Bağlama, yapısal bağlantı olmaksızın tümce ötesinde mantıksal ilişkiler kurmayı ve sözcüksel bağlaşımsa göndergesel özdeşlikten bağımsız olarak sözcük kümeleri içindeki anlambilimsel ilişkileri çözmeyi gerektirdiğinden çocuk okuyucuya en çok güçlük sunan bağlaşıklık araçlarıdır. Bu çalışmada, 11-14 yaş grubuna yönelik İngilizce ve Türkçe anadili ders kitaplarından seçilen türdeş (öyküleyici, biyografik, bilgilendirici) okuma metinleri söylemsel-anlambilimsel çözümlemeye tabi tutulmuş; bağlayıcı ve sözcüksel birimler işlevsel dilbilgisine göre sınıflandırılıp sıklık-yüzde hesapları yapılmıştır. Sonuçta, türden bağımsız olarak, Türkçe okuma metinlerinde (1) daha sık bağlamaya başvurulduğu, (2) bağlama kullanımının ders kitabının düzeyine göre gelişim göstermediği, (3) bağlamanın ekleyici ve zamansal türlerinin baskın olduğu ve neden, sonuç, karşıtlık, koşul ilişkileriyle anlamı genişletmek yerine yeniden ifade etme veya açıklık getirmenin yeğlendiği belirlenmiş; İngilizce okuma metinlerinde ise (4) bağlamanın düzeyle birlikte arttığı, (5) bağlamanın ekleyici, çeliştirici, nedensel veya zamansal türleri arasında ayrım gözetilmediği, ama (6) bağlayıcı ilişkilerin yapısal kaynaklar yardımıyla bağımlı tümcecikler içinde gerçekleştirildiği ve böylece (7) çocuk okuyucuya daha çok dilsel ipucu sağlanarak metni çözmenin kolaylaştırıldığı anlaşılmıştır. Ayrıca metin büyüklüğü değişse bile Türkçede sözcüksel yoğunluğun hep daha yüksek olduğu, buna karşılık İngilizcedeki sözcüksel birimlerin ana diziler oluşturarak konusal odağı sağlamakta Türkçeden geri kalmadığı ve ders kitabının düzeyi yükseldikçe sözcüksel yoğunlukların arttığı gözlemlenmiştir. Sözcüksel bağlaşım sağlamada, Türkçe ve İngilizce metinler parça/bütün türündeki ilişkilerin kullanımı yönünden benzeşirken, İngilizce metinlerin beklentisel ilişkiler ve anlamca yakın adların eş dizimliliğinden zengin olduğu ortaya çıkmıştır

With the purposes of comparing the use of conjunction and lexical cohesion and evaluating the textual quality in first language coursebooks, texts of the same type, chosen from English and Turkish series for 11-14 year-olds, were subjected to discourse-semantic analysis. The categorization of the conjunctive and lexical items according to functional grammar revealed that Turkish texts made more use of conjunction, especially of the additive and temporal types, and were more inclined to restate or clarify, whereas English texts preferred to realize conjunctive relations with structural resources of dependent clauses. Also, despite the variation in the text size, English texts managed to maintain topical focus by entering into major lexical strings, and both corpora were similar in terms of metonymy, while English texts were richer in expectancy relations and collocations of meaningfully-related nouns. Introduction Several studies of learner products in the native language supported that cohesion, coherence and the overall quality of the text are in interaction, and when poor writers were found to use fewer devices of cohesion, the teaching of their use by presenting conjunctives in reading texts became the purpose (Connor, 1999; Lenk, 1998; Lunsford, 1990). As a result, while first language learners trace the cohesive elements in the reading texts, they learn as readers how to solve text difficulties, and also as writers how to produce meaningful and connected sentences (Gamble & Yates, 2002). It is the coursebook that forms the primary resource for presenting the functioning of cohesive devices most comprehensively and intensively. The cohesive features of the reading texts in the coursebook has a great influence especially on the child reader’s comprehension, and the perceptive capacity for cohesive elements in the text gradually improves by making a big leap between the ages of 8-13. (Chapman, 1982; Gamble & Yates, 2002). For this reason, this study aimed to reveal the patterns of use for the cohesive devices in the reading texts chosen for the coursebook, the child reader’s immediate source of reference. Homogeneous texts, extracted from first language English and Turkish coursebooks, were subjected to systemic-functional analysis in line with Halliday and Hasan’s (1985) model of cohesion, and the cohesive items were compared in the bilingual corpora of pedagogical texts. While coherence is simply defined as the unity and totality of meaning in the text, the textuality of a group of sentences can only be claimed, if it is compatible with its environment; that is, it has texture (Halliday & Hasan, 1985; Stoddard, 1991). Eggins, (2000), on the other hand, argued that coherence is concerned with the contextual properties of a text, whereas cohesion is related to its internal properties, and if there is a text as part of discourse, this text must be in harmony with both the İngilizce ve Türkçe Anadili Ders Kitaplarındaki Türdeş Okuma Metinlerinin… 863 Turkish Studies International Periodical For the Languages, Literature and History of Turkish or Turkic Volume 10/7 Spring 2015 situational context and with itself; in other words, it must be cohesive (Halliday & Hasan, 1985). According to functional grammar, there are five cohesive devices that compose a text by interpreting the preceding and following elements in a semantic relationship of dependency above the sentence level: reference, substitution, ellipsis, conjunction, and lexical cohesion. Method The corpus of the study was comprised of the reading texts, derived from the local, MoNE-approved coursebook, “İlköğretim Türkçe 6-7-8”, (İÖT) and the famous series of an international publisher, “Cambridge Checkpoint English 7-8-9” (CCE) (Cox, 2012, 2013, 2014; Deniz, 2013a, 2013b; Şahin, 2013). The qualitative data from the narrative, biographical, and informative texts was examined with the descriptive analysis method. While comparing the density of the cohesive devices, their proportion to the total number of sentences or words were take into consideration, as in Eggins’ (2000) example. After the texts were transferred to Word, word counts and numeration of sentences were completed, the conjunctive items were labelled with respect to their functional categories and the lexical strings were identified. Finally, the frequencies and percentages were calculated by Excel and tabulated. Findings In the current study, undertaken for explicating the patterns of conjunction and lexical cohesion in use, and evaluating the quality of similar reading texts from first language English and Turkish coursebooks on the basis of the operation of cohesion, it was found that regardless of the text type, (1) Turkish texts used conjunction more frequently, (2) conjunction use did not develop in line with the textbook grade, (3) the additive and temporal types of conjunction dominated and the sentences tended to restate or clarify instead of expanding the meaning by means of reason, result, contrast and condition relations, while in English texts (4) conjunction increased with the level, (5) no distinction was made in the use of the additive, adversative, causal and temporal types of conjunction, but (6) conjunctive relations were realized with structural resources, within dependent clauses, and thus, (7) textual interpretation was facilitated by providing the child reader with more linguistic cues. Also, even though the text size changed, the lexical density was always higher in Turkish texts, whereas English texts were equally efficient in maintaining topical focus by entering into major lexical strings, and the lexical density in both texts appeared to increase with the grade. As for the lexical cohesion, Turkish and English texts were similar in terms of metonymy, whereas English texts were richer in expectancy relations and collocations of meaningfullyrelated nouns. Discussion and Conclusion The systemic-functional analysis of cohesion in homogeneous texts from first language English and Turkish coursebooks provided the following results: 864 Meliha R. ŞİMŞEK Turkish Studies International Periodical For the Languages, Literature and History of Turkish or Turkic Volume 10/7 Spring 2015 1. Regardless of the generic difference among the narrative, biographical and informative texts, Turkish texts had a denser concentration of conjunction. 2. Semantic relationships were formed between the larger parts of discourse by moving beyond the sentence boundaries. 3. Statements had the tendency to make restatements or clarifications, for the propositions were predominantly modified in terms of time relationships rather than reason, result, contrast and condition. 4. No steady increase was observed in the use of conjunction, as the level of the coursebook changed. 5. Conversely, in English texts, conjunction increased with the grade of the textbook. 6. No distinction was made in the use of the additive, adversative, causative or temporal types of conjunction. 7. However, such structural resources as dependent clauses were preferred in the lexico-grammatical realization of conjunctive relations. 8. Hence, it was understood that the use of more linguistic materials as cues facilitated the child reader’s interpretation of the text. When the distributive properties of the lexical relations were contrasted in the same texts, the below results were obtained: 1. Despite the varying volume of the texts, the lexical density in Turkish was always higher for all the three, narrative, biographical, and informative types. 2. Nevertheless, the lexical items in English were as efficient at forming long strings as those in Turkish, and the topical focus was equally maintained in each case. 3. Moreover, the lexical density of both corpora rose, as the level of the coursebook increased. 4. When the head items of the three longest strings were compared, the words describing the protagonist, the setting or the major processes were foregrounded independent of text type. 5. In the instances of lexical cohesion through (i) metonymy and (ii) expectancy relationships, Turkish and English texts resembled in the former, whereas English texts were richer in the latter, for they embodied collocations composed of meaningfully-related nouns, rather than made out of grammatical co-occurrence of verbs and nouns. In the light of these findings, it is evident that using conjunctions besides connectors is important for building up interdependency relationships on the sentence level. Gamble and Yates (2002) also noted that the cohesive device of conjunction adds elegance to composition as well as multiple levels of meaning, but it is, at the same time, the one that the child reader suffers from the most developmentally due to comprehension difficulties, and that therefore requires the highest level of support. Similarly, Crossley and McNamara (2010) found that not all the cohesive devices were equally efficient at maintaining cohesion in text production, and only subordinate conjunctions indicated a positive İngilizce ve Türkçe Anadili Ders Kitaplarındaki Türdeş Okuma Metinlerinin… 865 Turkish Studies International Periodical For the Languages, Literature and History of Turkish or Turkic Volume 10/7 Spring 2015 relationship with cohesion by determining text quality as well. Another striking conclusion is that increased lexical cohesion eases text comprehension by providing more clues for grasping the topical content of the text and its instruction with all the subtypes deserves special attention in reading and writing classes. Witte and Faigley (1981), too, revealed that lexical cohesion was more frequently used in highlyscored texts, and identified it as a sign of the writers’ capacity for expanding and connecting their ideas. In the same way, Hoey (1991) discovered that lexical cohesion alone could account for at least 40% of all the cohesive devices and owed its dominance in the text to its ability to develop multiple relationships. In summary, it is essential that in order for the young, inexperienced, naive reader to cope with processing difficulties, pedagogical texts have greater amount of cohesion and are adequately equipped with explicit textual clues, that is, conjunctives, especially in the face of challenging readings (McNamara et al., 2011; O’Reilly & McNamara, 2007). Consequently, writers should revise their conjunctive choices for the reading texts in first language English and Turkish coursebooks, and systematically employ structural and non-structural linguistic devices, namely connectors and conjunctions, by taking the student’s level into consideration. Also, as in the case of the reading texts found in first language English coursebooks, more lexical items in first language Turkish coursebooks need to enter into lexical cohesion by representing relationships other than simple repetition, and therefore, provide students with more samples of concise expression in discourse.