Directionality in Translation


Czech Science Foundation (GA ČR), project nr. 16-03037S


Institute of Translation Studies, Charles University, Prague


Tomáš Duběda,

David Mraček,

Vanda Obdržálková,



Languages covered



About the project

Our project seeks to explore the issue of non-native translation (also known as inverse or L2 translation), which has so far received limited attention from researchers in Translation Studies and has even been rejected by some as unnatural and unprofessional, despite being an increasingly common practice in numerous countries, including the Czech Republic.

A large sample of translations between Czech and English/French will allow for extensive empirical research into the quality of inverse vs. direct translation and how quality is dependent on factors such as text genre, time allotted for the task, and years of experience.

Further data will be gleaned from a set of pre- and post-testing questionnaires obtained from the participants, shedding more light on the preferences, perceptions and habits of Czech translators regarding inverse translation, in addition to identifying the specific challenges involved in inverse translation and strategies employed to overcome them. Complementary methods, such as interviewing native speaker revisers and analysing the ways local translation agencies communicate with clients and potential translators about inverse translation, are expected to contribute to a better understanding of the specificities of the Czech market.

The sample

160 translations (each 1 standard page long) obtained from 40 Czech translators (20 professionals and 20 M.A. students; 20 translating from/into English and 20 translating from/into French).

Two text genres were used: promotional (web presentation of a kitchen producer), and legal (terms and conditions of a web server).

Each tested person produced translations in both directions, i.e. two direct and two inverse translations.

The participants filled in one pre-testing questionnaire and two post-testing questionnaires.

The tested subjects worked in their homes and were allowed to use any resources except for communication with other people. 90 minutes per page were allowed. All translations are anonymous and will be used exclusively for non-commercial purposes. The translators were remunerated for their work.

The translation database is a highly relevant empirical resource, as it is based on carefully calibrated texts, has been obtained under controlled conditions, and contains both translated texts and questionnaire data.

Quality assessment

A detailed guide was elaborated to guarantee cross-assessor compatibility. Two parallel approaches were used: global (overall mark), and local (error type and frequency).

Global assessment scale: A – excellent, B – acceptable, C – borderline, F – unacceptable.

Error categories: Ⓥ – meaning (gross error), V – meaning, T – terminology, S – style, G – grammar, F – form.

All translations into Czech were assessed by a panel of three native revisers. Translations into English and French were assessed by two independent native assessors for each language.

Selected results of the survey

  • The translators attain better quality when translating into their mother tongue, the average grade obtained being B for translations into CS and C+ for translations into EN or FR.

  • The two directions also differ in the total number of errors. Surprisingly perhaps, this difference can be accounted for mainly by a higher number of grammatical errors rather than errors in the category of style.

  • Genre has more impact on the quality of translations into a foreign language than on that of translations into Czech, confirming the assumption that factual texts are easier to handle for non-native translators than expressive texts.

  • Professionals perform better than students when translating into Czech (B+ vs. B–), while the quality of the translations in the opposite direction is comparable in both groups (C+). This paradoxical finding could be explained by a fossilisation of translation competences and an improving level of foreign language skills achieved by students over the past ten to fifteen years.

  • Translators who attended a course in legal translation do not achieve better overall quality in translating legal texts than those who did not; however, their translations contain fewer errors of meaning.

  • The overall quality of translations into English is higher than that of translations into French.

  • The practice of non-native translation is linked to cultural and economic factors and does not seem to be on the decline in the Czech Republic. This direction of translation is associated with a certain “social permissiveness” shared by all participants in the translation process, from clients down to target-text recipients.


Conferences attended

  • Profilingua 2016, Plzeň (Czech Republic)

Vanda Obdržálková: Strategie řešení problémů v překladu neliterárního textu do mateřského a nemateřského jazyka

  • XColloque sur la traduction, la terminologie et l’interprétation, Varadero (Cuba), 2016

Tomáš Duběda : La traduction vers la langue B : entre qualité du produit et attitudes des traducteurs

  • Forum HERACLES, Prague (Czech Republic), 2017

Tomáš Duběda: La traduction vers une langue étrangère et son rôle dans la formation des futurs traducteurs

  • Quo vadis, legal translation?, Bratislava (Slovakia), 2017

Tomáš Duběda: Quality assessment in legal translation: structuring a multidimensional space

  • FIT Congress 2017, Brisbane (Australia)

Tomáš Duběda: Translation Quality and “Cross-Cultural Allowance” in Translating into a Foreign Language

  • Liletrad, Sevilla (Spain), 2017

Vanda Obdržálková: Translation from and into English as Non-Mother Tongue: Attitudes, Problem-Solving, Quality

  • Teaching Translation and Interpreting Conference, Łódź (Poland), 2017

David Mraček: Inverse translation: comparing student and professional perspectives

  • Vědecký výzkum a výuka jazyků, Hradec Králové (Czech Republic), 2017

David Mraček: Gramatické a stylistické chyby v překladu do angličtiny jako cizího jazyka

  • TIFO Olomouc (Czech Republic), 2017

David Mraček: Directionality in Translation: What Makes Translating into L2 Difficult?

  • Translation Research – Translator Training, Budapest (Hungary), 2018

David Mraček: The Role of Native Speakers in L2 Translation

Tomáš Duběda: Native Revision in L2 Translation: A Panacea or a Risk?

  • Transius 2018, Geneva (Switzerland)

Tomáš Duběda: Assessing Terminological and Textual Competence in Legal Translation: A Case Study

  • Postgraduate Conference in Translation, Leeds (United Kingdom), 2019

Tomáš Duběda: Non-Native Translation in Central Europe: Between Past and Present

  • Translation and Interpreting Forum Olomouc (Czech Republic), 2019

Tomáš Duběda: Assessing the Quality of Non-native Legal Translation: Evidence from Revision and Metarevision

  • Translation and Interpreting Forum Olomouc (Czech Republic), 2019

David Mraček: L2 Translation: the Perspective of Native Speakers

  • Traduction et traductologie : la fin de l’Histoire ?, Geneva (Switzerland, online), 2020

Tomáš Duběda: La traduction non native en Europe centrale : entre passé et avenir

Conferences organized

  • L2 Translation: Getting Out of the Grey Zone, Prague (Czech Republic), 2019


Úvod > Věda a výzkum > Projekty > Directionality in Translation