written and a spoken text, in the end there were 27 written and 27 spoken texts for analysis. The texts were coded and categorized. Qualitative and quantitative analyses showed that there were significant relationships between the frequency of four types of cohesive devices (lexical cohesion, reference, ellipsis, and conjunction) and the complexity of the written texts. However, in spoken texts, only in the case of lexical cohesion and conjunction, significant relationships were observed between the frequency of cohesive devices and the texts’ complexity level. Overall, in the analyzed texts, lexical cohesion was the most frequent cohesive device while ellipsis and substitution were the least frequent cohesive devices. In both written and spoken texts, with regard to the frequency of the devices, a nearly identical rank order was observed which was as follows: lexical cohesion, in-text reference, conjunction, ellipsis, and substitution. This pattern of occurrence could inform the development and evaluation of level specific texts.

Expanded Abstract
Research literature has so far focused on representing cohesive devices in various language instruction texts based on Halliday and Hassan's (1976) Systemic Functional Grammar. However, these studies restricted themselves to identifying only one or two cohesive devices in these texts. Hence, studies need to adopt a practical comprehensive framework which would cover all cohesive devices especially in Persian language instruction texts. Through an in-depth contrastive analysis, this study aimed to address the relationship between the frequency of cohesive devices in written and spoken texts of “Let’s learn Persian” textbook and the difficulty level of its texts. The textbook series authored by Zolfaghari, Ghaffari, and Mahmoodi Bakhtiari (2008) includes three main levels: elementary, intermediate, and advanced. They were selected since they had both written and spoken texts. Nine written and spoken texts in each level (in total 27 written and 27 spoken texts) were selected through systematic random sampling. Having categorized and codified the selected texts, researchers analyzed the data quantitatively and qualitatively.  The quantitative and qualitative analyses of the data indicated that both written and spoken texts of “Let’s Learn Persian” textbook had various cohesive devices. As higher level texts were presupposed as more difficult ones, frequency of lexical cohesion, reference, ellipsis and conjunction increased in the higher level written texts.  It can be argued that there existed a significant relationship between the frequency of the mentioned cohesive devices and difficulty level of written texts. There was no significant relationship between substitution frequency and difficulty level of the texts. The frequency analysis of cohesive devices in spoken texts of all levels showed that there existed significant relations between conjunction, lexical cohesion, and text difficulty. The overall data analysis indicated that patterns of use of cohesive devices were highly similar in spoken and written texts. Indeed, both text types followed similar patterns of cohesive devices application, namely lexical cohesion, reference, conjunction, ellipsis and substitution. The only difference was related to the place of reference in the identified pattern. This difference can be due to endophric and exophoric reference manifestation in both text types. Regarding spoken texts, it can be argued that as the difficulty level increased, the frequency of endophoric reference also increased. However, the exophoric reference has not been observed as frequently in spoken texts. This can be explained in terms of the textbook developers’ assumption that non-Persian language learners may not enjoy enough background knowledge about the texts. Furthermore, the identified patterns can be indicative of the fact that lexical cohesions are the first and most significant cohesive devices in spoken and written texts. Of all types of lexical cohesions, repetition can be regarded as an effective cohesive device for non-Persian language learners which could aid them to transfer meaning more effectively. It can influence the tone of speech when the speaker wants to put more emphasis on meaning. Substitution and ellipsis were the least frequently observed cohesive devices in written and spoken texts as well. This showed that the texts were developed based on circumlocution. Indeed, the textbook developers selected the written and spoken texts which were appropriate for each instructional level. The low frequency of substitution and ellipsis showed that the textbook developers presupposed that the non-Persian learners and native Persian speakers had little or no shared background knowledge about the text contents. The normal distribution of cohesive devices in written and spoken texts of “Let’s learn Persian” textbook showed that the textbook developers could apply a systematic cohesive devices pattern. In other words, this sequential pattern of cohesive devices can be taken into account by textbook developers in designing or evaluating standard TPSOL textbooks. It is suggested that further studies explore how cohesive devices are used in other textbooks developed for Persian language learners. Comparison of the findings of these studies could help us to identify a more comprehensive pattern of cohesive devices which could aid in the development of standard Persian language teaching textbooks.