Ordering of language learning material and applying this claim provide the presupposition that it is possible to find the cognitive and linguistic demands of tasks so they could be arranged in a way that they eventually support and promote EFL learning process and help the development of their inter-language. The present study explored the effect of task complexity on the written performance of upper intermediate Iranian EFL learners. Task complexity operationalized in terms of the resource–directing variable of reasoning demands (causal and intentional) based on Robinson’s (2005) Triadic Componential Framework. To this end, a total of 55 Iranian EFL learners from Salis institute in Tabriz, Iran participated in this quasi-experimental study. To make sure of their appropriate level, the researcher distributed homogeneity test: Oxford placement Test (OPT). Finally, 40 participants of 16-24 years of age were selected as the sample. All of the subjects were the members of identical cultural societies. They were also from a bilingual community (Turkish as their mother tongue and Farsi as their second language). Participants did writing tasks, simple and complex narration, intentional and causal, and simple and complex narration tasks. In simple narration prompts, learners were given 10 pictures and were required to just narrate them but in reasoning demand prompts, six middle pictures were scrambled and learners were asked to put them in the correct order and narrate them: to differentiate reasoning demands the intentional and causal, a set of pictures from the adult Wechsler Intelligence Scale was used. The syntactic complexity, accuracy, and lexical variation of written performance were surveyed. The analysis of data by Paired samples t-test and MANOVA revealed that task complexity had a differential effect on lexical variety, syntactic complexity and accuracy of upper-intermediate EFL learners. Complex intentional reasoning of task complexity did not affect learners’
