Extensive research on university student writing (as detailed in the first chapter) has shown us that there is variation, and that it is useful to capture that variation not only as a matter of research interest, but also because of those with a vested interest in university student writing – those who teach it on pre-sessional and in-sessional courses and those who write materials which are used in these courses. So do we really need another approach? A communicative text type approach? Goulart clearly thinks so. She acknowledges disciplinary differences and differences in register, and argues that a communicative approach should focus on purpose. This is a detailed and persuasive argument, so it is worth considering the steps she goes through.
In her introductory chapter (pp. 1–33), Goulart expertly reviews what we know about university writing and how it has been analysed. She ties her definition of assignment type to register, following Biber & Conrad (Reference Biber and Conrad2019). A register can be characterised by the situation in which it is used. For Goulart, essays, critiques, laboratory reports and case studies are all assignment types, or registers. Thus what others have called genres or tasks or assignment types, she calls register. (The term register is also used with multiple senses by different authors and therefore could easily be critiqued as she critiques the term genre, but she does not do that.) Her aim is to classify the communicative purposes of texts (e.g. ‘to argue’, ‘to explain’, ‘to recount’), as these influence the language used. Communicative text types share the same major communicative purpose. In order to illustrate how one register can correspond to several communicative purposes, she decides to look at the various communicative purposes of essays, and examine their situational and textual characteristics. She is interested in how essays are similar across disciplines as well as how they differ. So she has clarified her intent, and the focus on communicative text types that share the same communicative purpose will be intuitively familiar to teachers of English trained in a communicative approach that focuses on language functions (‘to argue’, ‘to explain’, ‘to recount’) or to those familiar with Rose & Martin’s school genres (e.g. see Rose & Martin Reference Rose and Martin2012: 128, where the main purposes are described in more detail, but similar terms). These questions around terminology and the challenges of describing university student writing are picked up in subsequent chapters.
Chapter 3 (pp. 45–65) takes an in-depth look at the challenges of classifying university writing into register categories by examining studies that describe university writing. Here Goulart brings together various classifications of university writing (e.g. Römer & O’Donnell Reference Römer and O’Donnell2011 – Michigan Corpus of Upper-level Student Papers, MICUSP – and Nesi & Gardner Reference Nesi and Gardner2012 – British Academic Written English Corpus, BAWE) and shows both the overlaps between these studies of student texts with other studies of writing prompts as well as the ultimate lack of alignment and the challenges that arise if, for example, student explicit purpose is taken as critical. She crafts such arguments well in favour of a communicative purpose approach, which can describe (a) variation of purposes (e.g. ‘to argue’, ‘to explain’) within the same register (e.g. case study, critique, lab report), and (b) the extent to which the same text combines more than one communicative purpose.
In chapter 2 (pp. 34–44) she describes the corpus and how it is tagged. It contains 347 L1 and L2 undergraduate third-year texts written by 207 students at 60 British universities. The texts are taken from BrAWE (Brazilian Academic Written English Corpus, a corpus of Brazilian study-abroad texts mostly from Life and Physical Sciences), BAWE (a corpus of successful assessed student writing) and the ETS corpus (ETS Corpus of Non-Native Written English, from which texts from Social Sciences and Arts & Humanities in the UK were selected; see Biber et al. Reference Biber, Reppen and Staples2017). By combining these three corpora, a good spread of third-year undergraduate texts is achieved, each averaging more than 2,000 words, across four disciplinary groups: most texts come from Physical Sciences (PS, 36.6%), followed by Life Sciences (LS, 22.0%), Social Sciences (SS, 21.4%) and Arts & Humanities (A&H, 19.8%). The description of the tagging is comprehensive, starting with the Biber tagger, which adds a basic set of tags, then extending this to features found in academic English not originally covered, and adding complexity features (using a Python script). Each text is coded for its communicative purpose (see chapter 4) and its layout (also called textual analysis; see chapter 5), and a multidimensional analysis follows (see chapter 6). This is summarised in table 2.7 and figure 2.2.
Chapters 4, 5 and 6 form the heart of this book, where the major communicative purpose of each text is central. Eight communicative purposes were identified: to argue/take a position, to compare, to describe a tangible object, to explain/analyse, to give personal advice, to narrate a personal event, to give a procedural recount and to propose/express possibility. These are described in table 4.1. For example, ‘to explain/analyze’ is defined as ‘combines information from different sources to demonstrate knowledge of a topic’. The purposes are also compared, for example, ‘To explain involves creating connections between theoretical knowledge and an object of study or situation, while to describe details the characteristics of a concrete/tangible object of study with no need to situate it in the field.’
These major communicative purposes are used to identify text types, but, interestingly, minor purposes are also identified, as presented in table 4.3, where we see, for example, that of all the texts with ‘to give a procedural recount’ as their major purpose, 79% have ‘to explain/analyze’, 20.6% have ‘to argue/take a position’ and 10.1% have ‘to compare’ as minor communicative purposes. This is a level of detail that is novel and demonstrates the value and complexity of identifying both major and minor communicative purposes. Differences are also shown across disciplinary groups. Also, communicative purposes are compared to MICUSP, BAWE and ETS classifications, and found to cut across them significantly; for example, the major communicative purpose ‘to compare’ is found in Essays, Informative Essays, Argumentative Essays and Critiques (table 4.8).
Chapter 5 (pp. 92–113) includes a situational analysis of student writing in general (following Conrad Reference Conrad1996) and a new framework of textual/layout characteristics. This framework includes Text length, Layout (e.g. abstract, headings, non-integral citations), Visual elements (figures, tables) and Explicitness of purpose (stated or not). This perspective provides an interesting overview of the texts (table 5.3) and again this can be presented by disciplinary group (e.g. tables 5.4–5.8). These features are summarised in tables 5.9 and 7.5. Arts & Humanities and Social Sciences texts tend to be longer with more explicit statements of purpose; Life Sciences and Physical Sciences texts tend to have abstracts, headings and visual elements. Life Sciences has the most in-text citations, followed by Arts & Humanities and Social Sciences, with Physical Sciences having the least. None of these findings are particularly novel, but it is useful to see them together.
Chapter 6 (pp. 114–51) presents two multidimensional (MD) analyses of the corpus, by disciplinary group and communicative text type. An MD analysis (which Goulart explains in good detail) involves grouping linguistic features that co-occur in texts and interpreting the outcomes. The first ‘additive’ MD analysis involves comparison with Biber’s (Reference Biber1988) features which were proposed to differentiate spoken and written language. The second ‘new’ MD analysis involves developing new dimensions based on the corpus and its tagging not only for 98 parts of speech, but also for complexity features. Of these, 38 features remained (see table 6.2), from which three dimensions were identified: Dimension 1 ‘Elaborated Discourse vs Condensed Style’, Dimension 2 ‘Production of Possibility vs Content-Focused Description’ and Dimension 3 ‘Informational Density vs Involved, Academic Narrative’.
The first dimension is similar to that found in other studies of academic writing (Gray Reference Gray2015; Gardner et al. Reference Gardner, Nesi and Biber2019), where Arts & Humanities (and to a lesser extent Social Sciences) texts are at the Elaborated Discourse end, while Physical Sciences (and to a lesser extent Life Sciences) texts are at the Condensed Style end. Within disciplinary groups it is also evident that assignments which ‘argue’ or ‘compare’ are more elaborate, while assignments which ‘propose’ or ‘give a procedural recount’ are more condensed.
The second dimension includes Production of Possibility, which is similar to that proposed by Hardy & Römer (Reference Hardy and Römer2013) and Gardner et al. (Reference Gardner, Nesi and Biber2019), where features such as modal verbs, the verb to be, subordinating conditionals and mental verbs co-occur to express possibility. Unlike in Dimension 1, here there seems to be more variation with communicative text type than with disciplinary group. Thus, ‘to explain’, ‘to argue’ and ‘to compare’ load mainly on the negative side of Dimension 2, and ‘to propose’ and ‘to give a procedural recount’ load on the positive side of this dimension.
Dimension 3 with its focus on Informational Density includes features such as word length, nominalisations, attributive adjectives and prepositional phrases. Goulart’s Dimension 3 shows some overlap with Dimension 4 in Gardner et al. (Reference Gardner, Nesi and Biber2019). In both there is disciplinary variation, with Social Sciences and Life Sciences loading together on the positive side (Informational Density), and Arts & Humanities and Physical Sciences loading together on the negative side.
Goulart is to be praised for her discussion and exemplification of the findings of this MD analysis along these three dimensions. Her Dimension 2 has shown that variation occurs not only according to disciplinary group (as in Dimension 1 and Dimension 3) but also according to communicative text type, or purpose. To explore this further she considers variation across Essays.
In chapter 7 (pp. 152–85) Goulart tries to make sense from her communicative purpose perspective of what have been called ‘Essays’. To do this, she identifies 373 texts classed as Essays in the BrAWE, BAWE and ETS corpora and explores variation across communicative text type and discipline.
Goulart then uses her classification of communicative text types to differentiate these texts. She finds that in essays which have ‘to explain’ as their communicative purpose, students reproduce knowledge or display familiarity with a topic, while in essays with ‘to argue’ as their communicative purpose, students give their own point of view about a topic. She also finds variation across disciplinary groups: Essays in Life Sciences are more likely ‘to explain’, whereas essays in Social Sciences are more likely ‘to argue’ (see table 7.3). Most essays have ‘to explain’ as either a major or minor communicative purpose, but the most common combination for essays is ‘to argue’ as the major purpose and ‘to explain’ as the minor purpose. This is important because it reminds us that, at university, students are expected to build arguments and display their understanding of key topics. The combination of purposes varies across disciplinary groups with ‘to argue – to explain’ as most frequent in Arts & Humanities and Social Sciences, but ‘to explain’ alone being most common in Life Sciences and Physical Sciences.
Chapter 8 (pp. 186–208) as the final chapter brings together conclusions and implications. This volume represents a huge amount of work, as each of the 347 texts were annotated for major and minor communicative purposes as well as for textual characteristics, alongside the linguistic tagging conducted for the MD analyses. Table 8.1 presents a good overview of five of the major communicative purposes, with their textual characteristics and linguistic realisations (through Dimensions 1, 2 and 3). These purposes are ‘to give a procedural account’, ‘to argue’, ‘to explain’, ‘to propose’ and ‘to compare’.
The conclusions show that analysis of both communicative text type and discipline is worthwhile. In fact, Goulart suggests that discipline has a greater impact on the language patterns observed in university writing than communicative text type. Consequently, discipline-specific language patterns should be taken into consideration when teaching university writing.
Future suggestions for teaching include paying attention to disciplinary differences, and for research include comparing L1 and L2 writers, and looking at sections of texts to see whether they express different communicative purposes. The Appendices present a huge amount of detail for anyone wishing to follow up on aspects of this study.
So, do we need another classification? There is no doubt that this volume contributes to what we know (and don’t know!) about university writing. The descriptions with examples from specific texts are excellent, and the corpus procedures are detailed, but I am not sure that I would be able to describe a text as ‘to argue’ or ‘to argue – to explain’ and why/how did we reduce the eight communicative text purposes to five text types? I am also surprised that the focus on situational features did not extend to the three corpora, which were built in different contexts.
For me, the two main contributions are (a) a focus on major and minor communicative purposes, and (b) a systematic analysis of textual layout characteristics. The former is complex, initially intuitive but perhaps difficult to apply. It is certainly worth reviewing for those with a detailed interest in university writing, and the writing examples are well chosen. The latter is much easier to understand and apply. Both provide new insights, and whether they are worth the effort depends on your reading purpose, as would be consistent with a communicative approach.
