Intercultural Faultlines

Research Models in Translation Studies: v. 1: Textual and Cognitive Aspects

Maeve Olohan

• Publisher: Routledge
• Year: 2017
• ISBN:
  9781317640714
• Edition: 1
• Topic: Langues et linguistique
• Subtopic: Linguistique

Intercultural Faultlines offers an exploration of research models and methods in translation studies, as implemented, discussed and critically evaluated by some of the leading researchers in the field of translation and interpreting.

While the focus throughout is on textual and cognitive aspects of translation and interpreting, the objects of study and consequently the methodological considerations are wide-ranging. The volume contains chapters focusing on research conducted in areas as diverse as corpus-based translation studies, dialogue interpreting, simultaneous interpreting, acquisition of translation competence, cognitive processes in translation, translation into the L2, creativity in translation and translation quality assessment. Some research models and methods are applied to translation for the first time, while others are more established and can be assessed in terms of their reliability and the generalizability of the results they yield. Issues of research design and methodology are addressed, and interesting questions are raised which are likely to become the focus of attention in future research, for example with regard to causal models of translation, translational ethics, collaborative research and issues of power in interpreting research.

1 Shifts, But Not As We Know Them? Research Models and Methods in Translation Studies

MAEVE OLOHAN

Abstract: Using the models of translation proposed by Chesterman (1998 and this volume) – comparative, process and causal – as a means of Structuring the discussion, this chapter presents an overview of the research models and methods in translation and interpreting employed by some of the leading researchers in the field. While the focus throughout is on textual and cognitive aspects of translation and interpreting, the objects of study and consequently the methodological considerations are diverse. Some models and methods are applied to translation for the first time, while others are more established and can be assessed in terms of their reliability and the generalizability of the results they yield. Issues of research design and methodology are addressed, and interesting questions are raised which are likely to become the focus of attention in future research, for example with regard to causal models of translation, translational ethics, collaborative research and issues of power in interpreting research.

1. Introduction

In the terminology of Lakoff and Johnson (1980), the metaphor in the title of this volume is ontological, enabling us to view activities and ideas as physical entities and thus to refer to, identify and quantify them. The natural concept of the geological fault may help us conceptualize the virtual boundary between cultures and languages – volatile and shifting – as a line along which minor or major displacement can occur, horizontally or vertically, suddenly and with drastic consequences or very gradually and almost imperceptibly.

The metaphor of bridge-building is often used to refer to translation and interpreting. Drawing attention to faultlines may be seen as taking us a stage deeper both physically and figuratively, to the fracture in the earth’s crust and the forces of compression or tension which cause one section of the crust to be displaced relative to another, with the consequences at the surface of the earth which that might have, from earthquakes to the formation of mountain ranges. The study of these faultlines and their effects may be to bridge-building (and tunnel-drilling) what translation research is to translation. We use faultlines to indicate the location of past, present or future displacement; the effect on the surface of any such movements might be modelled, investigated, measured, experienced, described, explained or predicted. The aim of seismological research is to gain a better understanding of the origin of seismic vibrations and the structure of the earth’s crust using sophisticated models, measuring instruments and techniques. In areas prone to such earthquakes, much effort is expended, not just on studying and trying to understand the phenomenon but on designing (and producing) structures, including bridges, which will withstand the force of the tremors.

2. Models of translation

As the less cryptic subtitle of the volume indicates, we are concerned here with research models in translation studies. Andrew Chesterman in Chapter 2 draws our attention to the elusive nature of concepts such as ‘theory’ and ‘model’. He focuses on ‘models of translation’, defining these as “preliminary, pretheoretical ways of representing the object of research”. The choice of a particular model of translation will have implications for the research model and methodology used to investigate it, and for the subsequent construction of translation theories.

Three types of models of translation are proposed: comparative, process and causal. Firstly, comparative models involve discovering correlations between two sides of the relation of equivalence, or more appropriately, similarity. While exemplified by contrastive linguistic approaches where source language system is equated with target language system, we make use of this approach in translation studies in positing relations of similarity between source and target texts. Secondly, process models are dynamic models representing a change of state through time, applied by those interested in sequential relations between different phases of the translation process. Causality may be implicit in comparative and process models but is not central, overt or explicit. Using comparative and process models we can describe translation products and processes, but we cannot explain why they look the way they do, or what effects they may have. However, cause and effect have figured in many approaches to translation and Chesterman proposes that these can be linked in a third model, a causal model of translation, according to which causal conditions produce a target text which, in turn, produces effects. Levels of causation to be considered include cognitive, situational and socio-cultural and these have corresponding levels of effect.

Comparative and process models are contained within the causal model, making it, in Chesterman’s view, the richest model, and the only one which accommodates all four types of hypotheses. All three models make use of interpretive hypotheses and descriptive hypotheses. However, without introducing the notion of causality, the comparative model does not allow explanatory or predictive hypotheses, and the process model does not allow explanatory hypotheses. Both explanatory and predictive hypotheses figure within the causal model in two ways, firstly with regard to the relation between the target text and the causal conditions (i.e. when we propose that a translation has particular features because of a particular cause), and secondly with regard to the relation between effects and translation (i.e. when we propose that a given effect was caused by a particular feature of the translation). Chesterman concludes by advocating a causal model for empirical translation studies research, for the formulation and testing of specific explanatory and predictive hypotheses from research already done, the development of improved conceptual and empirical tools for defining and analyzing translation effects, and the creation of new hypotheses that link causal conditions, features of translations and translation effects.

3. Research methods

As highlighted by Chesterman, the choice of research method will be influenced by the way in which the object of study is represented. While the object of investigation in the chapters of this volume is translation or interpreting, specific textual and cognitive aspects of these activities are focused on in the various studies, and the approach to their investigation is influenced by the underlying model of translation or interpreting, whether comparative, process or causal. The study of textual aspects often involves a comparison between source and target texts, or between translated and original language, whereas investigation of cognitive aspects would seem to fit within a process model. Within research based on these models, we can have a number of different objects of investigation, research models and methods. In this volume, for example, studies of translation and interpreting as a process are presented which focus on:

• the application of models of cognitive processing to the description of how creative translations are produced
• choice network analysis as a method for identifying and describing decision-making nodes and paths in translation behaviour
• design of empirical experimental research to investigate the sub-com petencies of translation competence and their acquisition
• the benefits and drawbacks for translation into the L2 of the use of monolingual L1 corpora in the translation classroom

Studies presented here which are based on the comparative model of translation deal with:

• a methodology for investigating lexical creativity in parallel corpus data
• design issues in compiling and using corpus resources for translation research
• the use of comparable corpora to compare the occurrence of specific linguistic features in translated language with their occurrence in original language
• the application of the repetition model to analyze lexical repetition and cohesion in translation with a view to assessing translation quality

A causal model is one which goes beyond description and which deals explicitly with causes (e.g. why a translation is the way it is) and effects (e.g. why people react in this way to a translation). In this volume, causality is treated in the following studies:

• presentation of research questions which arise when the inferential paradigm of communication is applied to translation
• discussion of translation as cooperation based on the model of the prisoner’s dilemma
• application of personal construct psychology in collaborative research on simultaneous interpreting
• interactionistic approach to research in dialogue interpreting
• consideration of ethical, advocating and empowering dimensions of research in sign language interpreting

Both comparative and process models can be used to describe translation products and translation processes. Chesterman points out that descriptive hypotheses can have explanatory force, and causality may be implicit in both models.

4. Process model of translation: research models and methods

Most of the empirical translation process research of recent years has used introspective methods– the think-aloud method in particular – adopted from psycholinguistics in order to elicit data from translating subjects (e.g. Krings 1986). Stuart Campbell in Chapter 3 acknowledges both the achievements and limitations of think-aloud and word-based experimental techniques, but wishes to propose a third, complementary method of investigating mental processes. Choice network analysis involves analyzing translations of a source text produced by a number of translators working into the same target language; these are seen as a source of evidence of mental processing. A range of differences and similarities in translation behaviour is observed, a complete range of behaviours is approached as the sample becomes larger, and a model of the mental processing can be inferred through comparison and classification of the behaviours observed.

Campbell presents examples of how this method has been applied to the translation of nouns from Arabic into English, passive verbs from English into Arabic and Spanish, cross-clause ellipsis and relative clauses from Arabic into English and complex noun phrases from English to Spanish. He suggests that choice network analysis may also be used to reveal contrasts between the translatability of specific language pairs, to estimate the relative difficulty of parts of source texts, to compare strategies of novice and expert translators, and of those translating into L1 and L2. Campbell also sees it as a useful tool for generating (predictive) hypotheses about translation. He discusses the principles for building networks and the issue of linearity. While he admits that we need to find ways of representing processes which accommodate both linear and simultaneous processing of multiple inputs, choice network analysis represents a method for analysis of multiple translation products which enables decision-making nodes to be identified, graphically represented and described.

The PACTE group at the Universitat Autònoma de Barcelona, whose work is reported on by Allison Beeby in Chapter 4, is interested in gaining a greater understanding of translation competence and its acquisition in order to design teaching methods and materials. Beeby guides us through a flowchart depicting the stages involved in choosing a research model and in designing empirical experimental research, and she presents the decisions made by the PACTE group at each stage. Translation competence, as the object of study, is seen as consisting of six sub-competencies, each of which is defined. Hypotheses are then put forward relating to the nature of these sub-competencies and their interrelations. In order to give this theoretical model scientific value and to validate the suppositions about translation competence, it is deemed necessary to isolate the sub-competencies through empirical research. Working hypotheses are formulated and an experiment is designed to confirm or refute the hypotheses. In designing the experimental research, the PACTE group wished to ensure objectivity, repeatability, validity, generalizability and quantifiability. Beeby details the measuring instruments, subjects and tasks of the experiments. The focus here is on research design and the decisions which this involved, with results from the experiments not yet available.

Paul Kussmaul in Chapter 5 is interested in investigating creativity in translation. Drawing on Preiser’s (1976) definition of a creative product as one which is both novel and appropriate for the task, Kussmaul defines a creative translation as one which “involves changes when compared with the source text, bringing in something that is novel […] and which […] is also appropriate for the task which was set”. Dismissing as non-rational and pre-scientific the notion that creativity is a mysterious and inexplicable phenomenon, Kussmaul advocates the u se of cognitive models and notions to consider questions about translation processes and strategies. Using data from translation classes as examples of creative translation, he applies Fillmore’s scenes and frames semantics (e.g. 1977), together with the notion of prototypicality, to explain the production of translations which are both creative and adequate. Similarly, Langacker’s concepts of figure, ground and canonical viewpoint (1987) and Schank’s thematic organization points (1982) are used to explain creative techniques. Some broader research questions are raised in relation to the applicability of general theories of language processing to translation and the potential of these theories to explain how translation differs from other types of language processing. Empirical investigations of translation processes may provide answers here. Kussmaul also suggests that translation could be used to show what happens during language processing generally. He concludes by suggesting ways in which we could investigate whether knowledge of these cognitive models could improve translators’ ability to be creative.

The issue of creativity is also of interest to Dominic Stewart in Chapter 6, whose starting point is the notion that translations may be more conventional than original texts, as suggested by some corpus-based investigations of translations (e.g. Baker 1993). Stewart focuses his attention on translation into the L2 and proposes the hypothesis that translation into the L2 may exhibit even greater reliance on conventional and recurrent linguistic patterns than translation into the L1. In addition, he questions the effect that the use of corpora during the translation process may have on the translation product in terms of conventionality and creativity. His discussion of creativity and conventionality refers, on the one hand, to the infinite creative potential offered by language and, on the other hand, to our actual use of language which conforms to patterns and does not exploit that creative potential to the full.

Like Kussmaul, Stewart draws his data from the translation classroom, in which his students made use of the British National Corpus for translation into English as the L2. From observing their work, Stewart concludes that this use of the BNC brings clear benefits to non-native speakers of English from a language-learning perspective. However, the benefits for the translation process are, in Stewart’s view, less clear-cut, and it appears that using the BNC may be neither the most profitable nor the most economical way of arriving at acceptable translations for the cases investigated. Moreover, students tended to be heavily influenced in their choice of translation solution by the frequency of occurrence of patterns in the data, not the appropriateness of those expressions for the context in question. The study raises interesting questions as regards the use of corpora in translating. While corpora may compensate for lack of native-speaker knowledge or intuitions, their use tends to highlight habitual usage and recurrent patterns. On the one hand, translated language is perceived as tending to be more conventional than original language; on the other hand, using corpora to assist in the translation process as described above may actually contribute to that normalizing tendency. Stewart reflects on the possibility of corpus-processing software being used in the future (much as we use grammar-checkers and spell-checkers today) to identify unconventional collocations, a technological aid which could hinder or help the production of creative translation products. He concludes by returning to the broader question of whether L2 translation is more conventional than L1 translation or original writing and outlines a research project designed to investigate this hypothesis empirically.

5. Comparative model of translation: research models and methods

Lexical creativity and conventionality also form the basis for Dorothy Kenny’s work, reported on in Chapter 7. Here, as in Stewart’s study, the starting point is an interest in investigating whether linguistic normalization is a feature of translation. Kenny derives her data, not from the translation classroom, but from a parallel corpus consisting of literary texts in German and their English translations (GEPCOLT). Her analysis focuses on hapax legomena (i.e. word forms that occur only once) and writer-specific forms (i.e. forms which occur in the works of a single author only) as these are forms which may be lexically creative and which can be readily identified and retrieved from texts using corpus-processing tools. Similarly, the capability of these tools to produce quantitative and frequency-based data also makes it possible to investigate those collocations which occur only once in a corpus or in an author’s writing, since their use may be considered lexically creative. As highlighted by Stewart in Chapter 6, data retrieved from corpora, for example in the form of concordances, instantly highlight recurrent usage or linguistic patterns, and studies in corpus linguistics and corpus-based translation studies are usually interested in such patterns. Kenny’s work contrasts with this, since her focus is on what is not lexically routine, rather than on what is routine. She presents computer-assisted methods of identifying and accessing these creative features, with examples from her study of lexical creativity in GEPCOLT. The difficulties involved in applying these methods are discussed and suggestions made for refining the methodology. The methods are evaluated in terms of their degrees of precision and recall, and we are reminded of the potential disparity between ...

Table of contents

• Cover
• Title
• Copyright
• Contents
• 1. Shifts, But Not As We Know Them? Research Models and Methods in Translation Studies
• 2. A Causal Model for Translation Studies
• 3. Choice Network Analysis in Translation Research
• 4. Choosing an Empirical-Experimental Model for Investigating Translation Competence: The PACTE Model
• 5. A Cognitive Framework for Looking at Creative Mental Processes
• 6. Conventionality, Creativity and Translated Text The Implications of Electronic Corpora in Translation
• 7. Lexical Hide-and-Seek: Looking for Creativity in a Parallel Corpus
• 8. Parallel Corpora in Translation Studies Issues in Corpus Design and Analysis
• 9. Strange Strings in Translated Language A Study on Corpora
• 10. The Text-organizing Function of Lexical Repetition in Translation
• 11. Issues of Translation Research in the Inferential Paradigm of Communication
• 12. On Cooperation
• 13. Mediating Castles in the Air Epistemological Issues in Interpreting Studies
• 14. Models and Methods in Dialogue Interpreting Research
• 15. Co-constructing Yeltsin – Explorations of an Interpreter-Mediated Political Interview
• 16. Issues of Power and Method in Interpreting Research
• Notes on Contributors
• Name and Subject Index