2These constraints allow physicists, for example, to read articles in physics journals much the same way that we read newspapers (Bazerman 1985). Physicists ‘window shop’ for potential problems to explore. They read selectively, suing the abstract, the headings, seeding the news as they jump back and forth or skip sections depending on their interests or as one section raises questions about earlier ones (Bazerman 1895: 11). They are able to do this because the constraints imposed by the rewriting and editing process mentioned above ensure the certain information will be found in certain parts of the passage.

3In analyzing how scientists succeed in reading in the manner described by Bazerman, Atkinson (1991) described the conventions as guideposts through a text. Macro-rhetorical-spatial conventions such as introductions, methods and results sections, titles, headlines, or chapter breaks divide the information for the reader and can be considered as the “design coherence or architectural structure of the text” (Atkinson 1991: 65). The reader realizes that the physical break and the new heading mean a different type of information will be introduced. The architecture of the text works in a similar manner to various rooms in a house each of which have different purposes.

4Within each section on the other hand, the purpose is to integrate material rather than separate it like the beams of a house. Sections may be organized by ‘top-level organization principles’ such as the problem solution structure or comparative and contrast. Within each top-level organization principle the author uses phrasal-clausal material to define the structure in the manner or bricks and mortar.

5This paper will focus on the research article, which is one of the genres that students will face most frequently in their studies and which professionals must master to become members of the discourse community. We will focus on one of the macro-structural-spacial sections —the introduction— not only because this has been the section of the research article which has received the most attention in the literature but because an understanding of the working of the introduction can set the scene for the reader and allow the writer to communicate more effectively. If the reader understands how the introduction works he or she will access the appropriate schema and look for relevant information to interpret the text.

6We will assume that one of the best ways of demonstrating understanding of a text is an ability to form a working summary. Therefore if we can find a theory which provides a useful summary, we may have a sharper awareness of how information in a text is packaged and how to help our students identify that information. We will begin with Swales’ analytical approach described in Genre Analysis (1990), which provides a schema for introductions. We will then see how a summary derived from this approach corresponds to one produced by Rhetorical Structure Theory (1988), which focuses on the relations between propositions. Finally we look at a theory which produces summaries based on collocation introduced in Hoey’s (1992) bonding theory. The fact that the three theories converge suggests how different levels of text structure contribute to signalling meaning.

Swales’ Introduction Analysis

7Swales has identified three moves and steps for each move within the introduction of the scientific research article (see Table 1).

Table 1. Swales’ moves and steps

Move 1 Establishing a territory Step 1 Claiming centrality and/or Step 2 Making topic generalizations and/or Step 3 Reviewing items of previous research	Declining rhetorical effort
Move 2 Establishing a niche Step 1A Counter-claiming or Step 1b Indicating a gap or Step 1C Question-raising or Step 1D Continuing a tradition	Weakening knowledge claims
Move 3 Occupying the niche Step 1A Outlining purposes or Step 1B Announcing present research Step 2 Announcing principal findings Step 3 Indicating RA structure	Increasing explicitness

Swales 1990: 141

8The introduction becomes more explicit as it progresses. In move 1 the author establishes territory, defining the subject to be dealt with. One can boldly assert the centrality of the subject at hand or merely review previous research and by implication delimit the topic. However, if much research has already been conducted on a particular topic, one has to justify the present study. If much research has already been devoted to one topic, how can one justify yet another paper? One establishes a niche by showing the limitations of previous studies. One thus creates research space by showing how the present study will fill a gap in the previous knowledge. There is a scale of strength of claim. One can make a very strong claim by contradicting previous research or by indicating that past research has not covered a particular issue. One could make a much weaker claim by merely continuing a tradition. In the final move the writer describes how he or she will occupy the niche by outlining the purposes of the paper or indicating the paper’s structure. If all steps are included, there is a strong tendency for them to be in the order described above.

9Although Swales’ analysis was based on scientific introductions, one can see that it works well for many humanities introductions as well. Using Swales’ approach we will analyze an introduction from an article from the Journal of Second Language Writing (see table 2 and appendix 1).

Table 2

Move 1 Establishing territory Step 2 Generalization(s)	1. Many ESL learners come to the learning task already literate in their first language.
Step 3 Review of previous research	3. Much has been written about the difference between the forms and the functions of first and second language literacy.
Step 1 Generalization(s)	4. These studies focus on products, on what learners already know about writing/reading that may affect their second language acquisition efforts.
Move 2 Establishing a niche Step 1B Indicating a Gap	5. However, little attention has been paid to the process of literacy acquisition, to what second language learners have learned about LEARNING to read and write in their L1 that may affect how they approach literacy acquisition in ESL writing classrooms.
Move 3 Occupying the Niche Step 1B Announcing present research	6. In this paper I will examine L1 writing development in two contexts: Japanese and Chinese primary and secondary educational settings.
Step 3 Indicating RA Structure	8. Using this perspective, I will examine three aspects of literacy acquisition that affect the ways in which Japanese and Chinese students learn to read and write:

11Since these moves and steps tend to reoccur in the hundreds of articles Swales has analyzed, writers, editors and ultimately readers must be using them as ’signposts’ throughout the text. One can form a workable summary of introductions by including the moves and steps mentioned above. Thus sentences 1, 3, 4, 5, 6 and 8 form a coherent summary. If we add a 4th move justification, then sentences 9 and 12 would be included. As we will see in the next section, the summary corresponds to summaries created by different approaches.

Rhetorical Structure Theory

12In Rhetorical Structure Theory, Mann and Thompson focus on the relations between clauses. Using our building metaphor above, these relations could be considered the beams of the text. According to Mann and Thompson, as much meaning is contained in the relationships between clauses as in the clauses themselves. If one recognizes the basic relations between clauses, one can construct many inferences. For example, if one recognizes that there is a problem/solution relationship between the sentences below, one can construct many inferences. Each inference depends on the inference above it:

I’m hungry. Let’s go to the Fuji Gardens.
1. Our going to Fuji Gardens is a (partial) solution to my problem of being hungry.
2. I cannot eat here.
3. I want to be somewhere where I can eat.
4. I want to go somewhere where I can eat.
5. The Fuji Gardens is an eating establishment.
6. The Fuji Gardens is open.
7. The Fuji Gardens serves Japanese food.
8. I like Japanese food.
9. I do not mind eating raw fish.
10. I am partial to Japanese beer.
11. I can use chopsticks. (Mann Thompson 1986: 68)

13One can infer that the speaker cannot eat here (2) because, otherwise, there would be no motivation for going to Fuji Gardens as a solution to the problem of being hungry (1). If one cannot eat here (3), then one would logically want to go where one could eat (4). If one wants to eat, then the Fuji Gardens must be an eating establishment (5). However, there would be no logical reason to suggest going there if it was not open (6).

14Mann and Thompson list 22 basic relations upon which inferences such as those in 1–11 above can be made (see table 3).

Table 3

Subject Matter	Presentational
Elaboration Circumstance Solutionhood Background (increases ability) Volitional Cause Volitional Result Non-Volit.Cause Non-Volit.Result Purpose Condition Otherwise Interpretation Evaluation Restatement Summary Sequence Contrast	Motivation (increases desire) Antithesis (increases positive regard) Antithesis (increases positive regard) Enablement (increases ability) Evidence (increases belief) Justify (increases acceptance) Concession (increases positive regard)

Mann Thompson 1988: 250

15It is not only the relationships between the adjacent sentences which help create the meaning of the text but also that between larger blocks. These blocks may in turn have relations with sentences higher in the structure. Let us examine the introduction to highlight the relations between clauses:

16The overall relationship according to this analysis is of problem/solution. Clauses 1-5 indicate a gap in previous research that the present paper will try to fill. Clauses 1-4 depend on clause 5, which is the key sentence in the problem portion of the text (nucleus according to Mann and Thompson). One could re-phrase the material as a concession: Although much research has been conducted, little attention has been paid... Clauses 1-2a provide background to clause 3, which introduces relevant research, while clause 4 evaluates the research. Clauses 2 and 2a give additional information about the students introduced in clause 1. Clause 2 provides the circumstance within which one can interpret clause 2a. As can be seen, there is a hierarchy of relationships, which contribute to the meaning of the text as a whole.

17One can produce a workable summary by including only those clauses which have several satellites. If one includes only clauses which have two or more dependent attachments one has the following 1, 3, 5, 6, 8, 9, 11a, 12. Notice that up to sentence 9, one has almost the identical summary as in Swales’ analysis. One could object however, that since we are familiar with Swales approach, we will read the introduction in the same manner and thus form an identical summary. In order to provide validation, we must find a theory with a more objective approach to forming summaries.

Bonding Theory

18Bonding theory focuses on semantic relations to understand coherence. In our architecture metaphor, bonding theory looks at the bricks of our structure to see what the surface, i.e. the words, tells us about the relations of coherence in the text. Links are created between sentences if they share synonyms, or have a synonyms/antonyms relationship, or have the same basic word form (write - writer), etc. Sentences with three or more links are bonded. Hoey has found that bonded sentences are coherent even if separated by hundreds of pages. In other words, a reader will usually be able to construct a coherent text out of any two sentences from the same or even different texts if they have at least three links. The first step is to list the links between sentences (see table 5).

19There are four links between sentence 1 and sentence 2 (learning, language, first - which is a synonym in this passage with native and literature - see appendix 2). One can count the number of links and put it in a chart form (see chart 1).

Chart 1

	1
2	4	2
3	5	3	3
4	3	4	4	4
5	4	5	7	8	5
6	2	3	3	3	3	6
7	1	1	2	3	4	3	7
8	2	2	3	5	6	7	6	8
9	1	0	1	0	1	5	1	5	9
10	0	0	1	0	0	0	0	0	0	10
11	1	1	1	2	2	3	0	5	5	0	11
12	2	3	4	3	4	6	2	5	3	0	2	12

20Following sentence 1 down, one can see that there are four links between 1 and 2 as mentioned earlier, five links between sentence 1 and 3 and so on.

21Hoey compares sentences to oeuvres. A work that is very important will be quoted by other authors just as a sentence that is important in the text will have links to many other sentences. One can create a workable summary by choosing an arbitrary number of links of three and over and combining those sentences. If one looks at links of five or more for example, one can focus on areas of heavy collocation and thus by implication great coherence. The sentences which have five or more links to the most sentences are 8 and 5. Sentence eight has five or more links with 5, 6, 7, 9, 11, 12. Sentence five has five or more links with 2, 3, 4, and 8. By combining these linked sentences, one has a summary very similar to that produced by Swan’s and Mann and Thompson’s approaches:

Table 6

Conclusion

22As can be seen from the analyses above, different approaches arrive at similar results. Swales’ theory corresponds closest to traditional schema theory in which one searches for the prototypal script to interpret a text. Understanding occurs when the reader identifies a real-world frame in which to interpret the text. If students can be taught to identify the gap for example, they will be well on their way to Mann and Thompson’s theory focuses on the relations between propositions. The hierarchy of relations contributes as much to the meaning of the text as the words themselves. Finally, Hoey’s bonding theory shows us how the individual building blocks contribute to the organization of the text. Perhaps the reader subconsciously takes all of these relationships into account when comprehending a text. If we as reading teachers can become more aware of the intertwined relationships that create textuality, we may be better able to locate sources of reader incomprehension and identify appropriate remedies.