Overuse of Discourse Markers in Turkish English as a Foreign Language (EFL) Learners’ Writings: The Case of ‘I Think’ and ‘in My Opinion’ Ihsan Unaldi University of Gaziantep, Faculty of Education, Gaziantep, 27310, Turkey E-mail: ihsanunaldi@gmail.com KEYWORDS Learner Corpus. Discourse. Overuse. Language ABSTRACT This study tries to dwell on overuse of two discourse markers I think and in my opinion in Turkish EFL learners’ written productions. The data was collected from 161 Turkish EFL learners and a corpus of 58.046 words was compiled and it was compared with a native English corpus of 54.285. The focus of the comparison was the frequency of the phrases I think and in my opinion. A raw frequency calculation of the phrases revealed that the Turkish EFL learners actually used a considerable amount of them when compared with the written productions of native speakers of English. In the inferential analysis process, the variances of these phrases in the two corpora were calculated and since the calculations yielded statistically significant differences, a non-parametric test, Mann Whitney U-test, was employed. The results validated the obvious difference of the phrases in terms of frequency, which means that there is a plethora of the phrases of I think and in my opinion in Turkish EFL learners’ written productions. Address for correspondence: Dr. Ihsan Unaldi University of Gaziantep, Faculty of Education (27310), Gaziantep, Turkey Telephone: +90 (342) 360 12 00 / 3795 E-mail: ihsanunaldi@gmail.com INTRODUCTION Written or spoken productions of learners of English as a foreign language (EFL) have been analyzed from different aspects with different concerns. Throughout years, data gleaned from EFL learners from many different L1 backgrounds have been subject to numerous quantitative analyses. These analyses focused on issues such as perspectives on grammar (Biber and Reppen 1998; Meunier 2002), error analysis (Dag- neaux et al. 1998; Flowerdew 1998; Granger 1999; Flowerdew 2000; Abe and Tono 2005), chunks and phraseology (Granger 1998b; De Cock 2000), pragmatic developments (Flowerdew 1998; Belz and Vyatkina 2005; Callies 2009;) discourse (Aarts and Granger 1998; Mulak 2000; Aijmer 2001; Pulcini and Furiassi 2004; Gilquin 2008) and even on very specific concerns like punc- tuation marks (Celik and Elkatmis 2013). When the registers are taken into account, it is not surprising that written collections of L2 produc- tions outnumber those of spoken ones (O’Keeffe et al. 2007). Common Features of EFL Written Productions In a well-known meta-analysis, Silva (1993) compared L1 essays with L2 written productions collected from EFL learners coming from differ- ent language backgrounds such as Arabic, Chi- nese, Japanese and Spanish. L2 writing appeared to be distinct from and less effective than L1 writing. Moreover, L2 writings appeared to have certain organization issues. Similarly, Hinkel (2001) compared essays of native speakers of English with writings collected from speakers of Chinese, Japanese, Korean, Vietnamese, and In- donesian. Frequency rates of overt exemplifica- tion markers in the texts such as (as) an exam- ple, for example, for instance, in (my/our/his/ her/their) example, like, mainly, namely, such as ..., that is (to say) were calculated and ana- lyzed via non-parametric statistical techniques. The results showed that the non-native group employed far more example markers (conjunc- tions), first person pronouns, and past tense verbs in their academic texts. Again, in another study Hinkel (2002) analyzed 68 lexical, syntac- tic and rhetorical features of L2 text. The related corpus included texts written by advanced learn- ers of English from six different L1 backgrounds: Arabic, Chinese, Indonesian, Japanese, Korean and Vietnamese. The results of the study indi- cate that L2 writers have a severely limited lexi- cal and syntactic repertoire. This led the learn- ers to produce simplistic texts which are rooted © Kamla-Raj 2013 Anthropologist, 16(3): 575-584 (2013)