Abstract:
A topic-type description accounts for the nature of the information contained in a text, as well as the way in which the information is organised and realised linguistically (Johns and Davies, 1983). Using a topic-type perspective, two writing programmes were planned to draw attention to conventional categories of information in texts which describe physical structure, namely the parts of the structure, the location of the parts, and the function of the parts. In an experimental programme, the semantic features of physical structure texts were brought to the conscious attention of secondary school students for whom English is a second language. This was done through the use of a self-questioning scale (Hillocks, 1984) based on an ideal set of "information constituents". The self questions were applied not only to input which preceded writing, but also during the writing process itself. A comparison group used prototypical texts based on the physical structure topic-type, but students did not have features of the topic-type drawn to their conscious attention.
The results of a nine week instructional programme showed that both the experimental and the control groups made similar gains in text fluency as measured by the number of physical structure information constituents produced, and in text complexity as measured by the type of physical structure information included in the texts. Similar gains were also made in grammatical accuracy by both groups.
In a post-test, when all the support of input was removed (teacher input, reading cards, practice activities and writing scales), the experimental group wrote more conventional texts than the control group in terms of information contained in the text and in the way the information is organized.
The results suggested that reading input, in particular, provided both groups with opportunity for acquisition. However explicit learning of text structure, as experienced by the experimental group, was of benefit when input through reading was removed. It was concluded that provisions to acquire and to learn text structure knowledge appear to have positive effects on text production.
