Literature review

Models of Reading
Many researchers have tried to explain the reading process and have arrived at various reading models. Researchers who have reviewed the processes involved in reading distinguished two kinds of processing, bottom-up and top-down processes. The bottom-up model emphasizes focusing exclusively on what is in the text itself, especially on the letters, words and sentences in the text (Aebersold & Field, 1997; Carrell, 1984). It is also called the text-based or data-driven reading model. Supporters of this approach focus on how readers extract information from the printed page (Samuels & Kamil, 1998). In the top-down model, on the other hand, the processing of a text begins in the mind of the reader (Bernhardt, 1998). Readers make predictions about what they will encounter by using their background knowledge, their experiences and their knowledge of how language works (Carrell & Eisterhold, 1983; Grabe & Stoller, 2002). In these two models, the term ‘top’ refers to higher order mental concepts such as the prior knowledge of the reader whereas the term ‘bottom’ refers to the physical text (Urquhart & Weir, 1998).
Theories that stress bottom-up processing indicate that language consists of sounds and letters, and decoding begins with the smallest units, letters, and works up to words, phrases and sentences. Proponents of this model indicate that written texts are hierarchically organized so the readers need to first identify letters, then words, and then proceed to sentence, paragraph and text level to construct meaning (Aebersold & Field, 1997; Anderson, 1999). Therefore, this model focuses on helping students decode the smaller units that make up a text.
As the traditional view changed over time, researchers started to consider reading as an active rather than a passive process. Thus, the bottom-up model was criticized for underestimating the contribution of the reader (Eskey, 1998). The importance of active readers and the use of background knowledge began to have an impact on theories of the reading process. These concepts did not play an important role in the bottom-up reading theory, in which the reader mainly needed to use textual clues to comprehend the text (Eskey & Grabe, 1998). Contrary to the bottom-up model, the top-down model involves knowledge that the reader brings to the text which enables the reader to actively participate during the reading process, making and testing hypotheses about the text (Carrell, 1998b; Goodman, 1998; Urquhart & Weir, 1998).
Goodman’s “psycholinguistic model of reading” is also considered as a topdown model. This model views readers as active participants who make predictions and verify them by processing the printed information (Goodman, 1998; Samuels & Kamil, 1998). However, the top-down model does not work well to describe what less proficient and developing readers do, and it seems to describe what skillful and fluent readers, for whom decoding has become automatic, do (Eskey, 1998).
As the importance of both the text and the reader was realized, the interactive model, which combines the prior knowledge and textual information, emerged (Eskey & Grabe, 1998; Urquhart & Weir, 1998). This model stresses both what is on the written page and what the reader brings to it (Aebersold & Field, 1997; Urquhart & Weir, 1998). The bottom-up and top-down processes work together in order to facilitate comprehension. Studies have shown that effective reading requires both background information and linguistic knowledge functioning together (Bimmel & Schooten, 2004; Grabe, 1998). If there is a problem with either one of them, the other compensates. For instance, when the linguistic ability of the reader is poor, top-down processing is likely to be used, or if the reader does not have the necessary background knowledge to interpret the new text, he allows the meaning come from the text itself. In the interactive model, the bottom-up process, which emphasizes textual decoding, and the top-down process, which emphasizes reader interpretation and prior knowledge, function simultaneously to help readers perceive meaning from a text (Aebersold & Field, 1997).