The (non-)ageing of non-retranslations? The alleged ageing of Swedish non-retranslations

ABSTRACT

Recently, scholars have pointed to the need for exploring non-retranslations, i.e. translations that are continuously being published in a target culture but not retranslated. Yet, there has not been a thorough examination of the phenomenon to date. This article aims to explore this phenomenon in greater depth by reporting on a pilot study from a bibliography on Swedish non-retranslations. More specifically, the article investigates the concept of (alleged) ageing in relation to fourteen Swedish non-retranslations that have been published, in the same translation, in more than ten editions over the course of 45–87 years. The non-retranslations are discussed through the parameters of publication history, agents of non-retranslations, and textual features. The article’s findings contradict some established views regarding the (alleged) ageing of texts as a motive for retranslation and complement others.

KEYWORDS:

• Non-retranslation
• retranslation
• ageing
• great translations
• translation history

Introduction

Several retranslation scholars have pointed out that retranslations are surrounded by different assumptions, such as that “every generation deserves its own Dostojevsky” (Van Poucke 2017, 93) or that a retranslation of a title should be issued “for each generation, that is every twenty or thirty years” (Vanderschelden 2000, 5). Yet, these assumptions are contested by the existence of non-retranslations, i.e. translations that are continuously being published without prompting a retranslation. Recently, scholars have argued for the need to explore “[t]he curious concept of non-retranslation” (Van Poucke and Sanz Gallego 2019, 14) in greater depth. For example, Kaisa Koskinen and Outi Paloposki (2019, 31–32) write that “more scholarly attention might need to be directed at cases of non-retranslation, i.e. works that remain relevant and read but do not get retranslated or revised” and suggest that “[a] corpus of these special cases of “non-aging” translations might allow us to advance our understanding of the processes of aging”. With this article, I wish to answer their call for an exploration of a corpus of these presumably “non-aging” translations.

Piet Van Poucke (2017, 92) refers to the “(alleged) aging of (literary) translations” as the way translated texts become old or outdated. Ageing, which includes linguistic, cultural and translational aspects, has often been taken for granted in research on retranslation. As Koskinen and Paloposki (2015b, 27) note: “The process [of retranslation] is initiated because there is an understanding that the existing translation is somehow faulty: too old, too outdated, too free, too domesticated or too foreignized”. A discussion of ageing in relation to non-retranslations – translations that seemingly are not deemed faulty despite their age – may counter some of these assumptions. Furthermore, it can help answer important questions about what has been retranslated, as proposed by Koskinen and Paloposki (2019, 31). So far, non-retranslations have mainly been touched upon in research concentrated on a single author’s work (Bollaert 2019). I wish to broaden this scope by reporting on a pilot project, which is the first part of a larger systematic empirical investigation on non-retranslations in twentieth-century Sweden. The pilot project focuses on 14 titles that have all been published in the same translation more than ten times during the last 90 years. As such, they present a counterexample to the often taken-for-granted perspective on retranslation.

The aim of the article is twofold. First, it aims to expand our understanding of non-retranslation as a phenomenon in literary culture and object of study in translation studies. Second, it seeks to provide empirical evidence for the growing body of research on the “(alleged) aging of (literary) translations” (Van Poucke 2017, 92) through an in-depth investigation of 14 non-retranslations.

First, I define some of the key terms for the study. I then introduce how temporal aspects have been conceptualized in relation to retranslations and the implications for this in relation to non-retranslations, with an emphasis on Antoine Berman’s (1990) notion of “great translations”, and linguistic ageing. Thereafter, I present the overall project and the corpus used in this article. The next section analyses three parameters of relevance for non-retranslations. The results are discussed in the final section.

Notes on terminology

A few comments on terminology are needed at this point, particularly to clarify the use of the terms non-retranslation, edition, reprint, and (overt and covert) revision. I define non-retranslation as a literary title that is continuously being published without being retranslated. Edition is a tricky term in this context. In her work on Nobel laureates in Swedish translation, Jana Rüegg (2021, 30) defines an edition (Swedish utgåva) as a “new version of a title”. In my material, a “new version” involves a new format and possibly a new publishing house. An edition can, in turn, be printed in several reprints (Swedish upplaga). The terms edition and reprint present a challenge when working practically with the Swedish National Library Catalogue Libris, which is where my material is derived from. As Rüegg (2021, 30–31) notes, it is hard to distinguish between an edition and a reprint. The usage in Libris is inconsistent and the two terms are used interchangeably and sometimes together (for example “[ny utg.][ny uppl.]”, i.e. “[new edition][new reprint]”). Like Rüegg, I have chosen a pragmatic path and included all the posts that are listed in Libris, which in practice means that there are instances where a reprint may be included as an edition. This choice is motivated by the fact that the posts in Libris reflect a long-term interest in the title, which is my main focus in this study. For example, Stäppvargen by Hermann Hesse has been published in seventeen editions using the same translation by Sven Stolpe (and Karin Stolpe¹) from 1938. However, as we shall see, it has been covertly revised, i.e. revised without it being acknowledged in either the colophon or the bibliographical data. In my view, this covert revision does not counter its status as a non-retranslation. This implies a wider definition than Koskinen and Paloposki (2019, 31–32), who exclude revisions by defining non-retranslations as “works that remain relevant and read but do not get retranslated or revised”. My reason for including revisions is above all practical, as very few revisions are overtly acknowledged in my study. It would present a major challenge to exclude revisions as it would mean that every edition of a title would have to be manually checked for revisions before it could be included in the bibliography. Including revisions in the definition of non-retranslation is also interesting since revision has been proposed as the first step toward retranslation (Vanderschelden 2000, 1).

Ageing in retranslation … and non-retranslation

Within retranslation studies, ageing – and, more generally, temporal aspects of retranslation – is seen as a recurrent motive for retranslation. In this section, I will discuss two temporal aspects in relation to non-retranslation: the concept of “great translations” and ageing. The notion of “great translations” (“grandes traductions”) stems from Berman’s (1990) influential essay on retranslation. Commonly overshadowed by the thought that later formed the basis of the retranslation hypothesis, the concept of great translations deserves further attention, not least in relation to non-retranslations. According to Berman, translations cannot keep up with “a given state of language, literature and culture” (1).² Consequently, Berman put forward the idea of a great translation, which refers to the one translation that will provide the target culture with a satisfactory text “when the translation of a text becomes vital for a culture’s being and history” (6). An interesting feature of great translations, Berman posits, is that they are a product of “great translators” (6). In his rather idealistic view, these translators possess “a translating force” (6) that goes beyond a simple desire to translate.³ This viewpoint may be associated with a longstanding view on the translator as being driven by vocation, which may be mirrored in contemporary research conducted by Rakefat Sela-Sheffy (e.g. 2010) on Israeli star translators who emphasize their personal aptness and their vocation to translate, which indicates a vaguely professionalized field.

Although commonly associated with retranslations, Berman’s initial writing on great translations is not as clear-cut:

If all retranslations are not great translations (!), every great translation is, in fact, a retranslation […] For this correlation to hold empirically, it needs to be nuanced in two ways. To begin with, it is not … universal. A first translation can be a great translation. (Berman 1990, 3; cited in Lane 2020, 307)

From this quotation, it is clear that the hypothesis that all great translations are retranslations is more ambiguous than it is generally perceived to be. In particular, Véronique Lane (2020, 307) has pointed out that the ellipsis shows an uncertainty whether “universal” (French absolue) is the right word, which she sees as “a tangible trace within the text of the distinctions Berman sought to emphasize” (307). Moreover, Lane (307) explains how these punctuation marks inflict on the interpretation of the quotation:

While Chesterman, adopting a descriptive approach, attempted to universalize Berman’s thoughts on retranslation, scholars adopting an ideological approach have pointed to the subjective judgements called for by his use of adjectives: if only retranslations can be “great”, then first translations are bound to be “poor”. Berman strikingly preempted such “universalizing” and “ideological” impulses by adding here that “first translations can be great” and that “all retranslations are not great translations (!)”.

The first distinction – that first translations can be great – enables non-retranslations to be considered great translations. More recently, great translations have been described as “[a] TT which has acquired a long-term status in the TL” (Vanderschelden 2000, 11) or as translations that “resist time” (Van Poucke 2017, 96), which is perhaps a more pragmatic perspective. These descriptions bear obvious resemblances to the concept of non-retranslation: a non-retranslation is the only translation of a given source text in a target culture and has therefore likely achieved a long-term status in the target language as a translation that has “restist[ed] time”.

In addition, great translations are linked to canon formation and the accumulation of status of a target text in a target culture. The relationship between retranslation and canon formation has been described by Koskinen and Paloposki (2015b, 27) as “a complex circular interdependency: classics are considered worthy of reinterpretation and are thus often retranslated, but retranslation is also considered a signal of the work’s status as a classic, and the act of retranslation can be considered to raise it into the class of classics”. Following this line of thinking, the absence of retranslations could, at first glance, indicate that non-retranslations have not (yet) reached canonical status in the target culture. As Isabelle Vanderschelden (2000, 6) points out, it takes time “before a literary text can establish itself as a classic abroad”. It might, however, be worth pondering whether a non-retranslation can acquire canonical status in the target culture precisely because it has not been retranslated. The texts in this study have all been published more than ten times, which suggests that they have acquired some sort of standing in the Swedish target culture. Moreover, as Koskinen and Paloposki (2015b, 27) note, a common assumption regarding retranslations is that certain titles are deemed “worthy” of being retranslated. Van Poucke (2017, 3) also discusses retranslation of canonical works in terms of worthiness:

Where (sic) we are dealing with canonical literary works, which is very often the case as those are usually the works that are considered to be worthy of retranslation, the readers acknowledge that the language of the ST is old and tend to accept its outdated character in translation as well.

In relation to non-retranslations, the first thought would be that these translations have not been considered worthy enough to be retranslated. However, these titles have obviously been considered worthy enough to be published, sometimes in numerous editions over a long period of time, which indicates that there is some kind of value attached to these titles in the target culture.

The quotation above comes from Van Poucke’s (2017) meta-analysis of 70 case studies on retranslations in which he explores ageing as a motive for retranslation. He defines “ageing” as an umbrella term covering “all aspects of translations that comply with (linguistic, translational, cultural) norms and conventions that are replaced in retranslation because of the (alleged) outdated character of the feature, whether that be part of the lexicon, syntax, style, register or translation strategy” (95). Thus, an aged lexicon in itself does not necessarily prompt a retranslation, possibly because the readers are more ready to accept an aged and outdated lexicon in the translation of a canonical work. Instead, Van Poucke finds that the ageing occurring in most of the case studies could be referred to as a sort of “higher level” category, such as stylistic ageing or translational ageing, but that it is “not empirically proven on the lexical and syntactic level” (110). The issue of an outdated lexicon is noteworthy in relation to non-retranslations in several respects. Firstly, several scholars have brought the question of the dividing line between a revision and a retranslation into the limelight (e.g. Paloposki and Koskinen 2010; Koskinen 2018). If Van Poucke’s claim that an aged lexicon does not prompt a retranslation holds true, this might indicate that possible revisions in non-retranslations are performed in “lower level” categories, e.g. grammatical, syntactical, and lexical. Secondly, this issue naturally touches on another temporal aspect, namely the language variation that occurs in any given language over time. Regarding Swedish, Mats Larsson (2007, 54), building on Lars Melin (1998), writes that the Swedish written language since the 1950s has undergone “relatively major changes”, which has led to Swedish becoming “shorter, more informal and more intimate, more visual, more international and more oral”. In times of rapid language change, the changes may be mirrored in literature. Koskinen and Paloposki (2015a) have found compelling evidence that different eras favor retranslation. A surge of retranslations is noticeable, for example, in late 1940s Finland, foreshadowing the modernist boom of the 1950s.

Creating a bibliography of non-retranslations

The present article is a part of a larger project entitled “Non-retranslations in twentieth century Sweden”. The project consists of two parts: the first part entails the creation of a bibliography of non-retranslations published in Sweden from 1900 onwards, and the second part comprises a case study of Lily Vallquist’s (1897–1986) Swedish translation of Françoise Sagan’s Bonjour tristesse, which has been published eight times in the same translation since 1955 (Svahn 2022). The bibliography aims at gaining a macro perspective of the phenomenon of non-retranslations in Sweden, thus answering the recently expressed call for macro perspectives on retranslations (Berk Albachten and Tahir Gürçağlar 2018; Paloposki 2018; Koskinen and Paloposki 2019). Hopefully, it will encourage future research on non-retranslations in Sweden, and eventually lead to more case studies on non-retranslations.

The creation of the bibliography is ongoing. Currently, it contains 150 titles and 766 editions. The selection criteria used for the bibliography of non-retranslations stipulate that each title has been published at least three times in the same translation, and the first edition must have been published in Sweden between 1900 and 1990. The year 1990 was chosen as a suitable ending point considering that a title could hypothetically have been published three times during the thirty years between 1990 and 2020, when this project began. Only novels and short story collections are included. The search has been conducted in the National Swedish Library Catalogue Libris. According to Johan Svedjedal (2012, 37–38), Index Translationum (2014), the UNESCO database of book translations, should be treated cautiously when working with Swedish translations, and Libris is deemed more reliable for the project in terms of locating titles. There is, however, some inconsistency in what kind of bibliographical data is available, which is especially apparent in the early decades of the twentieth century.

My initial interest in non-retranslations as a phenomenon began as a result on my work on the translator of Bonjour tristesse, Lily Vallquist (Svahn, n.d.). I then started looking into translations from the same time and found more non-retranslations than I had expected. This discovery created a sort of snowball effect, where I searched very broadly for authors and non-retranslations. The process included examining all the other translations of an author in the bibliography, going through book series as well as publishing houses. Because of the problems of locating non-retranslations, the bibliography does not aim to be representative of all non-retranslations in Sweden but, hopefully, can provide a sufficient sample from which to draw meaningful conclusions.

Out of the 150 titles in the bibliography, 14 have more than ten editions (i.e. more than ten editions have been published in the same translation). As a first step, this article focuses on these 14 titles, which together include a total of 176 editions (see Table 1), representing 22.85% of all the editions in the bibliography.

Table 1. The 14 titles ordered chronologically after the first edition’s publication year.^a

Swedish title	Original title	Author	Translator	First edition	Last edition	Number of editions
Hattmakarens borg	Hatter’s castle	Archibald Joseph Cronin	Louis Renner	1931	1985	11
Stäppvargen	Der Steppenwolf	Hemann Hesse	Sven Stolpe	1932	2019	17
Den goda jorden. En kinesisk folklivsskildring	The good earth	Pearl Buck	Sven Rosén	1932	2018	11
Riddarna kring Dannys bord	Tortilla flat	John Steinbeck	Sven Barthel	1938	1996	11
Vredens druvor	The grapes of wrath	John Steinbeck	Torsten Jonsson	1940	2019	17
Siddharta. En indisk berättelse	Siddhartha, eine indische Dichtung	Hemann Hesse	Nils Holmberg	1946	2012	13
Muren	Le mur	Jean-Paul Sartre	Eyvind Johnsson and Eva Alexandersson	1946	2005	11
Äcklet	La nausée	Jean-Paul Sartre	Eva Alexandersson	1949	2012	11
Ett mord annonseras	A murder is announced	Agatha Christie	Britt-Marie and Sven Bergström	1951	2015	12
Öster om Eden	East of Eden	John Steinbeck	Nils Holmberg	1953	2018	13
Flugornas herre	Lord of the flies	William Golding	Sonja Bergvall	1959	2016	13
Moment 22	Catch 22	Joseph Heller	Torsten Blomqvist	1962	2012	12
Spionen som kom in från kylan	The spy who came in from the cold	John le Carré	Lena and Cai Törne	1964	2017	11
Glaskupan	The bell jar	Sylvia Plath	Christina Liljencrantz	1974	2019	12

^a Johnson translated the title story and Alexanderson translated the remaining short stories.

From Table 1, it is clear that the titles have been written by a diverse group of authors. The source languages are English, French, and German. In the following section, the material is explored in depth according to three parameters: publication history, agents of non-retranslations, and textual features. The first part investigates the publication history and aims to dismantle some of the assumptions regarding non-retranslations. The second part contextualizes the material by situating the authors and translators in the era when the translations were first published in Sweden, which is illuminating in terms of the title’s status development in the target culture. The third part focuses on linguistic ageing and contains a small-scale textual investigation detecting revisions between the different editions. Together, these three sections offer a first glimpse into the phenomenon of non-retranslation in relation to ageing.

Parameters of non-retranslations

Publication history: publication timespans and intervals, and publishers

The publication timespans and intervals are, of course, of paramount importance in an exploration of non-retranslations in relation to ageing. As shown in Figure 1, the 14 titles display a wide timespan between the first and most recent editions: Stäppvargen⁴ has the longest span (88 years; 1932–2020) and Glaskupan the shortest one (45 years; 1974–2019). The mean length of the publication timespan of the 14 titles is 63.2 years, showing that they have indeed acquired longstanding status in Sweden as the only translations of the source texts. Moreover, considering the common perception that a retranslation is needed “every twenty or thirty years” (Vanderschelden 2000, 5), this number clearly shows that such a claim does not hold true for all translations.

Figure 1. Time span between the first and most recent editions.

It is also worth considering the mean length of each title’s publication interval, i.e. the average number of years between editions of the same title. As displayed in Figure 2, the mean publication interval runs from 8.6 years for Den goda jorden to 4 years for Glaskupan. These numbers show a continuous interest in the investigated titles; the mean interval of all the 14 titles is 5.5 years, indicating that a new edition appeared roughly every five years without prompting a retranslation.

Figure 2. Mean interval between editions for each title.

Another interesting aspect of the titles’ publication histories concerns their publishing houses. The 176 editions in the material have been published by 15 different publishing houses: Albert Bonniers förlag (94), Aldus (21), Forum (16), Wahlström & Widstrand (11), Tiden (10), Trevi (8), Hökerberg (5), Litteraturfrämjandet (2), Alba (1), Bra Böcker (1), Abcoude⁵ (1), Månpocket (1), Reader’s Digest (1), Repris (1), Åhlén & Åkerlund (1), and Bookmark (1).⁶ The majority (53.5%) were published by Albert Bonniers förlag, the leading publishing house in Sweden. However, considering that Aldus, Forum, Wahlström & Widstrand, Tiden, Alba, Repris, and Månpocket all belonged to the Bonnier cooperation at the time when the editions were published, the number of editions tied to Bonniers increases to 147 (83.5%). This suggests that non-retranslations are tied to prestigious publishing houses – possibly counter-intuitively, since larger publishing houses are more likely to have the financial means to commission a retranslation. Large and prestigious publishing houses might also be more well-suited to attract recognized translators, which could positively affect the book’s chances of becoming “longsellers”. When looking at the most recent edition of each title, 11 of 14 have been published by Albert Bonniers förlag, the exceptions being Hattmakarens borg in 1985 (Bra Böcker), Moment 22 in 2011 (Wahlström & Widstrand), and Ett mord annonseras in 2015 (Bookmark). In general, five of the 14 titles were published since 2018 and three of them (Glaskupan, Stäppvargen, Vredens druvor) in 2019. These numbers clearly show that the phenomenon of non-retranslation is not solely an issue of the past but is also present in the contemporary publishing industry.

A noteworthy tendency is the frequency of these titles being published as a part of larger series: 104 (59%) of the 176 editions have been published in some sort of series. The majority are published in prominent quality series aimed at high prestige literature, such as Delfinserien, Panache, and Svalan, all released by Bonnier publishing houses, indicating that the titles have been deemed high-prestige literature. The combination of a high concentration of publishing houses and non-retranslations being published in high-prestige literature series further points towards the publishing houses’ role in the creation and maintaining of non-retranslations.

To sum up, several common assumptions regarding retranslations are contradicted by the findings outlined above. First, the publication timespan displays a wide range, from 45 to 87 years, which suggests that these titles have indeed been deemed relevant for an extended period of time. Second, the consistency with which these titles have been republished – on average every 4.6 years  – further strengthens this view and also shows that their relevance has been consistent over time. Third, 11 of the 14 titles were most recently published in 2012 or later – five of them were even published since 2018, and three of them in 2019, showing that these titles are deemed relevant for a contemporary audience. In addition, the vast majority of the studied titles have been published by well-established and prestigious publishing houses, often in series designed to promote prestigious literature.

Agents of non-retranslations: authors and translators

Van Poucke (2017, 111) asserts that studying agents involved in retranslations can provide a fuller picture of the phenomenon. Similarly, shedding light on agents involved in non-retranslations – such as authors and translators – can provide a deeper insight into the phenomena.

The eight authors in the corpus – Hermann Hesse, Pearl Buck, John Steinbeck, Jean-Paul Sartre, Agatha Christie, William Golding, Joseph Heller, and Sylvia Plath – have some shared characteristics, although the differences are greater. The majority of them have only one title in the data set, the exceptions being Steinbeck (3), Hesse (2) and Sartre (2). The authors include Nobel laureates (Hesse, Buck, and Sartre) as well as other high-prestige authors such as Steinbeck, Golding, Heller and Plath. Two authors can be characterized as crime authors: Agatha Christie and John le Carré. The author that is the hardest to describe is Archibald Joseph Cronin, who is not well-known in Sweden today. The majority of the authors are male, with the exception of Buck, Plath, and Christie. The source languages are English (10), French (2), and German (2). It is interesting to note that the source languages can be characterized as hype-central or central according to Johan Heilbron’s (1999) categorization, which suggests that the practice of non-retranslation is not due to a lack of translators.

As we have seen in Berman’s discussion of great translators as a prerequisite for great translations, translators may influence how their translations will be received. Several of the fifteen translators in this group are well-known names in Swedish translation history for different reasons. Eyvind Johnson, for example, was a prominent writer, a member of the Swedish Academy, and received the Nobel prize for literature in 1974. Sven Stolpe was also a well-known writer and a public intellectual.

In 1954, at the instigation of the Swedish Association for Literary Translators, the Swedish evening paper Aftonbladet launched a series of interviews with twelve prominent literary translators (Liffner 2013). Out of these twelve, four appear in the dataset: Louis Renner (interviewed together with his wife and translation partner Lisbeth Renner), Sonja Bergvall, Torsten Blomqvist, and Nils Holmberg. Those interviews provide insight into the Swedish translator scene of the 1950s when most titles in this study were published. For example, it is interesting to note that both Holmberg and Blomqvist refer to themselves as “professional translators” (Swedish yrkesöversättare). At the time, Holmberg was engaged in the newly founded translator association as a delegate in their negotiations with the Swedish publishers’ association. In the interview, Blomqvist clearly positions himself as a “professional translator” who “should be able to handle everything – except poetry, which is an art on its own” (Liffner 2013, 39).

Furthermore, the interviews offer insight into what the translators and the milieu at the time considered to be their most prominent translations, which may be another indication of the status of the source texts when they were first published. In the case of Bergvall, we can read: “The literary finest and most important of her translated books are: Isabel Bolton: Sover jag eller vakar, A.L. Barker: De oskuldfulla, Saul Bellow: Offret, L.P. Hartley: Bror och syster and Elisabeth Bowen’s Middagshöjd” (Liffner 2013, 25). In the interview with Holmberg, his most important translations are discussed as follows:

 – I must have translated around two hundred fifty books, some of them anonymously, he says. Most of them were made after 1939 when he started translating professionally. And it’s no small potatoes that he has set his eyes on and managed: Thomas Mann’s Josef series, Lotte i Weimar, Doktor Faustus etc., Hermann Hesse’s Glaspärlespelet, Dickens’ Pickwickklubben, Winston Churchill’s monumental work on the Second World War … It is also possible to mention books by Somerset Maugham, Robert Graves, and Arthur Koestler. (Liffner 2013, 32–33)

Also, in the interview with Blomqvist, there is a passage when his most noteworthy translations are discussed:

His first book translation came in 1947. It was “an English miss novel”, not worth mentioning. Later, Blomqvist translated some ordinary entertainment novels. But his record also includes some quite exquisite things. The first significant work was Jean Stafford’s Puman (in the Panache series). Later he also dealt with Van Tillburg Clark’s Den store kattens spår; at that time he was a fully-fledged expert on pumas. Belonging to his “peaks” are also a couple of books by William Sansom, Compton-Burnett’s Tjänare och tjänarinnor, Ralph Ellison’s Osynlig man, Carl Sandburg’s memoirs, Charles Lindbergh’s The Spirit of St. Louis, which will be published this autumn, and a memoir volume by a professor of literature in Harvard, Alfred Kazin (this one is unfortunately still unpublished). One can also mention Irwin Shaw’s De unga lejonen and Stormvarning. (Liffner 2013, 39–40)

In sum, Bergvall, Holmberg, and Blomqvist discuss titles which are seldom well-known to a contemporary audience – in some cases, they seem to have fallen into complete oblivion. This aspect of ageing denotes the long-term diachronic consecration of literary titles and authors and how their status changes over time. Renner does not elaborate on his most essential translations in his interview, except that they particularly enjoyed translating Pearl Buck and Vicki Baum, two authors that are well-represented in the bibliography. In the case of Bergvall, Flugornas herre came out in Swedish five years after this interview; Blomqvist translated Moment 22 nine years later. Holmberg, however, had already translated both Hesse’s Siddharta and Steinbeck’s Öster om Eden at the time of the interview, yet he does not seem to hold them in high esteem.

Summing up, most of the authors in this project’s corpus are nowadays considered high prestige, but many of the titles do not seem to have enjoyed that status when they first came out in Sweden. Some of the translators are known to the public as writers, while some others are keen to emphasize their role as professional translators.

Textual features: covert and overt revisions

This section reports on a small-scale manual investigation of whether these titles have been covertly revised. Several scholars have pointed out the blurred boundaries between a revision and a retranslation (e.g. Vanderschelden 2000; Paloposki and Koskinen 2010; Koskinen 2018). The bibliographical data does not indicate that any of the 14 titles have been subject to revisions. To investigate whether any covert revisions had taken place in this study’s corpus, I randomly selected passages of roughly 150 words at five instances in each first edition, i.e. a total number of 750 words in each edition, in the 169 editions⁷ available at the National Library of Sweden in Stockholm. These samples were then compared through a close reading of all the subsequent editions in order to detect any revisions. Of course, this is a small sample and can only signal tendencies.

The investigation indicates that seven of the 14 titles – Vredens druvor, Siddhartha, Öster om Eden, Moment 22, Flugornas herre, Ett mord annonseras and Glaskupan – have not been subject to revision across the different editions (see Table 2). The remaining seven have been subject to revisions to various degrees. We can only speculate whether these revisions were made by the publishing houses alone or with the translators’ consent. The only indication of the latter being the case is a note in Libris accompanying the entry on Muren from 1987 and in subsequent editions stating that Alexanderson’s revised her translation, which gives the impression that she did not revise Johnson’s.⁸ Le Carré’s Spionen som kom in från kylan was revised by Mo in 2017, which is acknowledged in the colophon of that edition.

Table 2. Overview of all titles with revised editions, revision level, and examples.^a

Title	First edition	Latest edition	Revised editions	Revision level (number of revisions)	Example
Hattmakarens borg	1931	1985	1955, 1985	Grammar and syntax (2)	Mamma-mamma; tjuguåtta-tjugoåtta
Stäppvargen	1932	2019	1951, 1960, 1968	Grammar and syntax (20)	Punctuation; kommo-kom; är–blev
				Lexical (12);	Medge-säga; elastisk-spänstig; styv-stel
				Style and register (10)	Added words; split sentence
Den goda jorden	1932	2018	1938, 1942, 1945	Grammar and syntax (3)	Olan-O-lan; voro-var; icke-inte
				Style and register (29)	Changed chapter order
Riddarna kring Dannys bord	1938	1996	1950, 1962a, 1990	Grammar and syntax (27)	Dannys Bord-Dannys bord-Dannys Bord; fast-fastän; punctuation
Vredens druvor	1940	2019	–	–	–
Siddharta. En indisk berättelse	1946	2012	–	–	–
Muren	1946	2005	1987, 2004	Grammar and syntax (3);	röck-ryckte; sa-sade, tjugu-tjugo; utropar-utropade
				Lexical (2)	ställning-hållning; genevarna-geneveborna
				Style and register (1)	Removed sentence
Äcklet	1949	2012	1964b, 1982	Grammar and syntax (2);	utropar-utropade; tjugu-tjugo
Ett mord annonseras	1951	2015	–	–	–
Öster om Eden	1953	2018	–	–	–
Flugornas herre	1959	2016	–	–	–
Moment 22	1962	2012	–	–	–
Spionen som kom in från kylan	1964	2017	2017	Grammar and syntax (2)	jag-Jag; sen–sedan
				Lexical (3)	i sin koj-på sin brits; gastkramning-förlamande isolering
				Style and register (13)	du-ni; polisen-die Polizie
Glaskupan	1974	2019	–	–	–

^a The revision “genevarna” – “geneveborna” found in the 2004 edition of Muren likely occurred already in the 1987 edition, where all the other revisions were found. However, the pages 87–128 were missing from that edition.

Table 2 shows the revisions found in the different editions. The categorization follows Van Poucke’s (2017) division of linguistic ageing into grammatical/syntactical ageing, lexical ageing, and stylistic ageing. However, these categories do not align precisely with Van Poucke’s. For example, here, grammatical/syntactic revisions include changed spelling, changed punctuation, and changes in verb tense. Lexical revision includes changed lexicon, generally to a more contemporary or to a more colloquial one. Lastly, stylistic revisions include changes to forms of address, additions, deletions, and macro-scale changes related to the integrity of the oeuvre. These categories are used to enable comparison although they, naturally, overlap. For example, changes to punctuation or using a more modern synonym also affect the stylistic level.

As can be seen in Table 2, with the exception of Spionen som kom in från kylan, the titles first published in 1951 or later have not been revised; the only titles published before 1951 that have not been revised are Vredens druvor from 1940 and Siddhartha from 1946. This indicates that some sort of linguistic ageing has likely played a part in whether the titles have been revised or not. In six out of the seven revised titles, the most significant revisions have been carried out in the editions published in the 1950s–1960s. Vanderschelden’s (2000, 1) claim that revision tends to be the first step towards retranslation, thus, does not hold true for this corpus, with one exception which will be discussed below.

Most revisions are found at the grammatical/syntactic level, which can be partly explained by changing linguistic conventions, in particular changing conventions regarding spelling, punctuation, and plural verb forms (i.e. the replacement of plural verb forms with the corresponding singular forms). Most notable in this category are the changes to punctuation. Out of the 28 instances of grammatical/syntactical revisions in Riddarna runt Dannys bord, 24 refer to changing punctuation. To give an example, a comparison from the first and third edition of Riddarna runt Dannys bord reads:

1938 Nej, när man talar om Dannys hus, är det underförstått att man avser en enhet, vars beståndsdelar är män, från vilka utstrålade vänlighet och glädje, filantropi och, mot slutet, hemlighetsfull sorg.

[No, when you speak of Danny’s house, you are understood to mean a unit, of which the parts are men, from which came sweetness and joy, philanthropy and, in the end, a mystic sorrow] (Steinbeck 1935, 9)⁹

1950 Nej, när man talar om Dannys hus är det underförstått att man avser en enhet, vars beståndsdelar är män från vilka utstrålade vänlighet och glädje, filantropi och, mot slutet, hemlighetsfull sorg.

[No, when you speak of Danny’s house you are understood to mean a unit, of which the parts are men from which came sweetness and joy, philanthropy and, in the end, a mystic sorrow] (Steinbeck 1935, 9)

Conventions regarding the use of commas in the Swedish language changed in the 1950s, moving away from a clausal use in favor of commas used for clarity (Dahl 2016, 41). The revisions being made in the editions from 1950–1960 make sense in relation to these changing linguistic conventions.

There are some examples of lexical revisions in the material, especially in Stäppvargen where some constructions have been reformulated. For example, the 1951 edition reads “härjade och stod i med alla grejorna” [“ravaged and went about with all the things”], which in the 1960 edition was changed to “flyttade och skruvade på det hela” [“moved and screwed at the whole lot”], and, in the 1968 edition, was changed to “vred och vände och skruvade” [“twisted and turned and screwed”]. There are also examples of revisions including changes of words with very similar meaning (e.g. ställning-hållning).

There are some cases of stylistic or “higher-level” ageing such as a full sentence removed in Eva Alexanderson’s revision of Muren from 1987 and additions made in Stäppvargen from 1960. More overarching revisions are found in Buck’s Den goda jorden affecting the novel’s composition, where chapters 3 and 7 in the third edition from 1938 merge with the previous chapters, resulting in new chapter headings (i.e. chapter 3 became chapter 2). The fourth edition published in 1939 was made in the same fashion, whereas the fifth edition published in 1942 follows the original outline. Yet, the seven editions published since 1945 follow the merged style from 1938 and can hence be seen as a form of paratextual revision.

The most thorough revisions, affecting grammar/syntax, lexicon, and style, are found in the revisions of Stäppvargen and in Johanna Mo’s revision of Spionen som kom in från kylan from 2017. In that edition, there is a change in address (du-ni), inclusion of foreign words (polisen–die Polizei), and lexical changes. With this example, it might be worth pondering Vanderschelden’s claim of revisions being the first step towards retranslation, as a retranslation was issued during the completion of this article and published in 2021.

Interestingly, some publishing houses acknowledge and even emphasize that they are using old translations. The most recent editions of Vredens druvor, Stäppvargen and Glaskupan were published by Albert Bonniers förlag in 2019 with the following note in the colophon: “In order to let the book serve as a historical document, we have chosen to publish it in its original state. Therefore, words and expressions that are or may be offensive for today’s readers occur”. This message to the contemporary reader shows an awareness of the possibly outdated character of the translation. Yet, it may also serve as an indication of ageing as a positive feature that highlight the work’s perceived status as a classic.¹⁰

To summarize, the investigation shows that half of the investigated titles have been subject to revisions: covert revisions except from Muren and Spionen som kom in från kylan. Most revisions are found at the grammatical/syntactic level, but there are also examples of lexical and stylistic revisions. The revisions were mostly made in the 1950s–1960s, showing that revisions have generally not been performed in the most recent editions. Another conclusion is that none of the revisions found alters the source text to the point that a retranslation was thought necessary, which supports the decision to include revised editions in the definition of non-retranslations. However, Spionen som kom in från kylan is interesting in this regard since a retranslation was issued just five years after it had been revised.

Discussion

This article set out to explore ageing in relation to the phenomenon of non-retranslation. The findings both contradict and complement some of the previous research regarding the (alleged) ageing of translations as a motive for retranslation. First of all, this claim is challenged, since the material in this study clearly shows that ageing in itself does not necessarily prompt a retranslation. The fact that 11 of the 14 titles in this study were last published in 2012 or later – five of them were even published since 2018 and three of them in 2019 – indicates that, although a translation may be old, it can still be deemed relevant and publishable. The findings thus counter the assumption that ageing inevitably leads to retranslation. However, the fact that half of the titles examined here have been subject to revisions supports the claim that some sort of ageing has occurred. Next, I will address two major questions that have come to light, and point out future avenues for research.

Non-retranslations as great translations?

One question arising from this study relates to whether these 14 non-retranslations may be called “great translations”. For Swedish readers during the twentieth century and into the 21st, these non-retranslations are synonymous with the original literary titles, and, as such, they have shaped the image of their authors in Sweden. In this respect, they may be described as “great translations” that effectively have “resist[ed] time” (Van Poucke 2017, 96). Although this is not how Berman’s ideas have generally been understood, his initial writing on the subject does not exclude this interpretation. Following this line of thought, could then their translators be called great translators? It is true that some of them, mainly Johnson and Stolpe, deserve this designation due to their status as writers. In Berman’s view, great translators are characterized by a “translating force”, which reminds us of the translation sociological research carried out by Sela-Sheffy (e.g. 2010) in which high prestige literary translators emphasize their vocation to translate and mystify their skills. However, the interviews with some of the translators from the 1950s rather indicate a sort of emerging professionalization of the Swedish literary translation field, where some interviewees highlight being professional translators. In the future, archival research may provide more insight into who initiated the translations of the investigated titles, which could be indicative of the translators’ translating force.

Non-retranslations as ageless translations?

Another important question relates to what kind of ageing takes place in non-retranslations. Van Poucke found that ageing could most often be related to a “‘higher level’ category, such as stylistic ageing or translational ageing, but not on the lexical and syntactic level” (2017, 110). My limited textual investigation complements Van Poucke’s findings since most of the revisions were made at the grammatical/syntactical level. Taken together, these findings indicate that the dividing line between when a revision (and not a retranslation) is deemed necessary is drawn between grammatical/syntactical ageing (and to some extent lexical ageing) and stylistic ageing. This aspect needs to be investigated more thoroughly in future research. While Koskinen and Paloposki (2015a) find that some eras favor retranslation, the fact that most revisions were made in the 1950s and 1960s indicates that some eras instead favor revision, for example when the target language has experienced major recent changes.

Another aspect of ageing concerns the target readers’ expectations. It is more or less explicitly understood that ageing negatively affects the readers’ possibilities of accepting an aged translation. One idea worth pondering is to what extent reading a modern classic influences, and potentially enhances, the reader’s willingness to accept an aged lexicon, as Van Poucke (2017, 3) alludes to. Koskinen and Paloposki (2015b) argue that retranslations can be labeled classics because they have been retranslated. The results from this study suggest that the titles under investigation are granted their status as classics because of their status as non-retranslations – as “untouched” classics that mirror the linguistic and translational norms from the time when they were first translated. Finally, one temporal aspect that cannot be overlooked is time itself. It is possible that the authors discussed here have not yet reached the momentum required in order for their titles to be retranslated, echoing Berman’s idea about a favorable socio-cultural environment required for retranslation. The number of times the 14 non-retranslations in this study have been published in Sweden can arguably speak against this idea, but may not totally refute it. The retranslation of Spionen som kom in från kylan points in this direction.

Concluding remarks

This article answers the call for a corpus of non-retranslations, taking a macro perspective on the issue of non-retranslation and ageing. It points toward the complex process behind non-retranslation, which includes the diachronic consecration of authors and literary titles, the evolvement of the target language conventions, and the expectations of the readers of modern classics. More generally, the study has shown that non-retranslation is a recurrent phenomenon in the Swedish target culture to the present day, which is likely the case for other target cultures as well. While this study focused on ageing in relation to non-retranslation, it has also shown that non-retranslations may be explored from many different angles. Koskinen and Paloposki (2019) have suggested that looking into non-retranslations can help answer questions of what has not been retranslated, i.e. in relation to retranslations. To this I would like to add that non-retranslations can also be studied in relation to each other. This will help us to further advance our understanding of the curious concept of non-retranslation.

Acknowledgments

I wish to thank Prof. Kaisa Koskinen and Dr. Christopher Crocker for valuable comments on earlier drafts of this article. I am also grateful to the anonymous reviewers whose suggestions have ameliorated the article. The remaining faults are my own.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This project was supported by the Birgit and Gad Rausing Foundation.

Notes on contributors

Elin Svahn

Elin Svahn holds a PhD in Translation Studies from Stockholm University. She defended her PhD thesis The Dynamics of Extratextual Translatorship in Contemporary Sweden. A Mixed Methods Approach in 2020 and works as an assistant professor at the Institute for Translation and Interpreting Studies at Stockholm University. Her research interests include retranslation, translation sociology, and translation history. She is the co-editor of a special issue of Hermes dedicated to the translation profession together with Minna Ruokonen and Leena Salmi (2018) and a methodologically oriented volume co-edited with Lova Meister (Morfem, 2020).

Notes

1 Sven Stolpe often translated together with his wife Karin Stolpe, and Dahl (n.d.) mentions Stäppvargen as one of their collaborative translations.

2 All translations by the author unless otherwise indicated.

3 I am using Vanderschelden’s (2000, 11) translation for Berman’s term “surge de traduire”.

4 In the following, I refer to the Swedish titles in the text.

5 In cooperation with Bonnier.

6 Siddhartha from 1946 cannot be associated with any publishing house.

7 The editions that were not available at the National Library of Sweden in Stockholm are Öster om Eden (1957), Äcklet (1949), Muren (1970), and Glaskupan (1974 [a second edition published the same year as the first edition], 1975, 1978).

8 The note reads: “The title story was translated by Eyvind Johnson and the remaining by Eva Alexanderson, who revised her translation in 1987.”

9 I have added commas in order to reflect the use of commas in the Swedish translation from 1938; the example from 1950 follows the same rule.

10 In the case of Stäppvargen, the most recent edition is, however, not the original edition but the revised edition from 1968.