2.3.1 Bottom-up processing
Bottom-up processing is activated by the new incoming data. The features of the data pass into the system through the best fitting, bottom-level schemata. Schemata are hierarchically formed, from the most specific at the bottom to the most general at the top. It acknowledges that listening is a process of decoding the sounds, from the smallest meaningful units (phonemes) to complete texts. Thus, phonemic units are decoded and connected together to construct words, words are connected together to construct phrases, phrases are connected together to construct utterances, and utterances are connected together to construct complete, meaningful text. That is to say, meaning is arrived at as the last step in the process. A chain of incoming sounds trigger schemata hierarchically organized in a listener‘s mind— the phonological knowledge, the morphological knowledge, lexical and syntactical knowledge (syntactical knowledge aids to analyze the sentence structure). Thus, the listener makes use of ―his knowledge of words, syntax, and grammar to work on form‖ in the bottom-up processing (Rubin, 1994, p. 210). This process is closely associated with the listener‘s linguistic knowledge. However, bottom-up processing has its weak points. Understanding a text is an interactive process between the listener‘s previous knowledge and the text. Efficient comprehension that associates the textual material with listener‘s brain doesn‘t only depend on one‘s linguistic knowledge.