RUSSIAN TRANSLATED DISCOURSE RESEARCH: PATTERNS OF LEXIS USAGE AS LINGUISTIC INDICATOR OF TRANSLATION UNIVERSALS REPRESENTATION

 

YEKATERINA KRASNOPEYEVA^*

 

 

Summary

 

This paper reports on the results of a research project devoted to investigation into the nature of translations into Russian, within the framework of corpus-based Translation Studies. The paper discusses a number of theoretical aspects behind translation universals research, developing the concept of Translated Discourse. Case studies conducted as part of the research project are described. The case of explicitation is discussed in detail. This work is a part of the Russian Foundation for Humanities supported project (No. 15-54-00020).

 

Key words: Russian language, translationese, translated discourse, universals of translation, explicitation, corpus-based methodology.

 

Introduction

 

Studying universal inherent qualities of translations remains one of the key issues in today’s Translation Studies. The last two decades marked the emergence of a range of hypotheses discussing potentially uniform unique features of translations characterized as the universals of translation (e.g. simplification, explicitation, normalization, see Baker, 1993). In search of these features, a variety of languages and language pairs are being studied with the help of corpus-based methodology (Ibid.). Corpus-based Translation Studies and potential universals theory are evolving alongside each other. Translation universals are viewed as descriptive constructs, working hypotheses that require constant empirical verification. Detailed description and conceptualization of translated discourse characteristics in a variety of languages is essential for further development and refinement of theoretical concepts concerning the nature of translation.

Research project described in this paper studies Russian translated discourse and attempts to use lexis as linguistic indicator to test a number of hypotheses against the backdrop of Russian narrative prose corpora. The paper is organized as follows: in section Theoretical Framework the concept of Translated Discourse is discussed. The following sections describe qualitative and quantitative methodology of the project. The case study of lexeme который is presented in detail, exemplifying investigation into the tendency of explicitation.

 

Theoretical Framework: Translated Discourse as an Umbrella Term

 

The idea of unique nature of translations compared to texts originally written in a certain language has made a number of appearances through history within different theoretical frameworks. Third code (Frawley), third language (Duff, Toury), interlanguage (Toury), translationese (Gellerstam, Newmark), hybrid language (Trosborg), подсистема национального языка (Комиссаров, Умерова) – these terms^[1] in their entirety reflect a strictly linguistic view on the nature of translations. For a long time these ideas retained their hypothetical nature, since there existed no methodology able to empirically prove ontological difference between translations and non-translations.

Shift from linguistic outlook to ubiquitous acceptance of discursive ontology of translation process led to defining translated texts as the ones

“record[ing] genuine communicative events and as such are neither inferior nor superior to other communicative events in any language. They are however different […]” (Baker, 1993).

Search for this difference was proposed as a task for corpus-based translation studies (Ibid.). Today, its major branch of potential universals research has become the driving force in defining translated text as a category. In one of her recent works, Baker argues:

“[o]ne of the main strands of corpus-based research in Translation Studies involves examining similarities and differences between translated and non-translated text, in an attempt to demonstrate that translations form a distinctive textual system within a target culture. This type of research clearly makes a number of assumptions, the most important of which are that translated and non-translated texts exist as categories [...]” (Baker, 2007).

The question still pertinent, however, is whether the object of study defined as “the language of translation” truly coincides with discursive ontology. Translationese is still often used to explain its essence. To emphasize that translations are different from non-translated texts, and to point out a systematic inherent discursive nature of this difference, Garbovskiy introduces a concept of translated discourse (also translational discourse, Rus. переводной дискурс), which we use in this study.

Translated discourse is defined as a body of texts, which is generated in the process of translation, exists in a discursive space of a receiving culture and possesses a number of distinctive inherent qualities, differentiating it from authentic, original texts (Гарбовский, 2012). When one uses the term translated discourse, they refer to the texts generated in the (prototypical^[2]) communicative act of translation in connection with the network, the system of life circumstances that played their part in the formation of translations as they are.

Translated discourse can be viewed as a legitimate object of study for a wide range of versatile research – from quantitative to anthropological. What could be the role of alleged translation universals in this conceptual space? As part of this research project, we pursued the idea to trace interrelation of the concepts of translated discourse and translation universals. Supposedly, cross-fertilization of the said fields of study can lead to better understanding of the nature of translation process: corpus-based methodology carries potential in translated discourse research, which is discussed elsewhere (Краснопеева, forthcoming).^[3] In this study, corpus of translations is understood as a model of translated discourse.

 

Corpora Description and Quantitative Study Methodology

 

Methodology applied in the study combines a traditional corpus-based and a corpus-driven approaches. In the latter, the data under consideration plays the key part. When applying the approach, the researcher commits to the integrity of the data. Both presence and absence of recurrent patterns in the corpora are considered meaningful and cannot be ignored (Tognini-Bonelli, 2001).

Within Translation Studies, different types of corpora have been utilized to identify the peculiarities of translated discourse. For this experiment a specific set of corpora is used: the parallel subcorpus of Russian National Corpus (RNC) and a Comparable Corpus of Russian Literary Prose (a DIY corpus). The quantitative corpus-driven part of the study demanded a translational comparable corpora, and the qualitative part (which follows and develops the quantitative one) required examination of parallel concordance lines. The RNC parallel subcorpus contains translations of various genres into and from Russian. All the texts are morphologically annotated, translations are aligned with the originals. The following parameters were applied to single out a relevant subsection from the parallel subcorpus of RNC (11 texts, 1,770,365 tokens): language of the original text – English, language of translation – Russian, genre – novel (fiction), average length – over 50,000 tokens, time of creation – contemporary narrative prose. These parameters also apply to a relatively small DIY corpus created for the study specifically: a subcorpus of translations from various languages was added (see Table 1). The texts in the DIY corpus are lemmatized and annotated using the TreeTagger and a parametric file for the Russian language.^[4]

 

 

The first quantitative part of the study aims to track patterns that differentiate translated Russian-language texts from their non-translated variety. Both these findings and results of other comparative studies are used as a basis for further qualitative research.

We start by comparing frequency lists^[5] of NTTS, TTS(E) and TTS(V) sorted by calculated log-likelihood values (LL) (Rayson, Garside, 2000) and percentage of corpus coverage. A Three-Phase Comparative Analysis (Jantunen, 2004) allows both to avoid and identify language-pair influence on the process of translated text formation: NTTS list is compared to TTS(E) list, then to TTS(V) list. Then frequency lists of TTS(E) and TTS(V) are compared to each other. This step helps identify lexemes that can help track representation of certain potential universals of translation: corpus-driven approach is in action.

Microsoft Excel macro Visual Basic Find_Matches^[6] was used to perform comparison of 1000 most frequent lexemes in each of the three frequency lists to find lexemes present in each pair. 92-100 percent of the lexemes in all the three lists are present in all the 32 texts of the DIY corpus.

NTTS vs TTS(E). Sorting the lists by percentage of corpus coverage and LL values showed 432 lemmata with statistically significant difference in frequency (p<0,05). Within the framework of universals research, difference in content words usage is not as informative as the difference in structural words distribution. For example, noun земля is significantly overused in  NTTS, however this tendency is hard to interpret (see Table 2). Sorting the lists by percentage of corpus coverage helps draw structural words to the list head and track patterns of variation.

 

Table 2. Example of the First 15 Lines in NTTS vs TTS(E) Comparison List

 

 

Relative and interrogative pronouns overuse (который, что), as well as extensive usage of potential discourse markers (кстати, потому) in TTS(E) may be interpreted as a sign of explicitation of clausal relations in translations. To see whether the tracked preferences are indicative of language pair or not, one has to compare frequency lists of non-translated discourse and generalized translated discourse.

NTTS vs TTS(V). Sorting the lists by percentage of corpus coverage and LL values showed 576 lemmata with statistically significant difference in frequency (p<0,05). Relative and interrogative pronouns overuse (который, когда, что, если, чем, чтобы), as well as extensive usage of potential discourse markers (затем, однако, правда, разве) in TTS(V) are present in translations.

TTS(E) vs TTS(V). Interestingly, comparison of these frequency lists is the least informative out of thee phases of the analysis. Identified lemmata with statistically significant difference in usage are mostly content words (which depend on thematic difference of texts in subcorpora), so the usage of structural words may be considered more similar in TTS(E) and TTS(V). Less differences may be observed between translational subcorpora.

Quantitative data collected through the Three-Phase Comparative Analysis serves as a basis for qualitative research featuring more detailed study of comparable corpora and parallel concordances.

 

Qualitative Case Studies Methodology and Results

 

Further investigation into five hypothesized translation universals of explicitation, simplification, normalisation, interference and unique items hypothesis in Russian translated discourse is realized in seven case studies. Despite the initial conscious choice to exclude the originals from descriptive corpus-based Translation Studies research, appropriateness of parallel corpora use has been later justified:

“comparable and parallel corpora in fact offer complementary perspectives on translation norms/universals, such that neither would suffice in isolation to shed full light on this complex research topic” (Bernardini, 2011).

Qualitative part of the research uses a complementary to the DIY corpus RNC subsection, which helps verify initial speculation based on the quantitative data. The results of the case studies (both quantitative and qualitative) are summarized in Table 3.

Statistical hypotheses tested in case studies are based on existing assumptions about features of translated discourse:

- simplification in the Russian translated discourse research follows Laviosa’s (1998) findings^[7] ;

- normalisation study is based on the original definition by Baker (1996) – “the tendency to conform to patterns and practices that are typical to the target language”;

-interference is understood as a universal on a higher level of abstraction (Mauranen, 2004)^[8] ;

-unique items hypothesis case is built on Tirkkonen-Condit’s (2004) assumption about specific items underrepresentation in translations^[9] ;

-explicitation is defined as “spelling out in target text of information which is only implicit in a source text” (Olohan, Baker, 2000). The case of explicitation is discussed further in detail.

 

Table 3. Summary of the Research Project Results

 

 

Symbols used in Table 3

 

□         –          quantitative analysis of a comparable corpus

●         –          qualitative analysis of a parallel corpus

—       –           a hypothesis was not tested

Y         –          a hypothesis was verified

N         –          a hypothesis was not verified

Y/N     –          inconsistent result

 

Explicitation in Russian Translated Discourse

 

Since quantitative analysis showed overuse of lexemes что and который in translational component of the DIY comparable corpus (see Table 4), one of them, namely который was chosen as a formal operator for further qualitative investigation. Both lexemes are not optional in a Russian sentence, however they can introduce clauses of periphrastic nature, thus may indicate explicitation. Interestingly, when compared to each other, the two parts of translational corpora (TTS(E) and TTS(V)) show overuse of что in translations from English, which may indicate evidence of source language influence, or interference.

 

Table 4. Comparison of что and который Usage in DIY Comparable Corpus

 

 

RNC web interface provides a parallel concordance tool making it possible to query a subsection with relevant parameters. 550 bi-texts with который instances in 11 translations were randomly chosen and reviewed. The following constructs were found corresponding to который:

-  Calque: attributive clause with that, which, where, what, when, who, whose, whom, and zero conjunction (40 percent of 550 bi-texts);

-  Explicitation caused by grammatical asymmetry: verb forms: Pariciple I, Pariciple II, Gerund, Infinitive; adjectives with suffix –able, e.g. imaginable, unreadable (37 percent of 550 bi-texts);

-  Optional explicitation: fragments containing neither grammatical form absent from the Russian language, nor  attributive clause (23 percent of 550 bi-texts).

Optional explicitation, the most curious phenomenon relevant for universals research, is present in only 23 percent of reviewed cases. The following example illustrates optional explicitation. Translator adds description to define a phrase a global force, which may limit the amount of possible interpretations of the original fragment.

 

 

At the moment, the result of qualitative study may be considered inconsistent, and further investigation into the nature of elaboration and explicitation is required.

While looking at который as a formal operator for explicitation research, it was hard to overlook another tendency in который usage – leveling out, also a proposed universal defined as “a tendency to gravitate around the centre of any continuum rather than move towards the fringes” (Baker, 1996). Interrogative pronoun который sometimes is characterized as colloquial (e.g. В который вагон садиться?, МАС), and as an indefinite pronoun it may be also vernacular (e.g. Надо бы голубей Васиных сосчитать, не пропали бы которые. А. П. Чехов. Бабы, МАС). In the majority of reviewed bi-texts (543 out of 550), который is a relative pronoun. Который as an interrogative pronoun (e.g. Который час? В который раз...) and as an indefinite pronoun (e.g. Которые говорят, время пришло...) were used only 7 times (out of 550) and который of other grammatical meanings were not present, which may be considered a clue for further leveling out research.

 

Conclusion

 

The described study shows that potential universals of explicitation, simplification, normalisation, interference and the unique items hypothesis may be present in the corpora of translations into Russian, which may potentially be extended to Russian translated discourse and translated discourse as an abstract phenomenon.

Combination of a corpus-based and a corpus-driven approaches provides means to track not only the sought tendencies, but uncover interrelations of other universal features, the coexistence of which forms characteristic profile of translated discourse. In this paper, an attempt is made to present a short overview of the research project pursuing to uncover inherent lexical patterns of Russian translated discourse. Let us list some of them.

Quantitative study results:

-Lexemes который and что are more often used in translations, which may indicate potential universal tendency of explicitation.

-Lexemes что and сказать differentiate subcorpus of translations from English from subcorpus of translations from various languages, which may indicate potentially universal tendency of interference.

-Translations into Russian contain less content words and more frequent words than texts originally written in Russian, which may indicate potentially universal tendency of simplification.

-Diminutive forms are used in translation less often than in the texts originally written in Russian, which could be indicative of verification of the unique items hypothesis.

Qualitative study results:

-Lexeme который usage is associated with both grammatical (compulsory) and optional explicitation.

-Speech verb сказать commonly corresponds to say and tell in the original, which may indicate a potentially universal tendency of interference. The said tendency is balanced out by an opposing trend of normalisation that is evident when tracking translations of to say: calques are used less often than other various verbs and phrases.

To our knowledge, this is the first comprehensive study of Russian narrative prose using instruments and methods of corpus-based Translation Studies^[10] , which strengthens the evidence of Translation Universals existence.

 

Bibliography

 

Baker, M. “Corpus Linguistics and Translation Studies: Implications and Applications”, Text and Technology: In honour of John Sinclair, Amsterdam, Philadelphia : J. Benjamins, 1993.

Baker, М. “Corpus-based translation studies: The challenges that lie ahead”, Terminology, LSP and Translation: Studies in language engineering in honour of Juan C. Sager, Edited by Harold Somers, 1996.

Baker, M. “Patterns of Idiomaticity in Translated vs Non-Translated Text”, The Study of Language and Translation ; ed. W. Vandeweghe, S. Vandepitte, M. Van de Velde, Amsterdam, Philadelphia : J. Benjamins, 2007.

Becher, V. “Explicitation and implicitation in translation : A corpus-based study of English-German and German-English translations of business texts”: Dissertation zur Erlangung der Würde des Doktors der Philosophie, Hamburg, 2011.

Bernardini, S. “Monolingual Comparable Corpora and Parallel Corpora in the Search for Features of Translated Language”, SYNAPS – A Journal of Professional Communication, No. 26, 2011.

Jantunen, H. J. “Untypical patterns in translations. Issues on corpus methodology and synonymity”, Translation universals : Do they exist?  Amsterdam : J. Benjamins, 2004.

Kenny, D. Lexis and Creativity in Translation: A Corpus-Based Study, Manchester : St. Jerome Publishing, 2001.

Laviosa, S. “Core Patterns of Lexical Use in a Comparable Corpus of English Narrative Prose”, Meta : journal des traducteurs / Meta: Translators’ Journal, Vol. 43, 1998.

Mauranen, A. “Corpora, universals and interference”,  Translation universals : Do they exist? Amsterdam : J. Benjamins, 2004.

Olohan, M., Baker, M. “Reporting that in Translated English. Evidence for Subconscious Processes of Explicitation?”, Across Languages and Cultures 1(2), 2000.           

Rayson, P., Garside, R. “Comparing Corpora using Frequency Profiling”, WCC’00 Proceedings of the workshop on Comparing corpora (Volume 9), 2000.

Tirkkonen-Condit, S. “Unique items – over- or under-represented in translated language?”, Translation universals : Do they exist? Amsterdam : J. Benjamins, 2004.

Tognini-Bonelli, E. Corpus Linguistics at Work (Language Arts & Disciplines), Amsterdam, New York : John Benjamins Publishing Company, 2001.

Алексейцева, Т. А. Экспликация как способ преодоления межъязыковой и межкультурной асимметрии в переводе, дис. ... канд. филол. наук : 10.02.20, Санкт-Петербург, 2009.

Гарбовский, Н. К. “Русский переводной дискурс: миф или реальность”, Материалы III Междунар. науч.-практ. конф. “Язык и культура в зеркале перевода”, 25-29 апр. 2012.

Малый академический словарь (МАС), под ред. А. П. Евгеньева ; АН СССР, Ин-т рус. яз. Москва, 1957–1960, http://biblioclub.ru/?dict_id=118&page=dict

НКРЯ/RNC (Национальный корпус русского языка / Russian National Corpus), http://www.ruscorpora.ru.