1.Introduction

This study contributes to the development of methodologically sound interdisciplinary cooperation between translation studies (TS) and sociology by presenting a methodology of joint translatological–sociological (TRA–SOC) cooperation in data collection, analysis, and interpretation, for research in the area of translation strategies and norms. It draws on ongoing research that aims to identify and interpret norms of literary translation from English to Czech in the time period 2000–2016. Conceptually, the research relies on Toury’s notion of translation as a social activity and norms of translation as explanatory hypotheses arrived at by the generalisation of regularities in translation strategies (Toury 1995Toury, Gideon 1995 Descriptive Translation Studies – and Beyond. Amsterdam: John Benjamins. , 1999Toury, Gideon 1999 “A Handful of Paragraphs on ‘Translation’ and ‘Norms’.” In Translation and Norms, edited by Christina Schäffner, 9–31. Clevedon: Multilingual Matters., 2012Toury, Gideon 2012 “The Nature and Role of Norms in Literary Translation.” In The Translation Studies Reader, 3rd ed., edited by Lawrence Venuti, 168–181. Abingdon: Routledge.). Translatologically speaking, our ongoing research and this study are product-oriented in design, which has two consequences. First, we do not apply process-oriented methods such as think-aloud protocol analysis or other cognitively oriented research methods. Second, we encounter translation strategies in the form of shifts and in line with Toury we assume that “regularities of surface realization and/or translation-source relationships and shifts bear immediate witness to regularities of translational behaviour” (1995, 182). We also apply his distinction between obligatory and optional shifts and share his view that

since the need itself to deviate from source-text patterns can always be realized in more than one way, the actual realization of so-called obligatory shifts, to the extent that it is non-random, and hence not idiosyncratic, is already truly norm-governed. So is everything that has to do with non-obligatory shifts …(57)

For example, one of the key systemic differences between English and Czech is the nominal character of the former and the verbal character of the latter, but the tendency of Czech to prefer finite verb forms instead of non-finite ones (which do exist in the system) can be realised in many different ways by translators as the realisation is not rule-based (for further details see Tárnyiková [2007]Tárnyiková, Jarmila 2007 Sentence Complexes in Text: Processing Strategies in English and in Czech. Olomouc: Palacký University Olomouc.).11.On the contrary, it is norm-based and the fact that it can be relatively easily quantified even for large amounts of texts (supposing a corpus-based approach is used) makes it an ideal formal indicator that yields a clear picture of the strategies employed by translators and the resulting shifts in this area.

The principal research methods used in the study are translatological comparative textual analyses and sociological semi-structured interviews with translators. By means of textual analyses, we identify regularities in translation strategies which are assumed to be the observable results of the norms of translation. By means of interviews, we seek to (a) situate the translators in their respective social contexts (we analyse their biographies, family backgrounds, generational anchorage, education, preferences, attitudes, and value orientations), as we share Chesterman’s (2016Chesterman, Andrew 2016 Memes of Translation: The Spread of Ideas in Translation Theory. Rev. ed. Amsterdam: John Benjamins. , 86) presupposition that these social variables influence their choice of translation strategies; and (b) find out in what way the translators reflect on their practice.

Given its objectives, our research comprises three sub-tasks: (1) analyse the Czech field of literary translation including its position in the world system of translation, (2) determine representative samples of translations and translators, and (3) compare textual analyses of representative samples of translations and the social contextualisation of their translators and their practice.

The research project’s main focus is Task (3), but its completion requires that Tasks (1) and (2) be accomplished first. Tasks (1) and (2) are not included in the study presented in this article; however, their results are built into it where relevant. The results of Task (1) foster the interpretation of the results of the primary Task (3), as they reveal the indispensable socio-historical background of the translators’ decision-making. On the basis of our analysis of the Czech field of literary translation we can claim that within the world system of translation (Heilbron 1999Heilbron, Johan 1999 “Towards a Sociology of Translation: Book Translations as a Cultural World-System.” European Journal of Social Theory 2 (4): 429–444. , 2010Heilbron, Johan 2010 “Structure and Dynamics of the World System of Translation.” Paper presented at the International Symposium: Translation and Cultural Mediation, UNESCO , Paris, February 22–23.; Heilbron and Sapiro 2007Heilbron, Johan, and Gisèle Sapiro 2007 “Outline for a Sociology of Translation: Current Issues and Future Prospects.” In Constructing a Sociology of Translation, edited by Michaela Wolf and Alexandra Fukari, 93–107. Amsterdam: John Benjamins. ), Czech is placed at the border of semi-central and peripheral languages. Šimeček and Trávníček (2015)Šimeček, Zdeněk, and Jiří Trávníček 2015 Knihy kupovati … Dějiny knižního trhu v českých zemích [To buy books … A history of the book market in the Czech lands]. Prague: Academia. state that the Czech literary field is highly developed since more than ten times as many books are published than would correspond to the representation of the Czech population in the world’s population. Translations from English occupy a prominent position, accounting for approximately 50% of all translations (401–402). It can thus be expected that the translation strategies of translators from English to Czech will lean towards foreignising ones (Heilbron 2010Heilbron, Johan 2010 “Structure and Dynamics of the World System of Translation.” Paper presented at the International Symposium: Translation and Cultural Mediation, UNESCO , Paris, February 22–23., 6). The results of Task (2) comprise a methodology for creating representative samples of translations and translators (Zehnalová and Kubátová 2019Zehnalová, Jitka, and Helena Kubátová 2019 “From a Target Population to Representative Samples of Translations and Translators.” The Translator 25 (2): 87–100. ) and its application for the purposes of our research (i.e., the gradual development of representative samples of translations and translators as the object of the joint TRA–SOC analysis). Representativeness is necessary to ensure the generalisability of findings regarding regularities in the translation strategies and norms of Czech literary translation in the given period.22.Our target population consists of 15,381 translations. The calculated representative sample size is 5.6% (i.e., 867 translations). The research samples of translations and translators are established during the research on the basis of purposive and theoretical sampling. For further details see Zehnalová and Kubátová (2019)Zehnalová, Jitka, and Helena Kubátová 2019 “From a Target Population to Representative Samples of Translations and Translators.” The Translator 25 (2): 87–100. . To ensure the reliability and generalisability of the findings of Task (3), a sound methodology is needed. Within the proposed joint TRA–SOC methodology, analyses are carried out from the perspective of the translator and on three levels: doxa, results of practice, and the practice itself.

2.Theoretical framing

In sociological approaches to translation, the contextualisation of translators is pursued at both the individual and the social levels, as translation practice – that is, the use of translation strategies – is based on the individually and socially conditioned decision-making of translators:

Any translation is necessarily bound up within social contexts: on the one hand, the act of translating, in all its various stages is undeniably carried out by individuals who belong to a social system; on the other, the translation phenomenon is inevitably implicated in social institutions, which greatly determine the selection, production, and distribution of translation, and, as a result, the strategies adopted in the translation itself.(Wolf 2012Wolf, Michaela 2012 “The Sociology of Translation and Its ‘Activist Turn’.” In The Sociological Turn in Translation and Interpreting Studies, edited by Claudia V. Angelelli, special issue of Translation and Interpreting Studies 7 (2): 7–21. , 10)

TS scholars have drawn on the concepts of influential sociologists (for a survey, see Wolf [2010]Wolf, Michaela 2010 “Sociology of Translation.” In Handbook of Translation Studies, vol. 1, edited by Yves Gambier and Luc van Doorslaer, 337–343. Amsterdam: John Benjamins. ). The most frequently applied sociological theory is Bourdieu’s (1990)Bourdieu, Pierre 1990 The Logic of Practice [orig. Les sens pratique ]. Translated by Richard Nice. Stanford: Stanford University Press. theory of practice, mainly his notion of habitus, introduced to TS by Simeoni (1998)Simeoni, Daniel 1998 “The Pivotal Status of the Translator’s Habitus.” Target 10 (1): 1–39. and then applied and discussed by others (for surveys, see Sela-Sheffy [2005]Sela-Sheffy, Rakefet 2005 “How to Be a (Recognized) Translator: Rethinking Habitus, Norms, and The Field of Translation.” Target 17 (1): 1–26. , Vorderobermeier [2014]Vorderobermeier, Gisella M. ed. 2014 Remapping Habitus in Translation Studies. Amsterdam: Rodopi. , and Xu and Chu [2015]Xu, Minhui, and Chi Yu Chu 2015 “Translators’ Professional Habitus and the Adjacent Discipline: The Case of Edgar Snow.” Target 27 (2): 173–191. ), as it makes the connection of the individual and social levels possible:

In Bourdieu’s theory, the habitus concept stands for his heroic attempt […] to synthesize the zoom-in and zoom-out perspectives on culture, namely, to frame the actions of the individual on the micro-level within the grand-scale durable social structures.(Sela-Sheffy 2014Sela-Sheffy, Rakefet 2014 “Translators’ Identity Work: Introducing Micro-Sociological Theory of Identity to the Discussion of Translators’ Habitus .” In Vorderobermeier (2014, 43–58). , 43)

Pasmatzi (2014)Pasmatzi, Kalliopi 2014 “Translatorial Hexis and Cultural Honour: Translating Captain Corelli’s Mandolin into Greek.” In Vorderobermeier (2014, 73–92). argues that the social and individual aspects of habitus form the focuses of two trends within TS: the first investigates the collective attributes of habitus and their macro-manifestations, while the second deals with its micro-manifestations, namely the tendency of translators to translate particular texts and, to a certain extent also, their concrete linguistic, stylistic, and textual choices. Most Bourdiesian TRA–SOC studies belong to the first strand (73).

The relatively frequent use of the notion of habitus in TS has given rise to some critical objections as well, mainly concerned with its alleged determinism (Meylaerts 2008Meylaerts, Reine 2008 “Translators and (Their) Norms: Towards a Sociological Construction of the Individual.” In Beyond Descriptive Translation Studies: Investigation in Homage to Gideon Toury, edited by Anthony Pym, Miriam Shlesinger, and Daniel Simeoni, 91–102. Amsterdam: John Benjamins. ; Sela-Sheffy 2014Sela-Sheffy, Rakefet 2014 “Translators’ Identity Work: Introducing Micro-Sociological Theory of Identity to the Discussion of Translators’ Habitus .” In Vorderobermeier (2014, 43–58). ) and its internal inconsistency (Ben-Ari 2014Ben-Ari, Nitsa 2014 “Political Dissidents as Translators, Editors, and Publishers.” In The Sociological Turn in Translation and Interpreting Studies, edited by Claudia V. Angelelli, special issue of Translation and Interpreting Studies 7 (2): 144–160. ). Both can be related, on the one hand, to the way the concept is perceived and, on the other, to the two previously mentioned research trends and the resulting orientation of specific research projects. From the perspective of macro-approaches concerned with collective attributes, habitus may appear as too deterministic, as individual attributes and motivations are not the primary focus. Micro-approaches directed at individual attributes see many diverse influences that are not shared among translators and, as a result, perceive the concept of habitus as internally incoherent. It is also worth noting that translation strategies, the main area of translators’ decision-making and of translation practice, are not foregrounded in macro-approaches; and while they are investigated within micro-approaches, typically only a small number of translations are covered, which does not allow for comparisons and for placing individual characteristics into a broader, collective perspective on the decision-making of a larger number of translators.

Hanna (2014)Hanna, Sameh F. 2014 “Remapping Habitus: Norms, Habitus and the Theorisation of Agency in Translation Practice and Translation Scholarship.” In Vorderobermeier (2014, 59–71). points out that in TS, Bourdieu’s self-reflexive sociology is applied along with translatological research methods that have epistemological and methodological limits. These, according to him, are linked to the criteria for the selection of the researched materials, ways of categorisation, as well as the concepts used by researchers to interpret phenomena (70). In our research, we aim at the generalisation of regularities in translation strategies at the level of translation norms, and consequently we work with representative samples (which represent large target populations) of translations and translators. This means that we gradually obtain data on the collective level, but at the same time have access to data on individual aspects of translation strategies as well as the professional and other dispositions of translators. Therefore, we can perceive both the collective and the individual aspects – individuality does not disappear; on the contrary, it is foregrounded against the background of the collective, and vice versa.

According to Bourdieu’s theories of social fields and practice, objective and subjective conditions intervene in translators’ decision-making. In our context, the primary objective structure is the pressure of the Czech literary translation field which, in different forms, generates the possibilities and limitations of translation. These are mainly constituted by the norm of adequacy preferred by the field, the evaluative and assessment criteria of experts in the field, the policies and strategies of publishing houses, and the conditions of the Czech book market. As the fundamental subjective structure intervening in the decision process, we consider the translator’s habitus. In agreement with Bourdieu (1990)Bourdieu, Pierre 1990 The Logic of Practice [orig. Les sens pratique ]. Translated by Richard Nice. Stanford: Stanford University Press. , we conceive of habitus as a ‘mental’ individual and individualised structure, an ‘unconscious’ disposition to act and think in a certain way, which includes skills, abilities, worldview, expectations, schemes of perception, assessment, and behaviour. Through the habitus, agents perceive the world around themselves, think about it in a specific way and act in it, as well as react to the given context of specific social and historical conditions. However, because “as soon as he reflects on his practice, adopting a quasitheoretical posture, the agent loses any chance of expressing the truth of his practice, and especially the truth of the practical relation to the practice” (Bourdieu 1990Bourdieu, Pierre 1990 The Logic of Practice [orig. Les sens pratique ]. Translated by Richard Nice. Stanford: Stanford University Press. , 91), the researcher needs to be cautious when asking agents about their practice or about its impact (see further explanations concerning doxa in this section). Habitus can be arrived at by the analysis of practice and the results of practice. These are the reasons that underpin our decision to ground the research sociologically on the semi-structured interviews and translatologically on the textual analyses: by means of these two methods, we are able to ascertain the results of practice and from these, we can reach conclusions about the practice itself.

In research on the norms of (literary) translation within the sociology of translation, habitus can be applied as a principal notion because it allows translatological textual analysis to complement sociological analysis: “While the study of norms renders a map of ‘depersonalized translation tendencies’ so to speak, the study of the habitus should account for how these norms are incorporated in the minds and practices of individual translators” (Sela-Sheffy 2014Sela-Sheffy, Rakefet 2014 “Translators’ Identity Work: Introducing Micro-Sociological Theory of Identity to the Discussion of Translators’ Habitus .” In Vorderobermeier (2014, 43–58). , 43). As stated by Wolf, habitus, and textual and social analyses are interconnected:

The postulate that the habitus of an individual or a collective can be reconstructed through his or her various activities […] is of paramount interest for the understanding of the translation process because it helps trace the interaction between (translation) text analysis and social analysis.(Wolf 2012Wolf, Michaela 2012 “The Sociology of Translation and Its ‘Activist Turn’.” In The Sociological Turn in Translation and Interpreting Studies, edited by Claudia V. Angelelli, special issue of Translation and Interpreting Studies 7 (2): 7–21. , 13)

We agree that habitus and textual analyses are interconnected and that habitus can be reconstructed through various activities. However, these activities (i.e., the translator’s practice) are not directly observable, unlike, for example, cultural practice (table manners, sports, etc.) analysed by Bourdieu (1984)Bourdieu, Pierre 1984 Distinction: A Social Critique of the Judgement of Taste [orig. La Distinction: Critique sociale du judgement ]. Translated by Richard Nice. London: Routledge and Kegan Paul.. Therefore, we pay attention to the results of practice and assume that they might point to the practice itself. This should clear the way for establishing not only how norms are incorporated in the minds of translators, but also, and more importantly, how they are applied in practice.

At the same time, we take into account the significant fact that in the semi-structured interviews, the source of information about the practice of translators is the translators themselves. What they say they are doing is doxa (i.e., their idea about their practice), not the practice itself. Bourdieu and Wacquant (1992)Bourdieu, Pierre, and Loïc J. D. Wacquant 1992 An Invitation to Reflexive Sociology. Chicago: University of Chicago Press. define doxa as “an uncontested acceptance of the daily lifeworld” (73). Agents, in our case translators, take their lifeworld for granted and are thus not fully aware of the social, economic, political, and other conditions forming and limiting their perception, assessment, and acting: “It is because agents never know completely what they are doing that what they do has more sense than they know” (Bourdieu 1990Bourdieu, Pierre 1990 The Logic of Practice [orig. Les sens pratique ]. Translated by Richard Nice. Stanford: Stanford University Press. , 69). Similar to Sela-Sheffy (2014Sela-Sheffy, Rakefet 2014 “Translators’ Identity Work: Introducing Micro-Sociological Theory of Identity to the Discussion of Translators’ Habitus .” In Vorderobermeier (2014, 43–58). , 44–45), we accept that translators typically do not follow conscious and explicitly describable opinions, but instead the unconscious “feel for the game” (Bourdieu 1990Bourdieu, Pierre 1990 The Logic of Practice [orig. Les sens pratique ]. Translated by Richard Nice. Stanford: Stanford University Press. , 66). We thus do not mistake their narrative accounts of their translation practice for the theoretical explanation of these accounts. Instead, we analyse the narrative accounts in order to uncover the principles underpinning translation practice and the possibilities that translation practice encompasses. All this means that research should not confuse practice with doxa. Consequently, the question we pose is how to explore the practice of (literary) translators – whether it can ‘speak’ and how to ‘make it speak’.

Sociologically, our methodology relies on the definition of practical logic, which asserts that practice cannot speak in any discourse (Bourdieu 1990Bourdieu, Pierre 1990 The Logic of Practice [orig. Les sens pratique ]. Translated by Richard Nice. Stanford: Stanford University Press. , 91–92). At the same time, however, it proceeds from the assumption that by means of textual analyses, these limits can be surpassed. In other words, if we research translation practice – what translators do, and why and how they do it – and through the analysis of the semi-structured interviews we look into what they say they do (i.e., how they ‘narratively do’ their translation decisions), we have to ask ourselves if we can see anything relevant in terms of how they really do act. Had we at our disposal only the semi-structured interviews, we would have to answer that we cannot claim with any degree of certainty anything about the translation practice. As soon as agents start thinking about their practice, they lose the chance to express any truth about it, especially about their practical relationship to it, and take a quasi-theoretical position. An academic enquiry prompts them to take a stand on their practice. But this stand is no longer a stand of acting (neither is it a stand of science – it is a quasi-theory). Nevertheless, an approach connecting translatological and sociological analyses makes it possible to obtain both sociological data (doxa obtained from semi-structured interviews) and translatological data on the results of practice (regularities of translation strategies uncovered by textual analyses), and the fusion of sociological and translatological approaches puts us in a better position to deal with the translation practice – we know doxa and the results of practice. Based on this, is it possible to draw conclusions about the practice itself? Probably, yes. We assume that in this case, practice itself can speak through its observable and analysable results. From our perspective, it is reasonable to start with empirical evidence (observable results of translation practice), connect it with the results of sociological analysis and relate practice (acting) to the objective conditions of its genesis (i.e., to the conditions defining both the functions that practice performs and the means it uses to fulfil them). We are convinced that if we proceed this way, we can establish what influenced/determined the decision-making of translators and answer the questions: How do translators translate? Why do they translate the way they do? What do they really do when translating?

This brings us to another question: How do we handle the fusion of TRA–SOC data, methodologically? The following section introduces a methodology designed for this purpose and tested in our research.

3.The methodology

As already explained, the object of our research is the regularities of translation strategies investigated from the perspective of translators on three levels: doxa, practice, and results of practice. The presented methodology reflects these three levels by fusing sociological and translatological approaches to data collection, analysis, and interpretation. The individual methodological steps are described below and subsequently, the methodology is exemplified in one case of data collection, analysis, and interpretation (i.e., one translation and its translator). It is beyond the scope of the present article to report on all the results of the analysis of the given case, and hence only some examples will be given.

We use two key methods of data collection: (1) qualitative semi-structured interviews with the translator,33.We conceive of the qualitative interview as a dialogue between the researcher and a communication partner. The traditional procedure of narrative interviews that precludes additional questions concerning the reasons for the communication partner’s behaviour (see Bertaux and Kohli 1984Bertaux, Daniel, and Martin Kohli 1984 “The Life Story Approach: A Continental View.” Annual Review of Sociology 10: 215–237. ) was adjusted to the needs of our research. Here we were inspired especially by Kaufmann (2010)Kaufmann, Jean-Claude 2010 Chápající rozhovor [Understanding interview]. Prague: Sociologické nakladatelství.. and (2) comparative analyses of the source text (ST) and target text (TT). The qualitative semi-structured interview was selected mainly because it provides enough space for the narrative of the communication partner, and, at the same time, helps to maintain a clear orientation towards the research problem. The framework of the semi-structured interview is an interview guide, namely the list of interview topics. The guide does not allow the interview to diverge too much, but is used as an instrument to keep the conversation on the topic and not as an external pressure or determinant. Skipping topics and questions is not problematic, and we regard it as more advantageous to let the translator talk freely in the moment without prompting them to go back to the interview guide. It is necessary to be flexible and attuned to the given situation. If the translator spontaneously broaches questions that we intended to ask later, we let them speak, seeking to sustain spontaneity. It is important for the translator to have an opportunity to mention topics that we do not expect and that are not included in the interview guide. Likewise, it is also important to motivate the translator to talk about what is relevant for our research problem.

During data collection, two interviews with each translator are conducted. The interview guides combine the sociological and the translatological approaches (see Figure 1). The guide for the first interview covers two areas. The first draws on Bourdieu’s theories of social fields and of practice and, based on these, the following research question was formulated: What influences the decision-making of translators? The question thus focuses on both the social contextualisation of the translator (i.e., the world that has formed them – their family background, generational anchorage, motivation to become a translator, models, placement in the translation world, and selection of texts for translation) and the translator’s capitals (economic, cultural, social, and symbolic; for further details, see Bourdieu [1984]Bourdieu, Pierre 1984 Distinction: A Social Critique of the Judgement of Taste [orig. La Distinction: Critique sociale du judgement ]. Translated by Richard Nice. London: Routledge and Kegan Paul.). We investigate the translator’s habitus also through personal narratives on key events and persons in their life (McAdams 1993McAdams, Dan P. 1993 The Stories We Live by: Personal Myths and the Making of the Self. New York: Guilford Publications.) and through an analysis of their value preferences and frames (the significance of their family, leisure time, work, and religion; for further details, see Prudký [2009]Prudký, Libor 2009 Inventura hodnot: Výsledky sociologických výzkumů hodnot ve společnosti České republiky [A stocktaking of values: The results of sociological research on values in the Czech Republic’s society]. Prague: Academia.). In addition, we also pay attention to the translator’s interpretation of the pressures of the field and of the book market: their cooperation with the publishing house, contract conditions, gaining of commissions, the influence of literary agents, and so on (for further details, see Bourdieu [1996]Bourdieu, Pierre 1996 The Rules of Art: Genesis and Structure of the Literary Field [orig. Les Règles de l’art ]. Translated by Susan Emanuel. Stanford: Stanford University Press. ). We rely here on the analysis of the Czech field of literary translation and the Czech book market that we conducted.44.The results are not included in this study as they are too extensive and the specific conditions of the Czech book market are not directly relevant to the presented methodology.

The second area is related to the following research question: How do translators reflect on their translating, their practice? The topics related to this question are anchored in (a) Toury’s (1995)Toury, Gideon 1995 Descriptive Translation Studies – and Beyond. Amsterdam: John Benjamins. account of translation norms,55.In the present context, the initial norm and the operational norms are the most pertinent. Since Toury (1995Toury, Gideon 1995 Descriptive Translation Studies – and Beyond. Amsterdam: John Benjamins. , 60–61) highlights the interrelatedness of these norms, we refer here, generally, to ‘norms’. and (b) the results of the given ST/TT analysis. We derived the following questions from Toury’s account: What changes to the ST does the translator deem permissible? Specifically, we investigate the translator’s opinion on shortening or expanding the text, simplifications and explanations, shifts in paragraph or sentence boundaries, translator style, and opportunities to influence the selection of texts for translation.

The textual analysis preceding the first interview is performed on the basis of quantitative indicators (see Tables 1, 2, and 3), which were set up to reflect Toury’s account of translation norms (Toury 1995Toury, Gideon 1995 Descriptive Translation Studies – and Beyond. Amsterdam: John Benjamins. ; Lambert and van Gorp 2014Lambert, José, and Hendrik van Gorp 2014 “On Describing Translations.” In The Manipulation of Literature: Studies in Literary Translation, edited by Theo Hermans, 42–53. Abingdon: Routledge. ). However, a fundamental difference is that, unlike Toury, in whose approach “the choice of ST–TT coupled pairs (segments that are analysed) is still far from systematic” (Munday 2016Munday, Jeremy 2016 Introducing Translation Studies: Theories and Applications. 4th ed. New York: Routledge. , 194), we determined the indicators in advance (with a certain degree of flexibility) and arranged them systematically. They were selected in order to allow for the analysis of a larger number of texts, for the comparability of all texts, and for flexibility (allowing not only for comparability, but also for the inclusion of specific features of particular texts and translators). They also allow us to factor in both structural differences between English and Czech and the individual decision-making of translators (reflection and self-reflection on translating and on the given translation). Data obtained by means of quantitative indicators do not depend on the researcher’s interpretation, are easily quantifiable, and can mostly be collected via electronic tools.

Apart from quantitative indicators, we also established qualitative ones (see Tables 1, 2, and 3). The results of the textual analyses based on qualitative indicators are used to determine the topics of the second interview. Data collected on the basis of qualitative indicators depend on the interpretation of the researcher and collection requires a higher degree of researcher input.

The indicators in Tables 1 and 2 are amenable to standardisation and thus comparability of results obtained for all texts is possible. The qualitative indicators in Table 3 provide analytical flexibility and ensure an orientation of the textual analyses towards the interviews. The indicators are used on three levels:

1. Whole-text analysis (Table 1): A limited number of indicators with high information value.

2. 5% sample analysis (Table 2): A higher number of indicators permitting a more detailed analysis.

3. 1% sample analysis (Table 3): Indicators allowing for a very detailed analysis that would not be feasible for a larger number of longer samples or whole texts.

Table 1.Whole-text analysis indicators

Comparative analysis: Whole text
Type of translation (direct/indirect)			D/I
Quantitative indicators			Value
	ST length (characters)
	TT length (characters)
	ST/TT length ratio (characters)
	ST length (paragraphs)
	TT length (paragraphs)
	ST/TT length ratio (paragraphs)
	Material added (longer than one sentence)		Y/N
	Material removed (longer than one sentence)		Y/N
	Full translation		Y/N
Qualitative indicators
	Shifts in	main title*	Y/N
		subtitle/s	Y/N
		chapter titles	Y/N
Overall assessment
Adequate/acceptable**

*In some cultures, the agency of translators to translate the main title might be limited. If this is the case, the indicator is to be left out. In the Czech cultural environment, the agency of literary translators in this area is undisputed (this was confirmed by the interviewed translators).

**The judgement is based on the results of the analysis and on the expertise of the researcher. Toury (1995Toury, Gideon 1995 Descriptive Translation Studies – and Beyond. Amsterdam: John Benjamins. , 57) points out that the notion of the initial norm serves primarily as an explanatory tool and that the distinction is not an absolute one. That is why no margins can be predetermined.

The qualitative indicators in the most detailed 1% analysis in Table 3 require the greatest input from the researcher. They are specific to the given language pair and possibly to a specific ST, and the shift tendencies are to be described in terms of the initial norm applied as an explanatory tool. The suggested indicators are examples of shifts typically occurring in literary translation from English to Czech. Their common denominator justifying their inclusion as indicators is the fact that they leave scope for the translator’s decisions and thus provide an opportunity for the researcher to observe regularities in the translator’s linguistic behaviour in terms of either following the norms or practices of the source language and culture, or those of the target language and culture. For example, expressiveness (emotional appeal, evaluativeness) of Czech lexical units tends to be higher when compared with the corresponding English lexical units. The translator thus decides whether to replace an English lexical unit by a Czech lexical unit with a similar, higher or even considerably higher degree of emotional appeal or evaluativeness. The principle is the same for the other indicators as well, even if they reflect other linguistic areas.

The first interview is transcribed and analysed by the sociologist (Helena Kubátová) and the obtained results (data on the professional and life experience, opinions, and self-reflection on practice of the translator) provide input for the TS scholar (Jitka Zehnalová).

As an example of this procedure, we present data on one literary translator and one translation from English to Czech: Jan Jirák (JJ)66.We use ‘literary translator’ in the narrow sense as introduced by CEATL (Conseil Européen des Associations de Traducteurs Littéraires/European Council of Literary Translators’ Associations), defining literary translators as “translators of any work [of literature in the narrow sense of the word (fiction, poetry, plays, etc.)] published in book form and protected by copyright” (Fock, de Haan, and Lhotová 2008Fock, Holger, Martin de Haan, and Alena Lhotová 2008 Comparative Income of Literary Translators in Europe. Brussels: CEATL (Conseil Européen des Associations de Traducteurs Littéraires)., 5). According to the Union Catalogue of the Czech Republic (CASLIN), one of the databases of the National Library of the Czech Republic, Jan Jirák translated thirty-two works of literature published in fifty editions and re-editions. and Advokát chudých ‘The lawyer of the poor’ (Grisham 2011Grisham, John 2011 Advokát chudých [orig. The Street Lawyer ]. Translated by Jan Jirák. Prague: Euromedia Group/Knižní klub.), which is a translation of John Grisham’s The Street Lawyer (Grisham 1998Grisham, John 1998 The Street Lawyer. New York: Bantam Doubleday Dell.). Jirák holds an MA in English and Czech. Translating is not his profession and he considers it a hobby. On average, he translates one book per year. At the time of writing, he enjoys such a position in the Czech field of literary translation that publishing houses offer him translation tasks and he can make suggestions as to what books he would like to translate. Right from the start of his career as a literary translator, he preferred popular fiction,77.We distinguish between ‘popular fiction’ and ‘fiction’. For further details, see Zehnalová and Kubátová (2019)Zehnalová, Jitka, and Helena Kubátová 2019 “From a Target Population to Representative Samples of Translations and Translators.” The Translator 25 (2): 87–100. . particularly detective stories. He claims that he does not dare to translate fiction as it would require a more consistent approach, different from his present one where he translates once a year during the summer holidays.

In the interview, Jirák stated that he gained experience with translating dialogue in the area of audiovisual translation. He regards his experience in the translation of dialogue to be an advantage that helped him to learn how to translate descriptive or narrative passages. He also comments on the unit of translation and splitting of sentences:88.Interview samples are published with the translator’s permission. The interview was conducted in Czech and subsequently translated to English by the authors.

JJ:Dialogues are like bubbles in a fizzy drink, they simply bring the whole thing to life … When I worked on my first book translation, having ten years of practice with translating movie dialogues behind me, it was a love story … I was quite incompetent at writing descriptions … The dialogues, they were ok, I did them quickly, in a jiffy, and then a love scene came [laughter]. It was horrible … I kept re-writing it, I don’t know, maybe five, six times, before I got it, the way a narrative gets done … Gosh it took me quite some time to get it right … to advance from dialogues to descriptions and to narration.

JJ:I think that the unit of translation is the paragraph, not the sentence. And to skip a sentence is probably an extreme solution, but to cut a long sentence into shorter pieces, if I feel that in Czech it should keep some kind of dynamics and the long sentence hinders it, then well I do not hesitate to split it into two.

As a result, the TS scholar paid more attention to dialogues and discovered that apart from shifts that can be expected due to the verbal character of Czech (adding of verbs in direct speech and using semantically richer verbs in sentences that introduce them), the translator also strengthened lexical cohesion (replaced personal pronouns with nouns and adjectives) in sentences that introduce direct speech and used this strategy in descriptive passages as well. As he “did not hesitate to split sentences,” the strategy of adding words did not, however, result in the lengthening of sentences. By combining these strategies, the translator achieved naturalness of expression in the target language, comprehensibility of the TT (due to the shifts in cohesion – stronger lexical cohesion and adding of connectors – readers can orient themselves better in the text), and dynamism (the translator avoided long sentences). This combination of strategies was applied in both direct speech and narrative passages. Example (1) illustrates the combination of strategies, while Example (2) shows how cohesion is strengthened in the translation.99.In the examples, additions are indicated in grey highlighting, while the splitting of sentences is underlined. Back-translations (BT) from Czech to English are provided; they were constructed so as to make clear the semantic, grammatical, and stylistic differences between the ST and the TT. Where a difference occurred (for example, a ST pronoun was replaced by a TT noun, a ST sentence was split into two TT sentences, or a stylistically neutral ST unit was replaced by a stylistically marked TT unit) the English BT items are as close to the Czech TT as possible, rendering the TT noun by the semantically closest BT noun, the two TT sentences by two BT sentences, and the marked TT unit by a marked BT unit.

(1)

ST:“Put that gun down,” Rafter ordered from the doorway, and a split second later another shot rang through the reception area, a shot that went into the ceiling somewhere well above Rafter’s head and reduced him to a mere mortal. Turning the gun back to me, he nodded, and I complied, … (Grisham 1998Grisham, John 1998 The Street Lawyer. New York: Bantam Doubleday Dell., 3)

TT: „Dejte tu pušku pryč,“ poručil tulákovi Rafter stojící ve dveřích a o zlomek vteřiny později zaduněl vstupní halou další výstřel. Tentokrát mířil do stropu, hodně vysoko nad Rafterovu hlavu. Stačilo to, aby se Rafter změnil v nehybnou sochu. Chlápek v holinách otočil pušku zpátky ke mně, pokývl hlavou, já mu vyhověl a … (Grisham 2011Grisham, John 2011 Advokát chudých [orig. The Street Lawyer ]. Translated by Jan Jirák. Prague: Euromedia Group/Knižní klub., 4)

BT:‘“Put that gun down,” Rafter ordered the street bum from the doorway, and a split second later another shot rang through the reception area. This time it went into the ceiling somewhere well above Rafter’s head. It was enough to reduce him to an immobile statue. The guy in the rubber boots turned the gun back to me, nodded, I complied and …’

(2)

ST:Since I was the first one to come to her aid, he politely aimed it at me, and I too became rigid. (Grisham 1998Grisham, John 1998 The Street Lawyer. New York: Bantam Doubleday Dell., 2)

TT: A jelikož jsem byl první, kdo madam Devierové přispěchal na pomoc, tulák zdvořile namířil na mě a já ztuhl také. (Grisham 2011Grisham, John 2011 Advokát chudých [orig. The Street Lawyer ]. Translated by Jan Jirák. Prague: Euromedia Group/Knižní klub., 4)

BT: ‘And since I was the first one who hurriedly came to help Madam Devier, the street bum politely aimed it at me, and I too became rigid.’

These and similar types of data thus condition the selection of qualitative indicators to be used for the second textual analysis of the given translation, the results of which are used as topics for the second interview guide.

Therefore, the second interview guide follows up not only on what the translator said in the first interview (thus seeking further explanations and elaborations of the first interview topics), but also on the results of the comparative ST/TT analysis of a specific translation based on qualitative indicators. These results enable the sociologist to approach the second interview equipped with detailed knowledge of the translation strategies employed by the given translator in the given text. It thus presents the translator with concrete examples of the translation strategies which can help to ‘open the space’ of the interview, to increase the interest and trust of the translator and, consequently, to elicit and collect data that might otherwise not arise. Furthermore, the interview can be very precisely targeted at the practice of the particular translator, which would not be possible without this kind of information.

The aim of the second interview is to prompt the translator to reflect upon their own specific approach to translating (doxa) and, by pointing out the results of the textual analysis of a specific translation (results of practice), to compel the translator to consider their practice. The textual analysis revealed, for example, that Jirák tends to add signals of irony, as in Example (3). Here, we use the transfer of irony only in order to exemplify the suggested methodology (a discussion of this notoriously complex topic is not our aim as it would be far beyond the scope of this article). What is of interest to us is the fact that “the way speakers […] employ this pervasive tool to express their attitude reflects the linguistic and cultural distance between […] languages” (Chakhachiro 2018Chakhachiro, Raymond 2018 Translating Irony between English and Arabic. Newcastle upon Tyne: Cambridge Scholars., 1), which accounts for the occurrence of shifts, frequently explicitation shifts (see Hirsch 2011Hirsch, Galia 2011 “Explicitations and Other Types of Shifts in the Translation of Irony and Humor.” Target 23 (2): 178–205. ), and thus for its relevance in the exploration of translational norms.

(3)

ST:And then the voices disappeared into the background, growing fainter and fainter as my colleagues hit the back door. I could almost see them jumping out the windows. (Grisham 1998Grisham, John 1998 The Street Lawyer. New York: Bantam Doubleday Dell., 3)

TT: Načež se hlasy začaly vzdalovat a slábnout, a když za sebou moji drazí kolegové zabouchli zadní dveře, ztichly hlasy docela. Téměř bych se vsadil, že pánové vyskákali z oken. (Grisham 2011Grisham, John 2011 Advokát chudých [orig. The Street Lawyer ]. Translated by Jan Jirák. Prague: Euromedia Group/Knižní klub., 4)

BT: Whereupon the voices began to disappear and grow fainter and when my dear colleagues shut the back door closed, the voices grew completely silent. I felt almost like betting that the gentlemen jumped out the windows.

In Example (3), the distinction between formal and rhetorical ironic devices is pertinent: “The first is textual and identifiable in the text; the second is not subject to immediate identification by textual means” (Chakhachiro 2018Chakhachiro, Raymond 2018 Translating Irony between English and Arabic. Newcastle upon Tyne: Cambridge Scholars., 104). Hirsch (2011Hirsch, Galia 2011 “Explicitations and Other Types of Shifts in the Translation of Irony and Humor.” Target 23 (2): 178–205. , 183) points out that as “most accounts of irony claim, it necessarily involves criticism.” It is indeed the critical appraisal of his colleagues implied by the narrator in the extract that is ironic. However, this is inferable mainly from the situational context constructed within the fictional world and not immediately identifiable by linguistic means in the English ST; whereas in the Czech TT, the ironic effect is achieved also by linguistic means (more formal and more expressive lexical units, adding of lexical units, positive evaluation meant negatively, more complex sentence structure).

The sociologist (S) explained this result of the textual analysis during the second interview and asked the translator how he arrived at this solution. Jirák answered:

JJ:I don’t know … I have no idea, really, no idea at all. Definitely it’s not about any … I don’t tell myself something like “hey, this is irony, I need to put an irony signal in there.” It’s about one part of my brain reading the irony in English and the second one writing it in Czech, something like that, no kidding I don’t know. Really, no clue. Something in there … in my head, there is a kind of software and it reads it with English eyes and writes it for Czech eyes but what’s going on in there I don’t know.

The translator kept trying for some time to explain to the sociologist how he ‘does it’, only to finally give up (quite in accordance with Bourdieu’s assumption that people in their practice mostly do not follow any conscious and explicitly stated opinions but rather their unconscious feel for the game [Bourdieu 1990Bourdieu, Pierre 1990 The Logic of Practice [orig. Les sens pratique ]. Translated by Richard Nice. Stanford: Stanford University Press. ]). He moreover said that the more he thinks about his translation practice, the more afraid he is that his self-reflection will destroy his ability to translate (which he was not able to explain but which manifests in his translation practice):

JJ:… I am getting a bit scared … [laughter]

S:Of me? [both laugh]

JJ:Do you know the … it’s I guess in Tales from two pockets by Karel Čapek, the story about the gentleman who was good at tossing things? He took a stone and threw it far away across a brook; somebody saw it and thought, why, he is a natural for shot-putting, so he got him a coach and the coach told the guy: “You can’t do it this way! You need to put this arm forward, this leg backwards and the centre of gravity over there …” and it ended up by the man never ever tossing anything far away …

The main advantage of this way of collecting data is the possibility of comparing the translator’s doxa with the results of the textual analysis. As shown in Figure 1, the qualitative indicators can be further adjusted on the basis of the results of the second interview analysis and used for a third textual analysis.

Figure 1 shows just how important the close cooperation between the sociologist and the TS scholar is in developing the interview guide. Without the textual indicators, we would obtain data that socially contextualises the translator, but not the agent’s discourse on their translation practice (doxa). The TS scholar has to provide the sociologist with the results of analyses conducted on the basis of the indicators to enable the sociologist to ask the translator questions about them. If we want to obtain translatologically and sociologically relevant data, we cannot proceed separately; we have to advance jointly. As Figure 2 shows, if we were to collect data separately we would not obtain data suitable for joint interpretation.

4.Discussion

The nature of our research demands both standardised and flexible data collection. Standardisation guarantees data comparability; flexibility allows for obtaining data relevant for particular cases. Sociologically, this means that the first interview guide is standardised to a degree – it is the same for all translators (although it can be slightly adjusted during data collection due to newly gathered knowledge and empirical data) – while the second interview guide is targeted at a particular translator and translation. Translatologically, this means that we work with the indicators and samples set out in Table 1, 2, and 3, while being aware that we can keep adapting them. Flexibility also entails the necessity to adjust the methodology depending on the needs of a specific study, including the possibility of changing at the very beginning the indicators and/or sizes and numbers of samples. The selection of indicators depends on the source and target languages (in our case, given the verbal character of Czech and the nominal character of English, many indicators are related to the transfer of verbs), the genres under investigation, the research assumptions, and so on. The main factor in the choice of the number and size of samples is the feasibility of analyses. For example, when working with the 5% sample, we can react to the size and nature of the analysed text and to the data obtained from interviews. For shorter, stylistically hybrid texts or texts that are exceptional in some way, the percentage might be increased. Depending on the conditions of the specific study, the sample size can be initially set up at 20%. The proportion of standardisation and flexibility should certainly be well balanced; however, this has to be determined by the researcher for individual cases.

The presented methodology of interdisciplinary data collection, analysis, and interpretation has evolved in stages. First, we pilot-tested it through an interview in the field by the sociologist on the basis of topics inferred from Bourdieu’s theories of social fields and practice, and Toury’s theory of the norms of translation. Having analysed the interview and quantitatively analysed the given translation, we realised that had the sociologist led the first interview equipped with the topics derived from the textual analysis, we would have gained much more data. Consequently, we then adapted the methodology by including in the first interview guide topics that emerged from the quantitative textual analysis. Following the analysis of this interview, the qualitative indicators for the second textual analysis were set up. We reached the conclusion that the first interview raised many topics worth discussing with the translator, and so we decided to conduct a second interview. We included in the guide for the second interview not only these topics, but also topics obtained from the qualitative analysis.

In sum, we moved from the methodology to the field and then, informed by the field experience, back to the methodology and having adapted it, to the field again. We have gradually rendered the methodology more accurate, yet we do not consider it to be definitive. We expect that researchers will tailor it to their specific research projects. If a given research question does not require a detailed textual or social analysis, it does not have to be done in the way suggested and the second interview does not have to be conducted. If, on the other hand, an even more detailed textual or social analysis is needed, the indicators can be adjusted in the sense of making them more precise and/or numerous, a third textual analysis can be done, and possibly a third interview carried out.

5.Conclusion

The goal of this article was to present a methodology of joint TRA–SOC data collection, analysis, and interpretation, tailored to research on translation strategies and norms. We have been using and testing the methodology in ongoing research on the norms of literary translation from English to Czech. Our research assumption is that in order to identify norms of translation, it does not suffice to analyse translations only (i.e., the results of the translators’ practice); it is imperative to uncover the practice itself. For this reason we employ interdisciplinary cooperation, expecting it to facilitate an organic fusion of translatological and sociological data which we, given the objective of our research, consider necessary.

Interdisciplinary cooperation is not as easy as it sounds. Pursuing our research, we have realised that collaboration is more than providing each other with sub-tasks. We had to become acquainted with the fundamentals of each other’s discipline, and every task we worked on together had to be thoroughly discussed. Neither collaborator can enforce a solution, methodological procedure, or concept without explaining it to the other. Neither collaborator can claim ‘this is the way to get things done in my discipline and that’s it’. Collaborators have to explain why things are done in a certain way. It is a process that is demanding, sometimes to the point of being painful. The advantages are considerable, however. We found ourselves prompted to discuss issues, unable to rely on the tacit knowledge of the given field of study. In interdisciplinay collaborations, everything has to be spelled out, explained, and critically assessed before it can (possibly) be included and applied. We are confident that this kind of interdisciplinary cooperation can generate results which are adequate both translationally and sociologically and do not delineate norms of translation without considering translation practice and its sociocultural contexts.