attended and the students’ scores obtained on tests were the only representation of comprehension. Problems of reading comprehension were viewed as being essentially decoding problems, deriving meaning from print. This approach, however, could tell little to the researcher about how reading is successfully accomplished. Therefore, having felt the need for an insight into the process of reading, researchers shifted their focus of attention.
In the light of new orientation, the idea that reading proceeds word-by-word is rejected. In the early seventies, Goodman’s psycholinguistic model of reading (later named the top-down or concept-driven model) began to have an impact on views of second language reading. In this model the reader is active; makes predictions, processes information, and reconstruct a message encoded by a writer. As Clark and Silberstein (1987) denote: “Reading is an active process. The reader forms a preliminary expectation about the material, and then selects the fewest, most productive cues necessary to confirm or reject the expectation. This is a sampling process in which the reader takes advantage of his knowledge of vocabulary, syntax, discourse, and the real world.”
The top-down processing perspective into reading comprehension had a profound impact on the field, to an extent that it was viewed as a substitute for the bottom-up perspective, rather than its complement. Alderson and Urquhart (1984) contrast the two approaches as follows:
“A product view relates only to what the reader has got out of the text while process view investigates how the reader may arrive at a particular interpretation (p.150).” Smith (1973) emphasizes two important psycholinguistics contributions against word-by-word processing; first, it has been established that there is a severe limit to the amount of information that we are able to receive, process, and remember. Therefore, the reader does not use all of the information in the printed page, but he must select the most productive language cues in determining the message of the writer. As a result, the reading should necessarily be a rapid process which does not precede word-by-word.
Second, Smith holds, research has shown that actually reading is incidentally visual. More information is contributed by the reader than by the print on the page. In fact such a wordby-word reading has been recognized as a defective strategy used by native readers. Varzegar (1978), for instance, talks about the destroying effect of word-by-word reading on fluent reading. He regards reading as a multi-faceted process which entails different activities to be performed; a complex process of identification of letters, recognition of words, visual identification, and thereby sampling of ideas, predicting of forthcoming content and decoding the intended meaning of passage . Goodman (1971) considers reading as a psycholinguistic guessing game in which the reader constructs, as best he can, a message which has been encoded by a writer as a graphic display. The fact that reading is a selective process has been emphasized by Goodman (1973): “The reader does not use all the information available to him. Reading is a process in which reader picks and choose from the available information only enough to select and predict a language structure which is decodable.”
In this approach, reading is a process in which readers sample the text, make hypotheses about what is coming next, sample the text again in order to test their hypotheses, confirm or reject them, make new hypotheses, and so forth. The theory claims that this process of reading is a universal process, and posits the reading universal hypothesis which says that this model has been built through the study of English reading, but it must be applicable to reading in all languages and all orthographies (Goodman, 1975).
Smith (1971) describes reading as “the reduction of uncertainty”. Wallace (1992) elucidates that “as we progress through a text, our choices of what to select are constrained, often heavily, both by features within the text and those external to it”. He describes reading as a unitary aptitude, which is consistent with Smith’s and Goodman’s approach. They hypothesize that it is not possible to identify specific skills which can be built up in any hierarchical way. They believe that effective readers are characterized by ability and willingness to reflect on what they are reading. This unitary view of reading process, as Wallace (1992) contends, has led researchers to talk of reading strategies rather than distinctive skills. He states that “effective readers draw selectively on a range of strategies which are determined by reader purpose, text-type, and context”.
Finally Wallace (1992) proposes an interactive model of reading comprehension as follows: “Texts do not contain meaning; rather they have potential for meaning. This potential is realized only in interaction between text and reader. That is, meaning is created in the course of reading as the reader draws both on existing linguistic and semantic knowledge and input provided by the printed or written text. ”
However, as schema theory research has attempted to make clear, efficient and effective reading in L1 and L2 requires both top-down and bottom-up strategies operating interactively; that is interactive model (Rumelhart, 1977). Both top-down and bottom-up processes, functioning interactively, are necessary to an adequate understanding of second language reading and reading comprehension (Carrell, 1988).
As a matter of fact, the approaches that teachers employ to teach reading and students employ to read the reading materials depend on their functional definition of the reading, learning and language (Chastain, 1988). This means that there might be as many approaches as there are teachers, students, objectives, setting, and so on. But to save space, the major ones will be discussed briefly:
2.2.1) The Top down (Concept-Driven) Approach
The “top down” approach emphasizes readers bringing meaning to text based on their experiential background and interpreting text based on their prior knowledge (whole language).The word “top” refers to higher order mental concepts such as the knowledge and expectations of the reader, and “bottom” refers to the physical text on the page.
The top-down model of reading focuses on what the readers bring to the process (Goodman, 1967; Smith, 1971). The readers sample the text for information and contrast it with their world knowledge, helping to make sense of what is written. The focus here is on the readers as they interact with the text. This model starts with the hypotheses and predictions then attempts to verify them by working down to the printed stimuli.
Carrell and Eisterhold (1987) consider top-down processing as conceptually driven because:
“Top-down processing occurs as the system makes general predictions based on higher level, general schemata, and then searches the input for information to fit into these partially satisfied, higher order schemata. Top-down processing is therefore called conceptually driven (cited in Long & Richards, 1987, p. 221).”This view of reading was called the psycholinguistic guessing game (Goodman, 1967). According to Goodman (1967), readers employ 5 processes in reading:
Impact of Goodman’s model:
This model which has recently been characterized as a concept-driven, top-down pattern had the greatest impact on conceptions about native and second language reading instruction: it made the reader an active participant in the reading process ; from earlier views of SL reading as a passive linguistic decoding process to more contemporary views of SL reading as an active predictive process.
According to Stanovich (1980), this model has some disadvantages:
1. For many texts, the reader has little knowledge of the topic and cannot generate predictions.
2. Even i
f a skilled reader can generate predictions, this would take much longer than it would to recognize the words.
According to Eskey (1988), a top-down model of reading is essentially a model of the fluent reader and does not account for all the needs of students who are acquiring reading skills.
2.2.2) The Bottom up (Serial) Approach (Text-based)
The “bottom up” approach stipulates that the meaning of any text must be “decoded” by the reader and that students are “reading” when they can “sound out” words on a page. It emphasizes the ability to de-code or put into sound what is seen in a text. It ignores helping emerging readers to recognize what they, as readers, bring to the information on the page.
This model starts with the printed stimuli and works its way up to the higher level stages. The sequence of processing proceeds from the incoming data to higher level encodings.
Carrell and Eisterhold (1987) explain why bottom-up process is called data-driven: “Bottom-up process is evoked by the incoming data ; the features of the data enter the system through the best fitting, bottom-up schemata. Schemata are hierarchically organized from most general at the top to the most specific at the bottom. As these bottom-up schemata coverage into higher level, more general schemata, these too become activated. Bottom-up processing is, therefore, called data-driven. ” According to Stanovich (1980), this model has some problems:
1. This model has a tendency to depict the information flow in a series of discrete stages, with each stage transforming the input and then passing the recorded information on to the next higher stage.
2. An important shortcoming of this model is the fact that it is difficult to account for sentence-context effects and the role of prior knowledge of text topic as facilitating variables in word recognition and comprehension (because of lack of feedback).
According to Eskey (1988), the decoding model is inadequate because it underestimates the contribution of the reader who makes predictions and processes information. It fails to recognize that students utilize their expectations about the text, based on their knowledge of language and how it works (p. 3).
2.2.3) The Interactive Approach
For those reading theorists who recognized the importance of both the text and the reader in the reading process, an amalgamation of the two emerged the interactive approach. Reading here is the process of combining textual information with the information the reader brings to a text.
The interactive model (Rumelhart 1977; Stanovich 1980) stresses both what is on the written page and what a reader brings to it using both top-down and bottom-up skills. It considers reading as the interaction between reader and text. The overreliance on either mode of processing to the neglect of the other mode has been found to cause reading difficulties for SL learners (Carrell 1988, p. 239) The interactive models of reading assume that skills at all levels are interactively available to process and interpret the text (Grabe 1988). In this model, good readers are good decoders and also, good interpreters of text, their decoding skills becoming more automatic but no less important as their reading skill develops (Eskey 1988).
According to Rumelhart’s interactive model:
1. Linear models which pass information only in one direction and which do not permit the information contained in a higher stage to influence the processing of a lower stage contain a serious deficiency. Hence the need for an interactive model which permits the information contained in a higher stage of processing to influence the analysis that occurs at a lower stage.
2. When an error in word recognition is made, the word substitution will maintain the same part of speech as the word for which it was substituted, which will make it difficult for the reader to understand (Orthographic knowledge)