Significant considerations in esl/efl literacy
Theories of reading and their implications to the
teaching of reading in esl/efl classes & the place of
schemata theory on foreign language reading
by Hasan Bilokcuoglu
1.2.1. Bottom-up models: the traditional view
Bottom-up models argue that readers construct texts from the smallest to the largest units, like letters, words, phrases, clauses, sentences, texts as well as grammatical knowledge. In other words, bottom-up process refers to deriving the meaning of the text based on the incoming language data, from sounds, words, grammatical relationships, to meaning. Firstly, the reader is expected to become familiar with the words. Secondly, s/he is expected to get the meaning intended by the writer by combining the words that the reader recognized earlier.
Readers use top-down process when prior knowledge is used to understand the meaning of a text. Content words and contextual clues are used to form hypotheses for comprehension. On the other hand, readers as well use bottom-up process when they use linguistic knowledge to b understand the meaning of a text, which occurs in this way: meaning from lower level such as sounds, words, grammatical relationships to lexical meanings are built to reach at the final comprehension. Reading comprehension can be considered as an interactive process which requires both top-down and bottom-up processing, that is to say that readers use both prior knowledge and linguistic knowledge in understanding. For example, if someone asks us, 'have you got a watch?' and if we only employ the bottom-up process, and try to decode each individual word, then we may face a misunderstanding because of the missing top-down process. In the given example, the speaker asks the other person to tell him/her the time in fact.
According to Grabe and Stoller (2002), there is a mechanical pattern that the reader goes through by creating a piece-by-piece mental translation of the information in the text in which the interaction between the text and the reader contains little or no inference from the reader's own background knowledge. Anderson (1999) stresses that the reader is supposed to recognize letters at first, then recognize the words, and in the end get the meaning intended by the writer by combining the words the reader recognized earlier in this piece-by-piece mental translation process. Shortly, the bottom-up process of reading can be as a serial model where the reader begins with the printed word, recognizes graphics stimuli, decodes them to sounds, recognizes words, and decodes meaning (Paran, 1996; Alderson, 2000).
Barnett (1988:11) distinguishes the bottom-up process from the top-down process as following:
* Letters and characters
* The phonological component (letter and sound correspondences)
* Individual words
* The lexicon or vocabulary (words carrying meaning)
* Semantics (meaningful groups of words)
* Syntax or grammar (words functioning in relation to each other)
* Structure of sentences, paragraphs and the whole text.
Which elements of a written discourse are processed first determines the difference between bottom-up and top-down process. Jeremy Harmer (2001) describes this as: ''In metaphorical terms, this can be linked to the difference between looking down on something from above –getting an overview– and, on the contrary, being in the middle of something and understanding where we are by concentrating on all the individual features. It is the difference between looking at a forest, or studying individual trees in it.''(Harmer, 2001:201)
The reader reaches at understanding, in bottom-up reading, first, by putting different text elements together; that is, letters and words until a chunk of data is created (decoding stage). Just after the decoding stage, s/he goes through a stage of comprehension which transfers previous articulatory and phonological data into meaning. Bottom-up model of reading is characterized by a linear process of joining lower-level text items and gradually adding them together until a meaningful chunk of data is created (Samuels and Kamil in Barnett, 1988). A typical weak reader, when the texts are beyond her/his comprehension skills, operates the bottom-up decoding process as s/he is unable to draw meaning immediately. What some theorists believe is that bottom-up processing makes emphasis on how readers extract information from the printed page and letters and words are dealt with in a relatively complete and systematic fashion (Gough, 1972).
1.2.2. Top-down models: the cognitive view
In top-down processing, learners use their prior knowledge to make predictions about the text. Teachers, mostly, assume that when students hear (or read) every sound, word or sentence, they make out the general meaning of the passage. Nevertheless, in practice, they often take on a top-down approach to guess the probable theme and from there, they move on the bottom-up approach to check their understanding.
This process is called "conceptually driven" as it occurs when the system makes general predictions based on higher level, general schemata, and then looks for the input for the information to fit into these partially satisfied, higher order schemata. Considering the information mentioned above, it is possible to say that it is easier for students to comprehend the passage if students have background knowledge about it. It is believed that when students are familiar with the topic, then they are able to process it much better. Otherwise, if students have difficulties with it, then it is the teacher's job to help them with stimulating their schemata by some pre-reading activities. A top-down reading model can be described as a complete contradiction of a bottom-up model. Goodman (1970) calls the model "a psycholinguistic guessing game", which is quite exact that the reader does not immediately concentrate on the elements of text, but makes predictions about the meaning based on his/her knowledge of syntax and semantics. Goodman differentiates between four stages of top-down processing as seen following:
• predicting about grammar and meaning
• sampling the text to check whether the predictions are correct
Grabe and Stoller (2002) point out that reading is mainly directed by the readers' expectations and goals, that is why top-down models identify the reader as somebody who has expectations about the text information and grasps enough information from the text to confirm or reject these expectations.
Eskey (1988), however, emphasises the limitations of the top down model, pointing out that this model is dependent on the prediction of meaning (by using context clues) and combining contextual clues with background knowledge, so this model is only useful with skillful and fluent readers and does not work well with less proficient readers.
In short, readers principally need to train their higher order reading skills, i.e. improving predicting or guessing abilities, guessing meanings of words (by referring to the text), guessing topics of texts (by referring titles), skimming, scanning, critical reading, etc.
page 3 of 8
the article index