Károly, Krisztina (2017): Aspects of Cohesion and Coherence in Translation: The case of Hungarian-English news translation. Amsterdam/Philadelphia: John Benjamins, 269 p.

This book is a technical but engaging study of the way that certain types of cohesion and coherence are (re)produced in translation. Károly has developed an innovative approach to studying cohesion and coherence in translated texts. The approach stems from an interdisciplinary perspective which allows her to apply tools used in discourse analysis to the study of translated texts. The work will be of particular value for scholars interested in cohesion and coherence in translation, and I can easily imagine other scholars using the model developed and illustrated by Károly to analyze other corpora and text types. Scholars interested in news translation as a genre, however, have less to learn from this book. It does not, and nor is it meant to, constitute a study of Hungarian-English news translation per se.

The book begins with a short foreword which is itself followed by an equally short introduction. The foreword places the study in a wider research context, lying between translation studies and discourse analysis, while addressing “neglected or under researched” (p. IX) questions. It also outlines the main fields to which the work contributes: discourse analysis, genre studies and translation studies, wherein “[t]he most significant contribution of this study is to the field of target text oriented translation research” (p. XI). The introduction then offers an overview of the study by briefly outlining the research focus, the aims and research questions, the corpus, as well as the theoretical framework. The importance of the methodological contribution of this book clearly emerges at this stage:

[T]his book attempts to devise a translational discourse analysis model that is capable not only of offering a sufficiently comprehensive description of the various means of creating coherence in translation, but also of identifying the language pair specific shifts of coherence in translation and the discourse level strategies that accompany these.

p. 2

From a structural perspective, it would almost certainly have been preferable to combine the foreword and introduction, but they nevertheless perform the function of contextualizing and setting up the study.

The second chapter of the book provides a very detailed account of the theoretical background to the study and to the Complex Translational Discourse Analysis (CTDA) model that is developed here. The structure of the chapter is somewhat unusual, but it is broadly divided in two parts: the first part deals with translation and the second with cohesion, while the CTDA model is described right in the middle (section 2.3.2). As Károly notes, the model is based on theories that “were originally developed for the English language and for the analysis of ‘independent’ texts” (p. 52) so the main innovation lies in the development of a model capable of treating translated texts alongside the originals. It is also worth noting that Károly describes the CTDA model as “complex” “since it contains all of the discourse structural components of coherence that create continuity in discourse” and “includes all the theories and methods necessary for their identification and description” (p. 52). The wording is perhaps a touch hyperbolic, but it nevertheless underscores a major benefit of the model: it lets the scholar examine a range of different components of coherence at once, and allows us – as Károly does in the final chapter – to explore interactions between the various levels.

The corpus used to examine coherence and cohesion is described in the third chapter. It consists of 20 “‘summary’ sections of translated English analytical news articles and their corresponding Hungarian originals retrieved from the website of Budapest Analyses” published between 2004 and 2008 (p. 99). Translation scholars in general, and scholars of news translation in particular, are likely to find fault with the corpus design. Most obviously, at 6618 words, the corpus is very small. Károly acknowledges this and attributes it to “the extremely laborious and meticulous nature of the investigations” and to the “amount of data” that these texts represent for the kind of analysis being carried out (p. 99). The need for a small corpus is certainly understandable, but I am not convinced that it had to be this small. Moreover, the size of the corpus may be the main reason why very few of the results reported in the analysis chapters (4-7) turn out to be statistically significant. The other feature of the corpus that will surprise scholars of news translation is the source of the news articles. Although articles on the Budapest Analyses website are certainly translations of news from Hungarian to English, they definitely cannot be considered representative of the genre because, as Károly notes, the translations were done by translators, “contrary to the general practice in news translation where translation work is done by journalists” and it appears that they were produced by non-native speakers of English (p. 99). The study of non-standard practices such as these is definitely of value in our field, but the choice to use translations carried out by non-native speakers of English in particular was probably not the most sensible for Károly’s project; the reader cannot help but wonder whether she would have found greater changes in coherence and cohesion at the hands of native speakers.

The meat of the analysis is presented in chapters 4-7, each of which deals with a separate kind of cohesion. The aim of each chapter is to determine whether there are “shifts” in cohesion as the text is translated into English. Each chapter has the same structure: they start with an outline of the aims of the analysis which is explicitly linked to the general research questions; the methodology for the particular study is then presented; this is followed by the results of the analyses which include both quantitative and qualitative components; and the chapters end by assessing the contribution made to our understanding of translation universals.

Chapter 4 deals with four types of cohesive ties: reference, substitution and ellipsis, conjunction, and lexical cohesion. Károly finds no statistically significant quantitative evidence of changes to these ties in translation, but she does find qualitative evidence of shifts affecting three of the four types (reference, conjunction, and lexical cohesion) (p. 137). Chapter 5 focuses on lexical repetition and its function in source and target texts. Once again, the quantitative results showed no statistically significant evidence of shifts taking place in the translations although there are qualitative differences (pp. 143, 145). Károly highlights a potentially interesting phenomenon for translation scholars when she speaks of a ‘leveling up’ in some cases where there is no change in the absolute number of repetitions, but where one kind of repetition is substituted for another (p. 145).

The sixth chapter examines topical development and structure. Here, the quantitative analysis does reveal some significant differences between the source and target texts so Károly is able to “provide further justification to the claim that translation is generally accompanied by changes/shifts in the topical structure of the source text” (p. 167). The qualitative analysis sheds more light on the differences by showing that they are most frequent in the background section of the news story. Károly is surprised by this result because of the importance of the background section in translation “where the target audience […] may not possess the required, source culture and language related background knowledge to be able to understand the text” (p. 170). She therefore suggests that “changes in the topical structure of this component may […] make processing more difficult and endanger the correct/adequate interpretation of the translation.” I would have thought that the opposite would be true: changes are particularly frequent in the background section because it is here that the differences between the source and target readers’ knowledge would be the greatest and these changes in fact augment the reader’s comprehension of the text.

The final data chapter examines rhetorical structure through an analysis of the frequency, types and position of relational propositions. Once again, the quantitative results show no statistically significant difference between the source and target texts (p. 193). Qualitative analyses, on the other hand, do reveal some minor differences in both the type and position of relational propositions and it is interesting that “only eight of the 20 texts have completely identical rhetorical structures” (p. 196). Károly finds that shifts occur most frequently in the commentary section of the news stories and this, she argues, means that “consciously or unconsciously […] translators modify the relational propositional content of this component and thus, possibly, very subtly alter the intended message” (p. 200, author’s italics).

The eighth and final chapter brings the results of each of the data analysis chapters together. It reiterates the main finding across all of the chapters, namely that “[i]t is […] not so much the frequency, but rather the quality (the type, the location, the optional/automatic nature) of the shifts that produce changes at the various levels of discourse structure” (p. 213). As regards translation universals, there was no statistically significant evidence in support of either of the two main universals under consideration here (the explicitation hypothesis and the repetition avoidance hypothesis). However, Károly does rather go out on a limb, despite all the hedging, when she concludes, based on her study, that “shift of coherence as a phenomenon is special to the process of translation and may thus (possibly) be regarded as a translation universal” (p. 212, author’s italics).

Aspects of Cohesion and Coherence in Translation: The case of Hungarian-English news translation is certainly a useful contribution to the field of translation studies. Károly’s main contribution is methodological: she shows that it is possible to test empirically what happens to cohesion and coherence in translation. She has devised a model that other scholars can now use to explore the same phenomena in different corpora of translated texts. Her quantitative work is a particular strength: she is meticulous about reporting the results of statistical significance tests, and it is particularly laudable that she can quantify the validity of the manual coding by having each analysis performed by two different researchers. The main drawback of the study is without doubt the small size of the research corpus which unfortunately makes it difficult to know whether Károly’s conclusions really are a reflection of the wider reality. Readers will no doubt be convinced that there are qualitative shifts affecting cohesion and coherence in translation. However, the very small number of texts in the corpus means that it is impossible to know whether the lack of statistically significant differences in the frequency of the shifts is a true reflection of reality.