Bo Wang & Yuanyi Ma: 2022. Key themes and new directions in systemic functional translation studies

Research into Systemic Functional Linguistics (SFL) has a long history, with the construct of SFL being first articulated by Halliday (1961) in his article “Categories of the Theory of Grammar”, which introduced the systemic functional theory of grammar. In his An introduction to functional grammar (Halliday 1985b), the principles of this theory were then further explained. Halliday was interested in the link between grammar and the function and meaning of language, having been influenced by – among others – Malinowski (1923, 1935 and Firth (1950). Malinowski (1935), who had finished an anthropological work on the almost completely unknown (to the outside world) languages, cultures, and contexts in Papua New Guinea, considered translation as a process of interactive contextualization of linguistic structures. Building upon the notion expounded by Malinowski, Firth (1950) maintained that the study of meaning was the raison d’être of linguistics and that it should be viewed in terms of ‘function’ and ‘context’. Halliday took over these notions and initiated SFL, an applicable and holistic linguistic theory, which means it was designed to be applied and remains in constant dialogue with application to real world contexts (Halliday 1985a, 2007, 2008; Matthiessen 2014a). To date, numerous studies have expanded SFL, and now it is capable of examining phonetics, phonology, semantics, and context.

The application of SFL to translation studies dates as far back as the 1950s when SFL was still in its early days (for continuous development of SFL, see e.g. Halliday 1985b; Halliday and Matthiessen 2014; Matthiessen and Halliday 2009). The links between SFL and translation can also be seen in two other works. Hatim and Mason (1990) illustrated three dimensions of context (communication transaction, pragmatic action, and semiotic interaction) and a reconstruction of context through fields, tenor, and mode. Bell (1991), in his book Translation and translating: Theory and practice, introduced three macrofunctions and the different systems they each draw on, and these three macrofunctions developed into the metafunctions in SFL later. Drawing on the metafunctions in SFL (including the ideational, the interpersonal, and the textual metafunctions), Matthiessen (2014b: 272) considered translation as the “recreation of meaning in context through choice”.

Although several seminal studies in translation are informed by SFL theory, many scholars are still not familiar with the terminology of Systemic Functional Translation Studies (SFTS), with some of those even regarding SFTS as taboo. This edited volume, therefore, serves as a useful introduction to clarify the basic concepts of SFTS and to explore the potential areas for further investigation in this field. It contains eight interviews with scholars who have conducted seminal research in SFTS. In the introduction, the two editors provide the reasons for editing this book as well as an overview of its content. The interviews, as collected in Chapters 1–8, cover not only the basic concepts, theoretical developments, and future directions of SFTS but also the interviewees’ current engagement and contributions. The interviewees are all renowned researchers in the field, including Christian M. I. M. Matthiessen, Erich Steiner, Juliane House, Jeremy Munday, Adriana Pagano, and Akila Sellami-Baklouti. Finally, in the concluding chapter, the two editors summarize the chronological development of SFTS and outline future directions the field could move in.

Before embarking on a discussion of different topics in SFTS, the book (Chapters 1 and 2) first explores some key aspects pertaining to its development. Chapters 1 and 2 concentrate on the theoretical foundations of SFTS, providing new insights into translation studies. The opening chapter exchanges thoughts with Christian M. I. M. Matthiessen, who highlights the significance and application of SFL when studying translation. His key points focus on the idea of an intrinsic link between linguistics and translation studies, discussing the multilingual meaning potential within translation and the view of translation as a linguistic process. He also claims that all investigations that have engaged with SFL are, by their very nature, descriptive studies, as well as highlighting the deficiencies in Catford’s (1965) work, followed by an explanation of Matthiessen’s (2001) environments of translation in accordance with the six dimensions. Matthiessen posits that some professional translators have been so focused on ideational meaning with limited emphasis on textual meaning that it is necessary to verify the role of joint metafunctions. At the end of the chapter, he discusses the necessity for extensive text-based research and studies of other language pairs with very significant semiotic distances (e.g. Kiriwinan and English, Quechua and Chinese).

Chapter 2 records Christian M. I. M. Matthiessen’s opinions on SFTS in academia in a dialogic way, including his thoughts on other scholars’ acceptance of SFTS. Different from Holmes’s (1988) map of translation, where translation studies take on an institutional sense as a separate distinct discipline, another map in terms of the number of languages in focus and the cline of instantiation is introduced by Matthiessen. Additionally, he comments on the contributions made by J. R. Firth, Erich Steiner, Mona Baker, Juliane House, and J. R. Martin respectively, and underlines the relationship between SFL and Skopos theory, Polysystem theory, and Descriptive Translation Studies. In this chapter, Matthiessen makes it clear that, in his view, translation should not be separated from other aspects of linguistic theory; instead, translation and interpreting should be viewed together with other metalinguistic processes (drafting, editing, revising, and rehearsing). In this way, the investigation of translation studies supported by SFL will feed back into the development of SFL. Unlike traditional divisions of linguistics, he clarifies the relationship between pragmatics and semantics along the hierarchy of stratification, in that what is covered under the heading of pragmatics is part of a unified but metafunctionally diversified account of semantics. The importance of comparing systems in translation studies is also covered here. Moving on to consider the potential obstacles to and future advancements in SFTS, Matthiessen concentrates mainly on the realms of multilingual semantic potential, linguistic processes, and computational modeling. In terms of future development, he feels it is essential to find a solution to share experiences of working with different language pairs that might not include English and the distance of scholars. This conversation ends with the relationship between literary and non-literary translation in translation studies.

Chapter 3 engages in a conversation with Erich Steiner. Firstly, he gives a brief review of the history of machine translation projects, including Eurotra (Steiner 1986), a project he participated in, followed by an introduction of two corpora he has developed for research: CroCo (Crosslinguistic Corpora for Translation, which consists of eight written text types of original texts and their official translations to investigate explicitation as a possible property of translated texts) (see Hansen-Schirra et al. 2013 for more discussions on the CroCo project) and GECCo (German English Contrasts in Cohesion, the extension of CroCo comparing English and German registers along the whole scale of text types in terms of cohesion) (see Kunz et al. 2021 for more discussions on the GECCo project). Subsequently, derived from the assumption that texts instantiate specific registers, he discusses his view that cohesion specified for different registers helps to improve translation. Thinking about cultural differences, he also talks about the use of register in context depending on the configuration of the intended target language context. Before presenting his ideas related to grammatical metaphor, he first reviews the important concepts that originate from An introduction to functional grammar (Halliday 1985b). Steiner then goes on to explore the distinctions between English and German in terms of four perspectives in SFL, the conclusion of research in translation explicitation, the contributions SFL has made to translation studies, and a presentation of functionally-oriented tools for translation. He predicts there is a danger that translators will become merely semi-qualified appendices to translation technology. To address this, Steiner suggests that a lot of work on ergonomic workflow in translation-oriented technology, statistics-based tools (to help translators make lexicogrammatical decisions), and register-specific corpora merits further study. Finally, he wants to see more contrast studies of different language pairs and studies of the role of translators from two perspectives: research on the translation process and research into how people deal with sensitive ideological contexts, such as translation censorship.

Chapter 4 is a discussion with J. R. Martin, who models translation as re-instantiation. To begin with, he criticizes what he sees as the incorrectly theorized relationship between glossing and free translation and then states the availability of SFL in translation studies from the perspective of dimensions of manifestation (stratification, rank, axis, delicacy) and hierarchies in SFL (realization, instantiation, individuation), both of which make SFL an ideal theory for language typology and contrastive linguistics. Apart from these, the implications of the appraisal system (Martin and White 2005) to translation are displayed at the same time, which contributes to dealing with idiomatic and metaphorical language. According to Martin, there are two criteria to distinguish a translated text done by human beings or with one done by Google: one criterion is syntagms (the collocation and colligation in SFL terms), and the other is choices of meaning. He, therefore, feels that the importance of language description and typology, and of language description, typology, and modeling translation as re-instantiation should not be overlooked. Furthermore, Martin suggests that analyses concerning semantic systems, such as negotiation, ideation, and connexion will be needed in SFTS in the future. After praising Matthiessen for his seminal SFL theoretical framework, he pushes forward several questions on corpus linguistics, ranging from the extent to which machine-assisted analysis can help, and the extent to which computational linguistics can automatically analyze SFL patterns, to the gap between analyzing automatically and manually. In the future, Martin would like to see more theories about translation as sub-potentialization in SFL in two ways: the first is to investigate the meaning potential of culture distributed to different groups (gender, ethnicity, social class, generation, and disability), and the second is to look at individuation in terms of people combining attitude with triggers and targets.

Concerning the main point of developing translation studies as an applied linguistics discipline, Chapter 5 provides key ideas from a dialogue with Juliane House. She defines translation from her point of view and then asserts that the model with SFL as a structure concerned with context is pragmatic, but translation does not equal pragmatics. Through assessments of tenor, field, mode, and register (depending on text function, which divides into interpersonal, textual, and ideational perspectives), House’s model of translation quality assessment (House 2015) ultimately provides the register of the analyzed text. Taking IKEA’s different rules of personal address across languages (House and Kádár 2021) as an example, the term cultural filter is defined by her as a construct with four dimensions to keep the function of the original text equivalent with changes in field, tenor, and mode. When asked about the universals in translation, she rejects this possibility with the idea that translation is foregrounded by other variables, including translators, situations, and translation tasks, and there are no universals in translation. After the comparison of her three books, she states the conditions under which more detailed and serious corpus-based linguistics work would be done and more work about a particular phenomenon, like the shift in a particular register or genre in translation would be carried out.

Chapter 6 begins with a consensus between Jeremy Munday and Matthiessen on a linguistic turn in translation studies, as translation contributes to linguistics and applied linguistics. The discussion then moves on to a systemic model for Descriptive Translation Studies (Munday 1997, 2002) highlighting systemic differences between language and cohesion, as well as studies in translation with an appraisal framework. The conversation is followed by a comparison between the function of SFL and the appraisal system in translation, summarizing that the appraisal system may compensate for the deficiencies of SFL, as the latter does not take hidden meanings into account, whereas the former does. This problem is interpreted as a need for using one’s own pictures for translation studies in different contexts and different modes. Currently, Munday is engaging in the analysis of political speeches, multiple interpreting, and the archive work of literary translation. Concerning the future, he plans to extend discourse analysis beyond the textual perspective to include visual and philosophical perspectives, and there will be more systemic work, particularly the systematic comparison of different languages and multimodal analysis concentrating on discourse analysis and context. Furthermore, he adds that in the future there will be a need for a way for SFL to analyze digital media.

In Chapter 7, Adriana Pagano’s studies primarily center around the translation process, and she focuses first on an introduction and then on the latest developments in this field. Her investigation in process studies targets specific features of translators’ behavior to grasp what goes on in the translator’s thinking, and everything in the experiment will be recorded. Subsequently, she analyzes the potential of SFL for translation process studies with the help of advanced technology because SFL can explain many of the textual operations (modify, change, rewrite) and see text production as a text unfolds. Near the end of the chapter, she illustrates her plans to evaluate different types of texts for machine learning, which aims to train a machine to classify and evaluate texts, develop kinds of chatbots by analyzing records of people talking, and make semantic ontological representations to help machines translate into different languages. As Pagano states, the major focus should be put on human-machine interaction, especially text production processes (including translation, post-editing, and reviewing) in the coming age, and some attention to SFL needs to be transferred to language produced by the computer. Additionally, she notes that a transition of translation process exists in that sometimes we do not have translation at all but the production of a new text from language generation. This dialogue ends with Pagano’s current research on credibility assessments of machine-produced information.

In Chapter 8, Akila Sellami-Baklouti, a key SFL scholar in Tunisia, offers an outline of SFL in Africa. She first illustrates the motivation to apply SFL in translation studies with a focus on register and context, similar to the previous interviewees. Tunisian scholars’ application of SFL to different fields and the course arrangement of SFL and translation shared in the dialogue reflect the vivid developmental status in Africa. Following that, there are investigations (Sellami-Baklouti 2005, 2018) stemming from Sellami-Baklouti’s own experience of reading The adventures of Tom Sawyer. The aspect she discusses here is the clue-bearing function (a correlation exists between the way Twain’s characters use language and their social background), which is hidden in some translated versions with decreased aesthetic value. Thus, she researches the impact of context on the lexicogrammatical realizations in the source texts and target texts with recognition of the power of translators and the goal of translations. On future research needed, she agrees that academia should pay attention to machine translation and that SFL should be informed by translation studies.

The epilogue in this book unfolds from a detailed description of the development of SFTS in light of the timeline. After illustrating the function of Matthiessen’s (2001, 2021 environments of translation and the overall view of the recreation of meaning in translation studies, the two editors summarize some future directions of SFTS based on the previous chapters.

In conclusion, the book presents ideas in the form of interviews. Such an interpersonal presentation brings the readers closer to the scholars. From the editors’ inquiries, readers may have a general impression of the content of each section. Before starting the illustration of the concepts and viewpoints, scholars share their personal experiences that led to their contributions. As the interviews progress, the readers have an in-depth look at the interview content instead of only a simple introduction of the viewpoints of the interviewees.

These interviews contain discussions of all conceptualizations of translation from the SFL perspectives, helping readers to know about this subject quickly, as well as some detailed descriptions of experiment procedures (e.g. the cognitive experiment by Adriana Pagano), which could act as a model for future scholars, especially novices.

The first strength of the book, compared to others, is Chapter 8, which includes an introduction to SFL development in Africa. For a long time, English has dominated as the ‘metalanguage’ in which SFL is constructed in speech and writing (Matthiessen et al. 2010). However, every language has the same status, providing different featured insights. Therefore, the SFL developments in Africa will contribute to the overall construction of SFL.

The two editors also wrote a related book, Systemic functional translation studies: Theoretical insights and new directions (Wang and Ma 2021), which is complementary with this collection of interviews. Although the two books vary in terms of modes: the former is an introduction composed of six themed chapters, and the latter is the transcripts of interviews with six scholars, similar themes are discussed. The former book provides a deeper introduction to some key questions with translation examples and more detailed descriptions, such as the exhaustive design and implementation of CroCo and GECCo. In particular, the illustration of divergencies among schools within SFL (the Hallidayan model, the Sydney model, and the Cardiff model) will address the common perplexities of novices in SFL. However, the collection of interviews fills some gaps in Systemic functional translation studies: Theoretical insights and new directions (Wang and Ma 2021), as it illustrates SFL achievements in China rather than Africa. Overall, these two books complement each other, and no matter what order these two books are read in, it is likely that a fresh opinion on SFTS will be formed in the mind of readers.

In addition, the book possesses a significant strength in interdisciplinary approaches to translation studies, with a particular focus on the fields of neurolinguistics and computational linguistics. And the insights shared by the interviewee promise to have a significant impact on the field in the coming years. To begin with, this book refines the scope of translation studies as part of linguistic studies rather than as a separate field, unlike previous studies in which the understanding of translation was limited to Holmes’s (1988) map. In the interviews, Matthiessen (Chapters 1 and 2), Steiner (Chapter 3), and Pagona (Chapter 7) introduced several cutting-edge studies linking neuroscientific research to translation and the interpreting process with the help of eye tracking and keystroke logging devices. Neuroscience is an increasingly important area, particularly after the release of ChatGPT. This interdisciplinary research will not only improve the quality of translation, translators’ cultivation, and foreign language acquisition in the linguistic domains, it will also lift obstacles to a real Artificial General Intelligence (AGI). Additionally, the use of empirical and experimental research tools in collecting data provides strong support for the assumption of cognitive activity in the translation process, which leads linguistic studies to a more scientific image. It also discusses the development of computational linguistics in terms of SFL, including text understandability and simplification, multilingual text production, machine translation, and the post-editing process. Apart from the significance of the application, the need for theoretical development is also stated in the interviews, such as Pagano’s. Since SFL has always focused on language produced by humans, the text produced by machines or the combination of machines and human beings requires further investigation to find out to what extent human language will change as a result of the language produced by machines and what the impact of machine-produced language on human language is. It is essential for SFL, a descriptive theory, then, to analyze the newly arising language phenomenon that can improve human-machine interaction in turn, but this vital question seems to be overlooked nowadays. Additionally, unlike the traditional disciplinary boundaries with macro and micro perspectives, SFL offers a new map to present the linguistic track, and Matthiessen states that the relationship between pragmatics and semantics should be clarified and that these two disciplines share an overlapping scope. This book extends the realm of translation studies, providing more research methods to apply SFL research to solving practical problems.

Furthermore, another strength exhibited by the book is forward thinking for translation studies, especially the vision of translators under the profound impact of Large Language Models (LLMs). Although the famous artificial intelligent model ChatGPT has passed the Turing test, it still cannot replace human translators in translating, let alone in the more complicated interpreting. Compared to human translation, GPT’s translation quality needs to be improved further. From the criteria of human evaluation, it can be concluded that with GPT, some word choices are inaccurate or errors although the meaning of the sentence can be inferred from context. Besides, as for discourse awareness, some sentences may seem abrupt due to a lack of linkage, and tone and word choice are inconsistent, noticeably affecting the accuracy of reading comprehension. These problems that LLMs face completely respond to J. R. Martin’s method to distinguish translation from humans or Google mentioned before. Although SFL is not used in the currently dominant neuro-network technologies, it still has the potential to be applied in translation studies as a theory of context. Matthiessen (2014b: 272) redefined the meaning of translation as “recreation of meaning in context through choice”, raising a crucial question in the discourse awareness of machine translation, i.e. how different kinds of meaning unify and balance in text. In light of this, some of the interviewees imported the theories of discourse analysis into traditional translation studies. In terms of word choice, the frequency of every possible following word is very low in effect in a huge database like the one of GPT-4, but with register-specific corpora, the statistic-based model will possibly provide more dependent frequent choices just, like Erich Steiner’s solution to predicted danger mentioned in the interview. Additionally, another way to improve accuracy is the further development of multilingual studies (language comparison, contrastive analysis, and language typology) as Matthiessen mentions in the interview, by forming deeper and closer relations between more language pairs.

The book, as the first collection of interviews on SFTS, offers an overview of this detailed and holistic analysis system. Hence, it is essential reading for anyone seeking a deeper understanding of translation studies. It is hoped that a second volume will be published, which will incorporate more information and updates of research in China (as indicated in Huang et al. 2023), South America, and other regions around the world. In addition, other theoretical perspectives informed by SFTS can also be discussed in future volumes, such as the application of Rhetorical Structure Theory in SFTS, the balance between automated analysis and manual analysis, the influence of LLMs on translation studies, the interface between SFTS and other linguistic traditions (e.g. cognitive linguistics, neurolinguistics).