FOreIGNIZATION , AND leXICAl DeNSITY : eNGlISH TrANSlATIONS OF “ meTAI ” ( SeASONS OF THe YeAr ) BY KrISTIJONAS DONelAITIS

Department of English Philology, Vytautas Magnus University Donelaicio st. 52, LT-44248 Kaunas, Lithuania E-mail: v.kaledaite@hmf.vdu.lt Straipsnyje lyginami du K. Donelaicio Metų vertimai į anglų kalbą, atlikti Nado Rastenio (1967) ir Peterio Tempesto (1985), analizei pritaikant Lawrenco Venuti (1995) pasiūlytas savinimo33 ir svetiminimo34 vertimo strategijas. Tai viso Metų teksto vertimai, kuriuos skiria beveik 20 metų. Jie atspindi tuo laikotarpiu vyravusias vertimo procedūrų taikymo tendencijas. Taip pat pateikiamas ir aptariamas įvairių autorių požiūris į kultūrinių realijų vertimą. Strategijų lygmeniu empirinis tyrimas atliktas įvertinant vertimus pagal atskirus savinimo ir svetiminimo strategijų aspektus ir pacių vertėjų aiskinamuosius tekstus (vertėjo pastabas). Leksinio lygmens pokyciams nustatyti buvo pritaikyti tekstynų lingvistikos metodai ir „AntConc“ programa. Jie leido nustatyti du teksto parametrus originale ir vertimuose, t. y. leksinį tankumą ir leksine įvairove. Nors tiriamieji tekstai nedideli, taciau tyrimui panaudota kompiuterinė programa atskleidė, kad leksinis tankumas vertimuose yra beveik 8 procentais mažesnis nei originale. Tai galima vertinti kaip nemažą lietuvisko kultūrinio konteksto praradimą vertimuose.


inTROdUCTiOn
The rural epic Metai (Seasons of the Year) by K. Donelaitis, 1 a milestone of Lithuanian literature, has enjoyed the status of similar literary writings viewed as a 'must' for being translated into different languages. The website http:// www.mab.lt/Donelaitis/vertimai.html provides detailed information about the existing translations of Metai into 13 different languages (Armenian, Byelorussian, Czech, English, Georgian, German, Hungarian, Latvian, Polish, Russian, Spanish, Swedish, and Ukrainian). There are several different translations of the epic into English, German, and Russian.
As regards translations into English, a part of the epic (Joys of Spring, translated by Clark Mills) was published in the anthology of Lithuanian literature The Green Oak (1962). 2 In addition, the commemorative issue of Lituanus (1964) 3 devoted to the 250 th anniversary of K. Donelaitis included several excerpts from The Seasons translated by Clark Mills, Nadas Rastenis, Demie Jonaitis, and Theodore Melnechuk. However, the poem was not translated in full until 1967; this translation was carried out by Nadas Rastenis 4  Early translation reviews of Donelaitis include the first German review (1818) of Liudvikas Rėza's translation. The first evaluation of Donelaitis' poem in English appeared in 1869, in The Saturday review of politics, literature, science, and arts, under the title "Donalitius. The Lithuanian Poet". 6 It is a review of G. H. Nesselmann's edition of Donelaitis' poetry published in Koenigsberg in 1869.   Comments on the quality or analyses of K. Donelaitis' translations are not numerous, neither was I able to locate any of them on the translations by Clark Mills or Nadas Rastenis. One valuable source devoted to the analysis of Peter Tempest's translation of Metai (including an occasional reference to N. Rastenis' translation) is an article by Dalia Vabalienė (1993) entitled "K. Donelaičio Metai anglų kalba" 7 . Her article on Peter Tempest's translation was published two years earlier than the influential book by Lawrence Venuti, The Translator's Invisibility (1995), and D. Vabalienė in her analysis does not employ the terms 'foreignization' and 'domestication'. Even though D. Vabalienė explores the quality of P. Tempest's translation in a rather general way focusing mainly on the lexical level and the form of the epic (hexameter vs. iambic pentameter), she also provides interesting insights regarding the strategies applied in translating culture-specific items, the topic explored later by a number of researchers. This brings us to one of the key issues in recent translation theories, Lawrence Venuti (1995) in particular, on domesticating or foreignizing practices in translation.

THeOReTiCAl bACKGROUnd
Venuti 8 presents his ideas on foreignizing and domesticating (drawing on Schleiermacher (1813) 9 relating them to 'invisibility', i.e. "the translator's situation and activity in contemporary Anglo-American culture". Since then, the terms 'foreignization' and 'domestication' have been employed to denote the strategies applied by translators when rendering texts from one language into another. The main difference between the strategies is whether translation is oriented towards source-language or target-language cultural values. Venuti 10 also stresses the point that "the choice of whether to domesticate or foreignize a foreign text has been allowed only to translators of literary texts, not to translators of technical materials". Venuti 11 defines domesticating translation as [t]he forcible replacement of the linguistic and cultural difference of the foreign text with a text that will be intelligible to the target-language reader.
[…] The aim of translation is to bring back cultural other as the same, the recognizable, even the familiar; and this aim always risks a wholesale domestication of the foreign text […] Seen in this light, domestication serves bringing the translation closer to the target-language reader by using a more familiar language, with much of the 'otherness' taken away. It subsumes changes introduced on various levels of the text for the target readership to fully grasp it. This approach takes into account the S t r a i p S n i a i fact that a different nation, not sharing similar socio-historical experiences or a cultural background, may enjoy a fluent text better. Thus, according to Venuti,12 Fluency can be seen as a discursive strategy ideally suited to domesticating translation, capable not only of executing the ethnocentric violence of domestication, but also of concealing this violence by producing the effect of transparency, the illusion that this is not a translation, but the foreign text, in fact, the living thoughts of the foreign author.
Foreignization, on the other hand, recognizes 'otherness' and shows respect for this kind of specificity. This strategy strives to preserve in the translation the alien, the unusual for the target audience to appreciate: 13 To advocate foreignizing translation in opposition to the Anglo-American tradition of domestication is <…> to develop a theory and practice of translation that resists dominant targetlanguage cultural values so as to signify the linguistic and cultural difference of the foreign text.
what makes the translations of Metai particularly interesting is the amount of its culture-specific content portraying the daily life and the cultural background of Prussian Lithuanian serfs. The problems posed for translators by the cultural content can be approached from different perspectives, first, on the macro-level, which allows us to see the cultural content in terms of its contribution to the effect of the whole text. On the other hand, looking at the individual cultural components, problems arise at the lexical (semantic) level due to the references to a whole network of culture-specific entities such as customs, traditions, clothes, food, institutions, etc. Collectively, they are referred to as culture-specific items (CSIs, Davies 2003). Other terms include 'culture-specific concepts' (Baker 1992), 'realia' (Florin 1993;Robinson 1997), or 'culture-specific references' (Gambier 2004 Attempts to provide solutions for translation of CSIs have been made by many practicing translators and researchers, such as Newmark (1988), Baker (1992), Hervey and Higgins (1992), or Aixelà (1997), to mention just a few. This aspect has attracted the attention of Lithuanian translators and scholars as well and has been discussed, for example, by Balčiūnienė (2005), Mikutytė (2005), Leonavičienė (2014), Pažūsis (2014). Davies 14 notes that the treatment of CSIs in translation depends on 1) the distinction between two basic goals of the translator, i.e. "preserving the characteristics of the ST as far as possible, even where this yields an exotic or strange effect" or "adapting it to produce a target text which seems normal, familiar and accessible to the target audience" (this is reflected in Venuti's principles of domestication and foreignization), which relates to 2) the choice of particular translation procedures suggested for dealing with individual CSIs. It is possible to arrange the applicable translation techniques along the foreignization / domestication axis (see, e.g., Hervey and Higgins 1992, Aixelà 1996, Ghazala 2002). However, this aspect will not concern us here: more advanced corpus analysis tools are needed to analyse the full text of Metai and its translations in order to carry out a full-scale investigation of the treatment of CSIs in the two translations.

AiM, dATA And MeTHOd
Choosing a domesticating or foreignizing strategy for translating a particular text depends on a variety of factors. Holmes 15 believes (as quoted in Davies 16 ) "that among contemporary translators, while there is a general tendency to favour domestication at the linguistic level, there is an opposing trend towards 'historicizing and exoticizing in the socio-cultural situation' ".
One of the aims of the present analysis is to find out which of the opposing strategies is more consistently employed in the English translations by N. Rastenis (1967) and P. Tempest (1985), focusing on how much of the alien and exotic but typical for the source culture has been preserved in each translation (on the macro-level).
A more difficult question is how to measure the actual lexical changes in the treatment of individual CSIs spread across various networks of culture-specific content and to evaluate translation choices as either representing domesticating or foreignizing strategies. One possible way for approaching this complex issue is offered by corpus linguistics methodology (see Kenny 2001, Olohan 2004, Johansson 2007, Vaičenonienė 2011). A starting point can be generating word-frequency lists, which, as claimed by Bowker,17 "are simple yet powerful. In a translation context, they can help translators to determine which words seem to be 'important' on the basis of frequency, and translators can compare the frequencies of different words". Moreover, "the data generated by the frequency list can be further investigated using other features of corpus-analysis tools such as concordancers and collocation generators". 18 First of all, frequency lists are effective for identifying possible losses of important content words (especially CSIs) in translations. This line of research has been applied in this paper; the specific research questions on the micro-level have been narrowed down to investigating two properties, lexical density and lexical diversity (or richness), in the source-text and the two translations. The analysis was carried out using the AntConc programme. 19 Since this is a rather new approach to evaluating lexical changes in translations, it was decided to work with a small amount of data to find out if it is possible to identify any emerging patterns and to assess the potential of the method for large-scale research. Excerpts from the description of different seasons were chosen randomly. In total, 83 lines from the source-text and the corresponding number of lines from each of the two translations were analyzed. The total number of lines is 249; the total number of words is 2,004.

diSCUSSiOn OF FindinGS
In what follows, the specific aspects that are applicable to both English translations on the foreignization / domestication continuum (the macrolevel) will be focused on; next, the differences between the two English versions will be highlighted. The analysis shows that steps towards foreignization in the two translations are seen in the following treatment of the original text: 1) The author's name is written correctly in its complete form, showing respect for his identity.
2) The translators did not shorten the poem.
3) The content and the spirit of the poem is preserved, namely, each translation remains a poetical chronicle portraying one year of daily work of Prussian Lithuanian serfs and the concerns within that society in the mid-eighteenth century. 4) The presentation of another place and time, of a different, unfamiliar culture which is reflected in a specific way of life, customs and traditions speak for purposeful and conscious efforts to maintain the foreignness, strangeness, 'otherness' of the source text for the target reader. 5) In the Translator's note P. Tempest clearly states his position towards his task in translating Metai: "My overriding aim has been to let Donelaitis tell the story in his own words, at his own pace and without impediment"; 20 it is a very clear commitment to foreignization. The differences noticed in the two translations include the following aspects: 1) Spelling of proper names in their original forms is considered to be a sign of foreignization, and P. Tempest systematically follows this practice. N. Rastenis, on the other hand, chooses to drop Lithuanian diacritics while writing proper names, e.g., Prickus, Blekius, Maguze instead of Pričkus, Blėkius, Magužė, and this is a case of domestication. 2) There is one inconsistency, though, in P. Tempest's treatment of proper names. The translation retains the original Lithuanian names, except for Šlapjurgis and Susukatė, the names of a servant and a maid (Winter Cares, line 422). He thus renders them as Tippletom ('tipple' is to drink (alcoholic liquor) or engage in such drinking, especially habitually or to excess) and Fib ('a fib' or 'fibbing' is a type of deception; an untruth) providing the explanation that the Lithuanian variants, respectively, mean 'drunkard' and 'liar'. Conveying more explicitly the semantic value of these two names is a case of localization, and this procedure may be placed between foreignization and domestication on the continuum. The translation of these two "loaded names", 21 however, falls out from the general pattern of providing the original forms of the proper names. Apart from these two, P. Tempest did not attempt to translate other loaded names, such as Slunkius or Pelėda. Based on these observations, it can be assumed that P. Tempest's work is more foreignized than that of N. Rastenis. with respect to domestication, the following aspects should be highlighted: 1) Acknowledging the need for domestication is revealed in the prefaces of the translated texts. Introduction for N. Rastenis' (1967) translation was done by Elena Tumas, whereas Kostas Korsakas commented on P. Tempest's (1985) translation. The introductions present the biography of K. Donelaitis and survey his work, thus enabling the target audience to get to know the author.
2) A very valuable source documenting the translator's decision-taking strategies is translators' notes. N. Rastenis, however, does not offer any of his personal explanations or comments, but P.Tempest in Translator's note 22 reveals his ideas about the need to domesticate the source-text, where he stresses the fact that "Donelaitis wrote his poem in dactylic hexameters, a classical metre hardly used in English verse because of its unnatural lilt and length. So I have cast the poem in iambic pentameters, a metre familiar to English ears and well able to embrace the full text of the original". 3) As regards the changes on the lexico-semantic level, P. Tempest 23 admits: "I have made no attempts to reproduce the archaisms Lithuanian readers now find in the poem. It would be improper to do so. For the same reason I have avoided such historical terms as corvée, the unpaid day's labour a vassal owed his feudal lord, rendering it as 'work due to the lord'." 24 4) Further, P. Tempest explains his lexical choices for official names: "I have also used familiar English words as approximate equivalents for the names, mostly of German origin, of officials -'governor', 'steward', 'bailiff' and 'constable' for 'amstrotas', 'amtmonas', 'pakamorė', and 'vakmistras' respectively". 25 N. Rastenis, on the other hand, opts for neutral, stylistically unmarked variants, i.e. he renders them as 'squire', 'chief', 'overseer', and 'watchman'. 5) P. Tempest also provides Notes on the poem, where the cultural references, probably unfamiliar to an average reader, are given and explained. The Notes contain information about the relevant geographical, socio-historic and general cultural aspects. For example, the translator comments on food items ('šiupinys' -'pea-mush'), folklore phrases (' Jurgut, Jurgut, ...'), or national holidays (Midsummer Eve) 26 -all these items belong to CSIs. In this way such elements are presented as facts that the reader should be familiar with in order to understand the poem. The explanations reveal the translator's role of a mediator, which is, according to Davies, "to provide the target audience with whatever it is they need to know in order to be able to process the translation in a way similar to the way members of the source culture process the source text". 27

20
DOneL aiTiS, Kristijonas. Venuti claims that foreignization can take different forms, such as strict adherence to the target text and retention of cultural markers (or CSIs). As regards Metai, in addition to the culturally specific content, the language of the poem may be treated as stylistically marked; it is characterized as representing "the picturesque folk vernacular", full of "unexpected metaphors, expressive verbs" (Kubilius et al. 1997) 28 and expressive folk phraseology. The poem also contains archaisms and dialectal vocabulary, and all of this makes the translation of the poem into another literary system quite challenging. This aspect brings us to the next issue, that is, lexical density and lexical diversity (or richness) of the texts. Lexical diversity is measured by the number of different word types, while lexical density is counted by dividing the number of lexical (or content-carrying) tokens, such as nouns, lexical verbs, adjectives, and adverbs, by the total number of tokens in the text and multiplied by 100 to show the percentage. 29 The search results retrieved by the AntConc programme were looked through and adjusted manually to ensure more precise statistics. The results reflect the changes on the micro-level (presented in Table 1). As can be seen from Table 1, the lexical diversity of N. Rastenis' translation (a 789-word text) is higher than P. Tempest's with the total number of 654 words. As regards the lexical density, it is the highest in the Lithuanian data, reaching over 62%, while the two translations show statistically important lower density, roughly about 54% each. This means that about 8% of content words were lost in the translations. These findings, even though based on a small amount of data, are in line with the statistics concerning the full text of Metai and N. Rastenis' translation (unfortunately, P. Tempest's translation is not digitalized) 30 : the source text has 6160 word types whereas the translation only 4876. Keeping in mind that the number of function words in English texts, as a rule, is higher than in Lithuanian, this approach for exploring the loss of content-carrying lexis is quite promising. To quote Kenny, 31 "quantitative techniques, including techniques of statistical inference, can launch wider, more qualitative studies of texts and corpora".
As an example, consider lines 558-563 from Spring Joys; a comparison of the original and the two translations provides a preliminary explanation for at least some statistic differences. The first line in English is from N. Rastenis' (Lithuanian background) translation, while the second English line belongs to P. Tempest (British background).
O aruodų krūvos jau taipo pasibaidė, Our bins and barrels, filled in autumn to the brim And in the corn-bins too the mounds have shrunk Kad kisielių virt ir skanų šiupinį pleškyt With life-sustaining rye, wheat, barley, peas and oats, Till almost nothing's left from which to boil nieks nesiliko, kaip tiktai skūpa mažumėlė. now yield but little for our bread and oatmeal pap.
Oatmeal and pea-mush -only dribs and drabs. Ant pašaliai visi, kur ropės irgi repukai The corners where our turnips and potatoes were, look in the corners there, where heaps of swedes, Su kitais viralais zopostui buvo padėti, The barrels that contained our beets and sauerkraut, Turnips and suchlike vegetables lay.
As can be seen from these several lines, N. Rastenis opts for a more descriptive translation, which is reflected in the larger number of content words (see Table 1).
Yet another interesting finding regards the use of function words by the two translators. Table 2 presents the top ten items and their frequency in the respective translations.
The data show that 8 out of 10 top items in the group of function words coincide in both translations; the difference is noticed only in the use and frequency of 2 prepositions (for, of vs. in, with). In addition, N. Rastenis seems to favour the conjunction and a lot; in his translation there also appear more definite articles the (34 vs. 20), which implies a slightly different perspective on the texture of the original narrative.
Two comments follow from the discussion of the statistical counts. First, even though these numbers only highlight the situation regarding the loss of contentcarrying lexis in the translations, it clearly shows the potential for investigating which layer of the vocabulary has undergone most changes. On the other hand, mention should be made of a limitation in applying this method for evaluating lexical richness of the excerpt. Assessment of lexical diversity is related to the vocabulary size (the number of different word types) which, naturally, depends on the text length (the number of tokens). The size of the vocabulary analyzed here is not big; however, it was sufficient for establishing particular tendencies and patterns in the two translators' work.

Concluding remarks
The overall evaluation of the domesticating and foreignizing strategies employed in the two translations demonstrates that each of them is applied to a certain degree. It seems that the translators tried to balance the two approaches for the target reader to be able to appreciate both, the specificity of the cultural content and the fluent domesticated narrative. Their translation choices reflect Venuti's 32 theorizing on domestication / foreignization: "Translation, then, always involves a process of domestication, an exchange of source-language intelligibilities for target-language ones". Even though P. Tempest's translation is more foreignized, it may be suggested that there is a balance between the two strategies in both target texts. This outcome reflects the translators' twofold task (stated explicitly by P. Tempest) to appreciate and render the specificity of the original text and to produce an acceptable text for the target readership.
As regards analysis and comparison of changes on the lexico-semantic level in the two translations, a promising line of future research may be to investigate normalization and creativity of translated texts as proposed by Dorothy Kenny (2001)  in her book Lexis and creativity in translation: a corpus-based study. The remarkably expressive language of Metai deserves a thorough investigation of translation choices while rendering epithets, diminutives, and verb-participle clusters.