Racism, violence and disability in the Finnish (re)translations of Frances Hodgson Burnett's The Secret Garden

(1)

UNIVERSITY OF EASTERN FINLAND PHILOSOPHICAL FACULTY

SCHOOL OF HUMANITIES English language and translation

Anu Emilia Huovinen

Racism, Violence and Disability in the Finnish

(Re)translations of Frances Hodgson Burnett’s The Secret Garden

MA Thesis

March 2019

(2)

ITÄ-SUOMEN YLIOPISTO – UNIVERSITY OF EASTERN FINLAND

Tiedekunta – Faculty

Philosophical Faculty Osasto – School

School of Humanities Tekijät – Author

Anu Emilia Huovinen Työn nimi – Title

Racism, Violence and Disability in the Finnish (Re)translations of Frances Hodgson Burnett’s The Secret Garden Pääaine – Main subject Työn laji – Level Päivämäärä –

Date Sivumäärä – Number of pages

English Language and Transla- tion

Pro gradu -tutkielma x

11.03.2019 121

Sivuainetutkielma Kandidaatin tutkielma

Aineopintojen tutkielma

Tiivistelmä – Abstract

This study focuses on the translation of racism, violence, and negative representations of disability in the Finnish initial translation (Toini Swan 1920, re-edition 1994) and retranslations (Sari Karhulahti 2006, re-edition 2012 and Emilia Numminen 2008) of Frances Hodgson Burnett’s The Secret Garden (1911). The Finnish title of the (re)translations is Salainen puutarha.

The aim of the study is to compare the texts in order to find out if there are differences between the (re)translations in terms of the selected themes as well as to speculate on what is considered suitable in children’s literature based on the changes. The study presents a qualitative comparative analysis of selected parts of the texts. The study is partly material-driven and partly theory-driven.

The theoretical background of this study discusses specific features of children’s literature, retranslation as well as norms and translation norms, since the broad interest of this thesis is to study the translation of inappropriate elements in children’s literature. The theoretical notions concerning children’s literature rely, for example, on the texts of Oittinen (1993, 2000, 2006), Nikolajeva (2005) and Puurtinen (1995). In the study, interpretations of the novel are introduced bases on texts by Kutzer (2000), Kokkonen (2015) and Smith (2018).

Studying retranslations allows comparisons between different translations of a source text, and suggestions about the impact of the target culture and language norms on translations may sometimes be made. The theoretical discussion concerning retranslation relies on Koskinen and Paloposki’s (2015) book Sata kirjaa, tuhat suomennosta.

Kaunokirjallisuuden uudelleenkääntäminen and Deane-Cox’s (2014) book Retranslation. Translation, Literature and Reinter- pretation.

The results of this study show that there are differences between the (re)translations in terms of racist elements, but not much diversity between the translation of expressions of violence and disability. The results of this study do not point to major changes in what is suitable in terms of the themes examined. However, the two retranslations may represent different norms in terms of translating racist elements.

Avainsanat – Keywords

retranslation, children’s literature, norm, qualitative method

(3)

ITÄ-SUOMEN YLIOPISTO – UNIVERSITY OF EASTERN FINLAND

Tiedekunta – Faculty

Filosofinen tiedekunta Osasto – School

Humanistinen osasto Tekijät – Author

Anu Emilia Huovinen Työn nimi – Title

Racism, Violence and Disability in the Finnish (Re)translations of Frances Hodgson Burnett’s The Secret Garden Pääaine – Main subject Työn laji – Level Päivämäärä –

Date Sivumäärä – Number of pages

Englannin kieli ja kääntäminen Pro gradu -tutkielma x

11.03.2019 121

Sivuainetutkielma Kandidaatin tutkielma

Aineopintojen tutkielma

Tiivistelmä – Abstract

Pro gradu -työssäni tutkitaan, kuinka Francis Hodgson Burnettin teoksen Salainen puutarha (englanninkielinen läh- deteos The Secret Garden 1911) ensikäännöksessä (Toini Swan 1920, 10. painos 1994) ja uudelleenkäännöksissä (Sari Karhulahti 2006, uusintapainos 2012 ja Emilia Numminen 2008) on käännetty rasismia, väkivaltaa ja vammaisuuden halventavia kuvauksia sisältävät kohdat.

Tutkimuksen tavoitteena on vertailla käännöksiä ja tutkia mahdollisia eroavaisuuksia niiden välillä. Tavoitteena on myös pohtia, kuinka käännöksissä näkyvät käsitykset siitä, mikä on lastenkirjallisuudessa soveliasta sisältöä. Tutki- mus on laadullinen, ja siinä analysoidaan ja vertaillaan käännöksistä ja lähdetekstistä valikoituja tekstipätkiä toisiinsa.

Tutkimus on osittain aineistolähtöinen ja osittain teorialähtöinen.

Tutkimuksen teoriaosuudessa käsitellään lastenkirjallisuuden kääntämisen erityispiirteitä, uudelleenkääntämistä sekä normeja ja käännösnormeja, sillä tutkimuksen laajempi konteksti on epäsoveliaiden eli normien vastaisten elementtien kääntäminen lastenkirjallisuudessa. Lastenkirjallisuutta ja sen kääntämistä koskeva teoria nojaa muun muassa Oittisen (1993, 2000, 2006), Nikolajevan (2005) sekä Puurtisen (1995) tutkimuksiin. Tutkielmassa esitetään myös kaunokirjallisia tulkintoja teoksesta, ja osion keskeiset lähteet ovat Kutzer (2000), Kokkonen (2015) ja Smith (2018).

Uudelleenkäännösten tutkiminen mahdollistaa saman lähdetekstin eri käännösten vertailun. Lisäksi on mahdollista tulkita, kuinka kohdekielen ja -kulttuurin normit ovat mahdollisesti vaikuttaneet käännöksiin. Uudelleenkäännöksiä ja uudelleenkääntämistä koskevan teoriaosuuden keskeisiä lähteitä ovat Koskisen ja Paloposken (2015) teos Sata kirjaa, tuhat suomennosta. Kaunokirjallisuuden uudelleenkääntäminen sekä Deane-Coxin (2014) teos Retranslation. Transla- tion, Literature and Reinterpretation.

Tutkimustulosten mukaan rasististen elementtien käännöksissä oli vaihtelua eri käännösten välillä. Väkivallan ja vammaisuuden kuvausten käännöksissä ei sen sijaan ollut juurikaan vaihtelua. Tutkimus ei anna viitteitä siitä, että käännökset heijastelisivat tutkittujen osa-alueiden osalta muuttunutta käsitystä siitä, mikä on lastenkirjallisuudessa soveliasta. Uudelleenkäännökset mahdollisesti edustavat keskenään erilaisia normistoja, jotka ohjaavat rasististen elementtien kääntämistä.

Avainsanat – Keywords

uudelleenkäännös, lastenkirjallisuus, normi, kvalitatiivinen menetelmä

(4)

1 Introduction

The aim of my thesis is to study the translation of expressions related to racism, violence and disability in the Finnish initial translation and the retranslations of Frances Hodgson Burnett’s book The Secret Garden (1911), which is a children’s classic. The initial Finnish translation is from 1920 by Toini Swan. I use the 10^th edition of the book from 1994 in my study. The retranslations are from 2006 by Sari Karhulahti and 2008 by Emilia Numminen. This thesis presents a typical research on retranslation, since it focuses on a literary source text and its (re)translations (Koskinen & Paloposki 2015:10-11).

As many scholars (Alvstad 2010; Rastas 2013; Oittinen 2000; Nikolajeva 2005;

Puurtinen 1995) argue, children’s literature has several roles or functions. When children read and are read to, their linguistic skills and reading abilities develop. In addition, literature pro- vides children with information about the world and takes part in socializing them as members of community as well as gives aesthetic pleasure and entertains. As Alvstad (2010:22) points out, children’s literature often aims at conveying positive values, which in turn affects translation strategies because elements that do not fit the target-culture idea of suitable literature may be omitted or modified. Therefore, translated and non-translated children’s literature reflect the cultural values and norms about what is considered appropriate for children, what is expected of children, and what world view a culture and a society want to teach to children. This does not imply that children would lack the ability to critically reflect what they read, or that children would always absorb the values and information they come across while they read. Rather, this means that literature in its part reflects and reproduces representations, values and meanings.

Studying translated children’s literature can give information about what is regarded as suitable for children to read, and what kind of values and worldviews are and can be promoted. Retranslations make a good object of study since they allow comparisons to be made between translations published in different eras. An investigation into several (re)translations of a book may reveal something about the changes or the lack of changes in norms and values and what is familiar in a culture. However, retranslations may have sought justification by dif- ferentiating from their predecessors and in such cases differences in translation strategies may not directly mirror the cultural or linguistic changes (Deane-Cox 2014:13). Nevertheless, retranslations represent their time, cultural context and norms since texts are always produced in certain a culture and society.

(7)

2

The broad interest of this thesis is the translation of inappropriate elements in translated children’s literature. The Secret Garden contains several interesting themes, among which are death, class differences and dialect. However, considering the scope of this study, the topic had to be narrowed down. One of the reasons expressions related to racism, violence and disability were selected is that Minna Salomaa (2013) has previously studied The Secret Garden and its Finnish (re) translations in her master’s thesis Kolme eri avainta Salaiseen puutarhaan – suomennosten vertailua skoposteorian näkökulmasta, which focuses on skopos theory. In her study, Salomaa (2013), for example, studied dialect, gender and expressions that negatively described the appearance of the protagonist. Therefore, those themes are not discussed in detail in my thesis.

The research questions are following:

• How are features of racism, violence, and disability translated it the (re)translations of Secret Garden?

• Are there differences between the translation strategies of the different elements among the (re)translations and if so, what are the differences? What do these changes imply?

To answer these questions a qualitative, comparative study was conducted. The method con- sisted of comparing the selected text passages of the source text as well as the Finnish (re)translations. On the one hand, this study is material-driven as the research question has been formed based on the contents and themes of the source text. On the other hand, theoretical understand- ing of the typical features of translated children’s literature has influenced the research and collecting the material. My initial assumption was that the more recent retranslations would have omitted racism, violence as well as negative representations of disability more than the older initial translation, but this assumption was only partially supported by the material, as the results were more complex. The results will be discussed in detail in chapter 6.

As Koskinen and Paloposki point out (2015:68), retranslations are often justified because the language of the previous translation is considered aged or dated. According to Su- sam-Sarajeva (2003:4), retranslations are in such cases claimed to bring the text closer to the contemporary reader. Aged language needs not always refer to the linguistic aspects, but it can refer to culturally controversial vocabulary or themes. A good and a rather well-known example

(8)

3

of controversy related to editing and retranslating children’s literature is Pippi Långstrump (Pippi Longstocking) and the inappropriate and racist word negerkung that the original Swedish source text contains. In Finland, a revised version of Peppi Pitkätossu (by Päivö Taubert) was published in 2005 mainly because the publisher wanted to remove the racist language.

(Koskinen and Paloposki 2015:224). In Sweden, on the other hand, Astrid Lindgren’s family did not want the original texts to be edited because they considered the text as products of their time that should be interpreted as such. The family, however, changed their minds in 2015 (Koskinen and Paloposki 2015:228). The case is an example of how translated children’s literature can lead to discussions and negotiations about cultural, social and literary values and norms.

Retranslated children’s literature is also discussed briefly is a non-academic Finn- ish text Viisikko aiheuttaa yhä kääntäjille päänvaivaa – Eeva Joenpelto ei kehdannut käyttää omaa nimeään published by Yle (Vilkman 2016). The text focused on the Finnish retranslation of two books of the Famous Five series, Five Go Adventuring Again and Five Run Away To- gether (Viisikon uudet seikkailut and Viisikko karkuteillä in Finnish). It is explained in the text that during recent decades, Blyton’s texts have been modified so that, among other things, racist undertones have been removed and elements unfamiliar to the contemporary reader, such as old metaphors, have been changed. The Finnish publisher as well as the publishing editor, who were interviewed for the article, comment that there is no need to censor the Finnish retranslations. They argue that while in Britain there are versions that are purified, the Finnish publisher has not made such changes in the re-editions since the books are said to reflect the time when they were written.

The news article by Vikman (2016) introducing the retranslation of Blyton’s texts addresses many of the phenomena that I discuss in the theory section as well as the results of my thesis. Firstly, it is related to retranslated children’s literature and a source text which is old and set in a world that is possibly unfamiliar to the contemporary reader. Secondly, the interviewed professionals took a stance against modifying text in favor of following the original source text norm. This is connected to Toury’s (2012) distinction on initial translation norm, which determines whether the source-text norms or the target-text norms are followed. Thirdly, modification is connected to censorship. Fourthly, the notion of source text becomes problem- atic, as it is possible that there are several versions of the text available in the language of its origin. The text by Vikman (2016), despite being short, illustrates that retranslated children’s

(9)

4

literature is an interesting and relevant topic in public discussion, and therefore researched information is needed and welcome.

In this thesis, I will first discuss the theoretical background. In chapter 2, I address retranslation, norms and ageing of texts. In chapter 3, I discuss children’s literature and translating children’s literature. In connection with children’s literature I will address censorship, ideological manipulation, and multiculturalism. After that I will very briefly discuss violence in children’s literature and then discuss disability in literature.

In chapter 4, I introduce the novel The Secret Garden and its translations as well as the author and translators. In chapter 5, I discuss the material and the method. I will explain how the study was conducted and what are the shortcomings of the method. In chapter 6, I will present the results of the study in detail. Finally, chapter 7 presents the conclusion of this study.

(10)

5

2 Retranslation, translation norms and ageing of the translation and the source text

Next, I will present the theory of my thesis. In this chapter, I will first critically address the retranslation hypothesis and briefly present the idea of retranslation as rivalry. After that, I will discuss translation norms with a focus on Toury’s (2012) categorization of norms. In addition, I will briefly present the concepts of domesticating and foreignizing. Finally, I will move on to the ageing of the translation and the source text. Throughout the chapter, my aim is to discuss each theoretical approach from the perspective of children’s literature, which will then be discussed in detail in chapter 3.

2.1. Different perspectives on retranslation

Retranslation is a subsequent translation of a source text that has earlier been translated into a target language (Brownlie 2006: 146; Susam-Sarajeva 2003:2). As Susam-Sarajeva (2003:5) points out, retranslation is not limited to fiction because all kinds of texts are retranslated. How- ever, most research and theoretical discussion on retranslation focuses on literary texts and especially on canonical literature as well as sacred texts (Brownlie 2006:146; Koskinen and Pa- loposki 2015). Because of this, examples about retranslation are often adopted from literary translation and therefore the arguments and notions often concern the so-called great works of literature (Susam-Sarajeva 2003:2).

According to Susam-Sarajeva (2003:2), retranslations are usually studied to illustrate an aspect of translation such as chances in the target language or target culture. Susam- Sarajeva (2003:2) thus argues that while retranslations are studied, a theoretical discussion concerning retranslations is rare. In her text published in 2014, Deane-Cox (2014:1) refers to Su- sam-Sarajeva’s notion made in 2003 and further argues that the claim is still relevant a decade later. In addition, Koskinen and Paloposki (2015:8, 61) argue that retranslated texts are often studied as cases, and such studies do not focus on the retranslation process or phenomena that are typical of retranslation. In this respect, my study is not an exception, since the aim is to compare how features of racism, violence and disability are translated in three (re)translations of the F.H. Burnett’s novel The Secret Garden. However, since the material of my study con-

(11)

6

sists of the source text, its initial translation as well as two retranslations, the theoretical background of retranslation as well as the retranslation hypothesis are discussed. It also needs to be stressed that my thesis focuses on translation and retranslation of a literary text.

Retranslations are often approached from two different directions. On the one hand, emphasis is placed on how well the translation can convey the meaning of the source text.

From this perspective, a retranslation aims firstly at restoring those elements of the source text that the previous (re)translation(s) lacks or secondly, to correct any errors it contains. Therefore, the purpose of the retranslation is to bring the audience closer to the source text (Susam-Sara- jeva (2003:2; Deane-Cox 2014). On the other hand, the retranslation can be made to suit the requirements of the target language and culture and the expectations of the target audience. The underlying assumption is that while time has passed, changes in translation practices, norms as well as cultural and linguistic conventions have taken place and the previous translation has aged or become dated, especially in terms of language. Moreover, translations are considered to mirror their historical, cultural and political surroundings, not just the source text. This means that while the cultural context changes, the (re)translation ages. From this perspective, the goal of retranslation is to bring the source text closer to the contemporary reader (Susam-Sarajeva 2003:4).

2.2 Critical approach to the retranslation hypothesis

Source-oriented studies that deal with retranslation have commonly addressed the retranslation hypothesis, which originates from Antoine Berman’s theoretical discussion (Brownlie 2006:148). It is important to note that Berman did not present his ideas in the form of a hypothesis; the term has only been used later (Browlie 2006; Koskinen and Paloposki 2015:66: Deane- Cox 2014).

Koskinen and Paloposki (2015: 66) trace how Berman’s views have been transferred into the retranslation hypothesis. They mention Andreas Poltermann (1992:18) and Theo Hermans (1999: 139–140), and Andrew Chesterman (2000: 21–25) who referred to the translation hypothesis as an example of a descriptive hypothesis. Koskinen and Paloposki (2015:66) note that the hypothesis became widely known after Andrew Chesterman and Jenny Williams (2002: 72) presented it in their book The Map: A Beginner's Guide to Doing Research in Trans- lation Studies. The retranslation hypothesis in the book suggests that retranslations tend to be closer to the original text (Williams & Chesterman 2014: 72).

(12)

7

At the root of the retranslation hypothesis is Berman’s theoretical approach, most notably presented in the text La reduction comme espace de la traduction published in the 1990 issue of Palimpsestes which focused on retranslation (Deane-Cox 2014:3; Brownlie 2006:147).

In the article, Berman (1990 in Koskinen and Paloposki 2015:67) argues that no translation can be perfect, and the possibility or even the inevitability of a retranslation is an inherent part of translation.

Deane-Cox (2014:3) claims that progression through repetition is an integral fea- ture of Berman’s (1990) approach: the initial translation is defective, and its shortcomings can be overcome in a restorative retranslation, which is reflective and aware of the errors of its predecessor. Brownlie (2006:148) similarly argues that Berman favors a source-oriented translation, which meticulously conveys the patterns and features of the source text in its material and non-material aspects. Koskinen and Paloposki (2015: 68) however point out that while Berman argues that the initial translation is inadequate, he does not explain what constitutes a good translation and he does not provide any guidelines for comparing the source text with (re)translations.

Brownlie (2006:148) argues that German Romanticism and especially Goethe’s theory of a translation cycle have inspired Berman’s theoretical stance. Deane-Cox (2014:3) presents Goethe’s (1992:64–66) logic, according to which different kinds of translations are made to correspond with the receiving culture’s phase of reception of the source culture. The phases that the translations represent and reflect are called epochs (Deane-Cox 2014:3). The first epoch of translation acquaints the foreign country on the receiving culture’s terms, the second epoch presents a text in which the translator aims to place themselves into the foreign situation but end up appropriating the foreign idea and present it as their own (ibid.). The third epoch of translation is the final and highest (ibid.). During such an epoch, the translation aims to match the original text completely (ibid.). As Deane-Cox (2014:3) explains, the underlying idea of Goethe’s logic and the epochs is that of progression, in which the foreignness of the source text is first rejected, then appropriated and finally expressed in its own terms. As Deane- Cox (2014:3) stresses, in Goethe’s line of though retranslation’s purpose is to reveal the truth of the source text within a receiving culture. A similar idea that translation is a process which starts from an adaptive, defective initial translation and moves towards a source-oriented retranslation is, according to Deane-Cox (2014:3) and Brownlie (2006:147–148), characteristic of Berman’s (1990) approach. As interpreted and translated by Deane-Cox (2014:3–4), for Ber- man (1990:7) the culmination of retranslation is the great translation (grande traduction) which

(13)

8

restores the meanings of the original text that the previous translations have been unable to mediate. Since the great translation is formed in a reiterative process, an initial translation cannot be, according to Berman (1990 in Koskinen and Paloposki 2015: 67) a great translation. In Berman’s (2009:30) approach, great translations become remarkable works in their own right.

They are not regarded as secondary to the source text but become “a new original”, and in a way transcend the category of a translation.

According to Koskinen and Paloposki (2015: 67, 68), in Berman’s interpretation of Goethe’s model the first translation is a type of a gloss, a word for word translation, and the second translation is a free, adaptive version. The third translation is literal, which means that the translation conveys the cultural, textual and other special features of the text. In his own approach, Berman does not incorporate Goethe’s first epoch. As the first step Berman underlines the importance of introduction, paratexts, and paving the way for the initial translation (Koskinen and Paloposki 2015:67–68; Berman 2009:42).

Berman (2009:42) later expressed arguments that are similar to what he presented in his 1990 article, as he discussed the general transfer of literature in his book Toward a Trans- lation Criticism: John Donne (the original French text John Donne Pour une critique des tra- ductions was published in 1995). The theoretical view of transfer is presented as a part of crit- icism toward approaches, which focus on translation norms.

According to Berman (2009:42), the transfer can be divided into different mo- ments. Firstly, the source text becomes familiar in a culture as it is read and possibly integrated to a corpus of literary texts that are taught at educational institutions. This is followed by an introductory work about the text. After this comes the first translation, which has literary am- bition, but is defective. The first translation is followed by (re)translations and eventually a translation “of the whole oeuvre”. The process is accompanied by critical work, which, according to my interpretation, includes analyses and evaluations of the translations. Finally, there may be a canonical work, which stands out and stops the retranslation cycle for a long time.

Berman (2009:43–44) strongly underlines the source-text oriented approach, as he claims that translation’s purpose is to reveal the essence of the source text. Berman (ibid.) further argues that if the norms of a receiving system determine how a text is read and then translated, then the translation will not convey the essence of the source text.

Brownlie (2006:146) argues that Berman’s approach towards retranslation has similarities with the essentialist views of the narration theory. As presented by Brownlie

(14)

9

(2006:146), Chatman (1978) claims that a narrative has two parts, a story and a discourse. The story is the content, the deep structure of the narrative in which events take place in a certain order. Discourse is the superficial structure that is used to tell the story and it includes the narrative techniques such as first-person narration. The central idea is that there is a core of narration that can be conveyed in various ways on the level of the discourse. According Browlie (2006: 147), in such a view, the deep structure represents an essence, which can be transferred without changing it. Brownlie (2006:147) further argues that Berman’s (1990) theory on retranslation as a process of improvement also implies an essence, as the successive translation reveals the truth of the source text. However, as Brownlie (2006: 154) explains, Mcquillan (2000) has criticized the idea that a text has a non-historical essence. Mcquillan (2000 in Brown- lie 2006:154–155) rather argues that the meaning of texts, words and utterances are dependent on their context. Thus, if it is assumed that a text does not have a core meaning which is inde- pendent of its context, it is also difficult to claim that a retranslation will convey the true essence of a text.

Deane-Cox (2014:7) reminds that while the roots of retranslation hypothesis – the idea that first translations are target-oriented and later translations are source-oriented – can be traced back to the idea of a reiterative translation process that will reveal the source text and its cultural and linguistic specifics the retranslation hypothesis has not been purposely formed from Berman’s approach and his interpretation of Goethe. According to Brownlie (2006:148), retranslation hypothesis is an example of a transfer of an idea from one intellectual tradition (Ro- mantic Idealism) to another (Natural Sciences). In fact, Brownlie (2006:148) argues that because of this transfer, Berman’s (1990) theoretical approach is taken out of its context and therefore the retranslation hypothesis does not in itself contain the implication of an essence or an improvement over time in a similar way that Berman’s theoretical views do.

Koskinen and Paloposki (2015:84, 100) argue that the retranslation hypothesis is not adequate to explain retranslation and retranslation practices. By referring to the history of Finnish translation history and the emergence of Finnish literature and written Finnish, Koskinen and Paloposki (2015:84) suggest that the retranslation hypothesis rather describes a specific era of literature, during which a young, emerging literature incorporates adaptive translations, which are later followed by translations that are less adaptive. However, it might be further asked whether the retranslation hypothesis describes the emergence of a literature in surroundings where certain cultural, social and political conditions occur. There is research discussing political and ideological conditions that affect (re)translation. For instance, Pokorn

(15)

10

(2012) has discussed how the impact of the socialist regime and communist ideology material- ized as censorship in the (re)translation of children’s literature. A repressive political ideology is an extreme example of the external conditions of translation, but it highlights how cultural, political, ideological and historical factors can affect retranslation.

2.3 Retranslation as rivalry and differentiation

As Deane-Cox (2014:12) argues, retranslation can be regarded as a challenge and in such cases, the connections between the different (re)translations and the source text is central. According to Pym (1998:83), a retranslation undermines the validity of the previous (re)translation as it aims to bring something new to the text or to possibly correct errors in it. A re-edition, in its turn, tends to reinforce the authority of the (re)translation in question. Therefore, (re)translations which share the same or almost the same cultural location and emerge close in time to one another are in Pym’s (1998:82–83) approach active retranslations which rival each other.

In Pym’s view (1998:83), the study of active retranslations focuses on the translation itself. This means that the wishes of the publisher and readers as well as the agency of the translators, rather than cultural or translational norms or conventions, tend to be taken into consideration when the reasons behind retranslating are studied and when aspects of retranslations are analyzed. Deane-Cox (2014:13) refers to Venuti (2004:25), who regards retranslation as an act of differentiation and a conscious choice of reflecting certain values in the text. This purposeful differentiation can further be regarded as a way of justifying the existence of the retranslation (Deane-Cox 2014:13). Venuti (2004:26) further argues that when a retranslation seeks justification by claiming that it delivers the source text or parts of it adequately, there is a strong implication that the previous version is not acceptable, is in some way insufficient or lacks linguistic correctness.

Pym (1998:82) claims that retranslations which are separated by a long time or synchronic boundaries, such as geopolitical boundaries, are passive retranslations because they are not motivated by a rivalry between them. As Koskinen and Paloposki (2015:9) argue, passive retranslations do not compete over the same audience, because they are, for instance, targeted at a different readership. Children’s literature is a good example of this, since some versions are targeted at children and others at scholars, as (Oittinen 2006:36) notes. Pym (1998:83) is critical of the study of passive retranslations and implies that it runs the risk of

(16)

11

“blindly surrendering causality to target-culture norms.” According to Deane-Cox (2014:16) Pym’s view on passive and active translations does not have a solid empirical basis and is to some extent contested by research. Deane-Cox (2014:15–17) illustrates this by referring to, for instance, Seth’s (2010) research on the French retranslations of Adam Smith’s texts. The different translations studied were published two centuries apart, but the legitimatization of the retranslation is based on differentiation. In addition, the authority and validity of the latest version are underlined by mentioning and thus to some extent criticizing the choices made in the previous version. According to Deane-Cox (2014: 16–17), this goes against Pym’s idea that (re)translations published a long time apart are not motivated by rivalry and are therefore passive. In addition, it is worth noting that Koskinen and Paloposki (2015:56) argue against Tiit- tula’s (2013) view that an increased time gap between the Finnish (re)translations of German literary texts implies that the translations are passive and do not compete with each other. Ac- cording to Koskinen and Paloposki (2015:56), the difference in age alone is not enough evi- dence to make assumptions about the relationship between (re)translations.

Deane-Cox also presents cases that seem to fit with the idea of rivalry, such as St André’s (2003) discovery that Sir John Frances Davis’s translation on the Chinese novel Hao qiu zhuan in 1829 is a purposeful differentiation of the first translation by Percy in 1761. Deane- Cox (2014:15), however, points out that there are studies that show that retranslations sometimes deliberately repeat some elements of the previous translations and thus seek to do the opposite of differentiation. To illustrate this, Deane-Cox (2014:17) presents Mortier’s (2010:145) observation that Georges Proser’s French retranslation of Brecht’s Der Kaukasische Kreidekreis from 1978 follows some wordings, even mistakes of the previous translation by Jacob and Pfimmer.

2.4 Norms and translation

Deane-Cox (2014:9) points out that the social nature of translation as well as the network of social, cultural and political practices influencing translation can be discussed from the perspective of norms. Deane-Cox (ibid.) further notes that this is due to the popularity gained by Gambier’s (1994:416) argument that translations are reflections of the norms of their time. Re- translation is therefore an update, which is required because of the changes in the audience, their tastes, needs, and competences (Gambier 1994: 413 in Susam-Sarajeva 2003:4). Brownlie

(17)

12

(2006: 150) claims that retranslations are needed because the previous translation is unaccepta- ble, as it no longer fits the contemporary language or ways of thinking and behaving. The question of what is considered appropriate at a specific place at a specific time is a part of the academic and public debate about (re)translation, and appropriateness is in its turn linked to norms.

Brownlie (2006:150) argues that translations are affected by ideologies and norms of a specific culture at a specific time. According to Brownlie (2006:150), ideologies refer to sets of beliefs whereas norms are rather related to practices. Brownlie (2006:150) further argues that the definition of ‘norm’ is ambiguous. On the one hand, the term is used to refer to what is usual or habitual and, on the other hand, to describe what is acceptable. In Chesterman’s (2007:

357) definition, a norm is a society’s or a community’s sense of what is a correct way to act.

According to this view, a norm always contains a shared idea of appropriateness, and thus a typical practice is not the same thing as a norm. Moreover, a norm is characterized by regularity and intersubjectivity, and a community needs to have the means to criticize those who break norms (Chesterman 2007: 358–359). According to Puurtinen (1995:48), conventions are typical and preferred actions in a given situation. What differentiates them from norms is that ignoring them does not lead to sanctions while breaking a norm does. Norms, in general, are thus more binding and constraining than conventions. However, as Puurtinen (1995: 48) explains, there is not a full agreement on the distinction between a norm and a convention, since one scholar’s definition for a convention can be the same as another scholar’s definition for a norm.

According to Brownlie (2006: 150), the main sets of norms that affect translation of literary texts are literary, linguistic, and translation norms. Thus, translation and the selection of text to be translated is linked to the mainstream views of acceptable or good literature. In the case of children’s literature, translation is influenced by what kind of language and themes are considered appropriate for children. Therefore, the question is not just about translation or linguistic norms but also about cultural and social norms.

Toury (2012a:82) has classified translation norms into 1) initial norms, 2) pre- liminary norms and 3) operational norms. Preliminary norms guide the general translation policy as well as the directness of translation. Translation policy refers to the factors that govern the systematic selection of text types and even individual texts that are translated into a language and a culture at a particular time and place. Directness of translations refers to the tolerance of translation from an intermediate language, i.e. other languages than the source text: is it allowed to begin with, and if so, what languages and text types can be or cannot be translated from an

(18)

13

intermediate language? Preliminary norms thus influence what kind of texts are acceptable to be translated. As Chesterman (2007: 360) argues, the strongest preliminary norm determines whether a text can be translated, or more precisely, whether a translation can be published.

The initial norm is connected to the choice of whether the translation follows the norms of the source text and the source text culture or the norms of the target culture and language (Toury 2012b:170). As Puurtinen (1995:43) notes, the initial norm refers to the overall choice between adequacy and acceptability. The concept of adequacy refers to source-oriented approach, and it is used to describe how accurately the meanings of the source text are repro- duced in the translations. Acceptability, on the other hand is connected to a target-oriented approach, and it refers to whether a translation fits the norms and conventions of the target language and culture (Toury 2012:171; Puurtinen 1995:23, 41). Chesterman (2007: 360) reminds that the decision regarding the initial norm depends on several factors, such as the purpose of the translation, the commissioner’s expectation, and prevailing translation tradition. It needs to be remembered that different text types may favor or require different norms. For instance, the initial norm governing the translation of legal texts and advertisements may not be the same.

Operational norms govern the translator’s decisions during the translation pro- cess (Toury 2012b:172). Toury (2012b:172–173) makes a distinction between matricial norms and textual-linguistic norms. Matricial norms regulate how the linguistic material of the translation is presented, for example, the way a text is segmented into passages. Textual-linguistic norms control the selection of the source-text material that will be formulated into the target text. In other words, they determine what parts of the source text are included in the translation.

Toury (2012b:172–173) further argues that some textual-linguistic norms can apply to translations in general or to a specific text type. In addition, some norms may be the same that govern the production of non-translation text, but that is not necessarily the case.

As Toury (2012b: 171) points out, in actual instances of translation norms are used systematically, but not absolutely. This means that if the selected initial norm dictates that the norms of the source text are followed, there may nevertheless be instances in which the norms of the target language are followed.

(19)

14 2.5 Changing and coexisting norms

As Venuti (2004:28), Van Coillie (2006:132) as well as Chesterman (2007: 362) remind, translation norms are learnt from educational institutions, authors, publishers, and other possible agents involved in translation training and the translation industry. Because norms are inher- ently social and cultural constructs, they are not constant or static. The changing of norms re- quires, however, that they are made visible and contested. The same institutions that maintain and teach translation norms can also be the ones who question, transform, negotiate and challenge them (Toury 2012b: 174). The change may be gradual, which means that several norms can coexist or compete with one another. Toury (2012b:175) makes a distinction between dominant, prevailing mainstream norms, remnants of previous norms, and new norms that are forming.

Puurtinen (1995:44) points out that translation norms often change together with the constraints and convention of the target literature. As an example, she mentions the Finnish translations of William Faulkner’s novels. In the earliest versions of the 1940s, the translated text matched the traditional Finnish idea of what a novel is, and the language did not aim to create similar effects as the modernist techniques, such as deviant word formation, used in the source text. Modernization of the conventions of the Finnish language and literature in the 1960s influenced the later retranslations and the translators were able to introduce new techniques to reach a structural similarity with the source text. Therefore, it can be said that the target culture’s normative expectation and ideas of what a literary text is were mirrored the translation. In addition, since literature and translated literature are parts of society, cultural and social ideas of acceptable language, themes, and representations affect which type of literature is written and translated and which translation strategies are preferred.

Puurtinen (1995: 45) further reminds that translation norms do not necessarily match the literary norms of a given culture at a given time. For instance, in translated poetry, the norms of the target literature can be broken in order to achieve adequacy. Moreover, Koskinen and Paloposki (2015:75) point out that it is possible that norms are deliberately broken in a retranslation. Puurtinen (1995:45), however, claims that in this regard, children’s literature is less flexible, as the linguistic norms of the target language are more likely to be followed.

(20)

15 2.6 Studying translation norms

As Toury (2012b:170) argues, the problem with studying norms is that the norms which govern social behavior are rarely explicated or verbally formulated. This means that people know how to behave or how to not behave in a situation specific to their culture, but it is hard to define and to name the norm that guides their behavior and actions. According to Chesterman (2007:363), the normative approach can be criticized because norm as a concept is considered ambiguous and too abstract. If norms cannot be given a rigorous definition, and their existence is not always easy to empirically show, it is also very difficult to claim that a certain translation or translation strategy is a consequence of a specific set of norms.

According to Toury (2012b: 176), (translation) norms can be studied from peritextual and epitextual material. This means that in addition to the source texts and its translations, critiques, reviews as well as comments such as preface and footnotes can be included in the research material. In addition, Toury (2012b:177) argues that researchers need to be critical when they study, for instance, introductory chapters which may contain the explanations as well justifications for chosen translation strategies. Such texts may be studied as examples of what is said about the translation rather than as neutral accounts of the process. Lathey (2006:2) similarly reminds that what is said in an introductory text may not be the whole picture of the process or as the text represents the writer’s interpretation of the text and the translation project. The researcher should be skeptical, and the paratextual material can also be treated as texts that reflect, for example, certain discourses of translation. Lathey (2006:2) further notes that when it comes to children’s literature, translators rarely make themselves visible, and when they do, they usually justify their choices in the preface or introduction.

The translation process and the involvement and roles of different agents is rarely explicated in translated literary texts. For instance, a reader or a researcher does not necessarily have access to the editor’s or publisher’s comments and wishes or the details of the commission.

The finished, published book often contains no information about how the translator has been advised to translate, for instance, racist expressions, nor is there information about whether the translator considers some of their choices as the best possible alternative or a compromise.

Especially when there are no peritexts or other extratextual material available, the translated text needs to be considered a product in which the source text, source culture, target culture or (translation) norms and practices as well as the individual style and choices of the author and the translator intersect. However, as Brownlie (2006:156) warns, there is always the possibility

(21)

16

of oversimplification when trying to explain phenomena such as style, explicitation, and omis- sions as ideological, normative or typical of a specific culture. There may be many reasons for translating a book or several factors that affect what the translation is like, such as translators’

personal styles as well as publisher’s commercial interest. Therefore, it is not always possible to reduce the translation strategies or choices to the linguistic, cultural, and translation norms of the target culture. I would argue that this is applies especially when one or a few texts are studied, as it is hardly possible to make very broad generalizations. Nevertheless, translations are products of their culture and they do, to some extent, always either intentionally or uninten- tionally reflect the linguistic, cultural, literary and translation norms and conventions of their time and place.

2.7 Domesticating and foreignizing

The concepts of foreignizing and domesticating translation strategies often go hand in hand with the retranslation hypothesis. Simply put, foreignizing refers to preserving the elements of the original cultural context such as names, whereas domestication refers to the adaptation of the culture-specific terms and other elements (Paloposki 2011:39). Thus, domestication and foreignizing depict a similar idea of two opposing strategies as the contrast between a source- oriented and a target-oriented translation.

Lawrence Venuti (1995; 1998) criticizes the distinction between domesticating and foreignizing. He claims that all translations are to some extent domesticating since they are conveyed through the language, meanings, representations, and discourses of the target culture (Venuti 1995, 1998; Koskinen and Paloposki 2015:74). According to Venuti (1998:75), the values of the target-language culture are imposed on the translation, and translations construct a representation of foreign cultures rather than merely mediate them. As Faiq (2007: 211–213) argues, this means that foreign cultural values and elements are often expressed in ways that are familiar in the dominant western culture (Faiq 2007: 211-213). Therefore, translations may reflect and further reinforce cultural power relations, sometimes even stereotypical images of the foreign. Moreover, translation can reveal something about the dominant norms, hierarchies of the target language as well as the target language culture.

According to Venuti, (1995:20) foreignness is also context-bound as different elements can be unfamiliar in the target-language culture at a specific time. A rather concrete

(22)

17

example of this is how the names of food items in the initial Finnish translation of Carroll’s book Alice’s Adventures in Wonderland are translated differently than in the later (re)translation. This is partly because the food items and the source text-culture were unfamiliar and thus

“foreign” in the target culture at the time of the initial translation (Oittinen 1997:45).

According to Venuti (1995:20), a foreignizing translation can create an impression of foreign by disrupting the prevailing cultural codes of the target language. A foreignizing translation deviates from the norms of the target language by using, for example, a marginal discourse, but it still does this from within the target language. While foreignized translations can be considered to reflect the target-language values, they are nevertheless less transparent (Venuti 1995:34) than domesticating translations. Therefore, they may question the target-language norms as well as to make readers more aware that they are reading a translation. In Ve- nuti’s (1998:11) view, translations that aim to resist, break and uncover the dominant target culture values are called minoritizing. Minoritizing strategy, as the name implies, includes the use of non-standard and alienating language and it aims to disrupt the dominant discourse of translation and literature of the target culture. Venuti’s views of foreignizing and minoritizing thus differs from the idea of many source text oriented approaches, since the purpose would not be to reproduce the source text but to create an impression of foreignness.

Oittinen (2000:75) is critical towards the idea of foreignizing as promoting dis- turbance when it comes to children’s literature, since it does not take the special needs of a child audience into account. In addition, Oittinen (2000:75) criticizes the assumption that readers generally look for fluent translations because that view does not consider the possible multiple readers; for instance, scholars might find the domesticated children’s literature repulsive.

Koskinen and Paloposki (2015:75) also note that there are various and sometimes even con- flicting reasons as well as ways to use a foreignizing translation strategy. Sometimes foreignizing is not a conscious choice but is caused by the interference from the source text.

What needs to be remembered is that Venuti (1995; 1998) focuses mainly on the Anglo-American translation and publishing practices, which in his view predominantly prefer a translator’s invisibility, transparency and fluency. However, as Venuti (1995) himself points out, the amount of translations into English is considerably smaller than the amount of translations from English into other languages, which indicates an imbalanced cultural exchange. It might be claimed that domestication and foreignization have different implications in a culture that heavily exports its cultural products than in a culture that mostly receives influences, let alone cultures and languages that are marginal. Thus, it is important to discuss the choice of

(23)

18

foreign or domestic values in the larger context of cultural import as well as translation policy.

Emmerich (2017:7–8) refers to Venuti’s view as she describes how her own English translation The Few Things I Know About Glafkos Thrassakis from Greek was modified and shortened to meet the publisher’s needs. Emmerich (2017:8) further expresses that some might find it unset- tling that a work from a relatively minor language was reshaped for an American audience that represents a dominant target language and culture.

Both Venuti (1995, 1998) and Berman (2009) prefer foreignizing strategies, but while Venuti (1995:1998) stresses the importance of resistance and breaking the illusion of transparency, Berman regards foreignizing as an ethical choice. In Berman’s (1984:287 in Koskinen and Paloposki 2015:74) view, a bad translation is ethnocentric and adds domestic attitudes to foreign texts, whereas a good translation strives to limit ethnocentrism: a foreignizing strategy and retranslation are thus ways of resisting an inward-looking stance.

When foreignizing and domesticating are discussed on an abstract level, there is a danger of slipping into relativism. On a concrete level, researchers may sometimes be faced with problems when studying and defining, for example, which lexical, syntactical or semanti- cal features indicate whether a translated text or parts of it have been domesticated or foreignized. In addition, Kruger (2012:189) notes that translated text may contain, for example, both domesticating and foreignizing lexical elements, and it can be difficult to solve which category a (re)translation, as a whole, represents.

2.8 (Re)translation, time and ageing of a text

As Koskinen and Paloposki (2015:202–203) point out, time and ageing are integral parts of the discussion and discourses concerning retranslation. After all, retranslation as a term has a tem- poral dimension since it refers to a translated text that follows the initial translation or a previous retranslation or retranslations. As Koskinen and Paloposki (2015: 68) argue, on the one hand, (re)translations are justified because the previous (re)translation is considered aged or dated especially in terms of language. According to Susam-Sarajeva (2003:2), this means that as time has passed, culture and language have changed, and a translated text no longer fits the conventions of the target language and culture or the expectation on the target audience. On the other hand, retranslations are justified because the previous (re)translation(s) are not considered to convey the source texts properly (Koskinen and Paloposki 2015: 68).

(24)

19

Susam-Sarajeva (2003:2) claims that while the source text oriented approach is also connected to time, it contains an assumption of progress as well. As time passes, the initial translation is succeeded by a retranslation or retranslations that are usually better because the translators are more equipped and skilled to produce a translation which is more faithful and closer to the original than the previous (re)translations (Koskinen and Paloposki 2015: 236).

Susam-Sarajeva (2003: 2) argues that in this case the English verb to succeed carries both its meanings, i.e. to come after somebody or something and to be successful. However, it may be discussed whether the idea of progress is ingrained in the target text oriented approach as well.

Could it be so that the previous (re)translations are regarded as dated partly because language or society are thought to have been evolved since their publication?

Koskinen and Paloposki (2015: 68) note that the two justifications may contradict each other, as one of them suggests that a retranslation responds to the changes in a culture or a language, and the other implies that a retranslation aims at transferring the elements of the source texts as closely as possible. On a concrete level, this can mean, for instance, that following the source text as meticulously as possible might require syntactic or lexical accuracy, which in turn may contradict with the contemporary linguistic and literary practices. What the approaches have in common is the thought that the previous translation is somehow unfit.

However, as Koskinen and Paloposki (2015:231) point out, both written and spo- ken language changes. Therefore, texts that are not contemporary may deviate from the prevailing ideal of good or fluent language, when it comes, for example, to syntax, grammar or vocabulary. As Chesterman (2007: 360–362) points out, some changes are formalized as the institutions that have the authority to give guidelines regarding correct language update instructions.

Ageing of language is often discussed in the context of retranslation, but it is not always simple to determine what ageing of language is and how exactly it manifests itself.

Interestingly, the popular claim that a retranslation offers an update to the outdated language of the previous translation is not necessarily based on empirical material on linguistic and textual features of language change but more on an idea or an assumption that translations age and that the language of a retranslation is more modern than its predecessors. Such discursive features of the ageing of the language is illustrated by Koskinen and Paloposki’s (2015) research in which they studied the reception of retranslations by studying reviews of published literary retranslations. The reviews under scrutiny were mainly published in 2000 in Finnish newspa- pers such as Helsingin Sanomat. Koskinen and Paloposki (2015) found out that many of the

(25)

20

reviews contained similar notions and views regarding ageing of language (Koskinen and Pa- loposki 2015: 240). In many reviews, the language of the latest retranslation was considered modern, contemporary, often as opposed to the previous (re)translation, which were implicitly regarded as dated. However, empirical comparisons or examples were not used to support the claim (Koskinen and Paloposki 2015:235). Popular discussion on retranslation is in line with the academic discourse as, according to Koskinen and Paloposki (2015), another recurring theme in the reviews was the (re)translations’ closeness to the source text. Koskinen and Pa- loposki (2015:240) note that none of the reviews in their material claim that the latest (re)translation is further from the source text than a previous translation; on the contrary, there were examples where the reviewer argued that the retranslation was closer to the source text than the previous translation.

The view that translation is an evolving process may also affect the way retranslations are justified as well as how texts are translated or interpreted. In an issue of La retran- duction, Kahn (2010:216 in Deane-Cox 2014:6–7) comments on the German (re)translation of Proust and argues that while the latest versions may not have brought anything new to the text, there is a hope that a new translation will be made, which will meet the criteria of Berman’s great translation (Deane-Cox: 2014: 7). Similarly, in her text published in Autour de la reduc- tion, Béghain (2011:85 in Deane-Cox 2017:7) considers her own French (re)translation of Char- lotte Brontë’s Villette an improvement because it contains the accuracy that was lacking in the previous translation by Baccara (Deane-Cox 2014: 7).

Retranslations may not always be a consequence of aged language or they may not represent a stable, shared sets of norms. Susam-Sarajeva’s (2003) study shows that the vast number of Turkish retranslations of Roland Barthes’s works, which were published between 1960–1995, mostly between 1975–1995, reflect the process of forming a discourse in the receiving system. Retranslations thus represent negotiation, debate, rejection and acceptance of translated terminology. As Susam-Sarajeva (2003) points out, in Pym’s (1998) terms these retranslations that often coexist are active as they share the same cultural location and generation.

If scrutinized from the perspective of improvement or development, the abundance of retranslations shows rather “a spiral-like and vertiginous ‘evolution’” than a linear progress. (Susam- Sarajeva: 2003:6).

It is possible that a translation is not acknowledged as a translation that is affected by the target-language norms and the practices or cultural norms of a given society. The translation may be viewed as a transparent text no different from the source text. Therefore, there

(26)

21

can be a lack of discussion for the need of an updated or a revised translation (Deane-Cox 2014:

11–12). Susam-Sarajeva (2003:19) argues that the non-existence or rarity of retranslations can be indicative of something. She discusses the English translations of Hélène Cixous French works, which have rarely been retranslated. For instance, the English translation and reprints of Le Rire de la Méduse, (The Laugh of the Medusa) was of great significance since it became a representative of French feminism among the Anglo-American readership, yet it had been translated only once. Susam-Sarajeva (2003) considers it unlikely that the initial translation would be regarded as “accomplished” but rather suggests that the translation as well as the difficulties of the translation process were not widely problematized and discussed in connection to Cixous’s theories in the Anglo-American academic feminist system (2003:20). Susam-Sarajeva (2003:25) notes that problems were not explicated in the English American translations as, for instance, neologisms, footnotes, endnotes and alienating effects were avoided in order to create the effect of a transparent text.

What should also be noted according to Koskinen and Paloposki (2015: 129) is that (re)translations often coexist with one another, and a later retranslation does not necessarily replace the previous version or versions. The subject of my study, F.H. Burnett’s The Secret Garden is a good example of this as all the three (re)translations were available in the library.

This coexistence is possibly connected to the fact that different publishers may have their own translations of a classic the copyrights of which have been expired. From a commercial perspective, it is plausible that publishers have an interest to try to sell re-editions of their translations despite the fact that a new retranslation would have been published.

2.9 Ageing of the source text

As presented by Deane-Cox (2014:5), Berman (1990:4) considers the original text eternally young whereas ageing is characteristic of translations, apart from what Berman (1990, 2009) calls as the ‘grande transduction’ (the great translation). In Berman’s approach, ageing is a sign of the flawed nature of a (re)translation (Deane-Cox 2014:5). Moreover, Deane-Cox presents Topia’s (1990:48) notion that the source text does not age in a similar manner as (re)translations. This means that while the source text is constantly re-contextualized and re-interpreted, the (re)translation is dependent on the source text as well as the norms of its time and place and therefore becomes static and frozen (Topia 1990:48 in Deane-Cox 2014:6). From such a perspective, it could be then argued that a retranslation rather than a re-edition is a way to update

(27)

22

or re-contextualize a translated text in the target culture. However, Emmerich (2017:10) is critical towards the idea that a source text would be static or richer than a translation. According to her, translation is not just mediating, editing or manipulating a preexisting, stable source text but it rather continues a process of textual iteration started in the language of the initial composition (Emmerich 2017:10).

Arguments that the source text is eternally young are contested in the case of children’s literature. For instance, after a public discussion, racist expressions of the original Swe- dish Pippi Långstrump were removed (Koskinen and Paloposki 2015), which indicates that source texts do not remain eternally young but can be revised, if some parts of them are considered unfit in the contemporary culture or for their audience. This could be partly because conventions of translating children’s literature allow modification (Shavit 2006:26).

However, modification as well as the coexistence of multiple versions of the source text is not limited to children’s literature. Emmerich (2017:1–5) has criticized the underlying assumption that a source text would be a fixed, singular entity, and that the same text would serve as the source text to the initial translation and retranslation in one, let alone multiple languages. As Emmerich (2017:14) argues, a work may exist in multiple forms in the language of its initial composition. As an example, Emmerich (2017:2) mentions that Virginia Woolf’s novels are printed in the USA and the UK from different sets of proofs and that the editions differ slightly from one another. Moreover, texts can have a digital form and a paper form, different editions can have a different format, and different versions may have varying illustra- tion. In addition, excerpts of novels can be published in magazines; plays may have different wordings depending on production. Authors may also revise their works afterwards (Emmerich 2017:2). A good example of this is Finnish novel Totta (2010), to which the author Riikka Pulkkinen wrote an additional chapter in 2011. In this respect, the object of my study, The Secret Garden, is an interesting case in since it was initially published as a series called Mistress Mary in The American Magazine (Smith 2018:69) and was later published as a book. Thus, novel is not the initial format of the story.

Defining the source text that is translated to target languages becomes even more complex in the case of older classics that have multiple source versions, translations and adaptations, some of which may have been used as intermediary source texts or references when translated to other languages (Emmerich 2017:14). Emmerich (2017:15) illustrates this by pointing out that the university library that she uses has seventy-nine different holdings for

(28)

23

Robinson Crusoe. They include the initial imprint of 1719, but also numerous editions, some of which some are illustrated, abridged, include peretexts or are published under another title.

What this illustrates is that that if there is more than one version of the source text, it is easier to criticize the idea of retranslation as a progress, since the network and relations of the source text and its (re)translations becomes complex. Moreover, Emmerich (2017:13) problematizes the idea that a source text, in other words a book that is considered the original, would always predate the translation, since the original story may have been modified after the initial translation or (re)translations have been published. Hagfors (2003:125) notes that a passage which contains an allegory to anti-Semitism in the Swedish children’s book Pelle Svanslös och Taxen Max has been censored in the Swedish version from 1970s onwards. However, the said passage was included in the Finnish unabridged version (originally translated in 1945), which was sold and available in 2003. Similarly, Fernández-López (2006:47–49) notes that while racist and xenophobic elements were removed from Enid Blyton’s texts in the 1980s in Britain, in Spain the (re)translations of the 1990s were done based on the unpurified versions. For example, the racist expressions that referred to the Romani people in Blyton’s Five Fall into Adventure were removed from the 1986 British edition. However, the unmodified 1950 version was used as the source text for the Spanish translation of 1990, which includes the censored parts.

A text is influenced by the format in which it is published. Koskinen and Palopo- ski (2015:236) point out that the material form of a book, such as the cover and the lay-out, can strengthen the view that a (re)translation is aged. In addition, the cover image or other images of a publication may have intertextual links. If a book is adapted to a film, it is possible that a re-edition is published and the images from the film or a series are used in the cover of the book.

An example of this is Margaret Atwood’s The Handmaid’s Tale (1986), which was filmatized as television series in 2017 and then published as a re-edition (2017) with a cover picture of the lead character of the series. In addition, the re-edition has a new introduction in which the author discusses the book, its origin, reception as well as its adaptations to other media. When it comes to the material of my study, both the front and back cover of the 10th edition of Toini Swan’s (1994) Finnish translation has images of the film adaptation (1993) of the story. In addition, the book has an attachment of 16 pages that has pictures of the film.

If the source text cannot be regarded as a fixed, unchanged entity, defining a retranslation is not always simple, either. In their study, Koskinen and Paloposki (2015: 152) discovered that sometimes the line between a retranslation and revised edition is difficult to draw. For example, editions that were listed as revised editions were potentially modified in

Racism, violence and disability in the Finnish (re)translations of Frances Hodgson Burnett's The Secret Garden

UNIVERSITY OF EASTERN FINLAND PHILOSOPHICAL FACULTY

SCHOOL OF HUMANITIES English language and translation

Anu Emilia Huovinen

Racism, Violence and Disability in the Finnish

(Re)translations of Frances Hodgson Burnett’s The Secret Garden

MA Thesis

March 2019

ITÄ-SUOMEN YLIOPISTO – UNIVERSITY OF EASTERN FINLAND

ITÄ-SUOMEN YLIOPISTO – UNIVERSITY OF EASTERN FINLAND

Contents

1 Introduction

2 Retranslation, translation norms and ageing of the translation and the source text