Introduction

Academic texts are important because of their informative and referential functions (Montgomery, 2010; Olohan, 2020). Since English is today’s lingua franca, it is academic texts written in English that are usually being referred to (Pfau & Humblé, 2019). To enhance such knowledge dissemination,  academic texts are expected to be packaged in such a way that a reader can read them easily. In other words, academic text producers should consider both matter and manner of writing. In fact, experiences illustrate that an academic text can be read easier than others. Such experiences seem to be encountered by English native and non-native speakers alike. However, English non-native speakers, such as Indonesians who speak English as a foreign language (EFL), are naturally expected to experience more difficulty in reading English academic texts. In this study, the purpose is to attempt to relate textual properties to readability.

One of the textual properties responsible for readability is its thematic structures, a concept based on Systemic Functional Linguistics (SFL) (Halliday & Matthiessen, 2014; Webster, 2015). Thematic structure is made up of Theme, realised by the first element of the clause (point of departure), and Rheme, realised by other subsequent element(s) of the clause (temporary destination). The accumulative choices of Theme and Rheme  in a text contribute to its cohesion and readability respectfully. This suggests that a text with well-formed thematic structures contributes to its internal cohesion (Eggins, 2004) and naturalness or acceptability, thus enhancing readability (Butt et al., 2000; Baker, 2018; To, 2018). Considering its importance, scholars have studied thematic structures in various languages. The existing Theme and Rheme studies have examined parallel texts involving English and other languages (Kim & Matthiessen, 2015; Rahnemoon et al., 2017; Zulprianto, 2019) and monolingual texts in English (Dontcheva-Navratilova et al., 2020; Ghadessy, 1995; Hawes, 2015; Martin, 1995; Wei, 2016). While these scholars conclude the positive correlation between thematic structures and readability of a text, such a conclusion tends to be based on linguistic analysis per se.

In the Indonesian context, scholars have devoted studies on English thematic structures. These studies, however, are restricted to English texts written by students in the foreign language teaching settings (Gunawan & Aziza, 2017; Hendrawan et al., 2021; Yunita, 2018). Our study differs from these studies for two reasons. Firstly, the data were obtained from texts published and written by experts, not student writers learning the English language. Analyses of thematic structures of such texts have been ignored, particularly in the Indonesian context. This omission is likely because published academic texts tend to be considered well-written hence easy to read. Secondly, one of the purposes of the study was to apply a reader’s response to determine readability of the selected texts. Available studies on the topic merely rely on linguistic analyses. Surveying readers permits addressing the notion of readability in a more objective tone: not only from the (linguistic) thematic structures perspectives, but also from readers at a university.

University students in Indonesia are often required to read English academic texts for their courses. The advancement of information technology has made the procurement of such original English books easier. However, little is known about the readability degree of such English academic texts for Indonesian readers. This study intends to analyse the thematic structures of selected academic texts, seek out their readability by actual readers’ responses, and draw a possible link between readability and thematic structures. The study results are expected to inform stakeholders in English teaching and publishing industries, including teachers, lecturers, academics, writers, editors, and publishers regarding the readability of the English academic books they use or produce.

Here the thematic structures of fragments of three English academic books written by prominent academicians are examined: Language and Symbolic Power by Pierre Bourdieu (1991), Language and Power by Norman Fairclough (1989), and The Empire Writes Back: Theory and Practice in Post-Colonial Literatures by Ashcroft et al. (2002). The three books have enjoyed popularity, especially in the field of humanities around the world, including Indonesia. In fact, these three books are available in Indonesian translations. In this study, the purpose is to see the relationship between thematic properties and the degree of readability of the selected academic texts. Here readability is defined as the “ease of comprehension because of style of writing” (Harris & Hodges 1995, p. 203), while it is acknowledged that a writer’s choice of thematic structures is part of their own writing style.

Literature Review

Metafunctions in SFL

SFL was first developed by Michael Halliday. His seminal book, An Introduction to Functional Grammar (1985), is considered by many the main reference for SFL research. The book is so popular that it has been reprinted and revised since its first publication (Steiner, 2018).

What is ground-breaking about the SFL approach is that it sees language as functional. This paradigm is different from, for example, the generative theories pioneered by Noam Chomsky, which did not examine language in use or context but in mentalistic spheres (van Valin & laPolla, 1997). In SFL, text and context (of situation, i.e., field, tenor, mode, and culture) are inseparable and treated as having a realisational relationship.

A text does not occur in a vacuum. This means that if we know the context, we can deduce the kind of text in use or, conversely, if we know the text, we can predict the context in which the text is used (Eggins, 2004). The former shows a top-down analysis, while the latter involves a bottom-up analysis. The bottom-up approach is considered more practical since a text offers tangible linguistic data—such as words, phrases, clauses, and discourse patterns—that can be systematically analyzed. In contrast, context is often abstract and not as easily observable. SFL highlights the dynamic relationship between text and context, emphasizing their interdependence in meaning-making. In this, SFL necessitates linguistic evidence to substantiate any contextual interpretation (see for example Butt et al., 2000).

 In light of this, Eggins has noted that, with such an approach to language, SFL makes it possible to observe "realisational relationship extending all the way from the most abstract levels of context (ideology) to the very concrete words, structures, sounds and graphology of text" (p. 328). It is for this very reason that SFL has been applied as a working theory in Critical Discourse Analysis or CDA (O'Grady, 2019). In CDA, the SFL analysis takes a bottom-up approach in which the lexicogrammar of a text is analysed as the basis to describe, for instance, the ideology of the text. In the present study, a bottom-up approach was also used.

Another important differentiating characteristic of SFL is its theory, arguing that a clause simultaneously contains three kinds of meanings or metafunctions. The three meanings are ideational meaning (further divided into experiential and logical meaning), interpersonal meaning, and textual meaning (Halliday & Matthiessen, 2014). These three types of meanings occur simultaneously in a clause: while the ideational metafunction allows us to interpret a clause as representation of reality, the interpersonal metafunction views it as exchange between interlocutors and the textual metafunction considers it as message and these metafunctions can be examined through their transitivity and logico-semantics aspects, mood structures, and thematic structures, respectively.

This study focuses on textual metafunction, which Webster (2015) defines as "what makes a text into a coherent piece of language, as opposed to simply being an unorganized string of sentences" (p. 22). Therefore, textual metafunction concerns the manner rather than the matter of speaking/writing. In view of textual metafunction, the clause is seen as a message, suggesting that the clause is selectively constructed hence a meaningful choice. The textual meaning of a clause is analysed through its thematic structures, a more detailed explanation of which is given in the following section, but for more in-depth discussion (Butt et al., 2000; Eggins, 2004; Fontaine, 2013; Halliday & Matthiessen, 2014; Webster, 2015).

Thematic structures in SFL

As mentioned above, the textual meaning of a clause can be revealed by examining its thematic structure. This does not by no means infer that thematic structure analysis is the only way to uncover textual meaning: other approaches would include information structure and cohesion (Webster, 2015). Banks (2008), however, argues that thematic structure is the most important aspect of textual meaning and thus this study is limited to thematic analysis.

Thematic analysis is done by segmenting a clause construed as message into Theme and Rheme. Theme is defined as the point of departure of the message in the clause (Halliday & Matthiessen, 2014), and Rheme is the temporary destination of the message (Butt et al., 2000). By position, Theme is the first element in the clause that contains experiential meaning. Grammatically speaking, Theme can be realized by the subject, predicator, complement, or adjunct of the clause (Bloor & Bloor, 2004). Once the Theme has been identified, the rest of the clause serves as the Rheme.  This is shown in the example below:

The underlined constituent (Logical integration) is the Theme of the clause, while the rest (is the precondition of moral integration) is the Rheme (punctuation of the clause is ignored here). The Theme of the clause is logical integration, and the writer selected this as the point of departure of his message.

Theme is classified by (un)markedness and multiplicity. Given that Theme is the first element of the clause, its status, whether marked or unmarked, depends on the mood of the clause in question. In an interrogative English clause, the unmarked Theme is the WH-question or finite, while in an imperative clause, the Theme is the process/predicator in the clause (Eggins, 2004; Halliday & Matthiessen, 2014). However, the notion of marked or unmarked Themes is particularly worth examining in a declarative mood because a declarative clause allows reordering to some extent (Hasselgard, 2004).

In a declarative mood, an unmarked Theme is when the subject of the clause realizes the Theme. In English, such is expected and hence conveys the least meaningful choice. A marked Theme occurs when the first element (Theme) is realized by something other than the subject, namely, the predicator, complement, or adjunct. The constituent Logical integration from the example above is an unmarked Theme as it occupies the subject slot. An example of a marked Theme is given below:

In this example, In many post-colonial societies is the Theme. However, since it functions as the adjunct of the clause, the constituent is classified as a marked Theme. In English, adjuncts occur the most frequently as marked Themes compared to predicators or complements (Downing & Locke, 2006). This makes adjuncts less meaningful than predicators or complements as marked Themes. Adjuncts also seem to be thematized the most in Indonesian after subjects (Zulprianto, 2020). Despite so, other languages might not necessarily share the same textual characteristic.

In terms of multiplicity, Theme is divided into simple and multiple Theme. A simple Theme is solely realized by a topical Theme, while a multiple Theme is by a textual and/or interpersonal Theme(s), in addition to a topical Theme. The presence of a topical Theme is compulsory, while textual and/or interpersonal are optional.

A topical Theme is a constituent in the thematic element that carries a transitivity function. This can be realized by the subject, predicator, complement, or adjunct in the clause. In the two previous examples, Logical integration and In many post-colonial societies are both topical Themes. An interpersonal Theme occurs in the thematic element before the topical Theme. It instils an interpersonal, not experiential, meaning to the message. For instance, the word Undoubtedly in the clause Undoubtedly, logical integration is the precondition of moral integration is an interpersonal Theme. The inclusion of the interpersonal Theme results in subjectivity in the proposition in the clause. Another type of Theme is textual Themes. A textual Theme does cohesive work, and it does not carry experiential or interpersonal meaning (Eggins, 2004; Halliday & Matthiessen, 2014). For example, However in However, logical integration is the precondition of moral integration functions as the textual Theme of the clause: it connects opposing ideas between the current and the preceding clause. The types of Theme are shown in the figure below (all Themes are in italic and the Rhemes are written plainly:

Table 1: Types of Themes in the SFL framework

Hasselgard (2004) notes that when the three kinds of Theme co-occur, their order is textual^interpersonal^topical and they “do different jobs in shaping the thematic perspective of the clause” (p. 73). Analyses of these Themes are expected to shed light on the relationship between thematic structures and text readability. More specifically, this study attempts to answer these research questions: what are the thematic structures of the selected academic texts? Which text is more or less readable from the perspective of the sampled respondents? And what implications can be drawn between thematic structures and readability?    

Methodology

This study applied a mixed-method approach. The data were collected and analysed qualitatively and quantitatively in two ways through the selected texts and questionnaires.

Texts

Firstly, the texts are fragments purposively sampled from three academic books: Language and symbolic power by Pierre Bourdieu, translated from French and published in English in 1991 (Text 1), Language and power by Norman Fairclough, published in 1989/2013 (Text 2), and The empire writes back: theory and practice in post-colonial literatures by Bill Ashcroft, Gareth Griffiths, and Hellen Tiffin, also published in 1989/2002 (Text 3). Text 1 contained 38 sentences, Text 2 has 69 sentences, and Text 3 included 73 sentences. On average, each fragment was 1740 words long. The three books have been translated into Indonesian attesting to their popularity especially for students and academia in the field humanities although this does not necessarily mean they had read them. Content-wise, the three texts discussed relatively similar subject matters, that is, the relationship between language and ideology, thus resting in a relatively similar field of discourse.

The fragments were prepared as digital copies for ease of handling. Since the analysis was carried out at sentence level, each clause complex or clause simplex was segmented from each text. Each clause was divided into Theme and Rheme according to the SFL framework. The clause's Theme(s) were underlined while the Rheme was left in plain text. Since a clause may take a topical and textual and/or interpersonal Themes simultaneously, a further marker was added. For this purpose, a textual Theme is presented in italics, an interpersonal Theme is in small CAPITALS, and a topical Theme was typed in bold, as shown below:

In this example, But is the textual Theme, perhaps is the interpersonal Theme, and this is the topical Theme. The distributions of Themes in the three texts were presented in descriptive statistics, enabling us to see the general trends of the data before qualitative inquiry was applied (Saldanha & O’Brien, 2014). The descriptive statistics were simplified so as to allow us to show only relevant statistical information. 

Participants

Secondly, a questionnaire to figure out the readability of the selected texts was distributed. To this end, five respondents, including three males and two females and ranging in age from 32 to 44 years, were selected in a convenience sampling. These five respondents have a degree in English language and literature, suggesting they could read the texts with comprehension, and, at the time of completing the survey, were teaching in an English Department in a private university in Padang in Indonesia. We did not inform the respondents that the texts’ readability should be related to their thematic structures as this study partly aims to verify the notion of readability based on (linguistic) thematic analysis and responses by actual readers on a random basis.

In the introduction to the questionnaire, the participants were asked to give their consent, declaring their agreement to participate in the study. The written consent form was positioned as the first page of the questionnaire pages as a whole. Correspondences with the participants were done via email, including all the relevant attachments.

The questionnaire was prepared in Indonesian and consisted of attitudinal closed-ended and open-ended questions. The closed-ended questions involved semantic differential scales (see Dörnyei, 2007), asking the respondents to evaluate the readability of the shortened version of the three texts (see Appendix) on a scale of 1-4 (or a 4-point scale: 1 is the easiest to read; 2 is easy to read; 3 is difficult to read; 4 is the most difficult to read). The scores were analysed in simplified descriptive statistics that showed relevant statistical information. We also asked the respondents an open-ended question why they scored each text’s readability they did.

Results and Discussion

Thematic analysis

To begin with, the types or moods of clauses used in the three texts are examined. All of the clauses in Text 1 and Text 3 are of declarative mood. Text 2, in addition to declarative, constitutes imperative and interrogative clauses, although most of the clauses are in the declarative mood. In all three texts, declarative clauses predominate. This confirms what Mauranen (1996) has found that scientific texts are predominantly written in the declarative mood. After all, scientific texts are meant to give information rather than ask for information. The dominant use of declarative mood affects the thematic structure of the three texts. For example, it is expected that topical Themes realized by a subject will predominate. This is explained further below, where the distribution of the topical Themes, comparison of marked and unmarked Themes, and frequency of multiple and simple Themes in Text 1, Text 2, and Text 3 are demonstrated.

The topical Themes' distributions in the three texts are presented in Table 2. Note that since the thematic analysis is carried out at sentence level, the number of the sentences is similar to the topical Themes, whose presence is obligatory, in each clause (complex or simplex).

Table 2: Distribution of topical Themes in Text 1, Text 2, and Text 3.

As shown in Table 1, the topical Themes are only realized by subjects and adjuncts in Text 1 and Text 3. Text 2, interestingly, constitute predicators realizing the topical Themes. Most topical Themes are realized by subjects. Text 3 has the highest percentage, up to 85%, while Text 1 (68%) and Text 2 (68%) show the same percentage. These topical Themes are realized by nominal groups (or noun phrases) and pronouns. Such is the case in all three texts. An example of a topical Theme realized a nominal group (underlined and written in bold) is shown below:

Adjuncts are next. Text 1 has the highest percentage of topical Themes realized by adjuncts (32%), while Text 3 has the lowest percentage (15%). Text 2 falls in between the two texts. This study confirms what Downing and Locke (2006) and Hasselgard (2004) found that subjects and adjuncts frequently realize topical Themes. Prepositional groups and adverbial clauses realize the majority of these adjuncts.  An example of a topical Theme realized by a prepositional group in the three texts is provided below:

Interestingly, Text 2 contains topical Themes realized by predicators. This is absent in Text 1 and Text 3. The results of the analysis reveal that these topical Themes occur in imperative clauses. This suggests, apart from giving information to the reader, the writer of Text 2 (Fairclough, 1989) also demands services from the reader (see Butt et al., 2000). And he does this several times. Below is the example of a predicator imagine that realizes the topical Theme of the clause in question in Text 2:  

The predicator imagine in the example above serves as the topical Theme in the clause. The clause is in the imperative mood in which Fairclough demands the reader to do him a service, that is, to imagine something. The use of imperative mood signal unequal power between the writer and reader. On closer observation, all the predicators occupying a topical Theme are realized by mental Processes (verbs related to mental activities). This includes recall, look, think, see, and imagine. This is not surprising as the nature and mode of the texts prevent the writer from demanding, for example, material Processes from the reader as in a spoken text, like in a normal daily conversation.  

The distribution of marked and unmarked Themes can be deduced in the three texts from Table 2. As explained above, marked Themes are realized by other than the subject of the clause. Table 3 compares the marked and unmarked Themes in Text 1, Text 2, and Text 3.

Table 3: Comparison of unmarked and marked Themes in Text 1, Text 2, and Text 3.

As shown in Table 3, Text 3 has the lowest percentage of marked Themes while Text 1 and Text 2 have the same percentage of marked Themes. The considerably frequent occurrences of marked Themes (mostly realised by adjuncts) in this study are not surprising given that the data are written texts. In written texts, the writer expectedly carefully plan "the rhetorical development of to allow the foregrounding of circumstantial information’ (Eggins, 2004, p. 339). 

The use of multiple Themes warrants further explanation. The frequency of simple and multiple Themes is shown in Table 4:

Table 4: Frequency of simple and multiple Themes in Text 1, Text 2, and Text 3.

The multiple Themes in the data are comprised of textual and interpersonal Themes. On closer examination, it is found that Text 1 does not contain any interpersonal Themes; all of the multiple Themes in Text 1 are realized by textual Themes. Text 2 has four interpersonal Themes, and Text 3 one (one clause in Text 2 and Text 3 shows that a textual Theme and interpersonal Theme co-occur). Text 2 noticeably uses the most interpersonal Themes. Interpersonal Themes are by definition non-cohesive; they hint a perspective on the truth value of the clause message (Hasselgard, 2004) and an indicator to the reader that the propositional meaning of the message is solely taken from the writer's subjective opinion (Matthiessen, 1995) or the writer's point of view to influence the reader (Butt et al., 2000). Put differently, the use of an interpersonal Theme 'de-authorizes' the propositional meaning of the clause that follows. It is for this reason that interpersonal Themes seldom occurs in academic texts. The examples below show the use of interpersonal Themes in Text 2 and Text 3:

The expressions presumably and of course (presented in small capital) in both examples are the interpersonal Themes in Text 2 and Text 3, respectively. Both reduce the authority or truth value of the clause that follows and hint that each clause's propositional meaning is subjectively based on the writer's angle.

As a whole, the frequency of interpersonal Themes can lead to the conclusion that Text 1 is the most authoritative, Text 2 is the least authoritative, and Text 3 rests somewhere in between. However, it should be kept in mind that text authority or truth value of a text does not solely depend on the use of interpersonal Themes. Other aspects like modality, modulation, and polarity have to be considered as well (Butt et al., 2000). Covering these aspects of interpersonal meaning is beyond the scope of this study.

Furthermore, textual Themes occur considerably more frequently in the three texts compared to interpersonal themes. Text 2 has the most textual Themes (20), followed by Text 3 (17), while Text 1 has the least (7). The use of textual Themes is not unexpected, especially in academic texts explored in this study where ideas are often intricately linked. Textual Themes perform a cohesive function, explicitly marking how ideas in the clause and the preceding clause are connected (Hasselgard, 2004). However, the absence of textual Themes does not necessarily make a text lose its logical connection: it simply means that the writer makes implicit the logical connections of the clauses. It is noted that the decision on the part of the writer to include or exclude textual Themes contributes to the readability of the text: the more the textual Themes in the text, the easier it is to read or vice versa. It follows from this that Text 2 appears to have high readability compared to Text 1 with Text 3 falling somewhere between Text 2 and Text 1. The clauses below illustrate the use of textual Themes (in italic) in the three texts:

The textual Themes in these examples are but, furthermore, and accordingly. Each of these connects the ideas of the clause differently from the ideas in the preceding clause. In this sense, the textual Themes indicate continuity of ideas, which is determined by the meaning of the textual Theme in question. Next, the readability of the three texts from the perspective of the independent raters is addressed.

Readability

This study also distributed a questionnaire to five informants, asking them to score the degree of readability of the three texts. The results are shown in Table 5 below:Note: 1 is the easiest to read; 2 is easy to read, 3 is difficult to read, 4 is the most difficult to read.

Table 5: The readability score of Text 1, Text 2, and Text 3

The total score enables us to identify which one of the three texts is easier or more difficult to read. The higher the total or mean score of the text is, the more difficult it is to read or vice versa. The informants unanimously agreed that Text 1 by Bourdieu (1991) has the lowest readability (score = 18; mean (M) = 3.6) and that Text 2 by Fairclough (1989) shows the highest readability (score = 9; mean (M) = 1.8). Text 3 by Ashcroft et al. (2002) falls in between (score = 12; mean (M) = 2.4). The mean scores indicate that the Text 1 falls between ‘difficult to read’ and ‘the most difficult to read’, Text 2 is between ‘the easiest to read’ and ‘easy to read’, and Text 3 rests between ‘easy to read’ and ‘difficult to read’. In general, the scores for each text were relatively evenly distributed and therefore provided a general picture of the readability of the three texts.

The informants were also asked why they scored each text the way they did. They attributed their answer in various ways. This is not unexpected because they were not informed that the questionnaire was intended to explore the relationship between readability and thematic structures.  The informants’ comments can be summed up in terms of clarity of message, cohesive (between sentences/paragraphs), interactivity between the writer and reader, vocabulary (general and specific), sentences constructions (long or short), style (formal and informal) and background knowledge of the subject matter. One informant, for example, writes that “Text 1 is very difficult to read because it lacks cohesive devices between paragraphs and uses too many long sentences that prevent understanding” (our translation). Another informant notes that “Text 2 reads the easiest because of its interactivity, simple sentences, plain vocabularies, and informal style’ (our translation). Still, another informant says of Text 3 as being formal and clear but is suitable for a specific reader who is familiar with the subject matter.

Pedagogical implications

Our study has attempted to show the relationship between thematic properties of three academic texts and their readability. In this section, some pedagogical implications based on the findings of the study will be discussed. In particular, this study can be used as guidance for teachers and learners in their choice of academic texts used in the classroom. This study may inform teachers as to what thematic properties for texts to have so that their students may find them easier to read and comprehend. In this context, the thematic properties that were found to be important for enhancing text readability should demonstrate include interactivity and cohesiveness. These textual aspects will be explained by providing illustrations from this study.

Firstly, the notion of interactivity can be related to the choice of mood, use of first and second persons, and interpersonal Themes. On the one hand, Text 2 consists of declarative, interrogative (questions introduced by a question word or an auxiliary), and imperative clauses, contributing to its high readability. On the other hand, Text 1 and Text 3 are only made up of declarative moods. This suggests that, apart from giving information, Fairclough (1989), the writer of Text 2, is engaged with the reader by asking questions and demanding something to be done. The frequent use of interpersonal Themes also contributes to the degree of interactivity of Text 2. More specifically, this relates to the use of interpersonal Themes in the interrogative clauses, realized by the auxiliary as shown in the clause like:

The frequent use of interpersonal Themes in the declarative clauses weighs in the interactivity of Text 2. This is so because 'moderated' statements through the use of interpersonal Themes warrant a more symmetrical power between the writer and reader. Furthermore, the use of first-person pronoun (I) and second-person pronoun (you) as the topical Themes in Text 2 also makes it more interactive as if Fairclough and the reader were involved in a spoken intercourse. This is illustrated below:

In contrast, Text 1 and Text 2 also employ pronouns as the topical Theme. However, none of them is in the form of first and second persons; all of them are third person pronouns, which is more common in monologue texts. Overall, the variation of mood, the frequent use of interpersonal Themes, and the use of first person and second person led us to conclude that Text 2 is more spoken hence more interactive in nature than Text 1 and Text 2, although, in reality, the three texts are written.

Secondly, it can be seen that the frequent use of textual Themes contributes to the readability of Text 2, as illustrated in Table 4. Semantically, textual Themes explicitly connect the logical meaning of a current clause with the preceding clause, resulting in a cohesive presentation of ideational meanings. In addition, textual Themes serve as signposts for the reader as to what sort of propositional meaning comes next. Using textual Themes allows the writer to give directions for the reader during the course of their reading the text in question. Put differently, the more textual Themes are used, the more directions are given.

Since the presence of textual Themes is grammatically optional, it is reasonable to argue that their use is intentional on the writer's part. This means that the writer intentionally employs textual Themes such that their ideas or meanings are understood more easily, thus producing a text with better readability. While the writer(s) of three texts under examination uses textual Themes, they differ in frequency (see Table 4). Text 2 contains most of the textual Themes, Text 3 comes at a close second, and Text 1 has the least textual Themes. We argue that the frequency of textual Themes in Text 2 contributes to its readability.

The respondents have reported which one of the selected texts is more or less readable. The respondents reveal that interactivity and cohesiveness of the texts affect their readability, thus facilitating their understanding of it. This study has shown that the nature of a text’s interactivity and cohesiveness can be explicitly explained through its thematic structures. It should be noted that thematic structures of a text are not the sole textual property responsible for its cohesion. In addition to conjunctions (textual Themes), a lexicogrammatical resource that is important in distinguishing the selected texts in this study, other cohesion devices include reference, ellipsis, substitution, and lexical (Halliday & Hasan, 1975; Halliday & Matthiessen, 2014). The selections of topical Themes and Rhemes of the clauses, commonly referred to as thematic progression, also contribute to text cohesion (Butt, et. al., 2000; Eggins, 2004). These linguistic devices, however, are beyond the scope of this study.

These findings are expected to be of interest to writers, authors, editors, publishers, and other parties involved in books productions, in general. In particular, this study can help teachers and students, native or non-native, from different levels of education choose what texts they can best use in the classroom for reading activities. While this study focused on examining academic texts in a specific field of discourse and subject matter, its findings can be applied to other types of texts. This is because the notion of thematic structures is more about manner of writing rather than matter of writing. Teachers using English materials for their teachings, for instance, should consider the thematic structures of the texts they use with their students. By doing so, they can ensure that information in the text flows thematically and hence reads easily. Such awareness of thematic structures is also useful for other English skills other than reading. 

Conclusion

This study found that the thematic structures of the three academic texts under examination differ to some extent, despite discussing relatively similar subject matters. The thematic properties analysed include the choice of topical Themes, Theme multiplicity, and Theme (un)markedness. This study found that Text 2 (Language and power by Norman Fairclough) is the most interactive and cohesive, and Text 1 (Language and symbolic power by Pierre Bourdieu) is the least while Text 3 (The empire writes back: theory and practice in post-colonial literatures by Bill Ashcroft, Gareth Griffiths, and Hellen Tiffin) falls somewhere in between. Text 2 stands out as it has more varied thematic properties, followed by Text 3 and Text 1, respectively. For example, Text 2 has more varied Themes used in different clause moods. The variations of Themes of Text 2 include frequent uses of textual Themes, thus contributing to its cohesiveness and interpersonal Themes and topical Themes realized by first-person pronouns, both of which contribute to the degree of its interactivity. Furthermore, using varied moods (interrogative, imperative, and declarative) in Text 2 reinforces its interactivity.

The explanations given by the informants appear to support this as they find that the interactivity and cohesiveness affect their ease of comprehension or readability of the texts. The findings of this study shed light on the relationship between thematic structures (choices of Themes) of a text and its degree of readability: the more varied the Themes of a text, the more readable the text is. Teachers or lecturers are expected to be aware of thematic choices of the academic texts they choose for their course readings. This is to ensure that their students find them readable and hence understandable. Other parties such as writers, editors, and publishers of academic texts are expected to keep such thematic awareness in mind.  

Acknowledgments

We would like to thank the Lembaga Penelitian dan Pengabdian Masyarakat [Institute for Research and Community Service] of Universitas Andalas for supporting this research through the research grant with contract No.: T/29/UN.16.17/PT.01.03/Soshum-RT/2021.

References

Ashcroft, B., Griffiths, G., & Tiffin, H. (2002). The empire writes back: Theory and practice in post-colonial literatures. Routledge.

Baker, M. (2018). In other words: A coursebook on translation (3^rd ed.). Routledge.

Banks, D. (2008). The development of scientific writing: Linguistic features and historical context (Discussions in functional approaches to language). Equinox.

Bloor, T., & Bloor. M. (2004). The functional analysis of English: A Hallidayan approach (2^nd ed.). Arnold.

Bourdieu, P. (1991). Language and symbolic power (G. Raymond & M. Adamson, trans.).  Polity Press.

Butt, D., Fahey, R., Feez, S., Spinks, S., & Yallop, C. (2000). Using functional grammar: An explorer’s guide (2^nd ed.). Macquarie University.

Dontcheva-Navratilova, O., Jančaříková, R., Hůlková, I., & Schmied, J. (2020). Theme choices in Czech university students’ English-medium master's theses. Lingua, 243. https://doi.org/10.1016/j.lingua.2020.102892

Dörnyei, Z. (2007). Research methods in applied linguistics. Oxford University Press.

Downing, A., & Locke, P. (2006). English grammar: A university course (2^nd ed.). Routledge.

Eggins, S. (2004). An introduction to systemic functional linguistics. Continuum.

Fairclough, N. (1989). Language and power. Longman.

Fontaine, L. (2013). Analysing English grammar: A systemic functional introduction. Cambridge University Press.

Gadhessy, M. (1995). Thematic development and its relationship to registers and genres. In M. Gadhessy (Ed.), Thematic development in English texts (pp. 129-146.) Pinter.

Gunawan, W., & Aziza, F. (2017). Theme and thematic progression of undergraduate thesis: Investigating meaning making in academic writing. Indonesian Journal of Applied Linguistics, 7(2), 413-424. http://dx.doi.org/10.17509/ijal.v7i2.8350

Halliday, M. (1985). An introduction to functional grammar. Arnold

Halliday, M., & Hasan, R. (1975). Cohesion in English. Longman.

Halliday, M., & Matthiessen., C. (2014). An introduction to functional grammar. (4^th ed.). Arnold.

Harris, T. L., & Hodges, R. E. (Eds.). (1995). The literacy dictionary: The vocabulary of reading and writing. International Reading.

Hasselgård, H. (2004). Thematic choice in English and Norwegian. Functions of Language, 11(2), 187-212. https://doi.org/10.1075/fol.11.2.03has?locatt=mode:legacy

Hawes, T. (2015). Thematic progression in the writing of students and professionals. Ampersand, 2, 93-100. https://doi.org/10.1016/j.amper.2015.06.002

Hendrawan, W., Sujatna, E. T. S., Krisnawati, E., & Darmayanti, N. (2021). Thematic progression in uncovering paragraph coherence: Some evidences in Indonesian secretarial academy context. TESOL International Journal, 16(1), 179-209.

Kim, M., & Matthiessen, C. M. I. M. (2015). Ways to move forward in translation studies: A textual perspective. Target, 27(3), 335-350. https://doi.org/10.1075/target.27.3.01kim

Martin, J. (1995). More than what the message is about: English theme. In M. Gadhessy (Ed.), Thematic development in English texts (pp. 233-254). Pinter.

Matthiessen, C. (1995). Theme as an enabling resource in ideational ‘knowledge’ construction. In M. Gadhessy (Ed.), Thematic development in English texts (pp. 20-54). Pinter.

Mauranen, A. (1996). Discourse competence: Evidence from thematic development in native and non-native texts. In E. Ventola, & A. Mauranen (Eds.), Academic writing: Intercultural and textual issues. (pp. 195-230). John Benjamin.

Montgomery, S. L. (2010). Scientific translation. In Y. Gambier & L. van Doorslaer (Eds.), Handbook of translation studies (pp. 299-305). John Benjamin.

O’Grady, G. (2019). SFL and critical discourse analysis. In G. Thompson, W. L. Bowcher, L. Fontaine, & D. Schönthal (Eds.). The Cambridge handbook of systemic functional linguistics. (pp. 462-488). Cambridge University Press. 

Olohan, M. (2020). Scientific translation. In M. Baker & G. Saldanha (Eds.), Routledge encyclopedia of translation studies (3^rded.; pp. 510-514). Routledge.

Pfau, M., & Humblé, P. (2019). Análise tradutória de textos acadêmicos em ciências humanas: Um estudo de caso [Translation analysis of academic texts in the human sciences: A case study]. Cadernos de tradução, 39(2), 146-165. https://doi.org/10.5007/2175-7968.2019v39n2p146

Rahnemoon, S. N., Ahangar, A. A., & Nourmohammadi, E. (2017). Thematic structure and translation: A case study of the translation of English news into Persian. Lingua, 194, 26-50. http://dx.doi.org/10.1016/j.lingua.2017.05.003

Saldanha, G., & O'Brien, S. (2014). Research methodologies in translation studies. Routledge.

Steiner, E. (2018). A tribute to M. A. K. Halliday. Lingua, 216, 1-9. https://doi.org/10.1016/j.lingua.2018.10.009

To, V. (2018). Thematic structure in reading comprehension texts in English textbooks. International Journal of Languages, Literature and Linguistics, 4(1), 45-51.

van Valin, Jr. R. D., & laPolla, R. J. (1997). Syntax: Structure, meaning, and function. Cambridge University Press. 

Webster, J. (2015). Understanding verbal art: A functional linguistic approach. Springer.

Wei, J. (2016). Thematic choice in Chinese college students' English essays. English for Specific Purposes, 41, 50-67. https://doi.org/10.1016/j.esp.2015.09.003

Yunita, S. (2018). Theme and thematic progression in students’ recount texts. Indonesian Journal of Applied Linguistics, 7(3), 524-530.

Zulprianto, Fanany R., and Fanany, I. (2019). Thematic structures of paragraph-initial sentences in Animal Farm and its Indonesian translations. In: K. Rajandran and S. Abdul Manan, (Eds.), Discourses of Southeast Asia. (pp. 209-224). Springer.

Zulprianto. (2020). Two Indonesian translations of Animal Farm: Linguistics, ideology and equivalence [Unpublished doctoral dissertation], Deakin University. https://dro.deakin.edu.au/eserv/DU:30145560/zulprianto-twoindonesian-2020.pdf