Recent study has proposed that there is a central dichotomy in the modes of text incrementaion in written discourse (cf. Winter, Hoey, Coulthard, Sinclair, Tadros, amongst others). This core distinction is best described by Sinclair (1994) when he distinguishes between two types of in-discourse sentences, here referred to as the Encapsulating and the Prospected sentences. The encapsulating sentence (1b/2b), as Sinclair puts it, retains the ideational of the preceding text, so that at the time of its appearance, it is in itself the whole text. That is, each incrementation by means of the encapsulating sentence must evoke the preceding text as its 'encapsulation' (bracketed component in 2b) The prospected sentence, on the other hand, realizes a prediction made by its preceding sentence (3b).
1a. I was on sentry duty.
1b. I saw the enemy approaching
2a. I was on sentry duty.
2b. (While I was on sentry duty) I saw the enemy.
3a. I have only one reason for doing this.
3b. I need the money.
In this paper, I will suggest that the modes of text incrementation effected by means of these sentences are not just different, but in fact diametrically opposed to each other. This operational contrast in the modes of their implementation means that by learning to effect incrementation by the earlier acquired prospected sentence, the language learner has also, unwittingly, put into place the wherewithal to prevent implementation of incrementation by the encapsulating sentence. Hence, the learning challenge posed by incrementation by the encapsulating sentence.
In the first section of the paper, I will demonstrate how the ability to effect evocation of the preceding text as encapsulation is a defining constraint on incrementation by the encapsulating sentence, and represent the textuality that results from the recursive use of such incrementations in text construction.
In the second section of the paper, I will show that the evocation-of-the-preceding-text constraint directly impacts two key requisite competencies in text incrementation, namely, 1) effective fragmentation of the ideational whole, and 2) effective textualization of each resulting information unit. Message presentation calls for effective fragmentation of the nonlinear ideational whole into a linear succession of constituent units. In incrementation by the encapsulating sentence, proper fragmentation of the ideational whole must be into units that can effect evocation of the preceding text as encapsulation. Likewise proper textualization of these individual units is also constrained to effect evocation of the preceding text.
In the last section of the paper, I will examine representative samples of ESL writing, from a total collection of 35 essays written by high-intermediate learners whose L1 is Cantonese, to suggest that they have problems with effective fragmentation and effective textualization in building text by means of incrementation by the encapsulating sentence. I will argue that this is because their fragmentation and textualization strategies are equipped to facilitate the earlier acquired incrementation by the prospected sentence, which pointedly disallows evocation of the preceding text.
Newcastle, UK, 2004.